Indian J Med Res 119 (Suppl) May 2004, pp ix - xi Summary

XV Lancefield Symposium on Streptococci and Streptococcal Diseases

An impressive ceremony on Sunday 6th October work opens the door to comparative genomics and to marked the opening of the XV Lancefield International understanding the relationships and distinctions between Symposium on Streptococci and Streptococcal Diseases. and within species. It should now be possible to better The Governor of Goa, Mr Mohammed Fazal, spoke of understand cell physiology and metabolism, identify the challenge facing scientists in finding ways to fight bacterial virulence factors, determine antimicrobial against serious diseases by improving methods of resistance mechanisms, explore the impact of gene prevention and treatment. That scientists are responding acquisitions and deletions, identify phage determined to the challenge, I believe, has been demonstrated during activities, detect alternative vaccine candidates, and help the XV Lancefield Symposium. to explain bacterial evolution.

The opening lecture given by Dr Richard Krause, Additional to progress on bacterial genomic structure, entitled 'A Half Century of Streptococcal Research: Then we heard how Dr Pat Cleary was using Lux gene and Now' focused on landmark research undertaken technology to answer questions on the persistence of during the first half of the 20th Century. This research streptococci in nasal-associated lymphoid tissue. Another provided the foundations for understanding the new technology, microarray, had been used by Dr Larry pathogenesis and epidemiology of streptococcal McDaniel's group to analyse pneumococcal virulence diseases. Dr Krause particularly highlighted the work of factors. In a number of presentations amazing Rebecca Lancefield after whom the symposia are photographs were shown of streptococci attaching to, named. Lancefield identified structural differences in or internalising in, epithelial and endothelial cells. These organisms that allowed her to classify streptococci into photographs had been taken using scanning electron serogroups and serotypes and led her to define the microscopy. Dr Manfred Rodhe set a high standard with biologic significance of M protein on group A his extraordinary set of photographs suggesting that host streptococci. Dr Krause also stressed the importance cell caveole initiate invasion by group A streptococci. of studies undertaken by Wannamaker Rammelkamp, and Denny, who by careful epidemiologic study, An important focus of the Symposium was Rheumatic demonstrated that treatment of a streptococcal sore Fever (RF) and Rheumatic Heart Disease (RHD). Dr throat with penicillin could prevent rheumatic fever. The Padmavati set the scene by detailing the enormous burden second half of the 20th century saw increasing of RF and RHD that occurs globally but particularly in development of techniques for dissecting the structure developing countries. In India about six million children and function of streptococci, explaining the pathogenesis are suffering from RHD - a number nearly equivalent to of streptococcal disease and describing the epidemiology twice the total population of New Zealand. RHD is of disease manifestations. estimated to cause around 0.5 million deaths per year globally. Dr Ganguly highlighted the challenges facing One of the most striking impressions from the XVth India to control RF and RHD. He outlined some of the Lancefield Symposium was the apparent explosion of primary prevention strategies that were being new or advanced technologies being used to progress implemented at the community level. Dr Ganguly knowledge. At the time of the XIV Lancefield described research already undertaken in India that Symposium in 1999, the Group A Streptococcus had just identified a correlation between RF and HLA antigens been sequenced. Now at the XV Symposium Professor involving the HLA-DR locus. He stressed that if an M Jo Ferretti described the work that his laboratory and protein epitope-based vaccine was to be used in India it others had done in sequencing eight different strains would need to be tailor-made to take into account the M representing a number of streptococcal species. This types of streptococci circulating there. ix x INDIAN J MED RES (SUPPL) MAY 2004 Dr Madeleine Cunningham's most eloquent keynote of Streptococcus gordonii , an organism being tested address walked the participants through her group's for vaccine delivery. A third study reported, identified research. They have hypothesised that immune injury is C5a peptidase as a candidate vaccine antigen following caused by group A streptococci during the development demonstration that C5a peptidase elicits an immune of RF/RHD. She described their experimental evidence response in children naturally infected with group A that showed B cells activated by the carbohydrate of streptococci. The high conservation of C5a peptidase group A streptococci bind to the surface of the heart across all tested streptococcal serotypes suggests C5a valves and T cells activated by myosin-like peptides of peptidase may be an attractive vaccine candidate. M protein cause valvulitis with tissue scarring. Some subsequent presentations addressed the issue of The production and utilisation of vaccines first requires rheumatic carditis from the clinical perspective. well developed surveillance systems. Such systems Echocardiographic data, pathological findings, and inform on the frequency and distribution of strains causing surgical descriptions discussed were considered to disease, inform public health intervention strategies and support the importance of valvulitis rather than provide the information against which vaccines can be myocarditis in the pathology of RHD. Other presentations developed and their effectiveness measured. The and discussions associated with RF included treatment numbers of papers presented both orally and by poster regimen for secondary prophylaxis, the Jones criteria, that encompassed epidemiologic surveillance of clinical surgical and non-surgical interventions and disease and/or the serologic or molecular typing of strains immunogenetics of RHD. was proof that laboratories worldwide are very active in this area. The papers reported from Europe and North A landmark paper was presented by Dr Christine America mostly focused on dimensions of invasive Kirvan from Madeleine Cunningham's laboratory. The disease. In contrast, those from developing countries such paper entitled 'Antibody-mediated neuronal cell signaling as India, or from the Southern Hemisphere, focused in Sydenham's Chorea' demonstrated how anti- more on the surveillance of throat and skin infections streptococcal monoclonal antibodies derived from a and on Rheumatic Fever. It was apparent from the patient with Sydenham's Chorea were found to cross- presentations and discussions of data that differences react with ganglioside antigen and human caudate and exist in the clinical presenting features of the same brain sections. Furthermore, a monoclonal antibody disease in different parts of the world. Similarly, produced a significant increase in calcium/calmodulin- predominant group A strain types vary according to dependent protein kinase (CAM kinase II) in a human geography. Types occurring in the industrialised nations neuronal cell line. When acute phase serum from patients of the Northern Hemisphere are distinct from those with Sydenham's Chorea was tested, there was highly causing the same disease in developing countries of the increased CAM kinase activity that was not present in world. Within Australia, a similar dichotomy occurs with convalescent serum from the same patients. The data differences in disease presentations between the tropical suggest that antibody-mediated cell-signaling via CAM north and the more populated region of southern Australia. kinase II is important in the pathogenesis of Sydenham's Serogroup B streptococcal types and their disease Chorea. manifestations appear less diverse globally but this observation possibly reflects that information on In the absence of alternative population-based serogroup B disease and types causing it was not strategies for prevention of acute RF the focus is on the presented from the more tropical regions of the world. vaccine technology. The session devoted to vaccines The need to understand what strain types are causing provided fascinating listening. Dr Michael Good disease and where is of paramount importance for both described the development of chimeric peptides as serogroup A and serogroup B streptococci since potential vaccine candidates. He described how by vaccines based on type-specific polysaccharides or flanking the chimeric peptides with helical peptides the proteins would need to be tailor made to fit the structure was forced into a helical formation that induced circumstance. antibodies in a similar fashion to native M protein. In another presentation we heard results from studies that The range of techniques available and now used in had examined the kinetics of colonisation and elimination epidemiologic studies is providing a wealth of information XV LANCEFIELD SYMPOSIUM ON STREPTOCOCCI AND STREPTOCOCCAL DISEASES xi on strains types. However, to what extent the information resolved to make representation to WHO to ensure that is internationally comparable because of variations in recognition is given to the fact that any programme on techniques is uncertain. One typing system that has been the control of RF/RHD also requires quality surveillance standardised and quality controlled by a core of six of disease, including laboratory identification and typing reference laboratories is M typing. The move to a of streptococci. sequence-based emm typing, while permitting type description of all group A streptococci, is possibly Finally, the Organising Committee under the leadership differentiating streptococci that would be recognised as of Professor Ganguly and Professor Chhatwal must be biologically indistingushable in a functional assay. The congratulated both for the excellent conference and for need for evidence-based accurate information for the venue used. In preparing for the Symposium the determining strategic policies by WHO was stressed by Organising Committee had experienced considerable Dr Jonathan Carapetis. As an external consultant to difficulty in planning due to uncertainties of participation WHO he informed the conference of the current and international travel by overseas registrants in the organisational structure at WHO and of the fact that wake of terrorism and cross-border unrest. The fact that streptococcal disease had no current advocacy or many overseas scientists had travelled to India to programme in which it comfortably sat. During the participate indicated that the committee had made the conference the heads of national streptococcal correct decision to hold the XV LISSSD in Goa as laboratories present met to discuss the role for them in planned. The next meeting is to be held in Australia in addressing the burden of disease internationally. They September 2005.

Diana Martin President XV Lancefield International Symposium on Streptococci and Streptococcal Diseases Indian J Med Res 119 (Suppl) May 2004, pp 7-12

Short-chain oligosaccharide protein conjugates as experimental pneumococcal vaccines

W.T.M. Jansen & Harm Snippe

Eijkman-Winkler Institute, Heidelberglaan 100, 3584CX Utrecht, The Netherlands

Received August 7, 2003

Streptococcus pneumoniae remains a major cause of acute respiratory infections worldwide and is responsible for approximately 1 million childhood deaths each year. Despite the widespread use of antibiotics, the mortality and morbidity of pneumococcal disease remains high. Therefore, effective vaccines to prevent pneumococcal disease are needed. Bacterial vaccine development in general follows a similar track starting from large, rather crude vaccines (whole live, attenuated pathogen) towards smaller, better defined subunit vaccines. Ultimately, this track leads to the development and evaluation of minimal, highly defined subunit vaccines, based on a collection of single protective epitopes. This mini-review deals with capsular saccharide based vaccines. After a short overview of the development of pneumococcal vaccines from the 23 - valent polysaccharide vaccines to polysaccharide-protein conjugate vaccines, it focuses on the vaccine potential of synthetic oligosaccharides, conjugated to carrier proteins. Key words Oligosaccharide - Streptococcus pneumoniae - vaccine Worldwide, Streptococcus pneumoniae remains a polysaccharide (PS) 9. Pneumococcal capsular types major cause of acute respiratory bacterial infections, differ significantly in their virulence10. Of the 90 serotypes leading to approximately 1 million childhood deaths each known today, only a limited number of 20 serotypes 1 year . Antibiotic resistance is a growing problem in the cause the majority (90%) of disease10. Serotypes most 2 treatment of pneumococcal infections . Despite the wide commonly isolated from patients include serotypes 1 and spread use of antibiotics, the mortality and morbidity of 14 from the blood, serotypes 6, 10 and 23 from the pneumococcal disease remains high3. In addition, 3-5 cerebrospinal fluid and serotypes 3, 19 and 23 from the resistant pneumococci are increasingly observed . 11 Recently, a report describing vancomycin tolerant middle ear fluid of infants . Serotypes 6A, 6B, 14, 19F 9 pneumococci has alarmed the scientific community6. The and 23F are the main paediatric serotypes . These main risk groups for pneumococcal disease are infants serotypes are poorly immunogenic, which might be the below 2 yr and elderly above 60 yr7. Worldwide, reason for their success as a pathogen in infants. The approximately 1 million children die each year from capsular PS are very long polymers of either linear or pneumococcal pneumonia8. Therefore, renewed efforts branched repeating units consisting of 2 (serotypes 3, have been undertaken to develop effective vaccines 37) to 8 (serotype 17A) monosaccharides10. against pneumococci. Host protection against Streptococcus pneumoniae Capsular polysaccharide is mainly mediated by complement and antibody dependent phagocytosis. Capsular PS are the most Pneumococci can be subdivided into more than 90 important virulence factors of pneumococci9. Capsular different serotypes, based on their capsular PS provide resistance against complement mediated 7 8 INDIAN J MED RES (SUPPL) MAY 2004 phagocytosis12 and shield the inner structures of the pneumococcal pneumoniae24, but to a lesser extend to pneumococcus from the immune system13. As a otitis media26. It is therefore tempting to assume that the consequence, only antibodies against the capsular PS major obstacles in the prevention of pneumococcal have been shown to fully protect humans against invasive infections have been overcome. Despite the clinical and pneumococcal disease. However, PS are so called commercial potential of these vaccines, a number of thymus independent antigens, which means that they do major problems can still be envisioned. not recrute a T helper cell response14. As a consequence, the antibody response is mainly restricted to IgM and Protective capacities of these vaccines have not been IgA, affinity maturation does not occur, and B memory fully established yet, since information about protective cells are not produced14. opsonic antibody levels in humans is still lacking. In addition, the immunogenicity varies among the serotypes. 23-valent PS vaccine Especially serotype 6B may be a problematic serotype in currently tested conjugate vaccines27. Furthermore, The capsular PS of the 23 most prevalent serotypes multivalent conjugate vaccines are very expensive and have been included in the licenced 23-valent PS therefore not appropriate for vaccination programmes vaccine15,16. This vaccine provides protection against 73 in the developing world. The most important problem is per cent of the strains, and reduces the risk of systemic the number of serotypes that has to be included in the infection in the adult population to 83 per cent17. The vaccine. Although a multivalent conjugate vaccine that comprises the 9 different most virulent serotypes would efficacy of the vaccine, however is debatable in groups cover 80 to 90 per cent of the pneumococcal serotypes at risk for pneumococcal infections18,19. In the elderly, isolated worldwide from patients11, it is to be expected ineffective antibody responses after vaccination have that a shift in incidence towards non-vaccine serotypes been reported. However, a meta analysis of clinical PS may occur, or new serotypes may emerge. Since the vaccination studies including more than 65000 patients, 23-valent PS vaccine has proven to be safe, and the did not show differences in vaccine induced antibody 17 immunogenicity of each serotype seems not to be levels between adults and elderly . Nevertheless, in the suppressed by the other serotypes, similar benefits can elderly, avidity of these antibodies is lower, which might be expected from multivalent PS conjugate vaccines. correlate with an increased vulnerability to Nevertheless, the efficacy and safety of the vaccine 20 pneumococcal disease . For infants, the poor should be controlled when additional serotypes have to immunogenicity of the PS vaccine is beyond be incorporated in the vaccine. Immunogenicity may be 21,22 doubt and thereby is unsuitable. hampered by antigenic competition, carrier suppression28, unwanted cross reactions with host antigens may occur29 Polysaccharide-protein conjugate vaccines and finally, since capsular PS are indirectly linked to cell wall-PS30, increasing the number or doses of serotypes The conventional 23-valent PS vaccine provides in the vaccine, will result in a similar increase of cell insufficient protection in groups at risk for pneumococcal wall-PS content, which may affect the safety of the infection. To overcome the thymus independency of PS vaccine. antigens, PS are conjugated to carrier proteins, according to the immunological concepts developed in the mid Protein vaccines seventies. These conjugate vaccines are immunogenic in infants, but in general, antibody responses in adults Taken into account these obstacles towards a safe are not improved as compared to the conventional 23- and effective, worldwide pneumococcal vaccine, other valent PS vaccine23. Recently clinical phase III trials vaccine strategies should also be considered. To have been conducted for several conjugate vaccines24. overcome the disadvantages of serotype-specific These vaccines include the serotypes 4, 6B, 9V, 14, 18C vaccines, a number of pneumococcal proteins are being and 23F. As carrier proteins, either diphteria toxoid evaluated as vaccine candidates. These proteins include CRM197, meningococcus B outer membrane proteins pneumolysin, autolysin, pneumococcal surface protein or tetanus toxoid are used25. Several of these conjugate A (PspA), and pneumococcal surface adhesin A (PsaA). vaccines have been licenced and seem effective against Although most proteins induce a certain level of protection JANSEN & SNIPPE : PNEUMOCOCCAL VACCINES 9 against pneumococcal carriage and invasive disease in saccharides induced opsonic anti PS antibodies in mouse mice, their protective capacities in humans remain models and rabbit models33,39-43. For all serotypes tested uncertain31,32. so far, anti-PS responses could be induced by very small OS fragments comprising only 1 repeating unit or less of Synthetic oligosaccharides in pneumococcal the corresponding PS (Table). The serotypes listed in vaccines the Table belong to the most common ones isolated from adults and infants. Immune responses against the OS Pneumococcal capsular PS epitopes can be presented were highly specific and interestingly, some of the OS- by synthetic oligosaccharides (OS). The synthetic protein conjugate vaccines even induced higher levels components can be conjugated to a carrier protein with of anti-PS antibodies than the conventional PS-protein conventional coupling chemistry, to obtain semi-synthetic conjugate counterpart41. conjugate vaccines. The preparation of synthetic saccharides that represent fragments of pneumococcal Although the vaccine potential of small synthetic OS PS antigens is traditionally difficult. Nevertheless, the is clear from studies in murine models, their potential as combination of organic synthesis and biosynthesis using vaccine candidates in humans is still uncertain. It is known glycosyl transferases has lead to the production of that epitope size on pneumococcal PS can vary between numerous synthetic saccharides for serotypes 1-4, serotypes42,52, between species42 and with age53. In 6A/B, 7F, 8, 9A/V, 14, 17F, 18C, 19A/F, 22F, 23F, 27 and contrast to mice, humans may require longer saccharides 2933. These saccharides have been used as inhibitors to for an optimal immune response, presumably because determine the epitope specificity of anti-PS antibodies. of the better presentation of conformational epitopes53. In addition, synthetic OS can be used to study, and block, Again, this may vary per serotype since 6A and 6B41, pneumococcal adhesion to epithelial cells34-38. Finally, but not 23F42 OS fragments were recognized by human coupled to carrier proteins, a growing number of synthetic vaccination antisera, indicating the presence of epitopes saccharides is evaluated as a conjugate vaccine in murine for human anti-PS antibodies on the former but not the models. Within the vaccine department at the University latter OS fragments. Medical Center, Utrecht in collaboration with the Bijvoet Centre in Utrecht, several synthetic saccharides are A potential drawback of minimal synthetic OS under investigation. Synthetic 6B, 17F and 23F conjugate vaccines is the absence of displaying

Table. Minimal oligosaccharide (OS) fragments conjugated to carrier proteins induced (protective) anti-PS responses in mice and rabbits OS chain lengtha inducing (protective) anti-PS responses References Serotype Carrier- protein mice rabbits

Anti- PS protectionc Anti-PS protectiond responseb response 3 CRM 197 di, tri, tetra di, tri, tetra ND ND 40,43,44

6A KLH tri, tetra ND tri, tetra tri, tetra 41,45-47

6B KLH di, tri, tetra tetra di, tri, tetra di, tri, tetra 14 CRM 197 tetra, octa ND ND ND 39,48 17F KLH di, tri di, tri di, tri ND 44,49,50

23F KLH ND ND tetra ND 42,51 a, number of monosaccharides within 1 OS fragment, coupled to a carrier protein; b, as measured by ELISA using PS as coating antigen41; c, measured in a mouse challenge model41; d, passive protection of mice by anti OS-antibodies raised in rabbits41; ND, not done 10 INDIAN J MED RES (SUPPL) MAY 2004 conformational epitopes, which are present on the whole needed and when possible, natural PS may then be PS. On the other hand, the thymus dependency of an replaced sequentially by their (bio)synthetic OS OS-conjugate increases with shortening of the OS chain counterparts in future semi-synthetic pneumococcal length. Therefore, for PS containing conformational vaccines. epitopes, an optimum OS length has to be defined, long enough to express conformational epitopes, yet short References enough to induce an optimal Th cell response. To obtain OS that contain several repeating units of the PS, three 1. Pneumococcal vaccines. WHO position paper. Wkly Epidemiol methods can be applied: partial hydrolysis of the PS, Rec 1999; 74 : 177-83. 33 chemical synthesis, or biosynthesis . Several OS have 2. Tan TQ. Pneumococcal conjugate vaccines - Implications for been successfully prepared by hydrolysis. A community antibiotic prescribing. Curr Opin Microbiol 2000; disadvantage of this method is that the production of 3 : 502-7. these OS is less well controlled leading to less defined 3. Kellner JD. Drug-resistant Streptococcus pneumoniae infections: products (size and exact bound hydrolized) and the clinical importance, drug treatment, and prevention. Semin Respir presence of impurities. Chemical synthesis of OS of Infect 2001; 16 : 186-95. several repeating units starts to become possible, but is 4. Whitney CG, Farley MM, Hadler J, Harrison LH, Lexau C, still very complex, requiring a large number of synthesis Reingold A, et al. Increasing prevalence of multidrug-resistant Streptococcus pneumoniae in the United States. N Engl J Med steps. The combination of chemical synthesis and 2000; 343 : 1917-24. biosynthesis might be a promising alternative in the near future to produce large OS at sufficient amounts. 5. Schreiber JR, Jacobs MR. Antibiotic-resistant pneumococci. Pediatr Clin North Am 1995; 42 : 519-37. Information about the biosynthetic pathways of the capsular PS of serotypes 3, 14 and 19F has become 6. Wenzel RP, Edmond MB. Managing antibiotic resistance. N Engl J Med 2000; 343 : 1961-3. available33. Several glycosyl transferases are identified and cloned54-56 and one of them has already been 7. Rijkers GT, Sanders LA, Zegers BJ. Anti-capsular 57 polysaccharide antibody deficiency states. Immunodeficiency produced at a large scale . Even conjugation of certain 1993; 5 : 1-21. saccharides to a carrier protein or peptide might be 8. Sniadack DH, Schwartz B, Lipman H, Bogaerts J, Butler JC, performed in part enzymatically in the future, since Dagan R, et al. Geographic and temporal differences in serotype several enzymes have now been identified that can and serogroup distribution of sterile site pneumococcal isolates glycosylate specific amino acids residues 54,58. In addition from children - Implications for vaccine strategies. Pediatr Infect to new strategies to synthesize saccharides, recently Dis J 1995; 14 : 503-10. developed techniques such as Matrix-Associated Laser 9. AlonsoDeVelasco E, Verheul AF, Verhoef J, Snippe H. Desorption time of flight mass spectrometry (MALDI- Streptococcus pneumoniae: Virulence factors, pathogenesis, and tof) facilitate and improve the characterization of the vaccines. Microbiol Rev 1995; 59 : 591-603. product. 10. van Dam JE, Fleer A, Snippe H. Immunogenicity and immunochemistry of Streptococcus pneumoniae capsular polysaccharides. Antonie Van Leeuwenhoek 1990; 58 : 1-47. Our findings and the work of others on synthetic OS- protein conjugates encourage the exploration of synthetic 11. Hausdorff WP, Yothers G, Dagan R, Kilpi T, Pelton SI, Cohen R, et al. Multinational study of pneumococcal serotypes OS conjugate vaccines as a pneumococcal vaccine. causing acute otitis media in children. Pediatr Infect Dis J Nevertheless, to establish whether animals and humans 2002; 21 : 1008-16. can produce antibodies against minimal synthetic OS, is 12. Brown EJ, Joiner KA, Cole RM, Berger M. Localization of only the first step towards a synthetic pneumococcal complement component 3 on Streptococcus pneumoniae: anti- vaccine. Next step is to define the most immunogenic, capsular antibody causes complement deposition on the pneumococcal. Capsule Infect Immun 1983; 39 : 403-9. protective epitopes on PS and optimize the presentation of such epitopes to the immune system. 13. Nielsen SV, Sorensen UB, Henrichsen J. Antibodies against pneumococcal C-polysaccharide are not protective. Microb Pathog 1993;14 : 299-305. In conclusion, considering the large number of 14. Stein KE. Thymus-independent and thymus-dependent serotypes, the generation of new synthetic saccharides responses to polysaccharide antigens. J Infect Dis 1992; 165 should be stimulated. Depending on the serotype, when (Suppl 1) : S49-S52. JANSEN & SNIPPE : PNEUMOCOCCAL VACCINES 11

15. Requejo HI. Polyvalent pneumococcal polysaccharide vaccines; 28. Dagan R, Eskola J, Leclerc C, Leroy O. Reduced response to a review of the literature. Rev Hosp Clin Fac Med Sao Paulo multiple vaccines sharing common protein epitopes that are 1993; 48 : 130-8. administered simultaneously to infants. Infect Immun 1998; 66 : 2093-8. 16. Robbins JB, Austrian R, Lee CJ, Rastogi SC, Schiffman G, Henrichsen J, et al. Considerations for formulating the second- 29. Finne J, Leinonen M, Makela PH. Antigenic similarities between generation pneumococcal capsular polysaccharide vaccine with brain components and bacteria causing meningitis. Implications emphasis on the cross-reactive types within groups. J Infect for vaccine development and pathogenesis. Lancet 1983; 2 : Dis 1983; 148 : 1136-59. 355-7. 17. Hutchison BG, Oxman AD, Shannon HS, Lloyd S, Altmayer 30. Sorensen UB, Henrichsen J, Chen HC, Szu SC. Covalent linkage CA, Thomas K. Clinical effectiveness of pneumococcal vaccine. between the capsular polysaccharide and the cell wall Meta-analysis. Can Fam Physician 1999; 45 : 2381-93. peptidoglycan of Streptococcus pneumoniae revealed by immunochemical methods. Microb Pathog 1990; 8 : 325-34. 18. Avanzini MA, Carra AM, Maccario R, Zecca M, Pignatti P, Marconi M, et al. Antibody response to pneumococcal vaccine 31. Briles DE, Hollingshead SK, Nabors GS, Paton JC, Brooks- in children receiving bone marrow transplantation. J Clin Walter A. The potential for using protein vaccines to protect Immunol 1995; 15 : 137-44. against otitis media caused by Streptococcus pneumoniae. Vaccine 2000;19 (Suppl 1) : S87-S95. 19. King JC Jr, Vink PE, Farley JJ, Parks M, Smilie M, Madore D, 32. Zysk G, Bongaerts RJ, ten Thoren E, Bethe G, Hakenbeck R, et al. Comparison of the safety and immunogenicity of a Heinz HP. Detection of 23 immunogenic pneumococcal proteins pneumococcal conjugate with a licensed polysaccharide vaccine using convalescent phase serum. Infect Immun 2000; 68 : in human immunodeficiency virus and non-human 3740-3. immunodeficiency virus-infected children. Pediatr Infect Dis J 1996; 15 : 192-6. 33. Kamerling JP. Pneumococcal polysaccharides: A chemical view. In: Tomasz A, editor. Streptococcus pneumoniae, molecular 20. Romero-Steiner S, Musher DM, Cetron MS, Pais LB, Groover biology & mechanisms of disease. New York: Mary Ann Liebert; JE, Fiore AE, et al. Reduction in functional antibody activity 1999 p. 81-114. against Streptococcus pneumoniae in vaccinated elderly individuals highly correlates with decreased IgG antibody avidity. 34. Sharon N, Ofek I. Safe as mother’s milk: Carbohydrates as future Clin Infect Dis 1999; 29 : 281-8. anti-adhesion drugs for bacterial diseases. Glycoconj J 2000; 17 : 659-64. 21. Meyer M, Gahr M. Immunological principles of polysaccharide- protein conjugate vaccination. Monatsschr Kinderheilkd 1993; 35. Ukkonen P, Varis K, Jernfors M, Herva E, Jokinen J, 141 : 770-6. Ruokokoski E, et al. Treatment of acute otitis media with an antiadhesive oligosaccharide: a randomised, double-blind, 22. Pomat WS, Lehmann D, Sanders RC, Lewis DJ, Wilson J, Rogers placebo-controlled trial. Lancet 2000; 356 : 1398-402. S, et al. Immunoglobulin G antibody responses to polyvalent pneumococcal vaccine in children in the highlands of Papua 36. Tong HH, McIver MA, Fisher LM, DeMaria TF. Effect of New Guinea. Infect Immun 1994; 62 : 1848-53. lacto-N-neotetraose, asialoganglioside-GM1 and neuraminidase on adherence of otitis media-associated serotypes of 23. Briles DE, Ades E, Paton JC, Sampson JS, Carlone GM, Streptococcus pneumoniae to chinchilla tracheal epithelium. Huebner RC, et al. Intranasal immunization of mice with a mixture Microb Pathog 1999; 26 : 111-9. of the pneumococcal proteins PsaA and PspA is highly 37. Barthelson R, Mobasseri A, Zopf D, Simon P. Adherence of protective against nasopharyngeal carriage of Streptococcus Streptococcus pneumoniae to respiratory epithelial cells is pneumoniae. Infect Immun 2000; 68 : 796-800. inhibited by sialylated oligosaccharides. Infect Immun 1998; 24. Lee LH, Lee CJ, Frasch CE. Development and evaluation of 66 : 1439-44. pneumococcal conjugate vaccines: clinical trials and control tests. 38. Idanpaan-Heikkila I, Simon PM, Zopf D, Vullo T, Cahill P, Crit Rev Microbiol 2002; 28 : 27-41. Sokol K, et al. Oligosaccharides interfere with the establishment 25. Eskola J. Polysaccharide-based pneumococcal vaccines in the and progression of experimental pneumococcal pneumonia. prevention of acute otitis media. Vaccine 2000; 19 (Suppl 1) : J Infect Dis 1997; 176 : 704-12. S78-S82. 39. Mawas F, Niggemann J, Jones C, Corbel MJ, Kamerling JP, 26. Straetemans M, Sanders EA, Veenhoven RH, Schilder AG, Vliegenthart JF. Immunogenicity in a mouse model of a conjugate Damoiseaux RA, Zielhuis GA. Pneumococcal vaccines for vaccine made with a synthetic single repeating unit of type 14 preventing otitis media. Cochrane Database Syst Rev 2002; 2 : pneumococcal polysaccharide coupled to CRM197. Infect CD001480. Immun 2002; 70 : 5107-14. 27. Sorensen RU, Leiva LE, Giangrosso PA, Butler B, Javier FC III, 40. Benaissa-Trouw B, Lefeber DJ, Kamerling JP, Vliegenthart JF, Sacerdote DM, et al. Response to a heptavalent conjugate Kraaijeveld K, Snippe, H. Synthetic polysaccharide type 3- Streptococcus pneumoniae vaccine in children with recurrent related di-, tri-, and tetrasaccharide- CRM(197) conjugates induce infections who are unresponsive to the polysaccharide vaccine. protection against Streptococcus pneumoniae type 3 in mice. Pediatr Infect Dis J 1998; 17 : 685-91. Infect Immun 2001; 69 : 4698-701. 12 INDIAN J MED RES (SUPPL) MAY 2004

41. Jansen WT, Hogenboom S, Thijssen MJ, Kamerling JP, synthetic di-, tri- and tetrasaccharide conjugates in mice and Vliegenthart JF, Verhoef J, et al. Synthetic 6B di-, tri-, and rabbits. Vaccine 2001; 20 : 19-21. tetrasaccharide-protein conjugates contain pneumococcal type 50. Jones C, Whitley C, Lemercinier X. Full assignment of the 6A and 6B common and 6B-specific epitopes that elicit protective antibodies in mice. Infect Immun 2001; 69 : 787-93. proton and carbon NMR spectra and revised structure for the capsular polysaccharide from Streptococcus pneumoniae type 42. Alonso De V, Verheul AF, van Steijn AM, Dekker HA, Feldman 17F. Carbohydr Res 2000; 325 : 192-201. RG, Fernandez IM, et al. Epitope specificity of rabbit immunoglobulin G (IgG) elicited by pneumococcal type 23F 51. van Steijn AM, Kamerling JP, Vliegenthart JF. Synthesis of a synthetic oligosaccharide- and native polysaccharide-protein spacer-containing repeating unit of the capsular polysaccharide conjugate vaccines: Comparison with human anti-polysaccharide of Streptococcus pneumoniae type 23F. Carbohydr Res 1991; 23F IgG. Infect Immun 1994; 62 : 799-808. 211 : 261-77. 43. Snippe H, van Dam JE, Van Houte AJ, Willers JM, Kamerling 52. Alonso De V, Verheul AF, Veeneman GH, Gomes LJ, van Boom JP, Vliegenthart JF. Preparation of a semisynthetic vaccine to JH, Verhoef J, et al. Protein-conjugated synthetic di- and Streptococcus pneumoniae Type 3. Infect Immun 1983; 42 : trisaccharides of pneumococcal type 17F exhibit a different 842-4. immunogenicity and antigenicity than tetrasaccharide. Vaccine 1993;11 : 1429-36. 44. Lefeber DJ, Kamerling JP, Vliegenthart JF. Synthesis of Streptococcus pneumoniae type 3 neoglycoproteins varying in 53. Pichichero ME, Porcelli S, Treanor J, Anderson P. Serum antibody oligosaccharide chain length, loading and carrier protein. responses of weaning mice and two-year-old children to Chemistry 2001; 7 : 4411-21. pneumococcal-type 6A protein conjugate vaccines of differing 45. Thijssen MJ, Bijkerk MH, Kamerling JP, Vliegenthart JF. saccharide chain lengths. Vaccine 1998; 16 : 83-91. Synthesis of four spacer-containing tetrasaccharides' that 54. Shpak E, Leykam JF, Kieliszewski MJ. Synthetic genes for represent four possible repeating units of the capsular glycoprotein design and the elucidation of hydroxyproline-O- polysaccharide of Streptococcus pneumoniae type 6B. glycosylation codes. Proc Natl Acad Sci USA 1999; 96 : 14736- Carbohydr Res 1998; 306 : 111-25. 41. 46. Thijssen MJ, van Rijswijk MN, Kamerling JP, Vliegenthart JF. 55. Di Virgilio S, Glushka J, Moremen K, Pierce M. Enzymatic Synthesis of spacer-containing di- and tri-saccharides that synthesis of natural and 13C enriched linear poly-N- represent parts of the capsular polysaccharide of Streptococcus acetyllactosamines as ligands for galectin-1. Glycobiology 1999; pneumoniae type 6B. Carbohydr Res 1998; 306 : 93-109. 9 : 353-64. 47. Thijssen MJ, Halkes KM, Kamerling JP, Vliegenthart JF. Synthesis of a spacer-containing tetrasaccharide representing a 56. Elling L. Glycobiotechnology: enzymes for the synthesis of repeating unit of the capsular polysaccharide of Streptococcus nucleotide sugars. Adv Biochem Eng Biotechnol 1997; 58 : 89- pneumoniae type 6B. Bioorg Med Chem 1994; 2 : 1309-17. 144. 48. Niggemann J, Kamerling JP, Vliegenthart JF. Beta-1,4- 57. Yamamoto T, Nagae H, Kajihara Y, Terada I. Mass production galactosyltransferase-catalyzed synthesis of the branched of bacterial alpha 2,6-sialyltransferase and enzymatic syntheses tetrasaccharide repeating unit of Streptococcus pneumoniae type of sialyloligosaccharides. Biosci Biotechnol Biochem 1998; 62 : 14. Bioorg Med Chem 1998; 6 : 1605-12. 210-4. 49. Jansen WT, Verheul AF, Veeneman GH, van Boom JH, 58. Imperiali B, O’Connor SE. Effect of N-linked glycosylation on Snippe H. Revised interpretation of the immunological results glycopeptide and glycoprotein structure. Curr Opin Chem Biol obtained with pneumococcal polysaccharide 17F derived 1999; 3 : 643-9. Reprint requests: Dr W.T.M. Jansen, Eijkman-Winkler Institute, G04-614 UMC, Microbiology Heidelberglaan 100 3584 CX Utrecht, The Netherlands e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 13-16

Prevention of streptococcal pharyngitis by anti-Streptococcus pyogenes bacteriocin-like inhibitory substances (BLIS) produced by Streptococcus salivarius

J.R. Tagg

Department of Microbiology, University of Otago, P.O. Box 56, Dunedin, New Zealand

Received August 6, 2003

Background & objectives: Streptococcus salivarius is a numerically prominent member of the human oral microbiota that produces a variety of bacteriocin-like inhibitory substances (BLIS) having in vitro inhibitory activity against S. pyogenes. Our previous studies of S. salivarius isolates from children using a deferred antagonism BLIS production (P)-typing scheme showed that the 9 per cent of children having large populations of P-type 677 S. salivarius experienced fewer S. pyogenes acquisitions than either the 11 per cent of children having predominant P-type 226 populations or the 60 per cent of children with largely non-inhibitory (P-type 000) S. salivarius. Amongst the other BLIS P-types detected were a number of strongly-inhibitory (P-type 777) S. salivarius. In the present study the inhibitory agents produced by prototype strains of P-types 226, 677 and 777 S. salivarius are compared. Methods: The prototype BLIS-producing S. salivarius strains SN, 20P3, and K12 were isolated from tongue swabbings. BLIS P-typing was done using standard procedures. The BLIS molecules were purified and characterized. Results: S. salivarius SN (P-type 226) produces a heat-labile muramidase. S. salivarius 20P3 (P-type 677) produces the 2315 Da lantibiotic salivaricin A and S. salivarius K12 (P-type 777) produces two lantibiotics; salivaricin A2 (2368 Da) and salivaricin B (2733 Da). Interpretation & conclusion: The P-type 777 S. salivarius strain produced salivaricin A2 and salivaricin B. The combined production of two anti-S. pyogenes BLIS activities by this strain indicates that it could be adopted as a colonizing strain in bacterial interference trials. Key words Bacteriocin - bacteriocin-like inhibitory substance - lantibiotic - replacement therapy - Streptococcus salivarius

Streptococcus pyogenes is a major cause of acute possibility of adverse host reactions, severe disruption infections and of serious delayed sequelae, particularly of the indigenous microbiota and bacterial resistance in children. At present, the most effective strategy development. The implementation of bacterial available to deal with S. pyogenes is treatment of acute interference may offer a relatively specifically-targeted infections by administration of therapeutic doses of a alternative means of preventing the development of acute broad-spectrum antibiotic like penicillin. Also, since there S. pyogenes infections. Since commensal streptococci is currently no anti-S. pyogenes immunization available, are numerically predominant in the oral cavity, it seems the only means of protecting at risk individuals is antibiotic reasonable to speculate that they may be central to any prophylaxis. However, these approaches have many naturally occurring interference with colonization or inherent problems: the cost of the antibiotics, the infection by S. pyogenes. Children who acquire

13 14 INDIAN J MED RES (SUPPL) MAY 2004 S. pyogenes have been shown to have a lower Material & Methods proportion of throat cultures containing bacteria inhibitory or bacteriocidal for S. pyogenes than those who do not Bacterial strains and BLIS P-typing: The prototype become colonised1. The pantothenic acid antagonist BLIS-producing S. salivarius strains SN (P-type 226), enocin, produced by S. salivarius, was later speculated 20P3 (P-type 667)10 and K12 (P-type 777) were isolated to contribute to protection against S. pyogenes on Mitis Salivarius agar (Difco Laboratories, Detroit) infections2. Grahn and Holm3 established that the from tongue swabbings of human subjects. The prevalence of oral alpha-haemolytic streptococci having 9 standard indicators and the standard procedure used inhibitory activity against S. pyogenes was lower in those for the BLIS P-typing of streptococci have been children who became infected during an outbreak of described previously7. streptococcal tonsillitis. This group later recommended dosing with a mixture of alpha haemolytic streptococci Isolation and characterization of BLIS molecules: as a supplementary treatment of recurrent streptococcal The BLIS produced by S. salivarius strain SN was tonsillitis4. purified from 18 h 35C Todd Hewitt broth (Difco) culture supernatants using the procedures previously Of all the bacterial species known to regularly inhabit devised for the similar BLIS, zoocin A, produced by the human oral microbiota in large numbers S. salivarius S. equi subsp. zooepidemicus11. The procedures for is perhaps the most innocuous. There appear to be no isolation of the lantibiotic peptides salivaricin A2 and reports of this species causing infections within the oral salivaricin B from S. salivarius strain K12 were those cavity and the rare instances of its association with applied previously to the purification of salivaricin A bacteraemia or meningitis appear to have occurred in from S. salivarius strain 20P310. Amino acid immunologically-compromised patients or following compositions, mass spectrometry and N-terminal amino trauma to the patients tissues5. Since S. salivarius is acid sequencing were done by the Protein common, not only on the dorsum of the tongue but also Microchemistry Facility, Department of Biochemistry, on the oropharyngeal mucosa6 , it is well positioned to University of Otago. To enable Edman degradation to directly repel invasion by S. pyogenes. Our previous proceed through blockages caused by the presence of studies7,8 have demonstrated that approximately 45 per dehydro amino acids, these residues were first modified cent of S. salivarius strains inhibit the growth of one or by addition of thiol groups12. more members of a set of 9 strains used routinely as indicators of streptococcal BLIS. The pattern of inhibition Results of these indicators, when expressed in code form, is referred to as the BLIS production (P)-type of a A muralytic enzyme, salivaricin SN, was isolated bacterium. Subsequent studies of streptococcal from the supernatants of Todd Hewitt broth cultures pharyngitis in populations of school-aged children showed of S. salivarius strain SN. Tests of the inhibitory that some children were seldom infected. Many of these spectrum of strain SN using the deferred antagonism rarely-infected children were found to harbour large test established that 18 of 20 S. pyogenes were populations of S. salivarius producing anti-S. pyogenes susceptible, but none of a variety of other streptococci, BLIS activity. A study of 780 Dunedin school children9 including representatives of Lancefield groups B, C, found that there were two major types of BLIS activities D, F and G, S. salivarius, S. sanguinis and produced by the S. salivarius, the corresponding P-type S. mutans. The purified protein was inactivated on patterns being referred to as 226 (11% of children positive) heating at 80º C for 30 min. The N-terminal sequence and 677 (9% positive). A further 20 per cent of the of salivaricin SN was DINGGANTPGAYD……. children had S. salivarius of various other P-type designations, including some isolates producing S. salivarius strain 20P3 has been shown particularly strong (P-type 777) BLIS activity. The previously10 to produce the lantibiotic salivaricin A present study compares the BLIS activities produced (Table). Also shown in the Table is the sequence of by S. salivarius strains SN, 20P3 and K12, the salivaricin A1, the homologous lantibiotic produced by prototypes of BLIS P-types 226, 677 and 777 respectively. the serotype M4 S. pyogenes strain 2000613,14. TAGG : APPLICATION OF STREPTOCOCCUS SALIVARIUS BLIS 15 S. salivarius strain K12 was found to produce two in 15 probe-positive S. pyogenes revealed consistent anti-S. pyogenes peptides (Table). The detection of codon differences indicative of conservative amino acids lanthionine in amino acid analyses and stability to heating changes in positions 2 (K instead of R) and 7 (F instead at 100ºC for 30 min indicated that both are of the of I) of the salivaricin A propeptide. This variant form, lantibiotic class of bacteriocins. One of these was a T4S/ called salivaricin A1, appears to be expressed as a F7I variant of salivaricin A, named salivaricin A2. The functional product only in serotype M4 strains of second was quite unrelated to salivaricin A and is referred S. pyogenes but is sufficiently similar to effect induction to as salivaricin B. of salivaricin A production in salA-positive strains of S.salivarius. Discussion In the present study, yet another variant of salivaricin The present study shows that the anti-S. pyogenes A has been found to be produced by S. salivarius strain inhibitory activities previously shown to be relatively K12. Interestingly this variant, salivaricin A2, which commonly produced in vitro by S. salivarius are of three differs in two conservative amino acid substitutions markedly different molecular types. Strain SN, the from both salivaricin A and salivaricin A1, appears prototype of the P-type 226 S. salivarius produced a commonly to be produced together with salivaricin B heat labile lytic enzyme, salivaricin SN, the N-terminus in P-type 777 S. salivarius (data not shown). The of which was completely different from that of the P-type 777 S. salivarius strain K12 was shown to previously-described streptococcal muralysin, zoocin A produce in addition to salivaricin A2 an unrelated (ATYTRPLDTG…). Zoocin A has been shown to be lantibiotic salivaricin B. The combined production of a domain-structured enzyme similar to lysostaphin, with two anti-S. pyogenes BLIS activities by strain K12 the N-terminus responsible for catalysis and the was one consideration leading to its adoption as a C-terminal domain effecting target recognition15. Unlike colonizing strain in bacterial interference trials in zoocin A, which displays strong activity against S. mutans populations of school children and young adults. in addition to S. pyogenes, the S. salivarius agent appears relatively specifically active against Acknowledgment S. pyogenes. Cloning of the salivaricin SN structural gene is in progress and when completed will facilitate The financial support from the Marsden Fund, Royal Society of New Zealand and by the Health Research Council of New Zealand is direct comparison between the two enzymes for evidence acknowledged. of domain sharing. References Our previous studies13,14 have shown that the structural gene salA, which encodes the lantibiotic 1. Crowe CC, Sanders WE Jr, Longley S. Bacterial interference. II. salivaricin A, is the usual form detected in S. salivarius, Role of the normal throat flora in prevention of colonization by PCR amplification and sequencing of the salA-like gene group A streptococcus. J Infect Dis 1973; 128: 527-32.

Table. Primary sequence and molecular mass of lantibiotic salivaricins

Salivaricin Amino acid sequence Mass (Da)

A* KRGSGWIATITDDCPNSVFVCC 2315

A1* KKGSGWFATITDDCPNSVFVCC 2327

A2 KRGTGWFATITDDCPNSVFVCC 2368

B GGGVIQTISHECRMNSWQFLFTCCS 2733

* The data for salivaricin A10 and salivaricin A113 have been previously published 16 INDIAN J MED RES (SUPPL) MAY 2004

2. Sanders CC, Sanders WE Jr. Enocin: an antibiotic produced by and streptococcal diseases entering the new millenium. Streptococcus salivarius that may contribute to protection New Zealand: Securacopy; 2000 p. 81-5. against infections due to group A streptococci. J Infect Dis 10. Ross KF, Ronson CW, Tagg JR. Isolation and characterization 1982; 146 : 683-90. of the lantibiotic salivaricin A and its structural gene salA from 3. Grahn E, Holm SE. Bacterial interference in the throat flora Streptococcus salivarius 20P3. Appl Environ Microbiol 1993; during a streptococcal tonsillitis outbreak in an apartment house 59: 2014-21. area. Zentralbl Bakteriol Mikrobiol Hyg (A) 1983; 256: 72-9. 11. Simmonds RS, Naidoo J, Jones CL, Tagg JR. The streptococcal 4. Roos K, Holm SE, Grahn E, Lind L. Alpha-streptococci as bacteriocin-like inhibitory substance, zoocin A, reduces the supplementary treatment of recurrent streptococcal tonsillitis: proportion of Streptococcus mutans in an artificial plaque. a randomized placebo controlled study. Scand J Infect Dis 1993; Microb Ecol Health Dis 1995; 8: 281-92. 25: 31-5. 5. Carley NH. Streptococcus salivarius bacteremia and meningitis 12. Navaratna MA, Sahl HG, Tagg JR. Two-component anti- following upper gastrointestinal endoscopy and cauterization Staphylococcus aureus lantibiotic activity produced by for gastric bleeding. Clin Infect Dis 1992; 14 : 947-8. Staphylococcus aureus C55. Appl Environ Microbiol 1998; 64 : 4803- 8. 6. Frandsen EV, Pedrazzoli V, Kilian M. Ecology of viridans streptococci in the oral cavity and pharynx. Oral Microbiol 13. Simpson WJ, Ragland NL, Ronson CW, Tagg JR. A lantibiotic Immunol 1991; 6: 129-33. gene family widely distributed in Streptococcus salivarius and Streptococcus pyogenes. Dev Biol Stand 1995; 85 : 639-43. 7. Tagg JR, Bannister LV. “Fingerprinting” ß-haemolytic streptococci by their production of and sensitivity to 14. Upton M, Tagg JR, Wescombe P, Jenkinson HF. Intra- bacteriocine-like inhibitors. J Med Microbiol 1979; 12 : 397-411. and interspecies signaling between Streptococcus 8. Dempster RP, Tagg JR. The production of bacteriocin-like salivarius and Streptococcus pyogenes mediated by SalA substances by the oral bacterium Streptococcus salivarius. and SalA1 lantibiotic peptides. J Bacteriol 2001; 183: Arch Oral Biol 1982; 27: 151-7. 3931-8. 9. Dierksen KP, Tagg JR. The influence of indigenous bacteriocin- 15. Lai AC, Tran S, Simmonds RS. Functional characterization of producing Streptococcus salivarius on the acquisition of domains found within a lytic enzyme produced by Streptococcus Streptococcus pyogenes by primary school children in Dunedin, equi subsp. zooepidemicus. FEMS Microbiol Lett 2002; 215: New Zealand. In: Martin DR, Tagg JR, editors. Streptococci 133.

Reprint requests: Dr J.R. Tagg, Professor, Department of Microbiology, University of Otago, P.O. Box 56, Dunedin, New Zealand e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 17-21

Report of unusual clinical appearance in bacteraemia with non- haemolytic M-type 58 Streptococcus pyogenes

Nigel C. Weightman & Michael R.D. Barnham*

Department of Microbiology, Friarage Hospital, Northallerton & *Harrogate District Hospital North Yorkshire, UK

Received August 6, 2003

Background & objectives: The occurrence of haemolytic colonies on blood agar often provides the starting point for the laboratory diagnosis of pyogenic streptococci, while non-haemolytic variants could pass unrecognised, leading to a failure of diagnosis. We report the details of two epidemiologically unrelated patients with bacteraemia featuring M-type 58 Streptococcus pyogenes, a seemingly rare cause of human infection in the UK, and briefly review previous reports of infection with non-haemolytic strains of this species. Methods: Case notes of the two patients were reviewed. Isolates obtained from clinical specimens were recovered and identified and cultured on horse blood agar to observe pattern of haemolysis. Results: In the first case, of a 75 year-old man with leukaemia and a retropharyngeal abscess, the isolate was consistently non-haemolytic, probably due to a failure to produce streptolysin S as has been described before in a small number of reports involving various M-types. In the second case, of an 84 year-old woman with dermatitis and septicaemia, the organism was principally beta-haemolytic but with no haemolysis on aerobic culture where the colonies were well spaced, a phenomenon thought to be associated with abundant production of serum opacity factor (OF). Interpretation & conclusion: These cases are a reminder that misleading cultural appearances can occur with S. pyogenes and that OF positive strains can produce poor haemolysis on aerobic culture, or fail even to do so at all. Key words M-type 58 - non-haemolytic variants - Streptococcus pyogenes

The occurrence of haemolytic colonies in cultures on colonial appearances have led to its recognition, or where blood agar of clinical specimens from non-sterile sites it has been isolated in pure culture from otherwise sterile provides a common starting point for the laboratory sites1,2,4-6. To date there has been little systematic work diagnosis of Streptococcus pyogenes. The presence done to establish how prevalent non-haemolytic strains of non-haemolytic variants would most likely pass of S. pyogenes might be in their human host. A three- unrecognised in routine laboratory practice, leading to a month study involving 361 throat swabs from patients failure of diagnosis. There are occasional earlier with pharyngitis in the Dunedin district of New Zealand reports1-5 of non-haemolytic variants of this species in 19997 yielded 64 isolates of S. pyogenes of which isolated from patients with clinical infection. These have two (3%) were non-haemolytic. Non-haemolytic variants been found where the organism has been sought with of Lancefield group B, C, G streptococci and S. milleri particular diligence (as in outbreaks of rheumatic fever3 were also sometimes encountered in clinical specimens7 and sepsis on a plastic surgery unit1), where unusual and thought to occur more frequently in organisms from

17 18 INDIAN J MED RES (SUPPL) MAY 2004 animal sources8. We report the details of two patients changed to intravenous penicillin and metronidazole for with bacteraemia featuring M-type 58 S. pyogenes, a three days followed by oral coamoxiclav for one week. seemingly rare cause of human infection in the UK. On review one month later the patient had recovered with no further symptoms from the retropharyngeal Material & Methods abscess. Convalescent serology at two months after the illness showed the antistreptolysin O (ASO) titre Initial isolates from clinical specimens were recovered grossly raised at 640 units/ml and anti-DNAase B raised from commercially available or in-house prepared culture at 960 units/ml. The streptococcal isolate was confirmed media and diagnosed by standard laboratory techniques, by the Streptococcus Reference Unit, Colindale as and saved without delay at - 80oC in vials with beads S. pyogenes, Lancefield group A, T-type 8/25/Imp19, (Protect: Technical Service Consultants, Heywood, UK). M-type 58, Opacity Factor (OF) positive, anti-OF 58. Identity was confirmed and T/M serotyping performed at the Streptococcus Reference Unit, Laboratory of The second patient was a frail woman (84 yr old) Respiratory and Systemic Infection, Central Public admitted in spring 1997 with a one day history of Health Laboratory, Colindale. The streptococcal isolates worsening oedema and dyspnoea from left ventricular were studied fresh from the - 80oC saves using cultures failure, and a fever which rose to 38.5oC. She had long- on 5 per cent defibrinated horse blood in Columbia Blood standing and widespread exfoliative eczema with multiple Agar base (Bioconnections, Bardsey, Leeds code BC recent excoriations and a discharging, inflamed urethral 2001) in petri dishes incubated aerobically and meatus associated with an indwelling urethral catheter. anaerobically for up to 72 h to observe the patterns of During the previous eight months she had suffered a haemolysis around the colonies. A recent isolate of M- myocardial infarction and a stroke; on admission she was type 1 S. pyogenes, OF negative from a patient with taking aspirin amongst other medications. bacteraemia was grown in parallel for comparison. Investigations showed haemoglobin 9.5 g/dl, Study subjects: Clinical specimens were collected from circulating white cell count 18.7 x 109/l (neutrophils 13.6 two patients. The first patient was a 75 yr old man x 109/l, lymphocytes 1.4 x 109/l) and creatine admitted with a 12 h history of sore throat, dysphagia phosphokinase 739 IU/l (normal range 24-195 IU/l). and fever. His medical history included the development Blood culture and a peri-urethral swab yielded of hairy cell leukaemia 10 months before this admission, S. pyogenes which was confirmed by the Streptococcus and his medication included aspirin. On examination he Reference Unit, Colindale and shown to be of T-type had a fever of 38oC, pharyngeal erythema, swelling on 8/25/Imp 19, M-type 58, OF positive. She was treated the left side of the neck and enlarged cervical lymph initially with flucloxacillin, and intravenous penicillin G nodes. Flexible laryngoscopy revealed a paralysed left was added when the culture results were known. For a vocal cord. CT scan of the neck showed a thickening brief period she became hypotensive (blood pressure of the retropharyngeal tissue over a distance of 6 cm 68/31 mm Hg) but fever resolved, her dermatitis craniocaudally anterior to the 3rd and 4th cervical improved and she made a good recovery. The systemic vertebrae indicating a retropharyngeal abscess. infection was thought to have arisen from infected Investigations showed a haemoglobin of 8.7 g/dl, dermatitis, either from the recent excorations or from circulating white cell count of 0.8 x 109/l (neutrophils 0.4 the inflamed area around the urinary catheter. x 109/l, lymphocytes 0.2 x 109/l) and platelets of 52 x 109/l. Blood cultures were taken and he was commenced Results on broad spectrum antibiotics. Over the next 48 h he became apyrexial, his throat became less sore and the In aerobic culture, colonies from the first patient neck swelling decreased. Blood cultures yielded a non- were umbonate with an entire edge and colony diameters haemolytic catalase-negative, gram positive coccus up to 1mm by 48 h; from the second patient they were which gave a positive latex agglutination test for at first convex, becoming umbonate and of a similar size; streptococcus group A, and biochemical profile of the M-type 1 isolate was mucoid with some collapse to S. pyogenes. The patient’s antibiotic treatment was produce a heaped edge and colony size up to 2.8 mm WEIGHTMAN & BARNHAM : NON-HAEMOLYTIC VARIANT S. PYOGENES 19 diameter by 48 h incubation. In the first patient there their prevalence in clinical specimens. Streptolysin S is was a very narrow zone of greening around the colonies the product responsible for haemolysis around colonies but no haemolysis in either aerobic or anaerobic cultures; in aerobic culture and occasional non-producing strains failure to lyse red blood cells around the colonies in any have been reported. To date these have included isolates part of the culture plate was confirmed by low power of S. pyogenes of T-type 11, M-type 47, T-type 111, T- light microscopy. In the second patient the aerobic type 12 in a ward-based outbreak involving infants and cultures showed haemolysis over most of the blood agar nurses5, the same serotype causing acute post- plate by 48 h incubation, with zones of haemolysis up to streptococcal glomerulonephritis (PSGN) in patients in 2.8 mm diameter around the colonies located more closely Japan5, mucoid M-type 18 in the throats of a series of together. However, zones of haemolysis were smaller military personnel in Colorado during an outbreak of where the colonies were more widely separated and, in rheumatic fever3,4, T-type 251, M-type 49 from skin the most thinly spread parts of the plate, the colonies infection5, T-type 12 M-type 66 from an outbreak of showed small haloes of greening but no haemolysis sepsis in a plastic surgery unit1,5 and mucoid M-type 68 (confirmed by microscopy), resembling those of the first from an infection of the eye5. Non-haemolytic isolates patient. Colonial morphology was no different whether such as these are variable in their production of a the colonies were haemolytic or non-haemolytic. greening effect (alpha haemolysis) beneath colonies on Separate subcultures from haemolytic and from non- aerobic incubation2,4. The organism from the first patient haemolytic colonies of the second patient produced the in this study was most probably streptolysin S-deficient, same mixed picture with haemolysis predominant but although this was not studied at molecular level. Such some isolated non-haemolytic colonies by 48 h incubation. strains usually produce streptolysin O but, nonetheless, In contrast to the results with the first patient, all colonies may fail to produce haemolysis on anaerobic incubation2,5. from this patient showed haemolysis on anaerobic This effect has not been fully explained but might relate incubation. The M-type 1 cultures showed clear zones to the reported inactivity of streptolysin O in strictly of haemolysis around all colonies in both aerobic and anaerobic conditions2. In the Dunedin pharyngitis study7 anaerobic conditions, with diameters up to 5.8 mm by the use of a special streptolysin O-enhancing medium 48 h incubation in air. revealed group A, C and G streptococci, non-haemolytic on conventional laboratory media, in cultures from 2 per Discussion cent of the patients. Mutant strains with an inability to produce streptolysin O are rarely encountered4,5. The A review of previous studies involving streptococcal first patient in present study showed a high ASO titre in typing in these districts showed that there were no convalescence, suggesting that the infecting organism patients other than these with detected bacteraemia produced streptolysin O. featuring M-type 58 S. pyogenes during 22 yr of observation9. We detected two persons with The genetics behind the production of streptolysins asymptomatic pharyngeal colonisation with haemolytic O and S is under study5,12-14, and the role of these M-type 58 S. pyogenes in a screening programme of products as virulence factors in streptococcal infection, 1640 young adult applicants for poultry processing work as judged by in vitro laboratory and animal model in North Yorkshire during 1980-88 (i.e., two of 20 isolates research, appears to be substantial13,15,16. However, the saved for serotyping from 25 positive applicants)10. clinical significance of infection in man with organisms Haemolytic M-type 58 S. pyogenes was not found in a that fail to produce these substances is not known. The study of 263 typed isolates of this species from the limited data published to date suggest that various strains wounds of infected meat handlers in North Yorkshire, with this failure have been able, nonetheless, to produce 1978-8610, nor in a separate collection of 64 typed isolates local and systemic infection, PSGN and acute rheumatic from patients with sore throat studied here in 1981-8211. fever.

Non-haemolytic variants of S. pyogenes are thought The serum opacity factor (OF) of S. pyogenes is to occur rarely but they are hard to recognise in mixed closely related to M protein and occurs as a diffusable cultures and there have been few studies to determine substance17, described in earlier years as having 20 INDIAN J MED RES (SUPPL) MAY 2004 lipoproteinase activity but now thought to have the ability taken by our two patients might have predisposed to to produce agglomeration with lipoproteins8. Members serious, invasive forms of infection, as has been of some 40 M-types of S. pyogenes are producers of suggested earlier28. Whether M-type 58 streptococci OF, including the more frequent M-types 4, 49, 73, 81 have a special propensity to produce non-haemolytic and R28 as well as M-type 58 and others; these are forms, and how prevalent such strains might be, remain described as OF+ strains and many are particularly unknown. The detection of non-haemolytic M type 58 associated with streptococcal infections of the skin. streptococci resulted from the serendipity of isolating 18 Pinney and colleagues in 1977 showed that OF+ strains the organism from blood, when further tests are routinely tend to produce narrow, fuzzy zones of haemolysis in applied to identify organisms. The recognition of such comparison with the larger, clear-cut zones of OF strains in mixed cultures from body sites with a normal negative strains. Under some cultural conditions isolates flora would be unlikely, unless special techniques were can produce greening but no haemolysis, as noted by applied7. It is important that microbiologists recognise other workers19. The effect is thought to be due to disturbance of streptolysin S by OF. Streptolysin S this shortcoming in their reliance on haemolysis for the consists of a small, active polypeptide moiety induced detection of S. pyogenes in clinical specimens. These and carried by a range of larger molecules, such as RNA, cases also act as a reminder that OF+ strains of polyribonucleotides, alpha-lipoprotein and albumin15, 20. S. pyogenes may produce poor haemolysis on aerobic The isolate from our second patient which was of the culture, or fail to do so at all. With the general resurgence same serotype as that in the first patient, is an example of serious forms of streptococcal infection and their of an OF+ streptoccocus; in this case the isolate showed sequelae in many parts of the western world in recent non-haemolytic colonies only where they were well- years, the possibility of unusual cultural appearances in separated on the aerobic culture plate. Colony separation diagnostic tests, such as we describe here, should be provides less competition for nutrients from the underlying borne in mind. agar and presumably allows the vigorously growing organisms to produce high levels of diffusable OF. In References anaerobic culture such strains usually produce good zones of haemolysis. 1. Colebrook L, Elliott SD, Maxted WR, Morley CW, Mortell M. Infection by non-haemolytic group A streptococci. Lancet 1942; The index strain of M-type 58 S. pyogenes came ii : 30-1. from Trinidad, an island where streptococcal infection 2. Rammelkamp CH, Dingle JH. Pathogenic streptococci. has been intensively studied21. It was isolated in 1967 Annu Rev Microbiol 1948; 2 : 279-304. from a patient with pyoderma of the foot and PSGN 3. James L, McFarland RB. An epidemic of pharyngitis due to a (A. Efstratiou, personal communication). In subsequent nonhemolytic group A streptococcus at Lowry Air Force Base. times this serotype has been encountered only New Engl J Med 1971; 284 : 750-2. infrequently and, in a summary of isolates referred to 4. Chapman SS. Unusual group A streptococci: colonial appearance the Streptococcus Reference Unit, Colindale, UK over and hemolysis. In: Wannamaker LW, Matsen JM, editors. 11 yr, it accounted for less than 1 per cent of strains22, 23. Streptococci and streptococcal diseases: recognition, understanding and management. New York: Academic Press; Of the 139 cultures of this type received, 80 were from 1972 p. 211-33. the throat, 37 from skin and wounds, 12 from the vagina and only two isolated from the blood; single isolates were 5. Skjold SA, Maxted WR, Wannamaker LW. Transduction of the genetic determinant for streptolysin S in group A streptococci. from patients with suspected PSGN and rheumatic fever. Infect Immun 1982; 38 : 183-8. M-type 58 streptococci have a wide geographical 6. Taylor MB, Barkham T. Fatal case of pneumonia caused by a distribution with further reports including Czechoslovakia, nonhaemolytic strain of Streptococcus pyogenes. India and the USA24-27. Our local isolates were widely J Clin Microbiol 2002; 40 : 2311-2. separated in time and place. M-type 58 has been 7. Dierksen KP, Tagg JR. Haemolysin-deficient variants of generally regarded as a ‘skin-infecting type’ of Streptococcus pyogenes and S. dysgalactiae subsp. equisimilis S. pyogenes; bacteraemia with M-type 58 streptococci may be overlooked as aetiological agents of pharyngitis. appears to be rare and it is possible that the NSAIDs J Med Microbiol 2000; 49 : 811-6. WEIGHTMAN & BARNHAM : NON-HAEMOLYTIC VARIANT S. PYOGENES 21

8. Colman G. Streptococcus and Lactobacillus. In: Parker MT, 19. Martin DR. Laboratory evaluation of streptococci. In: Stevens Duerden BI, editors. Topley and Wilson’s principles of DL, Kaplan EL, editors. Streptococcal infections: clinical aspects, bacteriology, virology and immunity, vol.2, 8th ed. London : microbiology and molecular pathogenesis. New York : Oxford Edward Arnold; 1990 p. 119-59. University Press; 2000 p. 266-79. 9. Barnham M, Weightman N. Streptococcus pyogenes bacteraemia: 20. Berheimer AW. Hemolysins of streptococci: characterization a 16-year case sequence in North Yorkshire. Clin Microbiol and effects on biological membranes. In: Wannamaker LW, Infect 1997; 3 (Suppl 2) : 131. Matsen JM, editors. Streptococci and streptococcal diseases: recognition, understanding and management. New York : 10. Barnham M. A study of streptococcal skin sepsis in meat and Academic Press; 1972 p. 19-31. poultry handlers. MD Thesis. London, UK : University of London; 1989. 21. Potter EV, Poon-King T, Svartman M, Burt EG, Sharrett AR, Bray JP, et al. Unusual streptococcal strains in Trinidad and 11. Barnham M. Bacteraemia in streptococcal infections of the throat. their relation to nephritis and rheumatic fever. In: Haverkorn J Infect 1983; 7 : 203-9. MJ, editor. Streptococcal disease and the community. Amsterdam 12. Cleary PP, Pritchard KH, McShan WM, Ferretti JJ. Group A : Excerpta Medica; 1974 p. 285-300. streptococcal genetics and virulence. In: Stevens DL, Kaplan 22. Gaworzewska E, Colman G. Changes in the pattern of infection EL, editors. Streptococcal infections: clinical aspects, caused by Streptococcus pyogenes. Epidemiol Infect 1988; 100 : microbiology and molecular pathogenesis. New York : Oxford 257-69. University Press; 2000 p. 37-56. 23. Colman G, Tanna A, Efstratiou A, Gaworzewska ET. The 13. Betschel SD, Borgia SM, Barg NL, Low DE, DeAzavedo JC. serotypes of Streptococcus pyogenes present in Britain during Reduced virulence of group A streptococcal Tn916 mutants 1980-1990 and their association with disease. J Med Microbiol that do not produce streptolysin S. Infect Immun 1998; 66 : 1993; 39 : 165-78. 1671-9. 24. Bessen DE, Izzo MW, Fiorentino TR, Caringal RM, 14. Nizet V, Beall B, Bast DJ, Datta V, Kilburn L, Low DE, et al. Hollingshead SK, Beall B. Genetic linkage of exotoxin alleles Genetic locus for streptolysin S production by group A and emm gene markers for tissue tropism in group A streptococci. streptococcus. Infect Immun 2000; 68 : 4245-54. J Infect Dis 1999; 179 : 627-36. 15. Ginsburg I. Is streptolysin S of group A streptococci a virulence 25. Prakash K, Dutta S. Antibodies to streptococccal opacity factor in factor? APMIS 1999; 107 : 1051-9. a selected Indian population. J Med Microbiol 1991; 34 : 119-24. 16. Humar D, Datta V, Bast DJ, Beall B, De Azavedo JCS, 26. Facklam R, Beall B, Efstratiou A, Fischetti V, Johnson D, Nizet V, et al. Streptolysin S and necrotising infections Kaplan E, et al. emm typing and validation of provisional M produced by group G streptococcus. Lancet 2002; 359 : types for group A streptococci. Emerg Infect Dis 1999; 5 : 124-9. 247-53. 17. Top FH, Wannamaker LW. The serum opacity reaction of 27. Beall B, Facklam R, Hoenes T, Schwartz B. Survey of emm gene Streptococcus pyogenes: frequency of production of sequences and T-antigen types from systemic Streptococcus streptococcal lipoproteinase by strains of different serological pyogenes infection isolates collected in San Francisco, California; types and the relationship to M protein production. J Hyg Atlanta, Georgia; and Connecticut in 1994 and 1995. J Clin Camb 1968; 66 : 49-58. Microbiol 1997; 35 : 1231-5. 18. Pinney AM, Widdowson JP, Maxted WR. Inhibition of beta- 28. Barnham M. Nonsteroidal antiinflammatory drugs: concurrent haemolysis by opacity factor in group A streptococci. J Hyg or causative drugs in serious infection? Clin Infect Dis 1997; 25 Camb 1977; 78 : 355-62. : 1272-3.

Reprint requests : Dr N.C. Weightman, Consultant Microbiologist, Department of Microbiology, Friarage Hospital, Northallerton North Yorkshire, UK e-mail : [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 22-25

Enterococcal infections with special reference to phenotypic characterization & drug resistance

M.G. Karmarkar, Edwin S. Gershom & P. R. Mehta

Department of Microbiology, Seth G.S. Medical College & KEM Hospital, Mumbai, India

Received August 6, 2003

Background & objectives: Enterococci, classified as group D streptococci, are the second leading cause of nosocomial infections. The incidence of enterococcal infections and species prevalent in India is not thoroughly investigated. The present study was undertaken to isolate and characterize enterococci from clinical specimens and determine the antimicrobial susceptibility pattern of these isolates. Methods: Clinical specimens (blood, urine and swabs) were cultured on bile esculin azide agar (BEAA) for isolation of enterococci. The phenotype based scheme included Gram staining of growth on BEAA and subculturing of cocci on sheep blood agar plates for vancomycin disk diffusion and hydrolysis of L-pyrrolidonyl-b-napthylamide (PYR) testing. The phenotypic method was used to surveillance cultures that yielded growth on BEAA. Enterococcal strains were further identified to the species level by conventional biochemical tests. PYR positive isolates were further characterized into vancomycin resistant enterococci (VRE) and nonVRE depending upon vancomycin inhibition zone size. The isolates were characterized into vanA, vanB and vanC depending upon minimum inhibitory concentration (MIC) values. Conventional method was used to study the antibiogram of isolates. Results: A total of 52 isolates of enterococci (10 Enterococcus faecalis, 42 E. faecium) were isolated from 534 clinical specimens. Of the 52 isolates, 12 isolates were resistant to vancomycin with an MIC > 4 µg/ml but sensitive to teicoplanin (vanB isolates). Interpretation & conclusion: Our study reveals the problem of multiple drug resistant enterococci and emergence of VRE. Better susceptibility tests need to be used to measure the vancomycin resistance accurately. Key words Disk diffusion - enterococci - MIC - phenotype - vancomycin resistant enterococci

Enterococci are classified as group D streptococci. implicated in urinary, gastrointestinal tract and pelvic Enterococci are facultative anaerobes that are part of infections1,2,9. the normal intestinal flora in humans1,2. Though they are not considered to be highly virulent, their intrinsic Enterococci have an acquired resistance to several resistance and ability to acquire resistance to several classes of antibiotics either by mutation or by receipt of broad-spectrum antibiotics allows them to cause super foreign genetic material through the transfer of plasmids infections in patients already receiving antimicrobial and transposons2. The acquisition of high-level therapy1,3-5. The species most commonly implicated in aminoglycoside resistance and vancomycin resistance human infections is E. faecalis, the increasing has limited the therapeutic options available for clinicians. occurrence of E. faecium is of particular concern due The transfer potential of vancomycin resistant genes to high resistance to antibiotics especially in nosocomial from enterococci to Staphylococcus aureus which has settings6-8. Enterococcus species are most commonly been achieved in vitro but not yet reported in clinical 22 KARMARKAR et al : ENTEROCOCCAL PHENOTYPIC CHARACTERIZATION & DRUG RESISTANCE 23

settings, increases the importance of finding ways to limit Table I. Source and speciation of isolates the spread of vancomycin resistant enterococci (VRE)10. Source (n=534) E. faecalis E. faecium Total Enterococci are the second leading cause of Urine (253) 6 20 26 7 nosocomial infections , VRE have been reported from Blood (184) 4 14 18 worldwide3,6-8,11,12. The incidence of enterococcal infections and species prevalent in India is not thoroughly Others (97) - 8 8 investigated. Few studies from India reported E. Table II. Antibiotic resistance (%) in enterococci by Kirby - Bauer 1,9 faecalis as the most prevalent species , with high-level disk diffusion method16 resistance to aminoglycosides, but no resistance to vancomycin13,14. Antibiotic E. faecalis (n=10) E. faecium (n=42) Ampicillin 60 30 The present study was undertaken with the objective to study isolation, speciation and characterization of Penicillin G 40 71.43 enterococci from clinical specimens, to determine the Gentamicin 100 85.71 antimicrobial susceptibility pattern of the isolates, and to Vancomycin 10 28.57 determine the phenotypes of VRE. Teicoplanin 0 9.52 Nitrofurantoin 80 85.71 Material & Methods Ciprofloxacin 100 85.71 The present study was carried out at a tertiary care Table III. Antibiotic resistance (%) in enterococci by breakpoint hospital in Mumbai. A total of 534 clinical specimens method17 (urine, blood, pus and wound swabs) were included. Bile esculin azide agar (BEAA) was used for isolation of Antibiotic E. faecalis (n=10) E. faecium (n=42) enterococci from clinical specimens.The genus enterococcus was confirmed by Gram stain, esculin Ampicillin 60 30 hydrolysis, hydrolysis of L-pyrrolidonyl-b-napthylamide Penicillin G 40 71.43 (PYR) test and salt tolerance1,2,6,8. Gentamicin 100 85.71 Vancomycin 10 28.57 Speciation was based on Facklam’s conventional Teicoplanin 0 9.52 method15 : tolerance to tellurite for E. faecalis, fermentation of arabinose, mannitol, raffinose and sorbitol Nitrofurantoin 80 85.71 for E. faecium1,2,7,8. Antimicrobial susceptibility to Ciprofloxacin 100 85.71 penicillin, gentamicin, vancomycin, teicoplanin, ciprofloxacin and nitrofurantoin was determined by Kirby than E. faecalis but resistance to ampicillin was more Bauer disk diffusion16 and breakpoint method as per among E. faecalis than E. faecium. High level NCCLS standards17. The phenotypic characterization resistance to gentamicin was 100 per cent for both of VRE was based on the scheme outlined by Cetinkaya the species (Tables II, III). et al3. In this study we determined antimicrobial susceptibility Results & Discussion by disk diffusion and breakpoint method. Minimum inhibitory concentration by breakpoint method was Of the 534 samples tested, a total of 52 considered as the reference method. The correlation enterococci isolates (10 E. faecalis, 42 E. faecium) between the two methods was poor. Errors were were obtained (Table I). Species identification of determined by comparing the two methods (Table IV). isolates enabled to assess the species-specific The maximum errors were for ampicillin and vancomycin. antibiotic susceptibility patterns. In this study The utility of the disk diffusion method because of the resistance to penicillin was more among E. faecium errors associated with the susceptibility testing might 24 INDIAN J MED RES (SUPPL) MAY 2004

Table IV. Correlation between disk diffusion and breakpoint method18

Antibiotic Method used Kirby Bauer Breakpoint Error* RI S R I SVMMAMI Ampicillin 36 - 16 44 - 8 12 4 - Penicillin 34 - 18 34 - 18 - - - Gentamicin 42 - 10 40 - 12 - 4 - Vancomycin 14 22 16 12 - 40 4 8 26 Teicoplanin 4 - 48 - - 52 - 4 - Nitrofurantoin 44 - 8 44 - 8 - - - Ciprofloxacin 46 2 4 52 - - 2 - 2

* Error classification Error Kirby-Bauer Breakpoint Very major (VM) S R Major (MA) R S Minor(MI) I R/S R/S I

S, sensitive; R, resistance; I, intermediate result in unwarranted utilization or elimination of the In India the incidence of enterococcal infections antibiotics as part of the possible treatment regimens. is not thoroughly investigated. Enterococcus faecalis Emergence of vancomycin resistant enterococci has been is the most prevalent species cultured from humans attributed to imprudent use of cephalosporins, accounting for 80-90 per cent of clinical isolates in vancomycin, colonization pressure and noncompliance other studies5,7. We found E. faecium as the most with infection control measures. prevalent species in congruence with earlier studies on hospitalized patients6-8. This finding is of clinical All the 52 isolates in our study were sensitive to importance since E. faecium is more resistant to teicoplanin with an MIC < 0.5 µg/ml. Of the 52 isolates, penicillin than E. faecalis, thus limiting the therapeutic 12 were resistant to vancomycin with an MIC > 4 µg/ml. options. These 12 isolates belonged to VAN B phenotype i.e., variable resistance to vancomycin but susceptible to Our study signals the emergence of vancomycin teicoplanin. There are no reports of vancomycin resistant resistant enterococci. Thus, a more detailed study is enterococci in India. necessary using phenotypic and genotypic methods to have a better picture of enterococcal infections. The The results of our study are based on phenotypic present study reveals the problem of multidrug resistant methods alone. It was felt earlier that the use of both phenotypic and genotypic in conjunction with enterococci. To have an effective control of multidrug each other would provide a more accurate resistant enterococci would require better understanding information6-8. Susceptibility tests for enterococci do of the interaction between enterococci, hospital not correlate with each other4. Therefore environment and humans, prudent use of antibiotic, better susceptibility tests that detect vancomycin resistance isolation procedures in hospitals and other patient care accurately must be used or the prevalence of VRE environments and improved and rapid surveillance may be underestimated. measures. KARMARKAR et al : ENTEROCOCCAL PHENOTYPIC CHARACTERIZATION & DRUG RESISTANCE 25

References 11. Cereda RF, Gales AC, Silbert S, Joner RN, Sader HS. Molecular typing and antimicrobial susceptibility of 1. Desai PJ, Pandit D, Mathur M, Gogate A. Prevalence, vancomycin resistant Enterococcus faecium in Brazil. identification and distribution of various species of enterococci Infect Control Hosp Epidemiol 2002; 23 : 19-28. isolated from clinical specimens with special reference to urinary 12. Hsueh P, Teng LJ, Wang HL, Chang C, Chen YC. Emergence of tract infection in catheterized patients. Indian J Med Microbiol vancomycin resistant enterococci at a university hospital in 2001; 19 : 132-7. Taiwan: Persistence of multiple clones. Infect Control Hosp 2. Murray BE. The life and times of the enterococci. Clin Microbiol Epidemiol 1999; 20 : 828-33. Rev 1990; 3 : 46-65. 13. Agarwal VA, Jain YI, Pathak A. Concomitant high level 3. Cetinkaya Y, Falk P, Mayhall G. Vancomycin resistant resistance to penicillin and aminoglycosides in enterococci enterococci. Clin Microbiol Rev 2000; 13 : 686-707. at Nagpur, Central India. Indian J Med Microbiol 1999; 17 : 4. Jones RN, Sader HS, Erwin ME, Anderson SC and Enterococcus 85-7. Study group. Emerging multiple resistant enterococci. Diagn Microbiol Infect Dis 1995; 21 : 85-93. 14. Mathai E, Cherian BP, Chandy S. Determination of high level resistance to aminoglycosides among enterococci. Indian J Med 5. Jones RN, Sader HS, Erwin ME, Anderson SC and Enterococcus Res 1995; 102 : 255-7. Study group. Emerging multiple resistant enterococci. Diagn Microbiol Infect Dis 1995; 21 : 95-100. 15. Facklam RR, Collins MD. Identification of enterococcus species 6. Liassine N, Frei R, Jan I, Auckenthaler R. Characterization isolated from human infections by a conventional test scheme. of glycopeptide resistant enterococci from a Swiss hospital. J Clin Microbiol 1989; 27 : 731-4. J Clin Microbiol 1998; 36 : 1853-8. 16. Bauer AW, Kirby WM, Truck M. Antibiotic susceptibility 7. Nelson RRS, McGregor KF, Brown AR, Amyes GB, Young testing by a standardized single disc method. Am J Clin Path HK. Isolation and characterization of glycopeptide resistant 1966; 45 : 493. enterococcci from hospitalized patients over a 30-month period. J Clin Microbiol 2000; 38 : 2112-6. 17. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial susceptibility testing 8. Sahm DF, Fre L, Smith C, Eveland M, Mundy LM. Rapid M100-S9, National Committee for Clinical Laboratory characterization schemes for surveillance of isolates of Standards, Wayne, Pa. vancomycin resistant enterococci. J Clin Microbiol 1997; 35 : 2026-30. 18. Cotter G, Adley CC. Comparison and evaluation of antimicrobial susceptibility testing of enterococci 9. Mathai E, Margaret A, George V, Brahmadathan KN. Identification of Gram positive cocci from urine. Indian J Med performed in accordance with six national committee Res 1994; 100 : 10-4. standardized disk diffusion procedures. J Clin Microbiol 2001: 39 : 3753-6. 10. Noble WC, Virani Z, Cree RGA. Co-transfer of vancomycin and other resistance genes from Enterococcus faecalis 19. Woodford N, Johnson AP, Morrison D, Speller DCE. Current NCTC12201 to Staphylococcus aureus. FEMS Microbiol Lett perspectives on glycopeptide resistance. Clin Microbiol Rev 1992; 93 : 195-8. 1995; 8 : 585-615.

Reprint requests : Dr M.G. Karmarkar, Associate Professor, Department of Microbiology, Seth G.S. Medical College & KEM Hospital Acharya Donde Marg, Parel, Mumbai 400012, India e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 26-28

Evaluation & revaluation of upper limits of normal values of anti-streptolysin O & anti-deoxyribonuclease B in Mumbai

M.G. Karmarkar, Vineetha Venugopal, Leela Joshi & Riecha Kamboj

Department of Microbiology, Seth G.S. Medical College & KEM Hospital, Mumbai, India

Received August 6, 2003

Background & objectives: For the serodiagnosis of group A streptococcal (GAS) infections and their late sequelae, a battery of tests have been suggested. However, anti-streptolysin O (ASO) and anti- deoxyribonuclease B (ADNase B) are the most widely accepted tests for the same. It is essential to evaluate the upper limits of ASO and ADNase B levels in age and sex matched normal population before using them for the detection of patients. For this study, these values were determined in the past and were revaluated again in 2001-2002, in normal subjects. Methods: A total of 200 healthy individuals were included in the study in 1991-1992 and same number of age and sex matched healthy individuals were included in 2001-2002. The methodology used for determination of ASO and ADNase B levels were as per the techniques recommended by the WHO. Results: The findings show that the upper limits of normal ASO titers are 195 IU and 305 IU for adults and children respectively while the said levels for ADNase B for both adults and children were found to be 200 IU. Interpretation & conclusion: The findings of the present study will be helpful in the follow-up, better diagnosis and prognosis of group A streptococcal infections.

Key words ADNase B - ASO - GAS - ULN Group A streptococcal (GAS) infections and their rheumatic fever (ARF) or post-streptococcal late sequelae like rheumatic fever (RF) and rheumatic glomerulonephritis (PSGN), it is not always possible to heart disease (RHD) remain an important and major obtain an adequate clinical history or to recover the health problem in India. The incidence of RF varies from organism. In such cases, the presence of a host immune 0.2 - 0.75 per 1000 children per year in school children response is the only evidence of the recent infection that 5 - 15 yr of age. The prevalence of RHD varies from remains. Measurement of antibodies to specific 1.0-5.4 per 1000 in school children. Also among the streptococcal antigens is therefore necessary to confirm population 5 - 40 yr of age at risk for RHD, 1.4 million the diagnosis of the preceding GAS infection. This along are presently suffering from RHD 1. Such data have with the modified Jones criteria2 contributes to the final emphasized the importance of accurate clinical diagnosis, confirmation of clinical diagnosis. often requiring laboratory confirmation of the preceding GAS infection. The diagnosis of GAS infections is often Because RF/PSGN are non-suppurative sequelae of made by clinical observation and then confirmed by rapid GAS infection and there is a latent period between the antigen tests or by isolation of GAS from site of infection. streptococcal infection and onset of disease, serum taken However, when one is considering the diagnosis of at disease onset may really be convalescent; a rising post-streptococcal non-suppurative sequelae like acute titre may not be demonstrated. Hence, an upper limit of 26 KARMARKAR et al: ANTI-STREPTOLYSIN O & ANTI-DEOXYRIBONUCLEASE IN MUMBAI 27 normal or ULN value is useful when acute and Determination of ADNase B: Neutralization assay convalescent sera are unavailable3. For the purpose of method5 was used to determine the ADNase B levels. uniformity, ULN is defined as that titer exceeded by Microtiter dilutions of the serum samples were carried 20 per cent of a normal population4. ULN values are out in 2 sets (1:50 to 1:3200 in the first set and 1:75 to known to vary with respect to different geographical 1:4800 in the second set). 25 µl of deoxyribonuclease B locations as also with respect to different seasons3. (DNase B – 1U/ml) was added to all tubes and the Hence the need was felt to evaluate and revaluate the reaction was allowed to proceed for 15 min at 37°C. ULN values of ASO and ADNase B levels in Mumbai. Next, 50 µl DNA-methyl green complex (prepared using Sigma Type V, highly polymerised DNA, from calf Though diagnosis of GAS infection can be done by thymus, as per WHO Manual5) was added and the tubes a whole array of tests that are serodiagnostic, anti- were incubated at 37°C for 24 h. Results were recorded streptolysin O (ASO) and anti-deoxyribonuclease on the basis of digestion (leading to loss of green colour) B (ADNase B) are the most widely accepted tests for or no digestion (no loss of green colour) of substrate. the same. It is essential to evaluate the ULN of ASO The titre corresponds to the reciprocal value of serum dilution and ADNase B levels in age and sex-matched normal that causes an approximately 50 per cent digestion of population before using these for the detection of patients. substrate. In case of steep colour transition, the last well For this study, these values were evaluated in normal with total enzyme inhibition is the endpoint. The enzyme healthy subjects in Mumbai in the year 1990 -1991 and and substrate controls were put up along with the test. revaluated in the year 2001-2002. All methods used in ASO and ADNase B tests are Material & Methods in accordance with the techniques specified in the WHO manual on laboratory diagnosis of group A streptococcal Study subjects: A total of 200 normal healthy individuals infections (1996)2. (both children and adults) were included in the study in 1990-1991, and same number of age and sex matched Results normal, healthy individuals were enrolled in 2001-2002. These volunteers had no features suggestive of sore A total of 200 serum samples were available for the throat in the past three months, had no history of joint study, 40 of these from healthy children in the age group pains and had not taken any kind of antibiotic therapy 5 - 15 yr of which 20 were males and 20 were females, for the same. 160 samples were obtained from healthy, adults (80 males and 80 females). In both evaluation (1991-1992) and Blood samples (3-5 ml) were collected by revaluation (2001-2002), all individuals were followed up venepuncture, serum was separated and stored at -20°C thrice on quarterly basis to check for seasonal variation. till further use. The ULN of ASO in children during revaluation has Determination of ASO: Geometric dilutions of the shown a shift of one value upwards at 305 IU while the patient’s serum were carried out by tube dilution method5. value obtained during evaluation was 244 IU. The ULN The dilutions used were in the range of 1:100 to 1:744; of ASO in adults remained at 195 IU for both evaluation 0.5 ml of antigen i.e., streptolysin O (SLO – 2U/ml) and revaluation (Table I). The ULN of ADNase B for was added to each dilution and allowed to react for both children and adults remained fairly constant at 15 min at 37°C, 0.5 ml of the indicator system consisting 200 IU (Table II). No significant seasonal variation was of 5 per cent rabbit erythrocytes was added to all tubes. noted either in evaluation or revaluation. After a further incubation of 1 h at 37°C, the results were recorded on the basis of presence or absence of Discussion haemolysis. The highest dilution showing 50 per cent haemolysis is the end point and the corresponding Nearly two decades after declaring the year 1984 as reciprocal value is the ASO titre. The enzyme and buffer the year of the rheumatic child, infections due to GAS controls were put up along with the test. are still prevalent in India. Frequent exposure to the 28 INDIAN J MED RES (SUPPL) MAY 2004

Table I. Evaluation and revaluation of upper limit of normal values well as adults could be either a varied antigenic stimulus of anti-streptolysin O as compared to streptolysin O or due to limitations of the sensitivity of the test. ASO titres Evaluation Revaluation (IU) Children Adults Children Adults There is no information available in India on the ULN £100 44.66 44 35 27.5 values, a US study6 shows that the ULN values in 125 16 16 15 25 children of 2-12 yr were 240 for ASO and 640 for 156 5.33 20 22.5 13.125 ADNase B, respectively as against our study which has 195 13.33 8 5 17.5 a ULN of 305 for ASO and 200 for ADNase B in children 5-15 yr of age. The presence of a higher ULN 244 13.33 12 17.5 13.125 in case of ADNase B in the study conducted in the US6 305 5.33 0 5 3.125 may be due to a higher antigenic stimulus caused by 381 0 0 0 0.625 their GAS strains.

Values are expressed in percentages The present study opens further avenues to understand this microorganism called GAS and tells us Table II. Evaluation and revaluation of upper limits of normal values to constantly and periodically recheck on those very of anti-deoxyribonuclease B same values which we ourselves may have established. ADNase B Evaluation Revaluation Such checks may yield us results, which may serve as a titres (IU) Children Adults Children Adults foundation stone for further research on the changing antigenic stimuli of this organism. £50 31.54 42.5 62.5 65 50-75 16.66 25 12.5 4.375 References 75-100 16.66 15 7.5 100-150 8.33 7.5 5 3.125 1. Grover A, Vijayvergiya R, Thingam ST. Burden of rheumatic and congenital heart disease in India : Lowest estimate based on 150-200 16.66 7.5 10 6.25 the 2001 census. Indian Heart J 2002; 54: 104-7. 200-300 10.0 2.5 0 3.75 2. Gaasch WH. Guidelines for the diagnosis of rheumatic fever, 300-400 0 0 0 6.875 Jones criteria; 1992 update. JAMA 1992; 268 : 2069-73. 400-600 0 0 0 3.125 3. Shet A, Kaplan EL. Clinical use and interpretation of group A streptococcal antibody tests: A practical approach for the Values are expressed in percentages pediatrician or primary care physician. Pediatr Infect Dis J 2002; 21: 420-30. organism occurs due to a variety of highly predisposing factors which include overcrowding, bad hygienic 4. Wannamaker L, Ayoub E. Antibody titres in acute rheumatic fever. Circulation 1960; 21; 598 -614. practices, increased pollution leading to decreased immunity etc. The shift of ULN of ASO in children from 5. Johnson DR, Kaplan EL, Sramek J, Bicova R, Havlicek J, Havlickova H, Motlova J, Kriz P. WHO manual on 244 IU in 1991-1992 to 305 IU in the present scenario laboratory diagnosis of group A streptococcal infections. tells a similar story. The reason adults do not seem to Geneva: World Health Organization; 1996. show an increased ULN value may be due to the fact 6. Kaplan EL, Rothermel CD, Johnson DR. Antistreptolysin that GAS infections mainly affect children as they form O and anti-deoxyribonuclease B titers: Normal values for the susceptible group. The reasons for having no change children ages 2 to 12 in the United States. Pediatrics 1998; in the ULN values of ADNase B in both children as 101 : 86-8.

Reprint requests: Dr M.G. Karmarkar, Associate Professor, Department of Microbiology, Seth G.S. Medical College & KEM Hospital, Parel, Mumbai 400012, India e-mail: [email protected]

Indian J Med Res 119 (Suppl) May 2004, pp 37-43

Retraction

Identification & characterisation of the two novel streptococcal pyrogenic exotoxins SPE-L & SPE-M

Thomas Proft, Phillip D. Webb, Vanessa Handley & John D. Fraser

Department of Molecular Medicine & Pathology, Faculty of Medical & Health Sciences University of Auckland, Private Bag 92019, Auckland, New Zealand

Received August 7, 2003

Background & objectives: The streptococcal pyrogenic exotoxins (SPEs) are produced by Streptococcus pyogenes and belong to the family of bacterial superantigens, a group of highly mitogenic proteins. The aim of this study was to search unfinished streptococcal genomes for novel superantigens, to generate recombinant proteins from potential open reading frames (ORFs) and to analyse them for superantigen activity. Methods: The microbial genome database was searched using a TBLASTN search programme. Genotyping of S. equi and S. pyogenes isolates was done using the specific primer pairs. The spe-l and spe-m genes were amplified by PCR. Results: Two novel streptococcal superantigen genes (sepe-l and sepe-m) were identified from the Streptococcus equi genomic database at the Sanger Centre. Genotyping of S. pyogenes isolates resulted in the detection of the orthologous genes spe-l and spe-m in a restricted number of S. pyogenes isolates and revealed a link of spe-l to the M89 serotype. Recombinant SPE-L and rSPE-M were highly mitogenic for human peripheral blood lymphocytes with half maximum responses at 1 pg/ml and 10 pg/ml, respectively. The results from competitive binding experiments suggest that both proteins bind MHC class II at the b-chain, but not at the a-chain. The most common target for both toxins were human Vbl.1 expressing T cells. Seroconversion against SPE-L and SPE-M was observed in healthy blood donors. Interpretation & conclusion: The two novel ORFs identified in both, S. equi and S. pyogenes, code for proteins that show typical superantigen features. The seroconversion seen in some blood donors suggest that the proteins are indeed produced by the bacteria. Interestingly, the spe-l gene is highly associated with S. pyogenes M89, which is linked to acute rheumatic fever in New Zealand. Key words MHC class II binding - sero-conversion - streptococcal pyrogenic exotoxin -superantigen - T cell mitogen Bacterial superantigens (SAgs) are small highly Streptococcus pyogenes, a Lancefield group A mitogenic proteins that are produced as monomeric streptococcus (GAS), is a human pathogen that causes a precursor molecules with typical bacterial signal wide range of diseases including acute tonsillitis, scarlet peptides1. The mature secreted toxins simultaneously bind fever, necrotising fasciitis , bacteraemia, streptococcal toxic to MHC class II antigens and T cell receptor (TcR) shock syndrome and rheumatic fever7. The streptococcal molecules, which leads to the stimulation of large SAgs, which include streptococcal pyrogenic exotoxins numbers of T cells2-4. As a result, high systemic levels (SPE) A, C, G-J, streptococcal SAg (SSA), and several of the cytokines TNF-a and IL-1b and T cell mediators, allelic variants of streptococcal mitogenic exotoxin such as IL-2 and INF-g are released, which cause Z (SMEZ) are believed to play an important role as pathological conditions including fever and shock5,6. virulence factors in at least some of these diseases3,8,9. 37

38 INDIAN J MED RES (SUPPL) MAY 2004

Recently, two pyrogenic exotoxins (SePE-H and Generation of recombinant SPE-L and SPE-M: The SePE-I) were identified from S. equi10. S. equi is a spe-l and spe-m genes were amplified by PCR from Lancefield group C streptococcus (GCS) that causes 50 ng genomic DNA of isolates 10846 and FP 4223, strangles, a contagious inflammatory disease of the upper respectively, using the same primers as for genotyping. respiratory tract and associated lymph nodes of equids11. The PCR products encoding the mature proteins without The streptococcal pyrogenic exotoxins together with the the predicted leader sequences were cloned into the staphylococcal enterotoxin build a large family of BamHI-EcoRI cloning site of pET32b-3c expression bacterial superantigens. These proteins are highly vector. Recombinant SPE-L and rSPE-M were conserved in their protein structure, but possess only expressed in E. coli AD494 (DE3) (Novagene) as 12 limited primary sequence similarity . thioredoxin fusion proteins. The thioredoxin fusion proteins were purified on Ni2+ IDA sepharose (Sigma, We identified several novel streptococcal sag genes USA) and the mature proteins were cleaved off from earlier after searching the S. pyogenes M1 genome thioredoxin on the column using protease 3C13. database at Oklahoma University using a conserved 13,14 peptide sequence . In the present study this search Toxin proliferation assay: Human peripheral blood was extended to include non-group A streptococcal lymphocytes (PBLS) were purified from blood of a genomes. healthy donor by Histopaque Ficoll (Sigma,USA) fractionation. The PBLS were incubated in 96-well round

Material & Methods bottom microtiter plates at 105 cells per well with RPMI- 10 (RPMI with 10% foetal calf serum) containing

Bioinformatic tools: The microbial genome databases varying dilutions of recombinant toxins. After 3 days at the National Center for Biotechnology Information 0.1 µCi [3H]thymidine was added to each well and cells

(http://www.ncbi.nlm.nih.gov/Microb_blast/ were incubated for another 24 h. Cells were harvested unfinishedgenome.html) were searched with highly and counted on a scintillation counter. conserved b5 and a4 regions of streptococcal and staphylococcal SAgs, using a TBLASTN search TcR Vb analysis: Vb enrichment analysis was performed programme. The ORFs were defined by translating the by anchored multiprimer amplification. Human PBLs DNA sequences around the matching regions and aligning were incubated with 20 pg/ml of recombinant toxin at the protein sequences to known SAgs using ClustalW at 106 cells/ml for 3 days. A two-fold volume expansion of http://clustalw.genome.ad.jp. the culture followed with medium containing 20 ng/ml IL-2. After another 24 h, stimulated and resting cells Genotyping of S. equi and S. pyogenes isolates: The were harvested and RNA was prepared using Trizol New Zealand S. pyogenes and S. equi isolates reagent (Gibco, BRL, Gaithersburg, USA). A 500 bp (provided by Diana Martin, Institute of Environmental b-chain DNA probe was obtained by anchored Sciences and Research, Porirua and John Taylor, 16 Department of Microbiology, Otago University, Dunedin) multiprimer PCR as described previously , radiolabelled b b were grown in Brain Heart Infusion medium (BHI, Difco, and hybridised to individual V s and a C DNA region USA) at 370C and genomic DNA was prepared as dot blotted on a Nylon membrane. The membrane was described previously15. Genotyping was carried out by analysed on a Molecular Dynamics Storm Phosphor PCR with 50 ng of purified streptococcal genomic DNA imager using ImageQuant software. Individual Vbs were using the specific primer pairs sepe-l/spe-l.fw (G C G G expressed as a percentage of all the Vbs determined by A T C C G A T A C G T A C A A T A C A A A T G) hybridisation to the Cb probe. and sepe-l/spe-l.rev (G C G A A T T C A A T A G C A T T C G A C C), and sepe-m/spe-m.fw (G C G G A T Radiolabelling and LG-2 binding experiments: C C G A G G G G A C T A T T A A T A T T A A G) Recombinant toxin was radioiodinated by the chloramine and sepem/spe-m.rev (G C G A A T T C G G T T T C T method as previously described17. LG2 cells were used T T G A T A C T A A C). for cell binding experiments, as previously described17. Briefly, cells were harvested, resuspended in RPMI-10 A primer pair specific to a DNA region encoding the and mixed at 106 cells/ml with 125I-tracer toxin (1 ng) 23S rRNA was used as a positive control15. and 0.0001 to 10 µg of unlabelled toxin and incubated at

PROFT et al: NOVEL STREPTOCOCCAL SUPERANTIGENS 39

37° C for 1 h. After washing with ice cold RPMI, the that they are located on mobile DNA elements and thus pelleted cells were analysed in a gamma counter. restricted in their distribution to certain S. equi isolates (data not shown). For competitive binding studies, 1 ng of 125I-tracer toxin (rSPE-L or rSPE-M) was incubated with 0.0001 Screening of 40 S. pyogenes isolates from New to 10 µg of unlabelled toxin (rSPE-L, rSPE-M, rSPE-C, Zealand (including 29 M-types and 3 MNT) identified rSEB, and rTSST) for 1 h. the sepe-l gene in 6 of 40 isolates (15%) and the sepe- m gene in 2 of 40 isolates (5%). The sepe-l gene was Seroconversion experiments: Five µl of serum from restricted to only 5 M-types (M28, M41, M56, M59 and human healthy donors (n=20) were mixed with 10 µl of M89) and the sepe-m gene was found in only 2 M-types

PBS and 10 µl of 125I-labelled recombinant SPE-L or (M80 and M92). Interestingly, M89 strains are strongly rSPE-M and incubated for 1 h at 370C. As controls, associated with acute rheumatic fever in New Zealand18.

BSA or rabbit antiserum against SPE-C or SMEZ was New Zealand S. pyogenes M89 isolates were genotyped for the spe-l gene and 8 of 11 (73%) were positive used instead of human sera. After adding 50 µl of protein A - staphylococcal cells (Toxin Technology Florida, USA) indicating a strong association between this M-type and the samples were incubated for another 30 min at 40C the spe-l gene. before the cells were spun down, washed in 1 ml of PBS and counted in a Cobra Gamma Counter. Cloning and expression of spe-l and spe-m from S. pyogenes: The sepe-l gene was cloned from isolate Results 10846 (M89) and DNA sequence analysis revealed a strong sequence homology (99%) to the S. equi sepe-l. Identification of novel streptococcal superantigens: The gene was named spe-l according to the S. pyogenes The streptococcal genomes of S. gordonii, nomenclature for the streptococcal pyrogenic exotoxin S. pneumoniae, S. equi and S. mutants at the National family. Furthermore, the spe-l gene is 100 per cent Center for Biotechnology Information (NCBI) webpage identical to the hypothetical exotoxin gene spe-l found (http://www.ncbi.nlm.nih.gov/Microb_blast/ on prophage PhiHIH1.1 sequence (NC003157). The unfinishedgenome.html) were searched with a sepe-m was cloned from isolate FP4223 (serotype M80) superantigen peptide sequence corresponding to the and DNA sequence comparison showed strong highly conserved a4-region using the TBLASTN nucleotide homology to sepe-m from S. equi (98.1%) programme. Two ORFs that code for potential novel and was therefore labelled spe-m. The minor differences superantigens were found. One ORF showed strong between SePE-L and SPE-L and between SePE-M and sequence homology to the hypothetical exotoxin gene SPE-M suggest that the proteins are orthologous with spe-l found on the prophage PhiHIH1.1 sequence identical or very similar functions in the two streptococcal (NC003157) and was labelled sepe-l (S. equi pyrogenic species. exotoxin l). The other ORF showed no significant sequence homology to any known superantigen, but A revised streptococcal superantigen family tree, contained the highly conserved superantigen a4-peptide based on primary amino acid sequence homology is sequence and was named sepe-m. The sepe-l and shown in Fig. 1. All novel SAgs contain a zinc-binding sepe-m genes are adjacent at the 5’ end of a 4,470 bp motif near the C-terminus that is common among most contig separated by a 266 bp intergenic region. of the streptococcal SAgs, including all members of group A. Zinc is essential for binding to the polymorphic MHC Genotyping of S. equi and S. pyogenes isolates: class II b-chain via a divalent zinc ion19. Genomic DNA from 8 S. equi isolates were screened by PCR using specific primer pairs for the sepe-l and Recombinant proteins of SPE-L and SPE-M were sepe-m genes. In addition, a primer pair specific to a generated in E. coli AD494 (DE3) using pET32b-3c streptococcal DNA region encoding the 23S rRNA was expression vectors. To ensure the native conformation used as a positive control. Neither sepe-l nor sepe-m of the purified recombinant toxins, a standard could be detected in any of the S. equi isolates suggesting [3H]thymidine incorporation assay was performed to test

40 INDIAN J MED RES (SUPPL) MAY 2004

unlabelled toxins [ µg]

Fig.1. Family tree of streptococcal and staphylococcal superantigens. The tree was created using ClustalW and the TreeView computer program. The novel SAgs belong to group A, together with SPE-C, SPE-G, SPE-J and SMEZ, but build a separate branch within this group.

for their potency to stimulate peripheral blood lymphocytes (PBLs). Both toxins were active on human T cells with half maximum responses of 1 pg/ml unlabelled toxins [ µg] (rSPE-L) and 10 pg/ml (rSPE-M). Fig.2. Competition binding studies with rSPE-L and rSPE-M. LG-2 cells were incubated in duplicates with 1 ng of 125I-labelled Vb specificity of recombinant toxins: The human recombinant toxin and increasing amounts of unlabelled toxins. After TcR Vb specificity of the recombinant toxins was 1 h cells were washed and counted. Binding of radiolabelled rSPE- determined by multiprimer anchored PCR and dot blot L or rSPE-M was inhibited with excess amounts of unlabelled analysis using a panel of 21 human Vb DNA regions. rSPE-C, rSPE-L and rSPE-M. In contrast, excess amounts of The Vb enrichment after stimulation with toxin was unlabelled rSEB or rTSST did not influence binding of radiolabelled rSPE-L and rSPE-M. A. Competition assay with labelled rSPE-L. compared to the Vb profile of PBLs stimulated with B. Competition assay with labelled rSPE-M. the unspecific T cell mitogen ConA. Both toxins showed very similar results with Vb1.1 TcR being MHC class II binding: The protein sequences of the primary target. T cells carrying the Vb1.1 TcR SPE-L and SPE-M contain the conserved zinc binding were stimulated >80 times and nearly 70 times more motif that is required for binding to the polymorphic MHC with rSPE-L and rSPE-M, respectively, than with class II b-chain. In an attempt to determine the orientation ConA. of the toxins on MHC class II competition binding

PROFT et al: NOVEL STREPTOCOCCAL SUPERANTIGENS 41 experiments were performed. The recombinant toxins seroconversion in the normal population appeared lower were radiolabelled and tested with excess of unlabelled than to other SAgs, such as SMEZ15 indicating a lower toxin for binding to LG-2 cells. Recombinant SEB, distribution. This is consistent with the low frequency of rTSST and rSPE-C were used as reference proteins. genes found in clinical isolates from New Zealand. SPE-C binds exclusively to the MHC class II b-chain using the zinc binding motif19, while SEB and TSST both Discussion exclusively bind to the MHC class II a-chain in a 20 different, zinc-independent binding mode . The results Two novel sag genes, sepe-l and sepe-m, have been are shown in Fig. 2. These results suggest that both, identified from the S. equi genome database at the SPE-L and SPE-M, bind MHC class II exclusively to Sanger Centre. Subsequently, the orthologues spe-l and the b -chain in a zinc-dependent binding mode. This is a spe-m genes were identified from S. pyogenes. consistent feature among all group A members of the Genotyping of 40 S. pyogenes isolates showed that SAg family. spe-l and spe-m occurred in low frequencies (15 and 5%, respectively) and were therefore likely to be on

Seroconversion against SPE-L and SPE-M: Specific mobile DNA element(s). The S. equi orthologues sepe- antibodies against pathogen factors are often produced l and sepe-m could not be detected in any of the analysed after bacterial infections and can protect the individual S. equi isolates suggesting that they are rare in this from further encounter with bacteria producing these species. The spe-l gene is identical to a hypothetical toxins. Widespread seroconversion to staphylococcal and sag gene that has recently been identified on the streptococcal SAgs is common in the adult population. streptococcal prophage PhiHIH1.1. Ikebe et al21 showed Twenty sera from healthy donors were analysed for that all of 18 M3/T3 isolates rocovered from Japan during antibodies against SPE-L and SPE-M. 125I-labelled toxin or after 1992 have the spe-l gene, while none of 10 T3 was used to bind specific IgG antibodies, which were isolates recovered before 1973 have this gene. This then precipitated with staphylococcal cell bound protein suggests a very recent horizontal transfer of spe-l A (IgG receptor protein) (Fig.3). These results provided between different S. pyogenes isolates and between evidence that both, SPE-L and SPE-M, were expressed and secreted by some streptococcal isolates and generate S. pyogenes and S. equi. One of the most interesting effective antibody titres. However, the frequency of this outcomes of this study was the finding that the spe-l gene was linked to S. pyogenes M89, a serotype that is strongly associated with acute rheumatic fever in New Zealand17. Superantigens have been implicated in acute rheumatic fever, but their role in the pathology of the disease remains elusive18. Screening of a much larger panel of clinical isolates is currently underway to determine the strength of this linkage between spe-l and acute rheumatic fever causing strains. The novel toxins are closely related to the streptococcal and staphylococcal SAgs. Recombinant forms of SPE-L and SPE-M are extremely potent stimulators of human PBLs at nanomolar concentrations, confirming their role as SAgs. Fig.3. Seroconversion experiment shows the percentage of bound toxin compared to the total amount of toxin used. Seven subjects The data presented here suggest that SPE-L and were identified with seroconversion against SPE-L, of which four SPE-M bind MHC class II exclusively at the b-chain had a strong titre (sample 6, 11, 19, 20), one had a moderate titre (13) via a zinc binding complex. This assumption was and two had a low titre (2, 9). Six serum samples had significant supported by the fact that both toxins contained the antibody titres against SPE-M [two high (5,19), three moderate (6, 11, 15) and one low (9)]. Four sera showed conversion against characteristic zinc binding motif (H-X-D) close to the both toxins (6, 9, 11, 19). BSA and rabbit antibodies against SPE-C C-terminus and by the results from the competitive and SMEZ showed no binding to labelled SPE-L or SPE-M indicating binding assays. This MHC class II binding mode appears the strong specificity of the antibodies. , SPE-L; , SPE-M. to be conserved among all group A members of the SAg

42 INDIAN J MED RES (SUPPL) MAY 2004 family. Interestingly, SPE-L and SPE-M both stimulated comparative study of their molecular biology. Chem Immunol mainly Vb1.1 expressing T cells. Usually, SAgs can be 1992; 55 : 1-35. distinguished by their characteristic combination of Vb 2. Fraser JD, Arcus V, Kong P, Baker EN, Proft T. Superantigens targets. This finding might explain why none of the - powerful modifiers of the immune system. Mol Med Today analysed streptococcal isolates carried both genes, spe-l 2000; 6 : 125-32. and spe-m, as the need to activate a single Vb could be 3. Herman AJ. Kappler W, Marrack P, Pullen AM. Superantigens: met by a single toxin and two toxins with the same mechanism of T-cell stimulation and role in immune responses. function would be unlikely to offer a significant selective Annu Rev Immunol 1991; 9 : 745-72. advantage. 4. Marrack P, Kappler J. The staphylococcal enterotoxins and their relatives. Science 1990; 248 : 705-11. Seroconversion was observed in blood samples from 5. Bohach GA, Fast DJ, Nelson RD, Schlievert PM. Staphylococcal healthy donors at frequencies of 35 and 30 per cent and streptococcal pyrogenic toxins involved in toxic shock respectively. This correlated well with the low frequencies syndrome and related illnesses. Crit Rev Microbiol 1990; 17 : of the corresponding genes in S. pyogenes isolates. In 251-72. contrast, seroconversion against at least one SMEZ 6. Kotzin BL, Leung DY, Kappler J, Marrack P. Superantigens variant was observed in 100 per cent of 100 healthy and their potential role in human disease. Adv Immunol 1993; 54 : 99-166. donors from New Zealand15. The smez gene is chromosomally encoded and was found in all analysed 7. Cone LA, Woodard PM, Schlievert PM, Tomory GS. Clinical S. pyogenes isolates. Furthermore, seroconversion and bacteriologic observations of a toxic shock-like syndrome due to Streptococcus pyogenes. N Engl J Med 1987; 317 : against SPE-C was observed in blood samples at 146-9. 60 per cent (data not shown), and the spe-c gene was 8. Bernal A, Proft T, Fraser JD, Posnett DN. Superantigens in detected in 42 per cent of the S. pyogenes isolates. human disease. J Clin Immunol 1999; 19 : 149-57. Seroconversion against a secreted bacterial protein is not only an indicator of actual gene expression, but also 9. Kotzin BL, Leung DY, Kappler J, Marrack P. Superantigens and their potential role in human disease. Adv Immunol 1993; suggests a potential role as virulence factor. 54 : 99-166. 10. Artiushin SC, Timoney JF, Sheoran AS, Muthupalani SK. In conclusion, two novel superantigens were found Characterization and immunogenecity of pyrogenic mitogens in streptococcal strains belonging to 2 different SePE-H and SePE-I of Streptococcus equi. Microbial Pathol Lancefield groups, group A and group C. The sag genes 2002; 32 : 71-85. are most likely located on a mobile DNA element that 11. Harrington DJ, Sutcliffe IC, Chanter, N. The molecular basis of allows horizontal transfer of these genes between Streptococcus equi infection and disease. Microb Infect 2002; different streptococcal species, such as S. equi and 4 : 501-10. S. pyogenes. The linkage of spe-l to the acute rheumatic 12. Papageorgiou AC, Acharya KR. Microbial superantigens: from fever causing S. pyogenes M89 strain suggests a structure to function. Trends Microbiol 2000; 8 : 369-75. potential involvement of SPE-L in this disease. It remains 13. Proft T, Moffatt SL, Berkahn CJ, Fraser JD. Identification and to be seen whether sag genes will also be found in other characterization of novel superantigens from Streptococcus non group A streptococci. pyogenes. J Exp Med 1999; 189 : 89-101. 14. Proft T, Arcus VL, Handley V, Baker EN, Fraser JD. Acknowledgment Immunological and biochemical characterization of streptococcal pyrogenic exotoxins I and J (SPE-I and SPE-J) from Streptococcus pyogenes. J Immunol 2001; 166 : 6711-9. The authors acknowledge the S. equi Sequencing Group at the Sanger Institute for making their DNA sequences available on the 15. Proft T, Moffatt SL, Weller KD, Paterson A, Martin D, internet, and the Health Research Council, New Zealand and by an Fraser JD. The streptococcal superantigen SMEZ exhibits wide Early Career Excellence Grant from the University of Auckland for allelic variation, mosaic structure, and significant antigenic financial support. variation. J Exp Med 2000; 191 : 1765-76. 16. Hudson KR, Robinson H,Fraser JD. Two adjacent References residues in Staphylococcal enterotoxins A an E determine T cell receptor Vb specificity. J Exp Med 1993; 177 : 175-85. 1. Betley MJ, Borst DW, Regassa LB. Staphylococcal enterotoxins, 17. Li PL, Tiedemann RE, Moffat SL, Fraser JD. The toxic shock syndrome toxin and streptococcal exotoxins: a superantigen streptococcal pyrogenic exotoxin C (SPE-C)

PROFT et al: NOVEL STREPTOCOCCAL SUPERANTIGENS 43

exhibits a novel mode of action. J Exp Med 1997; 186 : 20. Seth A, Stern LJ, Ottenhoff TH, Engel I, Owen MJ, 375-83. Lamb JR, et al. Binary and ternary complexes between T-cell 18. Martin DR, Voss LM, Walker SJ, Lennon D. Acute rheumatic receptor, class II MHC and superantigen in vitro. Nature fever in Auckland, New Zealand: spectrum of associated Group 1994; 369 : 324-7. A streptococci different from expected. Pediatr Infect Dis J 1994; 13 : 264-9. 21. Ikebe T, Wada A, Inagaki Y, Sugama K, Suzuki R, Tanaka D, et al. Dissemination of the phage-associated novel 19. Roussel A, Anderson BF, Baker HM, Fraser JD, Baker EN. Crystal structure of the streptococcal superantigen SPE-C: superantigen gene speL in recent invasive and noninvasive dimerization and zinc binding suggest a novel mode of interaction Streptococcus pyogenes M3/T3 isolates in Japan. Infect Immun with MHC class II molecules. Nat Struct Biol 1997; 4 : 635-43. 2002; 70 : 3227-33.

Reprint requests: Prof. John D. Fraser, Department of Molecular Medicine & Pathology, Faculty of Medical & Health Sciences University of Auckland, Private Bag 92019, Auckland, New Zealand e-mail: [email protected]

Indian J Med Res 119 (Suppl.) 2004, pp 44-47

Promotion of fibronectin independent invasion by C5a peptidase into epithelial cells in group A Streptococcus

Sai Sudha Purushothaman, Hae-Sun Park & P. Patrick Cleary

Department of Microbiology, University of Minnesota, MN 55455, USA

Received August 6, 2003

Background & objectives: Group A Streptococcus, causative agent of several clinical manifestations codes for multiple protein invasins which help the bacterium to enter non-phagocytic cells. C5a peptidase (SCPA) is a surface protein conserved among different serotypes of M1 strain. The present study was taken up to study SCPA promoted fibronectin independent entry of GAS into epithelial cells.

Methods: An isogenic 90226 emm1DAB (M1– ) mutant was constructed with thermosensitive pGhost vector. This isogenic M1– mutant expressed SCPA on the surface as determined by Western blotting and immunofluorescence. Results: On preincubation with anti-SCPA serum, the isogenic M1– strain exhibited 54 per cent decreased invasion as compared to the bacteria incubated with control serum. Also, purified recombinant SCPA proteins blocked internalization of M1– streptococci into HEp-2 cells. The M1– strain invaded at the same efficiency in the presence or absence of fibronectin . Interpretation & conclusion: These results suggested that SCPA acted as a potential invasin of group A streptococcus and promoted invasion independent of fibronectin. Key words C5a peptidase - fibronectin - group A streptococcus - invasion - M1 strain

Group A Streptococcus (GAS) is an etiological agent C5a peptidase (SCPA) is 128 kDa cell surface of different clinical conditions ranging from pharyngitis, serine protease which belongs to the subtilisin group impetigo, scarlet fever to post infectious sequelae of of proteases5,6. It has substrate specificity for C5a rheumatic heart disease and glomerulonephritis1. and cleaves between His 67- Lys 68, thus reducing Although GAS has generally been considered an the ability of this anaphylatoxin to bind receptors on extracellular pathogen, it has evolved mechanisms to polymorphonuclear leukocytes (PMLs). SCPA penetrate the epithelial cells which lead to intracellular expressed on most GAS isolates independent of life. This helps the pathogen to evade the immune system, serotype, is highly conserved among different species antibiotics and also access to deeper tissues2. Group A of streptococcus. The C terminal region of the Streptococcus adherence and internalization into protein has a peptidoglycan anchor sequence, epithelial cells is a complex process which involves common with other Gram positive bacterial surface multiple adhesins/invasins. SfbI and its allelic variant proteins. C5a peptidase is a major virulence factor PrtF (SfbI/PrtF) and M proteins are some of the proteins of GBS that promotes invasion into epithelial cells which have been reported to promote invasion of and also depletes C5a concentration at the bacterial epithelial cells by GAS3,4. localization7.

44 PURUSHOTHAMAN et al: SCPA PROMOTED INVASION OF GAS INTO EPITHELIAL CELLS 45 As several surface components in GAS have been (SDS-PAGE), Western blotted and probed with anti- shown to participate in the interaction with eukaryotic SCPA antibodies8. cells, we attempted to demonstrate that SCPA promotes fibronectin independent entry of GAS into epithelial cells. Results & Discussion

Material & Methods The surface proteins expressed by GAS are highly variable among more than 100 serotypes and in fact Bacterial strains: GAS strain 90 226 was originally can vary within a particular serotype3. The per cent isolated from the blood of a patient with sepsis and was invasion contributed by M proteins in different genetic a serotype M1 strain. A deletion of the A and B repeats backgounds in shown in the Table. JRS4 (PrtF+/M6+ ) of emm1 (M1-) was made with thermosensitive pGhost invaded HEp-2 cells by 32.2 per cent and the major vector6. Surface expression of SCPA on this isogenic contributor to this activity was prtF as inactivation of M1- mutant was confirmed by Western blotting and this gene (SAM1) reduced invasion by 88 per cent. immunofluorescence. This was in contrast with the study done by Jadoun et al9 where they observed about 63 per cent decrease in Invasion assay: Hep-2 cells (ATCC CCL23) were invasion upon inactivation of prtF. Paradoxically, the grown in RPMI medium supplemented with 10 per emm6 gene (pVE5508) when expressed in cent foetal bovine serum ( Life Technologies, USA) and Lactobacillus lactis (VEL1122) promoted high antibiotics [5 µg/ml penicillin and 100 µg/ml streptomycin frequency invasion of this otherwise non-invasive (Sigma Chemical Co., USA)]. Cells were infected with bacterium (25%). M1 expressing strains, 90 226 and bacteria in antibiotics free media at a multiplicity of AP1 promoted internalization of GAS into HEp-2 cells infection of 1:5. The invasion assay was performed by 24 and 16.6 per cent and these lack prtF genes. according to the method of Cue et al3. For fibronectin Thus, M 1 protein was capable of mediating high independent experiments, invasion assay was done in frequency invasion independent of PrtF/Sfbl. These MEM medium without foetal calf serum which was the data confirmed that GAS coded for multiple invasins source of fibronectin. to become an insider and their activity was strain dependent10. Extraction of surface proteins: Stationary phase cultures were centrifuged and the bacterial pellet washed Our earlier study7 with GBS demonstrated that with 5ml of ice cold TE buffer [ 50mM Tris-HCl (pH7.5)/ C5a peptidase is a major invasin of this pathogen. In 1mM EDTA] containing 1 mM (PMSF) (Sigma, USA) this study, preincubation of bacteria with rabbit anti- and antipain (1 µg/ml).The bacterial pellet was then C5a peptidase of group B streptococcus (SCPB) reduced suspended in 1 ml of lysis buffer (TE buffer) containing invasion into A549 cells by 68 per cent and genetic 20 per cent sucrose, 10 mg/ml lysozyme, 100 units of inactivation of SCPB reduced invasion of Hep-2 cells mutanolysin, 1mM PMSF and antipain and incubated at by 81 per cent7. For comparison, this results is included 37ºC for 2 h. The cell free supernatant was concentrated in the Table. These results suggested that SCPB is the in Centricon 100 (Millipore, USA) according to the major invasin for GBS. Our interest thus became focused manufacturer’s instructions.

Table. Invasion of HEp-2 cells by different GAS strains Anti - SCPA antibodies: Rabbits were immunized with purified recombinant SCPA proteins. 10 per cent antisera Strains Characteristics % invasion was incubated with the bacteria for 1h at room JRS4 M6+/ PrtF+ 32.2 temperature prior to infection of monolayers to assess SAM1 M6–/ PrtF+ 3.7 its capacity to neutralize internalization of streptococci4. AP1 M1+/ PrtF– 16.6 90 226 M1+/ PrtF– 24.0 Western blotting: The surface proteins were electrophoresed and analyzed on 10 per cent sodium Per cent invasion is the intracellular cfu divided by the initial inoculum dodecyl sulphate polyacrylamide gel electrophoresis cfu 46 INDIAN J MED RES (SUPPL) MAY 2004 on the role of C5a peptidase in GAS invasion of epithelial cells. Preliminary studies demonstrated the ability of anti-SCPA antibodies to prevent invasion by GAS into HEp-2 cells.

Strain 90226 is a PrtF¯/SfbI¯/ M1+ that invades epithelial cells efficiently. M1 protein is the major invasin in this genetic background contributing to 90 per cent of invasion8. An isogenic M1– mutant invaded at a frequency of 2.3 per cent, which is 10 per cent of that exhibited by M1+ bacteria. This low level invasion by M1– streptococcus was equivalent to that observed for group B streptococcus. Invasion by this M1– strain must be promoted by invasins other than M protein. Fig.1. Western blot analysis of surface extract SCPA proteins of D – wild type and emm1 AB. Lane 1: molecular weight marker; Lane 2: To ascertain whether this isogenic M1 mutant Purified recombinant SCPA protein; Lane 3: wild type 90226; Lane expressed SCPA, surface proteins of both the wild type 4: emm1DAB 90-226 (M1–). and the M1– mutant were extracted and compared by Western blotting with polyclonal SCPA antibodies. recombinant SCPA proteins bound to epithelial cells Both the wild type and the isogenic mutant M1– with high affinity (data not shown). Preincubation of expressed equimolar concentrations of SCPA on their the epithelial cells with soluble recombinant SCPA surfaces (Fig.1). To determine, if invasion by M1– proteins competitively inhibited the invasion by streptococcus was due to the virulence contribution of bacteria. This demonstrated that the interaction SCPA protein, the capacity of antibody to inhibit M1– between SCPA and epithelial cells was essential for streptococcal invasion was tested. On preincubation the entry process. of M1- streptococci with SCPA antibodies, invasion was inhibited by 54 per cent confirming that SCPA was a Soluble fibronectin mediates the binding of PrtF1/ a b contributing factor for this strain (Fig. 2a). Soluble Sfb1, M protein to 5 1 integrins. High frequency 5

4 4

3 3

2 2 Per cent invasion Per cent invasion

1 1

2a 2b 0 M1- M1- M1- pre O90* O90R+anti O90R del* Fn no Fn anti-SCPA incubation with SCPB serum* serum rSCPA protein Fig.2a. Invasion of HEp-2 cells promoted by C5a peptidase of GAS ((M1– mutant) and GBS. Per cent invasion is the intracellular cfu divided by the initial inoculum cfu. The data are mean ± standard error of the mean of three different experiments. * The results were adapted from Cheng et al 20027. Fig.2b. SCPA mediated invasion of HEp-2 cells in the presence and absence of fibronectin. Per cent invasion is the intracellular cfu divided by the initial inoculum cfu. The data are mean ± standard error of the mean of three different experiments. PURUSHOTHAMAN et al: SCPA PROMOTED INVASION OF GAS INTO EPITHELIAL CELLS 47 invasion of epithelial cells by strain 90226 was Acknowledgment fibronectin dependent8. Anti-integrin antibodies and media without fibronectin reduced invasion by 85 per The authors thank Dr J.C. Piard for L. lactis VEL1122 and pVE5508, Dr M. Caparon for JRS4 strains and Tim Leonard for cent. The A and B repeats of M protein are essential help in figure preparation. This work was supported by Public Health for fibronectin binding, thus, the 90226 emm1D AB Services (PHS) grant AI34503, USA. (M1–) mutant could neither bind to fibronectin nor bind to integrin8. The 15 per cent residual activity of References the M1– mutant was independent of fibronectin/ integrins. SCPA expressing M1– streptococci invaded 1. Bisno AL, Stevens DL. Streptococcal infection of skin and soft tissues. N Engl J Med 1996; 334 : 240-50. Hep-2 cells at the same level in the absence/presence 2. La Penta D. Rubens C, Chi E, Cleary PP. Group A streptococcus of fibronectin (Fig. 2b). This is in contrast to the invasion efficiently invade human respiratory epithelial cells. PNAS promoted by M protein3. To confirm that Fn 1994; 91 : 12115-9. independent invasion is SCPA mediated, antibody 3. Cue D, Dombek PE, Hong L, Cleary PP. Streptococcus pyogenes inhibition experiments were done. Preincubation of M1– serotype M1 encodes multiple pathways for entry into human streptococci with anti-SCPA antibodies blocked epithelial cells. Infect Immun 1998; 66 : 4593-601. invasion by 65 per cent in the absence of fibronectin, 4. Molinari G, Rhode M, Guzman CA, Chhatwal GS. Two distinct pathways for the invasion of Streptococcus pyogenes in non- suggesting that Fn independent invasion was dependent phagocytic cells. Cell Micobiol 2000; 2 : 145-54. on SCPA (data not shown). Thus invasion promoted 5. Chen C, Cleary PP. Complete nucleotide sequence of the by SCPA contrasted to that of several fibronectin streptococcal C5a peptidase gene of Streptococcus pyogenes. dependent invasin molecules such as M, PrtF/SfbI J Biol Chem 1990; 265 : 3161-7. proteins. These data point to the fact that GAS codes 6. Ji Y, McLandsborough L, Kondagunta A, Cleary PP. C5a for a variety of invasins which lead to different peptidase alters clearance and trafficking of group A streptococci mechanism for host cell entry. by infected mice. Infect Immun 1996; 64 : 503-10. 7. Cheng Q, Stafslein D, Purushothaman SS, Cleary PP. The group B streptococcal C5a peptidase is both a specific protease Vaccine candidates include M protein and PrtF/SfbI, and an invasin. Infect Immun 2002; 70 : 2408-13. but the former exhibits a high degree of antigenic 8. Cue D, Lam H, Cleary PP. Genetic dissection of the Streptococcus variability and the latter is not produced by many pyogenes M1 protein: regions involved in fibronectin binding serotypes, including highly virulent M1 and M3 strains11. and intracellular invasion. Microb Pathol 2001; 31 : 231-42. The fact that SCPA is highly conserved among different 9. Jadoun J, Ozeri V, Burstein E, Skutelsky E, Hanski E, Sela S. serotypes and is on the cell surface makes it a novel Protein F1 is required for efficient entry of Streptococcus pyogenes into epithelial cells. J Infect Dis 1998; 78 : 147-55. vaccine candidate. Our results indicated that antibodies 10. Berkower C, Ravins M, Moses AE, Hanski E. Expression of against SCPA blocked internalization of streptococci different group A streptococcal M-proteins in an isogenic and neutralized the C5a peptidase activity, enabling background demonstrates diversity in adherence to and recruitment of polymorphoneutrophil leukocytes (PMLs) invasion of eukaryotic cells. Mol Microbiol 1999; 31 : 1463- to sites of infection. Thus, immunization with SCPA 75. could produce a dual protective response which acts at 11. Natanson S, Sela S, Moses AE, Musser JM, Caparon MG, Hanski E. Distribution of fibronectin binding proteins among different levels of infection to promote early clearance group A streptococci of different M types. J Infect Dis 1995; of bacteria. 171: 871-8.

Reprint requests: Dr Sai Sudha Purushothaman, Molecular Biophysics Unit, Indian Institute of Science, Bangalore 560012, India e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 48-56

Control of streptokinase gene expression in group A & C streptococci by two-component regulators

Horst Malke & Kerstin Steiner

Institute for Molecular Biology, Friedrich Schiller University Jena, D-07745 Jena, Germany

Received August 6, 2003

Background & objectives: Group A streptococci (GAS) and human isolates of group C streptococci (GCS) have the stable capacity to produce the plasminogen activator streptokinase, albeit with varying efficiency. This property is subject to control by two two-component regulatory systems, FasCAX and CovRS, which act as activator and repressor, respectively. The present work aims at balancing these opposing activities in GAS and GCS, and at clarifying the phylogenetic position of the FasA response regulator, the less understood regulator of the two systems. Methods: The GCS strain H46A and GAS strain NZ131 were used. Escherichia coli JM 109 was used as host for plasmid construction. Streptokinase activity of various wild type and mutant strains was measured. Phylogenetic trees of streptococcal FasA homologues were established. Results: The streptokinase activities of the GAS strain NZ131 and the GCS strain H46A were attributable to more efficient CovR repressor action in NZ131 than in H46A. The FasA activator, on the other hand, functioned about equally efficient in the two strains. Phylogenetically, FasA homologues clustered distinctly in the proposed FasA-BlpR-ComE family of streptococcal response regulators and used the LytTR domain for DNA binding. Interpretation & conclusion: Assessing the apparent streptokinase activity of streptoccal strains require the dissection of the activities of the cov and fas systems. Although experimental evidence is still missing, FasA is closely related to a widely distributed family of streptococcal response regulators that is involved in behavioral processes, such as quorum sensing. Key words CovR - FasA - phylogenetic tree of FasA - Streptococcus dysgalactiae subsp. equisimilis - Streptococcus pyogenes

The gene for the plasminogen activator streptokinase involved in the expression control of this important appears to be consistently present in all group A (GAS) virulence factor. Regarding the characterization of cis- and human isolates of group C streptococci (GCS). It is active sites, S1 nuclease experiments have identified the monocistronically expressed from a highly preserved core promoter and the major transcription initiation site2. chromosomal region in which it is interspersed among Circular permutation analysis combined with five unrelated genes transcribed in the opposite direction. determination of the activity of nested deletions in the Despite the omnipresence of the gene, its expression promoter-upstream region identified an intrinsic DNA levels can vary considerably among strains. Thus, bending locus which has a pivotal role in streptokinase individual isolates may differ in their streptokinase- (SK) gene expression3,4. In addition, the use of reporter determined capacity to generate an optimized proteolytic gene constructs in allele swap experiments between GAS habitat in the infected host and, consequently, may exhibit and GCS strains revealed that the host genetic different degrees of invasiveness1. Recent investigations background dictates the SK gene expression levels4. This have begun to shed some light on the regulatory systems suggested the existence of trans-acting factor(s) with

48 MALKE & STEINER : STREPTOKINASE GENE REGULATION 49 strain-specific activity that contact cis-active sites, for E. coli; kanamycin, 100 µg/ml for streptococci and thereby modulating streptokinase gene expression. 50 µg/ml for E. coli; spectinomycin, 100 µg/ml for both Subsequent work carried out by a number of different streptococci and E. coli. laboratories identified such trans-acting factors as components of two independent two-component signal Construction of cov and fas plasmids: General nucleic transductions systems, covRS regulating streptokinase acids techniques have been described previously1. gene expression negatively5,6 and fasCAX involved in Oligonucleotide primers were used in polymerase chain positive regulation of this gene7. Whereas initially the reactions (PCR) for the construction of recombinant action of these systems was studied independently of plasmids (Table I). Plasmids were electrotransformed one another, the balance between their opposing actions into strains H46A or NZ131 to insertionally inactivate in the GAS strain NZ131 to that in the GCS strain H46A the specified cov and fas genes, or to complement the was compared in the present study. The distribution and mutations. The construction and characteristics of the phylogenetic relationships of response regulator FasA resultant strain H46A and strain NZ131 derivatives are homologues in the finished and unfinished genome described in Table II. sequences of various species of the genus Streptococcus were also analysed. Streptokinase activity assay: The plasminogen activation assay on microtiter plates14 was used to Material & Methods measure streptokinase activity in BHI culture supernatant fluids of the various wild type and mutant Bacterial strains and growth: The GCS strain H46A strains. The release of para-nitroaniline from the and the GAS strains NZ131 and SF370 used in these chromogenic substrate H-D-valyl-leucyl-lysin p- experiments were grown in ambient air at 37 °C without nitroaniline (Sigma, USA) was measured at OD over agitation in brain heart infusion (BHI) broth (Difco, USA). 405 E. coli JM109 was used as host for plasmid time, and activity rates were calculated from the linear constructions and was grown at 37 °C in rotary flasks in parts of absorbance versus time plots. Standard standard Luria broth (LB). If appropriate, antibiotics were streptokinase was procured from Sigma, with 1 unit (U) used at the following concentrations: chloramphenicol, 3 being defined as the protein activity capable of liquifying µg/ml for streptococci and 100 µg/ml for E. coli; a standard clot of fibrinogen, plasminogen and thrombin erythromycin, 2.5 µg/ml for streptococci and 200 µg/ml at pH 7.5 and at 37 °C in 10 min.

Table I. PCR primers used for the construction of cov and fas plasmids

Primer pair Sequencea Template DNA Purpose

covRF1 ccggaattcCAAGGGTTGTTTGATGAATA SF370 pMcovRSF construction covRR1 ccggaattcATGACTTATTTCTCAC

covRF1 ccggaattcCAAGGGTTGTTTGATGAATA NZ131 pVAcovRSNZ construction covSR1 cgcctgcagCTTAAGCTACTCTAACTCTC F2 cgcggatccCATGATGATGACTGCGCGTGA NZ131 pFC1 construction R3 cgcaagcttCATGACACGATTCATATTAGTC fasAF1B cgcggatccCCAGCAGACACGCATAGAATC H46A pFW5fasA1 and pFW15fasA1 construction fasAR2H cgcaagcttCGGTGCTACTGCTTGAATCTCAG fasAF2B cgcggatccCTTGAATTAGCTGCAGCTATTCG H46A pFW14 fasA2 construction fasAR1H cgcaagcttCACAAGGTAGGATCTATGGC

fasAF3 tcccccgggGACAATTGTTAGAAAGGAGATAAAG SF370 pFasAXSF construction fasXR3 cgcaagcttGACGTCAGCTACTTATCCCTG a Lower case letters indicate sequences added to facilitate cloning 50 INDIAN J MED RES (SUPPL) MAY 2004 Sequence analysis of streptococcal FasA sequence of the DNA-binding domain derived from a homologues: Similarity searches with varying multiple alignment of 21 streptococcal FasA homologues parameters against publicly available databases was determined using the SeqLogo program in the (www.ncbi.nlm.nih.gov;www.sanger.ac.uk; implementation by SE Brenner (www.bio.cam.ac.uk/ www.tigr.org) containing finished and unfinished genomic seqlogo). Secondary structure predictions were sequences of Streptococcus species were performed performed using the programs of the PredictProtein using the TBLASTN program (www.ncbi.nlm.nih.gov) server (cubic.bioc.combia.edu/predictprotein). with Streptococcus dysgalactiae subsp. equisimilis FasA as a query1. The multiple alignment of the retrieved Results sequences and their phylogenetic trees were generated by distance matrix analysis using the AllAll program Balancing the action of the cov and fas systems in (cbrg.inf.ethz.ch/Server/AllAll.html). The consensus GCS and GAS: Streptokinase activities of cell free

Table II. Characteristics of the bacterial strains and plasmids used

Strain or plasmid Description or relevant phenotype

Plasmids pMG36, pMG36e8 Vectors; Km-r and Em-r, respectively pFW5, pFW14, pFW159 Vectors; Spc-r, Cm-r, Em-r, respectively pVA891210 Suicide vector; Em-r 1 pMcovRSF pMG36e containing covRSF370 under ist natural promoter; Em-r 1 pVAcovRSNZ pVA8912 containing covRSNZ131 under its natural promoter; Em-r 1 pVAcovRSNZdI, dII pVA8912 containing covRSNZ131 under its natural promoter

and the 3´end of dexBH46A (orientation dI and dII, respectively) for genomic integration into H46A; Em-r 11 pFC1 pFW5 containing a 313-bp internal fragment of covRNZ131; spc-r 1 pFW5fasA1 pFW5 containing an internal fragment of fasAH46A; Spc-r pFW15fasA11 As pFW5fasA1; Em-r 1 pFW14fasA2 pFW14 containing an internal fragment of fasAH46A; Cm-r 1 pFasAXSF pMG36 containing fasAXSF370 under vector promoter P32; Km-r Streptococcus dysgalactiae subsp. equisimilis Human serogroup C (GCS) H46A1,12 Cov- (K102amber) Fas+ 1 + + H46A(pMcovRSF) Cov Fas ; Em-r 1 + + H46AdexB::pVAcovRSNZdI Cov Fas ; Em-r 1 + + H46AdexB::pVAcovRSNZdII Cov Fas ; Em-r H46AfasA:: pFW15fasA11 Cov- Fas-; Em-r H46AfasA:: pFW14fasA21 Cov- Fas-; Cm-r 1 - + H46AfasA:: pFW15fasA1(pFasAXSF) Cov Fas ; Em-r Km-r + - H46AdexB::pVAcovRSNZdII, Cov Fas ; Em-r Spc-r fasA:: pFW5fasA11 Streptococcus pyogenes Serogroup A (GAS) SF37013 M-type 1; Cov+ Fas+ NZ131* M-type 49; Cov+ Fas+ NZ131covR::pFC11 Cov- Fas+; Spc-r 1 + + NZ131covR::pFC1, covR::pVAcovRSNZ Cov Fas ; Spc-r Em-r NZ131covR::pFC1, fasA::pFW15fasA11 Cov- Fas-; Spc-r Em-r *Obtained from Prof. J.J. Ferretti Superscript numerals denote reference numbers MALKE & STEINER : STREPTOKINASE GENE REGULATION 51 culture fluids obtained from saturated BHI cultures of strains H46A and NZ131 were about 80 and 3 U/ml, respectively. Since wild type H46A possesses a naturally acquired amber mutation at codon position 102 of covR and so actually proved to be a derepressed mutant for streptokinase production1, the question was whether or not the great difference in the streptokinase activities between H46A and NZ131 reflected solely the state of their covR alleles. Since the streptokinase alleles, skc and ska, respectively, of the two strains were also subject to positive control by the fas system1, a comparative mutational approach was used to analyse the differential contributions of the two regulatory systems to streptokinase production. This approach involved the creation of all possible combinations of wild type and mutant covR and fasA alleles in the two strains, including complementation of mutant alleles to rule out possible effects of polar mutations (Table II). The results of the streptokinase activity assays of the various strains in cultures with comparable cell density are given in Fig. 1. First, restoration of CovR repressor activity in H46A by introduction of the covRNZ131 allele decreased its streptokinase activity to approximately 50 per cent. Compared to the low streptokinase activity of wild type NZ131, H46A (Cov+ Fas+) released 13 times more streptokinase than NZ131. Inactivation of the CovR repressor of NZ131 resulted in a greater than 10-fold increase of its streptokinase activity which, Fig. 1. Streptokinase activities of the GCS strain H46A and GAS however, was still approximately 50 per cent lower than strain NZ131 carrying active (+) or inactive (-) alleles of response that of H46A (Cov- Fas+). This difference might be regulators covR and fasA. ND, no activity detectable. Values are shown as mean ± SEM of 3 samples. In H46A Cov-Fas+ values a attributable to a slightly lower stimulatory effect of the mean of 3 assays from the same sample. fas system in NZ131, as suggested by the streptokinase activity ratios of Fas+ versus Fas- strains in a Cov- regulatory system was the observation that there was 13 background (1.5 in NZ131 vs. 2.0 in H46A). Taken one region in the S. pyogenes SF370 genome that exhibited similarity values >34per cent to the together, these results showed that the opposing Staphylococcus aureus accessory gene regulator activities of the cov and fas systems were about equal AgrAC and the S. pneumoniae competence regulatory in H46A whereas in NZ131 the repressive CovR system ComDE7, both of which were involved in quorum activity excelled the stimulatory FasA activity by a sensing. Recently, sequence analysis of bacterial factor of about 9. In the absence of both repression genomes has placed the AgrA and ComE response - - and activation, i.e., in a Cov Fas background, the regulators in a family of transcriptional regulators that constitutive streptokinase activities did not differ bind DNA with a novel domain, designated LytTR, which substantially between the two strains, and, expectedly, is distinct from the classical DNA-binding helix-turn-helix both strains showed their lowest activities in a Cov+ or winged helix domain of the overwhelming majority of Fas- background (Fig. 1). the response regulators (including CovR) of the bacterial two-component signal transduction systems15. It was of Evolutionary relationships among streptococcal considerable interest, therefore, to study the distribution FasA response regulators: The starting point of recent and phylogenetic relationships of streptococcal FasA investigations that led to the identification of the fas homologues. 52 INDIAN J MED RES (SUPPL) MAY 2004

TBLASTN searches of the databases with FasAH46A indicating that it approximated the original distances as a query retrieved, from 12 Streptococcus species, a between each pair of nodes in a satisfactory manner. set of 21 proteins (as of September 01, 2002) that showed The tree also showed that the PAM (per cent accepted >54 per cent sequence similarity (random expectation mutations) distance between AgrA and any other value, E<7x10-29) to the query sequence. The indicated streptococcal sequence was greater than the divergence threshold values were observed for sequence similarity between the most distantly related streptococcal between FasAH46A and AgrA. The unrooted sequences. The most interesting topological feature of phylogenetic tree (Fig. 2) based on the distance matrices the tree was its branching pattern, which showed that between the sequences had a fitting index of 0.57, the streptococcal sequences form 3 distinct clusters,

Fig. 2. Unrooted phylogenetic tree of the FasA-BlpR-ComE family of streptococcal response regulators. Figures at the branches are PAM distances. The black circle in the Sau_1 branch indicates the weighted centroid of the tree. Species abbreviations and sources of the proteins are as follows: Sag_1, S. agalactiae, AF390107; Sag_2, AL766854; San_1, S. anginosus, AJ000864; Sau_1, AgrA from Staphylococcus aureus, AY082629; Sdy_1, S. dysgalactiae subsp. equisimilis, AY075107; Seq_1, Seq_2, Seq_3, S. equi, www.sanger.ac.uk/Projects/S_equi; Sgo_1, S. gordonii, www.tigr.org/tdb/ufmg; Sgo_2, CAA66788; Sin_1, S. infantis, AF498313; Smi_1, S. mitis, www.tigr.org/tdb/ufmg; Smi_2, AJ000871; Smi_3, AJ240795; Smu_1, S. mutans, AE014133; Sor_1, S. oralis, AJ240794; Spn_1, S. pneumoniae, AJ27830; Spn_2, AF067659; Spn_4, U33315; Spy_1, S. pyogenes, AE009971; Spy_2, AF493605; Sub_1, S. uberis, www.sanger.ac.uk/Projects/S_uberis. MALKE & STEINER : STREPTOKINASE GENE REGULATION 53 referred to as the BlpR, ComE, and FasA cluster, after widely distributed, with occurrences in species of the the S. pneumoniae BlpR and ComE, and the S. pyogenic, mitis and mutans group. It was interesting to pyogenes FasA response regulators, involved in note, however, that as a member of the pyogenic group, bacteriocin biosynthesis and virulence factor regulation S. agalactiae did not appear to contain FasA, as (BlpR), competence development (ComE), and the indicated by the complete genome sequences16,17 of two regulation of the production of streptokinase, streptolysin different serotypes, both of which contained members S and fibronectin/fibrinogen-binding proteins (FasA). The of the BlpR cluster. Similarly interesting was the fasX gene, thought to encode a non-translated RNA7 observation, that all finished13,18,19 and unfinished functioning as the terminal effector of the fasCAX genomic sequences (www.sanger.ac.uk/Projects/ operon1, 7 had no homologues in the operons of the ComE S_pyogenes) of different S. pyogenes strains and BlpR clusters, underscoring the separate position of contained a FasA protein but not necessarily a BlpR the FasA regulators. protein, as indicated by its absence from the genome of the M1 strain SF37013. The BlpR-like protein The distribution of members of the ComE cluster designated Spy_2 in Fig. 2 is identical to SilA, a response appeared to be restricted to the transformable species regulator in the sil locus which has been found to be of the mitis-anginosus group; members of the FasA responsible for the invasive properties of JS95, an M- cluster occurred preferentially in the pyogenic group, type 14 GAS strain isolated from a case of necrotizing and the proteins of the BlpR cluster seemed to be most fasciitis20.

Fig. 3. Consensus sequence of the DNA-binding domain of the streptococcal FasA-BlpR-ComE family of response regulators, based on a multiple alignment of these domains from the proteins included in Fig. 2. Residue number 1 corresponds to Glu138 of the S. dysgalactiae subsp. equisimilis H46A FasA protein (AY075107), designated Sdy_1 in Fig. 2. Triangles point to the positions of highly conserved positively charged residues. 54 INDIAN J MED RES (SUPPL) MAY 2004 Concerning the domain composition of the FasA- observed in field strains. Several reasons may be BlpR-ComE response regulators, the N-terminal advanced to explain the different strength of the CovR halves of all proteins included in the phylogenetic tree repressor in H46A and NZ131. First, there existed (Fig. 2) contained the response regulator receiver sequence differences between the two strains in their domain archetypically represented by the CheY wider promoter regions, which might influence CovR domain of E.coli (NCBI protein database at binding. The proposed short consensus sequence www.ncbi.nlm.nih.gov/Structure/cdd/). As a (ATTARA) for CovR binding to the hasA promoter22 peculiarity, the CheY-homologous domain of the RgfA was seen only once relatively far upstream of the protein (designated Sag_1 in Fig. 2) of S. agalactiae streptokinase core promoter in both H46A and NZ131. strain O90R21 was N-terminally truncated, lacking However, there were 4 more ATTA tetranucleotides, in about 36 amino acid residues when compared to the which the thymin pair was found to be necessary for other FasA-BlpR-ComE regulators. This deletion was CovR binding22 in NZ131 than in H46A. Thus, the ska not present in the full-length BlpR-homologues (Sag_2) promoter region of NZ131 might present a better target of the two completely sequenced S. agalactiae for CovR repressor action than the corresponding H46A strains, NEM31616 and 2603V/R17. A more extensive region. However, differential regulation could also be deletion was seen in the BlpS protein from S. caused by different affinity of CovR to individual binding pneumoniae (acc. number, AAK04633; excluded sites, or by different expression levels of CovR which from the tree in Fig. 2) which lacked the receiver autoregulates its own gene5. domain completely. It is noteworthy that H46A contains a naturally The output domain of the bacterial response acquired nonsense mutation in covR and thus adds regulators was typically a DNA-binding domain to the repertoir of strains that have been shown to contained in their C-terminal portions. Sequence carry spontaneous covRS mutations23. At present, analysis of the C-terminal halves of the streptococcal it is not clear whether this locus is hypermutable or FasA-BlpR-ComE proteins generated the consensus mutations occur at random but there is strong in vivo sequence presented as a sequence logo in Fig. 3. It selection for the loss of covR expression. Either showed that among the amino acids with the highest possibility is intriguing, particularly in view of the fact degree of conservation (indicated by the height of that CovR is a global regulator that targets several the letters) were many of the positively charged virulence genes1,5,6. In either case, the size of a residues, the position of which also tended to coincide more virulent subpopulation in vivo would thus with high columns, reflecting the statistical depend on the point in time where mutation occurs importance of the given position. The possibility in the course of infection and thus presumably existed that some of these positions were directly contribute to the outcome of streptococcal infections. involved in DNA binding. Secondary structure The potentiation of the mode of pathogenicity would prediction made it likely that the DNA-binding domain appear to be most severe in strains that, as covRS of the FasA-BlpR-ComE family contained 3 beta- wild types, carry efficiently repressed virulence strands, 2 alpha-helices, a beta strand, and an genes. Given the high frequency of covRS mutations, additional alpha-helix, in that order (Fig. 3). The order vigilance should be used when clinical strains are of these structures was similar to that predicted for established from single colony isolates and data the LytTR domain15, providing corroborative generated with them are extrapolated to a clinical evidence for the novelty of this domain. situation potentially determined by a diverse population of bacteria. Discussion Compared to what is known about the covRS Dissecting the two opposing regulatory systems system, our knowledge of the fasCAX system is much involved in the modulation of streptokinase activity by less advanced. Its growth phase-dependent control mutation enabled us to balance their activity and provide activity over multiple streptococcal virulence factors, possible explanations for different streptokinase activities which appears to be influenced by the nutritional MALKE & STEINER : STREPTOKINASE GENE REGULATION 55 environment1,7, makes it an important factor involved 9. Podbielski A, Spellerberg B, Woischnik M, Pohl B, Lütticken in the pathogenesis of streptococcal disease. Although R. Novel series of plasmid vectors for gene inactivation and initial investigations failed to find evidence for its expression analysis in group A streptococci (GAS). Gene 1996; involvement in quorum sensing7, encouraged by a way 177: 137-47. of reasoning known as guilt by association, one should 10. Steiner K, Malke H. Transcription termination of the keep an open mind regarding its participation in complex streptokinase gene of Streptococcus equisimilis H46A: behavioral responses. It will be interesting to find out bidirectionality and efficiency in homologous and heterologous whether the distinct clustering of FasA within the well- hosts. Mol Gen Genet 1995; 246: 374-80. defined FasA-BlpR-ComE family of streptococcal 11. Steiner K, Malke H. Life in protein-rich environments: response regulators is reflected by a similarly distinct the relA-independent response of Streptococcus pyogenes range of target genes. to amino acid starvation. Mol Microbiol 2000; 38 : 1004-16. Acknowledgment 12. Christensen LR. Streptococcal fibrinolysis: a proteolytic Authors acknowledge the financial support from the reaction due to a serum enzyme activated by streptococcal Deutsche Forschungsgemeinschaft and the Fonds der fibrinolysin. J Gen Physiol 1945; 28 : 363-83. Chemischen Industrie. 13. Ferretti JJ, McShan WM, Ajdic D, Savic DJ, Savic G, Lyon References K, et al. Complete genome sequence of an M1 strain of Streptococcus pyogenes. Proc Natl Acad Sci USA 2001; 98 : 4658-63. 1. Steiner K, Malke H. Dual control of streptokinase and streptolysin S production by the covRS and fasCAX two- 14. Tewodros W, Norgren M, Kronvall G. Streptokinase activity component regulators in Streptococcus dysgalactiae subsp. among group A streptococci in relation to streptokinase equisimilis. Infect Immun 2002; 70 : 3627-36. genotype, plasminogen binding, and disease manifestations. 2. Gase K, Ellinger T, Malke H. Complex transcriptional control Microb Pathog 1995; 18 : 53-65. of the streptokinase gene of Streptococcus equisimilis H46A. 15. Nikolskaya AN, Galperin MY. A novel type of conserved DNA- Mol Gen Genet 1995; 247: 749-58. binding domain in the transcriptional regulators of the AlgR/ 3. Gross S, Gase K, Malke H. Localization of the sequence- AgrA/LytR family. Nucleic Acids Res 2002; 30 : 2453-9. determined DNA bending center upstream of the streptokinase gene skc. Arch Microbiol 1996; 166 : 116-21. 16. Glaser P, Rusniok C, Buchrieser C, Chevalier F, Frangeul L, Msadek T, et al. Genome sequence of Streptococcus agalactiae, 4. Malke H, Steiner K, Gase K, Frank C. Expression and regulation a pathogen causing invasive neonatal disease. Mol Microbiol of the streptokinase gene. Methods 2000; 21: 111-24. 2002; 45 : 1499-513. 5. Federle MJ, McIver KS, Scott JR. A response regulator 17. Tettelin H, Masignani V, Cieslewicz MJ, Eisen JA, Peterson S, that represses transcription of several virulence operons in Wessels MR, et al. Complete genome sequence and comparative the group A streptococcus. J Bacteriol 1999; 181 : 3649-57. genomic analysis of an emerging human pathogen, serotype V Streptococcus agalactiae. Proc Natl Acad Sci USA 2002; 99 : 6. Miller AA, Engleberg NC, DiRita VJ. Repression of virulence 12391-6. genes by phosphorylation-dependent oligomerization of CsrR at target promoters in S. pyogenes. Mol Microbiol 2001; 40 18. Smoot JC, Barbian KD, Van Gompel JJ, Smoot LM, : 976-90. Chaussee MS, Sylva GL, et al. Genome sequence and comparative microarray analysis of serotype M18 group 7. Kreikemeyer B, Boyle MD, Buttaro BA, Heinemann M, A Streptococcus strains associated with acute rheumatic Podbielski A. Group A streptococcal growth phase-associated fever outbreaks. Proc Natl Acad Sci USA 2002; 99 : virulence factor regulation by a novel operon (Fas) with 4668-73. homologies to two-component-type regulators requires a small RNA molecule. Mol Microbiol 2001; 39 : 392-406. 19. Beres SB, Sylva GL, Barbian KD, Lei B, Hoff JS, Mammarella 8. van de Guchte M, van der Vossen JM, Kok J, Venema G. ND, et al. Genome sequence of a serotype M3 strain of group A Construction of a lactococcal expression vector: expression of Streptococcus: phage-encoded toxins, the high-virulence hen egg white lysozyme in Lactococcus lactis subsp. lactis. phenotype and clone emergence. Proc Natl Acad Sci USA Appl Environ Microbiol 1989; 55: 224-8. 2002; 99 : 10078-83. 56 INDIAN J MED RES (SUPPL) MAY 2004

20. Hidalgo-Grass C, Ravins M, Dan-Goor M, Jaffe J, Moses 22. Federle MJ, Scott JR. Identification of binding sites for the AE, Hanski E. A locus of group A streptococcus involved in group A streptococcal global regulator CovR. Mol Microbiol invasive disease and DNA transfer. Mol Microbiol 2002; 46 : 2002; 43 : 1161-72. 87-99. 23. Engleberg NC, Heath A, Miller A, Rivera C, DiRita VJ. 21. Spellerberg B, Rozdzinski E, Martin S, Weber-Heynemann J, Spontaneous mutations in the CsrRS two-component regulatory Lütticken R. rgf encodes a novel two-component signal system of Streptococcus pyogenes result in enhanced virulence transduction system of Streptococcus agalactiae. Infect Immun in a murine model of skin and soft tissue infection. J Infect Dis 2002; 70 : 2434-40. 2001; 183 : 1043-54.

Reprint requests: Dr Horst Malke, FSU Jena, Institute for Molecular Biology, Winzerlaer Strasse 10, D-07745 Jena, Germany e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 61-65

Characterization of the interaction of the pneumococcal surface protein SpsA with the human polymeric immunoglobulin receptor (hpIgR)

Christine Elm, Manfred Rohde, Jean-Pierre Vaerman, Gursharan S. Chhatwal & Sven Hammerschmidt

Department of Microbial Pathogenicity, GBF-National Research Centre for Biotechnology, 38124 Braunschweig, Germany

Received August 6, 2003

Background & objectives:The polymeric immunoglobulin receptor (pIgR) is produced by mucosal epithelial cells and plays a crucial role in mucosal immunity. At the basolateral surface of mucosal cells, the pIgR binds predominantly polymeric immunoglobulins, such as dimeric IgA and polymeric IgA (pIgA) and mediates their transport across the polarized cells. This results in apical release of secretory component (SC), either free or bound covalently to IgA, forming secretory IgA (SIgA). The choline-binding protein (Cbp) SpsA, also called PspC and CbpA, has been shown to interact with the pIgR. A hexapeptide motif in SpsA was identified as the minimal binding motif required for binding specifically to pIgR and SC. The present study was carried out to show that the hexapeptide motif in SpsA is crucial for the interaction of pneumococci and pIgR-expressing cells. Methods: Streptococcus pneumoniae were cultured to mid-log phase. Calu-3 cells and MDCK epithelial cells, stably transfected with the hpIgR cDNA in pCB6 were used in in vitro infection experiments. Pneumococcal adherence to and invasion of epithelial cells were assayed. Results: By the use of the N-terminal domain of SpsA and SpsA201, which exhibits a single amino acid substitution in the pIgR-binding motif, in vitro assays indicated the association of the identified hexapeptide motif, located between amino acid 198 and 203 in SpsA, with pneumococcal adherence to and invasion of hpIgR-expressing cells. Interpretation & conclusion: The present findings demonstrated not only the crucial role of the hexapeptide of SpsA, not only for the SpsA-pIgR interaction, but also for adherence and invasion of hpIgR-expressing cells. Key words Choline binding protein - polymeric IgA receptor - Streptococcus pneumoniae Streptococcus pneumoniae is an important pathogen pIgA-binding extracellular region, a membrane spanning colonizing the upper and lower respiratory tract of domain and a cytoplamic C-terminal tail2. The receptor humans and capable to cause relatively harmless local is sorted to the basolateral pole of mucosal epithelial cells infections as well as life-threatening pneumonia and and binds there non-covalently to pIgA. The complex is meningitis1. Dissemination of pneumococci starts at the endocytosed into vesicles which are translocated to and nasopharynx after penetration of epithelial cells. fuse with the apical surface of the cells. A proteolytic Recently, the Cbp SpsA was identified as the bacterial cleavage of the extracellular portion of the pIgR, also adhesin of the pIgR. The latter is synthesized as an known as secretory component (SC), then results in integral membrane protein in the rough endoplasmic release of SC, either free or covalently coupled to pIgA, reticulum of mucosal epithelial cells and comprises a forming SIgA2,3. Among known Cbps, the SpsA protein 61 62 INDIAN J MED RES (SUPPL) MAY 2004 exhibits a unique interaction with free SC or the SC SIgA and free SC were purified as described earlier portion of pIgR and SIgA4. The pathogenic significance from milk from humans. Antiserum against SpsA was of this interaction is underlined by the fact that SpsA generated by Eurogentec (Sart-Tilman, Belgium). provides a mechanism for adherence to nasopharyngeal Rabbits were immunized subcutaneously with 50 µg of pIgR-expressing cells. In addition to the efficient purified SpsA-derivate encoded by pQSH2 in 1 ml of colonization, the interaction was also associated with 1:1 emulsion of buffer and complete Freund´s adjuvant translocation of pneumococci across an epithelial barrier and boosted with 50 µg in incomplete Freund´s adjuvant by adopting in a reverse way the pIg receptor transcytosis at days 14, 28, and 56. machinery5. In addition, a pneumococcal mutant with deficient expression of SpsA displayed a reduced ability Epithelial cell adherence and invasion assay: to colonize the nasopharynx of rats6 and showed a Pneumococcal adherence to and invasion of epithelial significant decrease of nasopharyngeal colonization in cells were assayed in 24-well plates (Greiner, Germany). 5 pIgR-KO mice . A hexapeptide motif YRNYPT was Confluent epithelial cells (2 x 105) were inoculated with identified as the minimal pIgR/SC-binding motif using 5 x 106 pneumococci and incubated in Dulbecco´s synthetic peptide technology7. The present study shows Minimal Medium-HEPES at 37°C and 5 per cent CO2 that this hexapeptide is crucial for the interaction of for 4 h. Subsequently, the cells were rinsed several times pneumococci with the pIgR, and for their in vitro with PBS to remove unbound bacteria. Numbers of adherence to and invasion of epithelial cells expressing adherent and invasive pneumococci were measured by the hpIgR. microscopy. Extracellular bacteria, adhering to epithelial cells, were incubated for 15 min with a rabbit anti- Material & Methods pneumococcal antiserum, followed by fixation in 3.7 per cent paraformaldehyde and visualized by FITC-labelled Bacterial strains and cell culture : S. pneumoniae goat anti-rabbit Ig (Dianova). After this, intracellular were cultured in Todd-Hewitt-broth (Oxoid, Basingstoke, (=invasive) pneumococci were stained with the same England) supplemented with 0.5 per cent yeast extract anti-pneumococcal antiserum followed by TRITC- (THY) to mid-log phase or grown on blood agar (Merck). labelled goat anti-rabbit Ig (Dianova). Adherence (green Calu-3 cells (human lung epithelium; ATCC HTB-55) bacteria) and invasion (red bacteria) were scored on at and MDCK (Madin-Darby Canine Kidney) epithelial least 50 cells. Each experiment was repeated at least cells, stably transfected with the hpIgR cDNA in pCB68, five times; results were expressed as mean ± standard were used in in vitro infection experiments. Calu-3 and deviation. MDCK-hpIgR cells were cultured in Eagle´s Minimum Essential Medium supplemented with 10 per cent foetal calf serum, 2 mM glutamine, penicillin G (100 IU/ml), Results and streptomycin (100 µg/ml; all from GIBCO BRL) at pIg receptor-binding domains of pneumococcal SpsA 37°C under 5 per cent CO2. The medium for Calu-3 cells was further supplemented with 1 mM sodium protein: Binding of human pIgR and SIgA to the surface- pyruvate and 0.1 mM non-essential amino acids. displayed Cbp SpsA is mediated by SC. Our previous studies had mapped the binding region to the N-terminal Proteins and antisera: Escherichia coli M15[pREP4] part of SpsA. A hexapeptide YRNYPT was identified (Qiagen, Hilden, Germany) was used as host for as the minimal SC-binding motif. The gene encoding the recombinant pQE expression plasmids and cultured at SpsA protein has a mosaic structure and two major clades 37°C on Luria-Bertani (LB) agar or grown on LB-agar of the protein exist9. The hexapeptide occurs twice in containing 100 µg/ml ampicillin. Expression of His-tagged SpsA of most bacterial serotypes, at positions 125 and fusion proteins was induced with 1 mM isopropyl-b-D- 287, whereas only one copy is present, at position 198 to 203, in SpsA of seroytpe 1 strains. For investigation of thiogalactopyranoside after the culture reached an OD600 of 0.8 and continued to grow at 30°C for 4 h. The His- the SpsA-pIgR interaction in in vitro assays, functional tagged fusion proteins were purified by chromatography domains of the SpsA protein were expressed as His- under native conditions on Ni-nitrilotriacetic acid resins tagged fusion proteins. The SpsA proteins selected were according to the manufacturer (Qiagen). expressed by pQSH2, which expressed the mature ELM et al : SPSA-HPIGR INTERACTION FOR PNEUMOCOCCAL ADHERENCE & INVASION 63 N-terminal part of SpsA from amino acids 37 to 283, and pQSM1, which expressed a functional domain of SpsA from amino acid 159 to 324 and includes the proline rich sequence (aa 284 to 324)7. Cloned sequences were derived from a spsA gene of a serotype one strain. Fig.1. pIgR/SC-binding activity of SpsA-derivates. His-tagged SpsA proteins were analysed by Western blotting for their ability to interact Therefore, in both SpsA-derivates the hexapeptide occurs with SC. only once (Fig. 1). To validate the function of the hexapeptide in in vitro assays, the tyrosine at position 201 was selected for aspartic acid substitution. The resulting protein encoded by pQSH2201was designated SpsA201. Immunoblot analysis indicated that binding of SC to SpsA201 is completely abolished (Fig. 1).

Expression of human pIgR by human epithelial cells: In order to investigate the role of the SpsA-pIgR interaction for pneumococcal invasion, the expression of the human pIgR (hpIgR) was investigated among different respiratory cell lines, i.e., A549, HEp-2, Calu- 3 cells, which were compared by immunoblot analysis. Fig. 2. Human pIgR expression of epithelial cell lines. Expression of MDCK cells exhibiting a stable expression of hpIgR the pIgR was analysed by immunoblotting cell lysates using a (MDCK-hpIgR) were used as a positive control cell line. polyclonal anti-SC antiserum generated in rabbits. Immunoblot analysis of cells with anti-SC antiserum showed production of the pIgR in the MDCK-hpIgR and Calu-3 cells, and not in A549 and HEp-2 cells (Fig. 2).

SpsA-dependent adherence to and invasion of hpIgR-expressing cells: Previous studies showed that Streptococcus pneumoniae is able to adhere to and invade eukaryotic cells by using different adhesin- receptor interactions. In order to demonstrate the effect of the SpsA-hpIgR interaction for the pathogenesis of S. pneumoniae, the adherence to and invasion of hpIgR- expressing lung epithelial Calu-3 and MDCK-pIgR cell lines was investigated for wildtype and spsA-mutant pneumococci. Numbers of adherent and invasive pneumococci were measured by microscopy. When using SpsA-KO pneumococci in infection experiments, the bacteria substantially lost their ability to adhere to and invade Calu-3 and MDCK-hpIgR cells (Fig. 3). These results indicate that pneumococcal invasion occurs in a hpIgR-dependent manner as shown for the human Detroit nasopharyngeal cells8. The ability of spsA- Fig.3. Inhibition of pneumococcal adherence to pIgR-expressing deficient pneumococci to adhere to Calu-3 cells is cells. Blocking of bacterial binding to Calu-3 and MDCK-hpIgR substantially reduced and the ability to invade these cells cells was performed using SpsA-derivates and polyclonal anti- SpsA antiserum. Adherence (a) and invasion (b) of S. pneumoniae is completely abolished. These data, therefore, suggest NCTC 10319 was measured by microscopy. Results outline the that expression of the hpIg receptor is sufficient to crucial role of the hexapeptide motif in SpsA for attachment and mediate both adherence and invasion of pneumococci, internalization. 64 INDIAN J MED RES (SUPPL) MAY 2004 and that the SpsA-hpIgR interaction is early involved in The role of the SpsA/CbpA-pIgR interaction for pneumococcal invasion. pneumococcal pathogenesis was addressed in in vivo and in vitro studies. The in vitro studies demonstrated Inhibition of pneumococcal adherence to pIgR- that the SpsA/CbpA-pIgR interaction provides a expressing cells: Previous studies and our data mechanism not only for colonization, but also for revealed that the SpsA-pIgR interaction provides a invasion. The interaction is associated with the common mechanism for internalization. However, the translocation of pneumococci across nasopharyngeal contribution of the identified SpsA hexapeptide motif cells by co-opting the receptor transcytosis in the infectious process was not shown. In order to machinery5, albeit in a reverse apical to basolateral investigate whether the identified hexapeptide pIgR- direction. binding motif(s) are crucial for adherence and invasion of pneumococci, in vitro infection assays were Our data now demonstrate the crucial role of the performed using hpIgR-expressing epithelial cells and hexapeptide of SpsA, not only for the SpsA-pIgR SpsA-derivates. Blocking experiments were conducted interaction, but also for adherence to and invasion of using the N-terminal His-tagged SpsA-derivate SH2, hpIgR-expressing cells. The functional domain of SpsA encoded by pQSH2, and the mutated SpsA201 with a blocked the pneumococcal invasion of hpIgR-expressing critical Tyr/Asp amino acid substitution at position 201. cells, whereas SpsA201, with a single amino acid The N-terminal part of SpsA inhibited adherence of substitution at position 201 within the hexapeptide motif, pneumococci to the cells. In addition, also anti-SpsA did not strongly affect the invasion by pneumococci. antibodies substantially reduced adherence and invasion However, the pIgR-dependent invasion of human of pneumococci. In contrast, the recombinant mutated respiratory epithelial cells by S. pneumoniae was SpsA201 protein did not reduce adherence and invasion claimed to be a limited phenomenon, that would be of pneumococci to Calu-3 cells and MDCK-hpIgR restricted in cell type- and bacterial strain-specific cells. manners11.

Discussion In vivo studies using a model of colonization demonstrated a significant loss of nasopharyngeal The surface of S. pneumoniae is decorated with at colonization in rats for the pneumococcal cbpA- least nine different Cbps. SpsA, also designated CbpA mutant6. A further study apparently observed a and PspC, is one of the best characterized significantly reduced ability of pneumococci to colonize pneumococcal surface proteins. The protein was pIgR-KO mice5. In contrast, our biochemical analysis independently identified by different groups. SpsA was indicated that the interaction between pIgR/SC and identified by its ability to bind free SC and SIgA4. CbpA pneumococcal CbpA/SpsA was restricted to human was identified as the most abundant Cbp in a pspA- pIgR/SC. Binding assays indicated that SpsA binds mutant strain and shown to be responsible for binding only to SC and SIgA of human origin, but not to SC of pneumococci to eukaryotic cells6. In addition to its and SIgA from rabbit, rats, and mice7. These data function as SC-binding protein, this protein was recently are further supported by the fact that human, but not reported to bind human complement Factor H10. The rabbit pIg receptor are able to promote bacterial entry gene encoding the bacterial adhesin for SC/hpIgR has in stably transfected MDCK cells5. The mechanisms a mosaic structure, and comparison of the alleles underlying these in vivo effects are, therefore, unclear revealed that modular domains have contributed to gene in the light of the species-specific binding diversity during evolution9. However, the minimal SC- demonstrated in vitro. Therefore, it will be important binding motif was localized to a highly conserved region to localize the ‘domain´ within pIgR which is within the two different clades of the adhesin. A interacting with the pneumococcal SpsA/CbpA. The hexapeptide YRNYPT was identified to mediate determination of this ‘domain´ could help to elucidate binding of pneumococci to SC in human SC and SIgA7. whether a genetic diversity of the binding site in pIgR This hexapeptide motif is present once in one of the molecules among various mammals is responsible for two clades and twice in the other clade. the observed effects. ELM et al : SPSA-HPIGR INTERACTIONFOR PNEUMOCOCCAL ADHERENCE & INVASION 65 References adherence, colonization and immunogenicity of Streptococcus pneumoniae. Mol Microbiol 1997; 25: 819-29. 1. Musher DM. Infections caused by Streptococcus pneumoniae: 7. Hammerschmidt S, Tillig MP, Wolff S, Vaerman JP, Chhatwal clinical spectrum, pathogenesis, immunity, and treatment. Clin GS. Species-specific binding of human secretory component to Infect Dis 1992; 14 : 801-7. SpsA protein of Streptococcus pneumoniae via a hexapeptide motif. Mol Microbiol 2000; 36 : 726-36. 2. Piskurich JF, Blanchard MH, Youngman KR, France JA, Kaetzel CS. Molecular cloning of the mouse polymeric Ig 8. Tamer CM, Lamm ME, Robinson JK, Piskurich JF, Kaetzel receptor. Functional regions of the molecule are conserved CS. Comparative studies of transcytosis and assembly of among five mammalian species. J Immunol 1995; 154 : secretory IgA in Madin-Darby canine kidney cells expressing 1735-47. human polymeric Ig receptor. J Immunol 1995; 155 : 707-14. 3. Mostov KE. Transepithelial transport of immunoglobulins. 9. Brooks-Walter A, Briles DE, Hollingshead SK. The pspC gene Annu Rev Immunol 1994 ; 12 : 63-84. of Streptococcus pneumoniae encodes a polymorphic protein, PspC, which elicits cross-reactive antibodies to PspA and 4. Hammerschmidt S, Talay SR, Brandtzaeg P, Chhatwal GS. provides immunity to pneumococcal bacteremia. Infect Immun SpsA, a novel pneumococcal surface protein with specific 1999; 67: 6533-42. binding to secretory Immunoglobulin A and secretory component. Mol Microbiol 1997; 25 : 1113-24. 10. Dave S, Brooks-Walter A, Pangburn MK, McDaniel LS. 5. Zhang JR, Mostov KE, Lamm ME, Nanno M, Shimida S, PspC, a pneumococcal surface protein, binds human factor H. Ohwaki M, et al. The polymeric immunoglobulin receptor Infect Immun 2001; 69 : 3435-7. translocates pneumococci across human nasopharyngeal 11. Brock SC, McGraw PA, Wright PF, Crowe JE Jr. The human epithelial cells. Cell 2000; 102 : 827-37. polymeric immunoglobulin receptor facilitates invasion of 6. Rosenow C, Ryan P, Weiser JN, Johnson S, Fontan P, Ortqvist epithelial cells by Streptococcus pneumoniae in a strain-specific A, et al. Contribution of novel choline-binding proteins to and cell type-specific manner. Infect Immun 2002; 70 : 5091-5.

Reprint requests: Dr Sven Hammerschmidt, Research Centre for Infectious Diseases, University of Wurzburg, D-97070, Wurzburg, Germany e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 66-73

Interaction of human factor H with PspC of Streptococcus pneumoniae

Sandhya Dave*, Michael K. Pangburn#, Corunda Pruitt* & Larry S. McDaniel*†‡

Departments of *Microbiology, †Surgery & ‡Medicine, The University of Mississippi Medical Center, Jackson, MS 39216, & #Department of Biochemistry, The University of Texas Health Science Center, Tyler, TX 75708, USA

Received August 6, 2003

Background & objectives: Streptococcus pneumoniae has acquired virulence factors such as the polysaccharide capsule and various surface proteins, which prevent opsonization mediated by the complement system. PspC is one of the multi-functional pneumococcal surface proteins capable of eliciting an antibody response in mice. Our study further explores the role of pneumococcal surface proteins in resistance to complement mediated opsonophagocytosis by providing evidence that PspC binds human Factor H (FH), a regulatory protein of the alternative complement pathway. The present study was carried out to map the binding regions on PspC and FH, and to assess the functional activity of FH upon binding to PspC. Methods: FH binding to D39 and other pneumococcal strains was observed by flow cytometry. A series of FH truncated and deletion mutants and PspC mutants were used to localize binding regions within these molecules. The functional activity of FH upon binding to PspC was measured by a haemolysis assay. Results: FH binding to D39 and not to TRE108 (PspC-) cells was demonstrated by flow cytometry. Pneumococcal isolates of 14 different strains varied in their ability to bind FH. The binding region of FH within PspC to the first 225 amino acids of the a-helical domain was localized. The corresponding binding site for PspC is located within the SCR 6-10 region of FH. Haemolysis of rabbit red blood cells was inhibited by FH even in the presence of PspC. Interpretation & conclusion: FH binding is specific to PspC on the pneumococcal cell surface. The binding region on PspC mapped to the non-conserved N-terminal region of the a-helical domain. The binding site on FH to PspC is different from the active site that functions in degradation of C3b. A haemolysis assay provided evidence that the functional activity of FH was maintained upon binding to PspC. Thus, binding of FH to PspC might be an important mechanism by which S. pneumoniae resist complement activation and opsonophagocytosis. Key words Complement - factor H - PspC - Streptococcus pneumoniae The alternative complement pathway functions as a rapid degradation of bound C3b on the host cell nonspecific defense system against microbial pathogens1. surface1,2. The pathway is initiated by deposition of C3b on the surface of the invading organism resulting in either lysis, FH is a 150-kDa single glycoprotein that is found in neutralization, or phagocytosis (Fig. 1). Host cells are the plasma at a concentration of about 400 mg/l3. The protected from the complement cascade by cell surface protein is composed of 20 short consensus repeat molecules and complement regulatory proteins. One domains (SCRs) each containing approximately 60 amino regulatory protein is Factor H (FH) which functions in acid residues that fold into b-strands and sheets1,2. The 66 DAVE et al: HF BINDING TO PSPC OF S. PNEUMONIAE 67 functional properties of FH depend on the interaction of the complement system by binding to either the C3 individual SCR domains with each other. Micro- protein or components on the surface of activated human organisms have evolved various mechanisms of escaping epithelial and endothelial cells12,16,17. The focus of our the alternative pathway by minimizing deposition of research is the interaction of FH with PspC. activated C3b on their surface2. One such strategy involves expression of receptors that bind FH to the In this study, we localized the binding sites on both microbial surface. M6 protein of group A streptococci, FH and PspC. The PspC segment that binds FH was YadA of Yersinia enterocolitica, Por1A of Neisseria localized to amino acids 1 to 225 in the a-helical domain gonorrhoeae, CRASP1 of Borrelia burgdorferi, and of PspC. The corresponding binding site for PspC was Hic of type 3 Streptococcus pneumoniae all bind located within the SCR 6-10 region of FH. The functional FH3-9. We recently demonstrated that PspC of activity of FH in the presence of PspC was also assessed S. pneumoniae also binds FH10. using a haemolysis assay.

PspC, also known as CbpA and SpsA11,12 , is similar Material & Methods in structure to another pneumococcal surface protein, PspA13,14. It consists of an N-terminal a-helical domain, Bacterial strains, growth conditions, and cell lysates: a proline-rich domain, and a choline-binding domain. The The pneumococcal strains used in this study for flow a-helical domain of PspC is larger (50 to 100%) than cytometry and Western blot analysis were cultured as the a-helical domain of most PspAs15. The PspC choline previously described10. The pneumococci were grown binding and proline-rich domains are over 90 per cent on blood agar at 37ºC in 5 per cent CO2 overnight and homologous to the corresponding domains of PspA14. transferred to Todd-Hewitt broth with 0.5 per cent yeast Due to structural similarity, it is possible that the antibodies extract. The bacteria were harvested at late mid-log directed to the proline-rich or choline binding domains phase by centrifugation and suspended in phosphate- cross-react with both PspA and PspC molecules14. buffered saline (PBS; pH 7.2) for flow cytometry. The bacterial concentration (approximately 1 x 108 cfu/ml) was estimated by spectrophotometer (A ) and PspC is a potential virulence factor of S. pneumoniae 600 that functions by interacting with various components of confirmed by viable counts on blood agar plates. For Western blot analysis, pneumococcal cell lysates were the immune system. PspC binds to the secretory prepared as described previously18 and stored at -20°C component of immunoglobulin A11. It may also inhibit until use. Escherichia coli Y1090 was grown in Luria- Bertani medium and used to prepare a control lysate.

Flow cytometry: FH purified from human serum was biotinylated as described previously10 using the EZ-Link sulfo-NHS-LC-biotinylation kit (Pierce, Rockford, Il). The bacteria in 100 µl of 1 per cent bovine serum albumin-PBS were incubated with 100 µl of biotinylated FH (1/30) for 1 h at 37°C. The bacteria were washed with 1 ml PBS three times and suspended in 100 µl of fluorescein isothiocyanate-streptavidin (Southern Biotechnology Associates, Birmingham, Ala) for 30 min at room temperature. The cells were washed as before, suspended in 2ml of PBS, and analyzed by FACScan Fig. 1. Activation of the alternative pathway of complement in cytometer (Beckton Dickinson, USA). response to a pathogen. The pathway is activated by spontaneous hydrolysis of C3 to C3b in the fluid-phase. Deposition of C3b on Cloning and expression: Chromosomal DNA was the target leads to the formation of C3bBb, also known as the C3 - 19 convertase. The C3 convertase functions to increase deposition of isolated from LM91 (PspA ), an isogenic strain of D39 C3b on the activated-target surface resulting in opsonization and by DNeasy Tissue Kit (Qiagen Inc., Valencia, CA). phagocytosis of the pathogen by macrophages and neutrophils. Oligonucleotide primers LSM370 (5’ C A C C C G C G 68 INDIAN J MED RES (SUPPL) MAY 2004 A C A G A G A A C G A G G G A A G T A C3') and Haemolysis assay: The inhibition of the alternative LSM371 (5’C T A G G A T G A G C T T G G A A G pathway mediated by FH was examined by a haemolysis A G T 3') were used in a polymerase chain reaction assay9,22. C2-deficient serum (1/12 dilution) (Advanced (PCR) to amplify a 700 bp fragment, which encodes Research Technology, San Diego, California) was used amino acids (aa) 1 to 255 of the a-helical region of PspC. as a source of complement and rabbit red blood cells The PCR fragment was cloned into TOPO TA cloning (RBCs) (5 x 108) (Advanced Research Technology, San Version L (Invitrogen, Carlsbad, CA) and sub-cloned Diego, California) were used as target cells. A dilution into ProEx C for expression (Life Technologies, of serum was used that resulted in 80 per cent lysis of Rockville, MD). The ProEx vector produces a fusion target cells following incubation at 37°C for 40 min. protein with a six-histidine tag at the N-terminus of the Subsequently, C2-deficient serum with or without expressed protein. Expression was induced by the addition recombinant purified PspC (30 mM), FH (15 µM), or a of 60 mM isopropyl-b-D-thiogalactopyranoside (IPTG) pre-incubation mixture of PspC and FH (2:1 molar ratio) to exponentially growing E. coli TOP10 cells. Two other in 200 µl of veronal-buffered saline (VBS) with 13.5 fragments of PspC (aa 1 to 445 and aa 255 to 445) from mM EGTA, Mg2+ and 0.1 per cent gelatine was D39 (kindly provided by A. Brooks-Walter) were also incubated for 5 min at 37°C. The rabbit RBCs (50 µl) expressed as fusion proteins with a six-histidine tag in were added to the incubation mixture. The reaction was the pET20b expression vector (Novagen, Madison, Wis.) stopped with 750 µl of cold VBS, 10 mM EDTA at 5, as described previously14. The His-tagged proteins were column purified, and the purity of the proteins was 10, 20, and 40 min intervals. After centrifugation at 3000 confirmed by SDS-polyacrylamide gel electrophoresis. rpm for 5 min, the supernatants were removed to measure the amount of haemoglobin released by spectrophotometer at 412 nm. Western analysis: Western blot analyses were performed as previously described10. Purified PspC fragments and D39 lysate were separated on 4-20 per cent tris-glycine pre-cast gels (BioWhittaker Molecular Applications, Rockland, ME) and transferred to nitrocellulose membrane. The membranes were incubated with either biotinylated FH or polyclonal anti-PspC antibody14. Bound antibodies and FH were detected using strepavidin-conjugated horseradish peroxidase and a chemiluminescent substrate (Pierce, Rockford, Il).

Ligand dot blot: For ligand dot blotting, a procedure similar to the one described for interaction of FH with M protein20 was used. Briefly, 5 µg of purified PspC protein (aa 1 to 445), albumin, and E. coli lysate10 were dried onto a nitrocellulose membrane (Millipore Corporation, Bedford, MA ) at 37ºC for 30 min. After blocking in 1 per cent BSA-PBS overnight at 4ºC, the membranes were incubated with 0.3 µg of full-length FH or recombinant FH representing specific deletions of the twenty SCR domains that were cloned and expressed as previously described21 for 1 h at room temperature. The membrane was further incubated with Fluorescence intensity goat anti-FH antibody (Quidel, Mountain View, CA) Fig. 2. Binding of FH to exponentially growing capsular type 2 followed by rabbit anti-goat-biotinylated antibody S. pneumoniae (a) TRE108 (PspC-) compared to (b) D39 (PspC+) (Southern Biotechnology Associates, Birmingham, Ala) by flow cytometry. The bacteria were incubated with or without biotinylated FH and stained with fluorescein isothiocyanate- before detection with streptavidin-conjugated horseradish streptavidin. FH binding was measured by FACScan cytometer as peroxidase. Bound FH was visualized using a change in fluorescent intensity in comparison to the cells incubated chemiluminescent substrate (Pierce, Rockford, Il). without FH. DAVE et al: HF BINDING TO PSPC OF S. PNEUMONIAE 69

Fig. 3. FH binding to exponentially growing pneumococci of different capsular types by flow cytometry. This graph is a representation of the 14 different strains tested. Data expressed as the geometric mean of fluorescent intensity from at least three experiments. Results 1234 Binding of human FH to exponentially growing pneumococci: S. pneumoniae cells of various isolates were incubated with human FH, suspended in FITC- conjugated streptavidin, and examined by FACScan cytometer. Change in mean fluorescence intensity was observed upon incubation of wild-type D39 cells with FH (Fig. 2b). TRE 108 cells, an isogenic mutant of D39 that does not express PspC, did not bind FH (Fig. 2a). Variations in FH binding among the different pneumococcal isolates were also observed by flow cytometry (Fig. 3). These results were consistent with previous data obtained by Western analysis10 and immunofluorescence microscopy (unpublished data).

Mapping of the FH binding region on PspC: Western blotting was used to determine the region of PspC, which binds FH. Purified recombinant PspC proteins expressing either full-length or portions of the a-helix domain from D39 or its isogenic strain LM91 were FH incubated with biotinylated FH. Cell lysates from strain Fig. 4. Localization of FH binding site on PspC by Western blot D39 and E. coli were used as positive and negative analysis. Purified, recombinant, truncated PspC fragments and D39 controls, respectively. FH bound to the PspC (aa 1 to lysate were separated by gel electrophoresis, transferred to a 445), and PspC (aa 1 to 225) fragments from D39 and nitrocellulose membrane, and incubated with biotinylated FH. Lane LM91, respectively (Fig.4). PspC (aa 255 to 445) 1, PspC fragment lacking the amino acids 1-255 in the a-helical domain; lane 2, PspC fragment consisting of amino acids 1-445 of fragment from D39 lacked this binding activity. These the a-helical and proline-rich domains; lane 3, D39 lysate used as a results indicated that the FH binding region was located positive control; and lane 4, PspC fragment with 1-225 amino acids between amino acids 1 to 225 of PspC. of the a-helical domain. 70 INDIAN J MED RES (SUPPL.) 2004 Functional activity of FH in the presence of PspC: We examined the functional activity of FH in the presence of PspC using a haemolysis assay with C2 deficient serum and rabbit RBCs. The concentration of C2- deficient serum needed for approximately 80 per cent haemolysis of rabbit RBCs after 40 min was determined by preliminary experiments. Following incubation, 75.0 ± 7.64 per cent of rabbit RBCs were lysed by the C2-deficient serum. We observed 69.3 ± 8.9 per cent haemolysis of rabbit RBCs by the C2-deficient serum in the presence of PspC. As expected, the lysis was greatly inhibited (only 15 ± 0.58%) upon addition of purified FH to the C2-deficient serum (Table). Moreover, the inhibition of haemolysis of rabbit RBCs was maintained upon addition of a pre-incubated mixture of purified FH and PspC to the C2-deficient serum (12.0 ± 0.58%) suggesting that PspC bound FH is functional. Fig. 5. Mapping the PspC binding region on FH by ligand dot blot analysis. Recombinant PspC protein was adsorbed onto Discussion nitrocellulose membranes. E.coli lysate and albumin were used as negative controls. The membranes were incubated with either full- length FH or FH with specific deletions in the 20 SCR domains and PspC is a potential virulence factor of S. pneumoniae detected with anti-FH antibody. that functions by interacting with various components of the immune system. Recently, we showed that PspC binds Localization of the FH binding region to PspC: to the major regulatory protein of the alternative pathway The localization of PspC binding site on FH was of complement, FH10. In this investigation, we further 20 determined by a ligand dot blotting assay . PspC confirmed the binding of FH to exponentially growing D39 immobilized on a nitrocellulose membrane was and other pneumococcal strains by flow cytometry. We incubated with either full-length FH or recombinant also mapped the binding sites on both PspC and FH with truncated proteins with specific deletions in the SCR recombinant truncated mutant proteins. Moreover, we domains. BSA and E. coli lysate were used as provided evidence that FH activity is maintained in the negative controls (Fig. 5). FH and all recombinant presence of PspC in a haemolysis assay. The interaction proteins containing SCR 1-10 bound to purified PspC. of PspC with FH might play an important role in prevention The FH deletion mutant protein lacking SCR 6-10 of opsonization by minimizing C3b deposition on the (rFH D6-10) failed to bind PspC. The results pneumococcal surface. demonstrated that the PspC binding site for FH was located inclusively between the SCR 6 and SCR 10 FH bound to the surface of D39 cells whereas the domains of this complement regulatory protein. isogenic mutant strain of D39, TRE108 (PspC-) did not

Table. Haemolysis of rabbit red blood cells by alternative pathway of complement in the presence of FH and PspC

rRBCa Time (min) 5 10 20 40

Serum 14.00 ± 0.00 19.33 ± 0.67 41.33 ± 5.84 75.00 ± 7.64 Serum and PspC 13.67 ± 0.88 19.33 ± 0.67 33.33 ± 3.53 69.33 ± 8.97 Serum and FH 12.50 ± 0.76 10.00 ± 1.00 12.00 ± 1.00 15.00 ± 0.58 Serum and PspC-FH 12.53 ± 0.78 9.50 ± 0.87 11.00 ± 0.00 12.00 ± 0.58

a Data expressed as geometric mean of per cent haemolysis of rabbit red blood cells ± standard error of the mean from three experiments DAVE et al: HF BINDING TO PSPC OF S. PNEUMONIAE 71 bind FH. This suggests that FH binding is specific to according to variations in the amino acid sequence. The PspC on the surface of live D39 cells. As mentioned C-terminal end of the PspC variants contained either before, PspC shares over 90 per cent homology with the choline-binding domain or the LPXTG motif common PspA in the proline and choline binding domains. Since to many Gram positive bacteria. This ability to express TRE108 expresses PspA but not PspC, FH failing to and use two different anchorage domains may be unique bind to this strain further implies that the binding region to PspC of S. pneumoniae. The LPXTG motif was is present in the a-helical domain of PspC that does not present in PspC of type 3 pneumococcus and in the share homology with PspA. It has been reported that second protein of strains with two alleles of PspC. the interaction between FH and PspC maybe hydrophobic since binding was not inhibited in the presence of The binding of FH to exponentially growing increasing concentrations of salt23. pneumococci varied among the 14 different isolates tested by flow cytometry. These observations were similar to PspC is found on approximately 75 per cent of previous studies10 in that there was no correlation between pneumococcal strains14. Recently, the polymorphic pspC the intensity of FH binding and clade type of PspC or locus of 43 different pneumococcal strains was pneumococcal capsular type. It is well established that characterized by DNA sequencing of PCR fragments24. FH binds to a PspC variant, Hic, that contains an LPXTG The majority (approximately 86%) of strains tested motif in type 3 pneumococcus9. However, it is not clear contained single copy while some strains (14%) had two whether differences in FH binding are due to the highly tandem copies of the PspC gene. The DNA sequence variable nature of PspC among the different strains or of PspC varied in each of the 43 strains of pneumococci. presence of other FH binding molecules on the The PspC proteins were placed in 11 different groups pneumococcal surface.

Fig. 6. Model of the mechanisms of inactivation of the alternative pathway by FH. (A) A diagramme of FH showing the 20 SCR domains along with the three C3b (shaded regions) and the PspC binding sites. The active site on FH involved in degrading C3b or the C3b convertase is contained within the first four SCR domains. Note that the binding region for PspC is different than the active site for complement regulatory function. (B) Three different mechanisms by which FH has been shown to decrease the amount of C3b deposited on host cell surface. B1: FH binds to C3b on host cell and serves as cofactor for cleavage of C3b by FI. B2: FH functions in dissociation of Bb from C3bBb (C3 convertase) upon binding to C3b. B3: The presence of molecules such as sialic acid can result in preferential binding of FH than B to C3b on host cell surface. The effect on the activation of the alternative pathway may be similar upon binding of FH to PspC on the pneumococcal cell surface. 72 INDIAN J MED RES (SUPPL.) 2004 To further study this interaction, we mapped the control C3b degradation on or near the bacterial surface binding sites on FH and PspC. The FH binding was upon attachment since the binding site to the bacteria localized to amino acids 1 to 225 of PspC. This region (SCR 5-7) and active site for C3b inhibition (SCR 1-4) was located within the large a-helical domain of the are in separate regions within the molecule. molecule. According to sequence analysis, variations in PspC have been attributed to the a-helical domain. The The spontaneous hydrolysis of C3b in the fluid-phase first 100 to 150 amino acids of a-helical domain are of serum can also result in deposition of C3b on host cell hypervariable in both size and sequence. The surfaces. However, progression of the complement hypervariable region is followed by the first set of direct cascade is prevented on host cells at early stages of repeats ranging from 101 to 205 amino acids depending activation by complement regulatory molecules on the on the pneumococcal strain. Smaller amino acid repeats cell or in the serum. The three mechanisms of found in some strains are due to the lack of a sequence complement inactivation on host surface by FH are at the N-terminal end found in many strains with larger presented in Fig.6B. FH regulates the alternative amino acid repeats. This region of the molecule is highly complement pathway by serving as a cofactor for charged consisting of either lysine and glutamic acid cleavage of C3b by factor I, dissociating the C3bBb residues14. The binding of FH is also localized within the complex, or competing with factor B for binding to 1 first 261 amino acids of Hic in type 3 pneumococci9. C3b . The decrease in deposition of C3b on the cell 1 However, Hic did not inhibit binding of FH to PspC surface leads to inactivation of the alternative pathway . suggesting that both molecules are binding to unique sites on FH23. The binding site for the secretory component The binding of FH to PspC may be one of the of IgA is within the hexapeptide YRNYPT between mechanisms by which C3b deposition is also controlled amino acids 198 and 203 on PspC25. It will be important by pneumococcus to prevent complement attack. The to determine if FH or secretory IgA interferes with the mechanisms for C3b degradation on the pneumococcal binding of either molecule to PspC. surface may be similar to the host upon binding of FH to PspC (Fig. 6B). Thus, the interaction of FH with PspC may be an important mechanism by which pneumococci We localized PspC binding site on FH to SCR 6-10 escape opsonophagocytosis. It may also further explain domains. Other studies mapped the binding site on FH the variations in virulence among the different to be within SCR 8-15 for PspC and SCR 8-11 for Hic pneumococcal strains. of S. pneumoniae23,26. These findings are similar to studies involving interactions of FH and FHL-1/reconnectin with M protein of the group A streptococcus. The M Acknowledgment protein binding site in FH was first mapped within SCR 6- 1027. The binding of M protein since then is precisely Authors acknowledge Dr Edwin Swiatlo for his suggestions and 20 discussions, Dr Alexis Brooks-Walter for providing various mutant localized to SCR 7 of human FH . The complement strains of Streptococcus pneumoniae, Miss Nan Harvey for her resistant strains of Borreliae also consist of two surface assistance in flow cytometry and Alison Etheridge for purification proteins, CRASP-1 and CRASP-2, that bind to SCR 5-7 of PspC. This study was supported by NIH grant AI43653. of human FH and FHL-1/reconnectin, respectively8. References There are three sites on FH that bind C3b (Fig. 6A). Complement regulatory activity due to C3b binding is 1. Morgan BP, Harris CL, editors. Complement regulatory proteins. directly attributed to only the first four SCR domains of New York: Harcourt Brace and Company; 1999. FH. Thus, the binding of FH to PspC within SCR 6-10 2. Pangburn MK. Host recognition and target differentiation by implies that the active site is still available to bind C3b factor H, a regulator of the alternative pathway of complement. and inhibit complement activation. Immunopharmacology 2000; 49 : 149-57. 3. Kotarsky H, Hellwage J, Johnsson E, Skerka C, Svensson HG, Our data from the haemolysis assay provided evidence Lindahl G, et al. Identification of a domain in human factor H that FH was active upon binding to PspC. The regulatory and factor H-like protein-1 required for the interaction with activity of FH was also maintained after binding to Hic streptococcal M proteins. J Immunol 1998; 160 : 3349-54. of type 3 pneumococci, Yad A of Y. enterocolitica, and 4. Roggenkamp A, Ruckdeschel K, Leitritz L, Schmitt R, CRASP-1 of B. burgdorferi9,28,29. These proteins can Heesemann J. Deletion of amino acids 29 to 81 in adhesion DAVE et al: HF BINDING TO PSPC OF S. PNEUMONIAE 73

protein YadA of Yersinia enterocolitica serotype O:8 results in 17. Smith BL, Hostetter MK. C3 as substrate for adhesion of selective abrogation of adherence to neutrophils. Infect Immun Streptococcus pneumoniae. J Infect Dis 2000; 182: 497-508. 1996; 64: 2506-14. 18. McDaniel LS, Sheffield JS, Delucchi P, Briles DE. PspA, a 5. Ram S, Mackinnon FG, Gulati S, McQuillen DP, Vogel U, surface protein of Streptococcus pneumoniae, is capable of Frosch M, et al. The contrasting mechanisms of serum resistance eliciting protection against pneumococci of more than one of Neisseria gonorrhoeae and group B Neisseria meningitidis. capsular type. Infect Immun 1991; 59: 222-8. Mol Immunol 1999; 36 : 915-28. 19. McDaniel LS, Yother J, Vijayakumar M, McGarry L, Guild 6. Ram S, Mcquillen DP, Gulati S, Elkins C, Pangburn MK, Rice WR, Briles DE. Use of insertional inactivation to facilitate studies PA. Binding of complement factor H to loop 5 of porin protein of biological properties of pneumococcal surface protein A 1A: a molecular mechanism of serum resistance of nonsialylated (PspA). J Exp Med 1987; 165: 381-94. Neisseria gonorrhoeae. J Exp Med 1998; 188 : 671-80. 20. Blackmore TK, Fischetti VA, Sadlon TA, Ward HM, Gordon 7. Kraiczy P, Skerka C, Kirschfink M, Zipfel PF, Brade V. DL. M protein of the Group A streptococcus binds to the Mechanism of complement resistance of pathogenic Borrelia seventh short consensus repeat of human complement factor H. burgdorferi isolates. Int Immunopharmacol 2001; 1 : Infect Immun 1998; 66 : 1427-31. 393-401. 21. Pangburn MK, Pangburn KL, Koistinen V, Meri S, Sharma AK. 8. Kraiczy P, Skerka C, Kirschfink M, Brade V, Zipfel PF. Immune Molecular mechanisms of target recognition in an innate immune evasion of Borrelia burgdorferi by acquisition of human system: interactions among factor H, C3b, and target in the complement regulators FHL-1/reconectin and Factor H. Eur J alternative pathway of human complement. J Immunol 2000; Immunol 2001; 31 : 1674-84. 164 : 4742-51. 9. Janulczyk R, Iannelli F, Sjoholm AG, Pozzi G, Bjorck L. Hic, 22. Nydegger UE, Fearon DT, Austen KF. The modulation of the a novel surface protein of Streptococcus pneumoniae that alternative pathway of complement in C2-deficient human serum interferes with complement function. J Biol Chem 2000; 275: by changes in concentration of the component and control 37257- 63. proteins. J Immunol 1978; 120 : 1404-8. 10. Dave S, Brooks-Walter A, Pangburn MK, McDaniel LS. PspC, 23. Duthy TG, Ormsby RJ, Giannakis E, Ogunniyi AD, Stroeher a pneumococcal surface protein, binds human factor H. Infect UH, Paton JC, et al. The human complement regulator factor H Immun 2001; 69: 3435-7. binds pneumococcal surface protein PspC via short consensus 11. Hammerschmidt S, Talay SR, Brandtzaeg P, Chhatwal GS. repeats 13 to 15. Infect Immun 2002; 70 : 5604-11. SpsA, a novel pneumococcal surface protein with specific 24. Iannelli F, Oggioni MR, Pozzi G. Allelic variation in the highly binding to secretory immunoglobulin A and secretory component. polymorphic locus pspC of Streptococcus pneumoniae. Gene Mol Microbiol 1997; 25: 1113-24. 2002; 284 : 63-71. 12. Rosenow C, Ryan P, Weiser JN, Johnson S, Fontan P, Ortqvist 25. Hammerschmidt S, Tillig MP, Wolff S, Vaerman JP, A, et al. Contribution of novel choline-binding proteins to Chhatwal GS. Species-specific binding of human adherence, colonization, and immunogenicity of Streptococcus secretory component to SpsA protein of Streptococcus pneumoniae. Mol Microbiol 1997; 25: 819-29. pneumoniae via a hexapeptide motif. Mol Microbiol 2000; 13. McDaniel LS, McDaniel DO, Hollingshead SK, Briles DE. 36 : 726-36. Comparision of the PspA sequence from Streptococcus pneumoniae EF5668 to the previously identified PspA sequence 26. Jarva H, Janulczyk R, Hellwage J, Zipfel PF, Bjorck L, Meri S. from strain Rx1 and ability of PspA from EF5668 to elicit Streptococccus pneumoniae evades complement attack and protection against pneumococci of different capsular types. opsonophagocytosis by expressing the pspC locus-encoded Infect Immun 1998; 66: 4748-54. Hic protein that binds to short consensus repeats 8-11 of Factor H. J Immunol 2002; 168 : 1886-94. 14. Brooks-Walter A, Briles DE, Hollingshead SK. The pspC gene of Streptococcus pneumoniae encodes a polymorphic protein, 27. Sharma AK, Pangburn MK. Localization by site-directed PspC, which elicits cross-reactive antibodies to PspA and mutagenesis of the site in human complement factor H that provides immunity to pneumococcal bacteremia. Infect Immun binds to Streptococcus pyogenes M protein. Infect Immun 1997; 1999; 67: 6533-42. 65 : 484-7. 15. Briles L, Hollingshead SK, Swiatlo E, Brooks-Walter A, Szalai 28. China B, Sory MP, N’Guyen BT, De Bruyere M, Cornelis GR. A, Virolainen A, et al. Pneumococcal proteins PspA and PspC: Role of the YadA Protein in prevention of opsonization of Their potential for use as vaccines. In : Tomasz A, editor. Yersinia enterocolitica by C3b molecules. Infect Immun 1993; Streptococcus pneumoniae: molecular biology & mechanisms 61 : 3129-36. of disease. New York: Mary Ann Liebert; 2000 p. 253-60. 29. Alitalo A, Meri T, Ramo L, Jokiranta TS, Heikkila T, Seppala 16. Gordon DL, Rice J, Finlay-Jones JJ, McDonald PJ, Hostetter IJT, et al. Complement evasion by Borrelia burgdorferi: serum- MK. Analysis of C3 deposition and degradation on bacterial resistant strains promote C3b inactivation. Infect Immun 2001; surfaces after opsonization. J Infect Dis 1988; 157: 697-704. 69 : 3685-91.

Reprint requests: Dr Larry S. McDaniel,The University of Mississippi Medical Center, 2500 North State Street, Jackson, MS 39216 e-mail: [email protected]

Indian J Med Res 119 (Suppl) May 2004, pp 74-76

Retraction

New antimicrobial peptide active against Gram-positive pathogens

A. Jasir, F. Kasprzykowski*, V. Lindström, C. Schalén** & A. Grubb

Department of Clinical Chemistry, University Hospital, S-22185 Lund, Sweden, *Department of Chemistry University of Gdansk, Gdansk, Poland & **Department of Medical Microbiology Dermatology & Infection University Hospital, S-22185 Lund, Sweden

Received August 6, 2003

Background & objectives: Human and animal cystatins have been shown to inhibit the replication of certain viruses and bacteria, though it is not directly demonstrated that the effects are due to protease inhibitory capacity of the cystatins. We report antibacterial properties of a novel antimicrobial peptidyl derivative, (2S)-2-(Naaa-benzyloxycarbonyl-arginyl–leucylamido)–1–[(E)–cinnamoylamido]-3- methylbutane, structurally based upon the aminoterminal segment of the inhibitory centre of the human cysteine protease inhibitor, cystatin C. Methods: Clinical isolates of group A, B, C and G streptococci were collected. The antibacterial activity of Cystapep 1 derivative was tested by agar well diffusion method. Results: Cystapep 1, displayed antibacterial activity against several clinically important Gram-positive bacteria. It displayed minimal inhibitory and bactericidal concentrations of about 16 µg/ml for both Staphylococcus aureus and Streptococcus pyogenes. In radial agar diffusion assays, groups A, B, C and G streptococci as well as staphylococci were generally susceptible to the action of Cystapep 1, whereas pneumococci and enterococci were less susceptible. No activity against Gram-negative bacteria was observed. Interpretation & conclusion: Cystapep 1 also showed high activity against methicillin-resistant Staph. aureus (MRSA) and multi-antibiotic resistant coagulase negative staphylococci (CNS), suggesting its mechanism of action to be different from most currently used antibiotics. Key words Antimicrobial peptide - cystatin C - Staphylococcus aureus - Streptococcus pyogenes

A new group of potential antibacterial agents was The inhibitory centre of cystatin C comprises three described in 1989 when an oligopeptidyl derivative, peptide segments, Arg8-Leu9-Val10-Gly11, Gln55-Ile56- N-benzyloxycarbonyl-leucyl-valyl-glycyl-diazomethane Val 57- Ala58-Gly59 and Pro105-Trp106 6,7. This peptidyl (Z-Leu-Val-Gly-DAM), structurally based upon the derivative is based upon the aminoterminal segment of inhibitory centre of human cystatin C, was found to the inhibitory centre. While retaining capacity to inhibit suppress the growth of Streptococcus pyogenes1. The cysteine proteases, it shows a narrow antibacterial recently described cystatin superfamily of proteins spectrum by selectively inhibiting the growth of the comprises both eucaryotic and procaryotic cysteine cysteine protease producing bacterial species protease inhibitors2. Indeed, human cystatins C, D and Streptococcus pyogenes8. Subsequent analysis of a S, rat cystatins A and S, chicken cystatin and number of related compounds, however, showed that oryzacystatin have been described to inhibit the replication their antimicrobial effect was probably not ascribable to of certain viruses and bacteria3 although it has not yet any protease inhibition8. In addition, the antimicrobial been directly demonstrated that these effects are due to spectrum of both linear and cyclic compounds within the protease inhibitory capacity of the cystatins4,5. this class of compounds was found to differ from that of 74

JASIR et al : NEW ANTIMICROBIAL PEPTIDYL DERIVATIVE 75

Z-Leu-Val-Gly-DAM by comprising several Gram- most examined Gram-positive species showed good positive pathogens. For one of the most promising linear susceptibility (Table). Thus, all strains of S. aureus and derivatives, here named Cystapep 1, comparatively low CNS, including those being methicillin- or multi-resistant, MIC and MBC values for both S. pyogenes and were susceptible to Cystapep 1. Also b-haemolytic S. aureus were recorded8, and a strong mouse protective streptococci, including macrolide and tetracycline capacity of some of the compounds against lethal resistant strains, turned out susceptible; however, the streptococcal infections was noted8. In the present work inhibitory effect on group A was stronger than on group we studied the antimicrobial effect of Cystapep 1 against B, C and G strains. Pneumococci were slightly less a larger collection of clinically important pathogens, inhibited by Cystapep 1. Out of a-haemolytic including strains resistant to currently used antibiotics. streptococci both susceptible and resistant strains were found. E. faecalis strains were non-susceptible. In Material & Methods contrast, L. monocytogenes strains were clearly susceptible to Cystapep 1. Clinical isolates of group A, B, C and G streptococci,

S. aureus, CNS, Enterococcus faecalis, E. faecium, By broth dilution, two strains each of GAS and Streptococcus oralis, S. pneumoniae, Listeria S.aureus showed MBC and MIC of 16 mg/l, whereas monocytogenes, M. catarrhalis and Escherichia coli single strains of GBS, GCS, and GGS had MBC/MIC of 32 mg/l for Cystapep 1. were obtained from our Clinical Microbiology Department. The antibacterial activity of Cystapep1, 8 synthesized as described , was tested by agar well Discussion diffusion on PDM II agar. Holes of 5 mm diameter were punched and 50 µl of a solution of Cystapep 1 (1.0 g/L) Cystapep 1 was as effective against various antibiotic in DMSO was applied in each hole. The antibacterial resistant staphylococci and streptococci as against effect was classified arbitrarily based upon inhibition zone antibiotic susceptible strains of these species. Presently, diameters. resistant staphylococci represent leading agents in nosocomial and biomaterial-associated infections posing Results significant therapeutic problems due to shortage of effective antibacterial agents. In addition, the susceptibility A large number of Gram-negative isolates were found of b-haemolytic streptococci to Cystapep 1 may prove to be resistant to the action of Cystapep 1. In contrast, useful due to treatment problems both for invasive and superficially located infections. Table. Susceptibility of various Gram-positive species to Cystapep 1 Species (no. strains) Inhibition zone diameter (mm) The activity of Cystapep 1 appears limited to Gram- <7 8-11 12-15 >16 positive bacteria, a selectivity in its anti-bacterial spectrum which may be advantageous from an MSSA (57) 0 0 29 28 ecological point-of-view. The compound bears little MRSA (97) 0 0 1 96 resemblance to those of previously reported naturally CNS (56) 0 0 16 40 occurring antimicrobial peptides, e.g., defensins, since GAS (63) 0 0 0 63 Cystapep 1 is much smaller than these, contains extensively modified amino acid residues, and mimics GBS (33) 0 0 11 22 the active centre of a human major protease inhibitor, GCS (25) 0 0 21 4 though not retaining any protease inhibitory properties8. GGS (25) 0 0 22 3 Its mode of action therefore may differ from that of S. pneumoniae (34) 0 33 6 0 known membrane pore forming peptides but is still elusive. Chemical modification of Cystapep 1 in order a - streptococci (64) 15 4 33 12 to allow preparation of stable water solutions of it will L. monocytogenes (11) 0 0 11 0 probably be needed for its development into a clinically Total 15 32 150 268 useful anti-bacterial drug.

76 INDIAN J MED RES (SUPPL) MAY 2004

Acknowledgment 4. Aoki H, Akaike T, Abe K, Kuroda M, Arai S, Okamura R, et al. Antiviral effect of oryzacystatin, a proteinase inhibitor in rice, against herpes simplex virus type 1 in vitro and in vivo. This work was supported by grants from the Swedish Medical Antimicrob Agents Chemother 1995; 39 : 846-9. Research Council (Project 05196), the Medical Faculty of the University of Lund, A. Påhlsson’s, A. Österlund’s and G. and J. 5. Travis J, Potempa J, Maeda H. Are bacterial proteinases Kock’s Foundations. pathogenic factors? Trends Microbiol 1995; 3: 405-7. 6. Abrahamson M, Ritonja A, Brown MA, Grubb A, Machleidt References W, Barrett AJ. Identification of the probable inhibitory reactive sites of the cysteine proteinase inhibitors human cystatin C and chicken cystatin. J Biol Chem 1987; 262 : 9688-94. 1. Björck L, Åkesson P, Bohus M, Trojnar J, Abrahamson M, Olafssoni, et al. Bacterial growth blocked by a synthetic peptide 7. Bode W, Engh R, Musil D, Thiela U, Huber R, based on the structure of a human proteinase inhibitor. Nature Karshikov A, et al. The 2.0 A X-ray crystal structure of chicken 1989; 337 : 385-6. egg white cystatin and its possible mode of interaction with cysteine proteinases. EMBO J 1988; 7 : 2593-9. 2. Rawlings ND, Barrett AJ. Evolution of proteins of the cystatin superfamily. J Mol Evol 1990 ; 30 : 60-71. 8. Kasprzykowski F, Schalén C, Kasprzykowska R, Jastrzebska B, Grubb A. Synthesis and antibacterial properties of peptidyl 3. Collins AR, Grubb A. Cystatin D, a natural salivary cysteine derivatives and cyclopeptides structurally based upon the protease inhibitor, inhibits coronavirus replication at its inhibitory centre of human cystatin C. Dissociation of physiologic concentration. Oral Microbiol Immunol 1998; 13 : antiproteolytic and antibacterial effects. APMIS 2000; 108 : 59-61. 473-81. Reprint requests: Dr A. Jasir, Department of Clinical Chemistry, University Hospital, S-22185 Lund, Sweden e-mail: [email protected]

Indian J Med Res 119 (Suppl) May 2004, pp 77-79

Aminoglycoside resistance in enterococci isolated from paediatric septicaemia in a tertiary care hospital in north India

V.S. Randhawa, L. Kapoor, V. Singh* & G. Mehta

Department of Microbiology, Lady Hardinge Medical College & *Department of Paediatrics, Kalawati Saran Hospital, New Delhi, India

Received August 6, 2003

Background & objectives: Enterococci are important nosocomial agents and serious infections caused by them are often treated with a combination of cell wall inhibitor and aminoglycoside. However, the presence of high level aminoglycoside resistance in these isolates makes this treatment combination ineffective. The prevalence of such isolates in a tertiary care set up has important diagnostic and therapeutic implications. The present study was carried out to find out the occurrence of high level aminoglycoside resistant isolates of enterococci in paediatric septicaemia cases in a tertiary care set up in north India. Methods: Blood of paediatric cases with a clinical diagnosis of septicaemia was cultured to isolate and identify enterococci. Agar screen method was used to detect high level streptomycin and gentamicin resistance in these isolates. Vancomycin susceptibility of these isolates were determined as per the National Committee for Clinical Laboratory Standards (NCCLS) recommendations. Results: Fifty one enterococal strains were isolated from 21 neonates, nine infants and 21 children with a clinical diagnosis of septicaemia. Sixty eight per cent of these isolates had high level gentamicin resistance and forty three per cent had high level streptomycin resistance. All the isolates with high level streptomycin resistance also had high level gentamicin resistance. More than ninety five per cent of these isolates were sensitive to vancomycin. Interpretation & conclusion: The occurrence of high level gentamicin and high level streptomycin resistance in enterococcal isolates in our set up was high. This would require routine testing of the enterococcal isolates for high level aminoglycoside resistance. Alternative treatment regimes need to be sought, besides prudent use of antibiotics.

Key words Enterococci - high level aminoglycoside resistance - tertiary care hospital

Over the years, enterococci have become combination of cell wall inhibitors as penicillin, ampicillin increasingly resistant to antibiotics in terms of both the or vancomycin with aminoglycosides such as multiplicity of resistance and the level of resistance to streptomycin or gentamicin4. The addition of cell wall particular drugs1, and have emerged as important inhibitor agent helps in the penetration of the nosocomial agents due to their colonizing ability and multi aminoglycoside into the bacterial cytoplasm making the drug resistance2,3. intrinsically resistant organism as aminoglycoside sensitive. The presence of high level aminoglycoside A common regime for treatment of serious resistance (HLAR) in enterococci defined as minimum enterococcal infections such as septicaemia is the inhibitory concentration of aminoglycoside for the isolate

77 78 INDIAN J MED RES (SUPPL) MAY 2004

>2000 µg/ml, makes the synergism of cell wall inhibitor Table. Distribution of high level gentamicin and high level and aminoglycoside ineffective5. Among the many streptomycin resistance among the enterococcal isolates (n=51) mechanisms for the resistance in these isolates, one is No. of the secretion of enzymes; which inactivate the isolates (%) aminoglycoside by different mechanisms as adenylation HLGR (high level gentamicin resistance) 35 (68) and phosphorylation5,6. HLSR (high level streptomycin resistance) 22 (43) There is limited information on the presence of HLAR HLGR + HLSR 22 (43) enterococci isolated from paediatric septicaemia cases HLGR (not associated with HLSR) 13 (23) in a tertiary care set up. A high prevalence rate of these isolates would require routine screening for this characteristic and seeking of alternative regimes for the of the isolates was done and interpreted according to treatment of such cases. The present study was carried National Committee for Clinical Laboratory Standards out to determine the occurrence of HLAR in enterococci (NCCLS) recommendations9. isolated from paediatric septicaemia patients in a tertiary care set up in north India. Results

Material & Methods Fifty one enterococcal isolates were isolated from the paediatric septicaemia cases during the study period. Neonates (n=21) and children (n=30) including nine Of these, 21 isolates were isolated from neonates, nine infants with clinical diagnosis of septicaemia admitted from infants and 21 from children. Thirty two of these in Kalawati Saran Hospital, New Delhi, India between isolates were from males. October, 2000 and June 2001 were included in the study. Blood was cultured using brain heart infusion broth (Hi- There is a higher percentage positivity of the Media Laboratories, Mumbai, India). Subcultures were isolates having high level gentamicin resistance (68%) made from this broth to sheep blood agar at regular in comparison to the isolates having high level intervals for isolation of enterococci. streptomycin resistance (43%). All the isolates that had high level streptomycin resistance also had high Enterococci were identified using standard methods level gentamicin resistance but the opposite was not based on gram staining, catalase reaction, bile aesculin, true (Table). Of the 35 high level gentamicin resistant growth in 6.5 per cent NaCl and Sugar fermentation isolates, 34 (97%) were vanomycin sensitive. Similarly reactions7. Agar screen method was used to detect a high level of vancomycin sensitivity (95%, 21 of 22) high level aminoglycoside resistance amongst the was observed in isolates with high level streptomycin enterococcal isolates8. Briefly 5-6 colonies of the isolate resistance. were suspended in brain heart infusion broth and concentration adjusted to 0.5 McFarland standard (1.5 Discussion ×108 cfu/ml). It was subsequently diluted and 10 µl of the suspension was spotted on to the various quadrants High level aminoglycoside resistant enterococci of the Mueller Hinton plate such that each spot had often have plasmids which carry determinants 104 cfu. Acceptable growth on the control quadrants encoding resistance to other antibiotics, besides limiting (i.e., with no drug) indicated acceptable quality control. the option of using a combination of cell wall active Presence of growth in the drug quadrant with antibiotics and aminoglycosides2. This drug streptomycin (2000 µg/ml) indicated the strain to be combination depends on the synergistic bactericidal high level streptomycin resistant. Similarly growth in activity between the two antibiotic groups, is often the quadrant with gentamicin indicated the strain to be used empirically in serious infections5. A high high level gentamicin resistant. Growth in both the drug occurrence of high level aminoglycoside resistance quadrants indicated high level resistance to both necessitates routine testing of the enterococcal streptomycin and gentamicin. Vancomycin susceptibility isolates. RANDHAWA et al : HLAR IN ENTEROCOCCI IN PAEDIATRIC SEPTICAEMIA 79 High level aminoglycoside resistant enterococci References were first reported in France in 1979 and since then 5 1. Patterson JE, Zerros MJ. High-level gentamicin resistance in have been isolated from all the continents . They Enterococcus: Microbiology, genetic basis, and epidemiology. comprise significant fraction of the clinical enterococcal Rev Infect Dis 1990; 12 : 644-52. 3 isolates at some centres . A study by Zervos and 2. Facklam RR, Teixeira LM. Enterococcus. In : Collier L, Balows associates reported a prevalence of 55 per cent of high A, Sussman M, editors. Topley & Wilson’s microbiology and level gentamicin resistance in enterococci in an US microbial infections. London : Arnold; 1998 p. 669-82. centre3. Another study reported a prevalence rate of 3. Zervos MJ, Kauffman CA, Therasse PM, Bergman AG, high level gentamicin resistance in enterococci varying Mikesell TS, Schaberg DR. Nosocomial infection by gentamicin- from 1 to 49 per cent in the 27 European countries resistant Streptococcus faecalis. Ann Intern Med 1987; 106: 687-91. studied10. 4. Herman DJ, Gerding DN. Screening and treatment of infection caused by resistant Enterococi. Antimicrob Agents Chemother In India (Nagpur), a study by Agarwal et al11 1991; 35 : 215-9. reported a prevalence of high level gentamicin 5. Eliopoulos GM, Moellering RC. Antimicrobial combinations. resistance in enterococci to be 7.8 per cent whereas In : Lorian V, editor. Antibiotics in laboratory medicine. Maryland high level streptomycin resistance was reported to be : William and Wilkins; 1996 p. 330-96. 24.7 per cent. The higher rates in the present study 6. Combes T, Carlier C, Courvalin P. Aminoglycoside-modifying may be ascribed to the source of the isolates being enzyme content of multiple-resistant strains of Streptococcus from a tertiary care set up where chronic cases are faecalis. J Antimicrob Chemother 1983; 11 : 41-7. prevalent and a wider usage of broad spectrum 7. Facklam RR, Sahm DF. Enterococcus. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH, editors. Manual of antibiotics occurs. The reason for the higher prevalence clinical microbiology. Washington DC: ASM Press; 1995 p. of high level streptomycin resistance in comparison to 308-14. the high level gentamicin resistance in the enterococcal 8. Woods GL, Washington JA. Antibacterial susceptibility tests : isolates in this study11 was not clear. It may be related Dilution & disc diffusion. In: Murrary PR, Baron EJ, Pfaller to greater usage of streptomycin in comparison to MA, Tenover FC, Yolken RH, editors. Manual of clinical gentamicin12. Majority of the enterococcal strains in microbiology. Washington DC: ASM Press; 1995 p. 1327-42. our study were sensitive to vancomycin. Though the 9. Performance standards for antimicrobial susceptibility testing. treatment regime with vancomycin would be effective Eighth informational supplement M100-58, vol.18, No.1, 1998, National Committee for Clinical Laboratory Standards, but would not be practicable in our country as the drug Pennsylvania, USA. is expensive and would be beyond the reach of most 10. Schouten MA, Voss A, Hoogkamp- Korstanje JAA, and the patients. European VRE study group. Antimicrobial susceptibility patterns of Enterococci causing infections of Europe. Antimicrob To conclude, the present study highlighted the Agents Chemother 1999; 43 : 2542-6. importance of high occurrence of high level 11. Agarwal VA, Jain YI, Pathak AA. Concomitant high level aminoglycoside resistant enterococci in our set up. This resistance to penicillin and aminoglycosides in Enterococci at Nagpur, Central India. Indian J Med Microbiol 1999; 17: would necessitate routine testing of the isolates for high 85-7. level aminoglycoside resistance. Alternative regimes in 12. Mundy LM, Sahm DF, Gilmore M. Relationship between the management of enterococcal infection need to be Enterococal virulence and antimicrobial resistance. Clin Microbiol evaluated. Rev 2000; 13 : 513-22. Reprint requests: Dr V.S. Randhawa, Department of Microbiology, Lady Hardinge Medical College, New Delhi 110001, India Indian J Med Res 119 (Suppl) May 2004, pp 80-83

Impact of susceptibility to antibiotics of streptococci & enterococci isolated from patients with infective endocarditis on antibiotic treatment

Liliana Mihaila-Amrouche, Laurent Schlegel, Anne Bouvet & the Association pour l’Etude et la Prévention de l’Endocardite Infectieuse (AEPEI) Study Group*

Centre National de Référence des Streptocoques - Service de Microbiologie, Hôtel Dieu, Assistance Publique- Hôpitaux de Paris, Université Paris VI, 1 place du Parvis Notre-Dame, F-75181 Paris Cedex 04 & *AEPEI: Service de Maladies Infectieuses et Tropicales, Hôpital Bichat-Claude Bernard, 75018 Paris, France

Received August 6, 2003

Background & objectives: Streptococci and enterococci are the most frequent pathogens causing infective endocarditis. In order to update the recommendations for both curative and prophylaxis treatment, the susceptibility to antibiotics of the most prevalent species of Streptococcaceae isolated from the patients with infective endocarditis was determined. Methods: Streptococcal and enterococcal isolates (n=133) isolated from confirmed cases of infective endocarditis during a one-year prospective survey conducted in 1999 in France were studied. The identification of 106 streptococci and 27 enterococci to the species level was carried out by conventional methods. Their susceptibility to ten antibiotics used in curative or prophylactic treatment was measured. Minimal inhibitory concentrations were determined by agar dilution method. Results: All the streptococcal and enterococcal isolates were susceptible to 4 mg/l or less of penicillin or amoxicillin. High levels of resistance to aminoglycosides were observed in two species, Streptococcus gallolyticus subsp. gallolyticus and Enterococcus faecalis. All isolates were susceptible to glycopeptides. Resistance to erythromycin, clindamycin, and pristinamycin was restricted to some species. Interpretation & conclusion: Curative treatments recommended for streptococcal or enterococcal endocarditis, including penicillin, amoxicillin or vancomycin in association with gentamicin were found to be appropriate for 98.5 per cent of cases. The emergence of erythromycin resistance in oral streptococci led to the use of pristinamycin in oral prophylactic treatment in patients allergic to b- lactams.

Key words Antibiotic treatment - Enterococci - infective endocarditis - prophylaxis - Streptococci

Streptococcaceae represent 60 per cent of the In order to update the recommendations for the microorganisms isolated from blood or cardiac vegetations treatment and the prophylaxis of endocarditis, the of patients with infective endocarditis1. The common susceptibility to antibiotics of the most prevalent species condition found to be responsible for endocarditis streptococci was due to oral dental procedures. The of streptococci and enterococci was determined and bacteraemia due to group D streptococci or enterococci compared with the results of the previous national were mostly associated with intestinal diseases2. survey3,4.

80 AMROUCHE et al : SUSCEPTIBILITY TO ANTIBIOTICS OF STREPTOCOCCI & ENTEROCOCCI 81

Material & Methods Table I. Streptococci and enterococci isolates from patients with infective endocarditis Streptococcal and enterococcal isolates (n=133) were Microorganism No. of isolates isolated from patients with endocarditis during a one- year prospective survey conducted in 1999 by the Pyogenic streptococci ( n=7 ) Association for Study and Prevention of Infective Streptococcus agalactiae 5 1 Endocarditis (AEPEI) study group in France . The S. dysgalactiae subsp. equisimilis 2 diagnosis was confirmed according to the Duke Oral streptococci ( n=38 ) University criteria5. S. mitis 14 The 106 streptococcal and 27 enterococcal isolates S. oralis 9 were identified by conventional methods on the basis of S. sanguinis 7 morphology, haemolysis, Lancefield’s group antigen, S. gordonii 2 growth in NaCl 6.5 per cent (wt/vol) broth, and bile- esculine test6. Biochemical characters of identification S. parasanguinis 1 were obtained with the rapid ID 32 STREP and Api 20 S. anginosus 3 Strep systems (BioMérieux, Marcy l’Etoile, France). S. intermedius 1 Species terminology was adapted from the results of S. salivarius 1 recent taxonomic studies7-9. Group D streptococci ( n=61 ) Titrated powders of the ten antibiotics recommended S. gallolyticus subsp. gallolyticus 51 for the curative or the prophylactic treatment of S. infantarius subsp. infantarius 6 endocarditis were used : penicillin G, amoxicillin, S. gallolyticus subsp. pasteurianus 4 ceftriaxone, gentamicin, tobramycin, vancomycin, Enterococci ( n=27 ) teicoplanin, erythromycin, clindamycin, and pristinamycin. Minimal inhibitory concentrations (MICs) were Enterococcus faecalis 27 determined by the agar dilution method10 in Mueller- Hinton agar for enterococci and Mueller-Hinton agar the oral group (n=38), and the S. bovis / S. equinus supplemented with 5 per cent horse blood for complex (group D streptococci) (n=61) . The 27 streptococci. Susceptibility, intermediate susceptibility, enterococcal isolates were Enterococcus faecalis. and resistance to antibiotics were interpreted according The level of susceptibility to antibiotics differed to the recommendations of the Comité de according to the groups and the species (Table II). Most l’Antibiogramme de la Société de Microbiologie of the streptococcal isolates, 102 of 106 (96%) were Française (CA-SFM)11. As an additional marker, high susceptible to penicillin G (MICs < 0.25 mg/l). Four level of resistance to streptomycin was detected with out of 38 oral streptococci had an intermediate the disk diffusion method using paper-load disk with 500 susceptibility (0.5 mg/l < MICs < 16 mg/l); of these, mg of streptomycin (Bio-Rad, Ivry sur Seine, France). three were S. sanguinis and one S. oralis. All In addition, the phenotypes of resistance to erythromycin, enterococcal strains were also of intermediate observed on the basis of a double diffusion disk test and susceptibility. No strain was resistant to penicillin G. E-test strips (AB Biodisk, Sweden) with both Among streptococcal isolates, 103 of 106 (97%) were erythromycin and clindamycin, were indicative of their categorized as susceptible to amoxicillin (MICs < 0.5 mechanism of resistance12. mg/l) and three S. sanguinis as intermediate (1 mg/l < MIC < 16 mg/l ). No streptococcal isolates was found Results to be resistant to amoxicillin. All enterococcal isolates were susceptible to amoxicillin (MICs < 4 mg/l). All The 106 streptococcal isolates belonged to 13 streptococcal isolates were susceptible to ceftriaxone different species (Table I). They were distributed (MICs < 0.5 mg/l) and all enterococcal isolates were between three major groups : the pyogenic group (n=7), resistant (MICs > 32 mg/l). 82 INDIAN J MED RES (SUPPL) MAY 2004

Table II. Susceptibility to antibiotics of oral and group D streptococci

Penicillin G Amoxicillin Erythromycin Clindamycin Pristinamycin SIS I SI R S RSIR Oral streptococci 34 4 35 3 30 2 6 32 6 36 2 (n=38) (I+R%) (11) (8 ) (21) (16 ) (5)

Group D streptococci 61 20 41 20 41 50 8 3 (n=61) (I +R%) (67) (67) (18)

S, susceptible; I, intermediate susceptibility; R, resistant

High-level of resistance to streptomycin was observed their antibiotic susceptibilities. The curative treatment in two species : S. gallolyticus subsp. gallolyticus (20 of recommended for endocarditis combines penicillin or 51 strains) and E. faecalis (7 of 27). All streptococci had amoxicillin with gentamicin. In allergic patients, a low level of resistance to gentamicin and tobramycin. vancomycin or teicoplanin replaces b-lactams13. Our Only two of the 27 E. faecalis had high level of resistance results indicated that these recommendations were to gentamicin and tobramycin (MICs > 1024 mg/l). All appropriate in 98.5 per cent of the cases. The choice of isolates were susceptible to glycopeptides. antibiotics to prevent endocarditis in at risk patients depends on the streptococcal and enterococcal species The intermediate susceptibility or resistance to present on the portal of entry of the bacteria. For the erythromycin and clindamycin was restricted to some oral prophylaxis during dental procedures, erythromycin species: S. mitis (4 and 2 of 14, respectively), S. sanguinis (or other 14- and 15 C membered macrolides), (3 of 7), S. anginosus (1 of 3), S. dysgalactiae subsp. clindamycin, and pristinamycin are the alternatives to equisimilis (1 of 2), S. gallolyticus subsp. gallolyticus amoxicillin in allergic patients14,15. In 1990-91 national (41 of 51) and E. faecalis (12 of 27). Both the phenotype survey, 6 per cent of the strains of Abiotrophia defectiva of resistance to erythromycin and the high level of MICs (former S. defectivus) , Granulicatella adiacens of erythromycin and clindamycin indicated a constitutive (former S. adjacens) and oral streptococci were found mecanism of methylation for all isolates except two to be resistant to erythromycin 3. In the present study, 21 S. mitis. These two isolates were resistant to erythromycin per cent of the oral streptococci were resistant or of and susceptible to clindamycin due to an inducible intermediate susceptibility to erythromycin. This phenotype of methylation mechanism of resistance in one emergence of macrolide resistance led to the use of isolate and a phenotype of an efflux mechanism in pristinamycin ( 5% of resistant strains) 4. another. In addition, two isolates of S. sanguinis among the 38 oral streptococci and 11 of S. gallolyticus subsp. Among group D streptococci, S. gallolyticus subsp. gallolyticus among the 61 group D streptococci showed gallolyticus was the predominant species. It is intermediate susceptibility ( MICs = 2mg/l) or resistance associated with colonic cancer and endocarditis. ( MICs > 4mg/l) to pristinamycin. All isolates studied Although most of the isolates were resistant to several were susceptible to penicillin or amoxicillin levels reached antibiotics, they were susceptible to the antibiotics in the serum during the treatment. All isolates were recommended for prophylaxis during intestinal susceptible to vancomycin and teicoplanin. Only two procedures which included penicillin, amoxicillin, isolates of E. faecalis had a high level of resistance to 14,15 gentamicin. gentamicin and vancomycin . Less than 10 per cent of isolates of E. faecalis associated with endocarditis Discussion of intestinal or urogenital origin, had a high level of resistance to gentamicin which is recommended in Both curative and prophylactic treatment of combination with amoxicillin or vancomycin in endocarditis depends on the causative organisms and prophylaxis. AMROUCHE et al : SUSCEPTIBILITY TO ANTIBIOTICS OF STREPTOCOCCI & ENTEROCOCCI 83 The association observed between different species 7. Osawa R, Fujisawa T, Sly LI. Streptococcus gallolyticus and their susceptibility to various antibiotics confirmed sp. nov.; gallate degrading organisms formerly assigned to Streptococcus bovis. Syst Appl Microbiol 1995; 18 : the need of accurate identification of causative 75-8. microorganism for appropriate treatment of endocarditis. 8. Bouvet A, Grimont F, Collins MD, Benaoudia F, Devine C, Regnault B, et al. Streptococcus infantarius sp. nov. related to Acknowledgment Streptococcus bovis and Streptococcus equinus. Adv Exp Med Biol 1997; 418 : 393-5. The authors thank the colleagues of six French regions who 9. Schlegel L, Grimont F, Ageron E, Grimont PAD, Bouvet A. collected the strains and the members of the AEPEI Study Group: Reappraisal of the taxonomy of the Streptoccocus bovis / Leport C, Hoen B, Alla F, Béguinot I, Briançon S, Danchin N, Streptococcus equinus complex and related species description Delahaye F, Etienne J, Ruimy R, Selton-Suty C. of Streptococcus gallolyticus subsp. gallolyticus subsp. nov., S. gallolyticus subsp. macedonicus subsp. nov., and S. gallolyticus References subsp. pasteurianus subsp. nov. Int J Syst Evol Microbiol 2003; 53 (Pt 3) : 631-45. 1. Hoen B, Alla F, Selton-Suty C, Beguinot I, Bouvet A, Briançon 10. Courvalin P, Soussy CJ. Report of the Comite de l’ S, et al. Changing profile of infective endocarditis - results of a Antibiogramme de la Societe Francaise de Microbiologie. 1-year survey in France. JAMA 2002; 288 : 75-81. Technical recommendations for in vitro susceptibility testing. Clin Microbiol Infect 1996; 2 (Suppl 1) : 11-25. 2. Herrero IA, Rouse MS, Piper KE, Alyaseen SA, Steckelberg JM, Patel R. Reevaluation of Streptococcus bovis endocarditis 11. Comite de l’antibiogramme de la Societe Francaise de cases from 1975 to 1985 by 16s ribosomal DNA sequence Microbiologie report 2003. Int J Antimicrob Agents 2003; 21 : analysis. J Clin Microbiol 2002 ; 40 : 3848-50. 364-91. 3. Bouvet A. Endocardite Infectieuse : Enquête en France 1990- 12. Seppäla H, Nissinen A, Yu Q, Huovinen P. Three different 1991. Sensibilité aux antibiotiques des Streptocoques et phenotypes of erythromycin-resistant Streptococcus pyogenes Enterocoques. Méd Mal Infect 1992; 22 Suppl: 987-92. in Finland. J Antimicrob Chemother 1993; 32 : 885-91. 4. Mihaila-Amrouche L, Schlegel L, Collobert G, Bouvet A. 13. Wilson WR, Karchmer AW, Dajani AS, Taubert KA, Bayer A, Evolution de la sensibilité aux antibiotiques des souches de Kaye D, et al. Antibiotic treatment of adults with infective streptocoques et enterocoques responsables d’endocardites endocarditis due to streptococci, enterococci, staphylococci, infectieuses en France de 1990 à 1999. Méd Mal Infect 2002 ; and HACEK microorganisms. American Heart Association. 32 : 618-23. JAMA 1995; 274 :1706-13. 5. Durack DT, Lukes AJ, Bright DK. New criteria for diagnosis of 14. Cinquième Conférence de Consensus en thérapeutique infective endocarditis : utilization of specific echocardiographic anti-infectieuse de la Société de Pathologie Infectieuse de findings. Duke Endocarditis Service. Am J Med 1994; 96 : 200-9. Langue française Paris 27 mars 1992. Méd Mal Infect 1992; 22 (Suppl) : 1119-41. 6. Facklam RR, Washington J A II. Streptococus and related catalase - negative Gram-positive cocci. In: Balows A, Hausler WJ Jr, 15. Dajani AS, Taubert KA, Wilson W, Bolger AF, Bayer A, Herrmann KL, Isenberg HD, Shadomy HJ, editors. Manual of Ferrieri P, et al. Prevention of bacterial endocarditis. clinical microbiology, 5th ed.Washington DC : American Society Recommendations by the American Heart Association. JAMA for Microbiology; 1991 p. 238-57. 1997; 277 : 1794-801.

Reprint requests: Dr Liliana Mihaila-Amrouche, Service de Microbiologie, Hôtel Dieu, 1 place du Parvis Notre-Dame F-75181 Paris Cedex 04, France e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 95-98

Human immunogenicity studies on group A streptococcal C5a peptidase (SCPA) as a potential vaccine against group A streptococcal infections

Anita Shet, Edward Kaplan, Dwight Johnson & Patrick Paul Cleary*

Department of Pediatrics, World Health Organization Collaborating Center for Reference & Research on Streptococci & *Department of Microbiology, University of Minnesota Medical School, Minneapolis, MN 55455, USA

Received August 6, 2003

Background & objectives: Group A streptococcal C5a peptidase (SCPA) is a major virulence surface factor. Its highly conserved nature among all tested serotypes of group A streptococci (GAS) as well as animal protection studies make SCPA a prime vaccine candidate. The present study was undertaken to explore the human immunogenicity to SCPA using an indirect enzyme-linked immunosorbent assay. Methods: Children (n=72) who had signs and symptoms of acute pharyngitis and had GAS isolated from the throat at initial visit were included. Acute and convalescent sera were collected 4 weeks apart. ELISA was performed using recombinant SCPA peptide as antigen. Results: The mean convalescent anti-SCPA level was twice the level of mean acute anti-SCPA and the difference was statistically significant (P < 0.0001). There was a rise in convalescent anti-SCPA in all children aged 2-12 yr. Interpretation & conclusion: Our observations confirmed that SCPA was highly immunogenic in children infected with group A streptococcal pharyngitis. Further studies need to be done to characterize the immune response including antibody subclass.

Key words C5a peptidase - group A streptococcus - humoral immune response - vaccine Group A streptococcus is an important human and Europe1,2. Observations of increased incidence of pathogen that causes a variety of infections ranging in macrolide resistance among isolates of group A severity from pharyngitis and pyoderma to severe invasive streptococci3, as well as reports of clinical and diseases such as necrotizing fasciitis and toxic shock microbiological failures after penicillin treatment4, make syndrome. Streptococcal non-suppurative sequelae like the development of a safe and efficacious vaccine more acute rheumatic fever and rheumatic heart disease urgent than ever. remain major public health problems among most of the world’s population. The use of antibiotics and an Among the different surface proteins that have been improved standard of living have not significantly considered as potential candidate vaccines, the group A eliminated the high morbidity and mortality of group A streptococcal C5a peptidase (SCPA) plays a major role streptococcal infections as seen by many outbreaks of as a virulence factor by allowing the group A streptococci rheumatic fever and invasive disease in the United States to establish a mucosal site of infection5,6. The objective

95 96 INDIAN J MED RES (SUPPL) MAY 2004 of the present study was to explore the human When the immune response to SCPA was plotted immunogenicity of SCPA after natural infection with according to M type of the infecting strain it was found group A streptococcus using a standardized enzyme- that a mean rise in anti-SCPA occurred irrespective linked immunosorbent assay (ELISA) 7. of the M type, suggesting that all M types produced SCPA. Material & Methods Discussion The study population consisted of children (aged 2- 12 yr) residing in various states in the United States, Important features of a good vaccine candidate are who had signs and symptoms of acute pharyngitis and as follows: significant contribution to disease had group A streptococci isolated from the throat at pathogenesis, conservation among all the different the initial visit. Acute and convalescent sera were circulating strains, absence of ability to evoke human obtained 4 weeks apart. The ELISA was performed cross-reactive antibodies, protection in animals against as described previously7. Recombinant SCPA peptide infective challenge with the vaccine strains, and was used as the antigen and coated overnight on reproducible immunogenicity in the host. SCPA has microtiter plates. Acute and convalescent sera from fulfilled most of the above criteria. Our study clearly the same patient were always run adjacent to each showed that a strong immune response to SCPA was other on the same microplate under identical conditions. mounted in children following natural infection with group After incubation with secondary antibody, (alkaline A streptococcal pharyngitis. phosphatase-conjugated goat anti-human IgG) the substrate, p-nitrophenyl phosphate was added and the As a major streptococcal virulence factor, SCPA has enzyme reaction was stopped. Optical density at a been shown to contribute significantly to the organism’s wavelength of 405 nm was measured with an EL340 ability to colonize mucosal surfaces. By enzymatically BioTek microplate reader and analyzed to obtain cleaving C5a, the complement-derived chemotaxin, at antibody concentrations8. Paired Student’s ‘t’ test was the His67 – Lys68 bond, the leukocyte-binding site of C5a used to compare the means of the anti-SCPA responses was removed, and chemotaxis was inhibited5. The in acute and convalescent sera following group A resulting lack of host phagocyte infiltration at the initial streptococcal pharyngitis. The analyses were performed site prevents bacteria from being cleared, thus allowing using Instat for Macintosh version 2.0 (GraphPad streptococci to establish infection. This has been clearly Software, San Diego, CA). P<0.005 was considered shown in mouse experiments involving colonization of significant. the oral mucosa9. The presence of humoral and mucosal antibodies in mice, induced by using recombinant SCPA Results as an immunogen, was associated with protection against nasopharyngeal colonization after challenge with several A wide range in anti-SCPA concentrations was seen different streptococcal M types10. These results strongly in both acute and convalescent sera of 72 children suggest cross-protection. This feature of a potential included in the study. Overall, the mean convalescent vaccine candidate is especially important in view of anti-SCPA was twice the concentration of mean acute recent reports confirming the dynamic changes in anti-SCPA and this difference was statistically significant prevailing group A streptococcal serotypes within an (P < 0.0001). Of the paired acute and convalescent sera isolated community11, and also the well known fact that that were assayed, over two-thirds of the children studied most streptococcal isolates from developing countries demonstrated a significant rise, i.e., > 15 per cent rise are characterized as non-typeable strains12. SCPA plays in convalescent anti-SCPA concentration. When the an additional role in virulence by facilitating intracellular immune response to SCPA was examined according to invasion of epithelial cells13. This invasive property is age, it was seen that a mean rise in convalescent anti- blocked by the presence of antibody to SCPA. SCPA occurred in all of the represented age groups between 2 and 12 yr. It appeared that children below SCPA appears to be highly conserved among all the age of 5 yr had a higher magnitude of response. tested group A streptococcal strains as demonstrated SHET et al: HUMAN IMMUNOGENICITY STUDIES WITH SCPA 97 by the presence of cross-reacting antibodies among over 3. Martin JM, Green M, Barbadora KA, Wald ER. Erythromycin- 20 different serotypes of group A streptococci10,14. The resistant group A streptococci in schoolchildren in Pittsburgh. scpA gene is also found to be present among several N Engl J Med 2002; 346 :1200-6. different strains of group A streptococci15. A similar 4. Gastanaduy AS, Kaplan EL, Huwe BB, McKay C, peptidase is also expressed by human isolates of group Wannamaker LW. Failure of penicillin to eradicate group A B16-18, C (unpublished data) and group G streptococci19. streptococci during an outbreak of pharyngitis. Lancet 1980; ii : 498-502.

It has been previously shown that human sera 5. Cleary PP, Prahbu U, Dale JB, Wexler DE, Handley J. Streptococcal C5a peptidase is a highly specific endopeptidase. containing high titre of anti-SCPA IgG could successfully Infect Immun 1992; 60 : 5219-23. inhibit and neutralize SCPA activity20. More recently, it has been shown that inhibition of SCPA activity is not 6. Ji Y, McLandsborough L, Kondagunta A, Cleary PP. C5a peptidase alters clearance and trafficking of group A streptococci serotype specific. Rabbits vaccinated with SCPA were by infected mice. Infect Immun 1996; 64 : 503-10. found to have high titres of specific antibody, which was able to neutralize SCPA activity associated with several 7. Shet A, Kaplan EL, Johnson DR, Cleary PP. Immune response to group A streptococcal C5a peptidase in children: 10 different serotypes of group A streptococci . The strong Implications for vaccine development. J Infect Dis 2003; 188 : immune response generated in children infected with 809-17. group A streptococci in our study could potentially 8. Plikaytis BD, Holder PF, Carlone GM. Program ELISA for generate non-serotype specific neutralizing antibodies Windows User’s manual, version 1.00. Centers for Disease which could provide some degree of protection for the Control and Prevention, Atlanta, GA, USA, 1996. host against early establishment of infection by the 9. Ji Y, Schnitzler N, DeMaster E, Cleary P. Impact of M49, Mrp, bacteria. Enn, and C5a peptidase proteins on colonization of the mouse oral mucosa by streptococcus pyogenes. Infect Immun 1998; 66 : In conclusion, the present results showed that SCPA 5399-405. was highly immunogenic in children recently infected 10. Ji Y, Carlson B, Kondagunta A, Cleary PP. Intranasal with streptococcal pharyngitis and the magnitude of this immunization with C5a peptidase prevents nasopharyngeal immune response was independent of the infecting colonization of mice by the group A Streptococcus. Infect Immun serotype. These data, in addition to previous knowledge 1997; 65 : 2080-7. of the well-defined role of SCPA in virulence, lack of 11. Kaplan EL, Wotton JT, Johnson DR. Dynamic epidemiology of known cross-reactive immunity with human tissue, and group A streptococcal serotypes associated with pharyngitis. ability of neutralizing antibody to prevent bacterial Lancet 2001; 358 :1334-7. colonization, make SCPA an ideal vaccine candidate for 12. Kaplan EL, Johnson DR, Nanthapisud P, Sirilertpanrana S, combatting group A streptococcal disease. However, Chumdermpadetsuk S. A comparison of group A streptococcal further work characterizing the immune response serotypes isolated from the upper respiratory tract in the USA and Thailand: implications. Bull World Health Organ 1992; 70 : including antibody subclasses, cellular immunity, as well 433-7. as functional assays need to be done. Important questions 13. Purushothaman SS, Cleary P. Virulence of group A streptococcus: that need to be addressed include normal levels of anti- dual functions of the C5a peptidase [Abstract P2.34]. In: XV SCPA in different populations, and in populations Lancefield International Symposium on Streptococci and endemic for rheumatic fever and rheumatic heart Streptococcal Diseases, October 2002. Goa, India, 2002. disease. 14. O’Connor SP, Cleary PP. In vivo Streptococcus pyogenes C5a peptidase activity: analysis using transposon- and References nitrosoguanidine-induced mutants. J Infect Dis 1987; 156 : 495-504. 1. Veasy LG, Wiedmeier SE, Orsmond GS, Ruttenberg HD, Boucek 15. Chen CC, Cleary PP. Complete nucleotide sequence of the MM, Roth SJ, et al. Resurgence of acute rheumatic fever in the streptococcal C5a peptidase gene of Streptococcus pyogenes. intermountain area of the United States. N Engl J Med 1987; J Biol Chem 1990; 265 : 3161-7. 316 : 421-7. 16. Hill HR, Bohnsack JF, Morris EZ, Augustine NH, Parker CJ, 2. Stevens DL. Streptococcal toxic-shock syndrome: spectrum of Cleary PP, et al. Group B streptococci inhibit the chemotactic disease, pathogenesis, and new concepts in treatment. Emerg activity of the fifth component of complement. J Immunol 1988; Infect Dis 1995; 1: 69-78. 141 : 3551-6. 98 INDIAN J MED RES (SUPPL) MAY 2004

17. Cleary PP, Handley J, Suvorov AN, Podbielski A, 19. Cleary PP, Peterson J, Chen C, Nelson C. Virulent human strains Ferrieri P. Similarity between the group B and A of group G streptococci express a C5a peptidase enzyme similar streptococcal C5a peptidase genes. Infect Immun 1992; to that produced by group A streptococci. Infect Immun 1991; 60 : 4239-44. 59 : 2305-10. 18. Chmouryguina I, Suvorov A, Ferrieri P, Cleary PP. Conservation 20. O’Connor SP, Darip D, Fraley K, Nelson CM, Kaplan EL, of the C5a peptidase genes in group A and B streptococci. Infect Cleary PP. The human antibody response to streptococcal C5a Immun 1996; 64 : 2387-90. peptidase. J Infect Dis 1991; 163 : 109-16.

Reprint requests: Dr Anita Shet, Department of Clinical Investigation, Rockefeller University Hospital 500 E 63rd Street, 17C, New York, NY 10021, USA e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 104-107

Preclinical evaluation of a vaccine based on conserved region of M protein that prevents group A streptococcal infection

Michael Batzloff, Huaru Yan, Mark Davies, Jon Hartas & Michael Good

Cooperative Research Centre for Vaccine Technology, The Queensland Institute of Medical Research, Post Office Royal Brisbane Hospital, QLD 4029, Australia

Received August 7, 2003

Background & objectives: Infection with group A Streptococcus (GAS) may result in a number of human diseases ranging from the relatively benign pharyngitis to the potentially life-threatening invasive diseases and post-infectious sequelae. We have previously defined a minimal B-cell epitope from the conserved region of the M-protein. Here we report on the immunogenicity, opsonic potential of the resulting sera and the level of protection induced by this peptide in comparison to a pepsin extract of the M protein. Methods: Inbred mice were immunized with peptides derived from the M protein. Sera were collected from the immunized mice and its opsonic potential determined for M1 and M6 GAS strains. Mice were then intranasally challenged with a virulent M1 GAS strain to determine the protective efficacy of the peptides. Results: The peptides induced significant antibody responses when delivered subcutaneously and immunized mice demonstrated significantly enhanced survival compared to control groups following challenge. Interpretation & conclusion: The data obtained in the present study indicated that the chimeric peptide J8 from the conserved region of the M protein could form the basis for an anti-streptococcal vaccine in future.

Key words Chimeric peptides - group A streptococcus - immunogenicity - M protein

Infection with group A Streptococcus (GAS) may induced antibodies in vaccinated mice that opsonised result in a number of human diseases ranging from the multiple serotypes of GAS2,3. We also demonstrated that relatively benign pharyngitis to the potentially life- the prevalence of p145-specific antibodies amongst threatening invasive diseases and post-infectious humans living in GAS-endemic areas increased with age sequelae, such as rheumatic fever (RF) and rheumatic being found in approximately 40 per cent of children and heart disease (RHD). The resurgence in severe invasive over 90 per cent of adults4 - possibly increasing in disease and increased incidence of GAS infection, in parallel with the acquisition of natural immunity to GAS. combination with the dynamic epidemiology of GAS We also demonstrated that affinity-purified human serotypes highlight the need to develop a safe, broad- antibodies to p145 could directly opsonise multiple strains based, efficacious vaccine against this bacterium1. of GAS in vitro4,5.

We have previously described a 20-mer peptide Peptide p145 would be a vaccine candidate except (p145) from the conserved region of the M protein which for the fact that humans may develop cross-reactive 104 BATZLOFF et al : IMMUNOGENICITY & PROTECTIVE POTENTIAL OF J8 PEPTIDE AGAINST GAS 105 T cells to this region of the M protein6. We therefore Detection of murine antibodies: An ELISA was used determined the minimal antibody epitope from the to measure J8-specific murine serum IgG antibodies and p145 peptide and separated this from any potential the antibody isotypes were determined as described2,8. deleterious T cell epitope. We previously determined Titre was defined as the highest dilution that gave an that the protective epitope in p145 was conformational optical density (OD) reading of more than three standard with alpha helical propensity7 a technique was deviations (SD) above the mean OD of control wells developed to map the epitope by embedding small containing normal mouse sera. peptides from p145 within other longer and unrelated peptides which were also known to form a helical Indirect bactericidal assay: Murine anti-peptide sera coil. were assayed for their ability to opsonise GAS in vitro as previously described2,4,5,10. Briefly, bacteria were These peptides, referred to as chimeric peptides, grown overnight at 37oC in 5ml Todd-Hewitt broth maintained the appropriate tertiary structure, and (THB). GAS was then serially diluted to 10-5 in enabled us to map the minimal antibody epitope. The phosphate-buffered saline (PBS). For each individual antibody epitopes recognised by endemic human sera mouse, 50µl of fresh heat-inactivated serum was mixed were chimeric peptides J2, J7 and J84, whereas the with 50µl of the bacterial dilution and incubated for chimeric peptides recognised by p145 murine antisera 20 min at room temperature. After the incubation, 400µl were J7 and J87,8. Human T cells respond to J2 and of non-opsonic heparinised human donor blood was murine T cells from B10.BR mice immunized with p145 added. All donor blood was pre-screened prior to respond vigorously to J38. The aim of this study was performing the assay to ensure that it could support the to compare the immunogenicity and protective potential growth of the GAS strain by at least 32 times the of our chimeric peptide (J8) with p145 and a pepM inoculum in a 3 h incubation at 37oC. extract. The mixtures were incubated end-over-end at 37oC Material & Methods for 3 h and 50µl from each tube was plated out in duplicate on 2 per cent blood THB agar pour plates. Peptide synthesis: Synthetic peptides were produced The plates were incubated overnight and the number of as described9 and were purified by HPLC. The peptide colonies on each plate was determined. Opsonic activity sequences for the p145 and J8 peptides are LRRD of the anti-peptide sera [% reduction in mean colony LDASREAKKQVEKALEC and QAEDKVKQ forming units (cfu)] was calculated as [1-(cfu in the SREAKKQVEKALKQLEDKVQ respectively. presence of anti-peptide sera) / (mean cfu in the presence of normal mouse sera)] X 100. Immunization of mice: Peptides were administered subcutaneously in a volume of 50µl at the tail base to Intranasal GAS challenge procedure: GAS was B10.BR (inbred, H-2k) mice (Animal Resources Centre, cultured overnight in THB with 1 per cent neopeptone Western Australia). Each mouse received a total of and 200 mg/ml streptomycin, spun at 1500 g for 10 min 30µg of immunogen emulsified 1:1 in complete Freund’s and washed twice with THB containing 1 per cent adjuvant (CFA) (Difco, USA). Mice were also given neopeptone and resuspended in 25 per cent of the original subsequent booster injections at days 21, 31 and 41 post- volume. The inoculum dose (cfu/ml) was determined by primary immunization. optical density at 600nm and plating out 10-fold dilutions of bacterial suspensions on 2 per cent horse blood THB Serum collection: Blood was collected from mice via agar plates. Following overnight incubation of the plates the tail artery and allowed to clot at 37oC for at least at 37oC, colony counts were determined. Immunized 30 min. Serum was collected after centrifugation at and control mice were challenged intranasally with a 1000 g for 10 min, heat inactivated at 56oC for 10 min pre-determined dose of GAS in 30 µl (15 µl/nare). and stored at -20oC. Serum was collected from all of the mice 1-day prior to boosts and 15 days after the Statistical analysis: The T-tests, mean and the standard final boost. error of the mean were calculated using standard 106 INDIAN J MED RES (SUPPL) MAY 2004 formulae. The log rank or Mantel-Haenszel test was used to compare survival curves of challenged mice11. P< 0.05 was considered significant.

Results

Inbred B10.BR (H-2k) mice were immunized with the conserved region peptide p145 and its chimeric derivative J8. Control mice received PepM extract or PBS in CFA. The mice developed strong antigen-specific serum IgG titres (Fig.1). IgG isotypes were predominantly the IgG2a > IgG1 > IgG2b for both the J8 and p145 peptide. However for the pepM extract IgG1was the predominant isotype followed by IgG2a and Fig. 1. Antigen-specific serum IgG titres of immunized mice taken at IgG2b (data not shown). While IgA was not detectable day 56 post-primary immunization. Horizontal bar indicates average. in the serum, antigen-specific serum IgM was shown to PBS, phosphate buffer saline. be present albeit 5-fold lower then IgG1.

Serum collected from the immunized mice was capable of the in vitro opsonisation of the M1 GAS strain (Fig.2). Mice immunized with p145 (P < 0.001), J8 (P < 0.01) and pepM (P < 0.001) had significantly higher levels of opsonisation when compared to the PBS control group. Interestingly, serum from both peptide (p145 and J8) immunized mice was capable of the in vitro opsonisation of M1 GAS strain at similar levels to that of the pepM mice (P>0.05). The antiserum was also shown to opsonise the M6 GAS strain at similar levels to that observed for the M1 strain (data not shown). Fig.2. Average per cent opsonisation of an M1 GAS strain by serum taken at day 56 post-primary immunization. PBS, phosphate buffer Mice were then challenged intranasally with the M1 saline. GAS strain (Fig.3). Mice immunized with p145 (P < 0.05), J8 (P < 0.05) and pepM (P < 0.05) had significantly higher levels of survival compared to the PBS/CFA immunized mice. There was no significant difference in survival in the p145, J8 and pepM immunized groups, indicating the conserved region peptide was able to induce protection equal to that of the pepM immunized mice.

Discussion

This study was carried out to determine the immunogenicity of a conserved M protein epitope, J8, with CFA and compare the levels of opsonisation and protection induced by immunization with the parent peptide p145 and the M protein extract pepM. It was demonstrated that J8 Fig.3. Per cent survival of immunized mice following intranasal induced significant antibody responses when delivered challenge with an M1 GAS strain. BATZLOFF et al : IMMUNOGENICITY & PROTECTIVE POTENTIAL OF J8 PEPTIDE AGAINST GAS 107 subcutaneously and that immunized mice demonstrated fever: identification of a conserved target epitope on M significantly enhanced survival compared to control groups. protein of group A streptococci. Lancet 1994; 344 : 639-42. J8 when administered with CFA can protect up to 80 4. Brandt ER, Hayman WA, Currie B, Carapetis J, Wood Y, Jackson DC, et al. Opsonic human antibodies from an k per cent of H-2 responder strain mice (B10.BR) from endemic population specific for a conserved epitope on the virulent challenge12. However, minimal epitopes such as M protein of group A streptococci. Immunology 1996; 89 : J8 are unlikely to be immunogenic in an outbred population. 331-7. To render such epitopes immunogenic they must be 5. Brandt ER, Hayman WA, Currie B, Pruksakorn S, Good MF. conjugated to carrier proteins capable of inducing T cell Human antibodies to the conserved region of the M protein: help. Diphtheria toxoid (DT) has been used as a carrier opsonization of heterologous strains of group A streptococci. Vaccine 1997; 15 : 1805-12. protein in many systems and preliminary data from our own laboratory suggested that anti-DT antibodies may in 6. Pruksakorn S, Currie B, Brandt E, Phornphutkul C, Hunsakunachai S, Manmontri A, et al. Identification of T cell fact be capable of opsonising GAS making DT an obvious autoepitopes that cross-react with the C-terminal segment of choice for a carrier protein for J8 in future studies. the M protein of group A streptococci. Int Immunol 1994; 6 : 1235-44. The mechanism for J8 mediated protection of mice 7. Relf WA, Cooper J, Brandt ER, Hayman WA, Anders RF, from GAS challenge has not been completely defined. Pruksakorn S, et al. Mapping a conserved conformational epitope from the M protein of group A streptococci. Pept Res 1996; 9 : However it has been demonstrated that the N-terminal 12-20. epitopes of the M protein induce high levels of serum 8. Hayman WA, Brandt ER, Relf WA, Cooper J, Saul A, Good antibodies, which are highly opsonic to the corresponding MF. Mapping the minimal murine T cell and B cell epitopes GAS strain13-15. These epitopes have also been shown within a peptide vaccine candidate from the conserved region of to induce high levels of protection from GAS challenge the M protein of group A streptococcus. Int Immunol 1997; 9 : in mice12, indicating that protection is antibody mediated. 1723-33. 9. Houghten RA. General method for the rapid solid-phase In conclusion, the findings of the present study showed synthesis of large numbers of peptides: specificity of antigen- antibody interaction at the level of individual amino acids. Proc that our peptide from the conserved region of the M protein Natl Acad Sci USA 1985; 82 : 5131-5. could form the basis of an anti-streptococcal vaccine. 10. De Malmanche SA, Martin DR. Protective immunity to the group A streptococcus may only be strain specific. Med Acknowledgment Microbiol Immunol 1996; 183 : 299-306. 11. Altman D. Practical statistics for medical research. New York: The authors acknowledge the financial support by the NHMRC Chapman and Hall; 1991. (Australia), The Prince Charles Hospital Foundation, National Heart 12. Brandt ER, Sriprakash KS, Hobb RI, Hayman WA, Zeng W, Foundation of Australia and the Cooperative Research Centre for Batzloff MR, et al. New multi-determinant strategy for a group Vaccine Technology. A streptococcal vaccine designed for the Australian aboriginal population. Nat Med 2000; 6 : 455-9. References 13. Dale JB, Simmons M, Chiang EC, Chiang EY. Recombinant, octavalent group A streptococcal M protein vaccine. Vaccine 1. Kaplan EL, Wotton JT, Johnson DR. Dynamic epidemiology of 1996; 14 : 944-8. group A streptococcal serotypes. Lancet 2002; 359 : 2115-6. 14. Lancefield RC. Current knowledge of the type specific M antigens 2. Pruksakorn S, Galbraith A, Houghten RA, Good MF. of group A streptococci. J Immunol 1962; 89 : 307. Conserved T and B cell epitopes on the M protein of group A streptococci. Induction of bactericidal antibodies. J Immunol 15. Brandt ER, Teh T, Relf WA, Hobb RI, Good MF. Protective 1992; 149 : 2729-35. and nonprotective epitopes from amino termini of M proteins from Australian aboriginal isolates and reference 3. Pruksakorn S, Currie B, Brandt E, Martin D, Galbraith A, strains of group A streptococci. Infect Immun 2000; 68 : Phornphutkul C, et al. Towards a vaccine for rheumatic 6587-94.

Reprint requests: Dr Michael R. Batzloff, The Queensland Institute of Medical Research, PO Royal Brisbane Hospital Brisbane QLD 4029, Australia e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 108-114

Requests of medical examinations after pneumococcal & influenza vaccination in the elderly

D. D’Alessandro, S. Ciriminna†, A. Rossini*, M.C. Bossa*, & G.M. Fara*

DACP,*DPHS, “La Sapienza” University of Rome & †Sicily Region, Palermo, Italy

Received August 7, 2003

Background & objectives: There are no specific recommendations about pneumococcal vaccination (PV) in Italy. There are limited data on the requests of medical examination or hospitalization due to side effects of pneumococcal vaccination. In the present study we present our analyses on the requests of medical examination due to side effects of pneumococcal vaccination (PV) in 29,086 elderly aged ³65 yr. Methods: Vaccination with influenza vaccine (IV) alone or in association with PV was offered during the IV campaign in the winter 1999-2000 in 8 counties of Sicily. General practitioners (GPs) provided the vaccination and, 72 h later, recorded local and systemic side effects by a telephonic interview with the waccinees. GPs recorded also the requests of medical treatments. Results: A total of 16,601 subjects (57.1%) were vaccinated with IV alone and 12,485 (42.9%) with the association IV+PV. Adverse effects were reported by 14.7 per cent of vaccinees. Out of 4,289 subjects with side effects, 305 (7.1%) looked for a medical examination. The factors associated with a significantly increased risk for medical examination after vaccination were symptoms duration >72 h (OR 2.04; 95% CI 1.38-3.01; P=0.0002); local induration (OR 3.26; 95% CI 2.07-5.13; P<0.0001); general malaise (OR 2.05; 95% CI 1.34-3.13; P=0.0009); fever ³38°C (OR 3.34; 95% CI 1.90-5.86; P<0.0001); myalgia (OR 2.80; 95% CI 1.46-5.37; P=0.0018) and dyspnea (OR 37.86; 95% CI 9.65-148.60; P=0.0001). Interpretation & conclusion: The few requests for medical examination and the low rate of adverse effects confirmed the tolerability and safety of PV+IV; GPs consultation were related to the severity and duration of symptoms, but independent of the type of vaccine administered.

Key words Medical examinations - pneumococcal and influenza vaccine - tolerability Pneumococcal and influenza diseases are together invasive diseases, like meningitis, due to antibiotics among the major causes of morbidity and mortality in resistant strains have been shown to have a less the elderly and other at risk individuals, and are favourable outcome5,6. responsible for more number of deaths and hospitalisations1. The public health impact of these In Italy, there are specific national infections will be amplified in future decades for several recommendations about influenza vaccination and reasons. First, the proportion of the elderly is increasing; pneumococcal vaccine is offered only in few regions, at present in Italy the proportion of people aged > 65 yr in association with influenza vaccine, following local is about 18 per cent and as per the National Statistics indications. The limiting factors for giving national Office estimates for 2020 it will be 23 per cent2. Second, recommendations about this vaccine are the lack of in the last few years the antibiotics resistance of epidemiological data, and also the controversial results Streptococcus pneumoniae has increased3-5 and of published data on the effectiveness of

108 D’ALESSANDRO et al: MEDICAL EXAMINATION AFTER PNEUMOCOCCAL VACCINATION IN THE ELDERLY 109 polysaccharide vaccines against pneumococcal On the same questionnaire, after 72 h, GPs registered pneumonia among the elderly. by a telephonic interview about the side-effects if any, and use of drug due to onset of vaccine side effects. The simultaneous administration of pneumococcal and Further, GPs reported if vaccinated people had a medical influenza vaccines reduces the costs of its distribution examination also after vaccination. for health care organisation and increases the compliance in patients7-9. Further, this strategy does not significantly Survey responses were entered in a computerized increase the frequency of adverse effects7,9 or their database and statistical analysis were conducted in the severity, and does not affect the immunogenicity of either Department of Public Health Sciences at "La Sapienza" also10. There are limited data on the requests for medical University of Rome, with Statview 5.1® and Excel®. services due to the side effects of vaccination. Chi-squared test was used to compare proportions and trends, normal test and paired t-test were used to We present here our findings on the requests of medical compare means. Multivariate logistic regression was examinations, drug use and hospitalisation due to influenza used to evaluate the association of some characteristics and pneumococcal vaccinations among 29,086 persons (age, sex, living conditions, presence of more than one aged 65 yr or more who received either influenza and chronic illness, allergy, vaccine administered and pneumococcal vaccine or only influenza vaccine. symptoms) of the vaccinees with the request for medical examinations after vaccination. Material & Methods Results A cohort of 35,000 individuals aged ³ 65 yr was selected from the population living in 11 small cities A total of 29,086 individuals were vaccinated during (population of 40,000-70,000 inhabitants) of Sicily the campaign; of these, 16,601 (57.1%) were vaccinated Region. During the 1999-2000 influenza vaccination with IV alone and 12,485 (42.9%) were vaccinated with campaign, all General practitioners (GPs) of these cities IV+PV. Table I shows the major characteristics of the offered the influenza vaccine alone (IV) or in study population. IV+PV group had a higher proportion association with the 23-valent pneumococcal of subjects with chronic diseases, but the differences polysaccharides vaccine (IV+PV) to their patients of with IV group were not significant. this age group (³ 65 yr). Overall 4,289 subjects (14.7%) had at least one side The study included only the patients over 65 yr with effect due to the vaccination. The proportion of subjects with adverse effects showed a significant difference the condition to respond to a telephonic interview. All (P<0.0001) between IV group (2,145 subjects; 12.9%) patients received information about the aims of the study and IV+PV group (2,144 subjects; 17.2%). The and those enrolled provided written informed consent. significant difference persisted looking at the symptoms individually, with the exception of headache and dyspnea Both vaccines were administrated by the GPs; the (Table II). Local redness (6.2%) and local pain (5.0%) storage and handling of the vaccines was in accordance were most frequent local symptoms declared by with the manufacturers’ recommendations. The influenza vaccinees; respectively (5.9% in IV group versus 6.5% vaccine was given in the upper right arm, and the in IV+PV group for redness and 3.3% in IV group versus pneumococcal vaccine was given in the upper left arm. 7.2% in IV+PV group for local pain). Among systemic Both vaccinations were offered free of charge. adverse effects, general malaise (1.3% in IV group versus 2.6% in IV+PV) and fever (1.1% in IV group The elderly who received the simultaneous versus 2.6% in IV+PV) represented the most recorded pneumococcal and influenza vaccination, or influenza symptoms. In 39 per cent of cases fever exceeded vaccine alone, were interviewed during vaccination, 38.0°C. The proportion of persons with myalgias and using an ad-hoc questionnaire, about age, sex, living joint pain was more than three time higher in IV+PV conditions, baseline medical conditions (chronic and group versus IV group, nevertheless these symptoms immunosuppressive illness) and previous vaccinations. were reported infrequently. 110 INDIAN J MED RES (SUPPL) MAY 2004

Table I. General characteristics of subjects by type of vaccination administered

IV groupIV+PV groupTotal No. % No. % No. %

Subjects 16,601 57.1 12,485 42.9 29,086 100 Male sex 7,336 44.2 5,689 45.6 13,025 44.8 Mean age±SD (yr) 74.5 ±6.8 74.6 ±6.9 74.6 ±6.9 Living conditions: alone 2,607 15.7 2,122 17.0 4,729 16.3 with family 10,609 63.9 8,346 66.8 18,955 65.2 other conditions 949 5.7 534 4.3 1,483 5.1 not specified 2,436 14.7 1,483 11.9 3,919 13.5 Allergy: Yes 554 3.3 397 3.2 951 3.3 No 12,826 77.3 10,344 82,9 23,17 79.7 Uncertain 3,221 19.4 1,744 14.0 4,965 17.1 Heart disease 9,177 55.3 7,267 58.2 16,444 56.5 Chronic lung disease 4,633 27.9 4,161 33.3 8,794 30.2 Diabetes 2,308 13.9 1,955 15.6 4,263 14.6 Having at least 1 comorbidity 5,677 34.2 5,337 42.7 11,014 37.9 Previous influenza vaccination: Yes 10,815 66.7 8,056 64.5 18,871 64.9 No 5,400 18.9 4,213 33.7 9613 33.1 Uncertain 386 2.3 216 1.7 602 2.1

IV, influenza vaccination; PV, pneumococcal vaccination

Table II. Proportion of adverse effects due to vaccination

IV group IV+PV group Total No. % No. % No. %

At least one adverse effects 2,145 12.9 2,144 17.2† 4,289 14.7 Local symptoms: Local redness 980 5.9 816 6.5** 1,796 6.2 Local swelling 407 2.5 486 3.9*** 893 3.1 Local pain 555 3.3 902 7.2† 1,457 5.0 Local induration 158 1.0 161 1.3* 319 1.1 General symptoms: General malaise 222 1.3 323 2.6*** 545 1.9 Fever 185 1.1 329 2.6† 514 1.8 Headache 75 0.5 74 0.6 149 0.5 Shivers 80 0.5 120 1.0*** 200 0.7 Joint pain 55 0.3 128 1.0*** 183 0.6 Myalgia 52 0.3 135 1.1† 187 0.6 Dyspnea 8 0.3 9 0.1 17 0.1 Itch 28 0.2 49 0.4*** 77 0.3 Other 86 0.5 51 0.4 137 0.5

IV, influenza vaccination; PV, pneumococcal vaccination P * < 0.01, ** < 0.05, *** < 0.001, † < 0.0001 compared to IV group D’ALESSANDRO et al: MEDICAL EXAMINATION AFTER PNEUMOCOCCAL VACCINATION IN THE ELDERLY 111

Table III. Characteristics of symptomatic subjects classified by medical examination and type of vaccine

IV IV+PV Total symptomatic total examined % total examined % total examined % Subjects 2,145 142 (6.6) 2,144 163 (7.6) 4,289 305 (7.1) Mean age (yr) 74.4 75.5 73.7 73.4 74 74,4 Male sex 824 56 (6.8) 876 59 (6.7) 1,700 115 (6.8) Female sex 1,321 86 (6.5) 1,268 104 (8.2) 2,589 190 (7.3) Living conditions (%) -alone 414 24 (5.8) 418 30 (7.2) 832 54 (6.5) -with family 129 71 (5.5) 1,448 93 (6.4) 2,738 164 (6.0) -institutionalized* 188 10 (5.3) 77 11 (14.3) 265 21 (7.9) Previous influenza vaccination 1,431 105 (7.3) 1,431 108 (7.5) 2,862 213 (7.4)

*Subjects living in residential homes or long term care facilities

Table IV. Percentage of subjects with a medical examination by symptoms and type of vaccination

Symptoms IV IV+PV Total Total no. with symptoms(%) Total no. with symptoms(%) Total no. with symptoms(%) Local redness 980 69 (7.0) 816 56 (6.9) 1,796 125 (7.0) Local swelling 407 32 (7.9) 486 33 (6.8) 893 65 (7.3) Local pain* 555 21 (3.8) 902 57 (6.3) 1,457 78 (5.4) Local induration 158 25 (15.8) 161 17 (10.6) 319 42 (13.2) General malaise 222 31 (14.0) 323 44 (13.6) 545 75 (13.8) Fever 185 24 (13.0) 329 39 (11.9) 514 63 (12.3) Headache 75 9 (12.0) 74 6 (8.1) 149 15 (10.1) Shivers 80 6 (7.5) 120 20 (16.7) 200 26 (13.0) Joint pain 55 9 (16.4) 128 16 (12.5) 183 25 (13.7) Myalgia 52 4 (7.7) 135 24 (17.8) 187 28 (15.0) Dyspnea 8 3 (37.5) 9 6 (66.7) 17 9 (52.9) Conjunctivitis 4 1 (25.0) 6 2 (33.3) 10 3 (30.0) Itch 28 4 (14.3) 49 5 (10.2) 77 9 (11.7) Nettle rash 5 0 (0.0) 5 1 (20.0) 10 1 (10.0) Other 77 6 (7.8) 40 8 (20.0) 117 14 (12.0) Total 2,145 142 (6.6) 2,144 163 (7.6) 4,289 305 (7.1)

* Only for the symptoms local pain the proportion of subjects who asked for a medical examination was significantly different in IV and IV+PV group (Chi square test 1.968; P < 0.05)

Table V. Average score of symptoms and chronic diseases in examined and not examined subjects

IV IV+PV Total Yes N o Yes N o Yes N o Symptoms 1.7 1.3** 2.1 1.6** 1.9 1.5** Chronic diseases 1.4 1.3 1.4 1.6* 1.4 1.5 P * < 0.05, ** < 0.001 compared to examined 112 INDIAN J MED RES (SUPPL) MAY 2004

Table VI. Factors associated with request of medical examination after vaccination

Factor OR 95% CI P value Sex (male) 0.78 0.55-1.11 0.1696 Age ³ 74 yr 1.04 0.74-1.45 0.8150 IV+PV vaccine 0.93 0.66-1.32 0.7229 Living conditions 1.07 0.51-2.24 0.8469 (institutionalized vs at home) Allergy 1.25 0.66-2.35 0.4858 Comorbidity 0.95 0.68-1.33 0.7948 Duration >72 h 2.04 1.38-3.01 0.0003 Local redness 1.03 0.71-1.50 0.8494 Local swelling 0.89 0.57-1.38 0.6182 Local pain 0.60 0.40-0.89 0.0119 Local induration 3.26 2.07-5.13 <0.0001 General malaise 2.05 1.34-3.13 0.0009 Fever ³C 38° 3.34 1.90-5.86 <0.0001 Headache 0.79 0.32-1.97 0.6256 Shivers 1.09 0.49-2.40 0.8233 Joint pain 0.75 0.34-1.67 0.4916 Myalgia 2.80 1.46-5.37 0.0018 Dyspnoea 37.86 9.65-148.60 <0.0001 Itch 1.23 0.37-4.12 0.7254

OR, odds ratio; CI, confidence interval

Severe systemic adverse effects (e.g., anaphylactic reactions) were not reported after administration of vaccines. Four subjects (0.0014% of vaccinees), all vaccinated with IV+PV, were admitted to district hospitals due to the following clinical pictures: dyspnoea with general malaise; persistent fever (more than 72 h) over 38°C; broncopneumonia; circulatory collapse.

In 4,289 individuals with side effects, 305 (7.1%) looked for a medical examination: 1.3 per cent of IV+PV Fig. Medical examination by symptoms duration and vaccination group and 0.9 per cent of IV group. The proportion of group. examined subjects in the symptomatic group did not show significant differences after stratification by type of drugs, 4 (2.8%) in IV group and 4 (2.4%) in IV+PV vaccination and by several socio-demographic conditions group. (Table III). Between symptomatic subjects, 11 (0.25%) took medicaments: 54.5 per cent non-steroidal anti- With the exception of local pain (3.8% versus 6.3%; inflammatory drugs (NSAIDs), 36.4 per cent cortisone, P<0.05) the difference between IV and IV+PV groups 9.1 per cent cortisone associated with antibiotics. Among was also not significant looking at the proportion of adverse those who requested medical examination, 8 (2.6%) took effects that led to a medical examination (Table IV). D’ALESSANDRO et al: MEDICAL EXAMINATION AFTER PNEUMOCOCCAL VACCINATION IN THE ELDERLY 113 It is interesting to notice that the symptom dyspnoea led results are more valid for predicting future vaccination to a medical exam in more than 50 per cent of cases and responses7. the proportion overcomes 65 per cent of cases in IV+PV subjects. As already described7,11,13, mild local side effects (e.g., redness, pain at the injection site and swelling) The average symptoms score was significantly higher were the most frequent adverse effects observed in both in examined versus not examined persons (1.9 versus vaccination groups. Moderate systemic reaction (general 1.5; P < 0.001), whereas the average score of the number malaise and fever) and more severe local reactions of chronic diseases was significantly lower only for the (induration) occurred in less than 2 per cent of the examined subjects from IV+PV group (1.4 versus 1.6; vaccinees. P < 0.05) (Table V). Although 14.7 per cent (4,289 subjects) vaccinees The proportion of medical examinations increased had at least one side effect, only 7.1 per cent of them significantly with the increase in symptoms duration (305 subjects) had a medical examination. The risk (P < 0.0001). The difference was significant looking at for medical examination was independent from the IV+PV group alone (P<0.0001) (Fig.). vaccine administered and from other individual characteristics, like age, sex, living conditions and baseline medical conditions. On the contrary it was Multivariate logistic regression was used to identify related to the duration of adverse reactions and to the the determinants independently associated with the needs occurrence of severe local side effects or systemic of medical examinations in subjects with side effects symptoms, like general malaise; fever ³ 38°C; myalgia after vaccination (Table VI). The factors associated with and dyspnoea. a significantly increased risk for medical examination after vaccination were symptoms duration > 72 h The few cases of drugs prescriptions (about 4 [adjusted odds ratio (OR) 2.04; 95 % CI 1.38-3.01; prescriptions per 10,000 vaccinees) and their typology P = 0.0002]; local induration (OR 3.26; 95% CI 2.07- supported the safety and tolerability of the administered 5.13; P < 0.0001); general malaise (OR 2.05; 95% CI vaccines. Of the hospitalized cases reported, only 2 1.34-3.13; P = 0.0009); fever ³ 38°C (OR 3.34; 95% CI clinical pictures (dyspnoea with general malaise and 1.90-5.86; P <0.0001); myalgia (OR 2.80; 95% CI 1.46- fever over 38°C > 72 h) could be related to the 5.37; P=0.0018) and dyspnoea (OR 37.86; 95% CI 9.65- vaccination administered, leading to an incidence rate of 148.60; P=0.0001). Local pain results associated with a about 0.7 cases per 10,000 vaccinees. significantly smaller risk for medical examination (OR 0.60; 95% CI 0.40-0.89; P=0.0119). In conclusion, GPs consultation were related to the severity and duration of symptoms, but independent of Discussion the type of vaccine administered. The scarce use of drugs and the low proportion of patients consulting a GP confirmed that the combination of pneumococcal and The severity of the adverse effects observed in this influenza vaccine was safe and well tolerated. study, after simultaneous IV+PV vaccination or IV vaccine alone, was similar to that described in other Acknowledgment studies7-9. The lower occurrence of local symptoms observed in comparison with previous studies7-9 could Authors thank the Public Health Doctors and the general be related to older age of the vaccinees (74.6 yr). Several practitioners of Sicily Region for the assistance in vaccination and 7,11,12 studies described a decreasing in the frequency of collection of data. local reactions with the increasing age due to lower preexisting antibody levels in older persons or to weaker References response to the vaccine. It could also be related to the method of data collection. The self-reporting data may 1. Influenza and pneumococcal vaccination levels among persons not be precisely comparable with the survey in which aged ³65 years - United States. 2001. MMWR Morb Mortal the data are collected by an external observer, but the Wkly Rep 2002; 51 : 1019-24. 114 INDIAN J MED RES (SUPPL) MAY 2004

2. http://demo.istat.it/previsioni/index.html 8. Fletcher TJ, Tunnicliffe WS, Hammond K, Roberts K, Ayres JG. Simultaneous immunisation with influenza vaccine and 3. Pantosti A, D’Ambrosio F, Tarasi A, Recchia S, Orefici G, pneumococcal polysaccaride vaccine in patients with chronic Mastrantonio P. Antibioic susceptibility and serotype respiratory disease. BMJ 1997; 314 : 1663. distribution of Streptococcus pneumoniae causing meningitisi in Italy, 1997-1999. Clin Infect Dis 2000; 9. Toniolo-Neto J, Weckx LW, Halker E, Lopes CH, de Menzes 31: 1373-9. Succi RC, de Paiva TM, et al. Safety of simultaneous pneumococcal and influenza vaccination in elderly patients in 4. Marchese A, Schito GC. Resistance patterns of lower respiratory Brazil. Drugs Aging 1999; 15 (Suppl 1) : 43-5. tract pathogens in Europe. Int J Antimicrob Agents 2000; 10. Grilli G, Fuiano L, Biasio LR, Pregliasco F, Plebani A, 16 Suppl 1 : S25-S29. Leibovitz M, et al. Simultaneous influenza and pneumococcal 5. Nuorti JP, Butler JC, Crutcher JM, Guevara R, Welch D, Holder vaccination in elderly individuals. Eur J Epidemiol 1997; 13 P, et al. An outbreak of multidrug-resistant pneumococcal : 287-91. pneumonia and bacteremia among unvaccinated nursing home 11. Fedson DS, Musher MD, Eskola J. Pneumococcal vaccine. In: residents. N Engl J Med 1998; 338 : 1861-8. Plotkin SA, Orenstein WA, editors. Vaccines, 3rd ed. 6. John CC. Treatment failure with use of a third generation Philadelphia: Saunders; 1999 p. 553-607. cephalosporin for penicillin resistant pneumococcal meningitis: 12. Ponka A, Leinonen M. Adverse reactions to polyvalent case report and review. Clin Infect Dis 1994; 18 :188-93. pneumococcal vaccine. Scand J Infect Dis 1982; 14 : 67-71. 7. Honkanen PO, Keistinen T, Kivela S. Reactions following 13. Prevention of pneumococcal disease: recommendations of the administration of influenza vaccine alone or with pneumococcal Advisory Committee on Immunization Practices (ACIP). vaccine to the elderly. Arch Intern Med 1996; 156 : 205-8. MMWR Recomm Rep 1997; 46 (RR-8) : 1-24.

Reprint requests : Prof. D. D’Alessandro, Dip. Scienze di Sanità Pubblica “G. Sanarelli” Università “La Sapienza” P.le Aldo Moro, 5, 00185 Rome (Italy) e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 115-120

Immune responses of a liposome/ISCOM vaccine adjuvant against streptococcal fibronectin binding protein 1 (Sfb1) in mice

J. McArthur*, K. Schulze‡, J. Chin**, B.J. Currie*,**‡, K.S. Sriprakash*,**†‡, S.R. Talay‡‡, G.S. Chhatwal‡‡, C.A. Guzmán‡‡ & M.J. Walker*

*Department of Biological Sciences, University of Wollongong, **NSW Agriculture, Elizabeth Macarthur Agricultural Institute, Camden, ***Menzies School of Health Research, Darwin, †Bacterial Pathogenesis Laboratory, Queensland Institute of Medical Research, ‡Co-operative Research Centre for Aboriginal & Tropical Health, Darwin, Australia & ‡‡GBF National Research Center for Biotechnology, Braunschweig, Germany

Received August 7, 2003

Background & objectives: The fibronectin binding protein Sfb1 of Streptococcus pyogenes is a well characterised antigen which induces protection against lethal challenge with group A streptococcus (GAS) when adjuvanted with cholera toxin B-subunit (CTB). As an alternative to CTB adjuvanted intranasal immunisations we investigated the immune responses generated in mice using Sfb1 incorporated in to the skin and mucosal adjuvant SAMA4. Methods: Mice (BALB/c) were vaccinated intradermally with 100 µl of either SAMA4 (adjuvant only group) or SAMA4/Sfb1 and were boosted 7 days later. Mice vaccinated with CTB based vaccines were immunised by intranasal inoculation with a mixture containing 30 µg Sfb1 and 10 µg CTB on days 1, 3, 5 and 15. At 14 days after the last booster immunisation the immune response was characterised and mice were challenged with 108 CFU of S. pyogenes. Results: Mice vaccinated with SAMA4/Sfb1 elicited a Sfb1-specific IgG response in the sera that was significantly higher than that seen in control mice and mice immunised with the adjuvant only (P<0.05). No significant differences were seen for specific IgA antibodies in the sera in all groups examined. Compared with non-immunised and adjuvant only immunised controls, mice immunised with the Sfb1/SAMA4 vaccine exhibited a significant increase (P<0.05) in the number of Sfb1 reactive spleen cells in lymphoproliferation assays which were three fold higher than those seen for mice vaccinated with the Sfb1/CTB vaccine. Mice vaccinated with CTB/Sfb1 had the highest level of protection (80%) as where mice vaccinated with SAMA4 and SAMA4/Sfb1 displayed no protection (20% and 40%). Interpretation & conclusion: These data suggest that the SAMA4 adjuvant used in this study fails to elicit protective immunity in BALB/c mice when used to adjuvant the known protective antigen Sfb1.

Key words Sfb1 liposome/ISCOM vaccine adjuvant Fibronectin binding proteins of Streptococcus has also been shown to interfere with host clearance pyogenes are believed to mediate adhesion of the mechanisms being mediated by macrophages by binding pathogen to host tissue, a critical step in the initial stages to the Fc fragment of human immunoglobulins3. of infection. Sfb1 is a well characterised fibronectin binding protein of group A streptococcus (GAS). This Fibronectin binding proteins have been proposed as protein mediates bacterial attachment to host cells and potential vaccine targets for preventing GAS internalisation of GAS into nonphagocytic cells1,2. Sfb1 infections4. Antibodies directed against such adhesins 115 116 INDIAN J MED RES (SUPPL) MAY 2004 may prevent bacterial attachment and inhibit a final concentration 1 mM. Liquid cultures were colonisation5. When adjuvanted with cholera toxin B- agitated during incubation in an orbital-shaking incubator subunit (CTB), intranasal immunisation with Sfb1 (Paton Industries) at 180 rpm when required. induces protection against lethal challenge with GAS6. Other characteristics which make Sfb1 an attractive Vaccine preparation and immunisation schedule: candidate antigen are the presence of highly conserved Fusion proteins consisting of N-terminally hexa-histidyl epitopes and the fact it is expressed on the surface of tagged Sfb1 were purified under denaturing conditions GAS in 70 per cent of clinical isolates belonging to using Ni-NTA agarose according to manufacturers different serotypes and strains, independent of their instructions (Qiagen, Chatsworth CA). Formulation of geographic origin7,8. Anti-Sfb1 antibodies do not cross- SAMA4 adjuvant and SAMA4-containing Sfb1 vaccines react with heart proteins and therefore may not trigger were performed as previously described12. Each dose autoimmune reactions that are responsible for post of vaccine contained 50 µg Sfb1 and 0.6 µg of QuilA streptococcal sequelae8. (Advet, Lulea Sweden). Female BALB/c mice aged 10 wk were vaccinated intradermally with 100 µl of either Although ADP-ribosylating bacterial toxins such as SAMA4 (adjuvant only group) or SAMA4/Sfb1 and cholera toxin and Escherichia coli heat-labile were boosted 7 days later. Mice vaccinated with CTB enterotoxin are excellent mucosal adjuvants capable based vaccines were immunised by intranasal inoculation of inducing both systemic and mucosal immune (10 µl/nare) with a mixture containing 30 µg Sfb1 and responses, currently none of these toxins or their 10 µg CTB on days 1, 3, 5 and 15.For the characterisation derivatives have been approved for use in humans9. of the immune response serum samples, spleens and lung Recently these toxins have been shown to target and washes were collected 14 days after the last booster accumulate in olfactory nerves/epithelium and olfactory immunisation. Lung washes were obtained by tracheal bulbs of the central nervous system which has canulation and gentle washing with 0.7 ml of cold PBS highlighted the possibility for neuronal damage through containing 2 mM PMSF. the intranasal use of these proteins in humans10,11. As an alternative to CTB adjuvanted intranasal Enzyme linked immunosorbent assay (ELISA): immunisations we investigated the immune responses Immunoassay plates (96 flat bottom wells, Greiner) were generated in mice using Sfb1 incorporated in to the coated by overnight incubation at 4°C with 50 µl of skin and mucosal adjuvant SAMA4 (a liposome/ISCOM recombinant Sfb1 at 5 µg/ml in carbonate buffer (15 mM combination adjuvant12). The immune responses and sodium carbonate, 35 mM sodium bicarbonate, pH 9.6). protective potential were compared to those elicited Unbound antigen was removed and plates were washed by CTB/Sfb1 antigen/adjuvant combination which is three times with PBS containing 0.05 per cent Tween known to be protective. 20 (PBS-T20). Wells were blocked with 100 µl of PBS containing 0.05 per cent v/v Tween 20 and 1 per cent Material & Methods w/v bovine serum albumin (Sigma, USA) for 1 h at 37°C. After washing three times with PBS containing 0.05 per Bacterial strains, media and growth conditions: cent Tween 20 (PBS-T20), plates were incubated for S. pyogenes NS192 is a blood isolate from Northern 2 h at 37°C with serial dilutions of individual serum or Territory, Australia. GAS was cultured overnight lung wash samples. The plates were again washed, and without shaking at 37ºC in Todd-Hewitt broth (Difco, incubated for 1 h at 37°C with 50 µl of horseradish Detroit, MI) supplemented with 1 per cent yeast extract peroxidase-labelled goat anti-mouse IgG (Kirkegaard and or on 5 per cent sheep blood agar (Microdiagnostics, Perry Laboratories) or IgA (Sigma) secondary antibodies Australia). E. coli strains were routinely grown on Z diluted 1:2000 in PBS-T20. After washing, 50 µl of agar13 or in Luria Bertani (LB) broth14. When required, orthophenylenediamine substrate solution (Sigma) was antibiotics were added to the medium at a concentration added to each well and the reaction was allowed to of 100 µg/ml for ampicillin and 50 mg/ml for kanamycin. proceed at room temperature for 15 min. The reaction

Recombinant protein expression was induced by the was stopped by the addition of 25 µl of 2M H2SO4 and addition of Isopropyl-b-D-galactopyranoside (IPTG) to read at 490 nm with a SpectraMax 250 microplate MCARTHUR et al: STREPTOCOCCAL FIBRONECTIN BINDING PROTEIN 1 (SFB1) IN MICE 117 spectrophotometer (Molecular Devices). Titres for anti- volume, 14 days after the last boost. Mortality was Sfb1 antibodies were determined using the SoftMax recorded daily in all groups. program (Molecular Devices). Results To determine the concentration of total IgG or IgA immunoglobulin present in lung washes, plates were Antigen-specific humoral immune response: coated with rat anti-mouse IgG and IgA (Pharmingen) Intradermal immunisation with SAMA4/Sfb1 or as capture antibodies and were incubated with serial intranasal CTB/Sfb1 elicited a significant serum IgG dilutions of lung wash samples or purified mouse IgG or antibody response to the vaccine antigen when compared IgA to generate standard curves. Captured to the non-vaccinated control mice (P=0.038 and Immunoglobulin was then detected as described above. P=0.029). Anti-Sfb1 IgG titres in serum were 3 times higher in mice vaccinated with CTB/Sfb1 than those seen Lymphocyte proliferation assays: Splenocytes were in mice vaccinated with SAMA4/Sfb1. There was no isolated and released by teasing spleens into 5 ml 1 x significant increase of specific serum IgA in vaccinated RPMI 1640 (Life Technologies). Cells were collected animals (Fig.1a). by centrifugation at 800 g for 10 min and red blood

cells were lysed with 5 ml of 0.83 per cent NH4Cl2, 0.1 per cent KHCO3 and 0.01 M EDTA. Cells were washed twice in 5 ml PBS containing 1 per cent glucose and recovered by centrifugation (800 g for 5 min). Cells were then resuspended in 1 ml Complete Medium, RPMI containing 100 unit Penicillin/100 µg Streptomycin, 2 mM L-Glutamine, 1 mM Sodium Pyruvate, 0.06 mM b-Mercaptoethanol, 10 per cent fetal calf serum. An aliquot of the suspension was diluted tenfold in Trypan Blue and counted using a haemocytometer. The cell concentration was then adjusted to 5 x 106 cells/ml in Complete Medium. A 50 µl volume of the cell suspension (2.5 x 105 cells) was added in triplicate to a 96 well flat bottom cell culture plate (Interpath) containing 200 µl of complete media alone or with 25 µg/ml of Sfb1 antigen. The plate was

incubated for 72 h at 37°C in 5 per cent CO2 in air. At the end of the incubation period, 50 µl (0.5 µCi) of 3H thymidine (Amersham, TRK120, UK) was added to each well and the plate was incubated for a further 18 h. Cells were harvested with a PHD cell harvester (Cambridge Technology) onto glass fibre filter strips. The filters were punched out into insert vials (Pony vials, Packard, 6000286) containing 2 ml scintillant (Ultima Gold scintillation fluid, Packard, 60133). Tubes were then mixed vigorously and radioactivity were Fig.1. The humoral immune response directed against Sfb1 at counted for 1 min with a Packard 4000 series liquid 14 days post last booster immunisation. (a) Specific IgG and IgA scintillation counter (model 4530). present in the serum of control and vaccinated mice. Results are expressed as the geometric means of 5 mice per group. SEM are GAS challenge procedure: S. pyogenes NS192 was indicated by vertical bars. The results are statistically significant (Student’s t test) when compared with the values for control mice passaged through mice 6 times to enhance virulence. (P<0.05) (*). (b) Specific antibodies in lung washes of control and For intranasal challenge, all mice were challenged with vaccinated mice. Results are expressed as the per cent Sfb1-specific 108 CFU of S. pyogenes NS192 delivered in a 40 µl antibodies with respect to total immunoglobulin isotype. 118 INDIAN J MED RES (SUPPL) MAY 2004 Mucosal immune responses were assessed by responses when compared to untreated controls examining the production of Sfb1-specific antibodies in (P=0.039 and P=0.01) (Fig.2). Proliferative responses the respiratory mucosal. Only mice immunised with the were 6.5 times higher for cells isolated from mice CTB based vaccine elicited an increased Sfb1-specific vaccinated with SAMA4 containing Sfb1 when antibody response in the lung. Both IgA and IgG isotypes compared those seen for mice vaccinated with the were seen to increase by a small amount (0.5% to 7%), SAMA4 adjuvant alone. Cells isolated from mice however these increases may not be significant vaccinated with CTB/Sfb1 did not show a significantly (P=0.076) due to the low numbers of mice per group higher proliferative response (Fig.2). (Fig.1b). Determination of protective immunity : After intranasal Antigen-specific cell-mediated immune responses: challenge with S. pyogenes, mice vaccinated with Lymphocytes isolated from the spleens of mice CTB/Sfb1 had the highest level of protection (80% immunised with the SAMA4 adjuvant or SAMA4 survival). Mice vaccinated with SAMA4 and SAMA4/ containing Sfb1 showed significantly higher proliferative Sfb1 showed no protection with only 20 per cent and 40 per cent survival respectively. These were both lower than the level of protection seen in control mice (60%) (Fig.3).

Discussion

ISCOM-based adjuvants have been shown to induce both strong antibody and cellular immune responses in a variety of animal species15. ISCOMs are active after mucosal delivery and induce both local and systemic immune responses and are believed to be the most potent adjuvant for the induction of cell mediated immunity16. A licensed and commercially available ISCOM-based Fig.2. Sfb1-specific proliferative responses of spleen cells. The vaccine is currently used to protect horses from equine results are statistically significant (Student’s t test) when compared influenza and another ISCOM-based influenza vaccine with the values for control mice (P<0.05) (*). Results are expressed is being tested in humans17. The liposome/ISCOM as the mean counts per minute (CPM) of triplicate samples for combination adjuvant, SAMA4 used in the present 3-4 mice per group. Standard error of the mean is indicated by vertical lines. study has been shown to produce high specific antibodies in serum and at mucosal sites such as the lung and large intestine12,18. SAMA4 has also been shown to promote good memory responses and cytotoxic T-cell activity12. When used to adjuvant outer- membrane proteins from Actinobacillus pleuropneumoniae, a SAMA4 based vaccine was shown to induce protective immunity against intranasal challenge with the wild-type organism18. Although the exact mechanisms for adjuvant activity of ISCOM- based vaccines is not known, the synergism between SAMA4 and antigen components is not dependent on the antigen type and is not a particular attribute of the antigen used19. Fig.3. Survival of mice after intranasal challenge with S. pyogenes strain NS192. Animals (n=10) were challenged with 108 CFU of NS192, 14 days post last boost and mortality was recorded Intradermal immunisation with SAMA4/Sfb1 resulted daily. in a significant increase in antigen specific serum IgG. MCARTHUR et al: STREPTOCOCCAL FIBRONECTIN BINDING PROTEIN 1 (SFB1) IN MICE 119 However, this was 3 times lower than those seen in mice References vaccinated with CTB/Sfb1. SAMA4/Sfb1 vaccinated animals failed to elicit a mucosal immune response in 1. Molinari G, Talay SR, Valentin-Weigand P, Rohde M, Chhatwal the lung, but as expected exhibited a significant increase GS. The fibronectin-binding protein of Streptococcus pyogenes, in the number of Sfb1-reactive spleen cells in lymphocyte SfbI, is involved in the internalization of group A streptococci by epithelial cells. Infect Immun 1997; 65 : 1357-63. proliferation assays. Mice vaccinated with CTB/Sfb1 produced an increased Sfb1-specific antibody response 2. Talay SR, Valentin-Weigand P, Jerlstrom PG, Timmis KN, in the lung, where as cells isolated from the spleens of Chhatwal GS. Fibronectin-binding protein of Streptococcus pyogenes: sequence of the binding domain involved in adherence these mice failed to show a significant proliferative of streptococci to epithelial cells. Infect Immun 1992; 60 : response. This is comparable to the results reported by 3837-44. Schulze et al20 where mice were intranasally immunised 3. Medina E, Molinari G, Rohde M, et al. Fc-mediated nonspecific with various fragments of Sfb1 using CTB as an adjuvant binding between fibronectin-binding protein I of Streptococcus also failed to produce a significant increase in reactive pyogenes and human immunoglobulins. J Immunol 1999; 163 : lymphocytes isolated from spleens. 3396-402. 4. Courtney HS, Dale JB, Hasty DL. Strategies for preventing Mice vaccinated with CTB/Sfb1 have previously group A streptococcal adhesion and infection. In: An YH, been shown to be protected against lethal intranasal Friedman RJ, editors. Handbook of bacterial adhesion: 6 principles, methods and applications. Totowa, NJ: Humana challenge with GAS . This was again demonstrated in Press Inc; 2000 p. 553-79. this study where mice vaccinated with CTB/Sfb1 displayed the highest level of protection (80%). 5. Wizemann TM, Adamou JE, Langermann S. Adhesins as targets for vaccine development. Emerg Infect Dis 1999; 5 : 395-403. Protection in these animals have been correlated with the level of Sfb1-specific antibodies in serum and lung 6. Guzman CA, Talay SR, Molinari G, Medina E, Chhatwal GS. Protective immune response against Streptococcus pyogenes in washes with secretory IgA believed to be critical for mice after intranasal vaccination with the fibronectin-binding immunity6. Mice receiving SAMA4 based vaccines protein SfbI. J Infect Dis 1999; 179 : 901-6. displayed no mucosal immunity and had the lowest 7. Talay SR, Valentin-Weigand P, Timmis KN, Chhatwal GS. levels of protection. As these mice also displayed a Domain structure and conserved epitopes of Sfb protein, the significant increase in the number of Sfb1-reactive fibronectin- binding adhesin of Streptococcus pyogenes. spleen cells, cell mediated immunity does not appear to Mol Microbiol 1994; 13 : 531-9. play a role in protective immunity against an intranasal 8. Valentin-Weigand P, Talay SR, Kaufhold A, Timmis KN, GAS challenge. In endemic areas, the incidence of Chhatwal GS. The fibronectin binding domain of the Sfb protein streptococcal skin and respiratory infections decrease adhesin of Streptococcus pyogenes occurs in many group A streptococci and does not cross-react with heart myosin. with increasing age and this corresponds to increasing Microb Pathog 1994; 17 : 111-20. levels of anti-streptococcal serum antibodies21. 9. Kenney RT, Regina Rabinovich N, Pichyangkul S, Price VL, However, mice vaccinated with SAMA4 based Engers HD. 2nd meeting on novel adjuvants currently in/close vaccines also displayed an increased susceptibility to to human clinical testing. World Health Organization- GAS infection as demonstrated by having lower levels Organization Mondiale de la Sante Fondation Merieux, Annecy, of protection than those seen for control mice. A reason France, 5-7 June 2000. Vaccine 2002; 20 : 2155-63. why SAMA4 may exacerbate GAS infections may be 10. Van Ginkel FW, Jackson RJ, Yuki Y, McGhee JR. Cutting edge: due to the systemic toxic effects of QuilA (a semi- the mucosal adjuvant cholera toxin redirects vaccine proteins into olfactory tissues. J Immunol 2000; 165 : 4778-82. purified saponin component of the adjuvant) which has been reported by others19,22,23. 11. Fujihashi K, Koga T, van Ginkel FW, Hagiwara Y, McGhee JR. A dilemma for mucosal vaccination: efficacy versus toxicity using enterotoxin-based adjuvants. Vaccine 2002; 20 : 2431-8. These data suggests that the skin and mucosal 12. Chin J, San Gil F, Novak M, et al. Manipulating systemic and adjuvant, SAMA4 used in this study fails to elicit mucosal immune responses with skin-deliverable adjuvants. protective immunity in BALB/c mice when used to J Biotechnol 1996; 44 : 13-9. adjuvant the known protective antigen Sfb1 and cannot 13. Walker MJ, Birch R, Pemberton JM. Cloning and be considered as an alternative for the mucosal adjuvant, characterisation of an albicidin resistance gene from Klebsiella CTB in this situation. oxytoca. Mol Microbiol 1988; 2 : 443-54. 120 INDIAN J MED RES (SUPPL) MAY 2004

14. Sambrook J, Fritsch E, Maniatis T. Molecular cloning: a with a liposome-iscom adjuvanted vaccine. Scand J Immunol laboratory manual. 2nd ed. Cold Spring Harbour Press; 1989. 1998; 47 : 243-53.

15. Sjolander A, Drane D, Maraskovsky E, et al. Immune responses 20. Schulze K, Medina E, Talay SR, et al. Characterization of the to ISCOM formulations in animal and primate models. Vaccine domain of fibronectin-binding protein I of Streptococcus 2001; 19 : 2661-5. pyogenes responsible for elicitation of a protective immune 16. O’Hagan DT, MacKichan ML, Singh M. Recent developments response. Infect Immun 2001; 69 : 622-5. in adjuvants for vaccines against infectious diseases. Biomol Eng 2001; 18 : 69-85. 21. Brandt ER, Hayman WA, Currie B, et al. Opsonic human antibodies from an endemic population specific for a conserved 17. Rimmelzwaan GF, Nieuwkoop N, Brandenburg A, et al. A epitope on the M protein of group A streptococci. Immunology randomized, double blind study in young healthy adults 1996; 89 : 331-7. comparing cell mediated and humoral immune responses induced by influenza ISCOM vaccines and conventional vaccines. Vaccine 22. Watson DL, Lovgren K, Watson NA, et al. Inflammatory 2000; 19 : 1180-7. response and antigen localization following immunization with 18. San Gil F, Turner B, Walker MJ, Djordjevic SP, Chin JC. influenza virus ISCOMs. Inflammation 1989; 13 : 641-9. Contribution of adjuvant to adaptive immune responses in mice 23. Fossum C, Bergstrom M, Lovgren K, Watson DL, Morein B. against Actinobacillus pleuropneumoniae. Microbiology 1999; Effect of iscoms and their adjuvant moiety (matrix) on the initial 145 : 2595-603. proliferation and IL-2 responses: comparison of spleen cells 19. San Gil F, Turner B, Mullbacher A, et al. Flow cytometric from mice inoculated with iscoms and/or matrix. Cell Immunol analysis of cellular changes in mice after intradermal inoculation 1990; 129 : 414-25.

Reprint requests: Dr M. J. Walker, Department of Biological Sciences, University of Wollongong, N.S.W. 2522, Australia e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 126-130

Antibody classes & subclasses induced by mucosal immunization of mice with Streptococcus pyogenes M6 protein & oligodeoxynucleotides containing CpG motifs

R. Teloni, C. von Hunolstein, S. Mariotti, S. Donati, G. Orefici & R. Nisini

Istituto Superiore di Sanità, Rome, Italy

Received August 7, 2003

Background & objectives: Type-specific antibodies against M protein are critical for human protection as they enhance phagocytosis and are protective. An ideal vaccine for the protection against Streptococcus pyogenes would warrant mucosal immunity, but mucosally administered M-protein has been shown to be poorly immunogenic in animals. We used a recombinant M type 6 protein to immunize mice in the presence of synthetic oligodeoxynucleotides containing CpG motifs (immunostimulatory sequences: ISS) or cholera toxin (CT) to explore its possible usage in a mucosal vaccine. Methods: Mice were immunized by intranasal (in) or intradermal (id) administration with four doses at weekly intervals of M6-protein (10 µg/mouse) with or without adjuvant (ISS, 10 µg/mouse or CT, 0,5 µg/mouse). M6 specific antibodies were measured by enzyme linked immunosorbent assay using class and subclass specific monoclonal antibodies. Results: The use of ISS induced an impressive anti M-protein serum IgG response but when id administered was not detectable in the absence of adjuvant. When used in, M-protein in the presence of both ISS and CT induced anti M-protein IgA in the bronchoalveolar lavage, as well as specific IgG in the serum. IgG were able to react with serotype M6 strains of S. pyogenes. The level of antibodies obtained by immunizing mice in with M-protein and CT was higher in comparison to M-protein and ISS. The analysis of anti-M protein specific IgG subclasses showed high levels of IgG1, IgG2a and IgG2b, and low levels of IgG3 when ISS were used as adjuvant. Thus, in the presence of ISS, the ratio IgG2a/IgG1 and (IgG2a+IgG3)/IgG1 >1 indicated a type 1-like response obtained both in mucosally or systemically vaccinated mice. Interpretation & conclusion: Our study offers a reproducible model of anti-M protein vaccination that could be applied to test new antigenic formulations to induce an anti-group A Streptococcus (GAS) vaccination suitable for protection against the different diseases caused by this bacterium.

Key words Anti GAS vaccination - M6 protein - mucosal immunity - Streptococcus pyogenes

M protein is considered the main determinant factor two main factors: the high number of different, non cross- present on the surface of group A Streptococcus reactive serotypes and the fact that some portions of (GAS)1. In humans and animals only M protein type- the protein (B repetitive domain) contain autoantibodies- specific antibodies have the capacity to opsonize inducing epitopes against cardiac and skeletal tissues2. streptococci and to override the anti-phagocytic property An ideal vaccine for the protection against GAS would of M protein. The development of an efficacious and warrant muscosal immunity, but mucosally administered safe vaccine against GAS infections was prevented by M-protein has been shown to be poorly immunogenic in 126 TELONI et al: MUCOSAL IMMUNIZATION OF MICE WITH S. PYOGENES M6 PROTEIN 127 animals3,4. Several adjuvants or carrier proteins such as µg/mouse of ISS or M-ODN in a total volume of 66 µl Freund adjuvant (Fa), cholera toxin (CT), diphteria toxoid (33 µl/naris). Drops of vaccine were applied to external (Dt) have been used earlier to increase the antigenicity nares of mice after a light anaesthesia with 10 µl/g of GAS surface antigens. Because of the relative toxicity intraperitoneal (ip) administration of a solution containing of Fa and CT, only Dt could be used in humans. It has 5 mg/ml ketamin (Ketavet; Farmaceutici Gellini S.p.A., been shown that synthetic oligodeoxynucleotides (ODN) Latina, Italy) and 0.3 mg/ml xilazin (Rompun; Bayer containing unmethylated CpG dinucleotides in particular S.p.A., Leverkusen, Germany). base contexts (immuno-stimulating sequences, ISS) cause activation of B and NK cells as well as antigen presenting Collection of samples: From mice immunized three cells. When co-administered to experimental animals with times on days 0, 10 and 20; sera and bronchoalveolar proteins or peptides, ISS act as potent adjuvants for a lavage (BAL) were collected on day 30. Plasma was type-1 immune response. A particular advantage of ISS collected after a light anaesthesia by retroorbital puncture as adjuvants is their activity both in parenteral 30 days after the first inoculum. Samples were 5 (intramuscular, subcutaneous, sc and intradermal, id) immediately frozen at -35ºC until use. and mucosal (intranasal, in, oral and intrarectal) administration6. The aim of the present study was to BAL fluids were obtained through the insertion of a define the possible use of an adjuvant that could be used cannula into the trachea after cervical dislocation. Lungs without conjugation, both by parenteral and mucosal were then flushed with 0.6 ml of saline. The return was immunization, to increase the immunogenicity of M protein. We explored the potential of ISS to induce anti- spun to remove cellular debris and immediately frozen M6 antibody response when used as a systemic and at -35ºC until IgA assay. mucosal adjuvant7,8. The local and systemic immune response was studied to compare the activity of ISS to Evaluation of the immune response: The antibody that of cholera toxin, a widely used adjuvant in response to M6 protein in sera was determined by experimental mucosal vaccination that induces a Th-2 enzyme linked immunosorbent assay (ELISA) as type of immune response. previously described10. Briefly, M6 protein was used at 1µg/ml in PBS (100 µl/well) to coat flat-bottomed 96- Material & Methods well plates (Dynatech, Chantilly, VA) overnight at +4º C. After a postcoat of 1 h at +4ºC with PBS containing 2 Reagent and antigens: Single-stranded, per cent bovine serum albumin (BSA), sera (non-immune phosphorothioated CpG motif containing ODN and control and test sera) were added to appropriate wells control ODN (M-ODN) were synthesized (M-Medica at different dilutions in PBS with 0.5 per cent BSA Firenze, Italy) according to published sequences9 (5’- containing 0.05 per cent Tween 20 (PBS-T). After TGACTGTGAACGTTCGAGATGA-3’ and 5- overnight incubation at +4ºC and washing, plates were TGACTGTGAAGCATCGAGATGA-3’ respectively). incubated with anti-IgG (g chain specific) goat anti-mouse The cholera toxin (CT) purified from Vibrio cholerae antibodies conjugated with alkaline phosphatase was purchased from Sigma (St. Louis, MO, USA). The (Southern Biotechnology Associates, Birmingham, AL). recombinant M6 protein was kindly provided by V. After 1h incubation and washings, p-nitrophenylphosphate Fischetti4. (Sigma, St. Louis, MO, USA) was added as a substrate. The reaction was blocked with 3 M NaOH and the Immunization protocols: Groups (n=6) of female absorbance evaluated at 405 nm using the Microtitre BALB/c mice (6-8 wk old; Charles River, Calco, Lecco, reader Victor-1420 multilabel counter (EG & G/Wallac, Italy) were immunized by intradermal (id) injection of Turku, Finland). All tests were designed to compare in M6 protein alone (10 µg/mouse) or in combination with the same ELISA plate sera from mice belonging to ISS or M-ODN at 50 µg in a total volume of 50 or 100 different experimental groups. µl. ISS were administered in the first inoculum only. Groups (n=6) of the same animals were immunized by Subclasses of M6 protein IgG positive sera were intranasal (in) inhalation of 10 µg/mouse of M6 protein determined using rat anti-mouse IgG subclass biotinilated alone or in combination with 0.5 µg/mouse of CT, 10-20 monoclonal antibodies (PharMingen, San Diego, CA, 128 INDIAN J MED RES (SUPPL) MAY 2004 USA) and horseradish peroxidase-conjugated response. IgG titres were very high, when ISS were streptavidin (PharMingen). After the addition of used as adjuvants for a systemic (id) vaccination (Fig.1). o-phenylenediamine (Sigma fast; Sigma, USA) the However, in mice vaccinated id with ISS, M6 specific

reaction was stopped with 2N H2SO4 and the absorbance IgA were not detected in BAL. The specificity of the (ABS) was evaluated with the plate reader at 490 nm. response was determined using both ELISA (Figs 1-4) and radial immunodiffusion methods (Table). The anti-M6 protein IgA production in sera and BAL was evaluated using a rat anti-mouse IgA biotinilated Mice vaccinated in with recombinant M6 protein in monoclonal antibody (PharMingen, San Diego, CA) and the absence of adjuvant did not develop a specific Ig incubated for 1 h at +37ºC. After washings, horseradish response. Intranasal M-6 vaccination in the presence peroxidase-conjugated streptavidin was added and the of ISS induced a good systemic IgG response (Fig.1) reaction was evaluated as subclass IgG assay. Endpoint and mucosal IgA response (Fig.2). When using ISS, titres were expressed as the reciprocal log2 of the last however, mean titres of BAL anti-M-6 IgA were lower serum or BAL dilution, which gave an optical density than those obtained immunizing mice in the presence of (OD), at 405 or 490 nm higher than the cut-off. The CT (Fig.2) using the same immunization protocol. The ABS cut-off value was determined as the double of the use of ISS induced a relative higher synthesis of anti-M- background and it was always higher than the mean ABS 6 IgG2b than IgG2a. This phenomenon, that is indicative value of sera from untreated mice at the lowest dilution. of a Thl-like immune response, was observed when ISS were used as systemic adjuvants (Fig.3). When ISS were Reactivity of anti-M protein serum: The reactivity of used as mucosal adjuvants, IgG2a had the same mean the anti-M protein serum was tested by the Ouchterlony double diffusion test performed with 1 per cent agarose Table. Specificity of M protein response by immunodiffusion and gel with sodium barbital buffer, pH 7.4, using the hot ELISA acidic M protein extract (according to the M serotyping M antigen extract from Non immune serum Anti M protein procedure of Lancefield)11 prepared from the strains GAS M1 – ISS 650 negative negative belonging to the Istituto Superiore di Sanità (ISS) collection which included Group A streptococcus M1: GAS M3 – ISS 651 negative negative ISS-650, Group A streptococcus M3: ISS-651, Group A GAS M6–ISS 654 negative positive streptococcus M6:-ISS-654, Group G streptococcus: ISS- GGS –ISS 3833 negative negative 3833, Group B streptococcus serotype III: ISS-2601. GBS serotype III - ISS 2601 negative negative Statistical analysis: The data were expressed as arithmetic means of at least four independent experiments ± standard error of the mean (SEM). The data were analyzed for normal distribution and the statistical significance of the difference between groups was determined by the two-tailed unpaired Student’s t-test using the Statview 4.1 program (Abacus Concepts, Berkeley, CA, USA). Differences were considered significant with P < 0.05.

Results

Mice vaccinated id with recombinant M6 protein in the absence of adjuvant or in the presence of CT or M- Fig. 1. Anti M6 protein serum IgG titres. Mice were immunized with M6 in the absence (no treatment) or in the presence of ISS or ODN did not develop a specific IgG response (data not CT as systemic (id) or mucosal (in) adjuvant. Data are expressed as shown). One the other hand, when ISS were used, mice mean log2 reciprocal titres ± SEMn of three independent experiments developed a detectable anti-M6 protein specific immune using six mice for each experimental group. TELONI et al: MUCOSAL IMMUNIZATION OF MICE WITH S. PYOGENES M6 PROTEIN 129 titre than IgG2b. However, the ratio IgG2a/IgG1 (Fig.4) as well as the ratio (IgG2a+IgG3)/IgG1 (data not shown) were >1 when ISS were used both as systemic and mucosal adjuvant, indicating that ISS induce a Th1-like immune response10. On the other hand, CT induced a relative higher synthesis of anti M-6 IgGl and IgG2a than IgG2b (Fig.3). This pattern of M6 specific IgG subclass and the ratio IgG2a/IgGl < 1 (Fig.4) confirmed that CT favored a Th2-like immune response.

Discussion Fig 2. Anti M6 protein IgA titres. Mice were immunized with M6 in the absence (no treatment) or in the presence of ISS or CT as The wide range of streptococcal diseases, which mucosal (in) adjuvants. BAL were obtained by tracheal washing. include suppurative infections, toxin mediated diseases Data are expressed as mean log reciprocal titres ± SEM of three 2 and non-suppurative sequelae makes it difficult to clearly independent experiments using six mice for each experimental group. define the relevance of the different role of specific immunity in protection or autoimmunity induction12-15.

The main GAS virulence factor, M protein, is known to have anti-opsonic activity and some portions of this protein (B repetitive domain) contains autoantibodies inducing-epitopes against cardiac and skeletal tissues. The design of a vaccine to prevent GAS mediated diseases should take into account all these problems and together with the research for suitable antigens, it should also consider the prevalent type of immune response (Th1 or Th2) raised by the antigen-adjuvant combination Fig. 3. Anti M6 protein specific IgG subclasses. Mice were used. immunized with M6 in the absence (no treatment) or in the presence of ISS or CT as systemic (id) or mucosal (in) adjuvants. Data are In the present study it was confirmed that CT is a expressed as mean log reciprocal titres ± SEM of three independent 2 strong mucosal adjuvant16 for GAS M protein. In experiments using six mice for each experimental group. addition, we demonstrated that ISS showed strong systemic and mucosal adjuvant activity when co- administered with M protein. Both adjuvants did not require conjugation with the protein to induce significant antibody responses. CT induced a Th2 response, as indirectly demonstrated by the prevalent production of IgG1 over IgG2a specific anti-M protein IgG subclass. On the other hand, the IgG2a/IgG1 ratio was clearly >1 in mice receiving ISS as adjuvants, thus indicating that ISS favor an anti-M protein Th1 immune response. In addition, ISS, but not CT, were shown to strongly induce anti-M protein specific IgG response irrespective of the systemic or mucosal route of vaccination. In conclusion, our study offers a reproducible model of Fig 4. ISS induce a Thl-like immune response. The mean titres of anti-M protein vaccination that could be applied to test IgG2a and IgG1 obtained in three independent experiments of systemic and mucosal immunization using ISS or CT as adjuvants. new antigenic formulations to induce an anti-GAS The ratio >1 indicates a Thl immune response, while the ratio <1 is vaccination suitable for protection against the different indicative of a Th2 immune response. diseases caused by this bacterium. 130 INDIAN J MED RES (SUPPL) MAY 2004 References motif induces an anti-polysaccharide type 1-like immune response after immunization of mice with Haemophilus influenzae type b conjugate vaccine. Int Immunol 2000; 12 : 1. Saint S, Matthay MA. Risk reduction in the intensive care unit. 295-303. Am J Med 1998; 105 : 515-23. 9. Roman M, Martin-Orozoo E, Goodman JS, Nguyen MD, 2. Dale JB. Group A streptococcal vaccines. Infect Dis Clin Sato Y, Ronaghy A, et al. Immunostimulatory DNA sequences North Am 1999; 13 : 227-43. function as T helper-1-promoting adjuvants. Nat Med 1997; 3. Hayman WA, Brandt ER, Relf WA, Cooper J, Saul A, Good 3 : 849-54. MF. Mapping the minimal murine T cell and B cell epitopes 10. Mariotti S, Teloni R, von Hunolstein C, Romagnoli G, Orefici within a peptide vaccine candidate from the conserved region of G, Nisini R. Immunogenicity of anti-Haemophilus influenzae the M protein of group A streptococcus. Int Immunol 1997; type b CRM197 conjugate following mucosal vaccination with 9 : 1723-33. oligodeoxynucleotide containing immunostimulatory sequences 4. Bolken TC, Franke CA, Jones KF, Bell RH, Swanson RM, as adjuvant. Vaccine 2002; 20 : 2229-39. King DS, et al. Analysis of factors affecting surface expression 11. Cayeux P, Wahl R. New method of extraction and purification and immunogenicity of recombinant proteins expressed by of the M protein of Streptococcus pyogenes of the a group Gram-positive commensal vectors. Infect Immun 2002; 70 : applied to type 24. Ann Inst Pasteur (Paris) 1963; 105 : 1063- 2487-91. 70. 5. Kobayashi H, Horner AA, Takabayashi K, Nguyen MD, 12. Weiss KA, Laverdiere M. Group A streptococcus invasive Huang E, Cinman N, et al. Immunostimulatory DNA infections: a review. Can J Surg 1997; 40 : 18-25. pre-priming: a novel approach for prolonged Th1-biased immunity. Cell Immunol 1999; 198 : 69-75. 13. Norrby SR, Norrby-Teglund A. Infections due to group A streptococcus: new concepts and potential treatment strategies. 6. Horner AA, Raz E. Immunostimulatory sequence Ann Acad Med Singapore 1997; 26 : 691-3. oligodeoxynucleotide: A novel mucosal adjuvant. Clin Immunol 2000; 95 : S19-29. 14. Gonzalez MH. Necrotizing fasciitis and gangrene of the upper extremity. Hand Clin 1998; 14 : 635-45. 7. von Hunolstein C, Mariotti S, Teloni R, Alfarone G, Romagnoli G, Orefici G, et al. The adjuvant effect of synthetic 15. Bannan J, Visvanathan K, Zabriskie JB. Structure and function oligodeoxynucleotide containing CpG motif converts the anti- of streptococcal and staphylococcal superantigens in septic Haemophilus influenzae type b glycoconjugates into efficient shock. Infect Dis Clin North Am 1999; 13 : 387-96. anti-polysaccharide and anti-carrier polyvalent vaccines. Vaccine 16. Marinaro M, Staats HF, Hiroi T, Jackson RJ, Coste M, 2001; 19 : 3058-66. Boyaka PN, et al. Mucosal adjuvant effect of cholera toxin in 8. von Hunolstein C, Teloni R, Mariotti S, Recchia S, Orefici G, mice results from induction of T helper 2 (Th2) cells and IL-4. Nisini R. Synthetic oligodeoxynucleotide containing CpG J Immunol 1995; 155: 4621-9.

Reprint requests: Dr Roberto Nisini, Lab Batteriologia e Micologia Medica, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161, Rome, Italy e-mail Indian J Med Res 119 (Suppl) May 2004, pp 131-135

Receptor-mediated endocytosis of biofilm-forming Enterococcus faecalis by rat peritoneal macrophages

Baldassarri Lucilla, Bertuccini Lucia, Ammendolia Maria Grazia, Cocconcelli Pierluigi* Arciola Carla Renata**, Montanaro Lucio**, Creti Roberta† & Orefici Graziella†

Laboratorio di Ultrastrutture and †Laboratorio di Batteriologia e Micologia Medica, Istituto Superiore di Sanità, Rome; **Istituti Ortopedici Rizzoli, Bologna; & *Università Cattolica, Piacenza, Italy

Received August 7, 2003

Background & objectives: Enterococci are important nosocomial pathogens that are increasingly difficult to treat due to intrinsic and acquired resistance to antibiotics. Studies were taken up to identify virulence factors and to characterise pathogenic mechanisms of such infections to evaluate potential targets for treatments alternative to antibotic therapy. This study was carried out to evaluate the contribution of extracellular polysaccharide expressed by Enterococcus faecalis to resistance to phagocytosis and survival within rat peritoneal macrophages. Methods: Six E. faecalis clinical isolates were tested for their ability to survive within rat peritoneal macrophages. Cytochalasin D, colchicine and monodansylcadaverine were used to investigate the route of enterococcal entry inside macrophages. Results: Four of the isolates were able to produce extracellular polysaccharide and form biofilm after growth in glucose-supplemented medium, while no production could be detected in glucose deficient medium. Two isolates were polysaccharide-negative in both conditions. Isolates expressing extracellular polysaccharide were able to survive for more than 24 h compared to polysaccharide-negative bacterial cells of the same strain grown in glucose-deficient medium, which were readily cleared. Cytochalasin D virtually abolished the number of viable intracellular bacteria, after growth in either trypticase soy broth (TSB) or TSB supplemented with glucose; colchicine and monodansylcadaverine strongly affected survival of polysaccharide-positive bacteria, significantly more than that of polysaccharide-negative ones. Interpretation & conclusion: Biofilm-forming E. faecalis survived within rat peritoneal macrophages significantly better than polysaccharide-negative isolates. Perturbators of cytoskeleton and of surface receptors turnover, indicated receptors-mediated endocytosis as the most likely route for enterococcal entry into macrophages.

Key words Biofilm - endocytosis - enterococci

Enterococci are important opportunistic pathogens, Virulence mechanism of this microorganisms is not representing one of the leading causes of endocarditis well understood although some putative virulence factors and nosocomial bacteremia in the United States and have been described. Besides surface proteins and northern Europe1-3. Almost all such infections are caused polysaccharide possibly involved in virulence, the ability by the two species Enterococcus faecalis and of E. faecalis to survive inside polymorphonuclear Enterococcus faecium, responsible for 80 and 20 per leucocytes (PMN) and macrophages has been cent of infections, respectively. examined4-6. Translocation through the intestinal barrier 131 132 INDIAN J MED RES (SUPPL) MAY 2004 has long been indicated as one of the preferential portal of 5 ml of Hank’s balanced salt solution (HBSS). of entry of enterococci for spreading to distant sites7; Recovered cell viability was determined by trypan blue the ability of E.faecalis to survive within macrophages dye exclusion. Cell suspensions, normalised to 2.5X105 may contribute to pathogenesis by allowing translocation cells/ml in RPMI 1640 supplemented with 1 per cent of bacteria through the intestinal mucosa and transport foetal calf serum (FCS), were dispensed into 24 well to different body parts. Further, resistance to killing by plates and incubated for 30 min at 37°C under 5 per the PMN may represent a way to escape host defenses. cent CO2. Unattached cells were discarded by washing and macrophages further incubated for 24 h at 37°C Aggregation substance (AS) has been indicated to with RPMI 1640, 1 per cent FCS, before infection. In be responsible for internalisation within PMN through a experiments with inhibitors, all compounds were added preferential route4, allowing resistance to killing. We to macrophages 30 min before infection and were suggested that the ability to synthesise extracellular present in the cell/bacteria suspensions throughout the polysaccharides (slime) might mediate such alternate experiments. phagocytosis leading to enterococcal survival within peritoneal macrophages6. Electron microscopy: Infected macrophages were fixed with 2.5 per cent glutaraldehyde in 0.1 M sodium This study was done to examine the ability of slime- cacodylate buffer (pH 7.4) for 1 h and postfixed with forming and non-slime forming E. faecalis isolates to 1 per cent OsO4 in 0.1 M sodium cacodylate buffer for 1 survive within rat peritoneal macrophages, and to h at room temperature8. Fixed specimens were washed, characterise the route of entry by using perturbators dehydrated through a graded series of ethanol solutions of cytoskeleton and inhibitors of surface-receptor (30 to 100% ethanol) and embedded in Agar 100 (Agar turnover. Scientific Ltd., UK). Ultrathin sections obtained by a MT-2B Ultramicrotome (LKB Pharmacia, Sweden) Material & Methods were stained with uranyl acetate and lead citrate and examined by an EM 208 Philips electron microscope Bacterial strains and media: Six E. faecalis isolates (The Netherlands). of different origin (endocarditis 2, catheter 2, biliary stent 1, environment 1) were examined in this study. The Chemicals: Cytochalasin D and colchicine, diluted in isolates were grown in plain trypticase soy broth (TSB, phosphate buffered saline (pH 7.4), and OXOID, Basingstoke, UK) or in TSB supplemented with monodansylcadaverine (MDC), dissolved in dimethyl 1 per cent glucose (TSBG). sulphoxide (DMSO), were prepared as stock solutions and added to the cell culture medium at 1 and 5, 5 and Biofilm formation: To test for biofilm formation a 100 µg/ml, respectively. quantitative adherence assay6 was used. Briefly, 1:100 dilution of overnight cultures in TSB was used to inoculate All inhibitors were tested for possible effects on cell wells in a microtiter polystyrene plate containing different and bacterial viability by trypan blue exclusion test and media. After growth for 18 h at 37°C, plates were gently cfu counting, respectively, of treated and untreated washed three times with phosphate buffered saline (PBS) samples. pH 7.4, the adherent bacterial film was fixed by air drying at 60°C for 1h and then stained with Hucker’s crystal Results violet; excess stain was washed off with tap water. The optical density of the biofilm was measured at 570 nm Biofilm formation: Ability to form biofilm was tested in an automatic spectrophotometer (NovapathTM by the quantitative plate test after growth in plain TSB Microplate Reader, BioRad Laboratories Inc., CA, USA). or in TSBG medium. Four strains were slime producers when grown in TSB supplemented with 1per cent glucose Murine macrophages: Peritoneal macrophages were (TSBG), but slime-negative in absence of additional harvested from adult female rats as described6. Briefly, glucose; isolates EFS 38 and EFS 90 were biofilm a peritoneal lavage was performed with 2 applications negative in all conditions (Table I). BALDASSARRI et al : MACROPHAGE PHAGOCYTOSIS OF E. FAECALIS 133

Table I. Source and virulence factors pattern of the six E. faecalis isolates used in this study

Isolates Source pAD1 pAM373 CylA esp gelE Slime TSB TSBG EFS 9 Endocarditis + - - + - NP FP EFS 38 Biliary stent - - - + - NP NP EFS 44 Catheter - + - - - NP FP EFS 57 Environment - - - + - NP FP EFS 85 Catheter + - + - + NP FP EFS 90 Endocarditis - - - - + NP NP

pAD1, pAM 373-aggregation factor; CylA, cytolisin; esp, enterococcal surface protein; gelE, gelatinase; TSB, tripticase soy broth; TSBG, TSB supplemented with 1% glucose; NP, non-biofilm producing; FP, biofilm producing

Survival of E. faecalis inside rat peritoneal macrophages: The isolates examined belonged to our laboratory collection, which had been characterised by PCR for the presence of genes coding putative virulence factors including the aggregation factor (AS) pAD1 and pAM373, cytolisin (cylA), gelatinase (gelE) and enterococcal surface protein (esp) (Table I). Survival ability, expressed as percentage of the inoculum recovered from infected macrophages at 24 h, varied Fig.1. Well preserved enterococcal cells visible inside macrophages after 24 h infection, either in single-cell tight vacuoles (a) or in larger between isolates. In all cases, however, survival was vacuoles containing numerous bacteria (b). enhanced for glucose-grown bacteria compared to those grown in glucose-free medium (Table II). We could not find correlations with any of the genetic traits considered.

Well preserved bacterial cells were observed inside macrophage vacuoles at 24 h after infection, either in single-cell or in multiple cells-containing vacuoles (Fig.1).

Route of enterococcal entry into macrophages: Different inhibitors of cellular structures were used to examine the mechanism of enterococcal entry into the cells. Isolate EFS 85, which showed the highest survival Viable bacteria (% of control) levels, was used; EFS 85 was grown in TSB and/or TSBG to determine the effect of the extracellular polysaccharides on phagocytosis. Cytochalasin D, at TSB TSBG concentration of 5 µg/ml, virtually abolished the number Fig.2. Effect of surface receptors turnover and cytoskeleton of viable intracellular bacteria, after growth in either TSB pretubators on EFS 85 survival after growth in TSB (polysaccharide- or TSBG (97.5 and 99%, respectively, Fig.2). Colchicine negative cells) and TSBG (polysaccharide-expressing cells). Cytocalasin D at 1 µg/ml .. ; cytocalasin D at 5 µg/ml ; colchicine 123 reduced TSBG grown-EFS 85 entry into macrophages 123 5 µg/ml123 ; monodansylcadaverine (MDC) 100 µg/ml ; control by 95 per cent, while the effect on polysaccharide-lacking bar . Control is normalised to 100% and the effect of chemicals cells was weaker (27%). MDC also dramatically redued indicates the percentage of inhibition. Results shown are the mean of the number of viable intracellular polysaccharide-positive at least three experiments performed in duplicate. 134 INDIAN J MED RES (SUPPL) MAY 2004

Table II. Survival of different E. faecalis isolates inside rat peritoneal polysaccharide induced by the presence of an additional macrophages carbohydrate source in the medium. Isolate Medium T T T 0 3 24 To determine the route of entry of enterococci, (% of the inoculum) inhibitors of the surface-receptors turnover and EFS 9 TSB 1.2 0.002 - cytoskeleton perturbators were used. As known, TSBG 16.2 0.0125 0.002 polymerized filamentous actin contributes to the process EFS 38 TSB 0.33 0.009 - of phagocytosis, while microtubules take part in the transportation of endosomes and lysosomes within the TSBG 0.40 0.014 0.012 cells9,10. MDC is a frequently used inhibitor of receptor- EFS 44 TSB 0.53 0.006 - mediated endocytosis, which acts through the increase TSBG 0.60 0.09 0.02 of lysosomal pH and the decrease of receptor turnover EFS 57 TSB 1.3 0.006 - rate9. As expected, cytochalasin D, which inhibits the 11 TSBG 12.2 0.022 0.004 polymerisation of filamentous actin , dramatically affected the entry of enterococci within the macrophages. EFS 85 TSB 5.0 0.011 0.003 Interestingly, colchicine, which blocks the elongation of microtubules11, and MDC9, which slows down the TSBG 18.5 0.65 0.076 surface receptors turnover, strongly inhibited entry/ EFS 90 TSB 0.13 0.0056 - survival of polysaccharide-positive bacterial cells. The TSBG 0.2 0.006 - effect on cells of the same strain not expressing the TSB, tripticase soy broth; TSBG, TSB with 1% glucose extracellular polysaccharide was much weaker. These data suggested that a receptor-mediated endocytosis E. faecalis by 95 per cent (Fig.2). Again, TSB grown might be involved in macrophage entry of cells were affected at a lower extent (75%). polysaccharide-positive enterococcal cells leading to bacterial survival; interaction of specific macrophage Discussion receptors with polysaccharide moieties might mediate phagocytosis without inducing bactericidal functions. Enterococci are opportunistic pathogens which can cause severe infections in different body systems, Acknowledgment particularly among immunocompromised patients. Translocation through the intestinal barrier has been Authors thank Mr D. Lombardi for assistance, and the Italian indicated as one of the preferred portals of entry of National Research Council, Agenzia 2000/2001 project n. C003E9004 enterococci for spreading to distant sites; Gentry-Weeks for financial support. et al5 showed that E. faecalis was able to survive within peritoneal macrophages, which might contribute References to pathogenesis by allowing translocation of bacteria through the intestinal mucosa and hindering 1. Moellering RC. Emergence of enterococcus as significant antimicrobial therapy. We reported6 on the ability of nosocomial pathogen. Clin Infect Dis 1995; 21 : 1173-6. E. faecalis to survive within rat peritoneal macrophages 2. Huycke MM, Saham DF, Gilmore MS. Multiple-drug resistant up to 48 h, and suggested extracellular polysaccharide enterococci: the nature of the problem and an agenda for the future. Emerg Infect Dis 1998; 4 : 239-49. could be involved in extended survival. Rakita and colleagues4 showed that strains bearing AS stimulated 3. Murray BE. The life and time of the Enterococcus. Clin Microbiol their own phagocytosis by PMNs into larger Rev 1990; 3 : 46-65. phagosomes allowing survival, compared to 4. Rakita RM, Vanek NN, Jacques-Palaz K, Mee M, AS-negative cells. The present findings confirmed that Mariscalco MM, Dunny GM, et al. Enterococcus faecalis bearing aggregation substance is resistent to killing by human the survival of E. faecalis seemed to be related to the neutrophils despite phagocytosis and neutrophil activation. ability of bacteria to synthesise extracellular Infect Immun 1999; 67 : 6067-75. BALDASSARRI et al : MACROPHAGE PHAGOCYTOSIS OF E. FAECALIS 135

5. Gentry-Weeks CR, Karkhoff-Schweizer R, Pikis A, 8. Bertuccini L, Ammendolia MG, Superti F, Baldassarri L. Estay M, Keith JM. Survival of Enterococcus faecalis Invasion of HeLa cells by Enterococcus faecalis clinical in mouse peritoneal macrophages. Infect Immun 1999; isolates. Med Microbiol Immunol (Berl) 2002; 191: 25-31. 67 : 2160-5. 9. Wileman T, Harding C, Stahl P. Receptor mediated endocytosis. 6. Baldassarri L, Cecchini R, Bertuccini L, Ammendolia MG, Biochem J 1985; 232 : 1-14. Iosi F, Arciola CR, et al. Enterococcus spp. produces slime and survives in rat peritoneal macrophages. Med Microbiol Immunol 10. Swanson JA, Watts C. Macropinocytosis. Trends Cell Biol (Berl) 2001; 190 : 113-20. 1995; 5 : 424-8. 7. Wells CL, Jechorek RP, Erlandsen SL. Evidence for the 11. Rosenshine I, Ruschkowski S, Finlay BB. Inhibitors of translocation of Enterococcus faecalis across the mouse cytoskeletal function and signal transduction to study bacterial intestinal tract. J Infect Dis 1990; 162 : 82-90. invasion. Methods Enzymol 1994; 236 : 467-76.

Reprint requests: Dr Lucilla Baldassarri, Laboratorio di Ultrastrutture, Istituto Superiore di Sanità Viale Regina Elena, 299, 00161 Rome, Italy e-mail : [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 136-140

Genetic analysis of Streptococcus uberis plasminogen activators

Philip N. Ward & James A. Leigh

Institute for Animal Health, Compton Laboratory, Compton, Berkshire, RG20 7NN, UK

Received August 7, 2003

Background & objectives: Streptococci produce a diverse range of secreted plasminogen activators capable of converting mammalian plasminogen to plasmin in a species-specific manner. In all examples to date, the host animal’s plasminogen and that of a number of additional species have been shown to interact with these molecules leading to the conclusion that the pathogenesis of streptococci is in some way dependent upon activation of host plasminogen. PauA was the first plasminogen activator described from Streptococcus uberis, a pathogen frequently isolated from cases of bovine mastitis. Recently, a second S. uberis plasminogen activator (PauB) was identified from a Danish mastitis isolate. Interestingly, the pauB open reading frame occupied the locus normally filled by pauA. In the present study a genetic screen of streptococcal and field isolates frequently associated with mastitis was undertaken to assess the distribution, chromosomal location and sequence variation of these putative virulence factors. Methods: Southern analysis of a diverse panel of streptococci and additional bacterial isolates frequently associated with bovine mastitis was performed using pauA and pauB probes. Sequence variation of PauA was assessed at the protein level following nucleotide sequence analysis of pauA alleles amplified from isolates picked from different geographical locations. Results: We observed plasminogen activators to be universally distributed amongst S. uberis. A pauA allele was identified in all but one strain of S. uberis. This strain had a pauB allele substituted for pauA at the same locus. The remarkably low level of sequence variation demonstrated by PauA was further restricted to a limited number of residues within the molecule. Interpretation & conclusion: The high prevalence of PauA alleles in field isolates of S.uberis supported the observation that plasminogen activators are likely to confer an advantage with respect to colonization and growth. The findings of the present study support the theory that PauA plays a critical role in the pathogenesis of S. uberis.

Key words Bovine - mastitis - PauA - PauB - plasminogen activator - Streptococcus uberis The role of mammalian plasminogen activators and select group of pathogens including many streptococci plasmin in fibrinolysis, tissue remodelling, cellular and Staphylococcus aureus produce a diverse range migration and metastasis has long been established. The of secreted plasminogen activators capable of converting broad spectrum serine protease plasmin is formed mammalian plasminogen to plasmin in a species-specific following cleavage of the zymogen precursor plasminogen manner. Currently, six different streptococcal molecules by host activators tissue plasminogen activator and have been described that fulfil this function: the classical urokinase. Numerous pathogenic bacteria have evolved streptokinase (SK) family common to most group A, C mechanisms to take advantage of this protease activity and G streptococci of human origin 2, further plasminogen and bind plasmin(ogen) at the cell surface 1. In addition activators (Psk and Esk) identified from porcine and to being able to bind plasmin(ogen) at the cell surface, a equine isolates of S. equisimilis respectively3, the low 136 WARD & LEIGH : PLASMINOGEN ACTIVATORS OF S. UBERIS 137 molecular weight plasminogen activator (PadA) from by the manufacturer. Southern analysis was conducted S. dysgalactiae 4 and two plasminogen activators (PauA using standard procedures14. Digoxygenin-labelled and PauB) from S. uberis 5, 6. In all examples to date, probes were prepared by amplification of pauA (forward the host animal’s plasminogen and that of a limited number primer: 5’-AATAACCGGTTATGATTCCGACTAC, of additional species have been shown to interact with reverse primer: 5’-AAAATTTACTCGAGACTTCCT these molecules leading to the conclusion that the TTAAGG) and pauB (forward primer: 5’-CAAAGTA pathogenesis of streptococci is in some way dependent GAGGCCATGGCTTCAAAAGAAG, reverse primer: upon activation of host plasminogen. With respect to the 5'-CACTTTATTTCGGATCCAATTATTGC) using a world-wide involvement of S. uberis in the pathogenesis DIG PCR labelling mix (Roche, Germany). of bovine mastitis, the role of plasminogen activators PauA and PauB has yet to be defined. Previous studies PCR amplification and sequence determination suggested a widespread but not absolute distribution of of pauA alleles: The pauA alleles were amplified PauA among field isolates of S. uberis, however where and sequenced using flanking primers located in hexA PauA was absent, an alternative plasminogen activator (5’-GAGATTCCTCTCTAGATATCA) and orf1 (5’- (PauB) was found to be present7, 8. There are confusing GGGCTGCAGATCCGTTAAAAAATGACATTAATAT) reports in the literature regarding the invasive/adherent from a panel of UK mastitis isolates selected from 8 nature of the pathogen within the mammary gland and different locations. Amplified DNA was prepared the role of plasmin(ogen) in the establishment of foci of for sequencing using DNA Purification Kit II spin infection9, 10. An alternative role for cell surface- columns (Thermo-Hybaid, UK) and sequence associated plasmin(ogen) has been proposed from a determination was performed in both directions by nutritional standpoint. S. uberis lacks demonstrable Cytomyx, UK. secreted protease activity, which renders this auxotrophic bacterium at a distinct disadvantage when grown in milk. Results Plasmin(ogen) localised at the cell surface is thought likely to facilitate the procurement of essential amino ORF assignments for the plasminogen activator locus acids from complex milk proteins and peptides11. of S. uberis 0140J: A contiguous stretch of genomic DNA containing the pauA ORF was identified within The present study was undertaken to assess the the assemblies listed from the S. uberis strain 0140J distribution, chromosomal location and sequence variation sequencing project currently being undertaken at the of plasminogen activators PauA and PauB from Wellcome Trust Sanger Institute, UK12. Open reading Streptococcus uberis. frames were identified in the sequence flanking the pauA allele and assignments made for each putative gene Material & Methods based upon tBlastn results (Fig. 1). The pauA allele was flanked by DNA mismatch repair homologues hexA and ORF assignments: Open reading frame (ORF) hexB as previously reported8. These ORFs in turn were assignments at the PauA locus of S. uberis strain 0140J flanked by putative 2 DNA helicases, a group A were made using genomic sequence data from The Wellcome Trust Sanger Institute12. The PauA locus was scanned for predicted open reading frames using Artemis version 4.0 (Genome Research Ltd, The Sanger Centre, UK). Putative assignments for each ORF were made using “tblastn” (http://www.ncbi.nlm.nih.gov/).

Southern analysis: Genomic DNA was prepared using a previously described method13 from overnight cultures grown at 37oC. Two units of HindIII restriction endonuclease (Life Technologies, USA) were used to digest 5µg of each genomic DNA preparation as directed Fig. 1. pauA locus – S. uberis 0140J. 138 INDIAN J MED RES (SUPPL) MAY 2004 streptococcal (GAS) integral membrane protein conserved size (arrows in Fig.2) in all 20 Danish homologue and a putative 3-methyl-adenine DNA S. uberis isolates tested. Conversely, the pauB allele glycosylase I homologue. Further assignments were was only identified in the previously reported Danish made for putative arginyl tRNA synthetase and repressor isolate SK880 (data not shown) which lacked the pauA genes and a competence/damage inducible protein allele. homologue from S. pneumoniae. The abundance of pauA and pauB homologues in The abundance of PauA and PauB in field isolates other streptococci and other relevant species: A wider of S. uberis: A screen of Danish streptococcal field panel of streptococci and other species commonly isolates of bovine origin from 20 different farms was associated with bovine mastitis was also screened by undertaken by Southern analysis using pauA and pauB Southern analysis to determine the presence or otherwise coding sequence probes (Fig.2) and PCR (data not of homologous ORFs. The pauA allele was found to be shown) to assess the distribution and chromosomal restricted to, but universally present in S. uberis except location of these putative virulence factors. The pauA in one strain (SK880) where pauB was present (Results allele was found on genomic DNA fragments of not shown).

Sequence conservation of PauA and its geographic distribution: PauA sequence data were derived from a panel of S. uberis isolated from 8 different UK farms. The translation product of each allele sequenced was determined and compared with those already lodged on the GenBank nucleotide database. Multiple sequence alignment identified the positions within the coding sequence where heterogeneity was evident (Table). Variation that affected coding sequence was restricted to only 7 amino acids out of a possible 261. A novel combination of these 7 residues was identified (allele Fig. 2. Southern analysis of Danish S. uberis (numbered 1-20) isolated type 6) which was found in two discrete UK isolates. from 20 different locations. Panels were probed with pauA pauB probes respectively. Arrows indicate PauA specific bands. Tracks No evidence of geographical partitioning of strains was A and B represent the respective control strains S. uberis 0140J and found between isolates from Europe and USA. SK880.

Table. Sequence variation of PauA

Strain Accession Origin Allele Occurrence Residue (of mature sequence) 48 54 90 99 186 215 217 SK882 AJ131604 Dk 1 2 D VQQL Q H c216 AJ012549 UK 5 NCTC3858 AJ006413 2 1 D A R R R P D 95-140 AJ012548 USA 3 1 N V R R R P D ST10 UK 3 SK889 AJ131631 USA 4 1 D V R R S P D EF20 UK 2 SUPA70.2 AF283574 USA 5 2 D V R R R P D 0144 UK 4 C197 UK 6 2 N V R R R Q H WARD & LEIGH : PLASMINOGEN ACTIVATORS OF S. UBERIS 139 Discussion are finely tailored to suit the specific requirements for the range of hosts and disease scenarios encountered PauA was the first plasminogen activator described across this diverse family. from S. uberis, a pathogen frequently isolated from cases of bovine mastitis. Recently a second S. uberis Analysis of strains gathered from a variety of locations plasminogen activator (PauB) was identified from a within the UK enabled comparisons to be drawn with Danish mastitis isolate. Interestingly, the PauB open pauA alleles derived from other European or North reading frame occupied the locus normally filled by American sources. A very high level of amino acid PauA, between putative DNA mismatch repair genes sequence conservation was apparent, with the hexA and hexB/orf18. The chromosomal location of the identification of only one further coding variant (type 6) S. uberis plasminogen activators differs significantly which represented a different combination of amino acids from the situation observed previously for the classical at the 7 positions previously reported to vary within the streptokinase locus 15. The locus encoding streptokinase 261 residues of the mature polypeptide8. These in most group A, C and G streptococci was reported to observations further support the theory that pauA fulfils be flanked by lrp, encoding a leucine rich protein, abc, a critical role in the pathogenesis of S. uberis. an ATP binding cassette protein and rel, encoding a homologue of E. coli relA and spoT genes which Acknowledgment moderate levels of guanosine 5’,3’ polyphosphate during nutritional stress15. In contrast, the same locus in S. The authors acknowledge the generous gifts of field isolates from uberis was devoid of any plasminogen activator allele 8. Laust Johnsen (Department of Molecular and Structural Biology, University of Aarhus, Denmark); Lars Holst Pedersen (The Royal Veterinary and Agricultural University, Institute for Veterinary However, PauA is located elsewhere in the S. uberis Microbiology, Stigbøjlen, Denmark) and Elizabeth Berry (Institute chromosome, between hexA and hexB genes which for Animal Health, Compton, UK). This work was supported by display homology to the highly prevalent mutS and mutL funding from DEFRA, UK. DNA mismatch repair genes8. Further analysis of the S. uberis plasminogen activator locus in this study References identified further additional genes which indicated that this locus is primarily dedicated toward DNA replication 1. Lottenberg R, Minning-Wenz D, Boyle MD. Capturing host and repair functions. The absence of any bacteriophage- plasmin(ogen): a common mechanism for invasive pathogens? like sequences or insertion elements in these loci raise Trends Microbiol 1994; 2 : 20-4. interesting questions as to the origin and possible 2. Huang TT, Malke H, Ferretti JJ. Heterogeneity of the mechanisms of transfer of streptococcal plasminogen streptokinase gene in group A streptococci. Infect Immun 1989; activators between species that are not naturally 57 : 502-6. competent. 3. Caballero AR, Lottenberg R, Johnston KH. Cloning, expression, sequence analysis, and characterization of streptokinases secreted by porcine and equine isolates of Streptococcus equisimilis. The very high prevalence of pauA alleles in field Infect Immun 1999; 67 : 6478-86. isolates collected from various European locations supported the observation that plasminogen activators 4. Leigh JA, Hodgkinson SM, Lincoln RA. The interaction of Streptococcus dysgalactiae with plasmin and plasminogen. are likely to confer an advantage with respect to Vet Microbiol 1998; 61 : 121-35. colonization and growth. There remains only one 5. Leigh JA. Purification of a plasminogen activator from S. uberis isolate which lacks the pauA allele, and this Streptococcus uberis. FEMS Microbiol Lett 1994; 118 : 153- strain (SK880) is the only isolate identified with the pauB 8. allele. Intriguingly, the pauA allele appeared to be 6. Ward PN, Leigh JA. Characterization of PauB, a novel broad- restricted solely to S. uberis and was not detected in spectrum plasminogen activator from Streptococcus uberis. other streptococci or other bacteria commonly associated J Bacteriol 2002; 184 : 119-25. with bovine mastitis despite their occupation of the same 7. Johnsen LB, Poulsen K, Kilian M, Petersen TE. Purification environmental niche. In this respect, it is tempting to and cloning of a streptokinase from Streptococcus uberis. suggest that the streptococcal plasminogen activators Infect Immun 1999; 67 : 1072-8. 140 INDIAN J MED RES (SUPPL) MAY 2004

8. Rosey EL, Lincoln RA, Ward PN, Yancey RJ Jr, Leigh JA. 12. Genome sequence of Streptococcus uberis strain 0140J. http:// PauA: a novel plasminogen activator from Streptococcus uberis. www.sanger.ac.uk/Projects/S_uberis. FEMS Microbiol Lett 1999; 178 : 27-33. 13. Ward PN, Field TR, Ditcham WG, Maguin E, Leigh JA. 9. Thomas LH, Leigh JA, Bland AP, Cook RS. Adherence and Identification and disruption of two discrete loci encoding colonization by bacterial pathogens in explant cultures of bovine hyaluronic acid capsule biosynthesis genes hasA, hasB, and mammary tissue. Vet Res Commun 1992; 16 : 87-96. hasC in Streptococcus uberis. Infect Immun 2001; 69 : 392-9. 10. Matthews KR, Almeida RA, Oliver SP. Bovine mammary 14. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a epithelial cell invasion by Streptococcus uberis. Infect Immun laboratory manual. IInd ed. New York: Cold Spring Harbor 1994; 62 : 5641-6. Laboratory Press; 1989. 11. Kitt AJ, Leigh JA. The auxotrophic nature of Streptococcus 15. Mechold U, Steiner K, Vettermann S, Malke H. Genetic uberis. The acquisition of essential acids from plasmin organization of the streptokinase region of the Streptococcus derived casein peptides. Adv Exp Med Biol 1997; 418 : equisimilis H46A chromosome. Mol Gen Genet 1993; 241 : 647-50. 129-40.

Reprint requests: Dr Philip N. Ward, Institute for Animal Health, Compton Laboratory, Compton, Newbury, Berkshire RG 20 7HN, UK e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 141-143

Role of host genetic factors in susceptibility to group A streptococcal infections

Oliver Goldmann, Gursharan S. Chhatwal & Eva Medina

Division of Microbiology, Department of Microbial Pathogenesis & Vaccine Research, GBF-National Research Centre for Biotechnology, Braunschweig, Germany

Received August 7, 2003

Background & objectives: Epidemiological evidences indicate that host genetic factors might be critical in determining susceptibility to infection with group A streptococci (GAS). The objective of the present study was to determine the extent to which the genetic background of the mouse strain affected induction and resolution of GAS infection. Methods: Several inbred mouse strains were intravenously infected with Streptococcus pyogenes strain A20 and the mean survival times of mice was recorded overtime. Bacterial loads were determined in liver and spleen of infected animals at 48 h postinoculation. Results: Different strains of mice exhibited differential susceptibility to GAS infection. After systemic infection with S. pyogenes, inbred mice showed substantial differences in mortality and bacterial loads. Interpretation & conclusion: This study provides further evidences that a genetic component is associated with host susceptibility or resistance to GAS infection.

Key words GAS - host genetic factors - infection - susceptibility

Group A streptococci (GAS) infections have a wide study, the mouse model of S. pyogenes infection was spectrum of clinical presentations ranging from mild, self- used as an experimental tool to identify specific host limiting infections to very severe, life-threatening features that affected resistance/susceptibility to this disease1. That the same GAS strain can cause infection pathogen. of various severity in different hosts indicates that host genetic components seem to play an important role in Material & Methods determining susceptibility to GAS infections2-4. The investigation on the influence of host genetic factors on Mice: Inbred, pathogen-free, 7-wk old A/J, BALB/c, susceptibility to Streptococcus pyogenes and the DBA/2, C3H/HeN, and CBA/J mice were purchased identification of the corresponding genes will help to from Harlan-Winkelmann. clarify the molecular mechanisms involved in the pathogenesis of this infection. However, due to the Bacteria: S. pyogenes strain A20 (M type 23) was genetic heterogeneity of human populations and the fact obtained from the German Culture Collection that susceptibility to infection is usually a complex (DSM2071). Stock cultures were maintained at 70ºC multifactorial trait, animal models are needed to and were cultured at 37ºC in Todd-Hewitt broth understand the role of host genetic factors. In the present supplemented with 1 per cent yeast extract for about 141 142 INDIAN J MED RES (SUPPL) MAY 2004 6 h. Bacteria were collected in mid-log phase, washed different strains were infected intravenously with twice with sterile phosphate buffered saline (PBS), 105 cfu of S. pyogenes. After 48 h, mice were scarified

diluted to the required inoculum, and the number of viable by CO2 asphyxiation and the bacterial burden was bacteria was determined by counting colony-forming units quantified in the livers and spleens. A significantly (cfu) after diluting and plating in blood agar plates higher number of bacteria was found in the organs of (GIBCO, USA) containing 5 per cent horse blood. CBA, C3H/HeN and A/J than in the organs of DBA/2 and BALB/c mice. Infection protocol and organ preparation: Mice were inoculated with S. pyogenes in 0.2 ml of PBS via a lateral Discussion tail vein. Infected mice were euthanized by CO2 asphyxiation, and bacteria were enumerated after 48 h Epidemiological evidences indicated that host genetic in the liver and spleen by preparing homogenates of these factors might be critical on determining the outcome of organs in PBS and by plating 10-fold serial dilutions on GAS infection4. This strongly suggests that the presence blood agar. Colonies were counted after 24 h of or absence of specific host genes will determine how incubation at 37ºC. effectively GAS infection will be resolved. However, the molecular mechanisms and gene(s) involved in resistance/ Results susceptibility to GAS remain to be elucidated. The mouse model provides an optimal experimental environment in Different inbred mouse strains differed markedly in their survival time after infection with S. pyogenes: which the mechanisms responsible for susceptibility to After systemic infection with 105 cfu of S. pyogenes, GAS infection can be characterized by simple comparison inbred mice showed substantial differences in mortality. of the course of infection and immunological parameters Results showed that while some mouse strains were very in resistant and susceptible mice. resistant to GAS and survived infection (DBA/2 and BALB/c), others (CBA, C3H/HeN and A/J) were highly It is to be expected that the analysis of gene susceptible and their survival times after bacterial interactions in the control of pathogenesis of inoculation did not exceed 4 days (Table). S. pyogenes infection can be powerful means for understanding the interplay between host resistance and Bacterial burden in spleen and liver of mice after S. pyogenes virulent factors. A clearer delineation of 48 h of infection with S. pyogenes: Led to organ these events in the mouse will increase our colonization (Fig.) and tissue damage. Mice from understanding of streptococcal pathogenesis in humans Table. Mean survival times of different mouse strains after and help to the design of more effective therapeutical intravenous infection with S. pyogenes approaches. Mouse strain MST (days)

CBA 2.3±0.4 C3H/HeN 3.3±0.3 A/J 2.2±0.4 BALB/c * DBA/2 **

Female mice were used in the experiment. MST, mean survival time in days. *100 per cent survival. **90 per cent survival. Fig. Bacterial loads in the systemic organs of genetically different inbred strains of mice 48 h after infection with 105 cfu of S. pyogenes. GOLDMANN et al: HOST GENETIC FACTORS IN GAS INFECTION 143

References 3. Muotiala A, Seppäälä H, Huovinen P, Vuopio-Varkila J. Molecular comparison of group A streptococci of T1M1 1. Efstratiou A. Group A streptococci in the 1990s. J Antimicrob serotype from invasive and non-invasive infections in Finland. Chemother 2000; 45 (Suppl) : 3-12. J Infect Dis 1997; 175 : 392-9. 2. Cockerill FR 3rd, MacDonald KL, Thompson RL, Roberson F, 4. Medina E, Goldmann O, Rohde M, Lengeling A, Chhatwal Kohner PC, Besser Wiek J, et al. An outbreak of invasive group A streptococcal disease associated with high carriage rates of GS. Genetic control of susceptibility to group A the invasive clone among school-aged children. JAMA 1997; 277 streptococcal infection in mice. J Infect Dis 2001; 184 : : 38-43. 846-52.

Reprint requests: Dr Eva Medina, Division of Microbiology, Department of Microbial Pathogenesis & Vaccine Research, GBF-National Research Centre for Biotechnology, Moscheroder Weg 1, 38124, Braunschweig, Germany e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 144-147

The burden of group A streptococcal pharyngitis in Melbourne families

Margaret H Danchin*,**, Susan Rogers**, Gowri Selvaraj**, Loraine Kelpie*, Peter Rankin‡, Robert Vorich‡ Michael Howson‡, John B. Carlin*,**†‡‡, Nigel Curtis*,**†, Terrence M. Nolan‡‡ & Jonathan R. Carapetis*,**†

*University of Melbourne, Department of Paediatrics, **Murdoch Children’s Research Institute †Royal Children’s Hospital, ‡General Practitioner, ‡‡University of Melbourne School of Population Health, Melboune, Australia

Received August 6, 2003

Background & objectives: There are no recent data from industrialised countries documenting the incidence and costs of group A streptococcal (GAS) pharyngitis. Such data are important in developing policy regarding management (e.g., whether or not to use antibiotics to treat sore throat) and in planning preventive strategies, including preparing for the arrival of GAS vaccines. The present study was undertaken to estimate the incidence and costs of GAS pharyngitis in school aged children in Melbourne, Australia. We report here the results after initial 11 months of surveillance. Methods: A total of 202 families (852 individuals) with at least one child aged 3 to 12 yr were enrolled across Melbourne in a family-based cohort study, and are being followed prospectively for 24 months. Surveillance data for acute GAS pharyngitis (including serology), throat carriage, and costs of the disease were collected. Additional cases of GAS pharyngitis have been ascertained to improve the precision of costing estimates. Results: Cohort retention was 97 per cent. The spring, summer and winter carriage rates for children were 13.0, 8.0 and 16.0 per cent respectively. The incidence of GAS pharyngitis was 14 per 100 person- years for children. For every primary case there were 0.7 secondary cases and 24 per cent of families experienced at least one episode of GAS pharyngitis per year. Preliminary costing data suggest that 46 per cent of cases lead to school absenteeism and a high rate of antibiotic use. Interpretation & conclusion: The present data suggest that GAS pharyngitis remains very common in childhood, and that it has further implications in terms of secondary cases and costs. Key words Group A streptococcal - Melbourne - pharyngitis

In Australia, sore throat is the second most common epidemiology and costs of GAS pharyngitis. There have reason for which patients see their primary care doctor, been no prospective studies on the incidence or costs of and 89 per cent of general practitioners report that they GAS pharyngitis in populations at low risk of acute routinely prescribe antibiotics for sore throat1,2. The rheumatic fever in recent years. We undertook a study current debate over the merits of antibiotic treatment of aimed to estimate the incidence and costs of GAS sore throat or group A streptococcal (GAS) pharyngitis pharyngitis in school-aged children in Melbourne, in populations where acute rheumatic fever is now rare3, Australia. We used a family cohort design in order to and progress towards development of GAS vaccines4, also estimate the incidence of secondary cases, and to highlight the need for accurate data documenting the measure the burden of this disease on families, rather 144 DANCHIN et al: GROUP A STREPTOCOCCAL PHARYNGITIS IN MELBOURNE FAMILIES 145 than just on individual child. A two-year study is in anti-DNase B titre was > 300, or if there was a > two- progress; we report here the results after the initial fold rise in titre between acute and convalescent samples. 11 months of surveillance. This is part of a Individuals < 18 yr of age were defined as children and comprehensive study estimating the burden of all GAS those with ³ 18 yr of age as adults. diseases in Victoria. Enhanced surveillance: The same primary care Material & Methods practiced ascertain additional cases of sore throat, and perform throat swabs. When GAS was isolated, the study Two prospective surveillance projects were team visited the patient, and collected blood for serology, conducted a family-based cohort study (to determine the swabs from other household members, and diary data, incidence and costs), and enhanced surveillance of GAS as for the family cohort study. pharyngitis (to provide additional secondary transmission and costing data). For both studies, participants were Sample size determination: Data from the USA during recruited from 3 primary care practices in geographically the 1950s and 1960s suggested that 11-20 per cent of and socio-economically distinct regions of suburban school-age children developed symptomatic GAS Melbourne. This study was approved by the Ethics in pharyngitis per year5-7. We conservatively assumed a Human Research Committee of the Royal Children’s 10 per cent incidence per year, that families have an Hospital. average of 1.5 children in the 3-12 yr age group, and allowed for an intracluster correlation of 0.2. A sample Family cohort: Over a 4 month period starting from size of 160 families would lead to inclusion of August 2001, families with at least one child aged 3 to 240 children, or an effective sample size of 218 after 12 yr (the highest risk age group for GAS pharyngitis) applying the intracluster correlation. Using this sample were enrolled. All families meeting this criterion were size, the precision of estimation of a proportion of 10 per identified in each practice, and families were randomly cent would be ± 4 per cent. To accommodate up to a selected for recruitment. They are being followed for 20 per cent drop-out rate during the year, we planned to 24 months, during which time all episodes of sore throat enrol 200 families, divided evenly among the 3 General are being ascertained. When a case of sore throat Practices. occurred in any family member, families filled out a diary (for two weeks) including clinical and costing data, and Results presented to the general practitioner (GP) for a throat swab. If GAS was isolated, the study team visited the A total of 202 families were enrolled in the cohort family, swabed all household residents, and collected blood study, consisting of 852 individuals. The median duration for baseline serology from the index case. The study of follow up for this study was 11 months, during which team visited approximately 2 wk later to collect blood cohort retention was 97 per cent. Sixty three families for convalescent serology from the index patient, and (242 individuals), 66 families (296 individuals) and also from any other family members with a GAS-positive 73 families (314 individuals) were recruited from the three throat swab at the initial visit. In addition, all members regions, respectively. Throat carriage rates of GAS in of the cohort underwent throat swabbing on three spring, summer and winter were 13.0, 8.0 and 16.0 per occasions during the first year, to provide seasonal cent, respectively for children, and 2.0 per cent in each carriage rates of GAS. season for adults.

All swabs were placed in transport medium specify There were 242 presentations of index (primary) and cultured within 24 h. Standard culture and typing cases of sore throat (190 in children) and 243 secondary techniques5 were used. Anti-streptolysin O (ASO) and cases of sore throat in family members (119 children). anti-deoxyribonuclease B (anti-DNase B) titres were Approximately 20 per cent of primary cases were also performed using standard methodologies6,7. GAS GAS culture positive in children and adults, similar in infection was considered to be serologically confirmed secondary cases in children, but only 2 per cent in adult if either the first or second ASO titre was > 240 IU or secondary cases (Table I). Approximately 90 per cent 146 INDIAN J MED RES (SUPPL) MAY 2004 of all cases with sore throat and a positive GAS swab for a mean of 12 h, in order to care for their child. also had serological confirmation of GAS infection Adults with sore throat missed an average of 14 h of (Table I). work or their usual activity.

The incidence of all cases of GAS pharyngitis Discussion (primary and secondary) was 14 per 100 person-years for children and 4 per 100 person-years for adults The most comprehensive published study of sore (Table II). For every primary case there were 0.7 throat followed a group of Cleveland families from secondary cases. There were 44 index cases in 180 1948-52, and documented an incidence of symptomatic family-years at risk (24 cases per 100 family-years). GAS pharyngitis ranging from 1.1 per 1,000 person-days in 5-7 yr olds to 0.17 in adults, with an overall incidence Preliminary costing data were available only for cases of 0.20 per person-year in children8. Two other studies of sore throat, and were unable to be analysed separately from the USA9,10 during the 1960s reported incidences for proven GAS pharyngitis. Data were available for 99 of 0.15 and 0.22 per person-year in children. A recent cases in children and 24 cases in adults. Overall, 86 per prospective study came from a region of Auckland, cent of cases attended their general practitioner, and New Zealand11, with a large Maori and Pacific Islander 71 per cent of these visits led to the prescription of a population. In this population, where poverty and medication. Forty-six per cent of children with sore throat household crowding were greater than in suburban missed school (an average of 14 h of school time missed per child). Importantly, the adult carers of children with Melbourne, there were an average of 7 sore throats in sore throat were absent from work or their usual activities school children each year, and 50 per cent of children had a GAS culture-positive sore throat each year11. Another prospective study from India12 found an Table I. Number of positive throat swabs and proportion with positive anti-streptococcal serology, during 11 months of follow up incidence of GAS culture-positive sore throat in children of a cohort of 852 individuals of 0.95 per child-year. Neither of the latter two studies11,12 used serology to confirm culture-positive cases. Case group Swabs taken GAS culture GAS culture Therefore, there are no recent data from industrialised during sore positive and serology countries that allow an estimate of the incidence of throat (%)* positive (%)* proven GAS pharyngitis in school-age children, and no Index cases: data at all on the costs or burden of this disease on families Total 242 50 (21) 44 (88) and the wider community. Adult 52 12 (23) 11 (92) Children 190 38 (20) 33 (87) The present study demonstrated that 14 per cent of Secondary cases: children in urban Melbourne developed GAS pharyngitis Total 243 32 (13) 29 (91) each year, and that 24 per cent of families experienced Adult 124 3 (2) 3 (100) at least one episode of GAS pharyngitis each year. Children 119 29 (24) 26 (90) A high rate of secondary transmission was also found. Preliminary costing data suggested high rates of school *percentage of those already culture-positive. Serology positive – and work absenteeism, and that most cases of sore throat Either first or second anti-streptolysin O titre ³ 240 IU or anti- in this cohort received prescribed medications (in most DNase B titre ³ 300 IU, or ³ two-fold rise in titre between acute and cases, this was likely to be an antibiotic). Clearly, these convalescent titres data confirm that GAS pharyngitis remains an important Table II. Incidence of culture and serologically-proven group A problem for children and families in affluent communities, streptococcal pharyngitis in an 11 month cohort study and that the overall direct and indirect costs are likely to Case group Cases Person-years Incidence per 100 be substantial. These data will be important in at risk person-years determining the most appropriate recommendations Total 73 759 10 regarding antibiotic treatment of sore throat, and in Adult 14 324 4 developing cost-effective control strategies. These will Children 59 435 14 also be critical in developing models of strategies for the DANCHIN et al: GROUP A STREPTOCOCCAL PHARYNGITIS IN MELBOURNE FAMILIES 147 use of GAS vaccines, which are already in phase two 5. Shulman S, Tanz RR, Gerber MA. Streptococcal pharyngitis. clinical trials. In: Stevens D, Kaplan EL, editors. Streptococcal infections : clinical aspects, microbiology, and molecular pathogenesis. New York: Oxford University Press; 2000 p. 76-101. Acknowledgment 6. Rantz LA. Randall E. A modification of the technique for determination of the anti-streptolysin o titer. Proc Soc Exp Biol One of us (JRC) is supported by the National Health and Medical Med 1945; 59 : 22-5. Research Council (NHMRC) Career Development Award and MHD is supported by an NHMRC scholarship. The authors acknowledge 7. Johnson DR, Kaplan EL. Determination of anti-streptolysin O. the financial support by the NHMRC, Australia, and Dr Claire In: Johson DR, Kaplan EL, editors. Laboratory diagnosis of Harris for her assistance in establishing the network of general group A streptococcal unfections. Geneva: World Health practitioners and advice regarding study design. Organization, 1996. 8. Dingle JF, Badger GF, Jordan WS. Streptococcal infections. Illness References in the home: a study of 25,000 illnesses in a group of Cleveland families. Cleveland: Press of Western Reserve University; 1964 p. 97-117. 1. National Prescribing Service. Antibiotics in primary care. Prescribing practice review (PPR 18). Surey Hills, 9. Breese BB, Dicney FA, Talpey W. The nature of a small pediatric Australia: National Prescribing Service Limited; 2002; 18 : group practice 1. Pediatrics 1966; 38 : 264-77. p 2-3. 10. Cornfeld D, Hubbard JP, Harris TN, Weaver R. Epidemiologic 2. Britt H, Miller GC, Knox S, Charles J, Valenti L, Henderson J, studies of streptococcal infection in school children. Am J Public et al. General practice activity in Australia 2000-2001. Canberra: Health 1961; 51 : 242-9. Australian Institute of Health and Welfare (AIHW Catalogue 11. Lennon D. Rheumatic fever: a preventable disease? The New No.GEP 8), 2001. General Practice Series no.8. Zealand experience. Streptococci and streptococcal diseases: 3. Del Mar CB, Glasziou PP, Spinks AB. Antibiotics for sore entering the new millenium. Porirua : ESR; 2000 p. 503-12. throat. Cochrane Database Systematic Reviews 2001 : 2. 12. Nandi S, Kumar R, Ray P, Vohra H, Ganguly NK. Group A 4. Hu MC, Walls MA, Stroop SD, Reddish MA, Beall B, Dale JB. streptococcal sore throat in a periurban population of northern Immunogenicity of a 26-valent group A streptococcal vaccine. India: a one-year prospective study. Bull World Health Organ Infect Immun 2002; 70 : 2171-7. 2001; 79 : 528-33.

Reprint requests : Dr M.H. Danchin, Department of Paediatrics, Royal Children's Hospital, Flemington Road, Parkville Victoria, 3052, Australia e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 148-151

Invasive group A streptococcal disease in North Queensland (1996 - 2001)

R. Norton, H.V Smith*, N. Wood, E. Siegbrecht**, A.Ross & N. Ketheesan†

Microbiology, Queensland Health Pathology Services, Townsville Hospital, Townsville, Australia *Public Health Microbiology, QHSS Coopers Plains, Brisbane, Australia, **Institut fur Biotechnologie, Technische Universität Berlin, Berlin, Germany & †School of Biomedical Sciences & School of Medicine, James Cook University, Townsville, Australia

Received August 6, 2003

Background & objectives: The incidence of group A streptococcal (GAS) invasive infections have been increasing worldwide. The aim of this study was to characterize clinical and microbiological features of isolates obtained from invasive GAS infections in North Queensland, Australia between 1996 and 2001. Methods: Clinical and demographic data were collected prospectively. Isolates were biotyped, emm sequenced, M typed and tested for antibiotic sensitivity using E-test. Detection of the presence of the streptococcal pyrogenic exotoxin (spe) and fibronectin binding protein (prtF1) genes was also carried out. Results: There were 109 isolates from blood and sterile sites. All isolates were sensitive to penicillin. Tetracycline and erythromycin resistance was seen in 11 and 2.7 per cent of isolates respectively. The isolates were evenly distributed by age and sex. The overall mortality was 7 per cent and there were 18 cases of streptococcal toxic shock syndrome (STSS) in which the mortality was 22 per cent. Indigenous patients had a crude incidence rate of 82.5 per 100,000 per year compared with 10.3 per 100,000 per year in the non-indigenous patients. There was no predominance of emm / M type or association of spe type with STSS. There was also no relationship between the presence of the prtF1 gene and invasive disease. Interpretation & conclusion: Invasive group A streptococci from North Queensland are similar to those from the Northern Territory of Australia in that no single strain is predominant. The indigenous population is overrepresented. Invasiveness and the development of streptococcal toxic shock is not related to the presence of the prtF1 gene or spe a or c.

Key words Invasive group A streptococci - M type - streptococcal pyrogenic exotoxins - streptococcal toxic shock syndrome Invasive group A streptococcal (GAS) disease is surveillance scheme between January 1992 to May 1994 characterised by a variety of clinical presentations which however suggest that the proportion of reports of invasive include bacteraemia, necrotising fasciitis, myositis, scarlet GAS disease has remained stable since the beginning of fever and streptococcal toxic shock syndrome1. The surveillance in 19924. There are few published reports significant morbidity and mortality associated with this of invasive GAS disease in Australia5-7. More recent disease has led to both an increased awareness of reports are a local clustering of invasive GAS disease in possible pathogenetic mechanisms and incidence of this South Australia6 and a six year retrospective review from infection2,3. Australian figures taken from the LabDOSS the Northern Territory8. The former study used traditional 148 NORTON et al : GROUP A STREPTOCOCCAL DISEASE IN NORTH QUEENSLAND 149 M and T serotyping to determine relatedness while the by Stanley et al 10, excepting the minor modification of latter used Vir typing. A retrospective review of GAS using AmpliTaq Gold (Applied Biosystems by Roche, bacteraemia in Melbourne, Australia, showed that as in NJ, USA). this localised outbreak, an M1 clone was the most frequent isolate5. The relative predominance of the M1 The forward and reverse primers were spe A 5’- C T T A A and M3 serotypes in invasive disease has been noted G A A C C A A G A G A T G G C-3’ and 5’ A T A G G C worldwide8. This however was not demonstrated in the T T T G G A T A C C A T C G-3’, speB 5’T T C T A G G A T A C T C T A C C A G C-3’and 5’A T T T G A G C A review from the Northern Territory of Australia where G T T G C A G T A G C-3’ speC 5’-C A T C T A T G G no M type predominated7. A G G A A T T A C G C-‘3. and 5’ T G T G C C A A T T T C G A T T C T G C-3’. The present study was carried out due to the discrepancy between GAS isolates reported from the Antimicrobial sensitivity testing: Sensitivity testing was Northern Territory, which were not M type related and performed using Mueller-Hinton agar +5 per cent blood those from Adelaide and Melbourne which were M type using E-Test strips (AB Biodisc, Sweden) for penicillin, related. North Queensland has a similar population mix, chloramphenicol, erythromycin, tetracycline and vancomycin incubated for 24 h in a 5 per cent CO climate and environment to the Northern Territory. This 2 environment. The results were interpreted according to study is a prospective review of invasive GAS disease NCCLS Guidelines 200011. in North Queensland between 1996 and 2001. In addition to clinical and demographic data, the following Biotyping: Biotyping was performed according to the parameters for each isolate were determined: M typing, method of Bouvet et al12. emm sequencing, spe gene determination (a,b and c), Opacity factor (OF), antibiotic susceptibility, biotyping PrtF1 Detection: The presence of the putative and fibronectin binding protein gene (prtF1) detection. internalisation gene prtF1, was detected using the method described by Neeman et al13. Material & Methods Results Clinical and demographic data: All clinical and demographic data were prospectively collected. Cases There were a total of 109 GAS isolates from blood 9 were identified as STSS based upon set criteria . (94%) or sterile sites (6%). These represented 109 patients of which 59 were female and 50 male. The median age Isolation and identification of Streptococcus pyogenes: was 39 yr (Range: 1 month - 88 yr). Indigenous patients There were 109 isolates from blood or other sterile sites (defined as those identified as being of Aboriginal or Torres obtained from 109 patients. Identification was based on Strait Islander origin) comprised 28 per cent of the group characteristic haemolysis on Columbia horse blood agar studied while representing only 8 per cent of the population (Oxoid), sensitivity to bacitracin and detection of the in the region. The crude incidence rate was 82.5 per Lancefield GAS antigen (Wellcogen, Welcome, UK). 100,000 per year for indigenous and 10.13 per 100,000 per year for the non-indigenous patients. The overall M typing and opacity factor (OF) detection: M typing mortality was 7 per cent (Table I). and OF detection, was carried out by Dr Diana Martin of the Institute of Environmental Science and Research, The results of M/emm typing, OF, biotyping, spe a,b Porirua, New Zealand. c and prtF1 detection, are shown in Table II. All isolates that were M non-typable were emm sequenced. Of the emm sequencing: The emm sequencing was performed 18 isolates from cases of STSS, 9 were prtF1 positive. according to the protocols at http://www.cdc.gov/ncidod/ biotech/strep/protocols.htm All isolates tested were sensitive to penicillin. Erythromycin and tetracycline resistance among the Spe detection: For spe a, b and c toxin gene detection, isolates tested was uncommon at 2.7 and 11 per cent multiplex PCR was performed as previously described respectively. 150 INDIAN J MED RES (SUPPL) MAY 2004 Discussion A dominant M/emm type or clone could not be demonstrated in the present study which is similar to the The crude incidence rate in the North Queensland findings from the Northern Territory7. However, it is in population studied was found to be higher than previously contrast to two other Australian reports5,6 which showed reported from the Northern Territory, at 82.5 per 100,000 a clear predominance of M1 strains in cases of invasive per year7. The crude incidence rate in non-indigenous GAS disease. The population groups in these two 5,6 patients in both these regions, was between four to five studies were quite different to those of the Northern Territory study and the present study. The populations in times lower. Deaths occurred in 13 per cent of cases the earlier studies5,6 were surburban and non-indigenous from the Northern Territory compared with 7 per cent compared with the relatively high indigenous presence in the present study. in the Northern Territory7 and North Queensland populations (present study). The presence of risk factors was highlighted as a predisposing factor in the Northern Territory report, Table II. M/emm typing, opacity factor, biotyping and determination occurring in 82 per cent of patients studied7. We did not of spe and PrtF1 genes in isolates (n=109) look at specific risk factors apart from ethnicity which Number we feel is highly significant and encompasses a variety M/emm type of factors. M/emm 18 7 M/emm 80 7 In the present study, cellulitis and other soft tissue M/emm 1 6 infections were predictably the commonest source of M/emm 89 5 infection in 44 per cent of patients. No identifiable source emm 114 5 was found in 14 per cent of cases and highlights the M/emm NT/? New 13 presence of occult GAS disease. This has been noted in a previous study5 where identifiable focus could not be Others 66 found in 30-50 per cent of patients. Respiratory disease Opacity factor (OF) as a source of GAS disease, was observed in 9 per cent Negative 50 (46%) of patients with a median age of 1 yr in the present study. Positive 59 (54%) All of these were non-indigenous patients, emphasising Spe detection that this is essentially a disease of infancy. spe A 7 (6.4%) spe B 109 (100%) Table I. Clinical and demographic data of the patients (n=109) spe C 22 (20%) spe A & C 1 (0.9%) Total (%) Indigenous Non- n=33 (30%) indigenous spe and STSS 18 n=76 (70%) spe A 1(5.5%) Mortality 8 (7%) 2 (6%) 6 (8%) spe C 6 (33%) Biotype 3 61(56%) Cellulitis / soft tissue 48 (44%) 15 (45.5%) 33 (43.5%) Biotype 1 22 (20%) Pneumonia 10 (9%) 0 (0%) 10 (13%) Biotype 5 14 (13%) Postpartum sepsis 9 (8.5%) 1 (3.5%) 8 (10.5%) Biotype 2 8 (7%) Bone and joint 9 (8.5%) 3 (9%) 6 (8%) PrtF1 Positive 61 (56%) Surgical wound infection 7 (6%) 5 (15%) 2 (3%) PrtF1 fragment size 210 base pairs 11 Pharyngitis 5 (4.5%) 0 (0%) 5 (6.5%) 320 bp 14 No source evident 14 (13.5%) 6 (18%) 8 (10.5%) 430 bp 35 Other 7 (6%) 3 (9%) 4 (5%) 540 bp 1 STSS 18 (16.5%) 3 (9%) 15 (20%) Spe, streptococcal pyrogenic exotoxin gene STSS, streptococcal toxic shock syndrome PrtF1, fibronectin binding protein gene NORTON et al : GROUP A STREPTOCOCCAL DISEASE IN NORTH QUEENSLAND 151 The role of streptococcal pyrogenic exotoxins (spe) 3. Wessels MR. GAS disease in the 1990s : New clues to in the pathogenesis of STSS remains unclear. Of the pathogenesis. Int J Infect Dis 1996 ; 1 : 107-14. 18 patients of STSS, spe a was detected in only 1 isolate 4. Review of Group A streptococcus. Commun Dis Intell 1994; and spe c in 6 isolates. There was no difference in the 18 : 291-3. detection of spe a and c in isolates from STSS and 5. Carapetis J, Robins-Browne R, Martin D, Shelby-James T, Hogg G. Increasing severity of invasive group A streptococcal invasive GAS disease without STSS. disease in Australia: Clinical and molecular epidemiological features and identification of a new virulent M-non-typeable Internalisation of GAS into epithelial cells, has been clone. Clin Infect Dis 1995; 21 : 1220-7. shown to be associated with increased invasiveness14. 6. Norton R, Invasive GAS disease - A review. Commun Dis Intell The presence of adherence factors such as fibronectin 1995; 19 : 292-3. binding protein 1 (Fbp1) has been associated with 7. Carapetis JR, Walker AM, Hibble M, Sriprakash KS, increased invasiveness of Hep2 cells14-16. This protein Currie BJ. Clinical and epidemiological features of group A streptococcal bacteraemia in a region with hyperendemic superficial is encoded by the gene prtF1. In the present study, streptococcal infection. Epidemiol Infect 1999; 122 : 59-65. 56 per cent of isolates were prtF1 positive with between 8. Musser JM, Kapur V, Szeto J, Pan X, Swanson DS, 2 to 5 repeat regions. There was no relationship between Martin DR. Genetic diversity and relationships among isolates that were positive for prtF1, STSS and outcome. Streptococcus pyogenes strains expressing serotype M1 protein: recent intercontinental spread of a subclone causing episodes of invasive disease. Infect Immun 1995; 63 : 994-1003. This study confirms previous findings that isolates from Northern Australia do not form part of a common 9. Defining the group A streptococcal toxic shock syndrome. Retionale and consensus definitions. The Working Group on invasive strain but show considerable variability. The gene Severe Streptococcal Infections. JAMA 1993; 269 : 390-1. coding for fibronectin binding protein 1 (prtF1), is not 10. Stanley J, Desai M, Xerry J, Tanna A, Efstrations A, George R. essential for invasiveness or toxic shock syndrome. The High-resolution genotyping elucidates the epidemiology of group presence of spe a or c is not linked to STSS. Invasive A streptococcus outbreaks. J Infect Dis 1996; 174 : 500-6. GAS disease is an important cause of morbidity and 11. Performance Standards for antimicrobial disk susceptibility tests. mortality in Northern Australia. It has a particularly high National Committee for Clinical Laboratory Standards. In: Wayns PA. NCCLS. Approved Standards M2-A7 7th ed. 20 incidence in indigenous communities. Understanding both (1); 2000. host and bacterial factors involved in the pathogenesis 12. Bouvet A, Geslin P, Kriz-Kuzemenska P, Blanc V, Devine C, of this condition, will help towards the development of Grimont F. Restricted association between biotypes and strategies to prevent it. serotypes within group A streptococci. J Clin Microbiol 1994; 32:1312-7. Acknowledgment 13. Neeman R, Keller N, Barzilai A, Korenman Z, Sela S. Prevalence of internalisation-associated gene, prtF1, among persisting group- A streptococcus strains isolated from asymtomatic carriers. Authors thank Dr Diana Martin of the Institute of Environmental Lancet 1998; 352 : 74-7. Science and Research, Porirua, New Zealand, for typing the isolates and Ramaciotti foundation for financial support. 14. Jadoun J, Burstein E, Hanski E, Sela S. Proteins M6 and F1 are required for efficient invasion of group A streptococci into cultured epithelial cells. Adv Exp Med Biol 1997; 418 : 511-5. References 15. Jadoun J, Ozeri V, Burstein E, Skutelsky E, Hanski E, Sela S. Protein F1 is required for efficient entry of Streptococcus 1. Stevens DL. Invasive group A streptococcus infections. Clin pyogenes into epithelial cells. J Infect Dis 1998; 178 : 147-58. Infect Dis 1992; 14: 2-11. 16. Natanson S, Sela S, Moses AE, Musser JM, Caparon MG, 2. Stevens DL. Streptococcal toxic-shock syndrome: Spectrum of Hanski E. Distribution of fibronectin-binding proteins among disease, pathogenesis, and new concepts in treatment. Emerg group A streptococci of different M types. J Infect Dis 1995 ; Infect Dis 1995 ; 1: 69-78. 171 : 871-8.

Reprint requests: Dr Robert Norton, Director of Microbiology, The Townsville Hospital, Townsville, Queensland 4814, Australia e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 152-154

Epidemiological markers of group A streptococcal infections in France

Julien Loubinoux, Martine Florent, Bensalem Merad, Gislene Collobert & Anne Bouvet

Centre National de Référence des Streptocoques - Service de Microbiologie, Hôtel Dieu Assistance Publique - Hôpitaux de Paris, Université Paris VI, 1 place du Parvis Notre-Dame, F-75181 Paris Cedex 04, France

Received August 6, 2003

Background & objectives: A limited number of biotypes, T-types, and emm-types have been found to be associated with invasive isolates of group A streptococci, confirming the involvement of the M protein in virulence and its importance as an epidemiological marker. In this study, the epidemiological markers in the clinical isolates of group A streptococci were compared in invasive and non invasive isolates. Methods: From 1998 to 2001, 141 invasive and 353 non invasive isolates in France were studied and their biotype, T-type, and emm-type were determined. Results: The invasive isolates were mostly obtained from blood whereas the non invasive isolates were isolated from throat. Most of the isolates were of biotype 1. The invasive isolates were mostly of the T-type 1 associated with emm-type 1. The T-type 4 associated with emm-type 4 and the T-type 28 associated with emm-type 28 were also frequent. Invasive isolates responsible for puerperal sepsis and non invasive isolates were mostly of the T-type 28 associated with emm-type 28. Interpretation & conclusion: This study confirms the high prevalence of isolates of biotype 1, T-type 1, and emm-type 1 among invasive isolates of group A streptococci. The emm-type 28 associated with T-type 28 was frequently observed in non-invasive isolates. A prospective study is being conducted to update the prevalence of the different emm-types in France, which will be of importance for the development of future vaccines.

Key words Bacteraemia - emm-typing - pharyngitis - puerperal sepsis - Streptococcus pyogenes

Streptococcus pyogenes is responsible for against the prevalent strains. The present study was suppurative infections either local, such as pharyngitis undertaken to compare these epidemiological markers and pyoderma, or invasive (fasciitis, bacteraemia, and in invasive and non invasive clinical isolates in France. puerperal sepsis), associated or not with a streptococcal toxic shock syndrome1. A limited number of biotypes, Material & Methods T-types, and emm-types have been found to be associated with invasive isolates of S. pyogenes2, which confirms During 1998-2001, a total of 494 (141 invasive and the involvement of the M protein in virulence and its 353 non invasive) strains of S. pyogenes isolated in importance as an epidemiological marker. Information France were studied. The invasive strains were regarding the geographic distribution of emm-types would responsible for severe infections such as streptococcal be helpful in the development of vaccines directed toxic shock syndrome, necrotizing fasciitis, postpartum 152 LOUBINOUX et al: EPIDEMIOLOGICAL MARKERS OF GAS IN FRANCE 153

endometritis, and bacteraemia. These were referred Table. Differential enzymatic tests between major biotypes of to the French Reference Center for Streptococci. The S. pyogenes

non invasive strains were isolated from throat (n=247) Tests Biotypes during a prospective survey of pharyngitis in 1234 56 Bourgogne, a large region of France, or from various clinical specimens (n=106) collected from all over the Production of b-glucuronidase - + - - + - country. The biotypes were determined using rapid Acidification of ID 32 STREP strips (bioMérieux, Marcy l’Etoile, Mannitol - - - - - + b France) according to the possible presence of - Cyclodextrin + + - - - - glucuronidase and to the ability to hydrolyze four carbohydrates (Table)3. The serotypes T were Glycogen + + - - - - determined by means of slide agglutination of trypsin- Methyl-b-D-glucopyranoside + + + - + + digested suspensions of washed bacterial cells in the presence of type-specific antisera (Institute of Sera (7%), and T8/T25/Imp19 (7%). Common T-patterns of and Vaccines, Prague, Czech Republic). The emm- non invasive isolates were T28 (27%), T11/T12 (21%), types were determined by sequencing the variable 5’ T1 (11%), T3/T13/B3264 (11%), T8/T25/Imp19 (9%), end of the emm gene after amplification by PCR with and T4 (7%). Seven per cent of isolates were non T- the MF (5’-ATA AGG AGC ATA AAA ATG GCT- typeable. Most of invasive strains were of the emm-type 3’) and MR (5’-AGC TTA GTT TTC TTC TTT GCG- 1 (Fig.), associated with biotype 1 and T-type 1. The 3’) primers4. emm-type 28, associated with T-type 28, was frequently observed among the non invasive isolates. This Results & Discussion association was mostly found in non invasive isolates from throat during the prospective survey of pharyngitis, The 141 invasive isolates were obtained from blood but also in seven invasive isolates responsible for (n=103), skin (n=13), lung (n=12), synovial fluid (n=7), puerperal sepsis. cerebrospinal fluid (n=3), and peritoneal fluid (n=3). The 353 non invasive isolates were from throat (n=281), skin Consistent with the other reports from developed (n=61), and other sources (n=11). The most common countries2, 5, 6, this study confirms the high prevalence biotypes were biotype 1 (61% of invasive isolates, 43% of biotype 1, T-type 1, and emm-type 1 among invasive of non invasive isolates), biotype 3 (16% of invasive isolates of group A streptococci. The T-type 4 isolates, 27% of non invasive isolates), and biotype 2 associated with emm-type 4 and the T-type 28 (8% of invasive isolates, 17% of non invasive isolates). associated with emm-type 28 were also frequent. The Common T-patterns of invasive isolates were T1 (36%), type 28, which is the third most common emm-type T3/T13/B3264 (15%), T11/T12 (11%), T28 (10%), T4 among invasive group A streptococci infections, was

Invasive strains (n=62) Non invasive strains (n=151) also the major type in the isolates from postpartum endometritis. This predominance in puerperal sepsis is in agreement with previous studies7. A prospective study is being conducted to update the prevalence of the different emm-types in France, which will be of importance for the development of future vaccines. Percentage

References

1 2 3 4 6 11 12 22 26 58 68 75 77 78 1. Cunningham MW. Pathogenesis of group A streptococcal Others emm-types infections. Clin Microbiol Rev 2000; 13 : 470-511. Fig. Association of emm-types with invasive and non invasive isolates 2. O’Brien KL, Beall B, Barrett NL, Cieslak PR, Reingold A, Farley of S. pyogenes. MM, et al. Epidemiology of invasive group A streptococcus 154 INDIAN J MED RES (SUPPL) MAY 2004

disease in the United States, 1995-1999. Clin Infect Dis 2002; 5. Efstratiou A. Group A streptococci in the 1990s. J Antimicrob 35 : 268-76. Chemother 2000; 45 (Suppl) : 3-12. 3. Bouvet A, Geslin P, Kriz-Kuzemenska P, Blanc V, Devine C, 6. Varon E, Havlickova H, Gires A, Sarr A, Pitman C, Patey O, Grimont F. Restricted association between biotypes and et al. Group A streptococcal infections in France. Clinical serotypes within group A streptococci. J Clin Microbiol 1994; features and epidemiological markers. Adv Exp Med Biol 1997; 32 : 1312-7. 418 : 229-31. 4. Podbielski A, Melzer B, Lütticken R. Application of the 7. Chuang I, van Beneden C, Beall B, Schuchat A. Population- polymerase chain reaction to study the M protein(-like) gene based survseillance for postpartum invasive group A family in beta-hemolytic streptococci. Med Microbiol Immunol streptococcus infections, 1995-2000. Clin Infect Dis 2002; 35 : (Berl) 1991; 180 : 213-27. 665-70.

Reprint requests : Dr Julien Loubinoux, Service de Microbiologie, Hôtel Dieu, 1 place du Parvis Notre-Dame F-75181 Paris Cedex 04, France e-mail: [email protected]

c Indian J Med Res 119 (Suppl) May 2004, pp 155-159

Epidemiology of diseases caused by Streptococcus pyogenes in

Serbia during a nine-year period (1991-1999) ˆ Lazar Ranin, Nataša Opavski, Slobodanka Djukic´ & Vera Mijac

Institute of Microbiology & Immunology, School of Medicine, University of Belgrade, Dr Subotiæa 1, 11000 Belgrade, Serbia

Received August 6, 2003

Background & objectives: Streptococcus pyogenes (group A Streptococcus - GAS) is an important human pathogen which causes a variety of diseases, including tonsillopharyngitis, scarlet fever and rheumatic fever. It is important to understand the changes in epidemiology of the diseases caused by the pathogen for improved control of such infections. Hence, the aim of the present study was to carry out an epidemiological analysis of GAS infections in Serbia in a 9-yr period (1991-1999) and evaluation of susceptibility of GAS isolates obtained during the same period to penicillin and erythromycin. Methods: Occurrence of tonsillopharyngitis, scarlatina and rheumatic fever was analyzed and GAS carrier status in healthy children was examined over a 9-yr period from 1991 to 1999. Susceptibility to penicillin and erythromycin was determined for 1657 GAS isolates obtained from patients diagnosed with pharyngitis or scarlet fever and 512 isolates from healthy carriers. M-type antigen was also determined in these isolates. Results: The average incidences of tonsillopharyngitis and scarlet fever were 76.2 and 30.8 per cent respectively. A total of 166 cases of rheumatic fever were registered. Per cent of carriers varied from 5.5 to 11.4 per cent over the study period. Predominating M serotypes among GAS isolates tested were M1, M3, M4, M6, M11, M12 and M18, depending on the source of clinical material and period of isolation. Antimicrobial susceptibility testing showed susceptibility to penicillin in all isolates tested and resistance to erythromycin in 2.41 per cent of the isolates. Interpretation & conclusion: Although the fluctuations in incidence were noted during the nine-year period, the incidence of streptococcal tonsillopharyngitis is low but with a steady raise in Serbia. No significant changes in the incidence of scarlet fever and rheumatic fever were noted. Susceptibility to penicillin remained unchanged, but the number or erythromycin resistant strains have increased.

Key words Epidemiology - erythromycin - penicillin - resistance - Streptococcus pyogenes

Streptococcus pyogenes (group A Streptococcus - rheumatic fever, which usually follows throat infection, GAS) is one of the most common and ubiquitous human and glomerulonephritis. pathogens. It causes a wide variety of human diseases ranging from noninvasive diseases (tonsillopharyngitis, The epidemiology of diseases caused by GAS has cellulitis, impetigo) to severe invasive diseases changed, particularly in Western Europe and Northern (necrotizing fasciitis, streptococcal toxic shock syndrome, America1. In USA approximately 8,800 cases of invasive bacteraemia). The GAS infections can be further diseases (3.2/100,000 population) and over 10 million of complicated by nonsuppurative sequelae, such as noninvasive GAS infections occurred in 20002. In England

155 156 INDIAN J MED RES (SUPPL) MAY 2004 and Wales scarlet fever declined from 5,217 notified microbiological confirmation. Tonsillopharyngeal swabs cases in 1991 to 1,756 cases in 20013. Nonsuppurative were taken from all patients and inoculated onto complications of GAS infections, such as rheumatic 5 per cent sheep blood agar plates. After incubation of 24 fever, occur at extremely low rate in highly industrialized h at 37°C, preliminary identification of isolates as GAS countries4 which is primarily due to improved living was based on the Gram stain, morphological and cultural conditions and appropriate antibiotic therapy5,6. characteristics, negative catalase reaction, and Understanding the changes in epidemiology is important susceptibility to bacitracin. Definitive identification to for the recognition and improved control of streptococcal species level was done by the latex agglutination infections1. (Bio Mericux, France). Antistreptolyzin test (Torlak, Belgrade) was used to confirm the diagnosis of rheumatic Penicillin is the drug of choice for the treatment of fever. The carrier state was evaluated in a randomly chosen GAS infections. The resistance to penicillin has not been paediatric population including 8617 healthy children, aged described in GAS, but failure of penicillin treatment as 3-15 yr by examining the throat swabs. well as appearance of tolerance to penicillin has been reported7,8. Erythromycin is recommended as an Susceptibility of 1657 randomly chosen GAS isolates alternative antibiotic for the treatment of GAS infections obtained from patients with clinical signs of pharyngitis or in patients allergic to beta-lactam antibiotics. However, scarlet fever and 512 isolates from healthy carriers to the resistance to erythromycin and related antibiotics has penicillin (Galenika, Serbia) and erythromycin (Pierrel, been widely reported9-13. Italy), was performed by broth dilution and agar dilution methods, according to the NCCLS14 recommendations. The aim of the present study was to carry out an The isolates was defined as penicillin tolerant if minimal epidemiological analysis of registered GAS infections in bactericidal concentration/minimal inhibitory concentration Serbia in a 9-yr period and evaluation of susceptibility to (MBC/MIC) was ³16. Typing of M proteins for the same penicillin and erythromycin of GAS isolates obtained isolates was performed at the International Reference during the same period. Center in Prague, Chezk Republic.

Material & Methods The difference in the incidence rates of tonsillopharyngitis and scarlet fever was examined by During the 9-yr period, from 1991 to 1999, all new Student's t test, P<0.05 was considered significant. cases of GAS tonsillopharyngitis, scarlet fever and rheumatic fever (4-18 yr age) were registered in the three Results regional centres representative for all parts of Serbia (Central Serbia, Vojvodina and Kosovo & Metohija). The incidence rate (per 100,000 population) of scarlet Diagnoses of pharyngitis and scarlet fever were made fever was significantly (P<0.05) lower in comparison based on the clinical signs, laboratory evaluation and with the incidence rate of tonsillopharyngitis, 30.8 per

Table I. Number of patients with S. pyogenes tonsillopharyngitis, scarlet fever, rheumatic fever and incidence rates (cases per 100,000 population) during the period from 1991-1999 in Serbia

Disease Patients/ Years Incidence 1991 1992 1993 1994 1995 1996 1997 1998 1999

Pharyngitis Patients 6195 7260 7160 5997 5998 8826 9374 10597 5893 Pharyngitis Incidence 63.7 74.0 72.3 60.9 61.3 90.5 95.8 108.4 60.3 Scarlet fever Patients 3342 4029 3876 2649 1945 2168 3228 3745 2116 Scarlet fever Incidence 34.4 41.1 39.4 26.9 19.9 22.2 33.0 38.3 21.6 Rheumatic fever* Patients 22 33 30 17 17 4 18 15 10

*The data on incidence rates were not available RANIN et al: EPIDEMIOLOGY OF GAS IN SERBIA 157 cent vs. 76.2 per cent, respectively. During the study susceptible to penicillin, having MIC value £ 0.03 µg/ml period, 21 epidemics of scarlet fever comprising (Table III). However, a slight increase in penicillin MBC 285 patients were registered, predominantly in day care values (up to 0.125 µg/ml) was noted. An increase in centres and schools. Rheumatic fever (first attack) was number of isolates displaying tolerance to penicillin was registered only sporadically, with a total of 166 cases also observed over the years (Table III). In contrast to during the study period (Table I). The percentage of the results obtained with penicillin, MIC values for carriers recorded in paediatric population during the erythromycin increased 4 times (from 0.06 to 0.25 µl/ml) period of the study were: 8.5 per cent in 1991, 9 in 1992, over the study period (Table III). Among the isolates 8.7 in 1993, 7.6 in 1994, 5.5 in 1995, 5.9 in 1996, 8.9 in tested (predominantly serotypes M1, M3 and M6), 1997, 10.5 in 1998, and 11.4 per cent in 1999. 2.41 per cent expressed resistance to erythromycin, having MIC values higher than 1 µg/ml. According to the results of M-typing, the most prevalent GAS serotypes in Serbia were M1, M3, M4, Discussion M6, M11, M12 and M18 (Table II). The prevalence of the particular M types varied with respect to the year Analysis of the epidemiological characteristics of of isolation and source of the isolates. While M4, M11, streptococcal diseases in Serbia revealed the fluctuations and M12 serotypes were most prevalent among the in incidence rates. It is likely that this is the consequence isolates obtained from patients with pharyngitis and of pathomorphosis of streptococcal syndrome. Though healthy carriers during 1991-1993, M1 and M6 there were fluctuations during the study period, the overall serotypes became predominant during 1997-1999. The incidence of streptococcal tonsillopharyngitis is in low M type distribution was somewhat different for GAS but steady raise in Serbia. It should be noted that the isolates originating from patients with scarlet fever. M6 data shown for 1999 probably do not reflect the true and M12 types were predominant among isolates from incidence of streptococcal diseases, since the notification these patients during the study period. However, an of patients in that year was not complete. The increase increase in the number of M1 and M3 serotypes was in incidence of tonsillopharyngitis, may probably be noted during 1997-1999. attributed to the appearance and predomination of highly virulent serotypes (M1, M3) at the end of 1990s and, During the study period susceptibility of GAS isolates possibly, decreased acquired immunity in younger to penicillin was not changed. All isolates were population. No significant changes in the incidence of

Table II. M types of S. pyogenes isolates in Serbia during 1991-1999

Type Years 1991-1993 1994-1996 1997-1999 Pharyngitis Scarlet Carriers Pharyngitis Scarlet Carriers Pharyngitis Scarlet Carriers fever fever fever

M1 0 6.8 0 2.7 10.8 0 11.9 25.6 7.8 M3 0 4.6 0 2.1 7.5 0 6.8 16.3 1.5 M4 19.5 0 24.2 8.0 0 20.6 8.2 0 10.7 M6 10.7 13.6 8.3 18.5 20.4 13.3 20.3 22.8 17.7 M11 26.5 0 27.8 15.5 0 23.9 8 0 21.2 M12 23.2 26.3 15.8 25.7 25.2 17 26.6 21 22.3 M18 7.6 36.7 10.8 13.7 27.2 12.6 6.2 5 7.6 other types 7.2 5.8 6.9 7.3 3.3 6.5 5.5 3.6 5.2 non-typable 5.3 6.2 6.2 6.5 5.6 6.1 6.5 5.7 6.0

Values are given in percentages 158 INDIAN J MED RES (SUPPL.) 2004

17 Table III. Susceptibility of S. pyogenes isolates to penicillin and tolerance to penicillin , the present result is of erythromycin considerable interest. One of the most frequently used alternative drugs in curing streptococcal infections is Antibiotic MIC/MBC Patient Years erythromycin. Though the present analysis revealed a (µg/ml) status 1991- 1994- 1997- slight decrease in susceptibility of GAS isolates to 93 96 99 erythromycin, the prevalence of erythromycin resistant Penicillin MIC50 pharyngitis 0.015 0.015 0.015 GAS strains in Serbia remained at low level. It has been MIC90 pharyngitis 0.015 0.03 0.03 shown that the resistance of GAS strains to erythromycin directly correlates with macrolide consumption13. MIC50 scarlet fever 0.015 0.015 0.015 Erythromycin is not extensively used in Serbia, which MIC90 scarlet fever 0.015 0.03 0.03 probably explains the low level of resistance to this MIC50 carrier 0.015 0.03 0.015 antibiotic established among the GAS strains tested in MIC90 carrier 0.03 0.03 0.03 our study. Though the observed changes in susceptibility of GAS isolates to penicillin and erythromycin were not Penicillin MBC50 pharyngitis 0.03 0.06 0.125 highly significant, the results we obtained suggest the MBC90 pharyngitis 0.06 0.125 0.125 need for careful monitoring in the future. PTS (%) pharyngitis 3.2 4.8 7.2 MBC50 scarlet fever 0.03 0.03 0.03 References MBC90 scarlet fever 0.06 0.06 0.06 1. Efstratiou A. Group A streptococci in 1990s. J Antimicrob PTS (%) scarlet fever 1.8 2.1 2.1 Chemother 2000; 45 Topic T1:3-12. MBC50 carrier 0.03 0.06 0.06 2. www.cdc.gov, accessed in November 2002. MBC90 carrier 0.06 0.125 0.125 3. www.phls.co.uk, accessed in November 2002. PTS (%) carrier 2.5 2.4 3.7 4. Olivier C. Rheumatic fever - is it still a problem? J Antimicrob Erythromycin MIC50 pharyngitis 0.03 0.03 0.06 Chemother 2000; 45 Topic T1:13-21. MIC90 pharyngitis 0.06 0.06 0.125 5. Gordis L. The virtual disappearance of rheumatic fever in the United States: lessons in the rise and fall of disease. Circulation MIC50 scarlet fever 0.03 0.06 0.125 1985; 72 : 1155-62. MIC90 scarlet fever 0.06 0.125 0.25 6. Massell BF, Chute CG, Walker AM, Kurland GS. Penicillin MIC50 carrier 0.03 0.03 0.06 and the marked decrease in morbidity and mortality from rheumatic fever in the United States. N Engl J Med 1988; MIC90 carrier 0.06 0.06 0.125 318 : 280-6. PTS, per cent of tolerant strains; MIC, minimum inhibitory 7. Kaplan EL, Johnson DB. Unexplained reduced microbiological concentration; MBC, minimum bacterial concentration efficacy of the intramuscular benzathine penicillin G and of oral penicillin V in eradication of group A streptococci from children scarlet fever and rheumatic fever were noted. This is in with acute pharyngitis. Pediatrics 2001; 108 : 1180-8. contrast with results reported from highly industrialized 8. Michel MF, van Leenwen WB. Degree and stability of tolerance 4 countries . to penicillin in Streptococcus pyogenes. Eur J Clin Microbiol Infect Dis 1989; 8 : 225-32. No clinical isolates of GAS resistant to penicillin were 9. Arvand M, Hoeck M, Hahn H, Wagner J. Antimicrobial detected, which was an expected finding. These results resistance in Streptococcus pyogenes in Berlin. J Antimicrob are in accordance with the fact that despite the long Chemother 2000; 46 : 621-3. term application of penicillin in clinical practice, no GAS 10. Hsueh P-R, Chen H-M, Huang A-H, Wu J-J. Decreased activity 15,16 strains resistant to this antibiotic have been identified . of erythromycin against Streptococcus pyogenes in Taiwan. MIC values though remained within the susceptible Antimicrob Agents Chemother 1995; 39 : 2239-42. range, there was a slight increase in MBC values and 11. York MK, Gibbs L, Perdreau-Remington F, Brooks GF. per cent of penicillin tolerant isolates. Since it has been Characterization of antimicrobial resistance in Streptococcus shown that failures in treatment of infections caused by pyogenes isolates from the San Francisco bay area of Northern GAS are partially due to increased MBC values and California. J Clin Microbiol 1999; 37 : 1727-31. RANIN et al: EPIDEMIOLOGY OF GAS IN SERBIA 159

12. Bingen E, Fitoussi F, Doit C, Cohen R, Tanna A, George R, et al. 15. Kaplan EL. Recent evaluation of antimicrobial resistance in Resistance to macrolides in Streptococcus pyogenes in France in b-hemolytic streptococci. Clin Infect Dis 1997; 24 (Suppl 1) : pediatric patients. Antimicrob Agents Chemother 2000; 44 : S89-S92. 1453-7. 16. Horn DL, Zabriskie JB, Austrian R, Cleary PP, Ferretti JJ, 13. Baquero F, Garcia-Rodriguez JA, de Lomas JG, Aguilar L. and Fischetti VA, et al. Why have group A streptococci remained Spanish Surveillance Group for Respiratory pathogens. susceptible to penicillin? Report on a symposium. Clin Infect Antimicrobial resistance of 914 beta-hemolytic streptococci Dis 1998; 26 : 1341-5. isolated from pharyngeal swabs in Spain: Results of a 1-year (1996-1997) multicenter surveillance study. Antimicrob Agents 17. Sela S, Barzilai A. Why do we fail with penicillin in the treatment Chemother 1999; 43 : 178-80. of group A streptococcus infections? Ann Med 1999; 31 : 303-7. 14. National Committee for Clinical Laboratory Standards. 1997. Methods for dilution antimicrobial susceptibility tests for 18. Granizo JJ, Aguilar L, Casal J, Dal-Re R, Baquero F. bacteria that grow aerobically, 4th ed. Approved standard Streptococcus pyogenes resistance to erythromycin in relation M7-A4. National Committee for Clinical Laboratory Standards, to macrolide consumption in Spain (1986-1997). J Antimicrob Wayne. Pa. Chemother 2000; 46 : 959-64. Reprint requests: Dr Lazar Ranin, Institute of Microbiology & Immunology, School of Medicine, University of Belgrade Dr Subotica´ 1, 11000 Belgrade, Serbia e-mail: [email protected] [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 160-163

Changing incidence of detected streptococcal bacteraemia in North Yorkshire, England

Michael R.D. Barnham & Nigel C. Weightman*

Departments of Microbiology, Harrogate District Hospital & *Northallerton Friarage Hospital, North Yorkshire, UK

Received August 6, 2003

Background & objectives: Streptococcal bacteraemia remains a major and challenging clinical problem throughout the world. The epidemiology of these infections appears to be changing. In the present study we analysed the data collected over a period of 20 yr (1978-1999) to throw light on this. Methods: Laboratory records of all patients yielding clinically significant, non-pneumococcal, non- enterococcal streptococcal bacteraemia were reviewed for the decades 1978-1988 and 1990-1999 in the two districts (combined population 260,000). Results: From a total of 3134 patients with detected bacteraemia 338 (10.8%) yielded streptococci. The organisms comprised: in the beta-haemolytic group (n=169), GAS (32%), GBS (34%), GCS (5%), GGS (28%), GRS (1%); in the non-haemolytic/Streptococcus milleri group (n=160): S. milleri (21%), S. bovis (10%), S. sanguis (30%), S.mitis (10%), S. oralis (8%) and other ‘viridans’ streptococci (20%); nine isolates were of anaerobic streptococci. There was a steady increase in the number of blood cultures submitted for investigation during the period and streptococcal isolations of all species rose in proportion (approximately x 2.5) in the second decade compared with the first. Complementary data on the occurrence of necrotising fasciitis and toxic shock syndrome in these districts suggested a real increase in the occurrence of serious forms of GAS infection. Interpretation & conclusion: An increase in the level of detected streptococcal bacteraemia was seen in this part of England over 20 yr, involving a wide variety of different streptococcal species. This represented both improved ascertainment of infection and an increase in the burden of disease in the population.

Key words Bacteraemia - group C streptococci - group G streptococci - incidence of disease - Streptococcus agalactiae - S. bovis - S. milleri - S. pyogenes - S. suis

Despite the usual in vitro susceptibility to a wide 1999) in the same study population, to observe trends range of antimicrobial agents, streptococci remain and changing incidence of disease during the period 1978- frequent and important in clinical practice throughout the 1999. world. In our previous study over 10 yr (1978-1988) in the Harrogate and Northallerton districts of North Material & Methods Yorkshire we found that these organisms accounted for 11 per cent of clinically significant bacteraemias1. Since Clinical and laboratory records of all patients yielding then, detected infection has increased remarkably.The clinically significant, non-pneumococcal, non- aim of the present study was to give data on incidence enterococcal streptococcal bacteraemia for the decades of streptococcal bacteraemia for a second decade (1990- 1978-1988 and 1990-1999 in the Harrogate and 160 BARNHAM & WEIGHTMAN : CHANGING INCIDENCE OF STREPTOCOCCAL BACTERAEMIA 161 Northallerton districts (combined population 260,000) bacteraemia was 6.5 episodes per 100,000 population, were compiled from notes made at the time by the authors increasing from 2.8 in the first 5 yr period of the study to and from a review of casenotes. Streptococci were 9.5 in the last (Table III). Notable features of the identified by standard methods in the Yorkshire infections (as reported previously in many instances) laboratories and selected isolates referred for typing to include the following: the Streptococcus Reference Unit, Central Public Health Laboratory, Colindale, London. Table I. Bacteraemic episodes in the study area over two decades Organism Number of detected episodes in Results 1st decade 2nd decade Total 20 yr (1978-1988) (1990-1999) Over the period 1978-1999 there was a steady increase Any (i.e., all organisms) 902 2232 3134 of 3.5 per cent per year in all types of microbiological specimens submitted to the laboratories for investigation. Streptococci 100 238 338 Of these, the proportion of blood culture specimens Pneumococci 83 269 352 increased by 0.05 per cent per year from 1.95 per cent in Enterococci 27 77 104 1978 to 3 per cent in 1999. There was therefore a pronounced increase in the number of blood cultures Table II. Bacteraemic episodes per 100 blood culture sets received submitted, from under 1500 in 1978 to over 4500 in 1999. over the two decades The number of bacteraemic episodes rose markedly during Organism Number of detected episodes the second decade (1990-1999) (Table I). per 100 sets in period 1st decade 2nd decade Compared with the first decade, in the 1990s there (1978-1988) (1990-1999) was a rise in detected bacteraemic patients, streptococcal isolates, pneumococci and enterococci by factors of 2.5, Any (i.e., all organisms) 4.42 6.02 2.4, 3.2 and 2.9 respectively. An increase was also seen Streptococci 0.49 0.64 from the first to second decade in the bacteraemia Pneumococci 0.41 0.72 detection rate per 100 blood culture sets received (Table Enterococci 0.13 0.21 II). From a total of 3134 patients with detected bacteraemia 338 (10.8%) yielded streptococci compared Table III. Incidence of detected streptococcal bacteraemia (cases per with 352 (11.2%) yielding pneumococci and 104 (3.3%) 100,000 population per year) in four 5-yr periods, 1978-1999 enterococci. Analysing the 20 yr catchment of Organism Incidence in 5-yr period Total streptococci by organism, beta-haemolytic streptococci (no. of cases) 1st 2nd 3rd 4th 20 yr accounted for 50 per cent [in which group A streptococci (1978- (1983- (1990- (1995- period (GAS) made up 16%, group B streptococci (GBS) 17%, 1983) -1988) 1994) 1999) group C streptococci (GCS) 2.5%, group G streptococci Group A streptococci (54) 0.6 0.9 1.2 1.5 1.0 (GGS) 14% and group R streptococci (GRS) 0.5%]; the remaining half comprised non-haemolytic streptococci Group B streptococci (58) 0.2 0.8 1.5 1.9 1.1 including Streptococcus milleri (10%), S. bovis (5%), S. sanguis (14%), S.mitis (5%), S.oralis (4%), Group C streptococci (8) 0.08 0.3 - 0.2 0.2 anaerobic streptococci (3%) and miscellaneous (9%). Despite rising isolation rates seen in almost all species, Group G streptococci (47) 0.4 0.3 1.5 1.5 0.9 the relative proportions of organisms seen in the first and second decades of the study remained broadly steady; Group R streptococci (2) - - - 0.2 0.04 however, there was a disproportionate increase in GBS N/H streptococci (160) 1.4 2.5 4.5 4.0 3.1 and GGS infections and no substantial increase in GCS. Amongst the non-haemolytic streptococci, S. milleri Anaer streptococci (9) 0.08 0.2 0.2 0.3 0.2 isolates increased by a factor of 2.1 and S. bovis by 7.0 All streptococci (338) 2.8 4.9 8.8 9.5 6.5 between the first and second decades of the study. The N/H, non-haemolytic including S. milleri group; Anaer, anaerobic overall annual incidence of detected streptococcal 162 INDIAN J MED RES (SUPPL) MAY 2004 GAS: The number of infections detected per year varied Non-haemolytic streptococci: Detailed analysis of from nil to six, with two thirds occurring in the winter patients in the first decade of the study1 showed a months October to March2. Most infections were frequent association of viridans and entirely non- sporadic and community-acquired but three community- haemolytic streptococci with endocarditis/endarteritis (in based clusters occurred. Ages of patients ranged from two thirds of cases), while others yielding these organisms two days to 93 yr, with 24 per cent less than 15 yr of age demonstrated abscess, septicaemia, mucositis and and 47 per cent 70 yr or older, and there was a fatality infection associated with intravenous devices. S. milleri rate of 36 per cent. More than a dozen M-typing patterns was associated with serious pyogenic infections, abscess, were encountered in these patients but M-types 1, 3, 12 septicaemia and endocarditis. Similar findings were and R28 were predominant, varying over the years encountered in the second decade. Our review of patients broadly in agreement with reported national fluctuations. with S. bovis bacteraemia8 showed an age range of 40 Substantial increases were noted in the local incidence to 86 yr, endocarditis in 47 per cent (in whom the of detected streptococcal necrotising fasciitis3 and toxic mortality rate was 22%) and septicaemia in 53 per cent shock syndrome4 during this period, of which 50 and 86 (mortality 27%); five patients had cancer including three per cent of cases respectively were bacteraemic. The with colonic lesions. clinical epidemiology of non-invasive GAS infections varied in the study area during the period with detected Anaerobic streptococci: Bacteraemia was seen in pyoderma infections most frequent in the early years association with dental, intestinal and urinary tract and genital tract infections peaking in 19895. surgery, malignancy and alcoholism in patients ranging from 22 to 78 yr of age. GBS: Twenty one infections (36% of the total GBS isolations) occurred in infants, six in the first decade of Discussion the study and 15 in the second, and six (10%) occurred peri-partum in women aged 15-34 yr. For patients aged A steady but striking increase in the detection of ³ 35 yr, the incidence by sex was equal and detected clinically significant bacteraemia over the study period cases increased from five in the first decade (38% of of 20 yr included almost all species of streptococci in GBS isolations in that part of the study) to twenty five in proportion to increasing pneumococcal and enterococcal the second (56% of GBS isolations). isolates. Similar trends in detected Gram positive bacteraemias in this period have been reported by others 9,10 11,12 13,14 GCS: All bacteraemias were caused by the species in the UK , mainland Europe and the USA ; local rates of reported nosocomial bacteraemia appear S. equisimilis and none by S. zooepidemicus, although to vary according to the nature of the population served12. a few cases of superficial and locally invasive infection The rise in isolations in our series exceeded the rise in with the latter organism were found in the district during submitted blood cultures. Various technical factors can the study6. Clinical infection with GCS resembled that affect the results of blood culture15 but, although with GGS. automated growth detection systems for blood culture were installed during the period, there were no GGS: Infection occurred predominantly in the elderly, fundamental technical changes in our laboratories to 7 with mean age 74 yr and a fatality rate of 33 per cent . account for the increased yield of positives. Detected Amongst a wide variety of clinical features, most GAS bacteraemia increased 2.4-fold during the period frequent were cellulitis (in 53% of cases), septic arthritis and was matched by similar increases in streptococcal (11%) and pneumonia (11%), while endocarditis occurred necrotising fasciitis and toxic shock syndrome in these in 4 per cent and toxic shock syndrome in 4 per cent. districts3,4. This suggested that there was a real increase in the occurrence of serious forms of GAS infection here, GRS: S. suis infection was detected in a 44 yr old pig rather than just a simple improvement in ascertainment farmer with fever, headache and influenza-like of infection. In the case of detected non-GAS symptoms, and in a 71 yr old alcoholic woman with streptococcal bacteraemias, which also increased, septicaemia of undetermined origin. analysis showed little difference in the character of these BARNHAM & WEIGHTMAN : CHANGING INCIDENCE OF STREPTOCOCCAL BACTERAEMIA 163 infections over the two decades of the study. Infections 5. Barnham M. Pyogenic streptococci from genital tract infections. in adults with GBS and S. bovis appeared to increase In: Totolian A, editor. Pathogenic streptococci: present and future. St Petersburg: Lancer Publications; 1994 p. 42-3. disproportionately during the study while, for reasons unknown, GCS infections showed no change. Two cases 6. Barnham M, Kerby J, Chandler RS, Millar MR. Group C of the rarely encountered S. suis infection were noted streptococci in human infection: a study of 308 isolates with clinical correlations. Epidemiol Infect 1989; 102 : in patients towards the end of the study period. Improved 379-90. ascertainment of infection by increasing use of blood 7. Weightman NC, Barnham M. Group G streptococcal cultures may well have occurred over the 20 yr but we bacteraemia: a 20-year case sequence in North Yorkshire, suspect that there has also been a genuine increase in England. In: Martin DR, Tagg JR, editors. Streptococci and the burden of infection in the population. Factors that streptococcal diseases - entering the new millennium. New could influence this include increased survival into old Zealand: C/-ESR; 2000 p. 95-8. age, increasing use of intensive and immune-suppressing 8. Dealler S, Anderson AW, Weightman N, Johnstone DJ, treatments, increased survival of patients with disease Barnham M. Streptococcus bovis bacteraemia in North that predisposes to infection and increasing utilisation of Yorkshire 1982-1992. In: Totolian A, editor. Pathogenic health care facilities, with heightened opportunities for streptococci: present and future. St Petersburg: Lancer Publications; 1994 p. 52-3. hospital- and nursing home-acquired disease and cross- infection. In addition, there may have been changes in 9. Public Health Laboratory Service. Bacteraemia and bacterial the prevalence of certain streptococci with enhanced meningitis in England and Wales: 1982 to 1996. Comm Dis Rep 1997; 7 : 276-8. ability to cause disease, such as the M-type 1 and 3 GAS which have been more frequent here at times during 10. Reacher MH, Shah A, Livermore DM, Wale MCJ, Graham the study and have been associated with high rates of C, Johnson AP, et al. Bacteraemia and antibiotic resistance of its pathogens reported in England and Wales between 2,16 morbidity and mortality . The increases we note in 1990 and 1998: trend analysis. Br Med J 2000; 320 : detected streptococcal infection leave no room for 213-6. complacency in the diagnosis and treatment of these 11. Bouza E, Perez-Molina J, Munoz P. Report of ESGNI-001 and important infections. ESGNI-002 studies: bloodstream infections in Europe. Clin Microbiol Infect 1999; 5 (Suppl 2): 2S1-12. References 12. Correa L, Pittet D. Problems and solutions in hospital-acquired bacteraemia. J Hosp Infect 2000; 46 : 89-95. 1. Barnham M. Invasive streptococcal infections in the era 13. Bryan CS. Clinical implications of positive blood cultures. Clin before AIDS: a 10 years’ compilation of patients with streptococcal bacteraemia in North Yorkshire. J Infect 1989; Microbiol Rev 1989; 2 : 329-53. 18 : 231-48. 14. Banerjee SN, Emori TG, Culver DH, Gaynes RP, Jarvis WR, 2. Barnham M, Weightman N. Streptococcus pyogenes bacteraemia: Horan T, et al. Secular trends in nosocomial primary bloodstream a 16-year case sequence in North Yorkshire. Clin Microbiol infections in the United States, 1980-89. National Nosocomial Infect 1997; 3 (Suppl 2): 131 (abstract). Infections Surveillance System. Am J Med 1991; 91 (Suppl 3B) : S86-9. 3. Barnham M. Necrotising fasciitis: a perspective from North Yorkshire. In: Martin DR, Tagg JR, editors. Streptococci and 15. Mylotte JM, Tayara A. Blood cultures: clinical aspects and streptococcal diseases - entering the new millennium. New controversies. Eur J Clin Microbiol Infect Dis 2000; 19 : Zealand: C/-ESR; 2000 p. 47-9. 157-63. 4. Barnham M, Weightman NC, Anderson AW, Tanna A. 16. O,Brien KL, Beall B, Barrett NL, Cieslak PR, Reingold A, Farley Streptococcal toxic shock syndrome: description of 14 cases MM, et al. Epidemiology of invasive group A streptococcus from North Yorkshire, UK. Clin Microbiol Infect 2002; 8 : 174- disease in the United States, 1995-1999. Clin Infect Dis 2002; 81. 35 : 268-76.

Reprint requests: Dr M.R.D. Barnham, Somersby, Ripley Road, Knaresborough, North Yorkshire HG5 9BY, UK e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 164-167

Streptococcus milleri group bacteraemia in North Yorkshire, England (1989-2000)

Nigel C. Weightman, Michael R.D. Barnham* & Marion Dove

Department of Microbiology, Friarage Hospital, Northallerton & *Harrogate District Hospital North Yorkshire, UK

Received August 6, 2003

Background & objectives: Though carriage and local infection with organisms of the Streptococcus milleri group (SMG) are regular in clinical practice, bacteraemia is infrequent in man. The objective of the present study was to give an account of our experience with the SMG bacteraemia over a period of 12 yr in North Yorkshire. Methods. The laboratory and clinical records of all clinically significant cases of SMG bacteraemia in our district general hospital catchment (combined population 260,000) were reviewed for the 12 yr period from 1989 to 2000. Viable isolates were recovered, species identified, and minimum inhibitory concentration (MIC) determined. Results: Twenty nine episodes of infection gave an annual incidence of 0.93 cases per 100,000 population. Infections included abscess, pneumonia, septic arthritis, genital and urinary tract infections, endocarditis and diffuse septicaemia. Patient ages ranged from 18 to 90 yr but most patients were elderly, 82 per cent had evident predisposing conditions and mortality rate was 10 per cent. Species determination of the 22 isolates showed S. anginosus (64%), S. constellatus (27%) and S. intermedius (9%). Nearly all isolates were non-haemolytic and Lancefield grouping showed carriage of the F antigen (in 41%), C (14%) and no detected group (45%). Most of the isolates were susceptible to the antimicrobials tested. Interpretation & conclusion: Bacteraemia with SMG organisms was infrequent, often opportunistic and featured a low overall mortality rate. Intra-abdominal sepsis was the local feature in 16 (55%) of the patients and 9 (31%) of the total patient group showed abscess or empyema. There was a low rate of antibiotic resistance in these organisms from bloodstream infection.

Key words Bacteraemia - Streptococcus anginosus - S. constellatus - S. intermedius - S. milleri The Streptococcus milleri group (SMG) of organisms present study was to provide an account of 29 episodes is separated into 3 species, S. intermedius, of SMG bacteraemia in 28 patients during the 12 yr period S. constellatus and S. anginosus1. The organisms are from 1989-2000 in district general hospitals’ catchment regularly encountered in clinical specimens and typically population of 260,000 in North Yorkshire. produce small colonies on culture with variable haemolytic properties and variable carriage of Lancefield Material & Methods group antigens, including F, C, G and A. Bacteraemia with organisms of this group is infrequent in man2 and The records and the case notes of all patients with there have been few substantial case series or reviews diagnosed SMG bacteraemia detected in our laboratories within the last two decades2-6. The objective of the (two of the four laboratories serving North Yorkshire) 164 WEIGHTMAN et al: S. MILLERI BACTERAEMIA 165

Table I. Clinical features of patients and species identification

Feature Species

No. of patients S. anginosus S. constellatus S. intermedius Not speciated with feature+

Liver abscess 4 2 1 1 Other abscess* 422 Empyema 1 1 Pneumonia 3 1 1 1 Cholangitis/cholecystitis 96111 Infected skin lesion 4 2 1 1 Genital tract infection 1 1 Urinary tract infection# 1 1 Septic arthritis 1 1 Diffuse septicaemia 3 2 1 Infective endocarditis 1 1

* 1 dental abscess, 1 caecal abscess, 1 post-operative wound abscess, 1 abdominal abscess + 3 patients each had 2 features; #infected hydronephrosis

Table II. Potential predisposing factors in 28 patients yielding SMG (Oxoid Ltd., Basingstoke, UK) and minimum inhibitory organisms concentration (MIC) of selected antimicrobials 7 Factor No. of patients (%) with factor* determined by E test (AB Biodisk, Solna, Sweden) on isosensitest agar with 5 per cent whole horse blood Active cancer 7 (25) (Oxoid Ltd., Basingstoke, UK) at 37°C in 6 per cent + NSAIDs 6 (21) CO2. Isolates were deemed as susceptible or resistant Previous surgery 4 (14) to antimicrobial agents using British Society for Antimicrobial Chemotherapy (BSAC) criteria8. Chronic liver disease 3 (11) Diabetes mellitus 3 (11) Results Gall bladder disease 2 (7) Diverticular disease 2 (7) A total of 29 episodes of bacteraemia with SMG Congenital heart disease 1 (4)# were detected in 28 patients (male : female 14 : 14) during the study period, giving an annual incidence of None identified 5 (18) 0.93 per 100,000 population per year. In 4 episodes * some patients had more than one factor (14%), the bacteraemia was polymicrobial, together + NSAID - non-steroidal anti-inflammatory drugs taken at or just with one each of Proteus mirabilis, Haemophilus before onset of infection influenzae, S. equisimilis and Erysipelothrix #patient developed infective endocarditis rhusiopathiae. The mean age ± SD was 67 ±16.5 yr (range 18-90 yr). Most patients were elderly with 20 during a period of 1989 to 2000 were reviewed. All of 28 (71%) over 60 yr of age. In 22 (76%) patients viable isolates of SMG during the study period were the infection was community acquired while in 7 (24%) recovered from their Protect Bacterial Preserver beads it was hospital acquired. Of the 28 patients, 3 died. stored at -70oC (Technical Service Consultants, Twenty two isolates were available for further study Heywood, UK), species identity confirmed by Rapid ID32 and of these S. anginosus was most frequent (64%), Strep test (BioMerieux, Marcy L’Etoile, France), followed by S. constellatus (27%) and S. intermedius Lancefield grouped by latex agglutination test (9%). There appeared to be an association between 166 INDIAN J MED RES (SUPPL) MAY 2004

Table III. Haematological parameters in patients with SMG bacteraemia

Parameter Mean ± SD Range No. of patients with parameter* (% of tested patients)

Haemoglobin (g/dl) 11.4 ±2.2 7.3-15.7 <12 16 (59) 12-17 11 (41) White blood cell count (x 109/l) 16.2 ± 13.2 2-66.2 <3.6 1 (4) 3.6-10 7 (26) >10 19 (70) Neutrophil count (x109/l) 14.9 ± 13.1 0.8-63.7 <2 1 (4) 2-7.5 3 (12) >7.5 21 (84)

*results not available for haemoglobin and white blood cell count in 2 episodes, and for neutrophil count in 4 episodes

Table IV. Lancefield grouping of 22 available isolates of SMG abscess formation and bacteraemia with the species S. constellatus and S. intermedius (Table I). In 3 Species Lancefield group cases (10%), the infection was fatal, 2 female (age 76 C F Non-groupable* and 86 yr) and 1 male (age 62 yr), with one case each S. anginosus 36 5 of pneumonia, liver abscess and septicaemia from an infected intravenous access site. Two of these patients S. constellatus 02 4 had underlying malignant disease (myeloma and breast S. intermedius 01 1 carcinoma). Potential predisposing factors could not *No groups A, C, F or G antigens detected be identified in 5 patients (Table II). Haematological parameters were available for most of the patients (Table III). Of the 22 isolates, lancefield group C was Table V. Minimum inhibitory concentration-90 and antimicrobial susceptibility of the isolates found in three, F in 9 and 10 were non-groupable (Table IV). Beta haemolysis was shown by 3 strains Antimicrobial agent MIC range MIC %sensitive %resistant 90 (two S. constellatus and one S. anginosus) and only (µg/ml) (µg/ml) one strain formed mucoid colonies (S. constellatus). Benzylpenicillin 0.006-0.047 0.047 100 0 Resistance to the tested antimicrobial agents was found Tetracycline 0.016-3 0.38 95 5 in only 2 strains, both S. anginosus: one demonstrated resistance to macrolides, and one to tetracycline. Most Erythromycin <0.016-2 0.032 95 5 of the isolates were sensitive to the antimicrobials tested Clarithromycin <0.016-2 0.032 95 5 (Table V). Azithromycin <0.016-24 0.25 95 5 Clindamycin <0.016-0.47 0.094 100 0 Discussion Vancomycin 0.19-1 0.75 100 0 The incidence of clinically significant detected SMG Teicoplanin <0.016-0.16 0.064 100 0 bacteraemia, at 0.93 per 100,000 population per year in Moxifloxacin 0.006-0.32 0.094 100 0 this series, is similar to that found in a previous study2. A Linezolid 0.016-0.38 0.25 100 0 separate analysis of data over two decades showed SMG to make up 10 per cent of all clinically significant MIC, minimum inhibitory concentration streptococcal bacteraemias9. WEIGHTMAN et al: S. MILLERI BACTERAEMIA 167 Most of our patients were elderly and acquired their References infection in the community. The mortality rate for patients in the present study was only 10 per cent, a 1. Whiley RA, Beighton D. Emended descriptions and recognition figure similar to earlier reports2,5 but lower than that of Streptococcus constellatus, Streptococcus intermedius and reported by Jacobs et al4. Only 5 of 28 patients (18%) Streptococcus anginosus as distinct species. Int J Syst Bacteriol 1991; 41 : 1-5. had no apparent potential predisposing conditions, a similar rate to those given in earlier series4,5, suggesting 2. Casariego E, Rodriguez A, Corredoira JC, Alonso P, Coira A, that invasive SMG infection was usually opportunistic in Bal M, et al. Prospective study of Streptococcus milleri bacteremia. Eur J Clin Microbiol Infect Dis 1996; 15 : 194-200. nature. Active cancer was the commonest underlying condition, as reported by others2,4. Of the potential 3. Libertin CR, Hermans PE, Washington JA III. Beta-haemolytic predisposing factors, administration of aspirin or other group F streptococcal bacteremia. A study and review of the literature. Rev Infect Dis 1985; 7 : 498-503. NSAIDs on admission was notable (21% of patients), and worthy of further investigation as there is evidence 4. Jacobs J, Pietersen H, Stobberingh E, Soeters P. Bacteremia involving the “Streptococcus milleri” group: analysis of 19 cases. for the potential involvement of these drugs in severe Clin Infect Dis 1994; 19 : 704-13. streptococcal disease10. 5. Salavert M, Gomez L, Rodriguez-Carballeira M, Xercavins M, Freixas N, Garau J. Seven year review of bacteremia caused by 16 patients (55%) had intra-abdominal sepsis (liver Streptococcus milleri and other viridans streptococci. abscess, caecal abscess, abdominal abscess, cholangitis/ Eur J Clin Microb Infect Dis 1996; 15 : 365-71. cholecystitis, infected hydronephrosis), of which half 6. Bert F, Bariou-Lancelin M, Lambert-Zechovsky N. Clinical featured cholangitis or cholecystitis without abscess significance of bacteremia involving the “Streptococcus milleri” formation. These latter conditions have been noted earlier group: 51 cases and review. Clin Infect Dis 1998; 27 : 385-7. as the most common source of infection in SMG 7. Renneberg J, Niemann LL, Gutschik E. Antimicrobial 5 2 bacteraemia . Unlike a previous study in which liver susceptibility of 278 streptococcal blood isolates to seven abscess was a common source of infection, only antimicrobial agents. J Antimicrob Chemother 1997; 39 : 4 patients (14%) were found to have this aetiology. Of 135-40. particular interest was the association between the 8. MacGowan AP, Wise R. Establishing MIC break points and species S. constellatus and S. intermedius and abscess the interpretation of in vitro susceptibility tests. J Antimicrob formation. In this series, 45 per cent of patients with Chemother 2001; 48 (Suppl. S1) : 17-28. abscess yielded S. constellatus, compared with only 9. Barnham M, Weightman NC. Changing incidence of detected 10 per cent of patients without associated abscess. Both streptococcal bacteraemia in North Yorkshire, England. Indian J isolates of S. intermedius were also from patients with Med Res 2004; 119 (Suppl) : 160-63. abscess. This association of species with abscess has 10. Stevens DL. Could non-steroidal anti-inflammatory drugs been previously recognised11, both in bacteraemic and (NSAIDs) enhance the progression of bacterial infections to locally infected cases. Infective endocarditis has been toxic shock syndrome? Clin Infect Dis 1995; 21 : 977-80. reported as a notable feature in a substantial proportion 11. Clarridge JE III, Attorri S, Musher DM, Hebert J, Dunbar S. of patients yielding SMG from blood cultures2 but the Streptococcus intermedius, streptococcus constellatus and only case in our series was in a patient with congenital Streptococcus anginosus (Streptococcus milleri Group) are of different clinical importance and are not equally associated with heart disease (ventricular septal defect and bicuspid abscess. Clin Infect Dis 2001; 32 : 1511-5. aortic valve). 12. Jacobs JA, Stobberingh EE. In vitro antimicrobial susceptibility of the ‘Streptococcus milleri’ group (Streptococcus anginosus, Most of the isolates were sensitive to the Streptococcus constellatus and Streptococcus intermedius). antimicrobials tested similar to an earlier study12. J Antimicrob Chemother 1996; 37 : 371-5.

Reprint requests: Dr N.C. Weightman, Department of Microbiology, Friarage Hospital, Bullamoor Road, Northallerton DL6 1JG, UK e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 168-170

Comparison of Streptococcus pneumoniae serotypes causing acute otitis media & invasive disease in young children in the Czech Republic

R. Prymula, J. Motlova* & P. Kriz*

Purkyne Military Medical Academy, Hradec Kralove & *National Institute of Public Health, Prague, Czech Republic

Received August 6, 2003

Background & objectives: The availability of a type-specific pneumococcal vaccine for children is a worldwide problem. It is necessary to study the serotypes prevalent in a country before introducing a type-specific vaccine. The objective of the present study was to analyse the prevalence of Streptococcus pneumoniae serotypes in children suffering from acute otitis media or invasive pneumococcal disease and to compare a coverage of serotypes by individual pneumococcal vaccines. Methods: Children suffering from acute otitis media and invasive pneumococcal disease were analysed in the Czech Republic from October 1999 to November 2000. Serotyping was performed by the quellung technique using antisera from Statens Serum Institute (Denmark). Results: The most frequent serotypes in patients with acute otitis media were 3, 19F, 23F, 14, 9V, 1, 6B, 11A and 28F. Vaccine coverage for the identified serotypes in acute otitis media patients was 52.1 per cent for the 7-valent vaccine, 57.8 per cent for the 9-valent vaccine and 75.7 per cent for the 11-valent form of the vaccine. In 108 patients with invasive pneumococcal disease, the most frequent serotypes were 6B, 9V, 14, 19F, 3 and 23F. Vaccine coverage for the identified serotypes in patients with invasive pneumococcal disease was 62 per cent for the 7-valent vaccine, 66.4 per cent for the 9-valent vaccine and 77.5 per cent for the 11-valent form of the vaccine. Interpretation & conclusion: Vaccine coverage for the identified serotypes for the 11-valent pneumococcal vaccine was better than the other two vaccines. Key words Acute invasive disease - acute otitis media - serotypes - Streptococcus pneumoniae - vaccine Acute otitis media (AOM) is the most common Geographical differences in the distribution of serotypes disease diagnosed during early childhood in the developed as well as a large number of S. pneumoniae type countries. Based on various reports 50-60 per cent of antigens described emphasize the need why a study on children have at least one episode of acute otitis media the distribution of serotypes should be done prior to during their childhood. Approximately 88 per cent cases/ introduction of a type-specific pneumococcal vaccine in episodes of acute otitis media are bacterial and 25-50 a country. per cent of them are caused by Streptococcus pneumoniae. Invasive pneumococcal disease (IPD) is The present study was carried out to analyse the less frequent (incidence 160 per 100,000) but is more prevalence of S. pneumoniae serotypes in children serious compared to AOM. suffering from acute otitis media or invasive pneumococcal disease in Czech Republic and to compare The availability of a type-specific pneumococcal a coverage of serotypes by individual pneumococcal vaccine for children is a worldwide problem. vaccines. 168 PRYMULA et al: COMPARISON OF S. PNEUMONIAE IN CHILDREN 169 Czech Republic (October 1999 - November . Serotypes (%) of isolates obtained from patients with acute otitis media (n=141) and invasive pneumococcal disease (n=108) in 2000) serotype 1 3 4 5AOM 6A 6B 7F 5.7 8 15 9A 1.4 9N 9V 10A 11A 2.9 12F 14 5IPD 15A 2.9 15C 0.7 17A 18C 3.7 19A 7.4 19F 3.7 20 0.9 23F 4.6 6.4 24F 14.8 3.7 28F 2.8 0.9 29 1.8 31 14.8 2.8 33F 5 35F 0.7 9.3 37 38 1.4 40 0.7 11.1 0.7 6.4 1.4AOM, acute otitis media; IPD, invasive pneumococcal disease 14.3 1.4 9.3 1.8 5 3.7 0.7 10.2 0.9 0.7 5.6 0.7 0.9 0.7 1.4 1.8 0.9 0.9 Table 170 INDIAN J MED RES (SUPPL) MAY 2004 Material & Methods per cent for the 9-valent vaccine and 77.5 per cent for the 11-valent form of the vaccine. Young children (< 2 yr of age) suffering from acute otitis media were analysed prospectively in a multicentre Our finding on serotype distribution and vaccine study in the Czech Republic from October 1999 to coverage in children younger than 2 yr of age in Czech November 2000. Children with invasive pneumococcal Republic are in good agreement with other studies1-5. disease were also studied and isolates obtained from Both the acute otitis media and invasive pneumococcal these patients were sent for typing to the National disease serotypes showed the importance of using an Reference Laboratory for Streptococci and Enterococci 11-valent vaccine formulation. In fact, the serotypes 1 in Prague. Data on coverage of vaccine (7-valent, 9- and 3 not present in the 7-valent formulation accounted valent and 11-valent vaccine) were collected during the for 25 per cent of overall cases. Moreover, there is not a same time as for the prospective study. Typing was major difference in the serotypes represented in the performed by the quellung technique using antisera from vaccines as cause of acute otitis media or invasive Statens Serum Institute (Denmark). pneumococcal disease respectively but there is, instead, a different percentage of cases being attributed to each Results & Discussion vaccine-serotype.

Of the 143 patients with acute otitis media due to S. References pneumoniae, 141 isolates were serotyped and 140 confirmed. The most frequent serotypes in these 1. Konradsen HB, Kaltoft MS. Invasive pneumococcal infections patients were 3 (15.0%), 19F (14.3%), 23F and 14 (9.3% in Denmark from 1995 to 1999: Epidemiology, serotypes, and each), 9V (6.4%), 1 (5.7%), 6B, 11A and 28F (5.0% resistance. Clin Diagn Lab Immun 2002; 9 : 358-65. each) (Table). Vaccine coverage for the identified 2. Fraser D, Givon-Lavi A, Bilenko A, Dagan R. A decade serotypes in acute otitis media patients was 52.1 per (1989-1998) of pediatric invasive pneumococcal disease cent for the 7-valent vaccine, 57.8 per cent for the 9- in 2 populations residing in 1 geographic location: valent vaccine and 75.7 per cent for the 11-valent form implications for vaccine choice. Clin Infect Dis 2001; 33 : 421-7. of the vaccine. 3. Holm AM, Berild D, Ringertz SH, Haheim LL, Hoiby EA. Occurrence and clinical presentation of systemic pneumococcal A total of 108 isolates were obtained from children infections in an unselected population in Oslo, Norway, between (< 2 yr of age) with invasive pneumococcal disease. 1993 and 1997. Eur J Clin Microbiol Infect Dis 2002; 21 : There were 40 cases of meningitis, 16 of bacteraemia, 8 465-7. of sepsis/septicaemia/shock syndrome, 30 of lower 4. Kyaw MH, Clarke S, Edwards GF, Jones IG, Campbell H. respiratory tract infections and 14 other serious Serotypes/groups distribution and antimicrobial resistance pneumococcal infections. In these patients the most of invasive pneumococcal isolates: implications for vaccine frequent serotypes were 6B and 9V (14.8% each), 14 strategies. Epidemiol Infect 2000; 125 : 561-72. (11.1%), 19F (10.2%), 3 (7.4%) and 23F (5.6%) (Table). 5. Kyaw MH, Clarke S, Jones IG, Campbell H. Non-invasive Vaccine coverage for the identified serotypes in these pneumococcal disease and antimicrobial resistance: vaccine patients was 62 per cent for the 7-valent vaccine, 66.4 implications. Epidemiol Infect 2002; 128 : 21-7.

Reprint requests: Dr P. Kriz, National Institute of Public Health, WHO Collaborating Centre for Reference and Research on Streptococci in Prague, Srobarova 48, Prague10, 10042, Czech Republic e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 171-173

Group A streptococcal infections of the pharynx in a rural population in south India

Thangam Menon, S. Shanmugasundaram*, M. Palani Kumar & C.P. Girish Kumar

Department of Microbiology, Dr ALM PGIBMS, University of Madras & *Department of Cardiology, Stanley Medical College, Chennai, India

Received August 6, 2003

Background & objectives: There has been a resurgence in the incidence of rheumatic heart disease all over the world and hence surveillance and strain characterization are important. The aim of this study was to screen children in a rural community in south India for throat carriage of group A streptococci and to clinically assess them for signs of rheumatic heart disease. Methods: Throat swabs were collected from children (5-14 yr) in the village of Orathur, Tamil Nadu and cultured on tryptose blood agar plates. Beta haemolytic streptococci were serogrouped using Streptex kit and biotyped based on their ability to ferment carbohydrates and production of b-glucuronidase enzyme. Blood samples were also collected and antibodies to streptolysin O demonstrated by latex agglutination tests. All the children were examined by a paediatrician; ECG and echocardiography were performed to assess cardiac function. Results: Eighty of the 310 children included in the study had symptoms of acute respiratory infections; 16 of them grew beta haemolytic streptococci of which 8 belonged to group A (10%). Biotype 4 was most common. Antistreptolysin O (ASO) test did not correlate with culture results. Two of 310 children had rheumatic heart disease but both were culture negative. Interpretation & conclusion: Pharyngeal carriage of group A streptococci was common in this population. The prevalence of rheumatic heart disease was 0.6 per cent. The study emphasizes the need for active surveillance and characterization of GAS isolates.

Key words Biotypes - group A streptococci - rheumatic heart disease Group A streptococci are the most common disease is still prevalent in developing countries, aetiological agents of pharyngitis. Since the 1980s there particularly among young children who live in has been resurgence in the incidence of invasive S. communities that do not have adequate treatment pyogenes infections and rheumatic heart disease all over programmes. The present study was undertaken to type the world1 and hence strain characterization has assumed pharyngeal strains of group A streptococci isolated from great importance. Typing of strains is usually based on children in a rural population in a village in Tamil Nadu, variation in the M and T protein and production of serum south India. opacity factor (SOF). Genotypic methods such as emm and sof gene sequencing and vir typing, have been used Material & Methods and found to be highly discriminatory2. An association between serotypes and biotypes has been shown and A total of 310 children aged 5-14 yr attending a rural biotyping was found to be an easy method for rapid and health camp at Orathur near Chennai, Tamil Nadu, were preliminary characterization of strains3. Rheumatic heart included in the study. 171 172 INDIAN J MED RES (SUPPL) MAY 2004

Table I. Beta haemolytic streptococci from throat cultures

No. screened BHS No. (%) GAS No. (%) GCS No. (%) GGS No. (%) Symptomatic children 80 16 (20) 8 (50) 5 (31.2) 3 (18.7) Asymptomatic children 230 45 (19.5) 18 (40) 8 (17.7) 19 (42.2)

BHS, beta haemolytic streptococci; GAS, group A streptococci; GCS, group C streptococci; GGS, group G streptococci

Table II. Biotypes of group A streptococci isolated from children Blood samples were collected, serum separated and with acute respiratory infection (clinical isolates) and asymptomatic antibodies to streptolysin O were measured by ASO test children (carrier isolates) using a commercial kit (Tulip Laboratories Pvt. Ltd., India). Biotype Clinical isolates Carrier isolates (n=8) (n=18) Results

1- 2 Eighty of 310 children had symptoms of acute 3- 3respiratory infections (ARI) whereas 230 children were 43 3asymptomatic. Of these 80 children 16 (20%) with ARI 51 2grew beta haemolytic streptococci in culture and 8 of 16 (50%) isolates were group A streptococci. Of the 6- 1 remaining 8 isolates, 5 were group C and 3 were group 91 1G streptococci. Forty five of 230 (19.5%) of 10 2 - asymptomatic children grew beta haemolytic Untypable 1 6 streptococci and 18 of 45 (40%) were group A. Of the remaining 27 isolates, 19 were group G and 8 were group All children were examined by a paediatrician. A 12 C streptococci (Table I). Two of the 8 clinical isolates of lead ECG using a multi channel ECG recorder (Hewlett GAS and 6 of the 18 carrier isolates were positive for Packard, China) was done. Echocardiography was done SOF. using a Color Doppler (Esaote Megas, Italy) and a 3.5 MHz transducer. Apart from routine cavity Biotype 4 was the most common in the clinical measurements, valve morphology and function were also isolates. There were no striking differences in the overall recorded. frequencies of the biotypes. Biotype 3, which was common in carriers, was not seen in children with ARI Throat swabs were collected from tonsillar area and (Table II). posterior pharyngeal wall using sterile cotton tipped swabs. The material from the swab was transferred to One of the 18 asymptomatic children showed a titre sterile filter papers and transported to the laboratory, of 800 IU in the ASO test; 15 children were negative where each filter paper strip was incubated on the surface and 2 gave a titre of 200 IU. Two of the 8 culture positive of a tryptose blood agar plate for 24-48 h at 37ºC under children gave a titre of 400 IU and the remaining 6 were 5-10 per cent CO . Beta haemolytic streptococci were 2 negative. identified by Gram stain, colony morphology, and haemolysis on blood agar. The isolates were subcultured Two of the 310 children were diagnosed as having on Todd Hewitt broth and grouping was done by latex agglutination tests using Streptex kit (Abbott Murex, UK). rheumatic heart disease based on auscultatory findings Tests for SOF production was done using standard and echocardiography. Neither of them had positive throat methods4. cultures.

Biotyping was done by testing for the ability of the Discussion isolates to ferment mannitol, cyclodextrin, glycogen, pullulan, methyl b-D glucopyranoside and production of Pharyngeal carriage of group A streptococci was b-glucuronidase enzyme3. found to be common, occurring in 7.8 per cent of MENON et al : THROAT CARRIAGE OF GROUP A STREPTOCOCCI 173 asymptomatic children as compared to 10 per cent in community of south India, emphasizing the need for children with ARI in the present study, and rheumatic active surveillance. heart disease was diagnosed in 2 of the 310 children screened (0.6%). Koshi and Benjamin5 reported an References incidence of rheumatic heart disease in 4.5 per cent of rural south Indian children in 1981. In affluent developed 1. Kaplan EL, Johnson DR, Cleary PP. Group A streptococci countries the rates of rheumatic heart disease are usually serotypes isolated from patients and sibling and contacts during about 0.2 to 0.5 per 100,0006. Bouvet et al3 described the resurgence of rheumatic fever in United States in the mid biotyping as a useful method for preliminary 1980s. J Infect Dis 1989; 159 :101-3. characterization of strains of group A streptococci. 2. Gardiner DL, Sriprakash KS. Molecular epidemiology of However, in this study many isolates were found having impetiginous group A streptococcal infections in aboriginal no characteristics of biotypes 1-10, and were considered communities of Northern Australia. J Clin Microbiol 1996; 34 : untypable by this method. The common biotypes seen in 1448-52. this cohort of children were biotypes 4 and 3. Biotype 4 3. Bouvet A, Geslin P, Kriz-Kuzemenska P, Blanc V, Devine C, is commonly associated with M type 78; biotype 3 which Grimont F. Restricted association between biotypes and was encountered only among carriers, is known to serotypes within group A streptococci. J Clin Microbiol 1994; correspond with more than one M type (M3, M12, M78, 32 : 1312-7. and M nontypable strains)3. Antibody response to 4. Rehder CD, Johnson DL, Kaplan EL. Comparison of methods streptolysin O is known to occur in about 80 per cent of for obtaining serum opacity factor from group A streptococci. patients with acute group A streptococcal pharyngitis7. J Clin Microbiol 1995; 33 : 2963-7. Only 25 per cent of culture positive symptomatic children 5. Koshi G, Benjamin V. Rheumatic fever and rheumatic heart in this study showed a raised ASO titre. It is possible disease in rural south Indian children. Bull World Health Organ that some of these may be cases of viral pharyngitis in 1981; 59 : 561-99. patients who carried group A streptococci coincidentally. 6. Stollerman GH. Rheumatic fever. Lancet 1997; 349 : 935-41. In conclusion, pharyngeal group A streptococci were 7. Wannamaker LW, Ayoub EM. Antibody titers in acute rheumatic found to be common in this group of children in a rural fever. Circulation 1960; 21 : 598-614.

Reprint requests: Dr Thangam Menon, Professor, Department of Microbiology, Dr ALM PGIBMS, University of Madras, Taramani Chennai 600113, India e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 179-182

Evaluation of gastrointestinal symptoms as primary sign of severe invasive group A streptococcal infections

K. Ekelund, A. Lemcke* & H.B. Konradsen

Streptococcus Unit, Department of Bacteriology, Mycology & Parasitology, Statens & Serum Institut & *Department of Epidemiology, Statens Serum Institut, Denmark

Received August 6, 2003

Background & objectives: Severe invasive infections caused by group A Streptococcus (GAS) are often associated with shock and organ failure. We describe epidemiological and disease related data from the national surveillance of invasive GAS infection in Denmark in addition to three fatal cases that occurred in Denmark in 2002 with gastrointestinal (GI) symptoms as the dominating preliminary signs. Methods: As the National Streptococcal Reference Centre The Streptococcus Unit, Statens Serum Institut (SSI) receives the vast majority of the invasive GAS isolates from patients admitted to all the hospitals in Denmark. The isolates were T-typed by slide agglutination test emm-squencing and pulsed field gel electrophoresis (PFGE) were also performed. Results: During January 2002 three patients died at home and GAS were found at autopsy. Cases 1 (12 yr) and 3 (25 yr) had been ill for less than two days with nausea, diarrhoea and vomiting. Case 2 (25 yr) had the same symptoms for two weeks. None of the three had any underlying diseases. The GAS isolates from cases 1 and 2 were T-type 3-13-B3264, emm89 and SpeA-, SpeC-. The third isolate was T- type 1, emm1 and SpeA+, SpeC-. PFGE could not discriminate between the two isolates with T-type 3-13-B3264. The PFGE patterns of the three isolates were similar to those identified from GAS isolated elsewhere in Denmark at different times and from non-fatal cases. In 1999-2002, SSI received 409 isolates from patients with invasive GAS infection, and the mortality rate was 18 per cent. In 40 patients the primary symptoms were gastrointestinal, and in 30 per cent of these the outcome was fatal. Interpretation & conclusion: The various early clinical manifestations of severe GAS infections are still a major challenge for clinicians because of the importance of a fast and appropriate diagnosis and immediate start of treatment. Key words Denmark - diarrhoea - gastrointestinal symptoms - group A streptococcal - invasive streptococcal infections - Streptococcus pyogenes - surveillance - vomiting Streptococcus pyogenes, group A Streptococcus Denmark. None of the patients were admitted to the (GAS), has been described as an emerging cause of hospital and the three patients had predominated severe infections with septic shock, organ failure and gastrointestinal (GI) symptoms. In the present study, soft tissue infections [myositis and necrotizing fasciitis epidemiological and disease related data from the national (NF)]. The patients with streptococcal toxic shock surveillance of invasive GAS infections in Denmark are syndrome (STSS) were described as young without described to evaluate gastrointestinal complaints as a predisposing factors and had flu-like complaints as the primary sign of severe invasive GAS infections. primary symptoms1-4. In January 2002, three unexpected Information on the three fatal cases with invasive GAS deaths due to invasive GAS infections occurred in infections was also described. 179 180 INDIAN J MED RES (SUPPL) MAY 2004 Material & Methods The strain was mucoid and was SpeA-, SpeB+ and SpeC. T-type 1 emm1 was identified in tissue from heart, lungs, National surveillance: The invasive GAS isolates CSF, tonsils, skin and pharynx from the third patients. included in this study were from blood and cerebrospinal The strain was SpeA+, SpeB+ and SpeC-. fluid (CSF) collected from patients admitted to various hospitals in Denmark from January 1, 1999 to June 30, Epidemiological data on invasive GAS infections in 2002. The Streptococcus Unit serves as the National Denmark: During the study, the Streptococcus Unit at Streptococcus Reference Centre and receives the vast Statens Serum Institut received 409 isolates from patients majority of the invasive streptococcal isolates as part of with invasive GAS infections admitted to various hospitals the national surveillance. The 15 local clinical in Denmark. In relation to every invasive isolate, a microbiological departments voluntarily submit invasive questionnaire was sent out to the clinician responsible for GAS isolates to the Streptococcus Unit to be serotyped. the treatment of that patient and the Streptococcus Unit received 397 questionnaires in return. Fifty four per cent Typing of the isolates: After a confirmatory identification of the patients were females; the gender was not stated of the group A antigen5, the isolates were T-typed by in 10 cases (2%). The overall mortality rate was 18 per slide agglutination test with rabbit antibodies produced by cent, while the annual mortality rates for 1999, 2000, 2001 Statens Serum Institut, Denmark. Non-typeable strains and 2002 were 15, 19, 15 and 25 per cent, respectively. were designated NT5. emm-sequensing was performed The median age of the patients who died was 78.1 yr on the isolates from the three fatal cases and identification (range 4.4-95.7 yr). The survivors had a median age of of SpeA, SpeB and SpeC were performed on the 40 56.2 yr (range 0.4-97.4 yr). Stratification for age was not isolates from patients with gastrointestinal complaints as used in the analysis. T-types 1, 3-13-B3264, 28 and NT the primary symptoms. Both molecular methods were constituted 26, 22, 11 and 19 per cent respectively of the carried out as described elsewhere6. GAS isolates from dead patients, and 19, 24, 16 and 10 per cent respectively of the GAS isolates from the Questionnaire: A questionnaire was sent to the clinical survivors. Of the fatal cases, 96 per cent of the isolates wards where the patients were admitted and filled in by the were from blood in comparison with 99 per cent of the doctors responsible for the treatment of the patients. Of non-fatal cases. The remaining isolates were from CSF. the three patients (2 males, 1 female), the first was a 12 yr old boy (died on January 6, 2002). The second case (25 yr A significantly larger number of the fatal cases had old woman) was found dead at home on January 16, 2002, only GI-complaints as the primary symptom compared while the third case (25 yr old man) died on February 1, to the non-fatal cases (P<0.05). A similar pattern could 2002. All these patients had diarrhoea and vomiting. The be noticed for bacteraemia without any focus (P<0.05). isolates obtained from the patients were tested by T-typing. Septic shock occurred in 60 per cent of the fatal cases emm-sequencing and identification of the exotoxins (SpeA, compared to 17 per cent of the non-fatal cases (P<0.001) SpeB and SpeC). Pulsed field gel electrophoresis (PFGE) while STSS or NF occurred in 21 per cent of the fatal was also performed. and in 9 per cent of the non-fatal cases (P<0.05). Treatment in the intensive care unit and use of Statistical analysis: Fischer's exact test and Mann- mechanical ventilation occurred significantly more often Whitney U tests were used for statistical analysis. in the cases with fatal outcome (P <0.05 and P <0.01, P < 0.05 was considered significant. The SAS System, respectively) (Table). Release 8.02 (SAS Institute Inc. Cary, NC, USA) was used for all analysis. GI-complaints as the primary symptom: The primary symptoms were gastrointestinal in 40 patients, and in 12 Results (30%) of these the outcome was fatal. The outcome and primary symptoms were independent of female : GAS T-type 3-13-B3264/emm89 was identified in the male ratio. The median age was 75.3 yr (range 12.4- lungs, heart and CSF of the first patient. The strain was 85.0 yr) and 67.2 yr (range 5.5-89.8 yr) for patients who SpeA-, SpeB+ and SpeC-. At autopsy of the second died or survived respectively. T-type 1 constituted 25 patient, GAS T-type 3-13-B3264/emm89 was identified per cent, T3-13-B3264 14 per cent, T28 0 per cent and in tissue from CSF, lungs, peritoneum, blood and rectum. NT 8 per cent of the isolates from the fatal cases, in EKELUND et al: INVASIVE GAS INFECTIONS & GI SYMPTOMS 181 comparison 14, 29, 18 and 11 per cent from the non- SpeA, B and C: Ten isolates (25%) were identified as fatal cases. Three of the 12 cases with fatal outcome SpeA+, SpeB+ and SpeC- and this pattern constituted 21 had meningitis, and only 1 of 28 survivors had GAS and 33 per cent of the isolates from fatal and non-fatal isolated from CSF. Seventy five per cent of the isolates cases, respectively. In eleven isolates (28%) SpeA+, from the fatal cases and 96 per cent from the non-fatal SpeB+ and SpeC- were identified and this pattern was cases were from blood. There were no significant found in 33 per cent of the isolates from fatal cases and differences between the dead and survived patients in 25 per cent of the isolates from non-fatal cases. regarding other primary symptoms, clinical presentations or treatment not mentioned here, but included in the Discussion questionnaire (data not shown). Three fatal cases of invasive GAS infection occurred PFGE: The PFGE patterns of the isolates from the three in Denmark during January 2002. All patients were patients presented here were similar to the PFGE patterns young (12-25 yr old) and without predisposing factors. obtained from GAS isolated elsewhere in Denmark at Clusters of invasive GAS infections were primarily different times and from non-fatal cases (data not shown). described in institutions for elderly or mentally disabled PFGE could not discriminate between the two isolates but could be observed elsewhere7-10. Though PFGE could with T3-13-B3264. The 40 isolates from patients with GI- not discriminate two of the isolates with T-type 3-13- complaints as the primary symptom regardless of outcome B3264, it was concluded that there were no relatedness revealed 29 PFGE-patterns (data not shown). between the cases. The PFGE patterns were identical to the patterns of other both fatal and non-fatal isolates Table. Data from the national surveillance regarding invasive GAS identified in Denmark independent of time and place (data infections in Denmark (January 1999 to June 2002) not shown). PFGE analysis of the 40 isolates from All invasive GAS isolates patients with GI-complaints showed the presence of Dead Alive several clones rather than few virulent T-type-specific clones. Of the 40 infectious episodes, 12 had a fatal n (%) n (%) outcome and no clusters were found regarding to the T- Number 72 (18) 325 (82) type distribution and time of occurrence (data not shown). Female 42 (58) 169 (52) Median age (range) yr 78.1 (4.4-95.7) 56.5 (0.4-97.4) The patient related data were collected as part of the Microbiology national surveillance of invasive GAS infections. In 1999- T-type 1 18 (26) 63 (19) 2001 the response rate to the questionnaire was 99 per T-type 3-13-B3264 16 (22) 78 (24) cent. Because of the delay of clinicians answer the T-type 28 8 (11) 51 (16) response rate was 85 per cent in 2002 at the time of NT 14 (19) 33 (10) data counting. Questionnaire response rates in other Other T-types 16 (22) 100 (31) studies varied from 57-95 per cent11,12. Blood 69 (96) 321 (99) CSF 3 (4) 4 (1) The overall mortality was 18 per cent, which was Primary symptoms similar to what has been described earlier in Denmark, 13-15 GI-complaints 12 (17) 28 (9) Sweden and Norway but higher than described 16-19 Bacteraemia without focus 28 (38) 74 (23)* elsewhere . The mortality rate of 30 per cent among STSS and NF 15 (21) 30 (9)* patients with GI-complaints was comparable to what has Treated in ICU 25 (36) 68 (20)* been reported for patients with STSS or Surgery 15 (21) 93 (28)* meningitis4,18,20,21. Patients with STSS often have GI Artificial ventilation 17 (23) 33 (10)* complaints3,22, more often in patients in the middle age Septic shock 43 (60) 56 (17)* group than in the elderly and very young24. Only 18 per cent of our patients with GI symptoms had STSS or CSF, cerebrospinal fluid; GI, gastrointestinal; NF, necrotizing fasciitis; necrotizing faciitis (NF), regardless of the outcome. ICU, Intensive Care Unit Despite the young age of the three fatal cases, in general *P<0.05 compared to dead the group of patients with GI symptoms was older than 182 INDIAN J MED RES (SUPPL) MAY 2004 the patients with invasive GAS infections disregarding group A streptococcal illness - Spokane, Washington, 1999. the primary symptoms. Though this underlines the fact JAMA 1999; 282: 934-5. that increasing age might be a risk factor for fatal 9. Auerbach SB, Schwartz B, Williams D, Fiorilli MG, Adimora outcome, it emphasizes the characteristics of the three AA, Breiman RF, et al. Outbreak of invasive group A streptococcal infections in a nursing home. Lessons on prevention young patients presented here. The various early clinical and control. Arch Intern Med 1992; 152 : 1017-22. manifestations of severe GAS infections are still the major 10. Schwartz B, Elliott JA, Butler JC, Simon PA, Jameson BL, challenges for clinicians because of the importance of a Welch GE, et al. Clusters of invasive group A streptococcal fast and appropriate diagnosis and immediate treatment. infections in family, hospital, and nursing home settings. This also emphasises the importance of national and Clin Infect Dis 1992; 15: 277-84. international surveillance of invasive GAS infections. 11. Mascini EM, Jansze M, Schellekens JF, Musser JM, Faber JA, Verhoef-Verhage LA, et al. Invasive group A streptococcal disease in the Netherlands: evidence for a protective role of anti-exotoxin Acknowledgment A antibodies. J Infect Dis 2001; 181: 631-8. 12. Efstratiou A. Group A streptococci in the 1990s. J Antimicrob The authors thank Ms Kirsten Burmeister for T-typing of Chemother 2001; 45 Suppl: 3-12. streptococcal isolates, and Ms Aftab Jasir and Mr Jørgen Skov Jensen 13. Henrichsen J, Konradsen HB. Invasive infections caused by for establishing the molecular methods. Streptococcus pyogenes in Denmark 1990-1994. Adv Exp Med Biol 1997; 418 : 201-5. References 14. Svensson N, Oberg S, Henriques B, Holm S, Kallenius G, Romanus V, et al. Invasive group A streptococcal infections in Sweden in 1994 and 1995: epidemiology and clinical spectrum. 1. Cone LA, Woodard DR, Schlievert PM, Tomory GS. Clinical and Scand J Infect Dis 2000; 32 : 609-14. bacteriologic observations of a toxic shock-like syndrome due to Streptococcus pyogenes. N Engl J Med 1987; 317 : 146-9. 15. Hoiby EA, Loevoll OE. Alvorlig gruppe A-streptokokksykdom i Norge 2001. MSIS-Rapport. 2002; Uke 34. 2. Forni AL, Kaplan EL, Schlievert PM, Roberts RB. Clinical and 16. Kiska DL, Thiede B, Caracciolo J, Jordan M, Johnson D, Kaplan microbiological characteristics of severe group A streptococcus E, et al. Invasive group A streptococcal infections in North Carolina: infections and streptococcal toxic shock syndrome. Clin Infect epidemiology, clinical features, and genetic and serotype analysis Dis 1995; 21 : 333-40. of causative organisms. J Infect Dis 1997; 176 : 992-1000. 3. Hoge CW, Schwartz B, Talkington DF, Breiman RF, MacNeill 17. Moses AE, Goldberg S, Korenman Z, Ravins M, Hanski E, EM, Englender SJ. The changing epidemiology of invasive group Shapiro M. Invasive group a streptococcal infections, Israel. A streptococcal infections and the emergence of streptococcal Emerg Infect Dis 2002; 8 : 421-6. toxic shock-like syndrome. A retrospective population-based 18. O’Brien KL, Beall B, Barrett NL, Cieslak PR, Reingold A, Farley study. JAMA 1993; 269 : 384-9. MM, et al. Epidemiology of invasive group a streptococcus disease in the United States, 1995-1999. Clin Infect Dis 2002; 35: 268-76. 4. Stevens DL, Tanner MH, Winship J, Swarts R, Ries KM, Schlievert PM, et al. Severe group A streptococcal infections 19. Eriksson BK, Andersson J, Holm SE, Norgren M. associated with a toxic shock-like syndrome and scarlet fever Epidemiological and clinical aspects of invasive group A toxin A. N Engl J Med 1989; 321 : 1-7. streptococcal infections and the streptococcal toxic shock syndrome. Clin Infect Dis 1998; 27: 1428-36. 5. Johnson DR, Kaplan E, Sramek J, Kriz P, Motlova J, Bicova R, 20. van de Beek D, de Gans J, Spanjaard L, Sela S, Vermeulen M, Dankert et al. Determination of T-protein agglutination patterns. In: J. Group a streptococcal meningitis in adults: report of 41 cases and Laboratory diagnosis of group A streptococcal infections. a review of the literature. Clin Infect Dis 2002; 34: e32-6. Geneva: World Health Organization; 1997 p. 37-41. 21. Kaul R, McGeer A, Norrby-Teglund A, Kotb M, Schwartz B, 6. Jasir A, Tanna A, Efstratiou A, Schalen C. Unusual occurrence O’Rourke K, et al. Intravenous immunoglobulin therapy for of M type 77, antibiotic-resistant group A streptococci in streptococcal toxic shock syndrome - a comparative southern Sweden. J Clin Microbiol 2001; 39 : 586-90. observational study. The Canadian Streptococcal Study Group. Clin Infect Dis 1999; 28 : 800-7. 7. Kristensen B, Egelund B, Meyer M, Henrichsen J, Jepsen OB, 22. Wood TF, Potter MA, Jonasson O. Streptococcal toxic shock- Sievers CJ, et al. An outbreak of Streptococcus pyogenes like syndrome. The importance of surgical intervention. infections in institutions for the mentally retarded in Greater Ann Surg 1993; 217 : 109-14. Copenhagen 1995. Ugeskr Laeger 1996; 158: 1679-82. 23. Schellekens JF, Schouls LM, van Pelt W, Esveld M, van Leeuwen 8. Centers for Disease Control and Prevention. Use of pulsed- WJ. Group A streptococci: a change in virulence? Neth J Med field gel electrophoresis for investigation of a cluster of invasive 1998; 52 : 209-17.

Reprint requests: Dr Kim Ekelund, Streptococcus Unit, Department of Bacteriology, Mycology & Parasitology, Statens Serum Institut, Artillerivej 5, DK-2300 Copenhagen, Denmark e-mail : [email protected] Indian J Med Res 119 (Suppl) Mays 2004, pp 183-185

Induction of myocarditis in rabbits injected with group A streptococci

L.A.Burova, V.A. Nagornev, P.V. Pigarevsky, M.M.Gladilina, I.V. Molchanova, E.A. Gavrilova A.A.Totolian, A. Thern* & C. Schalén*

Institute of Experimental Medicine, Academy of the Medical Sciences, St.Petersburg, Russia & *Department of Medical Microbiology, Dermatology & Infectious Diseases University of Lund, Sweden

Received August 6, 2003

Background & objectives: We have earlier proposed that group A streptococcal (GAS) immunoglobulin binding surface proteins (IgGBPs) might trigger anti-IgG production and immune complex formation leading to glomerulonephritis. In the present study, cardiac tissue material from rabbits injected with heat-killed GAS was investigated. Methods: Rabbits were injected intravenously with 109 colony forming units of streptococci three times weekly for 8 wk. Cardiac tissue samples were obtained at different times and deposition of IgG, C3, TNF-a and IL-6 was studied. Results: After 8 or more weeks of intravenous (iv) injections, minimal changes were seen in animals receiving an IgG non-binding GAS strain, type T27, whereas in those animals receiving either of two IgG binding GAS strains, types M1 or M22, strong inflammatory and degenerative myocardial changes accompanied by deposition of IgG and C3 were noted. Furthermore, on injecting rabbits with defined mutants of a type M22 strain, the development of myocardial tissue damage proved to be dependent on the presence streptococcal IgGBPs. Interpretation & conclusion: The present data supported a role of streptococcal IgGBPs in the induction of myocardial tissue injury by GAS.

Key words Anti-IgG - experimental carditis - IgG binding protein - Streptococcus pyogenes Acute rheumatic fever (ARF) continues to account myocardial inflammatory changes in rabbits injected with for a major public health problem in developing countries various GAS strains. The heart tissue of rabbits injected and has also re-emerged recently in the US1. As with various GAS strains was examined morphologically commonly suggested, rheumatic carditis may be an and immunohistochemically. autoimmune phenomenon due to molecular mimicry between GAS and cardiac myosin or other tissue Material & Methods components2,3. There is ample evidence that both M protein and the hyaluronate capsule are of importance The following streptococcal reference strains were in the pathogenic events leading to ARF. Many GAS included: group A, types M1 (40/58), M22 (10/69) and M-related proteins are known to bind the Fc-part of IgG. T27 (SF40), which were originally obtained from the A role of these in experimental acute glomerulonephritis WHO reference laboratory, Prague. In addition, the triggered by GAS has been previously reported from clinical type M22 isolate AL168 and the following mutants our group4. The present study was carried out to examine were used: AL168mrp-emm-, AL168mrp-, AL168emm. the capacity of IgG-binding proteins (IgGBPs) to induce These mutants were isogenic derivates of AL168, 183 184 INDIAN J MED RES (SUPPL) MAY 2004 affected in either or both of the genes for the M family whereas strain T27(SF40) did not bind IgG (<3%). Both proteins Mrp22 and Emm22, and generated by transposon M1(40/58) and M22(10/69) induced circulating anti-IgG mutagenesis or gene replacement 5. All strains were from at agglutination titres exceeding 1:80 using Ripley-coated Dr G. Lindahl (Lund University, Sweden). Rabbits were cells, thus in contrast to T27(SF40) (<1:10). In three animals injected intravenously with 109 colony forming units (Cfu) injected with T27(SF40), no tissue deposition of IgG or of streptococci in 1 ml PBS three times weekly for 8 C3 was found. In contrast, four rabbits injected with M1 wk. After one month’s rest, injections were continued in (40/58) or M22(10/69) showed a high degree of deposition, some rabbits for a period of 6 more weeks. Thirty located in the sarcolemma, intermyofibrillar spaces and female rabbits of Chinchilla species weighing 2.5-30 kg, on the capillary basement membrane and reaching a obtained from the animal nursery “Rappolovo” (St. maximal intensity already at 4 wk of immunization. In the Petersburg) of the Russian Academy of the Medical same animals, monocytes/macrophages stained for IL-6, Sciences, were used. All experiments with rabbits were a carried out at the Department of Molecular Microbiology, but not for TNF- , whereas in the animals given Institute of Experimental Medicine according to Animal T27(SF40) no staining was observed. Welfare Assurance # A5243-01. Strain T27(SF40) did not induce any tissue alterations Heart tissue samples were obtained at different times (Fig. d). In contrast, after 4 wk of injections with from 7 animals immunized with M1(40/58), M22(10/69) or T27(SF40) and 23 rabbits injected with strain AL168 or its isogenic mutants by sacrificing randomly chosen rabbits after 4 and 8 wk of primary immunization or after 6 wk of reimmunization. Deposition of IgG was demonstrated, following fixation of the sample for 12 h in 4 per cent paraformaldehyde and deparaffinization, by goat anti-rabbit IgG antibodies followed by HRP- labelled rabbit anti-goat antibodies (Southern Biotechnology Associates, USA). Complement component C3 was detected by rabbit anti-human C3 (Daco, Denmark), cross-reacting with rabbit C3, followed by HRP-labelled goat anti-rabbit IgG antibodies (Sigma, USA). TNF-a and IL-6 were similarly detected by specific anti-rabbit TNF-a antibodies (AMS Biotechnology, USA) and goat anti IL-6 antibodies (Biosource International), respectively, followed by HRP-labelled rabbit anti-goat antibodies. The slides were viewed in an Axiomat (Opton) microscope. The details of these procedures were described previously4.

For transmission electron microscopy (TEM), tissue samples were incubated with 2.5 per cent glutaraldehyde solution in 0.1 cacodylate buffer, pH 7.4, at 4º C for 3 h. Following washes in the same buffer for 24 h the blocks were fixed with 1 per cent OsO4 in cacodylate buffer, dehydrated in increasing concentrations of alcohol and propylene oxide, and embedded in araldite. The prepared Fig. Degenerative changes of rabbit myocardial tissue after 8 wk material was visualized using a JEM 100B equipment. immunization and 6 wk of reimmunization with GAS type M1 (40/58). (a) Strongly hypertrophic mitochondria with partial or Results complete disintegration of cristae; (b) destruction of reticulum of reticulum and extension of channels into sarcoplasma; (c) disintegration and ruptures of myofibrils; (d) normal myocardial Strains M1(40/58) and M22(10/69) bound 78 and 35 morphology of rabbit immunized with IgG Fc-nonbinding GAS strain, per cent of radiolabelled IgG, respectively, in uptake tests type T27 (TEM X16,000 in a, b, d and X24,000 in c). BUROVA et al : MYOCARDITIS INDUCED BY GROUP A STREPTOCOCCI 185 M1(40/58), myocardial changes characterized by considered not to be rheumatogenic. Furthermore, IgG congestion and swelling of the endomysium, and at 8 wk, binding capacity of GAS strains is more common in OF+ intense degenerative changes involving sarcoplasma and than OF- serotypes7, and may therefore not be required myofibrils were found. Furthermore, disintegration of for rheumatogenicity. Nevertheless, a role of mitochondria, slight degeneration of myofibrils, and a streptococcal IgGBPs in the induction of myocardial pronounced cellular inflammatory reaction were noted. tissue injury by GAS, was strongly supported by our data. At 6 wk of reimmunization, mitochondrial hypertrophy with Though ARF and APSGN are distinct complications of distintegration of cristae, destruction of the sacroplasmatic acute streptococcal infection and probably linked to reticulum, and rupture of myofibrils were noted (Fig. a,b,c). different strains of GAS, our results suggest that, in both In addition, a strong cellular reaction in capillar and cases, bacterial binding of IgG might be of pathogenic precapillar vessels, with accumulation of lymphocytes, importance. Furthermore, previous4 and present data granulocytes, and activated monocytes, and a picture of indicate that induction of anti-IgG formation may be myofibrosis, including proliferation of perivascular crucial in this context. By such triggering, streptococcal connective tissue, damage of the capillary basement IgGBPs thus appear to be involved in the development membrane and endothelial cells, emerged. Similar tissue of tissue inflammatory damage common to these changes, compatible with severe myocarditis, were streptococcal sequelae. observed in a rabbit receiving M22(10/69). Acknowledgment Isogenic mutants of AL168, defective in the genes encoding one or both FcBPs were then tested. The The study was supported by the Russian Foundation for Basic parent strain yielded tissue deposition patterns and Research grant RFBI 02-04-49223), grant of the President of Russia myocardial changes in agreement with M1(40/58) and (NSh-2206.2003.4), the Royal Swedish Academy of Sciences, the M22 (10/69). In contrast, there were no morphological Crafoord and Österlund Foundations, and the Royal Physiographic alterations in four animals receiving the double mutant Society, Lund. AL168 mrp-; emm-. In those animals receiving mutants AL168 mrp- or AL168 emm-, both positive for binding References of IgG though to a lower level than AL168, development of myocarditis was observed in 6 out of 1. Veasy LG, Wiedmeier SE, Orsmond GS, Ruttenberg HD, 8 and 3 out of 7 rabbits injected, respectively. Of Boucek MM, Roth SJ, et al. Resurgence of acute rheumatic fever in the intermountain area of the United States. N Engl J interest, in the animals given either of the single Med 1987; 316 : 421-7. mutants but which did not develop myocarditis, 2. Quinn A, Ward K, Fischetti VA, Hemric M, Cunningham MW. circulating anti-IgG was not detectable, whereas in Immunological relationship between the class I epitope of those developing pathological changes anti-IgG titres streptococcal M protein and myosin. Infect Immun 1998; 66 : ranged between 1:20 and 1:320. 4418-24. 3. Swerlick RA, Cunningham MW, Hall NK. Monoclonal Discussion antibodies cross-reactive with group A streptococci and normal and psoriatic human skin. J Invest Dermatol 1986; 87 : 367-71. In rabbits injected by either of two IgG binding GAS 4. Burova LA, Nagornev VA, Pigarevsky PV, Gladilina MM, Seliverstova VG, Schalen C, et al. Triggering of renal tissue strains, pronounced myocardial changes, accompanied damage in the rabbit by IgG Fc-receptor- positive group A by heavy deposits of IgG and C3 and involving myofibrils, streptococci. APMIS 1998; 106 : 277-87. sarcolemma and other structures, were found. An IgG 5. Thern A, Wastfelt M, Lindahl G. Expression of two different non-binding strain did not elicit any tissue damage. antiphagocytic M proteins by Streptococcus pyogenes of the Moreover, using isogenic mutants of a type M22 IgG OF+ lineage. J Immunol 1998; 160 : 860-9. binding strain, induction of experimental myocarditis was 6. Johnson DR, Stevens DL, Kaplan EL. Epidemiologic analysis found dependent on the presence of at least one IgGBP of group A streptococcal serotypes associated with severe systemic infections, rheumatic fever, or uncomplicated in the test strain. pharyngitis. J Infect Dis 1992; 166 : 374-82. 7. Bessen D, Fischetti VA. A human IgG receptor of group A Though serotype M1 may be sometimes associated streptococci is associated with tissue site of infection and with ARF6, M22 and other OF+ serotypes are usually streptococcal class. J Infect Dis 1990; 161 : 747-54.

Reprint requests: Dr C. Schalen, Department of Medical Microbiology, Dermatology & Infectious Diseases, University of Lund, Sweden e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 186-190

Colour Doppler echocardiography in children with group A streptococcal infection related tic disorders

F. Cardona, A. Romano, G. Cundari, F. Ventriglia*, P. Versacci* & G. Orefici**

Departments of Child Neuropsychiatry & *Pediatrics, University La Sapienza of Rome & **Istituto Superiore di Sanità, Italy

Received August 6, 2003

Background & objectives: A possible relationship has been suggested between tic disorders and streptococcal infections. To understand the complex relationship between streptococcal infections and neuropsychiatric disorders in children the present study was done on colour Doppler echocardiography of patients with possible post-streptococcal tic disorders. Methods: The patients were 23 children (22 males, 1 female) affected by tic disorders, who at the time of the observation presented (or had presented in the past) signs of streptococcal infections temporally related to the onset or recrudescence of tic disorders. Echocardiographic examination and laboratory tests were performed on these children. Results: In 4 cases a mild mitral insufficiency and in 8 cases a minimal mitral insufficiency was seen, all haemodynamically not significant. Follow up studies (up to 1 yr) showed the consistency and persistence of these findings. Of the 12 patients with echocardiographic abnormalities, 10 displayed very high anti streptolysin O (ASO) titres, 5 showed positive cultures for GAS and 9 had abnormal ESR, even if no significant differences were found in respect to patients with tics and normal echocardiography. Interpretation & conclusion: With the caution due to the design of study and to low number of patients, our data seem to indicate that the pathophysiology of GAS-infection related tic disorders is similar to that SC, at least in some cases. Key words Children - echocardiography - GAS infection - tic disorder

In recent years many research groups have stressed exclusion criterion2. The absence of carditis in these a possible relationship between obsessive-compulsive patients has been one of the main criticisms to the disorder, tic disorders and streptococcal infections. In concept of PANDAS and to its putative pathogenetic particular, a post-streptococcal autoimmunity, similar mechanism3,4. to that of Sydenham's Chorea (SC) has been postulated as a possible pathophysiologic mechanism in a subset Here we report preliminary results of an open study of paediatric patients (paediatric autoimmune on color doppler echocardiography of patients with neuropsychiatric disorders associated with possible post-streptococcal tic disorder. This method of streptococcal infections-PANDAS)1. In these reports investigation has been chosen because it may detect the main difference between SC and PANDAS patients mitral regurgitation even when there is no murmur (silent was the absence of signs of arthritis or carditis in the mitral regurgitation), as happens in some cases of latter ones, since the presence of carditis was an rheumatic carditis5.

186 CARDONA et al: COLOUR DOPPLER ECHOCARDIOGRAPHY IN TIC DISORDERS 187 Material & Methods atrio-ventricular valves and a diastolic one across semilunar valves. Basically, the regurgitant jets have high A total of 23 children (22 boys and 1 girl) affected velocity as well as a variable degree of turbulence; by tic disorders were included from those followed at therefore they usually show aliasing inside the cardiac the outpatients division of Developmental chamber receiving the blood flow. Pathologic mitral Neuropsychiatric Department of University La Sapienza regurgitation was defined as meeting the following four of Rome. criteria: (i) length of colour jet > 1 cm; (ii) colour jet identified in at least two planes; (iii) mosaic colour jet; Inclusion criteria in the study were: to have a tic and (iv) persistence of the jet throughout systole. Jet disorder, age under 18 yr, and onset or recrudescence of orientation was also noted. tic disorder temporally related to signs of streptococcal infections (raise in ASO titre with or without GAS positive By means of direct observation of the regurgitant jet throat swab) at the time of study or in the past. All patients mistakes in diagnosis of valvular incompetence are underwent a complete physical and neurological unlikely to occur. Further such technique allows a assessment. semiquantitative assessment of regurgitation amount. The latter evaluation is related to the left atrial backflow The clinical diagnosis of tic disorder was made on magnitude in case of mitral incompetence; the left the basis of the criteria stated by the Tourette Syndrome ventricular outflow tract backflow magnitude in case of Classification Study Group. The actual severity of tic incompetence of aortic valve; and the regurgitant jet disorder was assessed by the Yale Global Tic Severity maximal area - chamber receiving retrograde blood flow Scale (YGTSS). On the basis of a global evaluation of area ratio. the tic disorder, the patients were classified as "mild tics", "moderate tics" and "severe tics". All echocardiographic examinations were separately reviewed by two cardiologists blinded for the clinical Echocardiographic examination and the laboratory diagnosis. tests including complete blood count and erytrocyte sedimentation rate (ESR) were performed the same Results day. The patients had a mean age of 9.6±2.4 yr (range 4 Specific tests for intercurrent (culture of throat swab yr 11 months-13 yr); the mean age of onset of tic disorder specimen) or antecedent (measurement of ASO titre) was 6.1±2.3 yr (range 2-12 yr); the duration of the GAS infections were also carried out. disorder ranged between 0.2 and 7.1 yr (mean 3.4 yr). Among 23 patients, 13 had a family history positive for ASO titres higher than 407 IU were considered tics, 17 were classified as Tourette syndrome, 5 as abnormal, since in a previous study this value was found multiple chronic tics and 1 as transitory tics. The mean to be 2 SD higher than the mean value obtained in a YGTSS score was 23.6 (range 3-52); on the basis of 6 normal control group seen at the time of the study . their overall tic history, 10 patients were classified as mild tics, 9 as moderate tics and 4 as severe tics. In one Colour Doppler echocardiography: The patient an "innocent" cardiac murmur had been reported echocardiographic data collection was performed with in the past during a routine paediatric follow-up. an Acuson Aspen ultrasound scanner with 2.5-5 mHz transducers incorporating colour flow, continuous and Eighteen patients had a recent onset or recrudescence pulsed wave Doppler. Standard views (parasternal long of tics temporally related with signs of streptococcal axis and short axis, apical four and five chambers) for infections and 5 reported this temporal association for mitral and aortic valve imaging were obtained. To assess the past, though not in the twelve months before the valvular incompetence colour-Doppler technique was study. None of the patients had a diagnosis of rheumatic carried out, as it reveals a systolic regurgitant jet across fever in the past. 188 INDIAN J MED RES (SUPPL) MAY 2004 With the exception of one patient diagnosed for innocent cardiac murmur, all patients had a normal physical and neurological evaluation.

Laboratory test showed that 17 of the 23 patients had ASO titres > 407 IU, 16 had ESR > 16 and 9 had throat swabs positive for GAS. Of the 9 patients positive for GAS, 8 had ASO titre >407 IU and ESR >16. In 12 of 23 patients (52%) color Doppler echocardiography showed abnormalities: a mild mitral insufficiency in 4 cases (Figs 1-2) and a minimal mitral insufficiency in 8 cases (Fig. 3), all haemodynamically not significant.

Fig. 3. Parasternal long axis view: minimal regurgitation of the mitral valve.

Table. Laboratory findings in patients with tic disorder with and without echocardiographic abnormalities

With echo Without echo Total abnormality abnormality

No. of patients 12 11 23

Mean ASO titre ±SD (IU) 912±598 617 ±414 729±543

ASO titre > 407 (IU) 10 7 17

ESR >16 9 7 16

Throat swab positive for GAS 5 4 9 Fig. 1. Parasternal long axis view: mild regurgitation of the mitral valve.

All echocardiography positive patients were retested between 6 to 12 months after the first examination. No significant variation of cardiac abnormalities were detected in any of these cases. Of the 12 patients with echocardiographic abnormalities, 10 showed ASO titre > 407 IU, five had a GAS positive swab and 9 had ESR >16 (Table).

All but one of these patients (the echo-positive children) had a recent onset or recrudescence of tics temporally related to streptococcal infection signs.

No difference in mean age, age at onset of tics, duration of symptoms, familiarity or results of the laboratory tests between patients with or without Fig. 2. Apical four chamber view: confirmed mild regurgitation of echocardiographic abnormalities were found the mitral valve. (Table). CARDONA et al: COLOUR DOPPLER ECHOCARDIOGRAPHY IN TIC DISORDERS 189 Discussion for their relationship with streptococcal infections. Should these two groups of patients considered different The population studied consisted of children affected or, if not, which treatment should be suggested? by tic disorders of different types and severity but all temporally related to streptococcal infections. In more This is the first report on echocardiographic than half of the children, there was evidence of examination of patients affected by tic disorder or by consistent and persistent echocardiographic PANDAS. In a previous study2 the authors specifically abnormalities, in particular mitral regurgitation but excluded patients with cardiac abnormalities, and the haemodynamically not significant. All these echo- method used to assess them was not reported. positive patients had a negative history for rheumatic fever and all but one had a silent physical examination In summary, with the caution due to the low number at the entry in the study. of cases studied, our data provide new insights on the complex relationship between streptococcal infections and neuropsychiatric disorders in children. The rate of echo-abnormalities found in our population (52%) was higher than that reported in the literature in References normal children. In one study7 mitral regurgitation was shown in 1.83 per cent of 329 normal children and 1. Garvey MA, Giedd J, Swedo SE. PANDAS: the search for adolescents (but never before the age of 7 yr) and in environmental triggers of pediatric neuropsychiatric disorders. another study8 in 2.4 per cent of 461 children with Lesson from rheumatic fever. J Child Neurol 1998; 13 : structurally normal hearts. 413-23. 2. Swedo SE, Leonard HL, Garvey M, et al. Pediatric autoimmune On the contrary a high incidence of valvular neuropsychiatric disorder associated with streptococcal incompetences, in particular mitral regurgitation, has infections: Clinical description of the first 50 cases. been described in patients affected by rheumatic fever, Am J Psychiatry 1998; 155 : 1122-4. also in those without a clinically evident carditis9. 3. Shulman ST. Pediatric autoimmune neuropsychiatric disorder Similar rates of echocardiographic abnormalities have associated with streptococci (PANDAS). Ped Infect Dis J 1999; 18 : 281-2. been reported in children with isolated rheumatic chorea10. 4. Kaplan EL. PANDAS? Or PAND? Or both? Or neither? Assessing a (possible?) temporal or pathogenetic relationship with the Group A "streptococcal diseases complex". It was suggested that a subclinical valvitis, Contemp Pediatr 2000; August: 81-96. echocardiographically distinguishable from "innocent" 5. Mimich LL, Tani LY, Pagotto LT, Shaddy RE, Veasy LG. Doppler 5 regurgitations , should be accepted as evidence of echocardiography distinguished between physiologic and carditis in the diagnosis of rheumatic fever11. If this pathologic "silent" mitral regurgitation in patients with rheumatic opinion is correct, we can assume that at least some of fever. Clin Cardiol 1994; 20 : 924-6. our patients were suffering by a tic disorder plus an 6. Cardona F, Orefici G. Group A streptococcal infections and tic actual or recent streptococcal infection plus a minimal disorders in an Italian pediatric population. J Pediatr 2001; carditis. Could the diagnosis of "rheumatic tic disorder" 138 : 71-5. be appropriate for these children? And, in this case, 7. Thomson JD, Allen J, Giggs JL. Left sided valvar regurgitation which type of treatment should be used? With these in normal children and adolescents. Heart 2000; 83 : 185-7. cardiac abnormalities long standing should these patients 8. Brand A, Dollberg S, Keren A. The prevalence of valvular receive anti-inflammatory drugs and antibiotic regurgitation in children with structurally normal hearts: a color Doppler echocardiographic study. Am Heart J 1992; 123 : prophylaxis? Or as suggested by Murphy and Pichichero 177-80. for children with PANDAS12, is it sufficient to treat 9. Folger GM Jr, Hajar R, Robida A, Hajar HA. Occurrence of the incoming streptococcal infection ? Our echo-positive valvar heart disease in acute rheumatic fever without evident patients were not distinguishable from the negative ones, carditis: color-flow Doppler identification. Br Heart J 1992; either for the clinical characteristics of tic disorders or 67 : 434-8. 190 INDIAN J MED RES (SUPPL) MAY 2004

10. Elevli M, Celebi A, Tombul T, Gokalp AS. Cardiac in the diagnosis of acute rheumatic fever? Cardiol Young 2001; involvement in Sydenham'chorea: clinical and Doppler 11 : 255-60. echocardiographic findings. Acta Paediatr 1999; 88 : 12. Murphy ML, Pichichero ME. Prospective identification and 1074-7. treatment of children with pediatric autoimmune 11. Ozkutlu S, Ayabakan C, Saraclar M. Can subclinical valvitis neuropsychiatric disorders associated with group A streptococcal detected by echocardiography be accepted as evidence of carditis infections. Arch Pediatr Adolesc Med 2002; 156 : 356-61.

Reprint requests: Dr Graziella Orefici, Istituto di Sanità, Viale Regina Elena 299, 00161-Rome, Italy e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 191-196

Diversity of surface protein expression in group B streptococcal colonizing & invasive isolates

P. Ferrieri, C.J. Baker*, S.L. Hillier† & A.E. Flores

University of Minnesota Medical School, Minneapolis, MN, *Baylor College of Medicine, Houston, TX, †University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

Received August 6, 2003

Background & objectives: The classification of group B streptococcal (GBS) isolates is based on the capsular polysaccharides (Ia-VIII), and antigenic characterization of clinical isolates is augmented by the detection of various surface-localized protein antigens. In our laboratory, all GBS isolates are routinely analysed for the alpha trypsin-resistant and the beta trypsin-sensitive c protein antigens, as well as other trypsin-resistant proteins R1, R3, and R4, as well as BPS. The purpose of this work was to study diversity of protein expression in colonizing isolates (vaginal and rectal sites) from non- pregnant women and from invasive isolates (blood or CSF) from mothers and their less than seven day old newborn infants. Methods: A total of 289 invasive isolates and 2660 colonizing isolates were collected between 1993-2002. All isolates were tested for polysaccharide serotype and cell surface-expressed protein profile by double immunoprecipation in agarose using monospecific antisera. Results: Among the 289 invasive isolates, 89.6 per cent expressed one or more trypsin-resistant proteins; 93 per cent of the colonizing isolates expressed one or more of these proteins. Overall, the most common surface protein expression profile by GBS serotype was: alpha in type Ia; alpha plus beta in type Ib; alpha and R4 in type II; R4 in type III; and co-expression of R1 plus R4 in isolates of type V. BPS was found in five (1.7%) invasive isolates, alone in two isolates and with other proteins in three isolates. Among 2660 colonizing isolates, BPS was found alone in 15 (0.6%) and in 57 additional isolates with other proteins. Among the total isolates, BPS was found predominantly in serotype Ia isolates, also expressing R1. Uncommon protein profiles of known serotypes included 11 type III isolates expressing alpha plus beta. Among 72 nontypable colonizing isolates, expression of R1 plus R4 was the commonest (33.3%) profile. Interpretation & conclusion: The GBS surface proteins and the common serotypes were distributed comparably in colonizing and invasive isolates. Trypsin-resistant, alpha and alpha-like proteins, R1 and R4 were the most prevalent. The phenotypic diversity of the surface-localized protein antigens of GBS is intriguing, and genotypic analysis will permit consensus in nomenclature from laboratory to laboratory. Key words Alpha protein - GBS serotypes - R proteins The importance of group B streptococci (GBS) as a phenotypic characterization of surface localized protein major perinatal pathogen for invasive disease has been antigens of GBS, such as the c protein, comprising the well documented1-3. There are now nine distinct capsular alpha (trypsin-resistant) and beta (IgA binding and trypsin polysaccharide (CPS) types of GBS, designated Ia, Ib, susceptible) antigens that are commonly found in GBS II, III, IV, V, VI, VII, VIII1,4-6. In addition to the CPS types Ia, Ib, and II7-12. Other trypsin-resistant proteins, types, we have investigated basic immunologic, and known as R proteins4,10,11,13,14 have been described by 191 192 INDIAN J MED RES (SUPPL) MAY 2004 us originally in types II and III GBS10. The R proteins Prototype strains and typing antisera: The prototype have been found to occur, with some exceptions, strains, with their original source designations, used as independently of the c protein4,10. We have also controls were from the University of Minnesota collection demonstrated that some molecular species of the R and are as follows: group B: O90R (Lancefield); Ia: S3- proteins occur in one of the more newly described CPS 3630P (U of M); Ib: H36B (Lancefield); II: 18RS21/19/ types, i.e., type V4,15. 2 (Lancefield); III: M781 (Houston); IV: Wilkinson 3139 (Rabinowitz, Israel); V: SS-1169 (Prague); VI: SS-1214 Characterization of the surface protein markers of (Prague); VII: Perch “7271” (Prague); VIII: JM9- the GBS clinical isolates is important since there is 130013 (Boston/Prague); c protein (alpha and beta evidence that the c protein, particularly the alpha protein, components): A909 (Lancefield); alpha only: 80-426 (U as well as other trypsin-resistant proteins that resemble of M); beta only: 81-418 (U of M); R1: D136C the alpha protein, may contribute to strain differences in (Lancefield); R3, R4, and BPS (group B protective virulence, are immunogenic, and antibodies to them may surface protein): Compton 25/60 (Prague). The typing be protective11,16-19. The designation of other alpha-like antisera, prepared as we have described before by repetitive molecules, such as Alp2 and Alp3, is under immunizing rabbits with formalin-killed whole bacterial study20, but at this time we shall refer to them as the R cells, were made monospecific by cross-absorption7,10,11. proteins, of which R1 and R4 are the major ones found in the clinical GBS isolates4,10,15,21. The Rib protein Antigen extraction and serotyping by described by Wästfelt et al22 has been thought by us4,10,15 immunoprecipitation in agarose: Testing for the and Norwegian investigators21 to be the same as R4 polysaccharide serotype and cell surface-expressed protein, based on epidemiological and antigenic studies, protein profile of each isolate was done, using a hot HCl but recently by means of gene sequencing we extract prepared from the cell pellet of overnight growth demonstrated that the R4 protein is identical to the in Todd-Hewitt broth3,7,10,11. Extracts were tested by 23 sequence of Rib . The gene sequence of R5 surface double immunoprecipitation in agarose using the protein has been reported, as well as, data indicating monospecific antisera to GBS capsular polysaccharide that antibodies to this surface protein are protective, and types Ia, Ib – VIII and to the c (alpha and beta components) this protein has been renamed group B streptococcal and R (R1, R3, R4 species) proteins. The extracts also protective surface protein (BPS)24,25. To further study were tested with antiserum to the group B surface the diversity of surface protein expression among GBS protective protein (BPS, formerly known as R5)24,25. isolates and to compare any differences between Extracts of control prototype strains were included in all colonizing isolates, and blood and cerebral spinal fluid tests and positive identification was based on observing a (CSF) isolates from invasive infections in mothers and their infants less than seven days of age, a total of 2660 precipitin band with a reaction of identity between the colonizing and 289 invasive GBS isolates were studied. extracts of the clinical and the control isolates.

Material & Methods Results

GBS clinical study isolates: A total of 2949 GBS isolates Serotypes Ia, III, and V were the three most common collected between 1993 and 2002 were examined. The serotypes recovered from nonpregnant women, followed 289 invasive isolates were collected in Houston, by serotype II. There were a few isolates of serotypes Minneapolis-St. Paul, Pittsburgh and Seattle from blood IV, VII, and VIII identified. A low number of the isolates, or cerebrospinal fluid (CSF) of mothers or newborn 2.2 per cent, were nontypable (Fig.1). Among invasive infants with early-onset disease while the 2660 colonizing isolates, serotypes Ia, III, and V also were the most isolates were collected in Pittsburgh from rectal (R) or prevalent. There was a single isolate of serotype VII vaginal (V) cultures of non-pregnant women. Isolates and three isolates (1.0%) were nontypable (Fig.1). were sent on blood agar slants to the GBS Molecular Reference Laboratory at the University of Minnesota The protein expression profiles of these 2660 GBS where they were stored frozen at -30°C in Todd-Hewitt colonizing isolates (Fig.2) revealed that approximately blood broth until they were serotyped. 92.6 per cent expressed one or more of the trypsin- FERRIERI et al: GROUP B STREPTOCOCCAL SURFACE PROTEINS 193 resistant proteins. The strains expressing BPS with examined phenotypically. For both colonizing and invasive another surface protein, usually the alpha protein were isolates, BPS expression was uncommon. The IgA also included. The most common pattern was expression binding beta protein26,27 was expressed in only a minority of alpha alone for 40.5 per cent or R4 alone for 27.3 per of invasive and colonizing isolates, 8.6 and 10.6 per cent, cent. Only 6.6 per cent of the isolates expressed none respectively. of the above described surface proteins. Among the 289 invasive isolates a similar distribution of surface protein The predominant surface protein expressions among expression was seen when compared to that of the the colonizing and invasive isolates, respectively, were colonizing isolates, and 88.9 per cent expressed one or alpha alone for 92 and 97 per cent in serotype Ia more trypsin-resistant proteins. Among these isolates isolates; alpha plus beta proteins for 86 and 81 per cent 10.4 per cent did not express the proteins that were for serotype Ib; alpha only in 64 and 57 per cent of

Fig.1. Comparison of CPS serotype distribution between invasive and colonizing isolates of group B streptococci. NT, nontypable.

Fig.2. Cell surface proteins expressed by invasive and colonizing isolates of group B streptococci. BPS, group B streptococcal protective surface protein. 194 INDIAN J MED RES (SUPPL) MAY 2004 serotype II; R4 for 95.6 and 96 per cent in serotype The great similarities in cell surface protein expression III; and R1 plus R4 for 71.4 and 53.8 per cent of profiles between the colonizing isolates and the invasive serotype V. GBS isolates suggest that there may be no single protein that enhances for virulence of the organism. Similarly, The surface-localized protein antigens were also although the women were not the mothers of the infected studied among 72 nontypable isolates and 79.2 per cent newborn infants, the CPS types were similar among the expressed one or more of the trypsin-resistant proteins, colonizing and invasive isolates. The higher prevalence with a predominance of R1 and R4 proteins in one-third of nontypable isolates among the colonized women of these isolates. compared to rare nontypable invasive isolates, may reflect loss of capsule in the genital tract of women Discussion colonized for long periods.

Prior to our studies over the past decade, isolates The dual presence of R1 and R4 proteins in GBS from neonates with early-onset infections were found isolates, predominantly in type V was first described in to be evenly divided among serotypes I, II, and III and our laboratory4,15. The R4 protein has been studied the majority of those with early-onset sepsis/meningitis extensively in our laboratory at the phenotypic, as well and the vast majority of those from late-onset infection as molecular level10,11,23. We have demonstrated its were caused by serotype III1-5. In our prospective immunogenicity in rabbits and humans and the surveillance studies, we have shown that there has been transplacental passage of anti-R4 antibody from mothers a shift with the current CPS types causing early-onset to infants16. In the present study, R4 was seen alone maternal and infant infection to be serotypes Ia, III, and primarily in serotype III (>95%), and in serotype II V1-5. Serotype V has emerged in the United States and (20-22%) and in serotype V (54-75%) either alone or in Europe not only as a cause of GBS genital colonization combination with R1. More recently, Smith et al23 in nonpregnant women, but it is a prominent cause of sequenced the R4 gene from a clinical isolate of serotype early-onset invasive disease in newborn infants and is III and showed that the nucleotide sequence was identical the major serotype recovered from invasive infections to the rib gene. in nonpregnant adults1-5,15. The reason for the emergence of serotype V over the past decade is unclear, but this is Over the past few years we have discovered a reminder that more newly described GBS uncommon protein profiles among GBS of known CPS polysaccharide serotypes, i.e., VIII, may emerge and type with patterns never described before. For example, replace the currently defined serotypes. For example, we have found the presence of the R1 and R4 proteins serotype VIII is a predominant cause of GBS colonization in CPS type Ib, and the alpha and beta proteins, expressed in women, as well as a pathogen for newborn infants in without R4, in serotype III (Ferrieri P, and Flores AE, Japan6. unpublished observations). It is possible, that significant genetic exchange is occurring within the genital tract of Our laboratory was the first to describe the ladder- women, particularly in the presence of simultaneous like appearance of the alpha protein by Western colonization with more than one serotype, and lengthy immunoblotting (WB)4,11. We also described the periods of colonization, which may enhance for these ladder-like effect of the R1 and R4 proteins by WB4,11. potential genetic events. The genetic basis for this observation was studied by Michel et al28, documenting the presence of tandem The importance of characterizing, at an antigenic level, repeating units in the alpha antigen gene. The alpha these other epidemiological protein markers of GBS is protein is the prototype for comparison among these that they may play a future role in development of trypsin-resistant proteins, which also include Alp2 and capsular polysaccharide protein conjugate GBS Alp3 that may be highly related to R1 and R4 proteins20. vaccines2. The alpha protein alone17 or alpha protein in Antigenic diversity, as well as similarity, within this family combination with Rib protein18 has shown of proteins is startling, and it may be due to genetic immunogenicity and protection in infant mice. Examples recombination. of other highly conserved surface-localized proteins FERRIERI et al: GROUP B STREPTOCOCCAL SURFACE PROTEINS 195 (expressed by 100% of strains tested) include the C3 7. Johnson DR, Ferrieri P. Group B streptococcal Ib/c protein binding protein29 and surface immunogenic protein Sip30, antigen: distribution of two determinants in wild-type strains which may hold promise as future vaccine candidates. of common serotypes. J Clin Microbiol 1984; 19 : 505-10. 8. Ferrieri P, Johnson DR, Flores AE. Ibc protein antigens in In summary, the GBS surface proteins and the culture supernatants of group B streptococci: comparison to common serotypes were distributed comparably in large HCl-extracted antigens. In: Kimura Y, Kotami S, Shiokawa Y, editors. Recent advances in streptococci and streptococcal numbers of invasive and colonizing isolates studied. diseases. Proceedings of the IXth Lancefield International Trypsin-resistant proteins, alpha and the alpha-like Symposium on Streptococci and Streptococcal Diseases; 1984 proteins, R1 and R4, were the most prevalent. September 10-13; Mt. Fuji, Japan. Reedbooks Ltd: Berkshire, Nontypable GBS isolates frequently express these England; 1985 p. 204-6. proteins, which can serve as molecular epidemiological 9. Ferrieri P. Surface-localized protein antigens of group B markers in the absence of a defined capsular streptococci. Rev Infect Dis 1988; 10 Suppl 2 : S363-6. polysaccharide antigen. Ultimately, resolution of 10. Flores AE, Ferrieri P. Molecular species of R-protein antigens phenotypic similarities among these surface proteins will produced by clinical isolates of group B streptococci. be determined by DNA sequencing. Combinations of J Clin Microbiol 1989; 27 : 1050-4. these protein, with or without polysaccharides, may prove valuable as GBS vaccines. 11. Flores AE, Ferrieri P. Molecular diversity among the trypsin resistant surface proteins of group B streptococci. Zbl Bakt 1996; 285 : 44-51. Acknowledgment 12. Russell-Jones GJ, Gotschlich EC. Identification of protein antigens of group B streptococci, with special reference to the Authors acknowledge the financial support from the National Ibc antigens. J Exp Med 1984; 160 : 1476-84. Institutes of Health (Institute of Allergy and Infectious Diseases), USA, Research Contract N01-AI-75326. 13. Lancefield RC, Perlmann GE. Preparation and properties of a protein (R antigen) occurring in streptococci of group A, type 28 and in certain streptococci of other serological groups. References J Exp Med 1952; 96 : 83-97. 14. Wilkinson HW. Comparison of streptococcal R antigens. 1. Baker CJ, Edwards MS. Group B streptococcal infections. In: Appl Microbiol 1972; 24 : 669-70. Remington JS, Klein JO, editors. Infectious diseases of the fetus and newborn infant. 5th ed. Philadelphia: WB Saunders; 2000 15. Benson JA, Flores AE, Baker CJ, Hillier SL, Ferrieri P. Improved methods for typing nontypeable isolates of group B streptococci. p. 1091-156. Int J Med Microbiol 2002; 292 : 37-42. 2. Ferrieri P. Prevention of group B Streptococcus infections. Semin 16. Fasola EL, Flores AE, Ferrieri P. Immune responses to the R4 Infect Dis 1997; 8 : 117-24. protein antigen of group B streptococci and its relationship to 3. Zaleznik DF, Rench MA, Hillier S, Krohn MA, Platt R, Lee M- other streptococcal R4 proteins. Clin Diagn Lab Immunol 1996; LT, et al. Invasive disease due to group B Streptococcus in 3 : 321-5. pregnant women and neonates from diverse population groups. 17. Gravekamp C, Kasper DL, Paoletti LC, Madoff LC. Alpha C Clin Infect Dis 2000; 30 : 276-81. protein as a carrier for type III capsular polysaccharide and as a protective protein in group B streptococcal vaccines. 4. Ferrieri P, Flores AE. Surface protein expression in group B Infect Immun 1999; 67 : 2491-6. streptococcal invasive isolates. In: Horaud T, Bouvet A, Leclercq 18. Larsson C, Stålhammar-Carlemalm M, Lindahl G. Protection R, de Montclos H, Sicard M, editors, Streptococci and the host. against experimental infection with group B streptococcus by Proceedings of the XIII Lancefield International Symposium on immunization with a bivalent protein vaccine. Vaccine 1999; Streptococci and Streptococcal Diseases; 1996 September 16- 17 : 454-8. 20; Paris, France. New York: Plenum Press; 1997 p. 635-7. 19. Michel JL, Madoff LC, Kling DE, Kasper DL, Ausubel FM. 5. Hickman ME, Rench MA, Ferrieri P, Baker CJ. Changing Cloned alpha and beta C-protein antigens of group B epidemiology of group B streptococcal colonization. Pediatrics streptococci elicit protective immunity. Infect Immun 1991; 1999; 104 : 203-9. 59 : 2023-8.

6. Matsubara K, Katayama K, Baba K, Nigami H, Harigaya H, 20. Lachenauer CS, Creti R, Michel JL, Madoff LC. Mosaicism in Sugiyama M. Seroepidemiologic studies of serotype VIII group the alpha-like protein genes of group B streptococci. B Streptococcus in Japan. J Infect Dis 2002; 186 : 855-8. Proc Natl Acad Sci, USA 2000; 97 : 9630-5. 196 INDIAN J MED RES (SUPPL) MAY 2004

21. Bevanger L, Kvam AI, MÆland JA. A Streptococcus agalactiae 26. Ferrieri P, Flores AE, Concepcion NF, Anthony BF. Bacterial R protein analyzed by polyclonal and monoclonal antibodies. surface expression and recognition of the IgA binding protein of APMIS 1995; 103 : 731-6. group B streptococci. In: Orefici G, editor. New perspectives on streptococci and streptococcal infections. Zbl Bakt 1992; 22. Wästfelt M, Stålhammar-Carlemalm M, Delisse A-M, Cabezon Suppl 22 : 186-8. T, Lindahl G. Identification of a family of streptococcal surface proteins with extremely repetitive structure. J Biol Chem 1996; 27. Russell-Jones GJ, Gotschlich EC, Blake MS. A surface 271 : 18892-7. receptor specific for human IgA on group B streptococci possessing the Ibc protein antigen. J Exp Med 1984; 160 : 23. Smith BL, Flores A, Dechaine J, Krepela J, Bergdall A, Ferrieri 1467-75. P. Gene encoding the group B streptococcal protein R4, its presence in clinical reference laboratory isolates & R4 protein 28. Michel JL, Madoff LC, Olson K, Kling DE, Kasper DL, Ausubel FM. Large, identical, tandem repeating units in the C protein pepsin sensitivity. Indian J Med Res 2004; 119 (Suppl) : alpha antigen gene, bca, of group B streptococci. 213-20. Proc Natl Acad Sci USA 1992; 89 : 10060-4. 24. Flores AE, Erdogan S, Chhatwal GS, Baker CJ, Hillier S, Ferrieri 29. Smith BL, Ferrieri P. Identification, conservation, and P. Expression of the R5 surface protein among polysaccharide suggested immunogenicity of GBbca, a group B streptococcal serotypes of group B streptococci. In: Martin DR, Tagg JR, surface protein that binds complement protein C3. In: Martin editors. Streptococci and streptococcal diseases entering the new DR, Tagg JR, editors. Streptococci and streptococcal diseases millennium. Proceedings of the XIVth Lancefield International entering the new millennium. Proceedings of the XIVth Symposium on Streptococci and Streptococcal Diseases; 1999 Lancefield International Symposium on Streptococci and Streptococcal Diseases; 1999 October 11-15; Auckland, New October 11-15; Auckland, New Zealand. Securacopy: Auckland, Zealand. Securacopy: Auckland, New Zealand; 2000 p. New Zealand; 2000 p. 191-4. 649-52. 25. Erdogan S, Fagan PK, Talay SR, Rohde M, Ferrieri P, Flores 30. Brodeur BR, Boyer M, Charlebois I, Hamel J, Couture F, Rioux AE, et al. Molecular analysis of group B protective surface CR, et al. Identification of group B streptococcal Sip protein, protein, a new cell surface protective antigen of group B which elicits cross-protective immunity. Infect Immun 2000; streptococci. Infect Immun 2002; 70 : 803-11. 68 : 5610-8.

Reprint requests: Dr Patricia Ferrieri, MMC 134, 420 Delaware St. SE, University of Minnesota Medical School Minneapolis, MN 55455, USA e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 197-200

Towards a Belgian consensus for prevention of perinatal group B streptococcal disease

P. Melin, G. Verschraegen*, L. Mahieu**, G. Claeys* & P. De Mol

Belgian Reference Laboratory for Group B Streptococci, Medical Microbiology, University Hospital of Liege, *Medical Microbiology, University Hospital of Gent & **Neonatology, University of Antwerp, Belgium

Received August 6, 2003

Background & objectives: In Belgium, as in many other countries, group B Streptococcus (GBS) is still the leading cause of sepsis and meningitis in neonates. In 2001, though no Belgian guidelines for their prevention were available, in some hospitals, obstetrical programmes included a GBS prevention policy. With an aim to reach a Belgian consensus for the prevention of perinatal group B streptococcal disease, a national consensum meeting was organized in 2001. We report here our experience and findings of this meeting. Methods: In November 2001, obstetricians, neonatologists, microbiologists and infectious diseases specialists were invited to participate in a GBS symposium. International and Belgian speakers presented epidemiological aspects, argued comparative cost-effectiveness of different approaches for prevention and debated technical and practical problems. Management of neonates with risk factors for GBS disease and progress in GBS vaccines were also included in the programme. Further results about Belgian obstetricians’ practice and compliance to a policy for prevention of neonatal GBS diseases, as answered in two mail surveys, were commented and discussed. In an interactive session at the end, each participant was asked to vote on the key points related to the different steps of the ideal prevention strategy to recommend. Results: For the main questions, 94 per cent of participants choose a screening-based approach and 94 per cent shifted from the current use of ampicillin to penicillin as first choice for antimicrobial prophylaxis. Further, 79 per cent voted for an approach with integrated neonatal prophylaxis for selected neonates at high risk for GBS disease and 47 per cent voted for a strategy based on an intrapartum rapid screening-based approach. Interpretation & conclusion: The state of the question by different speakers, the data from Belgian epidemiology, and the debate about cost-effectiveness of different approaches led to a massive vote in favour of the universal screening-based approach. Based on these results, a working group has been appointed by the Ministry of Health to draft and edit Belgian recommendations for the prevention of perinatal GBS disease. Key words Group B Streptococci - prevention - neonatal infection In Belgium, as in many industrialized countries, since More recently a short decline in GBS occurrence has the 1980s, group B streptococcus (GBS) has been and been reported (Public Health Institute and P. Melin, remains the most important pathogen causing neonatal unpublished data). invasive infections1-4. From 1985 to 1998, GBS caused 35 per cent of neonatal (0-28 days) sepsis and meningitis, Our interest in GBS research has expanded since next followed by Escherichia coli (18 %) and others4. the last two decades and this period was marked by key

197 198 INDIAN J MED RES (SUPPL) MAY 2004 steps. It started with a one-year prospective study When the risk-based approach was considered for the including clinical, bacterial and seroepidemiological decision to give an IAP, more than 90 per cent of the Fr aspects of 1350 pregnant women and their newborns5. obstetricians looked for intrapartum fever and prolonged Through the next 15 yr, several studies were completed. rupture of membranes, and 58 per cent considered They covered epidemiological and technological aspects prematurity also as a risk factor. Among Fl obstetricians and took place more often in the French community of respectively 57, 36 and 36 per cent considered the three Belgium6-11. In 1995, the Belgian Public Health Institute, more frequent risk factors to decide on IAP. For their epidemiology section, designated and identified a antibioprophylaxis, most of the obstetricians seemed to reference laboratory for GBS in University Hospital of have forgotten than penicillin G still existed, and preferred Liege. amoxycilline. Furthermore, regimen was often inadequate, and more frequently for the Fl obstetricians. In the early 1980s, the reported incidence of GBS Both surveys demonstrated room for improvement and early onset disease (EOD) was 3 per 1,000 live births5. need for coordination between medical partners. In 1990, as expected, it had not changed, but in addition the occurrence of 4 likely cases per 1,000 (clinically and To reduce the perinatal GBS burden, a need for biologically consistent with a GBS EOD, but not proved updated, widely accepted guidelines was obvious. by culture) was reported. In 1999, the estimated incidence Therefore we organized in November 2001 a national was 2 cases per 1,000 live births and 2 likely cases were consensus meeting for the prevention of perinatal GBS still reported for each confirmed case12. Through 1999 infections. Here we described how this meeting was to 2001, based on the data collected by the reference organized and we reported the conclusive results and laboratory, 4 EOD seemed to occur for 1 late onset consequences. disease. Among the EOD, 10 per cent of the neonates had a meningitis. Based on the reported outcomes of Material & Methods EOD, the mortality rate was 14 per cent13. In sixty per cent of cases with EOD no maternal risk factors except All obstetricians, neonatologists, microbiologists and 12 vaginal colonization was present . The GBS carriage infectious diseases specialists of the country were invited rate among pregnant women ranged from 13 to 25 per to participate at a consensus meeting in November 2001. 8 cent according to hospital location and type . The meeting was supported by the different Belgian professional societies in gynecology-obstetrics, in In 2000, as in many European countries, though no infectious diseases, in microbiology and in neonatology. guidelines for the prevention of perinatal GBS diseases were available, in some institutions, obstetrical The invited speakers were S. Schrag from the CDC, programmes included a GBS prevention policy. To Atlanta, USA, to state the global GBS neonatal burden, evaluate the Belgian GBS practices for prevention of to present the guidelines issued by the CDC in 1996, and perinatal GBS disease, two mail surveys, covering the to argue benefits and adverse effects observed or to whole country, were conducted: one in the French expect. D. Davies, from Canada, gave the neonatologist Community (Fr) in 1998-1999 and one in the Flemish view and reported the Canadian experience. B. Brodeur, Community (Fl) in 19996,11. For this evaluation, the also from Canada, stated the perspectives of the vaccine guidelines for prevention of GBS perinatal diseases, approach; and, M. de la Rosa, from Spain, reported the issued by the Centers for Disease Control (CDC) in 1996, successful Spanish experience ongoing since a few years were used as Gold standard14. CDC recommended an for the prevention of perinatal GBS diseases. Belgian intrapartum antimicrobial prophylaxis (IAP) for selected speakers presented national epidemiological aspects, women, identified either on a risk-based or a screening- argued cost-effectiveness of the different approaches, based approach. Through the answers of those who and debated microbiological, logistic and special problems. accepted to participate, the Belgian surveys showed Results of the two surveys, previously described, were significant geographical differences in the obstetricians’ reported, commented and discussed. clinical practice. Among the Fr obstetricians, 90 per cent had chosen the prenatal screening approach, whereas An interactive session was scheduled at the end of Fl obstetricians preferred the risk-based approach. the symposium. For this session, each participant had MELIN et al: BELGIAN CONSENSUS FOR PREVENTION OF PERINATAL GBS DISEASE 199 received an individual keypad to vote electronically “ in Voting for an alternative agent for penicillin-allergic real time” on each key point of a strategy to recommend women, 59.1 per cent participants voted for a first in Belgium for the prevention of perinatal GBS diseases. generation cephalosporin, 25.7 per cent clindamycin, To reach a consensus, an 80 per cent cut-off of 7.6 per cent erythromycin and 7.6 per cent had no agreement between participants was fixed. As many opinion. obstetricians from the French community had another important meeting the same day, a new conference, Many obstetricians from the Fr community attended summing up the consensus meeting, was scheduled three the conference summing up the consensus meeting. All weeks later. of them agreed with the perspective of a universal prenatal screening-based strategy and with penicillin G Results as first choice for IAP.

Before starting the vote for the strategy to Very soon after the meeting, the chosen universal recommend, the participants answered to questions screening-based approach and penicillin G for IAP for regarding their current practice. The majority of the the prevention of perinatal GBS disease were participants (67 %) practiced in the North of Belgium recommended by both Flemish and French Societies of (the Fl community), whereas only 18 per cent came from Gynecology and Obstetrics to their members. the South (Fr community) and 15 per cent from Brussels. Non-university hospitals or clinics were represented by These data and the evidence of GBS burden in the 48 per cent of the participants; 39 per cent practiced in Belgian epidemiology of neonatal infections drew some university hospitals and 13 per cent practiced in private attention and, a GBS working group was appointed by and non university hospitals or clinics. the Health Council (Belgian Federal Ministry of Health). The working group has to draft and edit guidelines to Seventy per cent participants answered that they recommend in Belgium for the prevention of neonatal already had a policy in agreement, or closed to the GBS infections by early 2003. screening-based approach of the CDC, 10 per cent had a mixed behavior according to the type of population, 3 Discussion per cent had no policy at all and the others did not know. Associated with high morbidity and mortality, infection Strategy to recommend: A great majority, 94 per cent due to Streptococcus agalactiae either in newborn of the participants, voted for a universal prenatal infants or adults and its prevention is a major public health screening strategy, based on screening cultures for problem. Since last two decades several Belgian studies vaginal and rectal GBS colonization of all women at 35- not only stressed the importance of GBS burden in 37 wks’ gestation. No one preferred a risk-based neonatal infections, but also the need for “Belgian” approach, 4 per cent chose another approach and 2 per recommendations for their prevention, to reduce their cent had no opinion. Seventy nine per cent participants incidence and related morbidity and mortality. voted for an approach which would integrate an antimicrobial prophylaxis for neonates at high risk, as In the consensus meeting organized in November those born to a mother with a chorioamnionitis. Nearly 2001, a consensus was reached for the two main half of the participants (47%) voted for a potential questions which were the strategy to recommend for strategy based on a rapid intrapartum screening when a identifying the pregnant women who should be given an clinically proven effective rapid test would become IAP and, the antimicrobial agent to be given for IAP. available. The strategy for which Belgian practitioners and experts voted was a universal prenatal screening based-strategy; For the intrapartum antibioprophylaxis, 92.4 per cent no one of the participants wanted a risk-based strategy. were convinced to give penicillin G as the drug of first With this vote, Belgium was a few months in advance, choice; only a few participants, 3.8 per cent did not in accord with the revised guidelines issued by CDC in agree with this choice and 3.8 per cent had no opinion. August 200215. The state of the question by the different 200 INDIAN J MED RES (SUPPL) MAY 2004 speakers, the data from Belgian epidemiology, and the processing. In: 40 th Intersciences Conference on Antimicrobial debate about cost-effectiveness of different approaches Agents and Chemotherapy (ICAAC), September 2000, Toronto, Canada, Abstract book, p. 459, no. 1746. were effective in gathering a majority of the participants for a massive vote in favour of the universal screening- 7. Melin P, Rodriguez Cuns G, Vicentino Fernandez W, De Mol P. based approach. This vote for future guidelines to be Antimicrobial susceptibility of Streptococcus agalactiae isolated recommended in Belgium showed the impact of the from patients in Belgium through 1989-1991 and 1996-1999. Proceedings of the XIV Lancefield International Symposium on programme of the meeting as 94 per cent participants Streptococci and Streptococcal Diseases; 1999 October, voted when only 70 per cent of the participants declared Auckland, New Zealand; November 2000, 305-9. to comply currently with this approach. By comparison 8. Melin P, Schmitz M, Tsobo C, Hayette MP, De Mol P. Rapid to the surveys evaluating Belgian practice for prevention intrapartum test (Strep B OIA) and prenatal cultures for of GBS diseases, the change was also well demonstrated, identification of group B streptococcal carriers at delivery: a as at that time, less that 50 per cent of Fl obstetricians prospective study. In: 40 th Intersciences Conference on complied with the screening-based approach6,11. At the Antimicrobial Agents and Chemotherapy (ICAAC), September time of these surveys, most of the obstetricians gave 2000, Toronto, Canada, Abstract book, p. 145, no. 357. amoxycilline for IAP, but at the consensus meeting, more 9. Melin P, Rodriguez Cuns G, Tsobo C, Hayette MP, Christiaens than 90 per cent voted for penicillin G. Both Belgian G, De Mol P. Prevalence of ermB, ermTR and mefA/B gene professional societies of gynecology-obstetrics and the classes among erythromycin-resistant group B streptococcus Ministry of Health reacted quickly. Within a few weeks isolates collected in Belgium. In: 39 th annual meeting of the following the meeting, societies have recommended the Infectious Diseases Society of America (IDSA), October 2001, San Francisco, USA, Abstract book, p.128, no. 523. universal screening based approach to their members and the Ministry of Health appointed a working group 10. Melin P, Keke D, Campo B, Hayette MP, Christiaens G, De for drafting and editing national recommendations for Mol P. Serotype distribution of group B streptococci (GBS) in Belgium : isolates from neonatal infection in adult or colonization the prevention of perinatal GBS diseases. in pregnant woman. In: 42 th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), September References 2002, San Diego, USA, Abstract book, no. 1416. 11. Mahieu L, De Dooy J, Leys E. Obstetricians’ compliance with 1. Zangwill KM, Schuchat A, Wenger JD. Group B streptococcal CDC guidelines on maternal screening and intrapartum disease in the United States, 1990 : report from a multistate prophylaxis for group B streptococcus. J Obstet Gynecol 2000; active surveillance system. MMWR 1992; 41 (SS-6) : 25-32. 20 : 460-4. 2. Schuchat A, Wenger JD . Epidemiology of group B streptococcal 12. Melin P, Epidémiologie et prevention des infections néonatales disease : risk factors, prevention strategies, and vaccine à streptocoques du groupe B en Communauté Française, development. Epidemiol Rev 1994; 16 : 374-402. Recherche en médecine preventive ou en promotion de la santé, 3. Schuchat A. Group B streptococcus. Lancet 1999; 353 : 51-6. Ministère de la Communauté Française de Belgique, Rapport 4. Ducoffre G. Surveillance des maladies infectieuses par un réseau d’activité 1999-2000. de laboratoire de microbiologie 2000 + Tendances 13. Melin P. Streptococcus agalactiae. In : Ducoffre G, Surveillance épidémiologiques 1983 - 1999 , Septicémies et méningites des maladies infectieuses par un réseau de laboratoire de néonatales. 2002, Institut Scientifique de la santé publique, microbiologie 2000 + Tendances épidémiologiques 1983 - 1999, section épidémiologie, Belgique. 2002, Institut Scientifique de la santé publique, section 5. Thesis : Melin P, Recherches sur l’épidémiologie du Streptococcus épidémiologie, Belgique. agalactiae B chez les parturientes et les nouveau-nés. , University of Liege, Belgium; 1987. 14. CDC. Prevention of perinatal group B streptococcal disease: a public health perspective. MMWR 1996; 45 : 1-24. 6. Melin P, Schmitz M, Heinrichs I, Hayette MP, Foidart JM, De Mol P. Prevention of neonatal group B streptococcal disease in 15. CDC. Prevention of perinatal group B streptococcal disease: Belgium : hospital policy, obstetricians practices and laboratory Revised guidelines from CDC. MMWR 2002; 51 : 1-22.

Reprint requests: Dr P. Melin, Belgian Reference Laboratory for Group B Streptococci, Medical Microbiology University Hospital of Liege, B23 Sart Tilman, Liege 4000, Belgium e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 201-204

Distribution of PFGE types of invasive Norwegian group B streptococci in relation to serotypes

N.K. Skjaervold, K. Bergh & Lars Bevanger

Norwegian University of Science & Technology, Department of Laboratory Medicine; Department of Microbiology, University Hospital of Trondheim, N-7006, Trondheim, Norway

Received August 7, 2003

Background & objectives: Pulsed-field gel electrophoresis (PFGE) has been used for characterisation of group B streptococci, non-typeable by serotyping. We wanted to compare PFGE with serotyping in order to see the how well the method discriminates between epidemiological unrelated strains. Methods: A total of 78 epidemiological unrelated invasive GBS strains were examined by PFGE using SmaI digested chromosomal DNA. Of these, 11 were nontypeable (NT) with regard to capsular polysaccharide (CPS) serotype. PFGE patterns were analyzed and classified in a dendrogram. Results: 75 strains were typeable by PFGE, and a total of 62 restriction profiles were identified. At an 85 per cent similarity level, 53 different PFGE patterns were identified. Within each serotype, PFGE patterns differed considerably, the largest degree of heterogeneity observed among type IV, Ia, and II strains. Serotype Ib, III, and V strains were more homogeneous. Strains with identical macrorestriction profiles belonged to the same CPS type, but varied with regard to serosubtypes. Any strain among the ones investigated showing a greater than 88 per cent similarity to a restriction profile in the database, could correctly be ascribed to a particular CPS type. Of the 11 NT strains 10 restriction profiles were found, two of which were identical to the PFGE profile for a cluster of type V strains, and one profile were identical to the profile showed by a cluster of 5 type Ib strains. Interpretation & conclusion: PFGE is a useful technique for classifying strains that are non-typeable by conventional serotyping. Key words Group B streptococci - PFGE - serotypes Pulsed-field gel electrophoresis (PFGE) has become serotypes (CPS) were selected: 12 Ia, 11 Ib, 12 II, 12 a powerful tool for typing of various bacteria, both in III, 8 IV, 12 V, and 11 NT (non-typeable) strains. The relation to outbreaks of infections in hospitals and in the isolates were consecutively received for serotyping from community. PFGE has also been used for characterisation hospitals in Norway during the years 1997-2001. In of group B streptococci, and compared with other typing addition, one reference strain of each of the serotypes methods to examine the discriminatory power when was included (Type Ia, ATCC 12400; type Ib, ATCC applied to strains that were epidemiological unrelated. 12401; type II, NCTC 11079; type III, ATCC 12403; We wanted to examine if the genetic diversity in various type IV, ATCC 49446; type V, ATCC 49447). Serotyping GBS serotypes isolated from invasive infections in was done by FAT with rabbit anti-serum against capsular Norway is comparable to that found by investigators in 5 1-4 polysaccharide types , and monoclonal anti-c protein a other parts of the world . and b and anti-R4 protein antibodies6-8. For PFGE Material & Methods analysis SmaI digests of chromosomal DNA were prepared9. PFGE was performed with a CHEF XA Of the 78 invasive GBS isolates, isolates representing Mapper (BioRad) in 1 per cent agarose gel with a voltage the most commonly found capsular polysaccharide gradient of 6 V/cm at 14°C, a run time of 24 h, and 201 202 INDIAN J MED RES (SUPPL) MAY 2004 switch time of 5-50 s. A lambda DNA ladder was used the 11 non-typeable (NT) strains, 10 restriction profiles as a size standard. Image capturing was performed with were found (Table). Non of the reference type strains the Gel Doc 200 Documentation system (BioRad). showed identical profiles with the clinical isolates. Pulsotypes were compared and classified in a Several clusters with identical macrorestriction profiles dendrogram by using the Jaccard coefficient with 2.5 were identified. Among the type Ib strains a cluster of per cent band tolerance and the unweighted pair group 5 strains-, among the type III strains a cluster of 4 method with arithmetic averages, provides by Molecular strains-, and among the type V strains a cluster of 5 Analyst Fingerprinting DST Software version 1.6 (BioRad). strains were identified. In addition, several pairs with identical profiles were found. Within each cluster or Results pair only strains with the same serotype (CPS type) were found. However, serosubtypes as defined by c 75 of the invasive strains and the 6 reference and R proteins were different for a few of the strains. serotype strains were typeable by PFGE. Two serotype Two of the NT strains showed an identical PFGE profile IV strains, and one serotype III strain were non-typable. Table. Genetic diversity in 75 invasive GBS strains grouped according PFGE patterns in 12 type Ia strains are illustrated in to capsular polysaccharide serotype (CPS) Fig.1. Among the 75 typeable strains, a total of 62 restriction profiles were identified (Fig. 2). At an 85- CPS type No. of strains No. of PFGE patterns and 80 per cent similarity level, 53 and 50 different within each CPS type PFGE patterns, respectively, were identified. PFGE Ia 12 10 patterns within each serotype (CPS) differed Ib 11 6 considerably. The largest degree of heterogeneity was II 12 10 observed in serotype IV, (6 profiles in 6 strains), and in III 11 7 serotype Ia and II strains, both with 10 profiles among IV 6 6 12 strains. Serotype Ib, III, and V were more V12 7 homogenous: 6 profiles in 11 strains, 7 profiles in 11 NT 11 10 strains, and 7 profiles in 12 strains, respectively. Among

Fig.1. PFGE patterns of 12 type Ia strains obtained after SmaI digestion of chromosomal DNA (lanes 1-12). SKJAERVOLD et al: GROUP B STREPTOCOCCI & PFGE 203

Fig. 2. Cluster analyses of PFGE patterns obtained after SmaI-digested chromosomal DNA from 75 invasive GBS strains and reference type strains Ia, Ib, II, III, IV, and V. 204 INDIAN J MED RES (SUPPL) MAY 2004 with the cluster of 5 type V strains, and one strain an 2. Benson JA, Ferrieri P. Rapid pulsed-field gel electrophoresis identical profile with the cluster of 5 type Ib strains. method for group B streptococcus isolates. J Clin Microbiol 2001; 39 : 3006-8. Discussion 3. Elliott JA, Farmer KD, Facklam RR. Sudden increase in isolation of group B streptococci, serotype V is not due to emergence of a new pulsed-field gel electrophoresis type. J Clin Microbiol Among 75 invasive strains, 62 different PFGE profiles 1998; 36 : 2115-6. were identified, confirming the genetic heterogeneity of 4. Rolland K, Marois C, Siquier V, Cattier B, Quentin R. Genetic 1-3 GBS as shown by other investigators . Certain features of Streptococcus agalactiae strains causing severe serotypes displayed a greater degree of homogeneity, neonatal infections, as revealed by pulsed-field gel especially type Ib, type III and type V. Strains with electrophoresis and hylB gene analysis. J Clin Microbiol 1999; identical macrorestriction profiles belonged to the same 37 : 1892-8. CPS type, but varied with regard to serosubtypes. Any 5. Bevanger L, Maeland JA. Type classification of group B strain among the ones investigated showing greater than streptococci by the fluorescent antibody test. Acta Pathol Microbiol Scand Sect B 1997; 85 : 357-62. 88 per cent similarity to a restriction profile in the database, could correctly be ascribed to a particular CPS 6. Naess AI, Bevanger L, Iversen O-J, Maeland JA. Evaluation of monoclonal antibodies in serovar classification of group B type. Some of the macrorestriction profiles of NT strains streptococci (GBS). APMIS 1991; 99 : 1058-60. were identical or similar to profiles displayed by strains 7. Bevanger L, Iversen O-J, Naess AI. Characterization of the a with known serotype, and PFGE is a useful technique antigen of the c protein of group B streptococci (GBS) using a for classifying these strains. murine monoclonal antibody. APMIS 1992; 100 : 57-62. 8. Bevanger L, Kvam AI, Maeland JA. A Streptococcus agalactiae References R protein analysed by polyclonal and monoclonal antibodies. APMIS 1995; 103 : 731-6. 1. Benson KD, Luchansky JB, Elliott JA, Degnan AJ, Willenberg 9. Moyo SR, Maeland JA, Bergh K. Typing of human isolates HJ, Thornbery JM, et al. Pulsed-field fingerprinting of vaginal of Streptococcus agalactiae (group B streptococcus; GBS) group B streptococcus in pregnancy. Obstet Gynecol 2002; strains from Zimbabwe. J Med Microbiol 2002; 51 : 100 : 545-51. 595-600.

Reprint requests: Dr L.S. Bevanger, Consultant/Professor, Department of Medical Microbiology, St. Olavs Hospital, University Hospital of Trondheim, N-7006, Trondheim, Norway e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 205-207

Active surveillance of early onset disease due to group B streptococci in newborns

L. Straková & J. Motlová

National Reference Laboratory for Streptococci & Enterococci, National Institute of Public Health, Prague, Czech Republic

Received August 7, 2003

Background & objectives: Early onset disease (EOD) due to group B streptococci (GBS) poses a serious threat in many countries. In the Czech Republic neither summarized data on the EOD incidence are available nor a nationwide prevention program has been initiated. The present surveillance was initiated to establish the incidence of EOD due to GBS in newborns in the Czech Republic, distribution of GBS serotypes and GBS susceptibility to antimicrobials. Methods: Both invasive and carrier GBS isolates from newborns and the data on the newborns' clinical status and maternal colonization and intrapartum prophylaxis were collected from 30 microbiological and clinical centres all over the Czech Republic within prospective active surveillance. HCl extracts of the GBS strains were precipitated with rabbit polysaccharide (I-VIII) and protein (c,R) antisera. Results: Between January 2001 and September 2002, GBS isolates from 239 full-term and 46 preterm newborns were collected. Of the 285 GBS positive newborns, 105 had invasive EOD, 42 showed suspected EOD, and in 56 clinical diagnosis was not specified. Eighty two GBS isolates were obtained from healthy colonized infants. The isolates obtained from newborns with confirmed invasive EOD were mostly of serotype III (42%), followed by serotypes V a Ia (13% each). Types Ia (26%), III (22%) and II (20%) were most frequent among the isolates from colonized individuals. Protein antigens (c protein, R protein) either coupled with polysaccharide or alone were found in 70 per cent (30 and 40 %, respectively) of the study isolates. Interpretation & conclusion: The incidence of EOD due to GBS found in the Czech Republic 0.7-1.0 per 1000 live births was comparable with the rates reported in the countries where the prevention programme has been implemented nationwide. Serotypes III, V and Ia prevailing among the isolates from Czech newborns with EOD belonged to those most frequently identified in the USA and Western European countries. Key words Active surveillance - Czech - group B streptococci - incidence - infection - neonate - serotype - Streptococcus agalactiae

Early onset disease (EOD) due to group B very limited data are available in some neonatology streptococci (GBS) poses a serious problem in many centres, and in most cases, microbiological details on countries. Most frequent clinical manifestations in EOD the causative agent are not specified. Thus the active are meningitis, sepsis or pneumonia difficult to GBS EOD surveillance was started in January 2001, differentiate from the respiratory distress syndrome. including EOD incidence monitoring, serotyping of the EOD incidence rates in other countries range between strains isolated and their testing for antimicrobial 0.4 and 5.4 per1000 live births1. In the Czech Republic, susceptibility.

205 206 INDIAN J MED RES (SUPPL) MAY 2004 Material & Methods from 56 newborns after a preterm or complicated birth whose diagnosis was not specified, were not included GBS isolates obtained from newborns with EOD and in the study. Of the total of 285 GBS isolates, 64 per colonized newborns together with the filled in forms with cent had both polysaccharide and protein (c or R) necessary data on their diagnosis or clinical condition antigens and 6 per cent had the protein alone. None of and possible maternal GBS carrier status and intrapartum the known antigens was found in 2 per cent of the GBS prophylaxis were referred to the National Reference isolates. Two deaths were reported during the study. Laboratory for Streptococci and Enterococci (NRL), Ib/c and c protein alone were demonstrated in the post National Institute of Public Health, Prague from 30 mortem isolates from these two newborns. The isolates hospital microbiological centres all over the Czech Republic. If available, maternal GBS isolates were sent together with those from newborns. The isolates confirmed as GBS were typed by precipitation reaction in gel with the use of HCl extracts prepared according to Lancefield's modification2 and rabbit antisera raised against polysaccharide (I-VIII) and protein (c,R) antigens. The absence of both polysaccharide and protein antigens was tested by the precipitation reaction in capillary tubes.

Results

Between January 2001 and September 2002, GBS isolates from 239 full-term and 46 preterm newborns from all over the Czech Republic were referred to the NRL. Of the 285 GBS positive newborns, 105 had invasive EOD (52 cases of sepsis, 7 of meningitis, 17 of respiratory disorders, in 29 cases infection was not Fig. 1. GBS serotypes from 105 neonates with proved invasive specified), in 42 infection was suspected, and in 56 with infection, Czech Republic, 2001(Jan) - 2002(Sep). hypoxia resulting from a preterm or complicated birth, clinical diagnosis was not specified. None of 82 GBS colonized infants was reported to develop EOD. As many as 170 isolates from the newborns were sent together with the maternal isolates from the vagina or lochia; the same serotype was always identified in the matched isolates. Mothers of 15 GBS positive newborns showed negativity in prenatal screening. Data on possible maternal colonization were not available for the remaining 100 cases. Among the strains isolated from 105 newborns with proved GBS EOD, serotype III prevailed widely followed by serotypes V and Ia (Fig.1). The same serotypes were also the most frequent serotypes to be identified among 42 isolates obtained from newborns with suspected infection. Serotypes Ia, III and II prevailed among isolates from Fig. 2. GBS serotypes from 82 colonized neonates, Czech Republic, 82 colonized individuals (Fig. 2). Serotypes of isolates 2001(Jan) - 2002(Sep). STRAKOVA & MOTLOVA : EARLY ONSET DISEASE DUE TO GBS IN NEWBORNS 207 from cerebrospinal fluid of 7 newborns with meningitis Acknowledgment were of serotypes III and V. The financial support by the Internal Grant Agency of the Discussion Ministry of Health of the Czech Republic is acknowledged. Authors thank all staff of microbiological and clinical centers involved in the active EOD surveillance for help in collecting strains. To obtain more accurate and extensive data, the GBS EOD surveillance has been continued and the References data reported are being verified. Although any nationwide EOD prevention programme has not been 1. De la Rosa Fraile M, Cabero L, Andreu A, Rao GG. Prevention implemented in the Czech Republic, the preliminary of group B streptococcal neonatal disease: a plea for a European estimate of the EOD incidence rate based on the consensus. Clin Microbiol Infect 2001; 7 : 25-7. present data, i.e., 0.96 per 1000 live births, was 2. Lancefield CR. Serological differentiation of specific types of comparable with the rates reported in the countries bovine haemolytic streptococci (group B). J Exp Med 1934; 59 : 441-58. where such programmes have been implemented1. In the Czech Republic, not all of the participating centres 3. Prevention of perinatal group B streptococcal disease. Revised guidelines from CDC. MMWR Recomm Rep 2002; 51 : 1-22. carried out prenatal screening and not all of them followed the CDC recommendations3. The culture 4. Radhakrishnan S, Brahmadathan KN, Mathai E, Jesudason MV, Lalitha MK. Changing patterns of group B streptococcal procedure optimizing the GBS detection was not used serotypes associated with human infections. Indian J Med Res for screening negative mothers of the GBS positive 1995; 102 : 56-9. newborns either. 5. Davies HD, Raj S, Adair C, Robinson J, McGeer A. Population- based active surveillance for neonatal group B streptococcal The serotypes prevailing in Czech newborns with infections in Alberta, Canada: implications for vaccine formulation. Pediatr Infect Dis J 2001; 20 : 879-84. invasive EOD were among those most frequently identified from such isolates in India and North 6. Kalliola S, Vuopio-Varkila J, Takala AK, Eskola J. Neonatal group B streptococcal disease in Finland: a ten-year nationwide 4,5 6,7 America and Western European countries . Type study. Pediatr Infect Dis J 1999; 18 : 806-10. VIII isolates frequently found in Japan were not 7. Berg S, Trollfors B, Lagergard T, Zackrisson G, Claesson BA. identified among the present GBS study isolates. Serotypes and clinical manifestations of group B streptococcal Prevailing serotypes as found among the GBS study infections in western Sweden. Clin Microbiol Infect 2000; 6 : isolates from the Czech newborns were consistent 9-13. with those identified from epidemiologically unlinked 8. Motlová J, Straková L, UrbáskovሠP, Sak P, Sever T. Vaginal and isolates from colonized pregnant Czech women8. In rectal carriage of Streptococcus agalactiae in the Czech Republic : incidence, serotypes distribution & susceptibility to antibiotics. the light of frequent antimicrobial resistance of Indian J Med Res 2004; 119 (Suppl) : 84-7. serotype V9, the relatively high proportion serotype 9. Blumberg HM, Stephens DS, Modansky M, et al. Invasive V among our isolates, including those isolated from group B streptococcal disease: the emergence of serotype CSF, is alarming. V. J Infect Dis 1996; 173 : 365-73.

Reprint requests: Dr J. Motlova, National Reference Laboratory for Streptococci & Enterococci, National Institute of Public Health Srobarova 48, Prague 10042, Czech Republic e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 208-212

Characterization of vaginal & rectal colonization with multiple serotypes of group B streptococci using multiple colony picks

P. Ferrieri, S.L. Hillier*, M.A. Krohn*, D. Moore*, L.C. Paoletti** & A.E. Flores

University of Minnesota Medical School, Minneapolis, MN, *University of Pittsburgh, School of Medicine Pittsburgh, PA & **Harvard Medical School, Boston, MA, USA

Received August 6, 2003

Background & objectives: There is paucity of information on vaginal and rectal colonization with multiple serotypes of group B streptococci (GBS). As part of an ongoing cohort study evaluating the natural history of vaginal and rectal colonization by GBS, the colonization with multiple serotypes was studied in 102 non-pregnant women aged 18-30 yr. Methods: Up to ten separate colony picks of beta-haemolytic streptococci (total 1515 isolates) were selected from vaginal and rectal primary culture plates. The colonies were identified as GBS, and their capsular polysaccharides (CPS) serotypes determined using monospecific rabbit antisera for types Ia-VIII by double immunodiffusion in agarose (DID). A colony dot immunoblot (DB) assay, using monospecific rabbit antisera to purified type polysaccharides conjugated to tetanus toxoid, was developed to serotype efficiently the multiple colony picks of GBS. Results: The CPS serotype distribution, examining only the 177 “a” or first colony picks from the 102 patients, was 30.5 per cent for Ia; 28.2 per cent for type III; 15.3 per cent for type II; and 13.6 per cent for type V. Only 2.8 per cent were nontypeable. Eighty of the 102 patients (78.4%) were colonized with only one serotype; 20 (19.6%) had two serotypes and two patients (2%) had three serotypes in their vaginal and/or rectal paired cultures. Overall, 91.9 per cent of the culture sites colonized with one to three CPS types (from the total number of colonies picked) were identified with a minimum of three colony picks. In 75 patients with vaginal/rectal pairs the GBS serotype concordance of only the “a” colony was 89.3 per cent and concordance decreased to 80 per cent when the serotype concordance of the total colony picks was analyzed. Interpretation & conclusion: In conclusion, there was a relatively high prevalence of serotype non- concordance in this population, and 21.6 per cent of patients had multiple GBS serotypes. Key words GBS colonization - multiple colony picks - serotype concordance

There are many unanswered questions regarding the picked from each culture, analysis of multiple serotypes epidemiology of vaginal and rectal colonization with isolated simultaneously from the same culture sites in group B streptococci (GBS) in women1,2. Among these patients was not done3. In epidemiological studies of an is the issue of whether women may be colonized with Australian aboriginal community, multiple Vir types of more than one GBS serotype when assessed at a specific GAS were isolated at the same time from pyoderma time. It has been documented that multiple serotypes of lesions in children4. However, vaginal and/or rectal group A streptococci (GAS) may simultaneously spread colonization with multiple serotypes of GBS has been within a family demonstrating positive skin or skin lesion rarely documented in women. The present study was cultures3. Although in that study three GAS colonies were designed to test the hypothesis that nonpregnant women

208 FERRIERI et al: VAGINAL & RECTAL COLONIZATION WITH GROUP B STREPTOCOCCI 209 may be colonized with more than one serotype were picked to Todd-Hewitt broth (THB) and after simultaneously. overnight incubation, were frozen at -30°C until they were processed for typing. Material & Methods Serotyping: Study population and specimen collection: The 102 Immunoprecipitation in agarose by double nonpregnant women (18-30 yr) were part of a cohort immunodiffusion (DID) – The “a” colony from each study evaluating the natural history of vaginal (V) and patient was grown in THB overnight, and the cell pellet 5 rectal (R) colonization by GBS . They were seen at was extracted with hot HCl6. These extracts were tested a primary health care center (n=20), a student health by immunoprecipitation in agarose (DID) with clinic (n=19) and a sexually transmitted disease clinic monospecific rabbit antisera to GBS polysaccharide (n=63). All were unmarried, 92 per cent currently serotypes Ia, Ib-VIII, the alpha and beta components of sexually active, and 46 per cent nonwhite. Separate the c protein, as well as the R1, R3, R4 and R5 group B sterile Dacron-tipped swabs were used to sample the streptococcal protective surface antigen (BPS) species vagina and rectum. Each swab was placed in a of R antigen7,8. separate Amies transport tube and transported to the laboratory within 12 h. Dot blot (DB) – A small aliquot from the THB cultures of all “a-j” isolates was spotted on a polyvinylidene Recovery and identification of GBS: The vaginal fluoride (PVDF) membrane (ImmobilonTM, Millipore, swab was inoculated to a Columbia agar plate with 5 Bedford, MA) marked with a grid pattern9. Spotted on per cent sheep blood and selective broth; the rectal each membrane also were appropriate positive and swab, to a colistin-nalidixic acid agar (CNA) plate and selective broth. Plates and broths (Prepared Media negative controls, including uninoculated THB. The Laboratories, Mississauga, ON) were incubated for membrane was washed, blocked and blotted for capsular polysaccharide type (CPS) only, using monospecific 18-24 h at 35°C in 5-7 per cent CO2. Plates were examined visually for beta haemolytic colonies; broths rabbit antisera to tetanus toxoid-conjugated purified GBS were subcultured onto 5 per cent sheep blood agar capsular polysaccharides (Channing Laboratory, Boston). plates (BAP). For each individual patient, the membrane with its “a-j” samples was blotted with antiserum to the serotype of To obtain a broad sample of GBS serotypes its own “a” pick as determined by DID. The membrane present, 10 separate colony picks (labeled “a-j”) were was then incubated with alkaline phosphatase-conjugated picked for identification from each V and R culture. F(ab´)2 goat anti-rabbit IgG, F(ab´)2 fragment (Jackson Colonies were selected from the original plate, but if ImmunoResearch Laboratories, West Grove, PA) and it had < 10 cfu (colony forming units) per plate, the developed with a nitroblue tetrazolium-3-bromo-4- remaining picks were made from the subculture plate chloroindolyl phosphate (NBT-BCIP) substrate. Results of the selective broth medium. Some V and/or R were read visually on a scale of 1+ to 4+, with readings culture sites were colonized very lightly with only one of ³ 2+ considered positive. or two cfu/ml of GBS. The DB result of the “a” isolate was read for Isolates were identified as GBS by agglutination technique using the PathoDx™ typing system concordance with the serotype as determined by DID; (Diagnostic Products Corporation, Los Angeles, CA). results of the “b-j” isolates, for concordance with the They were stored in litmus milk at -70°C until sent by CPS type of the “a”. The “b-j” isolates not blotted with overnight courier on blood agar slants (Gibson antiserum to the same serotype as the “a”, or giving Laboratories Inc., Lexington, KY) to the GBS Molecular weak reactions (£ 1+) were tested with antisera to the Reference Laboratory at the University of Minnesota. heterologous types. Those still not blotted by any antiserum or with weak reactions, were tested by Initial processing: All isolates (colony picks “a-j”) were immunoprecipitation in agarose by double subcultured on BAP to assess purity. Individual colonies immunodiffusion. 210 INDIAN J MED RES (SUPPL) MAY 2004 Total "a" colony picks Total "a-j" colony picks N = 177 N = 1515 Ia Ia 30.5% NT 34.0% NT 3.1% 2.8% Ib NT 9.0% 14.3% V 13.6% Ib 7.8% IV IV II 0.7% 0.6% 15.3% II III 15.4% III 24.7% 28.2%

Fig. Comparison of GBS serotype prevalence of total “a” vs. total “a-j” isolates. NT, nontypeable.

Results Table I. Serotype concordance of “a” vs. “a-j” colony picks in vaginal/ rectal (V/R) pairs* The pie diagrams (Fig.) depicted the distribution of serotypes for 177 “a” colony picks, the first colony or No. of V/R pairs (%) predominant colony pick, and the 1515 “a-j” colony picks. “a” colony “a-j” colonies In this comparison, the predominant serotypes for either set of data were Ia, III, and II. The fourth most prevalent Concordant** 67 (89.3) 60 (80) CPS type was V. The nontypeable isolates were very Non-concordant 8 (10.7) 15 (20) few. The cell-surface protein expression profiles of the N = 75 177 “a” colony picks were: alpha only, 40.7 per cent; alpha plus beta, 10.2 per cent; R1 only, 1.1 per cent; R4 only, 30.5 per cent; R1 plus R4, 13.6 per cent; and 4 per * 27 unpaired culture sites (19 rectal and 8 vaginal) were excluded cent of the isolates had none of these commonly studied from the analyses proteins (P Ferrieri and AE Flores, unpublished data). ** All the same serotype or same combination of serotypes

Analysis of the 1515 GBS isolates from 102 patients Table II. Distribution of patients with one or more serotypes in all revealed that 86 per cent of the vaginal and/or rectal “a-j” picks from vaginal and/or rectal culture sites sites sampled had six or more colony picks and 69.5 per cent of culture sites had ten colony picks of GBS. Of No. of serotypes Culture category No. (%) patients the culture sites sampled, there were 14.1 per cent with 1 with V/R paired cultures 57 four or fewer colony picks. The GBS serotype 1 R only 17 (78.4) concordance in 75 women who had vaginal and rectal } paired cultures obtained at the same visit was analysed. 1 V only 6 In analyzing these data, if only the “a” colony was 2 with V/R paired cultures 16 studied, the serotype concordance between the two 2 R only 2 (19.6) culture sites in each woman was 89.3 per cent. } 2 V only 2 However, when multiple GBS colonies were picked and serotyped, this concordance decreased to 80 per cent 3 with V/R paired cultures 2 (2.0) (Table I). N = 102 FERRIERI et al: VAGINAL & RECTAL COLONIZATION WITH GROUP B STREPTOCOCCI 211

Table III. Minimum number of colony picks required to detect all with evaluation of up to ten different GBS colonies per serotypes in a culture site culture site serotype concordance between the vaginal No. of serotypes detected by No. (%) of culture and rectal paired cultures was only 80 per cent. There designated number of colony sites with no. of were only minimal differences in the distribution of the picks serotypes identified various serotypes when comparing the single colony No. of “a” picks versus the “a-j” multiple colony picks. Further, serotypes found No. picks required* the predominant types were the same, and, of note, 1** 1 147 (85.0) was the predominance of serotype Ia followed by 228 (4.6) 91.9% serotype III. Further support for the prominent role of 233 (1.7) serotype Ia was that it was also the commonest 331 (0.6) serotype found in combination with one or more other } 241(0.6) serotypes. 252 (1.2) 261 (0.6) The finding of simultaneous colonization of women 273 (1.7) with multiple GBS serotypes has been reported rarely. 294 (2.3) In one report of pregnant women cultured at multiple 2103 (1.7) sites during pregnancy and/or delivery, 23 per cent of N=173† the GBS positive women had more than one serotype at the same time10. Among the 102 non pregnant women * To detect all serotypes/culture site ** Based on serotypes of b - j, “a” colony was concordant in the present study, multiple serotypes were detected in † Does not include four culture sites with only one isolate 21.6 per cent. It is curious that co-colonization with more than one GBS serotype was not uncommon in this Of great interest was the frequency of colonization sexually active group of women. The majority, 61.8 per with more than one GBS serotype among the total 102 cent of this cohort of relatively young, non pregnant patients. As seen in Table II, the commonest distribution women were seen at a sexually transmitted disease clinic. pattern was colonization with only one serotype, seen in 78.4 per cent. However, 20 patients (19.6%) were It is possible that the acquisition and colonization with colonized with two serotypes and two patients were multiple GBS serotypes may be related to the number of colonized with three serotypes. sexual partners. It has been demonstrated that frequent sexual intercourse and/or multiple sex partners in this In the 22 patients colonized with more than one serotype, group of women was associated with vaginal acquisition 5 there were ten different serotype combinations recovered of a new serotype of GBS . (data not shown). The commonest serotype combinations were Ia/Ib, Ia/III, Ia/NT, and III/V. The two patients It is not practical in routine microbiology laboratories colonized with three serotypes had Ia/II/III or II/III/V. to pick more than one GBS colony to pursue Serotype Ia was seen in 14 of the 22 (63.6%) patients. identification and antimicrobial susceptibility testing. Although in special epidemiological studies there may Of interest was the total number of colony picks that be a role for studying more than one bacterial colony were required to detect the total number of GBS based on morphology, degree of beta haemolysis or no serotypes in a culture site. As seen in Table III, haemolysis, or colony size, it is impractical to recommend cumulatively, 91.3 per cent of culture sites with one to selecting as many colonies as were done in the current two serotypes or 91.9 per cent of culture sites with up to study. Noteworthy was the finding that a minimum of three different serotypes were identified with a minimum three colony picks identified one to three GBS serotypes of three colony picks. in 91.9 per cent of the vaginal or rectal culture sites sampled. Discussion Detection of vaginal/rectal colonization with multiple In this study on GBS colonization among non- GBS serotypes may have relevance to the design of pregnant women, it was astounding to discover that clinical trials of specific GBS CPS vaccines and in 212 INDIAN J MED RES (SUPPL) MAY 2004 evaluating their efficacy in preventing homologous DR, Tagg JR, editors. Streptococci and streptococcal diseases: serotype colonization and infection in women and entering the new millenium. Proceedings of the XIV Lancefield infants2,8. In summary, the findings of the present study International Symposium on Streptococci and Streptococcal Diseases; 1999 October 11-15; Auckland, New Zealand. New demonstrated that colonization with more than one GBS Zealand: Securacopy; 2000 p. 155-8. serotype is not uncommon in women. 5. Meyn LA, Moore DM, Hillier SL, Krohn MA. Association of sexual activity with colonization and vaginal acquisition of group Acknowledgment B Streptococcus in nonpregnant women. Am J Epidemiol 2002; 155 : 949-57. The authors acknowledge the financial support from the National 6. Johnson DR, Ferrieri P. Group B streptococcal Ib/c protein Institutes of Health (Institute of Allergy and Infectious Diseases), antigen: distribution of two determinants in wild-type strains USA, Research Contract N01 AI-75326. of common serotypes. J Clin Microbiol 1984; 19 : 505-10. 7. Flores AE, Ferrieri P. Molecular diversity among the trypsin References resistant surface proteins of group B streptococci. Zbl Bakt 1996; 285 : 44-51. 1. Ferrieri P. Prevention of group B Streptococcus infections. Semin 8. Zaleznik DF, Rench MA, Hillier S, Krohn MA, Platt R, Lee M- Infect Dis 1997; 8 : 117-24. LT, et al. Invasive disease due to group B Streptococcus in 2. Hickman M, Rench MA, Ferrieri P, Baker CJ. Changing pregnant women and neonates from diverse population groups. epidemiology of group B streptococcal colonization. Pediatrics Clin Infect Dis 2000; 30 : 276-81. 1999; 104 : 203-8. 9. Ferrieri P, Flores AE, Concepcion NF, Anthony BF. Bacterial 3. Ferrieri P, Dajani AS, Wannamaker LW, Chapman SS. Natural surface expression and recognition of the IgA binding protein of history of impetigo. I. Site sequence of acquisition and familial group B streptococci. In: Orefici G, editor. New perspectives patterns of spread of cutaneous streptococci. J Clin Invest 1972; on streptococci and streptococcal infections. Zbl Bakt Suppl 1992; 51 : 2851-62. 22 : 186-8. 4. Carapetis JR, Currie BJ, Hibble M, Sriprakash KS, Bessen DE, 10. Jacomina A, Hoogkamp-Korstanje A, Gerards LJ, Cats BP. Mathews JD. Rapid turnover of multiple strains of group A Maternal carriage and neonatal acquisition of group B Streptococcus in an Australian aboriginal community. In: Martin streptococci. J Infect Dis 1982; 145 : 800-3.

Reprint requests: Dr Patricia Ferrieri, MMC 134, 420 Delaware St. SE, University of Minnesota Medical School, Minneapolis, MN 55455, USA e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 228-232

Construction of recombinant polypeptides based on beta antigen C (Bac) protein & their usage for protection against group B streptococcal infection

A. Suvorov, I. Ustinovitch, L. Meringova, K. Grabovskaya, G. Leontieva, E. Vorobieva & A. Totolian

Institute of Experimental Medicine, Academy of Medical Sciences, Pavlova 12, St. Petersburg 197376, Russia

Received August 8, 2003

Background & objectives: Group B Streptococcus (GBS) is an important human pathogen causing serious diseases of newborn and immunocompromized adults. Approximately 50 per cent of the GBS strains carry and express the gene of BAC antigen which is capable to bind IgA. Gene encoding for the BAC antigen has been cloned and sequenced but actual IgA binding region on the protein has not been detected. The aim of the present work was to localize the region of IgA binding on Bac protein, to evaluate the role of one of the Bac protein regions MLKKIE in IgA binding, and to investigate the ability of Bac based recombinant proteins to generate protective antibodies against GBS infection. Methods: Recombinant proteins based on beta antigen C were generated after PCR amplification of the fractions of bac gene with the following cloning of the PCR products into expression plasmids. Recombinant peptides were tested for IgA binding by immunoprecipitation and Western blot. One of the recombinant proteins expressing IgA binding was used as an antigen for immunization of mice and for GBS protection studies. Results: Several bac gene constructs were generated. Their ability to bind IgA varied dramatically depending on the size of the construct and location of the fragment on the bac gene map. The smallest peptide expressing IgA binding was 14 kD in size. Amino acid substitutions in MLKKIE region facilitated IgA binding ability. Immunization of mice with recombinant Bac based peptide induced the appearance of anti-GBS antibody with high affinity level providing protection against GBS infection. Interpretation & conclusion: Size dependence of Bac based recombinant peptides proved that the effective IgA binding required specific folding of the protein binding IgA. Region MLKKIE could not be considered as region, responsible for IgA binding. Generation of antibodies against Bac based recombinant peptides with high titre and affinity makes these proteins a potent candidates for generating a vaccine against GBS infection.

Key words Group B streptococcus - IgA binding - immunization - streptococcus - vaccine Group B Streptococcus (GBS) is the major cause of serotype III. Serotypes II and Ib are more commonly newborn mortality in USA and European countries. In found in adults. An increase was seen in GBS diseases spite of prophylactic procedures introduced in most among the adults, including the severe cases such as developed countries, newborn sepsis, pneumonia and necrotizing fasciitis1-3. One of the GBS protein virulence meningitis continues to be a major medical and factors is beta antigen C (Bac antigen)4. The bac gene epidemiological problem. Most of the severe newborn is present in about 50 per cent of GBS strains and is infections are caused by the GBS strains with capsule usually expressed in serotypes II and Ib5. Gene encoding 228 SUVOROV et al: BAC BASED RECOMBINANT PROTEINS & PROTECTION AGAINST GBS 229 for Bac protein has been cloned and sequenced6. An of p6 peptide 0.3, 0.6 or 1.2 mg/kg followed by second amino acid region MLKKIE on the molecule of Bac injection of 0.15, 0.3, or 0.6 mg/kg 30 days after the first protein was proposed to be responsible for IgA binding7. injection. On day 60 of the experiment mice were Based on previous findings the present study was carried infected with two different concentrations of GBS strain out to verify the IgA binding region on the Bac protein SU219 IIbc (3.5x107 and 3.5x106 cfu) by intraperitoneum by generating recombinant peptides and to investigate injection. Titre of antibodies was determined by ELISA9. the functional role of IgA binding peptides as potential Level of affinity of the antibodies to the antigen was vaccines against GBS infection. estimated according the computer programme Policonst10. Material & Methods Results Bacteria and plasmids: GBS strain SU219 IIbc was obtained from collection of strains in the Department of Several recombinant plasmids, containing the DNA Molecular Microbiology, Institute Experimental Medicine. fragments of bac gene were constructed. All these This strain was used as source of bac gene for DNA constructs contained authentic or modified MLKKIE amplification and cloning. The same strain was used for region (Table II). For modification of this region EcoRI animal studies. restriction site was introduced into two primers R2 and F7 (Table I). This modification resulted in substitution of DNA manipulations: Routine DNA manipulations such the region MLKKIE with region MLKRIQ in constructs as DNA isolation, digestion, ligation and transformation P2, P12 and P5. All the constructs were tested for were performed according to Maniatis et al8. DNA primers used for amplification and cloning of the bac Table I. DNA primers used for cloning of bac gene fragments gene fragments are shown in Table I. PCR fragments were cloned employing the plasmid vector pGemTEasy Primers DNA sequence of the primer 5-3’ Location on bac gene (Promega, USA). Plasmid vectors pQE30,31,32 F1 CGCAGAAAGCTTATCTAAT 688-708 (Qiagen, USA) were used for obtaining of the R2 CTTGAATTCTTTTCAGCATAG 791-812 recombinant proteins in all possible open reading frames. R3 CATAGCTTGAATTTCTTGA 914-934 F4 ACCTGGTGTTTGACCTGAACT 249-272 Protein studies: Recombinant proteins were purified by F5 GGAGGATATCGATATG 116-132 affinity chromatography employing Ni-NTA agarose R6 CGAAGACGACGTTGGA 1516-1532 columns (Qiagen, USA). Purified proteins were analyzed F7 CTATGCTGAAAAGAATTCAAG 791-812 by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and tested for IgA binding Table II. Construction and characteristics of recombinant peptides by immunoprecipitation or by Western-blot followed by based on GBS bac gene sequence detection by alkaline phosphatase labeled IgA (Sigma, Primers used for Plasmid and IgA MLKKIE Size of USA). cloning Size of the binding motive the (forward/ insert (bp) activity peptide Synthetic oligopeptide ITNEDKDSMLKKIEDIRKQA reverse) (kD) was kindly provided by Dr G.Vlasov, Institute of F5/R6 p4.2(1500) ++++ MLKKIE 35 Bioorganic Chemistry, Russian Academy of Sciences, F4/R6 P6(840) +++ MLKKIE 30 St.Petersburg. F1/R3 P10(246) + MLKKIE 14 Immunization and GBS protection studies: F4/R2 P2(562) _ MLKRIQ 20 Recombinant protein p6, carrying IgA binding domain F4/R3 P1(685) + MLKKIE 21 was used for immunization of mice. Animals used for F4/R2 + F7/R3 P5(685) ++ MLKRIQ 22 immunization were unbred male mice (16-18 g). Mice were infected subcutaneously with three different doses F1/R2 P12(123) _ MLKRIQ 7 230 INDIAN J MED RES (SUPPL) MAY 2004 expression of recombinant proteins by electrophoresis The peptide P6 used for immunization being and for IgA binding by ELISA and immunoprecipitation administrated in three doses gave the maximum rise (Fig.1). The intensity of IgA binding correlated with the of antibodies on day 48 from the first injection size of the peptide as well as with its location regarding (Fig.2 ) - 1:25000 for the smallest dosage and 1:100000 region MLKKIE. The highest degree of IgA binding was for the second and third dosages. Affinity of the anti- demonstrated employing the products of plasmids p4.2 Bac antibodies was the highest (constant of 13 and p6 (35 and 30 kD). Smallest recombinant peptide dissociation was equal to10 perM) for the smallest P12 (7kD) did not bind IgA at all. However, the size dosage of the antigen and minimal for the highest 8 was not the only limitation. Relatively large peptide P2 antigen dose (10 per M). The second dose of antigen (0.6 mg/kg on the first injection) gave rise to the (20kD) with modified MLKKIE region at the C end also antibodies with intermediate level of affinity did not bind IgA and peptide P1 (21 kD) with authentic (109 per M). This dosage of the antigen was MLKKIE region was able to bind IgA poorly. Synthetic considered the best dose for mice protection studies. oligopeptide, containing MLKKIE region did not bind IgA in any experimental set up. Interestingly, the peptide P5 GBS from strain SU219 being administrated in two which differs from peptide P1 only by two amino acids different concentrations 107 and 106 cfu was highly in MLKKIE region expressed relatively high level of virulent for the mice causing significant level of lethality IgA binding both in ELISA and immunoprecipitation during two first days of infection (Table III). Immunized (Fig.1).The smallest peptide, capable to bind IgA was animals demonstrated a significant level of protection P10 (14kD), but the level of binding was low and highly which was more distinct when lower bacterial depended on temperature and buffer conditions. concentration (106 cfu) was used.

Discussion

Bacterial proteins capable to bind human immunoglobulins are fairly common for pathogenic bacteria. Most of these proteins such as IgG binding staphylococcal protein A or streptococcal protein G share structural organization and these proteins are often homologous on DNA or amino acid level 11. Bac protein from GBS does not have any homologues Fig.1. IgA binding by different concentrations of recombinant among bacterial proteins discovered so far. The only peptides – derivatives of Bac protein. Peptides were administrated homologoue to Bac protein found so far was among in 3 ml drops in two times dilutions on nylon filter paper. Starting the proteins belonging to the superfamily of eukaryotic concentration was 1 mg/ml for all the peptides. Later on the filter immunoglobulins12. This protein is capable to bind Fc was incubated with myeloma IgA and treated with alkaline phosphatase labeled anti-IgA IgG. portion of IgA which makes bacteria invisible for the host immune system. This fact seems to be vitally

Table III. Protective effect of Bac antigen against GBS infection (cfu/ml in spleen) Immunized mice (cfu) Control group (cfu) Time after GBS infection s3.5x107 3.5x106 3.5x107 3.5x106 6 h 1 x102 0 6.2 x106 1.7 x105

1 day 3.5 x103 1 x102 >108 7 x105

Fig.2. Immune response after immunization with recombinant Bac 2 day 1.4 x103 0 >108 7 x102 peptide P6. SUVOROV et al: BAC BASED RECOMBINANT PROTEINS & PROTECTION AGAINST GBS 231 important for GBS colonizing rectum and vagina References where IgA represents about 90 per cent of total immunoglobulins, and provides defense against 1. Trivalle C, Martin E, Martel P, Jacque B, Menard JF, infection due to specific recognition of the antigens Lemeland JF. Group B streptococcal bacteraemia in the elderly. and due to recruitment macrophages to the site of J Med Microbiol 1998; 47 : 649-52. infection and binding to Fc, RI and CD89 receptors. 2. Gardam MA, Low DE, Saginur R, Miller MA. Group B Recent studies on the Ca2 (LLG) and Ca3 (PLAF) streptococcal necrotizing fasciitis and streptococcal toxic shock- like syndrome in adults. Arch Intern Med 1998; 158 :1704-8. loops on the IgA molecule demonstrated that Bac was capable to bind with both regions interfering with 3. Domingo P, Barquet N, Alvarez M, Coll P, Nava J, Garau J. 13 Group B streptococcal meningitis in adults: report of twelve the ability to bind CD89 . Our results corroborated cases and review. Clin Infect Dis 1998; 27 : 411-2. with this finding. The failure to limit the size of IgA binding region to couple amino acids can be explained 4. Russell-Jones GJ, Gotsclich EC, Blake MS. A surface receptor specific for human IgA on group B streptococci possessing the by the necessity to bind both loops on the IgA Ibc protein antigen. J Exp Med 1984; 160 : 1467-75. molecule. The requirement of the proper three- 5. Schalen C. Prevalence of IgA receptors in clinical isolates of dimentional folding of the IgA binding protein explains Streptococcus pyogenes and Streptococcus agalactiae: Serologic why the smallest IgA binding peptide P10 was not distinction between the receptors by blocking antibodies. FEMS able to provide efficient and stable binding of the IgA. Immunol Med Microbiol 1993; 7 : 39-45. Analysis of the region MLKKIE as potential IgA 6. Jerlstrom PG, Chatwal GS, Timmis KN. The Iga binding b binding site clearly demonstrated that this region alone antigen of the c protein complex of group B streptococci: or even together with couple of amino acids on both sequence determination of its gene and detection of two binding ends was not sufficient for IgA binding. Experiments regions. Mol Microbiol 1991; 5 : 843-9. with introducing the amino acid substitutions in 7. Jerlstrom PG, Talay SR, Valentin-Weigand P, Timmis KN, MLKKIE region which improved the IgA binding level Chatwal GS. Identification of an immunoglobulin A binding motif located in the b-antigen of the c protein complex of group can be explained by the modulatory function of this B streptococci. Infect Immun 1996; 64 : 2787-93. motif in the process of binding IgA. 8. Maniatis T, Fritsch EF, Sambrook J. Molecular cloning: a laboratory mannual. Cold Spring Harbor: Cold Spring Harbor Another set of experiments was related to the analysis Laboratory Press; 1982. of immune response to the Bac based peptides. 9. Friquet B, Charfotte AF, Djavadi-Ohaniance LJ. Measurement Considering that Bac antigen is not expressed by all GBS of the true affinity constant in solution of antigen-antibody strains we planned to use this peptide as a component of complexes by enzyme-linked immunosorbent assay. Immunol the conjugative peptide vaccine, which consists of several Methods 1985; 77 : 305-9. GBS immunogenic and protective peptides. 10. Agafonova II. Characteristic features of description of Administration of different doses of Bac peptide to experimental data in analysis of affinity distribution of serum laboratory animals demonstrated that recombinant peptide IgG specific to influenza virus. Vestnik AMN USSR 1991; 11 : used was highly immunogenic which was comparable 11-4. to the effect of other streptococcal proteins candidates 11. Sternberg L, O’Toole P, Lindahl G. Many group A streptococcal for vaccine production14-16. This fact together with a good strains express two different immunoglobulin-binding proteins, encoded by closely linked genes: Characterization of the proteins protection against GBS infection makes it possible to expressed by four strains of different M-type. Mol Microbiol consider Bac protein based peptides as candidates for 1992; 6 : 1185-94. generating an efficient conjugative peptide vaccine against 12. Bateman A, Eddy SR, Chorthyia C. Members of the GBS infection. immunoglobulin superfamily in bacteria. Protein Science 1996; 5 : 1939-41. Aknowledgment 13. Pleass RJ, Areschoug T, Lindahl G, Woof JM. Streptococcal IgA-binding proteins bind in the Ca2-Ca3 region and inhibit Authors acknowledge Dr G.Vlasov, Institute of Bioorganic binding of IgA to human CD89. J Biol Chem 2001; 276 : Chemistry, Russian Academy of Sciences for providing synthetic 8197-204. peptide and Dr Ferretti, University of Oklahoma for assistance in 14. Cleary PA, Stafslein D, Carlson B. A streptococcal C5a DNA sequencing of the genetic constructs. The work was supported peptidase vaccine induces protection to intranasal challenge with by RFFI grant 00-04-49360a. heterologous serotypes of group A streptococcus. Proceedings 232 INDIAN J MED RES (SUPPL) MAY 2004

of the XIV Lancefield International Symposium on Streptococci responses against Streptococcus pyogenes challenge in mice. and Streptococcal Diseases, October 1999, Auckland, New Infect Immun 2001; 69 : 924-30. Zealand, Martin, DR, editor. J.R.Tagg-p.437-40. 16. Martin D, Rioux S, Gagnon E, Boyer M, Hamel J, Charland N, 15. Kawabata S, Kunimoto E, Terao J, Nakagawa I, Kikuchi K, et al. Protection from group B streptococcal infection in neonatal Totsuka K, et al. Systemic and mucosal immunizations with mice by maternal immunization with recombinant Sip protein. fibronectin-binding protein FBP54 induce protective immune Infect Immun 2002; 70: 4897-1.

Reprint requests: Dr A. Suvorov, Institute of Experimental Medicine, Academy of Medical Sciences, Pavlova 12, St. Petersburg 197376, Russia e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 237-241

Influence of decreased penicillin susceptibility on growth rate

of beta haemolytic streptococci

´ ˆ Nataša Opavski, Slobodanka Dukic, Vera Mijac & Lazar Ranin

Institute of Microbiology & Immunology, School of Medicine, University of Belgrade Dr Subotica 1, 11000 Belgrade, Serbia

Received August 8, 2003

Background & objectives: Beta haemolytic streptococci (BHS), especially group A are still highly susceptible to penicillin. One possible explanation for this could be reduced growth capability in penicillin resistant BHS mutants. The present study was therefore undertaken to analyze the growth rates of BHS with decreased susceptibility to penicillin. Methods: Serial passages in the medium with subinhibitory concentration of penicillin were done to induce resistance to this antibiotic in 12 clinical isolates of BHS serogroups A, B, C, and G. Both penicillin susceptible (parental) and variants with decreased susceptibility to penicillin (laboratory strains) were grown in three different media and their growth rates were determined by counting the number of bacterial colonies and by measuring optical density of bacterial culture. Results: The lowest increase in minimal inhibitory concentration (MIC) value for penicillin (8-16 times) was obtained in BHS group A isolates, while the increase in MIC values of BHS groups B, C and G strains was higher (64-128 times) and they reached the level of complete resistance. Laboratory variants differed significantly from parental in their morphological and cultural characteristics. There were no statistically significant differences between the growth rates of penicillin susceptible and variants with decreased susceptibility to penicillin, though a delay in multiplication of the laboratory strains during exponential phase of growth was noted. Interpratation & conclusion: Though significant differences in phenotypic characteristics of penicillin susceptible and laboratory variants were noted, the results of this study provides no support to the assumption that variants of BHS with decreased susceptibility to penicillin of BHS were incapable for normal growth. Further studies needs to be done to find out the association between the decreased susceptibility to penicillin in the BHS and decreased growth capability in these bacteria. Key words Beta haemolytic streptococci - growth rate - penicillin - resistance - Streptococcus Since its descovery, though penicillin was extensively resistance to other antibiotics such as sulfonamides, used in the treatment of infections caused by tetracycline, erythromycin2. Also, resistance to penicillin streptococci, beta haemolytic streptococci (BHS), has been reported in other closely related species especially group A, remained highly susceptible to (S. pneumoniae, viridans streptococci, Enterococcus penicillin. However, failure in the penicillin treatment of spp.)3. It has also been shown that decreased infections caused by BHS group A isolates has been susceptibility to penicillin may be induced in BHS in reported in up to 30 per cent cases1, but no clinical isolate laboratory conditions4. It is generally assumed that resistant to penicillin has been identified so far. resistance to penicillin in BHS, as well as in other Streptococcus pyogenes successfully developed streptococci, is not mediated by b-lactamases, but by

237 238 INDIAN J MED RES (SUPPL) MAY 2004 low affinity penicillin binding proteins (PBPs)5. PBPs different media - THB, Brain heart broth (BHB) and act as enzymes, which catalyze the synthesis of the cell THB supplemented with horse serum (THB-HS). wall. Therefore, it is possible that changes in PBPs in Overnight culture of each strain was adjusted to an BHS result in development of mutants with severe optical density of 0.6 at 570 nm using an automated physiological defects that make them incapable for Multiskan EX reader (Labsystems, Helsinki, Finland), normal growth. The present study was undertaken with which corresponds to 3x108cfu/ml and further diluted to the objective to investigate growth capability of BHS 104cfu/ml. These suspensions were cultivated in THB, with decreased susceptibility to penicillin. BHB, and THB-HS, respectively, at 37 oC for 24 h. The samples (0.1 ml) were plated on the blood agar plates Material & Methods after 2, 4, and 24 h of incubation. The inoculated blood agar plates were incubated for 24 h at 37 oC and the Among 410 isolates of BHS examined in the colonies were counted and growth curves were laboratory of Institute of Microbiology and Immunology constructed. The second method comprised measuring during the period 1999-2001, twelve highly penicillin optical densities of bacterial cultures. Microtiter plates susceptible isolates of serogroups A, B, C, and G (three were filled with three different media (THB, BHB, THB- strains per serogroup) were selected. Two BHS group HS). Overnight cultures of each strain were diluted to A isolates were obtained from throat swabs and one 104cfu/ml and added into the wells (200 µl per well). from skin lesion; two BHS group B isolates from vagina The plates were incubated for 24 h at 37 oC and optical and one from urine; groups C and G included one isolate density was measured at 450 nm using Multiskan EX from throat and two from wounds, each. reader after 2, 4, and 24 h of incubation and growth rate curves were constructed. BHS were identified according to the current recommendation6. Isolates were cultivated in Todd Differences in growth rates of parental and laboratory Hewitt broth (THB) and stored at -70oC after adding variants were examined by Wilcoxon signed rank test, glycerol to a final concentration of 10 per cent. Minimal P < 0.05 was considered significant. inhibitory concentration (MIC) for penicillin (Galenika, Belgrade) was detected by broth dilution test, according Results to NCCLS7 recommendations. Serial passages of BHS in broth with subinhibitory Two groups of BHS were subjected to further concentrations of penicillin led to the development of analyses - 12 penicillin susceptible clinical isolates laboratory strains with 8-128 times raised MIC values. (parental) and 12 with decreased susceptibility to The lowest increase in MIC value (8-16 times) was penicillin (laboratory strains) that were obtained after obtained in three BHS group A isolates (Table), while 105 serial passages of parental isolates in THB the increase in MIC values of BHS groups B, C and G supplemented with subinhibitory concentration of strains was higher (64-128 times). The increase in the penicillin. Stability of decreased susceptibility to penicillin MIC values was discontinuous with periods of raise in laboratory strains was examined by serial sampling of alternating with periods of stagnation lasting for several the last subcultures onto the blood agar plates without passages. penicillin. Susceptibility of parental and laboratory strains to beta lactam antibiotics (ampicillin, amoxicillin, At the end of subcultivation, two group A strains cephalexin, cephachlor, ceftriaxone) was tested by the remained susceptible to penicillin (0.125 µg/ml), while agar dilution method according to the NCCLS7 the third one reached intermediate level of resistance recommendations. The morphological and cultural (0.25µg/ml). Other BHS strains developed complete characteristics of pen-s and pen-r strains were resistance to penicillin, having MIC equal or above 1µg/ml. compared. Growth rates of the two BHS groups were Laboratory strains also displayed decreased susceptibility determined by two different methods. The first one was to beta lactam antibiotics other than penicillin (Table). estimation of number of bacterial colonies on blood agar Evaluation of stability of decreased susceptibility to plates. Each strain of BHS was cultivated in three penicillin revealed different patterns in BHS group A OPAVSKI et al: DECREASED PENICILLIN SUSCEPTIBILITY & GROWTH RATE IN BHS 239 strains and other BHS strains tested. MIC values of laboratory variants were observable after additional 10 laboratory BHS group A strains were comparable to passages in medium without penicillin. that of parental isolates after 10 serial passages of the last subculture on the blood agar plates without penicillin. Growth rates of penicillin susceptible and those with The same procedure resulted in only one to two decreased susceptibility to penicillin were determined by concentrations decrease in BHS strains belonging to other two different methods. Both methods for evaluation of serogroups. growth rates disclosed similar patterns of growth dynamics in all parental and laboratory variants of BHS Comparison of morphological and cultural strains grown in THB and BHB. The representative characteristics between parental and laboratory strains growth curves obtained by measuring of optical density revealed significant differences. While penicillin of penicillin susceptible and variants with decreased susceptible isolates expressed all typical features of the susceptibility to penicillin of one BHS B strain grown in particular BHS groups, the cells of laboratory strains THB are shown in Fig.1. As compared to parental strains, were pale in Gram stained smears, their diameters were laboratory variants multiplied at slower rate during smaller and chains were longer. Also, the appearance exponential phase of growth, but there were no significant of their colonies was changed. The colonies of laboratory differences between OD values of parental and strains were small, white, and, most noticeable, without laboratory variants after 24 h of incubation. When the typical beta haemolysis. All these changes were strains were grown in media supplemented with horse particularly expressed in strains belonging to groups A serum, no delay in growth of variants with decreased and B. However, no differences between parental and susceptibility to penicillin was noted (Fig.2).

Table. MIC (µg/ml) values for penicillin, ampicillin, amoxycillin, cephalexin, cephachlor and ceftriaxon determined in 12 BHS strains before (parental) and after induction of resistance (laboratory) variants to penicillin

Penicillin Ampicillin Amoxycillin Cephalexin Cephachlor Ceftriaxon BHS groups P L P L P L P L P L P L

A1 0.015 0.125 0.030 0.125 0.030 0.25 0.060 1 0.125 1 0.030 0.25

A2 0.015 0.125 0.030 0.125 0.030 0.25 0.060 1 0.125 1 0.030 0.25

A3 0.030 0.250 0.015 0.125 0.015 0.25 0.060 0.5 0.125 2 0.015 0.125

B1 0.060 4 0.060 0.5 0.125 1 0.125 8 0.5 8 0.030 2

B2 0.125 2 0.125 0.25 0.060 1 0.125 8 0.5 8 0.060 1

B3 0.125 2 0.060 0.25 0.060 0.5 0.125 8 0.5 8 0.030 1

C1 0.015 2 0.030 0.25 0.030 0.5 0.125 8 0.125 8 0.030 0.25

C2 0.015 1 0.030 0.5 0.030 0.5 0.125 8 0.125 4 0.030 0.25

C3 0.30 1 0.030 0.25 0.030 0.5 0.125 8 0.125 2 0.030 0.25

G1 0.015 1 0.30 0.5 0.030 0.5 0.125 8 0.25 4 0.030 0.25

G2 0.030 0.5 0.030 0.25 0.030 0.25 0.125 8 0.5 2 0.030 1

G3 0.030 1 0.030 0.25 0.030 1 0.125 8 0.5 8 0.030 0.5

P, parental isolates; L, laboratory strains 240 INDIAN J MED RES (SUPPL) MAY 2004 stagnation periods were probably due to adaptation to changes PBPs. The application of serial passages in the previous study8 resulted in increase in MIC values in BHS group A strains, but they remained susceptible to penicillin. In the present study laboratory variants of BHS group A strains with intermediate level of resistance to penicillin were developed. In all laboratory variants of other BHS serogroups, MIC values reached the level of complete resistance. Similar differences between group A strains and those of other serogroups were noted when stability of decreased susceptibility to penicillin in laboratory variants was monitored. The decreased susceptibility to penicillin in laboratory variants of BHS group A strains was lost after serial passages on the blood agar plates without penicillin. The same procedure caused only one to two concentrations decrease in MIC values in penicillin resistant variants of BHS strains serogroups B, C, and G. There is no apparent explanation for the differences observed among group A strains and those belonging to the other serogroups tested. Evaluation Fig.1. Growth of penicillin susceptible (parental) and variants with of susceptibility to other beta lactam antibiotics showed decreased susceptibility to penicillin (laboratory) of BHS-B2 strain that induction of resistance to penicillin led to decreased in Todd Hewitt broth determined by measuring of optical density. susceptibility to other derivatives of penicillin and Fig.2. Growth of penicillin susceptible (parental) and with decreased cephalosporins in laboratory variants of BHS strains. susceptibility to penicillin (laboratory) of BHS-A2 strain in Todd Hewitt broth supplemented with horse serum determined by Probably subcultivation of the isolates in the presence measuring of optical density. of penicillin induced considerable multiple changes in PBPs. Discussion Significant changes were observed in morphological Since PBPs play a major role in the terminating phase and cultural characteristics of variants with decreased of cell wall synthesis5, the changes in these proteins susceptibility to penicillin. However, all the observed associated with decreased susceptibility to penicillin may alterations diminished during the subcultivation without hamper the normal multiplication of BHS and, thus, might penicillin. Reversibility of these changes indicated their be responsible for the absence of penicillin resistant BHS phenotypic origin. in nature. Thus, the possible influence of decreased susceptibility to penicillin on the growth of BHS was The growth rates of parental and laboratory variants explored through comparison of growth rates of penicillin of BHS strains were evaluated in three different media susceptible and variants with decreased susceptibility to by two methods in order to avoid possible errors. The penicillin. first method, based on counting colonies, registered the number of living bacterial cells, while the second Using the selective pressure with subinhibitory one, based on measurement of optical density, concentration of penicillin, BHS strains with decreased registered both living as well as dead cells. Moreover, susceptibility to penicillin were developed. Chemical the typical occurrence of streptococci in chains might mutagens or serial passages were employed earlier8 to have been a limitation for optical density decrease susceptibility to penicillin in BHS group A measurements. The differences observed between the strains. We chose the latter because it provided conditions growth rates of parental and laboratory variants were more similar to those associated with naturally occurring not significant and cultivation of penicillin susceptible resistance in bacteria. The observed increase in MIC and variants with decreased susceptibility to penicillin values in tested strains was not continuous and the in a highly nutritious supplemented medium such as OPAVSKI et al: DECREASED PENICILLIN SUSCEPTIBILITY & GROWTH RATE IN BHS 241 THB-HS, revealed that in the presence of high 3. Breiman RF, Butler JC, Tenover FC, Elliott JA, Facklam RR. quantity of nutritients laboratory variants grew as well Emergence of drug-resistant pneumococcal infections in the as penicillin susceptible. United States. JAMA 1994; 271 : 1831-5. 4. Gutmann L, Williamson R, Tomasz A. Physiological properties In conclusion, the results of this study provided of penicillin-binding proteins in group A straptococci. Antimicrob no support to the assumption that variants of BHS Agents Chemother 1981; 19 : 872-80. with decreased susceptibility to penicillin are 5. Severin A, Vaz Pato MV, Sa Figueiredo AM, Tomasz A. Drastic incapable for normal growth. Further studies changes in the peptidoglycan composition of penicillin resistant including large series of isolates analysed under both laboratory mutants of Streptococcus pneumoniae. FEMS in vitro and in vivo conditions are needed to be done Microbiol Lett 1995; 130 : 31-5. to explore the association between the decreased 6. Koneman EW, Allen SD, Janda WM, Schreckenberger PC, Winn susceptibility to penicillin in BHS, particularly in BHS WC. The Gram positive cocci, part II. In: Koneman EW, editor. group A, and possible defects in their growth Color atlas and textbook of diagnostic microbiology, 5th ed. capacity and virulence. Philadelphia: Lippincott-Raven Publishers; 1997 p.577-650. 7. National Committee for Clinical Laboratory Standards. 1997. References Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 4th ed. Approved standard M7-A4. 1. Sela S, Barzilai A. Why do we fail with penicillin in the treatment National Committee for Clinical Laboratory Standards, of group A streptococcus infections? Ann Med 1999, 31 : Wayne. Pa. 303-7. 8. Carsenti-Etesse H, Roger PM, Dunais B, Durgeat S, 2. Gerber MA. Antibiotic resistance: relationship to persistence Mancini G, Bensoussan M, et al. Gradient plate method to of group A streptococci in the upper respiratory tract. Pediatrics induce Streptococcus pyogenes resistance. J Antimicrob 1996; 97 : 971-5. Chemother 1999; 44 : 439-43.

Reprint requests: Dr Lazar Ranin, Institute of Microbiology & Immunology, School of Medicine, University of Belgrade Dr Subotica 1, 11000 Belgrade, Serbia e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 247-251

Genetic basis for mutacin N and of its relationship to mutacin I

J.D.F. Hale, M. Balakrishnan & J.R. Tagg

Department of Microbiology,University of Otago, Dunedin,New Zealand

Received August 6, 2003

Background & objectives: The mutans streptococci (MS) are a group of 7 species of dental caries- associated bacteria of which Streptococcus mutans and Streptococcus sobrinus are the most important in humans. Many MS produce bacteriocin-like inhibitory substances (BLIS), some of which have been characterised as small peptides capable of inhibiting the growth of closely-related species. These peptides have most commonly been referred to as mutacins. S. mutans strains N and UA140 appear to have closely similar BLIS activities. Both produce mutacins that seem to target the same species of bacteria. On closer analysis however, these two strains have been shown to produce distinctly different mutacins, known as mutacin N and mutacin I respectively. In the present study the mutacin N structural gene (mutN) was cloned and compared with the mutacin I structural gene (mutA). Methods: Cloning and sequencing of S. mutans N was done. The distribution of mutN using DNA from 216 streptococcal strains was determined by dot blotting. Results: Mut N was cloned and sequenced from an 1800 bp Bam HI/Eco RI fragment. PCR with the mutN primers mutNF and mutNR on the four mutN-positive strains identified identical bands to S. mutans N. The location of mutN differs significantly from that of mutA in that it is directly upstream of comC, a gene encoding a putative competence stimulating factor. Interpretation & conclusion: The close upstream proximity of mutN to comC suggests a link between mutacin N production and competence development. Further studies need to be done to detect competence-related genes in S. mutans strain N. Key words Bacteriocin - like inhibitory substance - mutacin - Streptococcus mutans Bacteriocins are typically defined as small ribosomally it was shown that S. mutans N produced mutacin N,3 a synthesised post-translationally modified proteinaceous 4806 Da peptide having partial homology with the molecules that generally inhibit the growth of closely b component of the two-peptide bacteriocin mutacin IV4. related bacteria. Bacteriocin-like inhibitory substances The aim of the present study was to clone the mutacin (BLIS) produced by the mutans streptococci (MS) have N structural gene and to identify other genes that may been referred to as mutacins1. Some of the strongly be implicated in mutacin N production. inhibitory mutacin-producing Streptococcus mutans have recently been categorised into four groups (A-D) Material & Methods according to their cross-immunity characteristics and their spectra of inhibitory activity2. This study2 further Bacterial strains and culture media: S. mutans N was established that the different bacteriocin activities of originally isolated from a supragingival plaque specimen S. mutans reflect distinct phylogenetic lineages. It is from an adult volunteer. The 216 streptococci screened becoming evident however that some strains that appear for the distribution of mutN included 106 isolates of to have similar spectra of inhibitory activity produce quite S. pyogenes, 82 S. salivarius, 14 S. mutans, different bacteriocins. For example, in a previous study 12 S. uberis, one S. dysgalactiae and one S. rattus 247 248 INDIAN J MED RES (SUPPL) MAY 2004 from J.R. Tagg’s collection. All strains were kept frozen Inverse PCR: DNA sequence outside of the cloned in skim milk at -70oC. S. mutans N was routinely grown contig was obtained by inverse PCR7. Briefly, in Todd Hewitt Broth (THB) (Difco, USA). approximately 5 µg of S. mutans N chromosomal DNA was digested for 4 h at 37oC with SpeI restriction enzyme. Cloning and sequencing of mutN: S. mutans N was 5 µl of this was then used for a ligation reaction using T4 o ligase [2 µl T4 buffer, 1 µl T4 ligase (Roche Diagnostics grown in 20 ml THB for 18 h at 37 C in 5 per cent CO2 in air. The chromosomal DNA was extracted using the Ltd., Lewes, England) and 12 µl MQ water] for 18 h at DNeasy Tissue Kit (Qiagen Ltd., Crawley, England) 12oC. 1 µl of this preparation was used in a PCR reaction according to the manufacturer’s instructions, with the using primers NinvF (5’-G C G T G C G C A A G T A following modifications. The THB overnight cultures C G G G G G T C T A-3’) and NinvR (5’-G A A A G T were pelleted by centrifugation at 1155xg and washed G T T C A T T A T C C A T T A C G-3’) designed to twice in 10 mM Tris-EDTA buffer pH. The cells were the cloned contig. The PCR was carried out using Taq resuspended in 140 µl lysis buffer containing 40 µl polymerase (Roche Diagnostics Ltd., Lewes, England) 50 mg/ml lysozyme and incubated for 30 min at 37oC. for 30 cycles, with denaturation at 95°C for 2 min, DNA was digested with BamHI and EcoRI (Roche annealing at 50°C for 30 sec and extension at 68°C for Diagnostics) and separated on a 1 per cent agarose 10 min. (Roche Diagnostics) TAE buffer gel. Southern hybridisation was carried out according to the Evaluation of the distribution of mutN in streptococci: manufacturer’s instructions [Hybond™-N+ (Amersham The distribution of mutN using DNA from Pharmacia Biotech Inc)]. 216 streptococcal strains of six species was determined by dot blotting. The DNA was derived by a method based on that of Upton et al8, but with use of only a The degenerate oligopeptide probe MutNC single phenol-chloroform extraction step. 5 µl of each (5’-GCWTATGGWGCWGCWAARGG-3’) was based DNA sample was applied to a nylon membrane on residues 12-18 (N-AYGAAKG-C) of the mutacin N [Hybond™-N+ (Amersham Pharmacia Biotech Inc)] 3 amino acid sequence . MutNC was end-labelled with via a vacuum manifold followed by 100 µl of 2 x SSC [g32P]ATP using T4 polynucleotide kinase and used to [0.3 M NaCl, 0.03 M sodium citrate (pH 7)]. screen Southern hybridisations using established Denaturation of the DNA was by exposure to two x 2 methodology5. min washes of 0.4 M NaOH followed by two x 2 min washes with 1 M Tris HCl. The membrane was then Competent Escherichia coli DH5a cells were exposed to UV light for approximately 5 min and probed prepared using the protocol described by Sheng et al6. with a digoxygenin-dUTP (Roche Diagnostics Ltd., Ligated pUC19 containing the mutN insert was Lewes, England)-labelled mutN probe, derived using electroporated into strain DH5a. Briefly, 3 µl of DNA PCR primers mutNF (5’-C C G T C A A G C T G C T G A T A C A T T C C-3’) and mutNR (5’ C G A T A was mixed with 40 µl of DH5a cells, thawed on ice TM T T T A A A C C T G C A C C G A C T A C G-3’). and placed in a chilled 0.1 cm electrode gap Pulsar The PCR was carried out using Taq polymerase (Roche cuvette (Bio-Rad Laboratories, California, USA) and Diagnostics Ltd., Lewes, England) for 28 cycles, with left for 1 min. A single pulse was applied (1.5 kV , denaturation at 95°C for 2 min, annealing at 55°C for 30 25 µF, 200 W ). Following electroporation, the cells sec and extension at 72°C for 30 sec. were resuspended in 1 ml of 2YT broth and incubated at 30oC with shaking for 1 h. Aliquots (50 µl) of the Deferred antagonism test: The P-type BLIS fingerprint cell suspensions were then plated onto Luria broth was as described by Tagg et al9 but S. mutans was (LB) agar plates containing 50 µg/ml ampicillin and o grown at 37 C in air with 5 per cent CO2. X-gal. Results DNA sequencing: Sequencing was carried out at the Centre for Gene Research at the University of Otago Cloning and sequencing of mutN: An 1800 bp BamHI/ using a Perkin-Elmer ABI 377A sequencer. EcoRI fragment containing mutN was identified (Fig.1), HALE et al: CLONING OF MUTN 249 cloned and sequenced. A further 900 bp was sequenced is MNVEENIMSFDVNSYNDLTRDELSQTIGG and following inverse PCR. Analysis of this sequence the predicted mutacin N propeptide is SRQAADTFLS revealed four open reading frames (ORF) (Fig.2). In SGAYGAAKGVTACASTGV YVVPATLVACGVV addition to mutN, there were ORF having homology to GAGLNIAFPH. Underlined amino acids indicate comC , a transposase gene (orfA ) and a bacteriocin- differences in the predicted sequence on the basis of like peptide gene (blpI). The translated product of mutN the nucleotide sequence when compared with the indicated a 79 amino acid precursor peptide having a sequence previously obtained by protein sequencing3. double glycine cleavage site preceding the mutacin N Directly preceding mutN was an ORF with homology to propeptide. The predicted mutacin N leader sequence the bacteriocin-like peptide gene blpI from Streptococcus pneumoniae strain KNR.7/8710. M12 Evaluation of the distribution of mutN in streptococci: The distribution of mutN was investigated using a probe generated by PCR using the mutN primers mutNF and mutNR. Probing was carried out using a dot-blot method with DNA preparations from 216 strains of streptococci. Of these, 4 of 14 S. mutans hybridised with the mutN probe. None of the 106 S. pyogenes, 82 S. salivarius, 12 S. uberis, or the single representative strains of S. dysgalactiae and S. rattus were positive for mutN. The mutN-positive strains were shown to produce a 1886 bp similar P-type pattern of 777 against nine standard 1700 indicators9. In addition, these mutN-positive strains produced an identical inhibitory BLIS pattern against an extended set of indicator strains (data not shown). MutN- positive strains were resistant to each other when cross- tested by deferred antagonism indicating they have cross- immunity and thus may be producing similar bacteriocins. PCR with the mutN primers mutNF and mutNR on the four mutN-positive strains identified identical bands to S. mutans N.

Discussion

Fig.1. Autoradiograph of an EcoRI/BamHI digest of S. mutans N 32 In this study the mutacin N structural gene was chromosomal DNA hybridised with µ P labelled mutNC probe. The 1800 bp fragment used for sequencing is indicated by the arrow. cloned and sequenced. MutN was sequenced from an Lane 1: l DNA digested with PstI; lane 2: S. mutans N digested 1800 bp BamHI/EcoRI fragment.The derived sequence DNA; lane 3: negative control S. mutans M46. established mutacin N to comprise a propeptide of 50 amino acids, one more than previously obtained by Edman N-terminal sequencing3. A 29-amino acid leader sequence with a typical double glycine cleavage site was identified upstream of mutN. Immediately adjacent to mutN is an ORF with homology to a bacteriocin-like peptide encoded by blpI10. Further upstream was an ORF identical to comC, a gene involved in competence Fig.2. Open reading frames (ORFs) found associated with mutN. development in S. mutans strain BM71. One further Nearest BLAST homology (% identity) match indicated in brackets ORF was identified downstream of mutN using inverse underneath. PCR which was found to have homology with a 250 INDIAN J MED RES (SUPPL) MAY 2004 S. agalactiae transposase (orfA). In an initial screen and competence development. A link between mutN was shown to be present in four strains of bacteriocin production and competence has already been S. mutans. shown in Streptococcus gordonii (Heng, personal communication). Further studies are required to detect The MS are known to frequently produce mutacins, the presence and activity of competence-related genes a number of which have been well characterised in S. mutans strain N. including: mutacin N3, mutacin I11, mutacin II12, mutacin III13, mutacin IV4 and B-Ny26614. S. mutans N has been In this study the mutN gene has been cloned and shown to produce a similar BLIS activity spectrum to sequenced. Homology of adjacent ORFs indicate that that of the mutacin I producer S. mutans UA140. mutN may be part of a two-component bacteriocin Mutacin I has been identified as a 2364 Da lantibiotic, a system and also that there may be some association with class of bacteriocins characterised by the presence of competence development. The fact that these two lanthionine residues and modified amino acids such as S. mutans strains produce markedly different 2, 3-didehydroalanine and 2, 3-didehydrobutyrine. MutA bacteriocins, yet have closely similar inhibitory spectra, the gene encoding mutacin I, is located within an operon requires further investigation. containing 14 genes involved in the processing, transport of the lantibiotic and the self immunity of the producing Acknowledgment cell11. The authors acknowledge the advice given by Dr Philip Wescombe. Support for these studies has been from the Marsden Mutacin N is clearly a different molecule to mutacin Fund and Health Research Council of New Zealand. I, despite the two producer strains showing identical inhibitory spectra. Mutacin N differs markedly from References mutacin I in that it is not a lantibiotic. The reason for the apparent close similarity in the activity spectra remains 1. Hamada D, Ooshima T. Inhibitory spectrum of a bacteriocin to be determined. like substance (mutacin) produced by some strains of Streptococcus mutans. J Dent Res 1975; 54 : 140-5. The discovery of a bacteriocin-like ORF with 2. Balakrishnan M, Simmonds RS, Kilian M, Tagg JR. Different homology to blpI was unexpected. Several two- bacteriocin activities of Streptococcus mutans reflect distinct phylogenetic lineages. J Med Microbiol 2002; 51 : 941-8. component bacteriocins15,16 and lantibiotics 17,18 have been reported but to date only one strain of S. mutans 3. Balakrishnan M, Simmonds RS, Carne A, Tagg JR. has been found to produce two bacteriocins4. BlpI is Streptococcus mutans strain N produces a novel low molecular mass non- lantibiotic bacteriocin. FEMS Microbiol Lett located adjacent to another bacteriocin-like ORF, blpJ 2000;183 : 165-9. in S. pneumoniae strain KNR7/87. The two peptides encoded by blpI and blpJ are thought to comprise a 4. Qi F, Chen P, Caufield PW. The group I strain of Streptococcus mutans, UA140, produces both the lantibiotic mutacin I and a two-component bacteriocin with homology to nonlantibiotic bacteriocin, mutacin IV. Appl Environ Microbiol thermophilin 13, a two-component bacteriocin from 2001; 67 :15-21. Streptococcus thermophilus strain Sfi1315. 5. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: A laboratory manual. New York: Cold Spring Harbor Laboratory Competence refers to the ability of a bacterial cell to Press; 1989. take up foreign exogenous DNA. ComC, comD and 6. Sheng Y, Mancino V, Birren B. Transformation of Escherichia comE respectively encode a quorum sensing peptide, coli with large DNA molecules by electroporation. Nucleic Acids histidine kinase and a response regulator. Their Res 1995; 23 :1990-6. involvement in the establishment of competence in 7. Triglia T, Peterson MG, Kemp DJ. A procedure for in vitro 19 S. mutans has been shown . ComC is found in the amplification of DNA segments that lie outside the boundaries opposite orientation to comD and comE in S. mutans19. of known sequences. Nucleic Acids Res 1988; 16 : 8186. The close upstream proximity of mutN to comC suggests 8. Upton M, Tagg JR, Wescombe P, Jenkinson HF. Intra-and that there may be a link between mutacin N production interspecies-signaling between Streptococcus salivarius and HALE et al: CLONING OF MUTN 251

Streptococcus pyogenes mediated by SalA and SalA1 lantibiotic 14. Mota-Meira M, Lacroix C, LaPointe G, Lavoie MC. peptides. J Bacteriol 2001; 183 : 3931-8. Purification and structure of mutacin B-Ny266: a new lantibiotic produced by Streptococcus mutans. FEBS Lett 1997; 9. Tagg JR, Bannister, LV. “Fingerprinting” b-haemolytic 410 : 275-9. streptococci by their production of and sensitivity to bacterocine like inhibitors. J Med Microbiol 1979; 12 : 15. Marciset O, Jeronimus-Stratingh MC, Mollet B, Poolman B. 397-411. Thermophilin 13, a nontypical antilisterial poration complex bacteriocin, that functions without a receptor. J Biol Chem 1977; 10. de Saizieu A, Gardes C, Flint N, Wagner C, Kamber M, 272 :14277-84. Mitchell TJ, et al. Microarray-based identification of a novel Streptococcus pneumoniae regulon controlled by an autoinduced 16. Moll G, Ubbink-Kok T, Hildeng-Hauge H, Nissen-Meyer J, Nes IF, Konings WN, et al. Lactococcin G is a potassium ion- peptide. J Bacteriol 2000; 182 : 4696-703. conducting, two-component bacteriocin. J Bacteriol 1996; 178 : 11. Qi F, Chen P, Caufield PW. Purification and biochemical 600-5. characterization of mutacin I from the group I strain of 17. McAuliffe O, Hill C, Ross RP. Each peptide of the two- Streptococcus mutans, CH43, and genetic analysis of mutacin I component lantibiotic lacticin 3147 requires a separate biosynthesis genes. Appl Environ Microbiol 2000; 66 : 3221-9. modification enzyme for activity. Microbiology 2000; 146 : 12. Chikindas M L, Novak J, Driessen JM, Konings WN, Schilling 2147-54. KM, Canfield PW, et al. Mutacin II, a bactericidal lantibiotic 18. Navaratna MA, Sahl HG, Tagg JR. Two-component anti- from Streptococcus mutans. Antimicrob Agents Chemother Staphylococcus aureus lantibiotic activity produced by 1995; 39 : 2656-60. Staphylococcus aureus C55. Appl Environ Microbiol 1998; 64 : 4803-8. 13. Qi F, Chen P, Caufield PW. Purification of mutacin III from group III Streptococcus mutans UA787 and genetic analyses of 19. Li YH, Lau PC, Lee JH, Ellen RP, Cvitkovitch DG. Natural mutacin III biosynthesis genes. Appl Environ Microbiol 1999; genetic transformation of Streptococcus mutans growing in 65 : 3880-7. biofilms. J Bacteriol 2001; 183 : 897-908.

Reprint requests: Prof. John R. Tagg, Department of Microbiology, University of Otago, PO Box 56, Dunedin, New Zealand e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 252-256

New group D streptococcal species

Laurent Schlegel*†, Francine Grimont†, Patrick A. Grimont† & Anne Bouvet*

*Centre National de Référence des Streptocoques, Service de Microbiologie, Hôtel Dieu AP-HP, Université Paris V & †Unité Biodiversité des Bactéries Pathogènes Emergentes, INSERM U389, Institut Pasteur, Paris, France

Received August 6, 2003

Background & objectives: The heterogeneity of group D streptococci led to the identification of various biotypes of Streptococcus equinus and Streptococcus bovis and to the description of new species. The objective of the present study was to improve the phenotypic delineation between species and to clarify their respective phylogenetic position. Methods: Physiological and genomic analyses were carried out in 84 representative strains of the group D streptococci. Biotypes were determined with the API 20 strep and rapid ID 32 STREP systems of identification. Quantitative DNA-DNA hybridization under stringent conditions and values of the D DTm allowed to delineate species and subspecies. The phylogenic position of the different genomic groups was determined by comparing the sequences of their 16S rDNA. Results: Four DNA-clusters, including seven species or subspecies, were characterized. Differential associations of biochemical characters allowed their identification. S. equinus and the type strain of S. bovis belonged to a single species. S. gallolyticus, S. bovis biotype II.2, and S. macedonicus formed a single DNA-cluster including three different subspecies. These were designated as S. gallolyticus subsp. gallolyticus, S. gallolyticus subsp. pasteurianus, and S. gallolyticus subsp. macedonicus. The two other DNA-clusters corresponded to the two subspecies of S. infantarius, and to S. alactolyticus. Interpretation & conclusion: This study presented a new classification associated with an identification scheme of group D streptococci. The changes in this classification demonstrate the interest of a polyphagic approach of the bacterial identification. Key words Group D streptococci - Streptococcus bovis - S. equinus - S. gallolyticus - S. infantarius - taxonomy Group D streptococci, the common inhabitants of the In 1984, a genomic classification of the ‘S. equinus / intestinal flora of vertebrates, were first identified by S. bovis’ complex delineated six different DNA- their antigenic and biochemical characteristics1,2. groups6. The group 1 included both S. equinus and some Genomic studies confirmed the separation of enterococci biotype II.1 S. bovis strains. The groups 2, 3, and 4 from non enterococci3,4. The group D streptococcal corresponded to different unnamed genospecies. The strains were then commonly designated as S. equinus group 6 was called S. alactolyticus, whereas the group 5 or S. bovis depending on their origin and on their (S. saccharolyticus) was further included in the biochemical characteristics. Identification of S. bovis Enterococcus genus7. strains was based on their phenotype. Biotypes I and II designated mannitol-positive and mannitol-negative Correlation studies between the genomic classification strains, respectively. Biotype II included two different and the biotypes led to identify phenotypic variants, which phenotypes according to the presence (biotype II.2) or were progressively described as individual species or absence (biotype II.1) of b-glucuronidase5. subspecies. The DNA-group 2 (S. bovis biotype I) was 252 SCHLEGEL et al: GROUP D STREPTOCOCCI 253 called S. gallolyticus8; it is associated with endocarditis with a selection of the available sequences of 16S rDNA and colonic carcinoma9. The DNA-group 4 was called from GeneBank and phylogenetic analysis of the 16S S. infantarius10; it includes two subspecies of medical rDNA data were performed by using the Clustal multiple importance11. Two additional species S. macedonicus12 sequence method with the Megalign® program from the and S. waius13 were further described. DNAstar® package. A distance matrix was computed using a Kimura model for nucleotide substitution. A Using DNA-DNA hybridization and 16S rDNA phylogenetic tree was generated from the distance sequencing, we confirmed that S. macedonicus and S. matrices using the neighbour joining method. waius were subjective synonyms, and demonstrated that S. bovis biotype II.2 and S. macedonicus were Results & Discussion subspecies of a same DNA-cluster, which included the previously described S. gallolyticus14. Here, we Phenotypic characterization: The phenotypic tests reappraise the classification of the S. bovis/S. equinus confirmed that all strains were group D streptococci. complex with the objective to improve the phenotypic None of them grew on bile-esculine agar and in 6.5 per delineation between species and to clarify their respective cent NaCl broth. They produced leucine aminopeptidase, phylogenetic position. alanyl-phenylalanyl-proline arylamidase, and acetoin. They did not produce pyrrolidonyl-arylamidase and arginine dihydrolase. Our results confirmed that S. Material & Methods equinus and S. bovis differed by classical characters. The species and subspecies exhibited noticeable Representative strains (n=84) of the DNA-groups 1, differences (Table). The major differential associations 2, 4, and 6, and of the new species were obtained from were: (i) acidification of mannitol and hydrolysis of gallate various collections. They were identified as group D by S. gallolyticus ; (ii) presence of b-glucuronidase and streptococci by conventional methods15. Production of b-mannosidase, and acidification of trehalose by enzymes and fermentation of carbohydrates were S. gallolyticus subsp. pasteurianus (S. bovis biotype determined using both the Api 20 Strep and II.2); (iii) absence of b-glucosidase and b-glucuronidase, rapid ID32 STREP identification systems (BioMérieux, with presence of b-galactosidase by S. gallollyticus France). Hydrolysis of methyl-gallate was assayed by subsp. macedonicus; (iv) absence of b-glucosidase by the colorimetric test8. most strains of S. infantarius subsp. infantarius, associated with the absence of b-glucuronidase and The DNA was extracted and purified using the b-galactosidase, and with the acidification of glycogen phenol/chloroform method16. DNA from type or and raffinose; (v) presence of b-glucosidase, and reference strains were labeled by using the Megaprim acidification of methyl-b-glucopyranoside, but not of DNA labeling® reaction kit (Amersham, UK). glycogen and raffinose by S. infantarius subsp. coli; Quantitative DNA-DNA hybridization was carried out (vi) absence of acidification of most of the in a liquid medium under stringent conditions of 60°C for carbohydrates, including lactose by S. alactolyticus. 16 h, according to a modification of the S1-nuclease/ trichloracetic acid precipitation method17. The DNA-DNA hybridization: The existence of four DNA temperature (Tm), at which 50 per cent of the reassociated homology groups within the S. bovis/S. equinus complex DNA is hydrolysed by the S1-nuclease, was determined. was demonstrated. DNA-cluster I is formed by the D The difference ( Tm) between the melting temperatures strains of S. equinus and the type-strain of S. bovis (91 of homo- and heteroduplexes allowed to quantify the to 100% DNA homology) Type strains of the former DNA divergence between strains with high level of DNA S. gallolyticus, S. macedonicus, and S. waius species, homology17. and the strains of S. bovis biotype II.2 belonged to the same DNA-cluster II. This cluster contained three Polymerase chain amplification of 16S rRNA different subgroups. The level of homology between D £ encoding DNA, and sequencing of amplified fragments strains was over 77 per cent ( Tm 2°C) within each 18 D £ were carried out as previously described . Alignment subgroup, and ranged from 48 to 93 per cent ( Tm 6°C) 254 INDIAN J MED RES (SUPPL) MAY 2004 ++ ++ S. alactolyticus coli subsp. V V S. infantarius macedonicus infantarius subsp. +---- pasteurianus V gallolyticus +++- -+----- +++- --+---- - / +---+--- --+v--- - +- / +- / +- / +- / + v-+----- ++++--+V +- / + + +v++---- v - - + - + + + v - - + + + + - - + + v - + - + ------Differential biochemical characteristics of group D streptococcal strains Differential equinus S. gallolyticus . / S. bovis S Table. Table. 20% of positive reaction; v, 21 to 79% of positive reactions 20% of positive reaction; v, £ -galactosidase ß-glucosidase a Aesculine Gallate ß-glucuronidase Starch ß-galactosidase (ß-GAR) ß-galactosidase (ß-GAL) ß-mannosidase Glycogen Inulin Lactose Mannitol Meth-ß-D-glucopyranoside Raffinose Trehalose No. of strains n = 9 + 8 n = 13 n = 21 n = 5 n =10 n = 14 n = 4 Production of Hydrolysis of Acidification of 80% of positive reactions; -, ³ +, SCHLEGEL et al: GROUP D STREPTOCOCCI 255 between the different subgroups. The first subgroup S. gallolyticus (DNA-cluster II). Within this cluster II, includes the reference strain of the DNA-group 2 of the divergence between the 16S rDNA sequences from Farrow et al6, and the type strains of S. gallolyticus the type or reference strains of S. gallolyticus, and S. caprinus. The second subgroup included the type S. macedonicus, and S. bovis biotype II.2 were in strains of S. macedonicus and S. waius. The last agreement with the delineation of the three subspecies, subgroup was formed by S. bovis II.2, including the type whereas the 99.5 per cent similarity of sequences strain of S. pasteurianus described independently as a between S. waius and S. macedonicus confirmed that 19 D £ new species . The Tm 6°C indicating the strong they belonged to a single genomic group, as recently affinity of heteroduplexes constructed with the strains reported by Manachini et al22. The third division included of the different subgroups, we proposed to name the S. alactolyticus and the subjective synonym three subspecies: S. gallolyticus subsp. gallolyticus, S. intestinalis (DNA-cluster IV). macedonicus, and pasteurianus. The DNA-cluster III included S. infantarius subsp. infantarius and subsp. This study presents a new classification associated coli. The DNA-cluster IV corresponded to with an identification scheme of group D streptococci. S. alactolyticus. Seven species or subspecies could be identified according to differential biochemical reactions. The changes in this 16S rDNA sequencing and phylogenetic analysis: We classification of the ‘S. bovis/S. equinus’ complex added in the Genbank databank the sequences of 16S demonstrate the interest of a polyphasic approach of rDNA that we have determined14. The 99 per cent the bacterial identification. similarity of the sequence of 1455 base fragments of the type strains of S. equinus and S. bovis confirmed them References to belong to the same species. Similarly, the 99.6 per cent homology between the two type strains of S. 1. Lancefield RC. A serological differentiation of human and other alactolyticus6 and S. intestinalis20 was in agreement groups of haemolytic streptococci. J Exp Med 1933; 57 : with their closely related whole-cell proteins patterns21. 571-95. We constructed a phylogenetic tree from the sequences 2. Facklam RR. Recognition of group D streptococcal species of human origin by biochemical and physiological tests. of 350 to 400 bp from our results and from those from Appl Microbiol 1972; 23 : 1131-9. previouly described type strains (Fig.). Three major 3. Kilpper-Bälz R, Fischer G, Schleifer KH. Nucleic acid divisions corresponding to the four DNA-clusters were hybridization of group N and group D streptococci. observed. The first division included the type strains of Curr Microbiol 1982; 7 : 245-50. S. equinus and S. bovis (DNA-cluster I), and the two 4. Bentley RW, Leigh JA, Collins MD. Intrageneric structure of subspecies of S. infantarius (DNA-cluster III). The Streptococcus based on comparative analysis of small-subunit second division includes the three subspecies of rRNA sequences. Int J Syst Bacteriol 1991; 41 : 487-94. 5. Coykendall AL, Gustafson KB. Deoxyribonucleic acid hybridizations among strains of Streptococcus salivarius and Streptococcus bovis. Int J Syst Bacteriol 1985; 35 : 274-80. 6. Farrow JAE, Kruze J, Phillips BA, Bramley AJ, Collins MD. Taxonomic studies on Streptococcus bovis and Streptococcus equinus: description of Streptococcus alactolyticus sp. nov. and Streptococcus saccharolyticus sp. nov. System Appl Microbiol 1984; 5 : 467-82. 7. Rodriguez U, Collins MD. Phylogenic analysis of Streptococcus saccharolyticus based on 16S rRNA sequencing. FEMS Microbiol Lett 1990; 71 : 231-4. 8. Osawa R, Fujisawa T, Sly LI. Streptococcus gallolyticus sp. nov.; gallate degrading organisms formerly assigned to Streptococcus bovis. System Appl Microbiol 1995; 18 : 74-8. 9. Klein RS, Recco RA, Catalano MT, Edberg SC, Casey JI, Fig. Simplified phylogenetic tree of group D streptococci based on Steigbigel NH. Association of Streptococcus bovis with comparative analysis of 350-400 bp 16S rDNA sequences. carcinoma of the colon. New Eng J Med 1977; 297 : 800-2. 256 INDIAN J MED RES (SUPPL) MAY 2004

10. Bouvet A, Grimont F, Collins MD, Benaoudia F, Devine C, 17. Grimont PAD, Popoff MY, Grimont F, Coynault C, Lemelin Regnault B, et al. Streptococcus infantarius sp. nov. related to S. M. Reproducibility and correlation study of three bovis and S. equinus. Adv Exp Med Biol 1997; 418 : 393-5. deoxyribonucleic acid hybridization procedures. Curr Microbiol 11. Schlegel L, Grimont F, Collins MD, Régnault B, Grimont PAD, 1980; 4 : 325-30. Bouvet A. Streptococcus infantarius sp. nov., Streptococcus 18. Janvier M, Grimont PA. The genus Methylophaga, a new line of infantarius subsp. infantarius subsp. nov., and Streptococcus g infantarius subsp. coli subsp. nov., isolated from humans and descent within phylogenic branch of Proteobacteria. food. Int J Syst Evol Microbiol 2000; 50 : 1425-34. Res Microbiol 1995; 146 : 543-50. 12. Tsakalidou E, Zoidou E, Pot B, Wassil L, Ludwig W, Devriese 19. Poyart C, Quesne G, Trieu-Cuot P. Taxonomic dissection of the LA, et al. Identification of streptococci from Greek Kasseri Streptococcus bovis group by analysis of manganese-dependant cheese and description of Streptococcus macedonicus sp. nov. superoxide dismutase gene (SodA) sequences : reclassification Int J Syst Bacteriol 1998; 48 : 519-27. of ‘Streptococcus infantarius subsp. coli’ as Streptococcus 13. Flint SH, Ward LJH, Brooks JD. Streptococcus waius sp. nov., lutetiensis sp. nov. and of Streptococcus bovis biotype 11.2 as a thermophilic streptococcus from a biofilm. Int J Syst Bacteriol Streptococcus pasteurianus sp. nov. Int J Syst Evol Microbiol 1999; 49 : 759-67. 2002; 52 :1247-55. 14. Schlegel L, Grimont F, Ageron E, Grimont PA, Bouvet A. 20. Robinson IM, Stromley JM, Varel VH, Cati EP. Streptococcus Reappraisal of the taxonomy of the Streptococcus bovis/ intestinalis, a new species from the colons and feces of pigs. Streptococcus equinus complex and related species description Int J Syst Bacteriol 1988; 38 : 245-8. of Streptococcus gallolyticus subsp. nov., S. gallolyticus subsp. macedonicus subsp. nov., and S. gallolyticus subsp. pasteurianus 21. Vandamme P, Devriese LA, Haesebrouck F, Kersters K. subsp. nov. Int J Syst Evol Microbiol 2002; 53 (Pt 3) : 631-45. Streptococcus intestinalis (Robinson et al. 1988) and 15. Facklam RR, Washington JA II. Streptococcus and related Streptococcus alactolyticus (Farrow et al. 1984) are catalase-negative Gram-positive cocci. In : Balows A, Hausler phenotypically indistinguishable. Int J Syst Bacteriol 1999; 49 : WJ, Herrmann KL, Isenberg HD, Shadomy HJ, editors. Manual 737-41. of clinical microbiology, 5th ed. Washington DC: American 22. Manachini PL, Flint SH, Ward LJH, Kelly W, Fortina MG, Society for Microbiology; 1991 p. 238-57. Parini C, et al. Comparison between Streptococcus macedonicus 16. Grimont F, Grimont PAD. Determination of rRNA gene and Streptococcus waius strains and reclassification of restriction patterns. In : Howard J, Withcombe DM, editors. Streptococcus waius (Flint et al. 1999) as Streptococcus Diagnostic bacteriology protocols. Totowa, NJ: Humana Press macedonicus (Tsakalidou et al 1998). Int J Syst Evol Microbiol Inc; 1995 p. 149-64. 2002; 52 : 945-51.

Reprint requests: Prof. Anne Bouvet, Centre National de Référence des Streptocoques, Service de Microbiologie, Hôtel Dieu Paris VI University,1, place du Parvis Notre-Dame, F-75181 Paris 04, France e-mail: [email protected] Indian J Med Res 119 (Suppl) May 2004, pp 257-261

Susceptibility testing of Streptococcus mitis group isolates

G. Bancescu, S. Dumitriu, A. Bancescu, C. Defta, M. Pana*, D. Ionescu, S. Alecu & M. Zamfirescu

Chair of Microbiology, University of Medicine & Pharmacy "Carol Davila" & *National Reference Laboratory for Streptococci, Cantacuzino Institute, Bucharest, Romania

Received August 8, 2003

Background & objectives: Suppurative oral and maxillofacial infections are usually mixed infections due to aerobic and anaerobic bacteria, most frequently by oral streptococci and antimicrobial treatment is necessary for such infections. The aim of this study was to investigate the antimicrobial susceptibility of Streptococcus mitis group strains isolated from Romanian patients with different oral and maxillofacial infections. Methods: Eighty-five isolates belonging to S. mitis group isolated from pus samples were identified at species level by the Rapid ID 32 STREP system. The E test was used to determine the susceptibilities of the isolates to penicillin, ampicillin, cefotaxime, erythromycin, clindamycin, chloramphenicol and tetracycline. Results: Of the 151 samples studied, 85 isolates belonged to S. mitis group. The minimum inhibitory concentration (MIC) values (mg/l) ranged from 0.016-0.75 for penicillin, 0.016-2 for ampicillin, 0.016- 1 for cefotaxime, 0.016-4 for erythromycin, 0.016-0.047 for clindamycin, 0.5-4 for chloramphenicol and 0.047-256 for tetracycline. Interpretation & conclusion: The low susceptibility and the resistance to some commonly used antibiotics found in this study indicated a need for a careful surveillance of the susceptibility pattern of oral streptococci isolates of clinical significance. Clindamycin and chloramphenicol might be suitable alternative agents in treatment of oral and maxillofacial infections involving penicillin-resistant bacteria and in case of patients with hypersensitivity to beta-lactam antibiotics. Key words Antimicrobial susceptibility - oral and maxillofacial infections - Streptococcus gordonii - Streptococcus mitis - Streptococcus oralis - Streptococcus sanguis The frequency of suppurative oral and maxillofacial treatment of such diseases1-4. However, the emerging infections in Romania is high1, and these are usually mixed resistance to beta-lactam antibiotics among viridans infections produced by an association of aerobic and streptococci may represent a potential clinical problem anaerobic bacteria. The most frequently isolated in the management of therapy. Therefore, the aim of the facultatively anaerobic microorganisms were the oral present study was to test the antimicrobial susceptibility streptococci, in particular those belonging to anginosus of Streptococcus mitis group isolates obtained from (the former Streptococcus milleri group) and mitis Romanian patients with different oral and maxillofacial groups. Although surgical drainage is of primary infections. importance, administration of antimicrobial therapy becomes necessary in serious infections and their Material & Methods complications, particularly in immunocompromised patients. Antibiotics are often prescribed empirically and Pus samples were obtained primarily by needle penicillin is still considered the drug of choice for the aspiration and occasionally by swabs dipped into the 257 258 INDIAN J MED RES (SUPPL) MAY 2004 surgical wound from 151 patients presenting at the chloramphenicol (100% susceptible isolates) and 0.047- Emergency Service of the Oral and Maxillofacial 256 mg/l for tetracycline (48.2% susceptible, 10.6% Surgery Clinic, Bucharest, with pyogenic oral and intermediate susceptible and 41.2% resistant isolates). maxillofacial infections (dento-alveolar abscess, All isolates classified as intermediate susceptible to periodontal abscess, fascial space abscess, pericoronitis, cefotaxime and/or to ampicillin were also found to be sialadenitis, osteomyelitis of the mandible and maxillary intermediate susceptible to penicillin. Regarding the sinusitis). Gram stained smears and cultures were association of resistance to antibiotics belonging to performed in all samples. Streptococcal strains were different classes, only one isolate of S. mitis and one of identified at species level using the Rapid ID 32 STREP S. sanguis was found with low sensitivity to penicillin system (BioMérieux, Marcy-l`Etoile, France). Isolates and resistance to tetracycline. In case of S. oralis a belonging to mitis group were further tested for single isolate with intermediate susceptibility to all susceptibility to penicillin, ampicillin, cefotaxime, beta-lactam antibiotics tested was found to be resistant erythromycin, clindamycin, chloramphenicol and to erythromycin, whereas another isolate was resistant tetracycline by the E test (AB Biodisk, Solna, Sweden). to erythromycin and tetracycline. S. pneumoniae ATCC 49619 was used for quality control. The inoculum was standardized at 0.5 Discussion McFarland turbidity in order to give semiconfluent growth on Müller-Hinton agar (Difco Laboratories, Detroit, USA) supplemented with 5 per cent sheep Though routine susceptibility testing of oral blood. The plates were incubated at 35ºC for 24 h in 5 streptococci involved in oral and maxillofacial infections is usually not performed, it is necessary to monitor from per cent CO2 atmosphere, except for erythromycin and clindamycin when ambient atmosphere was required. time to time the susceptibility patterns of the clinical Susceptibility breakpoint concentrations were those isolates belonging to these bacteria within a geographical recommended by the National Committee for Clinical area. This study was focused on the susceptibility testing Laboratory Standards5. of clinical isolates belonging to mitis group, since these microorganisms are frequently isolated from oral and maxillofacial infections. Unfortunately, these are often Results unrecognized pathogens due to the lack and difficulties in their identification and therefore, little information is Identification at species level of the streptococcal available on their antimicrobial susceptibility. isolates from the pus samples indicated that 85 isolates belonged to mitis group. Of these, 73 were S. oralis, In this study the sensitivity of the isolates was six were S. mitis, five S. sanguis and one isolate was determined by the E test which has been evaluated S. gordonii. Correlating the results of pus smears as an easy, accurate and standardized antimicrobial microscopy with those of primary cultures, S. oralis was susceptibility method6-8. Although most isolates were the only bacteria found to be involved in causing infection in sensitive to the beta-lactam antibiotics tested, 21, 7 seven abscess cases. The MICs values of antibiotics tested and 2 per cent isolates were intermediate susceptible and the distribution of the isolates of each species within the to penicillin, ampicillin and cefotaxime, respectively. three sensitivity groups are presented in Tables I, II. Isolates with reduced susceptibility to penicillin were found in each species, except for S. gordonii The ranges of the minimum inhibitory concentrations represented by a single strain which was susceptible (MICs) were: 0.016-0.75 mg/l for penicillin (78.8% to all seven antimicrobial agents. Low susceptibility susceptible and 21.2% intermediate susceptible isolates), to ampicillin, eventually associated with reduced 0.016-2 mg/l for ampicillin (93% susceptible and 7% susceptibility to cefotaxime, was found only among intermediate susceptible isolates), 0.016-1 mg/l for S. oralis and S. mitis isolates. In contrast, another cefotaxime (98% susceptible and 2% intermediate study reported more penicillin-resistant isolates among susceptible isolates), 0.016-4 for erythromycin mg/l S. sanguis from acute suppurative oral infections4. (90.6% susceptible, 3.5% intermediate susceptible and Highly penicillin-and cephalosporin-resistant S. mitis, 5.9% resistant isolates), 0.016-0.047 mg/l for clindamycin S. oralis and S. sanguis isolates have been found in (100% susceptible isolates), 0.5-4 mg/l for different countries7,9-11. BANCESCU et al: SUSCEPTIBILITY TESTING OF S. MITIS GROUP 259

Table I. Minimum inhibitory concentrations (MICs) of antibiotics tested against the Streptococcus mitis group isolates Antibiotic/species (no. of isolates) MIC (mg/l)

Range 50%* 90%** Penicillin / S. oralis (73) 0.016-0.75 0.064 0.25 S. mitis (6) 0.047-0.75 0.187 0.75 S. sanguis (5) 0.047-0.25 0.094 0.25 S. gordonii (1) 0.047-0.047 Ampicillin / S. oralis (73) 0.016-2 0.064 0.25 S. mitis (6) 0.047-2 0.22 1.125 S. sanguis (5) 0.047-0.25 0.064 0.125 S. gordonii (1) 0.023-0.023 Cefotaxime / S. oralis (73) 0.016-1 0.047 0.25 S. mitis (6) 0.032-1 0.297 0.75 S. sanguis (5) 0.032-0.25 0.064 0.125 S. gordonii (1) 0.023-0.023 Erythromycin / S. oralis (73) 0.016-3 0.016 0.25 S. mitis (6) 0.032-4 0.063 2.047 S. sanguis (5) 0.032-2 0.094 0.094 S. gordonii (1) 0.016-0.016 Clindamycin / S. oralis (73) 0.016-0.047 0.047 0.047 S. mitis (6) 0.047-0.047 0.047 0.047 S. sanguis (5) 0.023-0.047 0.047 0.047 S. gordonii (1) 0.016-0.016 Chloramphenicol / S. oralis (73) 0.500-4 1 2 S. mitis (6) 0.750-2 1.5 2 S. sanguis (5) 0.500-4 1 1.5 S. gordonii (1) 0.190-0.190 Tetracycline / S. oralis (73) 0.047-256 4 48 S. mitis (6) 0.125-48 2.19 32 S. sanguis (5) 0.125-8 0.38 1 S. gordonii (1) 0.064-0.064

*50% and **90%, concentrations inhibiting 50% and 90% of the isolates, respectively

Although erythromycin is considered the drug of 17-55%) to erythromycin for S. mitis, S. oralis and choice for patients known to be hypersensitive to S. sanguis isolates mostly obtained from pus samples penicillin4,8, the 5.9 per cent resistant and 3.5 per cent from oral and maxillofacial infections or in blood intermediate sensitive isolates confirmed that the cultures. In contrast, resistance to clindamycin and susceptibility to this macrolide could not be predicted chloramphenicol has been reported only occasionally before in vitro testing. Several published reports10-14 among oral streptococci isolates7,10,11,13. In the have indicated much higher resistance rate (between present study, all isolates were sensitive to both of 260 INDIAN J MED RES (SUPPL) MAY 2004

Table II. Distribution of Streptococcus mitis group isolates within the three antimicrobial sensitivity groups

Antibiotic/species (no. of isolates) No. of isolates

Susceptible Intermediate susceptible Resistant

Penicillin / S. oralis (73) 60 13 0 S. mitis (6) 3 3 0 S. sanguis (5) 3 2 0 S. gordonii (1) 1 0 0 Ampicillin S. oralis (73) 68 5 0 S. mitis (6) 5 1 0 S. sanguis (5) 5 0 0 S. gordonii (1) 1 0 0 Cefotaxime / S. oralis (73) 72 1 0 S. mitis (6) 5 1 0 S. sanguis (5) 5 0 0 S. gordonii (1) 1 0 0 Erythromycin / S. oralis (73) 67 3 3 S. mitis (6) 5 0 1 S. sanguis (5) 4 0 1 S. gordonii (1) 1 0 0 Clindamycin / S. oralis (73) 73 0 0 S. mitis (6) 6 0 0 S. sanguis (5) 5 0 0 S. gordonii (1) 1 0 0 Chloramphenicol / S. oralis (73) 73 0 0 S. mitis (6) 6 0 0 S. sanguis (5) 5 0 0 S. gordonii (1) 1 0 0 Tetracycline S. oralis (73) 33 8 32 S. mitis (6) 3 1 2 S. sanguis (5) 4 0 1 S. gordonii (1) 1 0 0 these drugs. The data suggested that clindamycin and In the present study, the resistance rate to chloramphenicol were suitable alternative tetracycline reached more than 40 per cent and that of antimicrobial agents in treatment of suppurative oral intermediate susceptibility more than 10 per cent. The and maxillofacial infections when penicillin-resistant same high frequency of tetracycline-resistance among bacteria were involved or in case the patient was mitis group isolates has been found in other studies allergic to penicillin. also7,11. The results indicate that this drug is clearly not BANCESCU et al: SUSCEPTIBILITY TESTING OF S. MITIS GROUP 261 recommended in the therapy of pyogenic oral and 6. Knapp CC, Washington JA. Comparison of the E test and maxillofacial diseases in which oral streptococci are microdilution for detection of ß-lactam-resistant mutants streptococci that are stably derepressed for type 1 ß-lactamase. involved. J Clin Microbiol 1992; 30 : 214-5. 7. Doern GV, Ferraro MJ, Brueggemann AB, Ruoff KL. Emergence In conclusion, the present findings indicated that the of high rates of antimicrobial resistance among viridans group susceptibility of oral streptococci isolates of clinical streptococci in the United States. Antimicrob Agents Chemother importance should be periodically tested, particularly to 1996; 40 : 891-4. the widely-used antibiotics, since there is considerable 8. Renneberg J, Niemann LL, Gutschik E. Antimicrobial susceptibility of 278 streptococcal blood isolates to seven concern about the increase in the frequency of penicillin- antimicrobial agents. J Antimicrob Chemother 1997; 39 : and multi-resistant strains. 135-40. 9. Konig A, Reinert RR, Hakenbeck R. Streptococcus mitis with Acknowledgment unusually high level resistance to beta-lactam antibiotics. Microb Drug Resist 1998; 4 : 45-9. The authors thank their colleagues from the Oral and Maxillofacial 10. Teng LJ, Hsueh PR, Chen YC, Ho SW, Luh KT. Antimicrobial Surgery Clinic, Bucharest, for their help in collecting samples. susceptibility of viridans group streptococci in Taiwan with an emphasis on the high rates of resistance to penicillin and References macrolides in Streptococcus oralis. J Antimicrob Chemother 1998; 41 : 621-7.

1. Ghitescu I. Antibiotherapy in dentistry. Bucharest: Cerma, 1997. 11. Wisplinghoff H, Reinert RR, Cornely O, Seifert H. Molecular relationships and antimicrobial susceptibilities of viridans group 2. Lewis MAO, Meechan CM, MacFarlane TW, Lamey PJ, Kay streptococci isolated from blood of neutropenic cancer patients. EJ. Presentation and antimicrobial treatment of acute oro-facial J Clin Microbiol 1999; 37 : 1876-80. infections in general dental practice. Br Dent J 1989; 166 : 41-5. 12. Alcaide F, Carrantala J, Linares J, Gudiol F, Martin R. In vitro activities of eight macrolide antibiotics and RP-59500 3. Lewis MAO, MacFarlane TW, McGowan DA. A microbiology (quinupristin-dalfopristin) against viridans group and clinical review of the acute dentoalveolar abscess. Br J Oral streptococci isolated from blood of neutropenic cancer Maxillofacial Surg 1990; 28 : 359-66. patients. Antimicrob Agents Chemother 1996; 40 : 4. Lewis MAO, Parkhurst CL, Douglas CWI, Martin MV, Absi 2117-20. EG, Bishop PA, et al. Prevalence of penicillin resistant bacteria 13. Traub WH, Leonhard B. Antibiotic susceptibility of alpha- and in acute suppurative oral infection. J Antimicrob Chemother nonhemolytic streptococci from patients and healthy adults to 1995; 35 : 785-91. 24 antimicrobial drugs. Chemother 1997; 43 : 123-31. 5. National Committee for Clinical Laboratory Standards. 14. Ono T, Shiota S, Hirota K, Nemota K, Tsuchiya T, Miyake Y. Performance standards for antimicrobial susceptibility testing; Susceptibilities of oral and nasal isolates of Streptococcus mitis and Ninth informational supplement: Approved Standard Streptococcus oralis to macrolides and PCR detection of resistance M100-S8. Wayne PA: NCCLS; 1999 p. 90-3. genes. Antimicrob Agents Chemother 2000; 44 : 1078-80.

Reprint requests: Dr G. Bancescu, Chair of Microbiology, University of Medicine & Pharmacy, "Carol Davila" Bucharest, Romania e-mail: [email protected]