Mutational Analysis of the Group a Streptococcal Operon Encoding Streptolysin S and Its Virulence Role in Invasive Infection
Total Page:16
File Type:pdf, Size:1020Kb
Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XBlackwell Publishing Ltd, 2005? 2005563681695Original ArticleSLS operon and invasive GAS infectionV. Datta et al. Molecular Microbiology (2005) 56(3), 681–695 doi:10.1111/j.1365-2958.2005.04583.x Mutational analysis of the group A streptococcal operon encoding streptolysin S and its virulence role in invasive infection Vivekanand Datta,1 Sandra M. Myskowski,1 Introduction Laura A. Kwinn,1 Daniel N. Chiem,1 Nissi Varki,2 Group A Streptococcus (GAS) is a leading human patho- Rita G. Kansal,3 Malak Kotb3 and Victor Nizet1* gen causing common infections such as pharyngitis 1Department of Pediatrics, Division of Infectious Diseases (‘strep throat’) and impetigo (Cunningham, 2000; Bisno and 2Department of Medicine, Glycobiology Research et al., 2003). During the last three decades, a resur- and Training Center, University of California, San Diego, gence of severe invasive GAS infection has been docu- La Jolla, CA, USA. mented worldwide (Efstratiou, 2000). Prominent among 3Veterans Affairs Medical Center, Research Service, invasive GAS syndromes are the destructive soft tissue Memphis TN, USA. infection necrotizing fasciitis (NF) and the multisystem disorder of streptococcal toxic shock syndrome (STSS), Summary each carrying significant risk of morbidity and mortality even with aggressive medical therapy (Stevens, 1999; The pathogen group A Streptococcus (GAS) pro- Sharkawy et al., 2002). While GAS strains of many M duces a wide spectrum of infections including necro- protein (emm) genotypes are capable of producing sig- tizing fasciitis (NF). Streptolysin S (SLS) produces the nificant disease, strains representing one globally dis- hallmark b-haemolytic phenotype produced by GAS. seminated clonal M1T1 GAS strain have persisted for The nine-gene GAS locus (sagA–sagI) resembling a over 20 years as the single most prevalent isolate from bacteriocin biosynthetic operon is necessary and invasive GAS infections (Cockerill et al., 1997; Cleary sufficient for SLS production. Using precise, in- et al., 1998; Murono et al., 1999; Chatellier et al., 2000), frame allelic exchange mutagenesis and single-gene including all nine surveillance centres of the United complementation, we show sagA, sagB, sagC, sagD, States Centers for Disease Control Emerging Infections sagE, sagF and sagG are each individually required Program Network in 2002 (http://www.cdc.gov/ncidod/ for SLS production, and that sagE may further serve dbmd/abcs). an immunity function. Limited site-directed muta- A hallmark phenotypic feature of GAS is the distinct genesis of specific amino acids in the SagA pre- zone of b-haemolysis surrounding colonies grown on propeptide supports the designation of SLS as a blood agar media. This phenomenon reflects complete bacteriocin-like toxin. No significant pleotrophic lysis of red blood cells produced by the potent oxygen- effects of sagA deletion were observed on M protein, stable cytolysin known as streptolysin S (SLS). SLS exists capsule or cysteine protease production. In a murine primarily in cell-bound form (Ginsburg, 1999) and is deliv- model of NF, the SLS-negative M1T1 GAS mutant was ered most effectively to target cells by direct contact with markedly diminished in its ability to produce necrotic GAS (Ofek et al., 1990). The cytolytic spectrum of SLS is skin ulcers and spread to the systemic circulation. broad including lymphocytes, neutrophils, platelets, The SLS toxin impaired phagocytic clearance and cancer cell lines and subcellular organelles (Keiser et al., promoted epithelial cell cytotoxicity, the latter pheno- 1964; Taketo and Taketo, 1966; Ginsburg, 1972; type being enhanced by the effects of M protein and Hryniewicz and Pryjma, 1977). Insertion of SLS into the streptolysin O. We conclude that all genetic compo- cell membrane results in the formation of transmembrane nents of the sag operon are required for expression pores and osmotic cell lysis, similar to that observed with of functional SLS, an important virulence factor in the complement-mediated cytotoxicity (Ginsburg, 1999; Carr pathogenesis of invasive M1T1 GAS infection. et al., 2001). The GAS chromosomal locus for SLS production was Accepted 14 January, 2005. *For correspondence. E-mail first identified by J. DeAzavedo and colleagues through [email protected]; Tel. (+1) 858 534 7408; Fax (+1) 858 534 5611. generation and analysis of SLS-deficient transposon © 2005 Blackwell Publishing Ltd 682 V. Datta et al. mutants (Borgia et al., 1997; Betschel et al., 1998). Sub- Results sequent chromosome walking studies performed by B. Requirement of individual sag locus genes in SLS Beall, and informed by the first GAS genome project production (Ferretti et al., 2001), recognized the existence of a nine- gene cluster (sagA–I for streptolysin-associated genes) in Earlier studies establishing the role of the sag locus in the region of transposon insertions. The functional bound- SLS production employed transposon mutants (Betschel aries of the sag locus were then defined by plasmid et al., 1998; Li et al., 1999) or insertional inactivation integrational mutagenesis and its organization as an mutants (Nizet et al., 2000; Biswas et al., 2001) likely to operon confirmed by RNA analysis (Nizet et al., 2000). exert polar effects on transcription of downstream genes Heterologous expression of the entire GAS sag locus in in the operon. To assess the specific requirement of indi- a non-haemolytic strain of Lactococcus lactis yielded vidual sag genes in SLS production, we combined precise robust b-haemolytic transformants, confirming that the in-frame allelic exchange mutagenesis with single-gene operon is both necessary and sufficient for SLS produc- complementation analysis. Bacterial strains used in this tion (Nizet et al., 2000). Individual gene homologies and study are listed in Table 1. GAS strain 5448 is a serotype structural features of the operon suggested that SLS is a M1T1 isolate from a patient with necrotizing fasciitis and bacteriocin-like peptide toxin, with structural gene sagA streptococcal toxic shock syndrome expressing SLS and encoding a 53-amino-acid prepropeptide precursor of the streptolysin O (SLO), cysteine protease, superantigens mature toxin (Nizet et al., 2000). This hypothesis was cor- SpeA, SpeF, SpeG and SmeZ and is genetically repre- roborated by the research group of B. Kreikemeyer (Carr sentative of the globally disseminated M1T1 clone that is et al., 2001), and later others (Dale et al., 2002), who the leading cause of invasive GAS infections (Kansal demonstrated that antibodies generated against synthetic et al., 2000). GAS strain NZ131, a serotype M49T14 skin peptides corresponding to the predicted SagA propeptide isolate from a patient with glomerulonephritis, expresses neutralized SLS activity. Highly homologous sag operons SLS and SLO, and is frequently used in genetic studies producing SLS toxins are responsible for the b-haemolytic because of its increased transformability (Simon and phenotypes of human isolates of groups C and G strep- Ferretti, 1991). Precise in-frame allelic replacement of the tococci (GCS and GGS) and the zoonotic pathogen sagA, sagB, sagC, sagD, sagE, sagF or sagG genes in Streptococcus iniae (Fuller et al., 2002; Humar et al., M49 GAS strain NZ131 produced a completely non- 2002). haemolytic phenotype (Fig. 1). Allelic replacement of Streptolysin S contributes significantly to GAS virulence sagA in the M1 GAS strain 5448 also eliminated SLS potential in animal models of infection (Betschel et al., production. Complementation experiments were per- 1998; Humar et al., 2002; Fontaine et al., 2003; Sierig formed to reintroduce the corresponding single gene to et al., 2003; Engleberg et al., 2004). Like the genetically each allelic exchange mutant. Full complementation was distinct b-haemolysin of the human pathogen group B noted for sagA, sagC and sagF, while partial complemen- Streptococcus (GBS), SLS could theoretically enhance tation was seen with sagB, sagD and sagG (Fig. 1). These GBS pathogenicity by causing direct tissue injury, promot- experiments demonstrate a unique requirement for each ing cellular invasion, impairing phagocytic clearance or of these six sag operon genes in production of functional additional yet to be defined mechanisms (Nizet, 2002). SLS toxin. We speculate that partial complementation in However, the analysis of SLS-associated virulence phe- the case of the individual genes mentioned could reflect notypes has been complicated by reports that various toxicity of overexpression (note SagB and SagD are pre- mutations in the region of the sagA gene could be dicted cytoplasmic proteins) or upsetting the stoichiome- associated with alterations in expression of other GAS try of complex assembly (SagG is predicted to participate virulence genes including M protein, cysteine protease with SagH and SagI in forming an ABC-type transporter). and streptokinase (Li et al., 1999; Biswas et al., 2001; Mangold et al., 2004). The sagE gene appears to encode an immunity function In the present study, we employ precise in-frame allelic replacement, single-gene complementation and limited Multiple initial attempts to complement the sagE allelic site-directed mutagenesis to test the specific requirement exchange mutant were unsuccessful. Because the pre- of individual GAS sag locus genes for the production of dicted gene product SagE shares homology to a bacteri- SLS. Using a precise allelic replacement DsagA mutant ocin immunity protein of Lactobacillus planterum,