Differentiation of pneumoniae and K. oxytoca by Multiplex Polymerase Chain Reaction Yogesh Chander1

M. A. Ramakrishnan1

Naresh Jindal1

Kevan Hanson2

Sagar M. Goyal1

1Department of Veterinary Population Medicine, College of Veterinary Medicine, University of Minnesota, Saint Paul, MN 55108

2Microbiology Clinical Laboratory, Hennepin County Medical Center, 701 Park Avenue, Minneapolis, MN 55415

KEY WORDS: Clinical isolates, INTRODUCTION Identification,K. pneumoniae, K. oxytoca, Members of genus Klebsiella are opportu- PCR. nistic pathogens that are frequently isolated ABSTRACT from various infections in humans and ani- 1 To differentiate Klebsiella pneumoniae from mals. In humans, Klebsiella spp in general, K. oxytoca isolated from clinical samples, a and Klebsiella pneumoniae and Klebsiella multiplex polymerase chain reaction (PCR) oxytoca in particular, are involved in noso- 1,2 was developed using specific prim- comial infections. In animals, Klebsiella ers. The position of bands on agarose gel, are mostly associated with sepsis, infec- 108 bp for K. pneumoniae and 343 bp for K. tions of urinary and respiratory tracts, and 3-5 oxytoca, confirmed strain identity. A total of mastitis. These disease syndromes cause 331 Klebsiella isolates were tested by this serious economic consequences in some 6 method, of which 219 (66%) were identified cattle herds. In fact, clinical mastitis due to as K. pneumoniae and 52 (16%) as K. oxyto- Klebsiella infections results in higher milk ca. Isolates, which tested negative with this losses than those due to Escherichia coli and PCR (n=60) were retested using Klebsiella may also result in the death of the affected 7 genus specific primers and were confirmed cows. Both K. pneumoniae and K. oxy- to be Klebsiella sp. Results obtained suggest toca are frequently isolated from domestic that this newly developed PCR can be used livestock. as a diagnostic tool for correct identification Phenotypic distinction between K. pneu- and differentiation of K. pneumoniae and K. moniae and K. oxytoca based on existing oxytoca, two clinically important Klebsiella biochemical tests is not very reliable, and species. is time consuming and laborious. Current

138 Vol. 9, No. 2, 2011 • Intern J Appl Res Vet Med. Table 1. List of primers used in this study.

Primer Sequence (5’-3’) Target gene Product size Reference name (bp) KP (F) CAA CGG TGT GGT TAC TGA CG rpo B 108 This study (R) TCT ACG AAG TGG CCG TTT TC KO (F) GAT ACG GAG TAT GCC TTT ACG GTG peh X 343 Kovtunovych et al. 2003 (R) TAG CCT TTA TCA AGC GGA TAC TGG Kleb. (F) CGC GTA CTA TAC GCC ATG AAC GTA gyr A 441 Brisse and Verhoef, 2001 (R) ACC GTT GAT CAC TTC GGT CAG G procedures such as Biolog and API systems Diagnostic Laboratory (MVDL), University often fail to differentiate between species of of Minnesota, Saint Paul, MN. Using the Klebsiella.8,9 Even the new CHROM Orien- API 20E identification system (BioMereiux, tation medium does not discriminate proper- France), 52 isolates were identified asK. ly between K. oxytoca and K. pneumoniae.10 pneumoniae, 19 as K. oxytoca and 33 as These difficulties often lead to mistaken Klebsiella sp. In addition, 115 Klebsiella identity between K. pneumoniae and K. oxy- isolates were obtained from the Mastitis toca.8,11 To understand the epidemiology of Laboratory at the College of Veterinary these and to control their spread, it Medicine, St Paul, MN. These isolates were is imperative to have easy means of discrim- identified simply asKlebsiella sp by colony ination between these two important species morphology on McConkey agar, motility of Klebsiella. Molecular methods such as test, and API 20E identification system (Bio- polymerase chain reaction (PCR) have been Mereiux, France). A total of 112 Klebsiella evaluated for the rapid identification of isolates from human clinical samples were Klebsiella in human clinical specimens.2,11 obtained from the Hennepin County Medical Kovtunovych et al11 developed a PCR assay Center, Minneapolis, MN. Using the Vitek 2 targeting the polygalacturonase gene that (BioMerieux, France) automated identifica- could discriminate K. oxytoca from other tion system and a set of biochemical tests, species of Klebsiella, and suggested its use 107 were identified asK pneumoniae and 5 for clinical and ecological monitoring of K. as K. oxytoca. oxytoca. Primers and PCR Amplification Neuberger et al2 reported the develop- DNA was isolated using PrepMan® reagent ment of a PCR assay that identified genetic (Applied Biosystems, Foster City, CA) ac- sequence of hemolysin gene of K. pneu- cording to the manufacturer’s instructions. moniae after an initial growth in blood Species specific primers were used for the cultures. These last two studies focused amplification ofK. pneumoniae and K. primarily on the identification of eitherK. oxytoca in a single reaction mixture (Table pneumoniae or K. oxytoca. In the present 1). PCR was performed using HotStarTaq™ study, we have developed a multiplex PCR DNA Polymerase kit (Qiagen, Valencia, CA) method to differentiate K. pneumoniae from and the reaction mixture consisted of: K. oxytoca isolated from clinical samples. 25 mL of master mix, 200 nM of each primer MATERIALS AND METHODS (forward and reverse), 100 ng of template, Bacteria and water to make final volume of 50m L. A total of 331 Klebsiella isolates were used Reaction conditions for PCR were: initial in this study (Table 1). Klebsiella isolates denaturation at 95°C for 15 min, 35 cycles of animal origin (n=104) were obtained of denaturation at 95°C for 1 min, annealing from the archives of Minnesota Veterinary at 55°C for 1 min, extension at 72°C for 2 min, and final extension at 72°C for 10 min.

Intern J Appl Res Vet Med • Vol. 9, No. 2, 2011. 139 Table 2. Speciation of Klebsiella isolated from different sources.

Sample origin Identification Identification by Number (%) identified as indicated species by PCR biochemical test K. pneumoniae K. oxytoca Untypeable* K. pneumoniae Animal clinical samples (n=52) 36 (69%) 0 16 (31%) (n=159) Human clinical samples (n=107) 88 (82%) 5 (5%) 14 (13%) K. oxytoca (n=24) Animal clinical samples (n=19) 0 17 (89%) 2 (11%) Human clinical samples (n=5) 0 4 (80%) 1 (20%) Klebsiella sp. Animal clinical samples (n=33) 19 (58%) 11 (33%) 3 (9%) (n=148) Mastitis samples (n=115) 76 (66%) 15 (13%) 24 (21%) *These isolates were tested with genus specific primers and were found to be Klebsiella sp

The PCR products were visualized on 1.5% 5 (3%) as K. oxytoca. agarose gel in Tris-acetate-EDTA buffer by The remaining 30 (19%) isolates could electrophoresis. not be identified with eitherK. pneumoniae A 100 bp DNA ladder was used as mo- or K. oxytoca specific primers and were lecular weight marker. The position of bands classified as untypeable (Table 2). Of the on agarose gel, 108 bp for K pneumoniae 24 K. oxytoca isolates (19 of animal origin and 343 bp for K oxytoca, confirmed strain and 5 of human origin), 21 (88%) were identity. PCR products were purified using identified asK. oxytoca while the remain- QIAquick PCR purification kit (Qiagen, ing three (12%) tested negative for both K. Valencia, CA) and sequenced both direc- pneumoniae and K. oxytoca (Table 2). Of tions with the same primers as used in PCR the 148 isolates identified only asKlebsiella reactions. Sequences obtained were matched sp. (33 from animal clinical samples and with online database (www.ncbi.nlm.nih. 115 from mastitis samples), 95 (64%) were gov) using BLAST. Isolates that tested nega- identified asK. pneumoniae and 26 (18%) as tive for either K. pneumoniae or K. oxytoca K. oxytoca by PCR. The remaining 27 (18%) by the newly developed PCR were retested isolates could not be identified as eitherK. with Klebsiella genus specific primers tar- pneumoniae or K. oxytoca (Table 2). geting the gyrA gene (Kleb F and R, Table All 60 isolates that could not be identi- 1) under the same reaction conditions as for fied as eitherK. pneumoniae or K. oxytoca PCR with species specific primers. The PCR by this multiplex PCR, were identified products were purified and sequenced. as Klebsiella sp when tested with genus RESULTS specific primers. For all these isolates, A total of 331 Klebsiella isolates were amplification products of expected size were screened using the newly developed PCR. obtained by PCR and all amplicons were se- To confirm primer specificity, PCR was car- quenced. Sequence analysis by BLAST did ried out on Salmonella typhimurium (ATCC confirm all of these isolates to beKlebsiella 14028), E. coli (ATCC 35248), Yersinia en- sp, but they could not be further speciated. terocolitica (ATCC 9610), Shigella flexneri All sequences were submitted to the Gen- (ATCC 12022), E. coli (ATCC 25922), K. Bank under accession numbers HM452905- pneumoniae (ATCC 13883), and K. oxytoca HM452944. (ATCC 8724). Only K. pneumoniae and K. DISCUSSION oxytoca strains were amplified. Of the 159 For the timely treatment of infections, it is K. pneumoniae isolates (52 of animal origin important that diagnostic tests used for the and 107 of human origin) screened, PCR identification of infectious microorganisms identified 124 (78%) asK. pneumoniae and are rapid, specific, and sensitive. Members

140 Vol. 9, No. 2, 2011 • Intern J Appl Res Vet Med. of the genus Klebsiella are opportunistic Of the 52 clinical isolates of animal pathogens which are difficult to identify and origin which were identified biochemically are often misclassified in clinical microbiol- as K. pneumoniae, 69% were confirmed ogy laboratories.9,12,13 Current biochemical as K. pneumoniae by PCR and, 89% of 19 methods of identification are time consum- K. oxytoca isolates were confirmed asK. ing and are often inconclusive because oxytoca. Of the 107 K. pneumoniae isolates related species eg, K. pneumoniae and K. from human clinical samples, 82% were oxytoca or even Enterobacter aerogenes, identified asK. pneumoniae and 80% of five often present similar biochemical patterns.14 isolates of K. oxytoca were correctly identi- Correct identification ofKlebsiella isolates fied. This could be because all the isolates is important for taxonomic and molecular used in this study were identified earlier by characterization.12 Molecular methods based either biochemical tests or automated iden- on amplification and sequencing ofrpoB, tification systems, and these methods are gyrA, parC genes, and 16S rRNA region not very accurate for differentiation among have recently become available for the Klebsiella species.8 Monnet et al8 reported identification ofKlebsiella sp.14-16 but there that up to 13% of Klebsiella isolates are mis- is no PCR based method for the identifica- identified using biochemical methods. These tion/differentiation of K. pneumoniae and K. results further indicate the limitations of the oxytoca. This study was designed to fill that automated identification systems in com- gap. parison to PCR. All isolates that could not The newly developed multiplex PCR be differentiated by species-specific prim- was specific for K. pneumoniae and K. ers were tested with Genus-specific primers oxytoca since no amplification was observed targeting the gyrA gene of Klebsiella. All of with other Gram negative bacteria tested these isolates were confirmed to beKlebsi - eg, S. typhimurium, E. coli, Y. enteroco- ella sp but could not be speciated. However, litica, and S. flexneri. The primers used in for most of the isolates, BLAST analysis this study yielded products of 108 bp for indicated a match with both K. variicola and K. pneumoniae and 343 bp for K. oxytoca, K. pneumoniae. This is probably because of enabling clear differentiation of these two very little difference in the sequence of this species. With this PCR, 78% of the isolates gene from these two isolates. that were identified asK. pneumoniae by Identification of any bacteria up to the spe- biochemical tests were identified asK. cies level is very important in clinical micro- pneumoniae, and 88% of K. oxytoca isolates biology so as to determine the association of could be confirmed asK oxytoca. Thus, 22% a particular species with a disease condition. of K. pneumoniae and 12% of K. oxytoca Though the present study helped in identifi- isolates were wrongly identified in bio- cation and differentiation of only two Kleb- chemical tests. This is not surprising because siella species, further studies are required R. planticola and R. terrigena are often mis- to differentiate other species of Klebsiella. diagnosed as K. pneumoniae or K. oxytoca The present study shows that PCR can be a by biochemical tests.9 Others have shown useful tool in diagnostic laboratories for cor- that isolates identified asK. pneumoniae by rect identification and differentiation ofK. biochemical tests can sometimes be R. plan- pneumoniae and K. oxytoca as this method ticola,8, 17 R. terrigena,8 or K. variicola.3,18 is rapid and specific for these two clinically Similarly, isolates classified asK. oxytoca important Klebsiella species. 8,17 19 are often R. planticola. Munoz et al REFERENCES reported that several isolates from mastitis 1. Podschun R, Ullmann U. Klebsiella spp. as nosoco- cases that were identified asK. pneumoniae, mial pathogens: epidemiology, ,typing were later identified asRaoultella sp. (ear- methods, and pathogenicity factors. Clin Microbiol lier named as R. planticola and R. terrigena) Rev. 1998;11:589-603. upon rpoB gene sequencing. 2. Neuberger A, Oren I, Sprecher H. Clinical impact of

Intern J Appl Res Vet Med • Vol. 9, No. 2, 2011. 141 a PCR assay for rapid identification ofKlebsiella 12. Alves MS, Dias RCS, Castro ACD, et al. Identifi- pneumoniae in blood cultures. J Clin Microbiol. cation of clinical isolates of indole-positive and 2008;46:377-379. indole-negative Klebsiella spp. J Clin Microbiol. 3. Brisse S, van Duijkeren E. Identification and anti- 2006;44:3640-3646. microbial susceptibility of 100 Klebsiella animal 13. Hansen DS, Aucken HM, Abiola T, et al. Recom- clinical isolates. Vet Microbiol. 2005;105:307-312. mended test panel for differentiation of Klebsiella 4. Kikuchi N, Kagota C, Nomura T, Hiramune T, Taka- species on the basis of a trilateral interlaboratory hashi T, Yanagawa R. Plasmid profiles ofKlebsi - evaluation of 18 biochemical tests. J Clin Micro- ella pneumoniae isolated from bovine mastitis. Vet biol. 2004;42:3665-3669. Microbiol 1995;47:9-15. 14. Lopes AC, Rodrigues JF, Clementino MB, et al. 5. Wilson DJ, Gonzalez RN, Das HH. Bovine mastitis Application of PCR ribotyping and tDNA-PCR pathogens in New York and Pennsylvania: preva- for Klebsiella pneumoniae identification.Mem Inst lence and effects on somatic cell count and milk Oswaldo Cruz 2007;102:827-832. production. J Dairy Sci. 1997;80:2592-2598. 15. Brisse S, Verhoef J. Phylogenetic diversity of Kleb- 6. Waller KP, Unnerstad H. Klebsiella mastitis: a poten- siella pneumoniae and clinical tial threat to milk production. Svensk Veterinärtid- isolates revealed by randomly amplified polymor- ning 2004;56:11-17. phic DNA, gyrA and parC gene sequencing and automated ribotyping. Int J Syst Evol Microbiol. 7. Grohn YT, Wilson DJ, Gonzalez RN, et al. Effect of 2001;51:915-924. pathogen-specific clinical mastitis on milk yield in dairy cows. J Dairy Sci. 2004;87:3358-3374. 16. Drancourt M, Bollet C, Carta A, et al. Phylogenetic analyses of Klebsiella species delineate Klebsi- 8. Monnet D, Freney J, Brun, Y., et al. Difficulties in ella and Raoultella gen. nov., with description of identifying Klebsiella strains of clinical origin. Raoultella ornithinolytica comb. nov., Raoultella Zentralbl Bakteriol 1991;274:456-464. terrigena comb. nov. and 9. Monnet D, Freney J. Method for differentiating comb. nov. Int J Syst Evol Microbiol. 2001;51:925- Klebsiella planticola and Klebsiella terrigena 932. from other Klebsiella species. J Clin Microbiol. 17. Westbrook GL, O’Hara CM, Roman SB, et al. Inci- 1994;32:1121-1122. dence and identification ofKlebsiella planticola in 10. Ohkusu K. Cost-effective and rapid presumptive clinical isolates with emphasis on newborns. J Clin identification of gram-negative bacilli in routine Microbiol. 2000;38:1495-1497. urine, pus and stool culture: Evaluation of the use 18. Rosenblueth M, Martínez L, Silva J, et al. Kleb- of CHROMagar orientation medium in conjunc- siella variicola, a novel species with clinical and tion with biochemical tests. J Clin Microbiol. plant-associated isolates. Syst Appl Microbiol. 2000;38:4586-4592. 2004;27:27-35. 11. Kovtunovych G, Lytvynenko T, Negrutska V, et al. 19. Munoz MA, Welcome FL, Schukken YH, et al. Identification ofKlebsiella oxytoca using a specific Molecular epidemiology of two Klebsiella pneu- PCR assay targeting the polygalacturonase pehX moniae mastitis outbreaks on a dairy farm in New gene. Res Microbiol. 2003;154:587-592. York state. J Clin Microbiol. 2007;45:3964-3971.

142 Vol. 9, No. 2, 2011 • Intern J Appl Res Vet Med.