- Email: [email protected]
Specificity of enterobacterial repetitive intergenic consensus and repetitive extragenic palindromic polymerase chain reaction for the detection of clonality within the Enterobacter cloacae complex
Diagnostic Microbiology and Infectious Disease 53 (2005) 9 – 16 www.elsevier.com/locate/diagmicrobio
Specificity of enterobacterial repetitive intergenic consensus and repetitive extragenic palindromic polymerase chain reaction for the detection of clonality within the Enterobacter cloacae complex Anita N. Stumpf a, Andreas Roggenkampa, Harald Hoffmannb,T a
Max von Pettenkofer Institute of Hygiene and Medical Microbiology, Ludwig Maximilian University Munich, Klinikum Groghadern, 81377 Munich, Germany b Institute of Microbiology and Laboratory Medicine at the Pneumological Hospitals Gauting, D-82131 Gauting, Germany Received 25 August 2004
Abstract An increasing number of clonal outbreaks caused by members of the E. cloacae complex is being reported. For the detection of clonality, pulsed-field gel electrophoresis (PFGE) is considered the golden standard, but PCR-based methods are cheaper, easier to perform, and provide faster results. One hundred ninety-five isolates of the E. cloacae complex isolated at the university hospital Groghadern, Munich, Germany, were assigned to their respective genetic cluster by partial hsp60 sequencing. All study isolates belonging to genetic clusters III and VI were selected to evaluate the specificity of the enterobacterial repetitive intergenic consensus (ERIC) and repetitive extragenic palindromic (REP) polymerase chain reaction (PCR) for the identification of clonal isolates belonging to the E. cloacae complex. For these 56 isolates, PFGE was performed, yielding 3 pairs of isolates with indistinguishable patterns. ERIC PCR resulted in 7 groups with identical patterns, together encompassing 49 study isolates. Comparing the ERIC PCR with the PFGE, a specificity of 14% considering the detection of bclonalQ isolates was calculated. In this respect, REP PCR performed much better, yielding a specificity of 90%. An unweighted pairgroup method with arithmetic averages tree based on ERIC PCR patterns allowed an accurate classification of the isolates to the respective genovars, suggesting that the ERIC PCR differentiates between genovars rather than between strains. In contrast, REP PCR differentiates better on the strain level. A proposed diagnostic system for the detection of subsumed outbreak strains of the E. cloacae complex is presented. It is based on an initial REP PCR, which should be confirmed by PFGE in cases of identical patterns, whereas ERIC PCR does not seem to be useful for the detection of outbreak strains when dealing with isolates of the E. cloacae complex. D 2005 Published by Elsevier Inc. Keywords: Enterobacter cloacae complex; Fingerprinting; PFGE; ERIC PCR; REP PCR
1. Introduction In a recent population genetic study, the nomenspecies Enterobacter cloacae turned out to represent, in fact, a large complex consisting of at least 12 genetic clusters (Hoffmann and Roggenkamp, 2003). An increasing number of outbreaks caused by the E. cloacae complex have been reported from several hospitals. For example, van Nierop et al. (1998) reported an outbreak in a neonatal intensive care unit with 9 deaths, Kuboyama et al. (2003) reported 3 outbreaks with 42 systemic infections and a crude T Corresponding author. Tel.: +49-89-85791-8230; fax: +49-8985791-8350. E-mail address: [email protected] (H. Hoffmann). 0732-8893/$ – see front matter D 2005 Published by Elsevier Inc. doi:10.1016/j.diagmicrobio.2005.04.003
mortality of 34%, and Fernandez-Baca et al. (2001) observed outbreaks affecting 15 neonates and causing 2 deaths. All of these lead the E. cloacae complex to be a serious hygienic problem. In a situation of questionable outbreaks, hygienic measurements should be initiated immediately. However, hygienic management is both expensive and demanding on manpower. Hence, an early answer is needed whether a series of infections with Enterobacter isolates represents a real outbreak or rather a polyclonal endemic infection due to different isolates of the same species. In the latter situation, changes in antibiotic strategies are certainly more adequate for the control than hygienic measures. Different techniques for the detection of clonality of isolates causing a series of infections are available. Pulsed-
10
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
field gel electrophoresis (PFGE) is considered the golden standard for genetic typing of bacteria (Olive and Bean, 1999; Tenover et al., 1995). It is highly discriminative but also expensive both in time and money. In contrast, polymerase chain reaction (PCR)–based methods are relatively cheap, easy to perform, and provide fast results. Different PCR fingerprint methods have been described, of which only few, such as enterobacterial repetitive intergenic consensus (ERIC) and repetitive extragenic palindromic (REP) PCR have so far been applied to the E. cloacae complex. In recent studies, both ERIC (Fernandez-Baca et al., 2001) and REP PCR (Shi et al., 1996) have been used for genetic differentiation of isolates. Both have been evaluated as highly discriminative methods. However, the studies did not take the genetic heterogeneity of the E. cloacae complex into consideration. Hence, the isolates of the respective studies might have belonged to several genovars of the E. cloacae complex. Then, the ERIC and REP PCRs would have differentiated on the species level rather than on the strain level, and a representative validation of both techniques would still remain elusive. In the present study, we focused on the evaluation of the specificity of ERIC and REP PCRs for the detection of outbreak strains within genetic clusters of the E. cloacae complex. PFGE was performed for each isolate as the golden standard reference method.
2. Materials and methods 2.1. Bacterial test collection One hundred ninety-five isolates were prospectively and randomly collected from clinical specimen at the Klinikum Groghadern, Munich, including only 1 isolate per patient. The collection comprised 55% of all isolates recovered from our diagnostic service from November 2001 to February 2002 and 90% of all blood culture isolates from 1997 to 2002. Isolates were biochemically identified as members of the E. cloacae complex (API 20E, bioMe´rieux, Marcy l’Etoile, France). They were assigned to the respective genetic cluster by partial hsp60 sequencing as previously described (Hoffmann and Roggenkamp, 2003). All isolates of genetic clusters III and VI were selected for further investigation. Cluster III consisted of 44 isolates, encompassing 2 intraabdominal isolates, 18 blood culture isolates, 1 isolate from a central venous line, 2 from pleural specimens, 9 from the respiratory tracts, 2 from skin swabs, 3 from stools, 5 from the urogenital tracts, and 2 from wound swabs. Twelve study isolates belonged to genetic cluster VI, encompassing 5 isolates from blood culture, 5 from the respiratory tract, 1 from a skin swab, and 1 from stool (Table 1). In one patient, 2 blood culture isolates with 3 days’ difference (isolates 268 and 425) were taken as internal control for the detection of clonality by the respective methods used. The author (AS), who performed ERIC and
Table 1 Study isolates and their characteristics Isolate Source of isolation
Genetic Designation of patterns and clustera groups of patterns
5 18 114 116 118 165 167 190 191 198 200 204 210 211 213 216 217 218 232 237 242 244 253 254 255 260 261 266 268 277 412 418 419 422 425 427 434 436 439 440 444 450 452 453 454 456 464 465 535 538 543 570 571 576 594 607
VI VI III III III III III VI III III III III VI III III III III VI VI III III III III III III VI III III III VI VI III III III III III III VI VI III III III III III III III III VI III III III III III III III III
PFGE (XbaI) PFGE (SpeI ) ERIC REP RT Stool RT Stool RT Pleura CVL RT RT RT UGT RT RT UGT RT UGT UGT Skin swab RT Pleura Wound swab Abdomen Skin swab RT UGT RT RT Abdomen BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx Skin swab Stool BCx BCx BCx RT Wound swab Stool
X-35 X-16 X-05 X-49 X-09 X-38 X-39 X-34 X-26 X-04 X-01 X-40 X-36 X-47 X-03 X-41 X-17 X-37 X-25 X-50 X-31 X-02 X-10 X-21 X-07 X-46 X-14 X-30 X-45 X-19 X-27 X-08 X-13 X-32 X-45 X-06 X-18 X-28 X-20 X-42 X-13 X-22 X-33 X-43 X-24 X-23 X-48 X-51 X-11 X-52 X-15 X-29 X-29 X-12 X-44 X-49
– – – S-05 – – – – – – – – – – – – – – – – – – – – – – – – S-03 – – – S-01 – S-03 – – – – – S-01 – – – – – – – – – – S-02 S-02 – – S-04
E-11 E-12 E-04 E-04 E-05 E-05 E-04 E-13 E-05 E-03 E-06 E-08 E-10 E-05 E-06 E-05 E-05 E-10 E-13 E-09 E-01 E-05 E-05 E-05 E-05 E-14 E-09 E-05 E-09 E-10 E-13 E-09 E-09 E-05 E-09 E-09 E-09 E-10 E-10 E-09 E-09 E-09 E-09 E-05 E-09 E-05 E-09 E-11 E-05 E-07 E-05 E-05 E-05 E-02 E-05 E-04
R-47 R-44 R-35 R-50 R-22 R-29 R-32 R-43 R-23 R-33 R-20 R-24 R-38 R-40 R-26 R-15 R-09 R-38 R-42 R-36 R-41 R-14 R-37 R-27 R-34 R-49 R-10 R-08 R-07 R-42 R-45 R-25 R-03 R-28 R-07 R-17 R-06 R-46 R-42 R-04 R-03 R-01 R-16 R-39 R-12 R-05 R-21 R-48 R-18 R-19 R-13 R-11 R-11 R-30 R-31 R-02
RT = respiratory tract; CVL = central venous line; UGT = urogenital tract; BCx = blood culture. Clusters rendered in boldface comprised more than 1 isolate. a Genetic cluster denominations following (Hoffmann and Roggenkamp, 2003).
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
11
Fig. 1. hsp60 neighbor joining tree (left) and UPGMA tree of ERIC PCR patterns (right) of 56 prospectively collected clinical isolates of the E. cloacae complex, including the internal control isolates (268 and 425) recovered from the same patient 3 days apart. Both trees have been rooted with the E. cloacaetype strain (ATCC13047T). The 2 major ERIC clusters i and ii corresponded to the 2 hsp60 sequence clusters analyzed (genetic clusters III and VI, respectively). Within the major ERIC clusters, 14 relatedness groups were observed (E-01 to E-14). TOutbreak causing strains.
REP PCR and PFGE, was not aware of this doubling of an identical strain.
DNA was prepared for PCR by quick-heat lysis. Isolates were grown in Luria-Bertani broth (LB) to an optical density of 1.0 at 600 nm. A 1.5-mL culture was centrifuged for 2 min at 10 000 rpm. The pellet was diluted in 0.5 mL of distilled water, boiled for 10 min, and centrifuged (5 min, 10 000 rpm, 4 8C).
ATC AG-3V). Reactions were carried out in a total volume of 50 AL and started after an activation time of 10 min at 95 8C. Samples were subjected to 30 cycles of 30 s at 95 8C, 30 s at 59 8C, and 1 min of extension at 72 8C. Cycles were followed by a final elongation step of 15 min at 72 8C. PCR products were visualized in 2% agarose gels after electrophoresis and ethidium bromide staining. DNA was cleaned using the NucleoSpin Extraction kit (Macherey-Nagel, Dqren, Germany) for further processing following the manufacturer’s guidelines.
2.3. HSP60 PCR
2.4. HSP60 sequencing
Amplification reactions were performed with an Applied Biosystems GeneAmp PCR system 2700 and Taq Gold DNA hot start polymerase (Perkin-Elmer, Branchburg, NJ) under standard conditions (500 pmol of each primer, 200 Amol/L nucleotide mix, 10 mmol/L Tris-Cl [pH 8.3], 5 mmol/L KCl, 1.5 mmol/L MgCl2). Primers were used as previously described (Hoffmann and Roggenkamp, 2003) (HSP-1218-F: 5V-GGT AGA AGA AGG CGT GGT TGC3V and HSP-1559-R: 5V-ATG CAT TCG GTG GTG ATC
PCR-amplified and cleaned DNA was sequenced using the dye terminator method. Purified DNA (0.2 Ag) was added to a 20-AL reaction solution containing 4 AL of BigDye dye terminator sequencing mix and 0.7 AL of a 5-pmol primer solution. Cycle parameters were 25 cycles with an initial 96 8C denaturation step of 10 s, followed by an annealing step at 55 8C of 10 s, and an extension step of 4 min at 60 8C performed in an Applied Biosystems GeneAmp PCR System 2700. Sequences were
2.2. Isolation of DNA
12
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
determined by electrophoresis with the ABI Prism 377 DNA sequencer. 2.5. ERIC and REP PCR The principles of the ERIC and REP PCR have been described in detail in previous publications (Shi et al., 1996; Versalovic et al., 1991). Primer sequences were as follows: ERIC-1R: 5V-ATG TAA GCT CCT GGG GAT TCA C-3V; ERIC-2: 5V-AAG TAA GTG ACT GGG GTG AGC G-3V; REP-1R-I: 5V-III ICG ICA TCI GGC-3V; REP-2-I: 5V-ICG ICT TAT CIG GCC TAC-3V(Versalovic et al., 1991). Primers were synthesized by Metabion, Munich, Germany. Amplification reactions were carried out in an Applied Biosystems GeneAmp 2700 PCR thermocycler. Each 25-AL PCR reaction contained 50 pmol of each of the 2 opposing primers, 1.25 mmol/L of each of the 4 dNTPs, 2 U AmpliTaq DNA polymerase (Perkin-Elmer), 1.5 mmol/L MgCl2, 5 mmol/L KCl, 10 mmol/L Tris-Cl (pH 8.3), and 1.6 AL DNA. Amplifications were performed with an initial denaturation step at 94 8C for 7 min, followed by 30 cycles of denaturation (90 8C, 30 s), annealing (ERIC: 52 8C/ REP: 40 8C, 1 min), and extension (65 8C, 8 min), completed by a single final extension step (65 8C, 16 min). Twenty microliters of each PCR reaction was electrophoresed on 2% agarose gels in 1 Tris acetate–EDTA buffer at 100 V for 1 h, stained with ethidium bromide, and visualized under UV light. In each run, ERIC and REP PCR were repeated for isolates 18 and 256, respectively, as internal control for the standardization of the PCR parameters. Molecular size markers (100-bp ladders) were applied to the first and the last lanes of each gel. The resulting PCR patterns were analyzed using the unweighted pair group method with arithmetic averages (UPGMA) algorithm of the GelCompar II software (Applied Maths, Sint-Martens-Latem, Belgium). UPGMA employs a sequential clustering algorithm, in which local topological relationships are identified in order of similarity, and the phylogenetic tree is built in a stepwise manner.
DRIII system (BioRad, Hercules, CA) in 1% w/v agarose gels (FMC InCert agarose, Biozym) at 9 8C in 0.5 TBE buffer. A constant voltage of 6 V cm-1 was applied with an increasing pulse time of 5 to 35 s over a period of 30 h. Gels were stained with ethidium bromide and photographed
2.6. Genomic DNA preparation and PFGE Preparation of chromosomal DNA, restriction endonuclease digestion, and analysis by PFGE was performed as previously described (Linhardt et al., 1992). In brief, strains were cultured overnight in brain heart infusion (BHI) medium, washed twice with TN buffer (10 mmol/L Tris, 150 mmol/L NaCl [pH 7.5]), and diluted to an OD600 of 0.8. Bacteria were mixed with equal amounts of 2% InCert agarose (Biozym, Hamburg, Germany). Lysis was performed for 72 h in lysis buffer (0.5 M EDTA, 1% laurylsarcosine, 2 mg/mL proteinase K) at 50 8C. Samples were washed 4 times in TE buffer (pH 8.0) and equilibrated in their respective enzyme buffer for 3 h at 4 8C. XbaI and SpeI (Fermentas, St. Leon-Rot, Germany) were used as restriction enzymes. Cleavage was performed with 33 U enzyme for 24 h. PFGE was performed with the CHEF-
Fig. 2. Identification of clonal isolates by PFGE analysis. Isolates 268 and 425 were recovered from the same patient as internal control for the detection of clonality in the present study. After digestion with XbaI (A), 4 pairs of identical PFGE patterns were found, of which only 3 were reproduced in a second PFGE after restriction with SpeI (B). In this respect, REP PCR patterns (C) corresponded to the second PFGE.
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
under UV illumination. PFGE Marker I E ladder (Boehringer, Mannheim, Germany) was used as molecular size standard. The resulting PFGE patterns were analyzed using the GelCompar II software (Applied Maths).
13
2.7. Nucleotide sequence accession numbers The hsp60 sequences newly reported in this study are accessible at the EMBL/GenBank databases under the accession numbers: AJ866391-AJ866530.
Fig. 3. UPGMA tree of REP PCR patterns. Similar to the NJ-tree of Fig. 1, 2 major clusters were produced. By REP PCR patterns, 7 isolates were not classified correctly to the genotype to which they belonged (isolates 116, 210, 211, 218, 242, 450, and 607). p, Patterns of clonal isolates; *, bfalsely Q identical patterns of nonclonal isolates.
14
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
3. Results 3.1. Classification of study strains to the genetic clusters of the E. cloacae complex After determination of partial hsp60 sequences, the study isolates were classified to the respective genetic cluster of the E. cloacae complex (data not shown). Forty-four isolates belonged to genetic cluster III and 12 isolates to genetic cluster VI (Fig. 1, left). For these 56 isolates, ERIC and REP PCR and PFGE were performed. 3.2. Pulsed-field gel electrophoresis PFGE, after XbaI digestion, generated fingerprints with approximately 7 to 12 bands per isolate (Fig. 2). Following the criteria of Tenover et al. (1995), isolates were considered as belonging to the same clone, if PFGE patterns differed by fewer than 2 restriction sites. UPGMA cluster analysis of the PFGE patterns yielded 52 different fingerprints (X-01 to X-52). Four fingerprints (X-13, X-29, X-45, and X-49) were identical for pairs of isolates each, whereas all the others were different. The 4 pairs with identical patterns belonged to genetic cluster III. They were reanalyzed in a second PFGE using SpeI instead of XbaI. Again, patterns were identical for 3 pairs (Fig. 2), whereas they were different for a fourth pair (isolates and 607) in the second run. One of the pairs with identical patterns served as internal control, the other two suggested transfer of the bacteria from one patient to another. 3.3. ERIC PCR The ERIC PCRs generated fingerprints with approximately 7 to 10 bands per isolate (Fig. 1). Reproducibility of the ERIC PCR was excellent because the internal controls of isolate 18 were identical in each run. Hence, repetitions of ERIC PCRs were not necessary. UPGMA analysis yielded 2 major ERIC clusters (ERIC cluster i and ERIC cluster ii), which corresponded to genetic clusters III and VI. Within the clusters, 14 relatedness groups of ERIC patterns (E-01 to E-14) were distinguished. Seven of them (E-01, E-02, E-03, E-07, E-08, E-12, and E-14) contained only 1 single isolate each, whereas 2 (E-06 and E-11) consisted of 2, 1 (E-13) of 3, 1 (E-04) of 4, 1 (E-10) of 5, 1 (E-09) of 14, and 1 (E-05) of 19 isolates. All isolates with identical PFGE patterns yielded identical ERIC PCR patterns but only 7 (14%) of the
Table 2 Differentiation of isolates by PFGE, ERIC PCR, REP PCR, and the combination of ERIC and REP PCR Method
Strains distinguished, n (%)a
PFGE ERIC PCR REP PCR ERIC and REP PCR
50 7 45 46
a
(100) (14) (90) (92)
Percentage of distinguishable isolates in comparison to PFGE results.
50 isolates differentiated by PFGE could also be distinguished by ERIC PCR (Table 2). 3.4. REP PCR Patterns obtained after REP PCR were more complex than the ERIC patterns. They consisted of 10 to 17 bands per isolate (Fig. 2). Similar to the ERIC PCRs, patterns of the internal control 256 were always reproduced identically, and repetitions of PCR runs were not necessary. Again, UPGMA cluster analysis of the REP PCR patterns yielded 2 major clusters (Fig. 3). With the exception of 7 isolates (116, 210, 211, 218, 242, 450, and 607), they corresponded to ERIC clusters i and ii and genetic clusters III and VI of the E. cloacae complex, respectively. Fifty different patterns (R-01 to R-50) were detected, of which 45 were obtained for single isolates only (R-01, R-02, R-04 to R-06, R-08 to R-10, R-12 to R-37, R-39 to R- 41, and R- 43 to R-50). Four REP patterns (R-03, R-07, R-11, and R-38) were obtained for 2 isolates each, and 1 (R- 42) for 3 isolates. There was no isolate of genetic cluster III with an ERIC or REP pattern identical to 1 of cluster VI and vice versa. In comparison to the PFGE results, 90% of isolates with different PFGE patterns also yielded different REP PCR patterns. This means that only 10% of the isolates that were distinguishable by their PFGE patterns were not distinguished by REP PCR. Similar to the ERIC PCR, all pairs of isolates with identical fingerprints in both PFGEs also had identical REP PCR patterns. The pair of isolates that showed identical patterns after the first PFGE (XbaI digestion) but different patterns in the second run (SpeI digestion) also showed different REP PCR patterns. Combining both the ERIC and REP PCR, 92% of the isolates with different PFGE patterns could be distinguished by the PCR-based methods. 4. Discussion The crude number of nosocomial infections with members of the E. cloacae complex has markedly increased over the last decades (Sanders and Sanders, 1997). The ability of the members of the E. cloacae complex to spread among patients has repetitively led to nosocomial outbreaks of infections, especially in neonatal intensive care units (DavinRegli et al., 1997; Fernandez-Baca et al., 2001; Kartali et al., 2002; van Nierop et al., 1998; Wenger et al., 1997). Valid and fast algorithms are necessary to identify outbreak strains as early as possible and to delineate real outbreaks from accumulations of polyclonal infections. For this purpose, the performance of ERIC and REP PCRs were compared for representative genovars of the E. cloacae complex (Olive and Bean, 1999). So far, conflicting data have been presented on the power of ERIC and REP PCR for the genetic discrimination of bacterial isolates. Some authors proclaimed the high discriminatory power of the ERIC PCR for Gram-negative rods in general (Ben-Hamouda et al., 2003; Chmielewski
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
et al., 2002; Szczuka and Kaznowski, 2004) and for Enterobacter species in particular (Fernandez-Baca et al., 2001), which, however, could not be reproduced by a number of studies for several other species (Jonas et al., 2003; Kerouanton et al., 1996; Rodriguez-Barradas et al., 1995; Ventura and Zink, 2002). In accordance to the latter’s, our analyses yielded several groups of isolates with identical ERIC patterns, of which clonality was only confirmed for 3 pairs of isolates by PFGE. Hence, a poor specificity of 14% was calculated for the ERIC PCR in respect to the detection of outbreak strains of the E. cloacae complex. In contrast, ERIC PCR reliably differentiated between the 2 genetic clusters analyzed. Hence, ERIC PCR seems to be a useful tool for the classification of isolates to their respective genetic clusters within the E. cloacae complex, whereas it has little to offer for the identification of outbreak strains. REP PCR has been applied to the genera Lactobacillus, Aeromonas, and Salmonella (Chmielewski et al., 2002; Szczuka and Kaznowski, 2004; Ventura and Zink, 2002), yielding poor patterns not allowing a valid differentiation of clones. However, when applied to species such as Enterobacter aerogenes, Listeria monocytogenes, or Bartonella species, REP PCR patterns were by far more complex and performed much better (Georghiou et al., 1995; Jersek et al., 1999; Rodriguez-Barradas et al., 1995; Shi et al., 1996). So far, 1 study analyzed the discriminatory power of ERIC and REP PCR for E. cloacae. In accordance to our results, REP PCR was superior to the ERIC PCR. However, at the time the study was conducted, there was no knowledge about the genomic heterogeneity of the nomenspecies. In a previous study, we could demonstrate that the nomenspecies E. cloacae in fact represents a complex of 12 different genovars of which some (E. asburiae, E. hormaechei, and E. kobei) have previously and some very recently (E. ludwigii, E. hormaechei subsp. oharae, E. hormaechei subsp. steigerwaltii, E. cloacae subsp. dissolvens, E. cloacae subsp. cloacae) gained species or subspecies status (Brenner et al., 1986; Hoffmann et al., 2005a, 2005b, 2005c; O’Hara et al., 1989). The 2 genovars analyzed in the present study were among the most prominent clusters in both the previous population genetic study and the present study of local clinical isolates. In view of these new data, ERIC PCR got a novel meaning because it allows fast and reliable classification of isolates to their respective genovars rather than genomic discrimination of isolates. PFGE is considered the golden standard method for DNA-based typing of isolates, but even this method has its pitfalls (Jonas et al., 2003). Our first PFGE run yielded 4 pairs of bclonal isolatesQ, for which clonality was confirmed only in 3 cases. One of them was our internal control consisting of 2 blood culture isolates (268 and 425) recovered from the same patient 3 days apart. The second pair consisted of 2 blood culture isolates (419 and 444) recovered from 1 patient nursed at the internal
15
medicine ICU (F2B) and from another 1 nursed at the secondary care ward to which the ICU patients are shifted once they have been weaned from the respirator. The third pair of clonal isolates (570 and 571) was recovered 4 days apart from blood cultures of 2 patients of the surgical ICU (G5). Facing a series of infections caused by an identical species, the major goal for the infectious disease professional is to identify outbreak strains as rapidly as possible. In situations of assumed outbreaks with members of the E. cloacae complex, adequate outbreak management should be initiated immediately and in parallel to the following diagnostic steps: first, a REP PCR should be performed of all boutbreakQ isolates, allowing a good genomic discrimination. If REP PCR patterns of different isolates are identical, a single PFGE analysis with XbaI should be performed. A confirmatory second PFGE is not needed in cases where REP and PFGE patterns are identical. ERIC PCR analyses might be performed if an identification of isolates on the level of genovars is needed (i.e., for research purposes).
Acknowledgments This study was funded by a grant from the Friedrich Bauer Foundation. The authors thank Beatrix Grabein, Andrea Haas, Waltraud Eder, Jana Bader, and Andreas Holzbach for their help in collecting study isolates.
References Ben-Hamouda T, Foulon T, Ben-Cheikh-Masmoudi A, Fendri C, Belhadj O, Ben-Mahrez K (2003) Molecular epidemiology of an outbreak of multiresistant Klebsiella pneumoniae in a Tunisian neonatal ward. J Med Microbiol 52:427 – 433. Brenner DJ, McWhorter AC, Kai A, Steigerwalt AG, Farmer Jr J (1986) Enterobacter asburiae sp. nov., a new species found in clinical specimens, and reassignment of Erwinia dissolvens and Erwinia nimipressuralis to the genus Enterobacter as Enterobacter dissolvens comb. nov. and Enterobacter nimipressuralis comb. nov. J Clin Microbiol 23:1114 – 1120. Chmielewski R, Wieliczko A, Kuczowski M, Mazurkiewicz M, Ugorski M (2002) Comparison of ITS profiling, REP- and ERIC-PCR of Salmonella enteriditis isolates from Poland. J Vet Med 49:163 – 168. Davin-Regli A, Bosi C, Charrel R, Ageron E, Papazian L, Grimont PA, Cremieux A, Bollet C (1997) A nosocomial outbreak due to Enterobacter cloacae strains with the E. hormaechei genotype in patients treated with fluoroquinolones. J Clin Microbiol 35:1008 – 1010. Fernandez-Baca V, Ballesteros F, Hervas JA, Villalon P, Dominguez MA, Benedi VJ, Alberti S (2001) Molecular epidemiological typing of Enterobacter cloacae isolates from a neonatal intensive care unit: Three-year prospective study. J Hosp Infect 49:173 – 182. Georghiou PR, Hamill RJ, Wright CE, Versalovic J, Koeuth T, Watson DA, Lupski JR (1995) Molecular epidemiology of infections due to Enterobacter aerogenes: Identification of hospital outbreak-associated strains by molecular techniques. Clin Infect Dis 20:84 – 94. Hoffmann H, Roggenkamp A (2003) Population genetics of the nomenspecies Enterobacter cloacae. Appl Environ Microbiol 69:5306 – 5318. Hoffmann H, Stindl S, Ludwig W, Stumpf A, Mehlen A, Heesemann J, Monget D, Schleifer KH, Roggenkamp A (2005a) Reassignment of
16
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
Enterobacter dissolvens to Enterobacter cloacae as E. cloacae subspecies dissolvens comb. nov. and emended description of Enterobacter asburiae and Enterobacter kobei. Syst Appl Microbiol 28: 196 – 205. Hoffmann H, Stindl S, Ludwig W, Stumpf A, Mehlen A, Monget D, Pierard D, Ziesing S, Heesemann J, Roggenkamp A, Scheifer KH (2005b) Enterobacter hormaechei subspecies oharae subsp. nov., E. hormaechei subspecies hormaechei comb. nov. and E. hormaechei subspecies steigerwaltii subsp. nov., three new subspecies of clinical importance. J Clin Microbiol 43:3297 – 3303. Hoffmann H, Stindl S, Stumpf A, Mehlen A, Monget D, Heesemann J, Schleifer KH, Roggenkamp A (2005c) Description of Enterobacter ludwigii sp. nov., a novel Enterobacter species of clinical relevance. Syst Appl Microbiol 28:206 – 212. Jersek B, Gilot P, Gubina M, Klun N, Mehle J, Tcherneva E, Rijpens N, Herman L (1999) Typing of Listeria monocytogenes strains by repetitive element sequence-based PCR. J Clin Microbiol 37:103 – 109. Jonas D, Spitzmuller B, Weist K, Ruden H, Daschner FD (2003) Comparison of PCR-based methods for typing Escherichia coli. Clin Microbiol Infect 9:823 – 831. Kartali G, Tzelepi E, Pournaras S, Kontopoulou C, Kontos F, Sofianou D, Maniatis AN, Tsakris A (2002) Outbreak of infections caused by Enterobacter cloacae producing the integron-associated beta-lactamase IBC-1 in a neonatal intensive care unit of a Greek hospital. Antimicrob Agents Chemother 46:1577 – 1580. Kerouanton A, Brisabois A, Grout J, Picard B (1996) Molecular epidemiological tools for Salmonella Dublin typing. FEMS Immunol Med Microbiol 14:25 – 29. Kuboyama RH, de Oliveira HB, Moretti-Branchini ML (2003) Molecular epidemiology of systemic infection caused by Enterobacter cloacae in a high-risk neonatal intensive care unit. Infect Control Hosp Epidemiol 24:490 – 494. Linhardt F, Ziebuhr W, Meyer P, Witte W, Hacker J (1992) Pulsed-field gel electrophoresis of genomic restriction fragments as a tool for the epidemiological analysis of Staphylococcus aureus and coagulasenegative staphylococci. FEMS Microbiol Lett 74:181 – 185. O’Hara CM, Steigerwalt AG, Hill BC, Farmer III JJ, Fanning GR, Brenner DJ (1989) Enterobacter hormaechei, a new species of the family
Enterobacteriaceae formerly known as enteric group 75. J Clin Microbiol 27:2046 – 2049. Olive DM, Bean P (1999) Principles and applications of methods for DNAbased typing of microbial organisms. J Clin Microbiol 37:1661 – 1669. Rodriguez-Barradas MC, Hamill RJ, Houston ED, Georghiou PR, Clarridge JE, Regnery RL, Koehler JE (1995) Genomic fingerprinting of Bartonella species by repetitive element PCR for distinguishing species and isolates. J Clin Microbiol 33:1089 – 1093. Sanders Jr WE, Sanders CC (1997) Enterobacter spp.: Pathogens poised to flourish at the turn of the century. Clin Microbiol Rev 10:220 – 241. Shi ZY, Liu PY, Lau YJ, Lin YH, Hu BS (1996) Epidemiological typing of isolates from an outbreak of infection with multidrug-resistant Enterobacter cloacae by repetitive extragenic palindromic unit b1-primed PCR and pulsed-field gel electrophoresis. J Clin Microbiol 34:2784 – 2790. Szczuka E, Kaznowski A (2004) Typing of clinical and environmental Aeromonas sp. strains by random amplified polymorphic DNA PCR, repetitive extragenic palindromic PCR, and enterobacterial repetitive intergenic consensus sequence PCR. J Clin Microbiol 42:220 – 228. Tenover FC, Arbeit RD, Goering RV, Mickelsen PA, Murray BE, Persing DH, Swaminathan B (1995) Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: Criteria for bacterial strain typing. J Clin Microbiol 33:2233 – 2239. van Nierop WH, Duse AG, Stewart RG, Bilgeri YR, Koornhof HJ (1998) Molecular epidemiology of an outbreak of Enterobacter cloacae in the neonatal intensive care unit of a provincial hospital in Gauteng, South Africa. J Clin Microbiol 36:3085 – 3087. Ventura M, Zink R (2002) Specific identification and molecular typing analysis of Lactobacillus johnsonii by using PCR-based methods and pulsed-field gel electrophoresis. FEMS Microbiol Lett 217:141 – 154. Versalovic J, Koeuth T, Lupski JR (1991) Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res 19:6823 – 6831. Wenger PN, Tokars JI, Brennan P, Samel C, Bland L, Miller M, Carson L, Arduino M, Edelstein P, Aguero S, Riddle C, O’hara C, Jarvis W (1997) An outbreak of Enterobacter hormaechei infection and colonization in an intensive care nursery. Clin Infect Dis 24:1243 – 1244.
Specificity of enterobacterial repetitive intergenic consensus and repetitive extragenic palindromic polymerase chain reaction for the detection of clonality within the Enterobacter cloacae complex Anita N. Stumpf a, Andreas Roggenkampa, Harald Hoffmannb,T a
Max von Pettenkofer Institute of Hygiene and Medical Microbiology, Ludwig Maximilian University Munich, Klinikum Groghadern, 81377 Munich, Germany b Institute of Microbiology and Laboratory Medicine at the Pneumological Hospitals Gauting, D-82131 Gauting, Germany Received 25 August 2004
Abstract An increasing number of clonal outbreaks caused by members of the E. cloacae complex is being reported. For the detection of clonality, pulsed-field gel electrophoresis (PFGE) is considered the golden standard, but PCR-based methods are cheaper, easier to perform, and provide faster results. One hundred ninety-five isolates of the E. cloacae complex isolated at the university hospital Groghadern, Munich, Germany, were assigned to their respective genetic cluster by partial hsp60 sequencing. All study isolates belonging to genetic clusters III and VI were selected to evaluate the specificity of the enterobacterial repetitive intergenic consensus (ERIC) and repetitive extragenic palindromic (REP) polymerase chain reaction (PCR) for the identification of clonal isolates belonging to the E. cloacae complex. For these 56 isolates, PFGE was performed, yielding 3 pairs of isolates with indistinguishable patterns. ERIC PCR resulted in 7 groups with identical patterns, together encompassing 49 study isolates. Comparing the ERIC PCR with the PFGE, a specificity of 14% considering the detection of bclonalQ isolates was calculated. In this respect, REP PCR performed much better, yielding a specificity of 90%. An unweighted pairgroup method with arithmetic averages tree based on ERIC PCR patterns allowed an accurate classification of the isolates to the respective genovars, suggesting that the ERIC PCR differentiates between genovars rather than between strains. In contrast, REP PCR differentiates better on the strain level. A proposed diagnostic system for the detection of subsumed outbreak strains of the E. cloacae complex is presented. It is based on an initial REP PCR, which should be confirmed by PFGE in cases of identical patterns, whereas ERIC PCR does not seem to be useful for the detection of outbreak strains when dealing with isolates of the E. cloacae complex. D 2005 Published by Elsevier Inc. Keywords: Enterobacter cloacae complex; Fingerprinting; PFGE; ERIC PCR; REP PCR
1. Introduction In a recent population genetic study, the nomenspecies Enterobacter cloacae turned out to represent, in fact, a large complex consisting of at least 12 genetic clusters (Hoffmann and Roggenkamp, 2003). An increasing number of outbreaks caused by the E. cloacae complex have been reported from several hospitals. For example, van Nierop et al. (1998) reported an outbreak in a neonatal intensive care unit with 9 deaths, Kuboyama et al. (2003) reported 3 outbreaks with 42 systemic infections and a crude T Corresponding author. Tel.: +49-89-85791-8230; fax: +49-8985791-8350. E-mail address: [email protected] (H. Hoffmann). 0732-8893/$ – see front matter D 2005 Published by Elsevier Inc. doi:10.1016/j.diagmicrobio.2005.04.003
mortality of 34%, and Fernandez-Baca et al. (2001) observed outbreaks affecting 15 neonates and causing 2 deaths. All of these lead the E. cloacae complex to be a serious hygienic problem. In a situation of questionable outbreaks, hygienic measurements should be initiated immediately. However, hygienic management is both expensive and demanding on manpower. Hence, an early answer is needed whether a series of infections with Enterobacter isolates represents a real outbreak or rather a polyclonal endemic infection due to different isolates of the same species. In the latter situation, changes in antibiotic strategies are certainly more adequate for the control than hygienic measures. Different techniques for the detection of clonality of isolates causing a series of infections are available. Pulsed-
10
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
field gel electrophoresis (PFGE) is considered the golden standard for genetic typing of bacteria (Olive and Bean, 1999; Tenover et al., 1995). It is highly discriminative but also expensive both in time and money. In contrast, polymerase chain reaction (PCR)–based methods are relatively cheap, easy to perform, and provide fast results. Different PCR fingerprint methods have been described, of which only few, such as enterobacterial repetitive intergenic consensus (ERIC) and repetitive extragenic palindromic (REP) PCR have so far been applied to the E. cloacae complex. In recent studies, both ERIC (Fernandez-Baca et al., 2001) and REP PCR (Shi et al., 1996) have been used for genetic differentiation of isolates. Both have been evaluated as highly discriminative methods. However, the studies did not take the genetic heterogeneity of the E. cloacae complex into consideration. Hence, the isolates of the respective studies might have belonged to several genovars of the E. cloacae complex. Then, the ERIC and REP PCRs would have differentiated on the species level rather than on the strain level, and a representative validation of both techniques would still remain elusive. In the present study, we focused on the evaluation of the specificity of ERIC and REP PCRs for the detection of outbreak strains within genetic clusters of the E. cloacae complex. PFGE was performed for each isolate as the golden standard reference method.
2. Materials and methods 2.1. Bacterial test collection One hundred ninety-five isolates were prospectively and randomly collected from clinical specimen at the Klinikum Groghadern, Munich, including only 1 isolate per patient. The collection comprised 55% of all isolates recovered from our diagnostic service from November 2001 to February 2002 and 90% of all blood culture isolates from 1997 to 2002. Isolates were biochemically identified as members of the E. cloacae complex (API 20E, bioMe´rieux, Marcy l’Etoile, France). They were assigned to the respective genetic cluster by partial hsp60 sequencing as previously described (Hoffmann and Roggenkamp, 2003). All isolates of genetic clusters III and VI were selected for further investigation. Cluster III consisted of 44 isolates, encompassing 2 intraabdominal isolates, 18 blood culture isolates, 1 isolate from a central venous line, 2 from pleural specimens, 9 from the respiratory tracts, 2 from skin swabs, 3 from stools, 5 from the urogenital tracts, and 2 from wound swabs. Twelve study isolates belonged to genetic cluster VI, encompassing 5 isolates from blood culture, 5 from the respiratory tract, 1 from a skin swab, and 1 from stool (Table 1). In one patient, 2 blood culture isolates with 3 days’ difference (isolates 268 and 425) were taken as internal control for the detection of clonality by the respective methods used. The author (AS), who performed ERIC and
Table 1 Study isolates and their characteristics Isolate Source of isolation
Genetic Designation of patterns and clustera groups of patterns
5 18 114 116 118 165 167 190 191 198 200 204 210 211 213 216 217 218 232 237 242 244 253 254 255 260 261 266 268 277 412 418 419 422 425 427 434 436 439 440 444 450 452 453 454 456 464 465 535 538 543 570 571 576 594 607
VI VI III III III III III VI III III III III VI III III III III VI VI III III III III III III VI III III III VI VI III III III III III III VI VI III III III III III III III III VI III III III III III III III III
PFGE (XbaI) PFGE (SpeI ) ERIC REP RT Stool RT Stool RT Pleura CVL RT RT RT UGT RT RT UGT RT UGT UGT Skin swab RT Pleura Wound swab Abdomen Skin swab RT UGT RT RT Abdomen BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx BCx Skin swab Stool BCx BCx BCx RT Wound swab Stool
X-35 X-16 X-05 X-49 X-09 X-38 X-39 X-34 X-26 X-04 X-01 X-40 X-36 X-47 X-03 X-41 X-17 X-37 X-25 X-50 X-31 X-02 X-10 X-21 X-07 X-46 X-14 X-30 X-45 X-19 X-27 X-08 X-13 X-32 X-45 X-06 X-18 X-28 X-20 X-42 X-13 X-22 X-33 X-43 X-24 X-23 X-48 X-51 X-11 X-52 X-15 X-29 X-29 X-12 X-44 X-49
– – – S-05 – – – – – – – – – – – – – – – – – – – – – – – – S-03 – – – S-01 – S-03 – – – – – S-01 – – – – – – – – – – S-02 S-02 – – S-04
E-11 E-12 E-04 E-04 E-05 E-05 E-04 E-13 E-05 E-03 E-06 E-08 E-10 E-05 E-06 E-05 E-05 E-10 E-13 E-09 E-01 E-05 E-05 E-05 E-05 E-14 E-09 E-05 E-09 E-10 E-13 E-09 E-09 E-05 E-09 E-09 E-09 E-10 E-10 E-09 E-09 E-09 E-09 E-05 E-09 E-05 E-09 E-11 E-05 E-07 E-05 E-05 E-05 E-02 E-05 E-04
R-47 R-44 R-35 R-50 R-22 R-29 R-32 R-43 R-23 R-33 R-20 R-24 R-38 R-40 R-26 R-15 R-09 R-38 R-42 R-36 R-41 R-14 R-37 R-27 R-34 R-49 R-10 R-08 R-07 R-42 R-45 R-25 R-03 R-28 R-07 R-17 R-06 R-46 R-42 R-04 R-03 R-01 R-16 R-39 R-12 R-05 R-21 R-48 R-18 R-19 R-13 R-11 R-11 R-30 R-31 R-02
RT = respiratory tract; CVL = central venous line; UGT = urogenital tract; BCx = blood culture. Clusters rendered in boldface comprised more than 1 isolate. a Genetic cluster denominations following (Hoffmann and Roggenkamp, 2003).
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
11
Fig. 1. hsp60 neighbor joining tree (left) and UPGMA tree of ERIC PCR patterns (right) of 56 prospectively collected clinical isolates of the E. cloacae complex, including the internal control isolates (268 and 425) recovered from the same patient 3 days apart. Both trees have been rooted with the E. cloacaetype strain (ATCC13047T). The 2 major ERIC clusters i and ii corresponded to the 2 hsp60 sequence clusters analyzed (genetic clusters III and VI, respectively). Within the major ERIC clusters, 14 relatedness groups were observed (E-01 to E-14). TOutbreak causing strains.
REP PCR and PFGE, was not aware of this doubling of an identical strain.
DNA was prepared for PCR by quick-heat lysis. Isolates were grown in Luria-Bertani broth (LB) to an optical density of 1.0 at 600 nm. A 1.5-mL culture was centrifuged for 2 min at 10 000 rpm. The pellet was diluted in 0.5 mL of distilled water, boiled for 10 min, and centrifuged (5 min, 10 000 rpm, 4 8C).
ATC AG-3V). Reactions were carried out in a total volume of 50 AL and started after an activation time of 10 min at 95 8C. Samples were subjected to 30 cycles of 30 s at 95 8C, 30 s at 59 8C, and 1 min of extension at 72 8C. Cycles were followed by a final elongation step of 15 min at 72 8C. PCR products were visualized in 2% agarose gels after electrophoresis and ethidium bromide staining. DNA was cleaned using the NucleoSpin Extraction kit (Macherey-Nagel, Dqren, Germany) for further processing following the manufacturer’s guidelines.
2.3. HSP60 PCR
2.4. HSP60 sequencing
Amplification reactions were performed with an Applied Biosystems GeneAmp PCR system 2700 and Taq Gold DNA hot start polymerase (Perkin-Elmer, Branchburg, NJ) under standard conditions (500 pmol of each primer, 200 Amol/L nucleotide mix, 10 mmol/L Tris-Cl [pH 8.3], 5 mmol/L KCl, 1.5 mmol/L MgCl2). Primers were used as previously described (Hoffmann and Roggenkamp, 2003) (HSP-1218-F: 5V-GGT AGA AGA AGG CGT GGT TGC3V and HSP-1559-R: 5V-ATG CAT TCG GTG GTG ATC
PCR-amplified and cleaned DNA was sequenced using the dye terminator method. Purified DNA (0.2 Ag) was added to a 20-AL reaction solution containing 4 AL of BigDye dye terminator sequencing mix and 0.7 AL of a 5-pmol primer solution. Cycle parameters were 25 cycles with an initial 96 8C denaturation step of 10 s, followed by an annealing step at 55 8C of 10 s, and an extension step of 4 min at 60 8C performed in an Applied Biosystems GeneAmp PCR System 2700. Sequences were
2.2. Isolation of DNA
12
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
determined by electrophoresis with the ABI Prism 377 DNA sequencer. 2.5. ERIC and REP PCR The principles of the ERIC and REP PCR have been described in detail in previous publications (Shi et al., 1996; Versalovic et al., 1991). Primer sequences were as follows: ERIC-1R: 5V-ATG TAA GCT CCT GGG GAT TCA C-3V; ERIC-2: 5V-AAG TAA GTG ACT GGG GTG AGC G-3V; REP-1R-I: 5V-III ICG ICA TCI GGC-3V; REP-2-I: 5V-ICG ICT TAT CIG GCC TAC-3V(Versalovic et al., 1991). Primers were synthesized by Metabion, Munich, Germany. Amplification reactions were carried out in an Applied Biosystems GeneAmp 2700 PCR thermocycler. Each 25-AL PCR reaction contained 50 pmol of each of the 2 opposing primers, 1.25 mmol/L of each of the 4 dNTPs, 2 U AmpliTaq DNA polymerase (Perkin-Elmer), 1.5 mmol/L MgCl2, 5 mmol/L KCl, 10 mmol/L Tris-Cl (pH 8.3), and 1.6 AL DNA. Amplifications were performed with an initial denaturation step at 94 8C for 7 min, followed by 30 cycles of denaturation (90 8C, 30 s), annealing (ERIC: 52 8C/ REP: 40 8C, 1 min), and extension (65 8C, 8 min), completed by a single final extension step (65 8C, 16 min). Twenty microliters of each PCR reaction was electrophoresed on 2% agarose gels in 1 Tris acetate–EDTA buffer at 100 V for 1 h, stained with ethidium bromide, and visualized under UV light. In each run, ERIC and REP PCR were repeated for isolates 18 and 256, respectively, as internal control for the standardization of the PCR parameters. Molecular size markers (100-bp ladders) were applied to the first and the last lanes of each gel. The resulting PCR patterns were analyzed using the unweighted pair group method with arithmetic averages (UPGMA) algorithm of the GelCompar II software (Applied Maths, Sint-Martens-Latem, Belgium). UPGMA employs a sequential clustering algorithm, in which local topological relationships are identified in order of similarity, and the phylogenetic tree is built in a stepwise manner.
DRIII system (BioRad, Hercules, CA) in 1% w/v agarose gels (FMC InCert agarose, Biozym) at 9 8C in 0.5 TBE buffer. A constant voltage of 6 V cm-1 was applied with an increasing pulse time of 5 to 35 s over a period of 30 h. Gels were stained with ethidium bromide and photographed
2.6. Genomic DNA preparation and PFGE Preparation of chromosomal DNA, restriction endonuclease digestion, and analysis by PFGE was performed as previously described (Linhardt et al., 1992). In brief, strains were cultured overnight in brain heart infusion (BHI) medium, washed twice with TN buffer (10 mmol/L Tris, 150 mmol/L NaCl [pH 7.5]), and diluted to an OD600 of 0.8. Bacteria were mixed with equal amounts of 2% InCert agarose (Biozym, Hamburg, Germany). Lysis was performed for 72 h in lysis buffer (0.5 M EDTA, 1% laurylsarcosine, 2 mg/mL proteinase K) at 50 8C. Samples were washed 4 times in TE buffer (pH 8.0) and equilibrated in their respective enzyme buffer for 3 h at 4 8C. XbaI and SpeI (Fermentas, St. Leon-Rot, Germany) were used as restriction enzymes. Cleavage was performed with 33 U enzyme for 24 h. PFGE was performed with the CHEF-
Fig. 2. Identification of clonal isolates by PFGE analysis. Isolates 268 and 425 were recovered from the same patient as internal control for the detection of clonality in the present study. After digestion with XbaI (A), 4 pairs of identical PFGE patterns were found, of which only 3 were reproduced in a second PFGE after restriction with SpeI (B). In this respect, REP PCR patterns (C) corresponded to the second PFGE.
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
under UV illumination. PFGE Marker I E ladder (Boehringer, Mannheim, Germany) was used as molecular size standard. The resulting PFGE patterns were analyzed using the GelCompar II software (Applied Maths).
13
2.7. Nucleotide sequence accession numbers The hsp60 sequences newly reported in this study are accessible at the EMBL/GenBank databases under the accession numbers: AJ866391-AJ866530.
Fig. 3. UPGMA tree of REP PCR patterns. Similar to the NJ-tree of Fig. 1, 2 major clusters were produced. By REP PCR patterns, 7 isolates were not classified correctly to the genotype to which they belonged (isolates 116, 210, 211, 218, 242, 450, and 607). p, Patterns of clonal isolates; *, bfalsely Q identical patterns of nonclonal isolates.
14
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
3. Results 3.1. Classification of study strains to the genetic clusters of the E. cloacae complex After determination of partial hsp60 sequences, the study isolates were classified to the respective genetic cluster of the E. cloacae complex (data not shown). Forty-four isolates belonged to genetic cluster III and 12 isolates to genetic cluster VI (Fig. 1, left). For these 56 isolates, ERIC and REP PCR and PFGE were performed. 3.2. Pulsed-field gel electrophoresis PFGE, after XbaI digestion, generated fingerprints with approximately 7 to 12 bands per isolate (Fig. 2). Following the criteria of Tenover et al. (1995), isolates were considered as belonging to the same clone, if PFGE patterns differed by fewer than 2 restriction sites. UPGMA cluster analysis of the PFGE patterns yielded 52 different fingerprints (X-01 to X-52). Four fingerprints (X-13, X-29, X-45, and X-49) were identical for pairs of isolates each, whereas all the others were different. The 4 pairs with identical patterns belonged to genetic cluster III. They were reanalyzed in a second PFGE using SpeI instead of XbaI. Again, patterns were identical for 3 pairs (Fig. 2), whereas they were different for a fourth pair (isolates and 607) in the second run. One of the pairs with identical patterns served as internal control, the other two suggested transfer of the bacteria from one patient to another. 3.3. ERIC PCR The ERIC PCRs generated fingerprints with approximately 7 to 10 bands per isolate (Fig. 1). Reproducibility of the ERIC PCR was excellent because the internal controls of isolate 18 were identical in each run. Hence, repetitions of ERIC PCRs were not necessary. UPGMA analysis yielded 2 major ERIC clusters (ERIC cluster i and ERIC cluster ii), which corresponded to genetic clusters III and VI. Within the clusters, 14 relatedness groups of ERIC patterns (E-01 to E-14) were distinguished. Seven of them (E-01, E-02, E-03, E-07, E-08, E-12, and E-14) contained only 1 single isolate each, whereas 2 (E-06 and E-11) consisted of 2, 1 (E-13) of 3, 1 (E-04) of 4, 1 (E-10) of 5, 1 (E-09) of 14, and 1 (E-05) of 19 isolates. All isolates with identical PFGE patterns yielded identical ERIC PCR patterns but only 7 (14%) of the
Table 2 Differentiation of isolates by PFGE, ERIC PCR, REP PCR, and the combination of ERIC and REP PCR Method
Strains distinguished, n (%)a
PFGE ERIC PCR REP PCR ERIC and REP PCR
50 7 45 46
a
(100) (14) (90) (92)
Percentage of distinguishable isolates in comparison to PFGE results.
50 isolates differentiated by PFGE could also be distinguished by ERIC PCR (Table 2). 3.4. REP PCR Patterns obtained after REP PCR were more complex than the ERIC patterns. They consisted of 10 to 17 bands per isolate (Fig. 2). Similar to the ERIC PCRs, patterns of the internal control 256 were always reproduced identically, and repetitions of PCR runs were not necessary. Again, UPGMA cluster analysis of the REP PCR patterns yielded 2 major clusters (Fig. 3). With the exception of 7 isolates (116, 210, 211, 218, 242, 450, and 607), they corresponded to ERIC clusters i and ii and genetic clusters III and VI of the E. cloacae complex, respectively. Fifty different patterns (R-01 to R-50) were detected, of which 45 were obtained for single isolates only (R-01, R-02, R-04 to R-06, R-08 to R-10, R-12 to R-37, R-39 to R- 41, and R- 43 to R-50). Four REP patterns (R-03, R-07, R-11, and R-38) were obtained for 2 isolates each, and 1 (R- 42) for 3 isolates. There was no isolate of genetic cluster III with an ERIC or REP pattern identical to 1 of cluster VI and vice versa. In comparison to the PFGE results, 90% of isolates with different PFGE patterns also yielded different REP PCR patterns. This means that only 10% of the isolates that were distinguishable by their PFGE patterns were not distinguished by REP PCR. Similar to the ERIC PCR, all pairs of isolates with identical fingerprints in both PFGEs also had identical REP PCR patterns. The pair of isolates that showed identical patterns after the first PFGE (XbaI digestion) but different patterns in the second run (SpeI digestion) also showed different REP PCR patterns. Combining both the ERIC and REP PCR, 92% of the isolates with different PFGE patterns could be distinguished by the PCR-based methods. 4. Discussion The crude number of nosocomial infections with members of the E. cloacae complex has markedly increased over the last decades (Sanders and Sanders, 1997). The ability of the members of the E. cloacae complex to spread among patients has repetitively led to nosocomial outbreaks of infections, especially in neonatal intensive care units (DavinRegli et al., 1997; Fernandez-Baca et al., 2001; Kartali et al., 2002; van Nierop et al., 1998; Wenger et al., 1997). Valid and fast algorithms are necessary to identify outbreak strains as early as possible and to delineate real outbreaks from accumulations of polyclonal infections. For this purpose, the performance of ERIC and REP PCRs were compared for representative genovars of the E. cloacae complex (Olive and Bean, 1999). So far, conflicting data have been presented on the power of ERIC and REP PCR for the genetic discrimination of bacterial isolates. Some authors proclaimed the high discriminatory power of the ERIC PCR for Gram-negative rods in general (Ben-Hamouda et al., 2003; Chmielewski
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
et al., 2002; Szczuka and Kaznowski, 2004) and for Enterobacter species in particular (Fernandez-Baca et al., 2001), which, however, could not be reproduced by a number of studies for several other species (Jonas et al., 2003; Kerouanton et al., 1996; Rodriguez-Barradas et al., 1995; Ventura and Zink, 2002). In accordance to the latter’s, our analyses yielded several groups of isolates with identical ERIC patterns, of which clonality was only confirmed for 3 pairs of isolates by PFGE. Hence, a poor specificity of 14% was calculated for the ERIC PCR in respect to the detection of outbreak strains of the E. cloacae complex. In contrast, ERIC PCR reliably differentiated between the 2 genetic clusters analyzed. Hence, ERIC PCR seems to be a useful tool for the classification of isolates to their respective genetic clusters within the E. cloacae complex, whereas it has little to offer for the identification of outbreak strains. REP PCR has been applied to the genera Lactobacillus, Aeromonas, and Salmonella (Chmielewski et al., 2002; Szczuka and Kaznowski, 2004; Ventura and Zink, 2002), yielding poor patterns not allowing a valid differentiation of clones. However, when applied to species such as Enterobacter aerogenes, Listeria monocytogenes, or Bartonella species, REP PCR patterns were by far more complex and performed much better (Georghiou et al., 1995; Jersek et al., 1999; Rodriguez-Barradas et al., 1995; Shi et al., 1996). So far, 1 study analyzed the discriminatory power of ERIC and REP PCR for E. cloacae. In accordance to our results, REP PCR was superior to the ERIC PCR. However, at the time the study was conducted, there was no knowledge about the genomic heterogeneity of the nomenspecies. In a previous study, we could demonstrate that the nomenspecies E. cloacae in fact represents a complex of 12 different genovars of which some (E. asburiae, E. hormaechei, and E. kobei) have previously and some very recently (E. ludwigii, E. hormaechei subsp. oharae, E. hormaechei subsp. steigerwaltii, E. cloacae subsp. dissolvens, E. cloacae subsp. cloacae) gained species or subspecies status (Brenner et al., 1986; Hoffmann et al., 2005a, 2005b, 2005c; O’Hara et al., 1989). The 2 genovars analyzed in the present study were among the most prominent clusters in both the previous population genetic study and the present study of local clinical isolates. In view of these new data, ERIC PCR got a novel meaning because it allows fast and reliable classification of isolates to their respective genovars rather than genomic discrimination of isolates. PFGE is considered the golden standard method for DNA-based typing of isolates, but even this method has its pitfalls (Jonas et al., 2003). Our first PFGE run yielded 4 pairs of bclonal isolatesQ, for which clonality was confirmed only in 3 cases. One of them was our internal control consisting of 2 blood culture isolates (268 and 425) recovered from the same patient 3 days apart. The second pair consisted of 2 blood culture isolates (419 and 444) recovered from 1 patient nursed at the internal
15
medicine ICU (F2B) and from another 1 nursed at the secondary care ward to which the ICU patients are shifted once they have been weaned from the respirator. The third pair of clonal isolates (570 and 571) was recovered 4 days apart from blood cultures of 2 patients of the surgical ICU (G5). Facing a series of infections caused by an identical species, the major goal for the infectious disease professional is to identify outbreak strains as rapidly as possible. In situations of assumed outbreaks with members of the E. cloacae complex, adequate outbreak management should be initiated immediately and in parallel to the following diagnostic steps: first, a REP PCR should be performed of all boutbreakQ isolates, allowing a good genomic discrimination. If REP PCR patterns of different isolates are identical, a single PFGE analysis with XbaI should be performed. A confirmatory second PFGE is not needed in cases where REP and PFGE patterns are identical. ERIC PCR analyses might be performed if an identification of isolates on the level of genovars is needed (i.e., for research purposes).
Acknowledgments This study was funded by a grant from the Friedrich Bauer Foundation. The authors thank Beatrix Grabein, Andrea Haas, Waltraud Eder, Jana Bader, and Andreas Holzbach for their help in collecting study isolates.
References Ben-Hamouda T, Foulon T, Ben-Cheikh-Masmoudi A, Fendri C, Belhadj O, Ben-Mahrez K (2003) Molecular epidemiology of an outbreak of multiresistant Klebsiella pneumoniae in a Tunisian neonatal ward. J Med Microbiol 52:427 – 433. Brenner DJ, McWhorter AC, Kai A, Steigerwalt AG, Farmer Jr J (1986) Enterobacter asburiae sp. nov., a new species found in clinical specimens, and reassignment of Erwinia dissolvens and Erwinia nimipressuralis to the genus Enterobacter as Enterobacter dissolvens comb. nov. and Enterobacter nimipressuralis comb. nov. J Clin Microbiol 23:1114 – 1120. Chmielewski R, Wieliczko A, Kuczowski M, Mazurkiewicz M, Ugorski M (2002) Comparison of ITS profiling, REP- and ERIC-PCR of Salmonella enteriditis isolates from Poland. J Vet Med 49:163 – 168. Davin-Regli A, Bosi C, Charrel R, Ageron E, Papazian L, Grimont PA, Cremieux A, Bollet C (1997) A nosocomial outbreak due to Enterobacter cloacae strains with the E. hormaechei genotype in patients treated with fluoroquinolones. J Clin Microbiol 35:1008 – 1010. Fernandez-Baca V, Ballesteros F, Hervas JA, Villalon P, Dominguez MA, Benedi VJ, Alberti S (2001) Molecular epidemiological typing of Enterobacter cloacae isolates from a neonatal intensive care unit: Three-year prospective study. J Hosp Infect 49:173 – 182. Georghiou PR, Hamill RJ, Wright CE, Versalovic J, Koeuth T, Watson DA, Lupski JR (1995) Molecular epidemiology of infections due to Enterobacter aerogenes: Identification of hospital outbreak-associated strains by molecular techniques. Clin Infect Dis 20:84 – 94. Hoffmann H, Roggenkamp A (2003) Population genetics of the nomenspecies Enterobacter cloacae. Appl Environ Microbiol 69:5306 – 5318. Hoffmann H, Stindl S, Ludwig W, Stumpf A, Mehlen A, Heesemann J, Monget D, Schleifer KH, Roggenkamp A (2005a) Reassignment of
16
A.N. Stumpf et al. / Diagnostic Microbiology and Infectious Disease 53 (2005) 9 –16
Enterobacter dissolvens to Enterobacter cloacae as E. cloacae subspecies dissolvens comb. nov. and emended description of Enterobacter asburiae and Enterobacter kobei. Syst Appl Microbiol 28: 196 – 205. Hoffmann H, Stindl S, Ludwig W, Stumpf A, Mehlen A, Monget D, Pierard D, Ziesing S, Heesemann J, Roggenkamp A, Scheifer KH (2005b) Enterobacter hormaechei subspecies oharae subsp. nov., E. hormaechei subspecies hormaechei comb. nov. and E. hormaechei subspecies steigerwaltii subsp. nov., three new subspecies of clinical importance. J Clin Microbiol 43:3297 – 3303. Hoffmann H, Stindl S, Stumpf A, Mehlen A, Monget D, Heesemann J, Schleifer KH, Roggenkamp A (2005c) Description of Enterobacter ludwigii sp. nov., a novel Enterobacter species of clinical relevance. Syst Appl Microbiol 28:206 – 212. Jersek B, Gilot P, Gubina M, Klun N, Mehle J, Tcherneva E, Rijpens N, Herman L (1999) Typing of Listeria monocytogenes strains by repetitive element sequence-based PCR. J Clin Microbiol 37:103 – 109. Jonas D, Spitzmuller B, Weist K, Ruden H, Daschner FD (2003) Comparison of PCR-based methods for typing Escherichia coli. Clin Microbiol Infect 9:823 – 831. Kartali G, Tzelepi E, Pournaras S, Kontopoulou C, Kontos F, Sofianou D, Maniatis AN, Tsakris A (2002) Outbreak of infections caused by Enterobacter cloacae producing the integron-associated beta-lactamase IBC-1 in a neonatal intensive care unit of a Greek hospital. Antimicrob Agents Chemother 46:1577 – 1580. Kerouanton A, Brisabois A, Grout J, Picard B (1996) Molecular epidemiological tools for Salmonella Dublin typing. FEMS Immunol Med Microbiol 14:25 – 29. Kuboyama RH, de Oliveira HB, Moretti-Branchini ML (2003) Molecular epidemiology of systemic infection caused by Enterobacter cloacae in a high-risk neonatal intensive care unit. Infect Control Hosp Epidemiol 24:490 – 494. Linhardt F, Ziebuhr W, Meyer P, Witte W, Hacker J (1992) Pulsed-field gel electrophoresis of genomic restriction fragments as a tool for the epidemiological analysis of Staphylococcus aureus and coagulasenegative staphylococci. FEMS Microbiol Lett 74:181 – 185. O’Hara CM, Steigerwalt AG, Hill BC, Farmer III JJ, Fanning GR, Brenner DJ (1989) Enterobacter hormaechei, a new species of the family
Enterobacteriaceae formerly known as enteric group 75. J Clin Microbiol 27:2046 – 2049. Olive DM, Bean P (1999) Principles and applications of methods for DNAbased typing of microbial organisms. J Clin Microbiol 37:1661 – 1669. Rodriguez-Barradas MC, Hamill RJ, Houston ED, Georghiou PR, Clarridge JE, Regnery RL, Koehler JE (1995) Genomic fingerprinting of Bartonella species by repetitive element PCR for distinguishing species and isolates. J Clin Microbiol 33:1089 – 1093. Sanders Jr WE, Sanders CC (1997) Enterobacter spp.: Pathogens poised to flourish at the turn of the century. Clin Microbiol Rev 10:220 – 241. Shi ZY, Liu PY, Lau YJ, Lin YH, Hu BS (1996) Epidemiological typing of isolates from an outbreak of infection with multidrug-resistant Enterobacter cloacae by repetitive extragenic palindromic unit b1-primed PCR and pulsed-field gel electrophoresis. J Clin Microbiol 34:2784 – 2790. Szczuka E, Kaznowski A (2004) Typing of clinical and environmental Aeromonas sp. strains by random amplified polymorphic DNA PCR, repetitive extragenic palindromic PCR, and enterobacterial repetitive intergenic consensus sequence PCR. J Clin Microbiol 42:220 – 228. Tenover FC, Arbeit RD, Goering RV, Mickelsen PA, Murray BE, Persing DH, Swaminathan B (1995) Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: Criteria for bacterial strain typing. J Clin Microbiol 33:2233 – 2239. van Nierop WH, Duse AG, Stewart RG, Bilgeri YR, Koornhof HJ (1998) Molecular epidemiology of an outbreak of Enterobacter cloacae in the neonatal intensive care unit of a provincial hospital in Gauteng, South Africa. J Clin Microbiol 36:3085 – 3087. Ventura M, Zink R (2002) Specific identification and molecular typing analysis of Lactobacillus johnsonii by using PCR-based methods and pulsed-field gel electrophoresis. FEMS Microbiol Lett 217:141 – 154. Versalovic J, Koeuth T, Lupski JR (1991) Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucleic Acids Res 19:6823 – 6831. Wenger PN, Tokars JI, Brennan P, Samel C, Bland L, Miller M, Carson L, Arduino M, Edelstein P, Aguero S, Riddle C, O’hara C, Jarvis W (1997) An outbreak of Enterobacter hormaechei infection and colonization in an intensive care nursery. Clin Infect Dis 24:1243 – 1244.