Molecular Epidemiology of Extended-Spectrum β-Lactamase-Producing Klebsiella pneumoniae Isolates in a District Hospital in Taiwan

Abstract

Thirty-one of 104 clinical isolates of Klebsiella pneumoniae collected over a period of 8 months were found to be putative extended-spectrum β-lactamase (ESBL) producers. Isoelectric focusing and an iodine overlay agar method were used for preliminary identification of the ESBLs. They were further identified by DNA sequencing. Seventy-one percent of the isolates were found to produce SHV-5. The variation in the ESBL patterns of these isolates was slight, with only five patterns being identified. The strains were typed by pulsed-field gel electrophoresis (PFGE), and 16 different genotypes were identified. When the PFGE patterns were analyzed by the algorithmic clustering method called the unweighted-pair group method using arithmetic averages, five clusters were found. However, significant genetic variations were found among 11 isolates and between each cluster. A plasmid of 36 kb was found in all clinical isolates and in the transconjugants. Our results indicate that the increase in the number of ESBL-producing K. pneumoniae isolates in this hospital is due mainly to the dissemination of a resistance plasmid rather than to the clonal spread of a few epidemic strains.

The introduction of extended-spectrum cephalosporins has facilitated effective treatment of severe infections caused by gram-negative bacteria. However, increasing use of these agents has been associated with the emergence of resistant bacterial strains (18). In 1983, Knothe et al. (13) described for the first time transferable resistance to broad-spectrum cephalosporins in clinical isolates of Klebsiella pneumoniae and Serratia marcescens. This new resistance phenotype was due to production of an extended-spectrum β-lactamase (ESBL), SHV-2, which evolved from the well-known SHV-1 enzyme (12). Subsequently, strains producing the TEM-derived ESBL CTX-1 were isolated in several French hospitals. Since then there have been rapid increases in the number and variety of ESBLs. Bacterial strains, especially K. pneumoniae strains, producing various types of ESBLs have spread, or independently evolved, worldwide (9, 15). Widespread dissemination of such strains within hospitals is reported with increasing frequency (9, 15). Recent surveys of hospital isolates of K. pneumoniae in England (14) and France (22) show that 14 to 16% produce ESBLs. Hospital colonization by these ESBL-producing strains usually is a complex phenomenon involving different mechanisms: dissemination of several epidemic strains (1, 2, 7, 10, 26), dissemination of plasmids and resistance genes (3, 11, 19–21), or concurrent dissemination of genes, plasmids, and strains (5). Moreover, identical ESBLs have evolved independently in different places at different times (8) and, occasionally, single isolates have carried multiple ESBLs (4).

In 1997, we noticed a marked increase in the number of ceftazidime-resistant K. pneumoniae strains isolated in our hospital, a 450-bed district teaching hospital in Taiwan. To analyze the epidemiologies of these putative ESBL-producing isolates, we used various molecular techniques to delineate their genetic relationships and identify their ESBL patterns. This is the first report of a study of the epidemiology of ESBL-producing K. pneumoniae isolates in a Taiwan hospital.

From January to August 1997, 104 nonrepetitive (one per patient) clinical isolates of K. pneumoniae were isolated consecutively from hospitalized patients. Among these, 31 were reported to be intermediate or resistant to either ceftazidime, cefotaxime, or aztreonam by our clinical microbiology laboratory (Table 1). These 31 resistant isolates were included in this study. Isolates were identified by using the API system (BioMerieux, Marcy l’Etoile, France). Production of ESBLs by these isolates was tested by an agar dilution method (16), the Etest ESBL screen (AB BIODISK, Piscataway, N.J.), and the double-disk synergy test (25).

TABLE 1.

Characteristics of 31 clinical isolates of ESBL-producing K. pneumoniae

Isolate	Wardb	Date of isolation (day/mo/yr)	Source	Macrorestriction genotypea	ESBL pattern(s)c	Plasmid profile(s)d
S1	MICU	1/1/1997	Urine	1	SHV-5	36.0, 4.0
S2	MICU	19/2/1997	Blood	2	SHV-5	36.0
S3	MICU	6/3/1997	Sputum	1	SHV-5	36.0, 4.0
S4	MICU	4/4/1997	Urine	3	SHV-5	36.0, 4.7
S5	MICU	16/4/1997	Urine	4	8.25	36.0
S6	MICU	18/4/1997	Sputum	5	SHV-5	36.0
S7	MICU	24/4/1997	Pus	7	7.9, 7.75	36.0, 9.8
S8	MICU	13/5/1997	Pus	8	8.25, SHV-2	36.0, 3.0
S9	MICU	20/5/1997	Pus	9	SHV-5	36.0, 5.2, 3.7, 2.9
S10	MICU	22/5/1997	Sputum	10	SHV-5, 7.9, 7.75	36.0, 9.8, 7.0, 4.0
S11	MICU	30/5/1997	Urine	11	SHV-5	36.0
S12	MICU	1/7/1997	Urine	9a	SHV-5	36.0, 3.7
S13	MICU	15/7/1997	Sputum	11	SHV-5	36.0
S14	MICU	21/7/1997	Sputum	9b	SHV-5, SHV-2	36.0, 3.7, 2.5
S15	SICU	17/1/1997	Sputum	1	SHV-5	36.0, 4.0
S16	NICU	26/3/1997	Sputum	1	SHV-5	36.0, 4.0
S17	NICU	31/3/1997	Sputum	10	7.9, 7.75	36.0, 16.1, 7.0
S18	NICU	21/4/1997	Sputum	7a	7.9, 7.75	36.0
S19	NICU	28/6/1997	Sputum	12	SHV-5	36.0, 10.0, 3.7
S20	W1	18/7/1997	Urine	11	SHV-5	36.0
S21	W2	21/1/1997	Urine	13	SHV-5	36.0
S22	W2	24/4/1997	Sputum	7b	8.25	36.0, 9.8
S23	W2	29/7/1997	Urine	9c	SHV-5	36.0, 3.7
S24	W3	8/5/1997	Sputum	10a	7.9, 7.75	36.0, 7.0, 4.5, 2.9
S25	W3	21/7/1997	Pus	14	SHV-5	36.0
S26	W4	14/5/1997	Sputum	15	SHV-5	36.0, 7.0
S27	W4	10/7/1997	Sputum	7c	SHV-5	36.0, 9.8, 4.0
S28	W4	18/8/1997	Urine	11	SHV-5	36.0
S29	W6	18/4/1997	Sputum	16	7.9, 7.75	36.0
S30	W6	23/4/1997	Sputum	6	SHV-5	36.0
S31	W6	24/4/1997	Pus	7	7.9, 7.75	36.0, 9.8

^aMacrorestriction genotypes were determined by PFGE after digestion with XbaI. 

^bMICU, Medical Intensive-Care Unit; SICU, Surgical Intensive-Care Unit; NICU, Neurological Intensive-Care Unit; W, ward. 

^cSome unknown enzymes are designated by their pIs as determined by isoelectric focusing. 

^dPlasmid profile, estimated plasmid size in kilobases. 

Pulsed-field gel electrophoresis (PFGE) was performed with a contour-clamped homogeneous electric field DRII apparatus from Bio-Rad Laboratories (Richmond, Calif.) as described previously (7). The chromosomal DNA was digested overnight with XbaI (GIBCO-BRL, Life Technologies, Gaithersburg, Md.). DNA was electrophoresed in 1.2% SeaKem GTG agarose (FMC) at 6 V/cm for 24 h; the pulse time was increased from 5 to 40 s. Because a single base mutation in the chromosomal DNA of an isolate is sufficient to introduce differences in three fragments in its restriction pattern, isolates with restriction patterns showing the same differences in one to three fragments were considered to belong to the same genotype (23). The PFGE patterns were also analyzed with the computer software Gelcompar for Windows version 3.1b (Applied Math, Kortrijk, Belgium). The PFGE patterns were compared by the algorithmic clustering method called the unweighted-pair group method using arithmetic averages) with the Dice coefficient of similarity (2 × number of matching bands/total number of bands in both strains). Isolates were considered to be within a cluster if the coefficient of similarity was >80%.

Analytical isoelectric focusing was performed with polyacrylamide gels (17) on crude cell-free sonic extracts. β-Lactamase activity was detected by the chromogenic nitrocefin test. Standard enzymes (including TEM-1, TEM-2, TEM-10, SHV-1, SHV-2, SHV-3, SHV-4, and SHV-5) were used as pI markers. In case more than one enzyme was found in the gels, the ESBL enzymes were identified by their ability to hydrolyze cefotaxime or ceftazidime in an agar overlay and by their susceptibility to inhibition by clavulanate (15). Molten Mueller-Hinton agar, containing 0.6% (wt/vol) ceftazidime or cefotaxime, 6% (wt/vol) potassium iodide, and 0.6% (wt/vol) iodine, was poured onto the isoelectric focusing gel and allowed to solidify and form an agar layer about 2 mm thick. The bands corresponding to ESBLs produced clear halos in the black background of agar within 30 min. The formation of these halos was inhibited when the molten Mueller-Hinton agar mixture was supplemented with clavulanate (2 mg/liter).

In case more than one ESBL was found in the same strain, plasmid transconjugation was performed to try to separate the plasmids bearing genes encoding different ESBLs. Escherichia coli K-12 J53-2 rif was used as a recipient for conjugation of the plasmids bearing genes encoding the ESBLs. Transconjugants were selected on plates containing Mueller-Hinton agar supplemented with ceftazidime or cefotaxime (2 mg/liter) and rifampin (250 mg/liter).

Plasmid-borne DNAs of clinical isolates and transconjugants were extracted with a plasmid extraction kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Plasmid DNA electrophoresis was performed with 0.6% agarose gel and visualized with ethidium bromide under UV light.

Plasmid DNA templates for PCR were prepared from the putative ESBL producers and the transconjugants as reported previously (24). PCR was used to amplify an SHV-specific product by using the following primers: 5′-TCAGCGAAAAACACCTTG-3′ and 5′-TCCCGCAGATAAATCACCA-3′. PCR conditions were those suggested by Perkin-Elmer (Applied Biosystems Division, Foster City, Calif.) for preparation of single-stranded DNA for automated sequencing, in which AmpliTaq DNA polymerase was used. Amplification was performed with a Perkin-Elmer DNA thermocycler, model 9600. DNA sequencing of both strands of the PCR products was performed with an ABI PRISM 310 Genetic Auto-analyzer (Perkin-Elmer).

Production of ESBLs was inferred in all 31 clinical isolates of K. pneumoniae on the basis of synergy between ceftazidime and clavulanate by the agar dilution method, double-disk synergy test, and Etest ESBL screen. All of these tests were found to be equally good for the detection of ESBL producers. The production of ESBLs of these isolates was further confirmed by isoelectric focusing and the agar overlay method. The iodine agar overlay method was found to be both sensitive and specific for the detection and preliminary identification of ESBLs (data not shown). About 30% (31 of 104) of the hospital isolates of K. pneumoniae were found to be ESBL producers. The prevalence rate is quite high compared with those of other reports (3.5 to 20%) (6, 14, 16, 22, 25). Until now, no significant information on the prevalance of ESBL-producing strains in other Asian countries has been available.

Because all ESBLs detected in these isolates have pIs above 7.0, we speculated that they are SHV derived. The SHV-specific genes of these isolates were amplified by PCR for DNA sequencing. Twenty-two isolates (71%) were found to harbor the gene encoding SHV-5, and two isolates (6%) were found to harbor the gene encoding SHV-2. Seven isolates were found to produce two unknown ESBLs (with pIs of 7.9 and 7.75). Because their β-lactamase-encoding genes could not be amplified by PCR with SHV-specific primers, they are not SHV derivatives. Plasmid transconjugation failed to separate the plasmids bearing the genes encoding these two enzymes, though these plasmids were successfully transferred to the E. coli recipients. These two enzymes may be encoded by genes located on the same plasmid or they may represent satellite bands of only one enzyme. There were another three isolates which produced an ESBL with a pI of 8.25. Similarly, they are also not SHV derived. The identification of these unknown ESBLs requires further molecular cloning and DNA sequencing.

PFGE analysis of these 31 isolates with XbaI revealed 16 distinct genotypes. The patterns obtained with XbaI are shown in Fig. 1, and the results are summarized in Table 1. Figure 2 is a computer-generated dendrogram that shows relatedness of the isolates by PFGE patterns. Five clusters, each containing isolates with coefficients of similarity of more than 80%, were identified among 20 isolates. A high level of genetic heterogeneity was found among the remaining 11 isolates and between each cluster. In the isolates from the Medical Intensive-Care Unit (MICU) where an outbreak of ESBL-producing K. pneumoniae was suspected, 10 genotypes were identified among 14 clinical isolates. Most strains were genetically unrelated, and no definite epidemic strain was found. On the contrary, the variation in ESBL patterns (Table 1) among these 31 isolates was minor, with only five patterns being identified. This suggests that an increase in the number of ESBL-producing K. pneumoniae isolates in this hospital is due mainly to dissemination of resistance plasmids or mutations in existing plasmid-mediated β-lactamases under the selective pressure produced by the overuse of third-generation cephalosporins. In contrast to our findings, the findings of previous studies indicated that the dissemination of strains producing SHV-derived ESBLs, especially SHV-5, in a hospital was due mainly to clonal spread (1, 2, 10). The epidemiological result of plasmid profile analysis of the clinical isolates was quite compatible with that of PFGE (Table 1). A plasmid of 36 kb was detected in all clinical isolates and transconjugants. This 36-kb conjugative plasmid may be responsible for the production of ESBLs in these strains. This result further supports the hypothesis that the increase of ESBL-producing K. pneumoniae in this hospital is due to dissemination of an epidemic plasmid. TEM-derived ESBLs seem to be rare in K. pneumoniae strains isolated in Taiwan. Our study also confirms that PFGE, when it is combined with plasmid profile analysis, is an effective tool for investigating the epidemiologies of ESBL-producing K. pneumoniae isolates.

FIG. 1.

PFGE of XbaI-digested genomic DNAs from ESBL-producing K. pneumoniae isolates. Lane M contains a lambda ladder (Bio-Rad) which served as molecular size marker; lanes 1 to 31 contain DNA digests of isolates S1 to S31, respectively (see Table 1 for the origins of the isolates).

FIG. 2.

Dendrogram of 31 clinical isolates of ESBL-producing K. pneumoniae based on PFGE results. Strains were clustered by the unweighted-pair group method using arithmetic averages (UPGMA). The scale indicates the percentage of genetic similarity. Max. tol., maximal tolerance in percent of the curve to match bands; Min. surf., minimal surface area of a band.