|Home | About | Journals | Submit | Contact Us | Français|
A collection of 40 confirmed isolates of Citrobacter spp., obtained from the culture of 5,732 clinical samples during 2006 to 2008, was studied for the presence of blaCTX-M, blaTEM, blaSHV, and blaampC by monoplex and multiplex PCRs. These isolates were also looked at for the presence of insertion elements IS26 and orf513, which play an important role in the resistance gene pool. blaCTX-M, blaTEM, blaSHV, and blaampC were noticed in 67.5%, 40%, 25%, and 40% of isolates, respectively. Sequencing for the specific CTX-M type revealed the presence of CTX-M-15-type extended-spectrum beta-lactamases (ESBL). sul1-type integrons were detected in 32.5% (13/40) of isolates, and a single plasmid of ca. ~22 kb was also noticed in all the isolates harboring bla genes. A total of 48.2% (13/27) of isolates harboring blaCTX-M alleles were found to carry IS26 elements and 53.4% (7/13) of isolates harboring the sul1-type integrons were found to carry orf513. Alarmingly, a few isolates simultaneously carried IS26 and orf513, reflecting the occurrence of complex evolution. All the Citrobacter isolates were noticed to produce an amplicon of 400 bp in an IS26 PCR, similar to that of the predominant UK CTX-M-15-producing Escherichia coli clone (clone A) and suggesting a probable genetic relatedness between the two, which could not be proved with British and Indian E. coli isolates in our earlier studies. Here, a first report is being presented describing the occurrence of blaCTX-M-15 in Indian Citrobacter spp. This is also the first report describing the simultaneous occurrence of blaTEM, blaSHV, and blaampC along with blaCTX-M in class 1 integrons. The recent findings indicate that the genetic environment of blaCTX-M-15 has changed through the insertion of IS26 and orf513, although the consequences of these events remain uncertain. It perhaps also suggests a “turnover” of these mobile elements in the population over time.
In recent years, infections with Citrobacter spp. have frequently been increasing as a cause of serious concern, especially for neonates and immunocompromised adults (11, 20). The importance of this infection lies in the fact that Citrobacter spp. tend to produce serious nosocomial outbreaks. A high degree of antibiotic resistance is also noteworthy for Citrobacter freundii and Citrobacter koseri. Globally, including in India, variability in the prevalence levels of Citrobacter infections has been reported (11-13, 16). Numerous international molecular studies have also described the occurrence of various CTX-M types in Citrobacter, viz., CTX-M-9 in the United Kingdom (14), CTX-M-30 in Canada (1), CTX-M-3 in Korea (10), CTX-M-2 in Japan (15), and CTX-M-14 in China (29). Nevertheless, it is unfortunate that no systematic molecular study describing at least the occurrence of blaCTX-M genes in Citrobacter spp. has ever been published in India, despite the fact that the first report of the detection of blaCTX-M-15 was from a small collection of Indian Enterobacteriaceae (9). Many plasmid-mediated beta-lactamase genes are located within or near mobile elements, such as integrons or transposons, which enhance their dissemination (8, 22). There are significant numbers of reports describing the presence of class 1 integrons in other bacterial species (19); however, the reports describing their occurrences in Citrobacter spp. are fragmentary (2). Moreover, we could not get any report describing the presence of blaCTX-M-15 in integrons in Enterobacteriaceae, including Citrobacter spp. Insertion sequence IS26 has also been reported in association with bla-CTX-M, including blaCTX-M-15 in Enterobacteriaceae, and, more specifically, has been found to insert within ISEcp1, although, interestingly, the insertion of IS26 differs from strain to strain (6). Frequent association of orf513 with sul1-type integrons has also been reported. However, the proper evaluation of these mobile elements is lacking for Citrobacter spp.
In light of the above facts, the present extensive study was planned to determine the prevalence of Citrobacter infections, their antibiotic susceptibility profiles, the occurrence of blaCTX-M, blaTEM, blaSHV, and blaampC, the presence of mobile elements, like sul1-type integrons and IS26 elements, and also the association of orf513 with sul1 integrons in Citrobacter isolates.
The present prospective study was carried out in the Department of Microbiology of Jawaharlal Nehru Medical College and Hospital, Aligarh Muslim University, U.P., India, over the period from September 2007 to February 2008.
From September 2007 to February 2008, a total of 5,732 clinical samples were subjected to routine culture and antimicrobial susceptibility testing. Of these, 1,763 samples yielded growths of Gram-negative bacteria, of which 37 produced growth of Citrobacter spp. (29 C. koseri isolates and 8 C. freundii isolates). Three (2 C. koseri and 1 C. freundii) isolates randomly collected during 2006 were also included for further analyses and molecular detection of bla alleles. All 40 isolates were identified by standard microbiological techniques (3, 7).
Antimicrobial susceptibility testing was carried out on Mueller-Hinton agar (HiMedia, India) by the standard disk diffusion method per the Clinical and Laboratory Standards Institute (CLSI; formerly NCCLS) guidelines (18). Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as controls. The antibiotics used and their amounts were as follows: cefotaxime, 30 μg; ceftazidime, 30 μg; cefoperazone, 75 μg; ceftriaxone, 30 μg; cefpirome, 30 μg; cefoxitin, 30 μg; aztreonam, 30 μg; gentamicin, 10 μg; ciprofloxacin, 1 μg; imipenem, 10 μg; piperacillin, 100 μg; piperacillin-tazobactam, 100/10 μg; cefoperazone-sulbactam, 75/10 μg; and nitrofurantoin, 300 μg. The antibiotic disks used were from HiMedia Laboratories, Ltd., India. Susceptibility to nitrofurantoin was tested only with urinary isolates. MICs for cefotaxime and ceftazidime were noted according to CLSI guidelines (17).
The potential presence of ESBLs was inferred phenotypically by combination disc methods using discs of cefoperazone versus cefoperazone-sulbactam and piperacillin versus piperacillin-tazobactam. An increase in zone diameters of ≥5 mm and ≥8 mm in combination discs containing inhibitors (cefoperazone-sulbactam and piperacillin-tazobactam, respectively) in comparison to their respective beta-lactams alone (cefoperazone and piperacillin) was taken as a positive result indicating the presence of ESBLs, as adopted in our previous study (23).
All 40 isolates were initially screened for the presence of blaCTX-M, blaTEM, blaSHV, and blaampC by PCR using a set of universal primers (Table (Table1)1) responsible for yielding amplicons of 593 bp, 936 bp, 930 bp, and 634 bp, respectively. The cycling conditions for detection of blaampC, blaTEM, and blaSHV were as follows: initial denaturation at 95°C for 15 min, 35 cycles of 94°C for 1 min, 58°C for 2 min and 72°C for 3 min, and a final elongation at 72°C for 10 min. The cycling conditions for blaCTX-M detection were initial denaturation at 94°C for 7 min, 35 cycles of 94°C for 50 s, 50°C for 40 s and 72°C for 1 min, and a final elongation at 72°C for 5 min. Subsequently, the bacterial isolates demonstrating the presence of blaCTX-M and blaampC alleles were characterized to the specific CTX-M genogroup and AmpC family by using the multiplex PCR protocol of Woodford et al. (27) and Perez-Perez and Hanson (21), as adopted in our previous studies (26).
Sequencing in the representative isolates was performed courtesy of Chromous Biotech, Ltd., Banglore, India.
sul1-type class 1 integrons were detected by PCR using 5′-CTT CGA TGA GAG CCG GCG GC-3′ and 5′-GCA AGG CGG AAA CCC GCG CC-3′ as forward and reverse primers, respectively. Cycling conditions were initial denaturation at 95°C for 3 min, 30 cycles of 95°C for 1 min, 63°C for 1 min and 72°C for 2 min, and a final elongation at 72°C for 10 min.
Recently, the presence of IS26 in the predominant United Kingdom CTX-M-15-producing E. coli clone, designated clone A, was reported (28). Subsequently, the high mobility and diversity of IS26 insertion into tnpA of ISEcpI were demonstrated in Indian Enterobacteriaceae (6). To determine whether Indian Citrobacter isolates also carried this particular insertion element, all CTX-M-15-carrying isolates were screened using the PCR protocol described by Woodford et al. (28).
The presence of orf513 was detected in isolates harboring sul1-type integrons by PCR using 5′-CTC ACG CCC TGG CAA GGT TT-3′ and 5′-CTT TTG CCC TAG CTG CGG T-3′ as forward and reverse primers, respectively. Cycling conditions were initial denaturation at 95°C for 3 min, 35 cycles of 95°C for 1 min, 58°C for 1 min and 72°C for 2 min, and a final elongation at 72°C for 10 min.
Plasmid analysis of the isolates harboring blaCTX-M was done by using the large-scale alkaline lysis method of Davis et al. (4), as adopted in our previous studies (24, 25). For demonstration of the presence of bla genes on plasmids, we adopted a direct approach by performing PCR utilizing the purified plasmids as templates. Plasmid samples were denatured at 95°C for 10 min prior to being used as templates. Plasmids obtained from other clinical isolates and found negative for respective bla genes in screening reactions were used as negative controls.
Randomly amplified polymorphic DNA (RAPD) typing of the isolates carrying blaCTX-M was done, as described previously (6, 26), to ascertain any specific clone circulating in the survey population. The results were analyzed using the gel documentation system (Bio-Rad), and the clustering was performed using the Quantity One software provided with the gel documentation system.
Of the various Gram-negative bacteria isolated from 2007 to 2008, a total of 2.1% (37/1,763) of isolates yielded growth of Citrobacter species, of which 29 were C. koseri and 8 were C. freundii.
The results for the antibiotic resistance rates are shown in Table Table2.2. All the isolates were multidrug resistant. Among the third-generation cephalosporins tested, maximal resistance was noticed against cefotaxime and cefoperazone in 80% of isolates each, while resistance to ceftazidime was recorded for 70% of isolates. More interestingly, the resistance to ceftriaxone was noticed only in 57.5% of isolates in vitro. The MICs of the resistant isolates for ceftazidime ranged between 16 and >1,024 mg/liter, whereas for cefotaxime they ranged between 64 and >1,024 mg/liter ( Table Table3).3). Concomitant high resistance to gentamicin, aztreonam, and ciprofloxacin was noticed in 70%, 70%, and 67.5% of isolates, respectively. Three isolates (7.5%) also demonstrated resistance to imipenem (C. freundii, 11.1% [1/9]; C. koseri, 6.45% [2/31]). The isolates were also examined for the presence of carbapenamases IMP, VIM, GIM, and SIM by PCR and were found negative by PCR (data not shown). The testing of isolates could not be attempted for KPC detection due to the nonavailability of control strains.
ESBLs were phenotypically detected in 80% (32/40) of isolates using piperacillin-tazobactam as a combination disk. All but one isolate were also detected by the cefoperazone-sulbactam combination disk method.
Out of 40 isolates studied, a total of 27 (67.5%) isolates demonstrated the presence of blaCTX-M in a monoplex PCR. In the collection of 37 Citrobacter isolates obtained from 2007 to 2008, a total of 25 (20 C. koseri and 5 C. freundii) isolates showed the presence of blaCTX-M, whereas in the 2006 collection of 3 isolates, 2 showed the presence of blaCTX-M, and both were C. koseri isolates.
By using the multiplex protocol of Woodford et al. (28), all the blaCTX-M-harboring isolates were found to possess group 1 CTX-M (Fig. (Fig.1A).1A). Detailed results demonstrating the occurrence of various combinations of genes are shown in Table Table44.
A total of 16 (40%), 16 (40%), and 10 (25%) isolates demonstrated the presence of blaampC, blaTEM, and blaSHV, respectively (Fig. 1B to D).
Based on multiplex PCR, all isolates harboring blaampC were found to belong to the CIT family of AmpC beta-lactamases.
Sequencing of the representative isolates (GenBank accession no. GQ117121 and GQ117122) revealed the presence of CTX-M-15-like beta-lactamases.
While screening the isolates harboring blaCTX-M alleles, a total of 48.2% (13/27) (1 C. freundii and 12 C. koseri) isolates were found to possess IS26 elements. Interestingly, no diversity in the insertion points of these elements was noticed in Citrobacter spp., as found previously with Indian E. coli and Klebsiella species. Rather, all the isolates were observed to produce an amplicon of 400 bp similar to that of the predominant United Kingdom CTX-M-15-producing E. coli clone (clone A).
Plasmid analysis of all the isolates harboring bla genes demonstrated the presence of a single plasmid of ca. ~22 kb (Fig. (Fig.2).2). A further PCR utilizing these plasmids as templates demonstrated the presence of respective bla genes on these plasmids.
sul1-type integrons were detected in 32.5% (13/40) (1 C. freundii and 12 C. koseri) of isolates (Fig. (Fig.3).3). When comparing the occurrence of sul1-type integrons and the presence of plasmids, 9 isolates demonstrated the presence of integrons and plasmids simultaneously, indicating the presence of integrons on plasmids, and 4 isolates (1 isolate having CTX-M only, 1 CTX-M + SHV + TEM, 1 CTX-M + AmpC + SHV + TEM, and 1 CTX-M + AmpC + TEM) were noticed to possess only integrons, indicating the presence of integrons on bacterial chromosomes.
A total of 53.4% (7/13) of isolates harboring sul1-type integrons were found to carry orf513. Six isolates were C. koseri, of which 2 isolates also carried the IS26 element. One isolate of C. freundii carrying orf513 was also found to carry the IS26 element.
RAPD typing revealed the presence of a diverse bacterial population, and no predominant clone was identified in the present Indian collection of Citrobacter spp. (Fig. (Fig.44).
A prevalence rate of 2.1% for Citrobacter isolation was observed in our North Indian tertiary hospital, calculated against the Gram-negative bacterial isolation at our hospital setup. This rate corresponded with some of the recent and previously published Indian and international reports (12, 13). On the other hand, a group at the Seattle Veterans Administration Medical Center, Washington, reported an occurrence rate of 6% of all Enterobacteriaceae isolated from the hospital (11).
We did find C. koseri to be a predominant isolate, as reported by other workers (5, 11). A high degree of resistance to third- and fourth-generation cephalosporins, as well as to piperacillin, gentamicin, and ciprofloxacin, was observed in this study. Despite such high resistance to cephalosporins, including cefotaxime, being reported in numerous Indian reports, none has even looked for the occurrence of blaCTX-M in those isolates. In the present study we not only looked at but also characterized the genogroup of CTX-M beta-lactamases by monoplex and multiplex PCR, respectively, and also the specific type detected by sequencing the representative isolates. A total of 67.5% of isolates harbored blaCTX-M, and all belonged to genogroup 1. This finding is important in light of the fact that blaCTX-M of genogroup 1 was earlier reported to be prevalent, and the only type, in a collection of E. coli and Klebsiella pneumoniae isolates from our institution (18). During that study we also noticed the occurrence of blaCTX-M in 2 out of 6 Citrobacter spp. tested (unpublished data), and, unfortunately, those isolates could not be processed further to characterize specific CTX-M since the original stock, transported to Birmingham, United Kingdom, was contaminated by an E. coli isolate. The issue of the real presence of CTX-M enzymes in Citrobacter also remained unaddressed due to this contamination. This is, therefore, the first report of the occurrence of blaCTX-M in Citrobacter spp. at such a high frequency on the Indian subcontinent. This finding is important because the first report of CTX-M-15 (genogroup 1) was with a small collection of bacteria from New Delhi, India, in 1999 (9) and because subsequently our report included a larger collection of E. coli and K. pneumoniae isolates from 2003 to 2005 (6), and it is plausible that these genes may have mobilized from these coliforms to Citrobacter. While analyzing our representative isolates of Citrobacter by sequencing, we found the presence of CTX-M-15 β-lactamases. We extended our study by looking for blaampC, blaTEM, and blaSHV in Citrobacter and found these bla genes in 40%, 40%, and 25% of isolates, respectively. Interestingly, simultaneous occurrences of various bla genes were noticed in the majority of the isolates; the most common combination was blaCTX-M + blaampC + blaTEM, followed by blaCTX-M + blaampC + blaTEM + blaSHV in 7 and 5 isolates, respectively (Table (Table44 shows the distribution of various bla genes).
By performing PCR, sul1-type integrons have been observed in 32.5% (13/40) of the isolates. sul1-type integrons were common in isolates harboring simultaneously blaCTX-M and blaampC (69.2%; 9/13), followed by isolates harboring blaCTX-M and devoid of blaampC (46.2%; 6/13). None of the isolates harboring only blaampC was found to have sul1 integrons.
Since we noticed the presence of the blaCTX-M alleles on a plasmid of ca. ~23 kb in our earlier collection of E. coli and K. pneumoniae isolates (26), this prompted us to look for the presence of plasmids in Citrobacter also. Interestingly, we found the presence of single plasmids of almost similar size in Citrobacter too. Again, by performing the PCRs for respective bla genes using the purified plasmids as templates, we demonstrated their presence on plasmids of Citrobacter spp. The dissemination of plasmids carrying these structures may explain the widespread distribution of these bla genes in Citrobacter spp., and we believe that these blaCTX-M genes and other bla genes in Indian Citrobacter spp. have probably mobilized through plasmids present in E. coli and K. pneumoniae in the human gut and that the two point mutations observed with blaCTX-M of Citrobacter reflect the phenomenon of evolution over the course of time.
The apparent acquisition of IS26 and orf513 by the more recent examples of integrons and plasmids in Indian Citrobacter strains is interesting. It perhaps suggests a “turnover” of these mobile elements in the population over time, the IS26 and orf513 insertion creating a selective advantage in those strains carrying these elements. Although the numbers are small, the geographical consistency of the observation could make it broadly applicable and worthy of further investigation.
In a nutshell, this is the first report indicating the occurrence of blaCTX-M at such a high frequency in Citrobacter spp. globally and in particular on the Indian subcontinent. This is also the first report describing the simultaneous occurrences of blaTEM, blaSHV, and blaampC along with blaCTX-M on class 1 integrons. The observed mutations in blaCTX-M-15 from these isolates are alarming and warrant urgently the surveillance of Citrobacter spp., bearing in mind the very large population in this region and the high prevalence of ESBL genes in Enterobacteriaceae, which represent a large, relatively poorly studied reservoir compared with Europe and the Americas. Moreover, due to increased transcontinental movement of the human population, it would be wise to know the prevalence and resistance complexity of these strains well in advance, so as to formulate a policy for empirical therapy. Since the Citrobacter isolates were observed to produce in an IS26 PCR an amplicon of 400 bp similar to that of the predominant United Kingdom CTX-M-15-producing E. coli clone (clone A), this suggests a probable genetic relatedness between the two which could not be proved with British and Indian E. coli isolates in our earlier studies and warrants further studies along this line. This study also urges the implementation of strict national and international antibiotic policies, keeping in mind the wide mobility of the human population globally.
M. Shahid is grateful to the Department of Science and Technology, Ministry of Science and Technology, Government of India, for the award of Young Scientist Project (SR/FT/L-111/2006).
Published ahead of print on 10 March 2010.