CTX-M type extended-spectrum β-lactamase in Escherichia coli isolated from extra-intestinal infections in a tertiary care hospital in south India.

Background & objectives: Infections caused by extended-spectrum β-lactamase (ESBL)-producing Escherichia coli carrying blaCTX-M genes have been spreading globally, but there are geographical variations in the type of blaCTX-Mgenes prevalent and there are scanty data from India. This study was conducted to determine the CTX-M type ESBLs in E. coli isolates obtained from clinical specimens from patients with extra-intestinal infections attending a tertiary care hospital in south India.
Methods: ESBL-producing E. coli isolated from patients with extra-intestinal infections were subjected to PCR using CTX-M group-specific primers. From a representative isolate, full-length CTX-M-15 gene was amplified and sequenced. An internal fragment of this gene was sequenced in 10 representative isolates.
Results: Of the 300 isolates of E. coli tested, 88 per cent carried CTX-M genes and blaCTX-M-15was the most dominant gene present in 90 per cent of the positive isolates. Most (91%) of the isolates positive for blaCTX-M were sensitive to meropenem. Interpretation & conclusions: Our findings showed blaCTX-M-15 to be the dominant gene. Based on the data on antimicrobial susceptibility, cefoperazone-sulbactum could be an antimicrobial of choice.

Extended-spectrum β-lactamases (ESBLs) are the most prevalent type of antimicrobial resistance mechanisms among Enterobacteriaceae such as Escherichia coli and Klebsiella pneumoniae1. There are over 200 enzymes characterized to be in the ESBL spectrum, which are encoded by plasmid or chromosomally carried genes1. Several studies indicate that CTX-M-type ESBLs are spreading globally and becoming dominant types in Enterobacteriaceae in many countries23. However, the diversity of CTX-M types occurring in India is yet to be fully understood. There are only a few reports of molecular identification of β-lactamases. For example, a few studies indicated that CTX-M-15 could be the predominant ESBL in northern India45. Report from south India56 has indicated the prevalence of CTX-M-1-type gene in 58.3 per cent of Klebsiella spp. and in 36 per cent of E. coli. This study was undertaken to estimate the presence of CTX-M type ESBLs in E. coli isolated from samples of extra-intestinal infections in patients attending a tertiary care hospital in south India.

Material & Methods

Three hundred isolates of E. coli obtained from urine, wound swab, sputum, pus, endotracheal secretions, bronchoalveolar lavage, bile fluids and other fluids from sterile body sites such as pleural fluid, bile, peritoneal fluid and blood cultures were collected over two and a half years (2013-2016) in the Microbiology department of The Madras Medical Mission, a tertiary care hospital in Chennai, India. Standard clinical microbiological procedures were used for the isolation of E. coli7 and identification was done through VITEK II compact (BioMérieux, France). Antimicrobial susceptibility testing was done in Mueller-Hinton agar according to the current Clinical and Laboratory Standards Institute (CLSI) guidelines891011.

Phenotypic testing for ESBLs: ESBL phenotypic screening was performed for all the isolates by the double-disk diffusion test using ceftazidime (30 μg) and ceftazidime/clavulanic acid891011. E. coli ATCC 25922 and E. coli ATCC BAA-2326 were used as the negative and positive controls and the zone diameter interpreted as per the CLSI M-100 recommended guidelines891011. ESBL phenotypic confirmatory test was performed by the double-disk diffusion method using antibiotic disks containing a combination of cephalosporin plus clavulanic acid in combination with a corresponding cephalosporin disk alone891011.

Storage of isolates: Isolates were transferred to semisolid agar and stored at −20°C. For long-term storage, cultures were suspended in Luria-Bertani (LB) broth (Himedia, India) with 30 per cent glycerol and stored at −80°C.

Molecular testing for CTX-M: As per the antibiogram analyzed over 2013 to 2016 the rate of ESBL-positive organisms was 70-80 per cent among various isolates from The Madras Medical Mission. All isolates were subjected to (PCR) using four sets of primers and PCR conditions described earlier for CTX-M groups12.

Sequencing of β-lactamase gene from a typical strain: To amplify the entire β-lactamase gene of Group I (CTX-M-15), the sequence from E. coli strain KS127 (accession number AB976567.1) was accessed from GenBank (). Primer design was based on primer search using NCBI Primer BLAST (). The annealing temperature used for amplification was 55°C. An isolate of E. coli, IRU1638, recovered from a patient with urinary tract infection was used to sequence the β-lactamase gene.

Sequencing internal region of β-lactamase of a few representative isolates: Ten representative isolates were selected based on source of isolation (7 urine, 2 blood and 1 bile fluid) and varying antibiogram for sequencing of internal region of blaCTX-M-15. The forward primer 5’ACGTTAAACACCGCCATTCC3’ and the reverse primer 5’TCGGTGACGATTTTAGCCGC3’ amplified a 356 bp fragment of the CTX-M-15 β-lactamase gene. PCR products obtained at 56°C were sequenced by Eurofins, Bengaluru, and the sequence was subjected to nucleotide BLAST search.

Results & Discussion

Of the 300 isolates tested, 238 (79.3%) were positive for Group I CTX-M-15 (Table I), and of these, majority 174 (73.1%) were urine isolates. Group IV had eight isolates and Group II and III had only one isolate each. Seventeen isolates showed mixed reactions, nine being positive for both Groups I and II, four with both Groups I and III and three isolates with Groups I and IV (Table I). Urine isolates accounted for 73 per cent of CTX-M Group I, while among CTX-M-negative isolates, urine accounted for 63 per cent (Table I). In a study of 140 Enterobacteriaceae ESBL producers from eastern India13, the most common gene was blaTEM (96.42%) followed by blaCTX-M (75%) and blaSHV (17.85%). Another study from Central India14 found that among the 526 urinary isolates of E. coli, the most common resistance gene detected was blaTEM (48.7%) followed by blaCTX-M (7.6%). Our study showed that 88 per cent of the E. coli isolated from extra-intestinal infections carried blaCTX-M (Table I). This finding was similar to a study from Chennai15 which looked at E. coli in HIV patients and detected blaCTX-M in 70.2 per cent of the isolates.

Table I

Prevalence of different groups of blaCTX-M in different clinical isolates of Escherichia coli

CTX-M types	Total number positive	Distribution in various samples (number positive)
Group I	238	Urine – 174 Blood – 22 Pus – 24 Sterile fluids – 9 Respiratory: 9
Group II	1	Urine - 1
Group III	1	Pus - 1
Group IV	8	Urine – 4 Pus – 1 Blood - 3
Group I and II	9	Urine – 6 Bile – 2 Sterile fluid - 1
Group I and III	4	Urine – 3 Blood - 1
Group I and Group IV	3	Urine - 3
Group I and II and IV	1	Blood - 1
Negative for all groups	35	Urine – 22 Pus – 5 Blood – 2 Fluid – 4 Respiratory specimens - 2

Our study also demonstrated that blaCTX-M Group I was the dominant group among extra-intestinal E. coli isolates. A study from Korea16 noted that of the 80 blaCTX-M carrying extra-intestinal E. coli studied, 36 carried blaCTX-M-15 (Group I), while 46 carried blaCTX-M-14 (Group IV). Predominance of Group IV over Group I has also been reported from China17. Among 201 E. coli isolates studied from Shandong Province in China, 116 (57.7%) carried blaCTX-M-14 (Group IV) and only 31 (15.4%) carried blaCTX-M-15 (Group I). In river water in India, blaCTX-M15 accounted for 46 per cent of isolates while 32 per cent isolates were positive for blaTEM18.

The antibiogram pattern of isolates positive and negative for blaCTX-M is indicated in Table II. CTX-M-positive isolates showed lesser sensitivity to clavulanic acid (7%) in comparison to sulbactam (62%) and tazobactam (55%). Majority (91%) of blaCTX-M-positive isolates showed meropenem sensitivity compared to those negative for this gene (72%).

Table II

Comparison of sensitivity of isolates positive for blaCTX-M and those negative for this gene

Antibiotic tested	Per cent sensitivity in CTX-M positive	Per cent sensitivity in all negative
Amikacin	94	93
Amoxicillin-clavulanic acid	7	13
Cefepime	4	3.4
Cefoperazone-sulbactam	62	55
Ciprofloxacin	18	13
Ertapenem	81	72
Meropenem	91	72
Imipenem	94	80
Piperacillin-tazobactam	55	54

The full-length blaCTX-M-15 was amplified and sequenced in one typical isolate and a 356 bp internal fragment was amplified and sequenced in 10 isolates. BLAST analysis showed that these sequences had 100 per cent similarity to blaCTX-M-15 from E. coli and Klebsiella spp. in GenBank. These data confirmed the results obtained with group-specific primers. To conclude, the most common type of ESBL in this small single-centre study was identified to be CTX-M-15. Cefoperazone-sulbactam may be a better choice than piperacillin-tazobactam in our setting. Further studies need to be done on a large number of isolates in different parts of the country.