Collection of samples

A total of 600 samples comprised of 400 raw bovine milk, 100 faeces, 40 feed, 40 environmental swabs and 20 water samples from 20 different bovine dairy farms in and around Hisar city, Haryana, India were collected under aseptic conditions in sterile containers. The samples were collected from different villages of Hisar viz. Village Post Office (V.P.O.) Dabra, Mirkan, Ladwa, Satrod, Niayana, Kamiri and from Sirsa city. Five apparently healthy lactating animals from each farm were selected randomly and included in the study. From each animal, raw milk samples were collected into sterile tubes from four quarters (10 ml each) separately. However, the faecal samples were collected by per rectal method. Two samples of feed and one sample of water also collected from each farm. Sterile cotton swabs moistened with normal saline were used for sampling of environment and from each farm, 2 such samples were taken from clean floor or wall of farms. The details of farms such as managemental practice, milking method, hygienic condition, housing system, water source, use of antimicrobial medicines at farm were collected in the questionnaire designed. All the samples were labelled properly and transported in ice box to the College Central Laboaratory, College of Veterinary Sciences, Lala Lajpat Rai University of Veterinary and Animal Sciences, Hisar, Haryana for bacteriological analysis. The ethical approval was obtained from the competent authority (IAEC/LUVAS/28/20).

Isolation and identification of E. coli

All the samples except faeces were pre enriched in Buffered Peptone Water (BPW) (1:10 ratio) at 37°C for 18 hours, while a loopful of faecal sample was directly streaked onto MacConkey agar (MLA). From BPW, a loopful of enriched inoculum was streaked onto MLA and incubated at 37°C for 18 to 24 hours and a single colony was obtained. The lactose fermenting pink to red colonies were further streaked onto Eosin Methylene Blue agar (EMB) and incubated overnight at 37°C. The dark centred colonies with metallic sheen were considered as E. coli presumptively. These isolates were further characterized by Gram staining and biochemical reactions viz. IMViC (Indole, Methyl red, Voges Proskauer, Citrate), catalase and oxidase tests.⁷ All the bacteriological media and reagents used for isolation were procured from Hi Media laboratories Pvt. Ltd, Mumbai, India.

Phenotypic detection of ESBL production

The E. coli isolates were checked for ESBL production on HiCrome ESBL agar (Hi Media, Mumbai, India) and incubated at 37°C for 24 hours. The colonies with purple colour were tentatively identified as ESBL producer. The ESBL producing isolates were further verified by Double-Disc Synergy Test (DDST) using ceftazidime (CAZ-30 µg) and ceftazidime with clavulanic acid (CAC-30/10 µg) as well as cefotaxime (CTX, 30 µg) and cefotaxime-clavulanate (CEC, 30/10 µg) discs. Briefly, E. coli isolates were inoculated into nutrient broth and incubated at 37°C. The bacterial culture with a turbidity equivalent to 0.5 Mac Farland standard unit was inoculated onto Mueller Hinton Agar (MHA) plates by spread plate method. The antibiotic discs were placed on the inoculated MHA plates at a distance of 20 mm apart and incubated overnight at 37°C. The inhibition zone diameter was measured for each antibiotic disc and its respective clavulanic acid containing discs. A difference of ⩾5 mm in the presence of clavulanic acid when compared to its absence was considered as positive for the production of ESBL.⁸

Determination of minimum inhibitory concentration

All the isolates which were positive for ESBL production on HiCrome ESBL agar were subjected to Minimum Inhibitory Concentration (MIC) test using MIC strips (EM132, Hi Media laboratories Ltd., Mumbai). The ESBL MIC (MIX+/MIX) strip was placed in the centre of MHA plates. The MIC values were noted at the point where ellipse of inhibition intersects the MIC scale on the strip. A ratio of MIX/MIX+ ⩾ 8 or ellipse of inhibition intersecting only on MIX+ side was considered as ESBL positive isolate.

Antimicrobial susceptibility testing

All the ESBL positive E. coli isolates obtained from various samples were subjected to antimicrobial susceptibility test according to CLSI guidelines.⁸ Cultures of all ESBL E. coli isolates were grown in nutrient broth upto 0.5 Mac Farland standard unit and tested against 16 antibiotics from 6 different classes on MHA plates. The antibiotics included were amikacin (30 µg), ampicillin (10 µg), amoxicillin (10 µg), cefoperazone (75 µg), ceftazidim (30 µg), cefotaxime (10 µg), ceftriaxone (30 µg), chloramphenicol (30 µg), enrofloxacin (10 µg), gentamicin (10 µg), kanamycin (30 µg), levofloxacin (5 µg), moxifloxacin (5 µg), neomycin (30 µg), oxytetracycline (30 µg) and streptomycin (10 µg).

Molecular characterization of the isolates

The phenotypically confirmed E. coli isolates were screened for specific genetic elements by molecular method. DNA extraction of the isolates was carried out by using QIAamp DNA blood mini kit (QIAGEN, Germany). E. coli isolates were confirmed by PCR targeting uidA gene using published primers.⁹ The reaction mixture was prepared using 12.5 µl of master mix (Promega, USA), 1 µl of each primer (10 µm), 2 µl DNA template and nuclease free water to make the final volume up to 25 µl. The E. coli isolates were also investigated for the presence of β-lactamase (bla) encoding genes (bla_CTX-M, bla_TEM, bla_SHV)¹⁰ by simplex PCR. The ATCC strains E. coli ATCC 25922 and K. pneumoniae ATCC 700603 strains were used as negative and positive control strains, respectively. The details of primers used in this study and the thermal profiling for different genes are given in Table 1. The amplified product was resolved on electrophoresis gel (1.5%) containing ethidium bromide (Sigma, USA) at a concentration of 0.5 µg/ml and the gel was run at 75 V for 1.5 hours using submarine horizontal electrophoresis system (Genetix Biotech Asia Pvt. Ltd, New Delhi) in 1× TAE buffer. The amplified PCR product was visualized using gel documentation system (Azure Biosystems, USA) and compared with the 100 bp DNA ladder (Promega, USA) as size marker.

Table 1.

Details of primers and thermal profiling of genes.

Statistical analysis

The farm data were analysed to find out the risk factors associated with presence of ESBL E. coli in dairy farms. Statistical analysis was done using chi-squared test at 95% confidence interval (IBM SPSS statistics version 21, New York). The difference was considered as statistically significant at P ⩽ .01 between variables.

Phenotypic characterization of ESBL E. coli

Out of 600 samples, a total 246 samples were shown pink colour colonies on MLA and greenish metallic sheen on EMB agar was observed from 240 samples. These colonies were biochemically characterized by IMViC tests and Gram’s staining. Pink coloured rods were observed upon Gram’s staining. All the isolates were positive for Indole and Methyl red tests. They were negative for Voges-Proskauer and Citrate test. All the phenotypically positive E. coli isolates were shown the presence of uidA gene (Figure 1).

Figure 1.

Agarose gel showing uidA gene amplified from E. coli isolates.

Lane M – 100 bp DNA ladder.

Lane 1 – Positive control.

Lane 2 – Negative control.

Lanes 3 to 6 – Samples.

Out of 240 isolates of E. coli, 74 (12.3%) isolates were found to be ESBL producers based on the growth on HiCrome ESBL agar (Figure 2). Maximum number of ESBL isolates were obtained from faeces (40.5%) followed by raw milk (37.8%), environmental swabs (17.5%), water (2.9%) and feed (1.3%) samples. All the isolates were given positive Double-Disc Synergy Test for CTX/CEC combination, but 91.89% isolates were found to be positive for CAZ/CAC combination (Figure 3), thus confirmed the production of ESBL in these isolates phenotypically. All the isolates showed zone of inhibition only on MIX+ side as it contained clavulanic acid and no zone of inhibition was found on MIX side (Figure 4). The MIC values of these isolates were noted and were ranged from 0.25 to 0.048 (Figure 5).

Figure 2.

ESBL producing E. coli on HiCrome ESBL agar.

Figure 3.

Double disc synergy test of ESBL E. coli.

Figure 4.

Minimum inhibitory concentration of ESBL E. coli.

Figure 5.

MIC values of ESBL E. coli isolates.

There is a wide variation in detection of ESBL E. coli in raw milk from different States of India. India is primarily divided into 6 physiographic areas, each of which has a distinct climate and culture. The husbandry methods also differ significantly from east to west and north to south. There are several factors that cause variations in the incidence of extended-spectrum beta-lactamases in different States and zones of India. The probable reason of these variations could include the sampling period, geographic location, antibiotic class and frequency of use in various zones, virulence factors, species variances, AMR patterns of each region etc. The meta-analysis study from India reported difference in ESBL prevalence percentages from north, east, south and central zones.¹¹ Comparable to the current investigation, 37.0% and 40.0% of ESBL E. coli from the raw milk and faeces, respectively is reported from Chhattisgarh, India.¹² There is no report available regarding the presence of ESBL E. coli from Indian dairy farm environments. However, the results were in agreement with studies of foreign countries. A prevalence of 10.0% and 5.0% of ESBL-producing E. coli from the animal drinking water and feed, respectively is reported in Indonesia.¹³ An occurrence of 28.6% ESBL E. coli from the environmental samples in Egypt is reported.¹⁴ A very high prevalence of 88.7% was noticed in Chiang Mai from the waste water of dairy farms¹⁵ and 13.0% ESBL E. coli occurrence from environmental samples.¹⁶

Antimicrobial susceptibility pattern

Most of the ESBL E. coli isolates were sensitive to chloramphenicol (82.4%) followed by gentamicin (77.0%) in antibiotic susceptibility test. These antibiotics could be a alternative drug of choice for treating the infections caused by ESBL E. coli. However, the lowest resistance to these antibiotics seems to be associated with the restricted usage along with the reduced preference of these drugs in the treatment of infections. Complete resistance (100.0%) was observed against cephalosporins class of antibiotics. This can be due to the production of β-lactamases which destroy the β-lactam rings of these antibiotics. A very high level of resistance was observed for fluoroquinolones also. This may be due to the co-existence of ESBL with plasmid-mediated quinolone resistance genes.¹⁷ The isolates from raw milk and faeces were 14.2% and 6.6% sensitive to enrofloxacin, respectively, while all the isolates of environmental swabs were 100% resistant. Similar findings were reported in other studies as well.^18,19 All the isolates of ESBL E. coli were found as multidrug resistant, with resistance to 3 or more antibiotics from 3 different antibiotic classes.²⁰ The antibiotic sensitivity pattern of isolates from various samples is given in Figure 6.

Figure 6.

Antibiotic sensitivity pattern of ESBL E. coli isolates.

Molecular characterization of ESBL E. coli

Presence of any one or more β-lactamase genes viz. bla_CTX-M, bla_TEM and bla_SHV were detected in 65 isolates, whereas 9 isolates were negative for all the β- lactamase genes investigated in this study. The bla_CTX-M gene was detected in 39 isolates (52.0%) and bla_TEM in 7 (9.45%) isolates. However, no sample possessed bla_SHV gene alone by simplex PCR. The co-expression of genes bla_CTX-M + bla_TEM was observed in 16 isolates (21.6%) and bla_CTX-M + bla_SHV in 3 (4.05%) isolates. The gene combinations of bla_TEM + bla_SHV as well as bla_CTX-M + bla_TEM + bla_SHV were not detected in this study. The PCR amplified products of bla genes and the source wise distribution of β-lactamase genes are given in Figures 7Figure 8. to 9.

Figure 7.

Agarose gel showing bla_CTX-M gene amplified from E. coli isolates.

Figure 8.

Agarose gel showing bla_TEM gene amplified from E. coli isolates.

Figure 9.

Source wise distribution of β-lactamase genes of ESBL E. coli.

Concerns have been raised in both human and veterinary medicine over the swift expansion and global dissemination of bla genes in ESBL E. coli. These genes can be transmitted to humans and animals, through the bovine milk, faeces and farming environments. It has been discovered that plasmids carrying bla genes are capable of transferring these determinants to other commensal bacteria or pathogens.²¹ There have been varying reports nationwide on the presence of bla_CTX-M in isolates of E. coli derived from milk. A higher carriage (54.54%) of bla_CTX-M gene from West Bengal²² is found. However, a low prevalence (2.09% and 3.12%) of bla_CTX in milk is also reported from Chhattisgarh and Orissa, respectively.^12,23

The findings of the current study regarding the detection of bla_CTX-M in faeces were comparable to that from Assam, who reported 28.57% bla_CTX-M in E. coli from cattle faeces.²⁴ Comparable to the current investigation, 19.0% prevalence of bla_CTX-M gene from feed of cattle in Bhopal is reported.²⁵ A prevalence rate of 5.4% and 93.4% is found for bla_CTX-M gene from the dairy farm environment, respectively.^26,27 A 95.9% prevalence of bla_CTX-M from raw milk and environmental samples is recently reported.²⁸

The current study’s percent prevalence for bla_TEM was higher than that of previous reports. A low prevalence of 3.66% was reported for bla_TEM in raw milk.²⁹ A comparable prevalence of 7.2% of bla_TEM in raw milk isolates as of our study is also available.³⁰ In another study, 28.57% isolates were harbouring bla_TEM in raw milk samples.³¹ The bla_SHV gene was found to be present in only 3 isolates obtained from faecal samples. Analogous reports can be found from North Macedonia³² with 2.56% occurrence of bla_SHV gene in faeces. There are reports of absence of bla_SHV gene in raw milk samples from Indian States.²⁹ However, another study from Anand, Gujarat, India reported 14.7% of isolates were possessing bla_SHV gene in raw milk samples.³³

Risk factors for occurrence of ESBL

Based on the results of the statistical analysis (χ2 test, P < .01), 7 variables have been identified to be associated with the presence of ESBL in dairy farms (Table 2). These variables are housing system, milking method, hygienic condition of farm, practice of washing of hands of milkmen as well as udder of animal and water sanitation at farm level. The significant variations in the populations and study methodology across different countries render it extremely challenging to compare the prevalence and risk factors for ESBL production in a reliable manner. Antimicrobial use and subsequent resistance on dairy farms is affected by various management factors including average herd age, disease outbreaks, cleanliness, the application of teat sealants and farmer’s attitudes about antimicrobial stewardship. The housing system of dairy farms can be correlated with the number of animals affected by crowdedness, which in turn leads to easy transmission of resistance genes through different routes. Intensive farming system especially indoor housing can lead to diseases and its associated antibiotic treatment might further leads to AMR.³⁴ In agreement with our study, other researchers also emphasized the importance of farm hygiene as risk factor for the prevalence of ESBL in dairy farms.³⁵ Source of drinking water was not found as a significant risk factor associated with the occurrence of ESBL E. coli in this study. It was similar to the findings of other studies.³⁶ Variables related to infrastructure and cleanliness of dairy farm is significantly associated with elevated chance of carrying ESBL as per the study conducted in Israel.³⁷ They also noticed an elevated carriage of ESBL according to the number of weekly visits of veterinarians in the farm for treatment. In another study, increased use of antibiotics especially cephalosporins are reported as a risk factor for ESBL producing Enterobacteriaceae on dairy farms,^36,38 however they were not find hygienic condition as a significant factor.³⁶ However, there are research study, which couldn’t find any correlation¹⁶ between the use of different types of antimicrobials including cephalosporins and ESBL herd status as that of us.

Table 2.

Risk factor analysis.