Identification of β-Lactamase-Encoding ( bla ) Genes in Phenotypically β-Lactam-Resistant Escherichia coli Isolated from Young Calves in Belgium

Abstract

The bla genes identification present in 94 phenotypically resistant Escherichia coli isolated from feces or intestinal contents of young calves with diarrhea or enteritis in Belgium was performed by microarrays (MA) and whole genome sequencing (WGS). According to their resistance phenotypes to 8 β-lactams at the disk diffusion assay these 94 E. coli produced a narrow-spectrum-β-lactamase (NSBL), an extended-spectrum-β-lactamase (ESBL) or a cephalosporinase (AmpC). All ESBL-encoding genes identified by MA and WGS belonged to the bla_CTX-M family, with a majority to the bla_CTX-M-1 subfamily. Two different genes encoding an AmpC, bla_CMY-2, and bla_DHA-1 were detected in isolates with an AmpC phenotype. The bla_TEM-1 and the bla_OXA-1 were detected alone in isolates with a NSBL phenotype or in combination with ESBL-/AmpC-encoding bla genes. Furthermore, the WGS identified mutations in the ampC gene promoter at nucleotides -42 (C>T) and/or -18 (G>A) that could not be identified by MA, in several isolates with an AmpC-like resistance phenotype. No carbapenemase-encoding gene was detected. To our knowledge this is the first survey on the identification of bla genes in E. coli isolated from young diarrheic or septicemic calves in Belgium.

Introduction

Nowadays, the increasing spread of bacterial acquired antimicrobial resistance (AMR) in animals and humans is a major public health hazard worldwide, especially when the resistance spectrum includes hospital last resort antibiotics, such as latest generation cephalosporin and carbapenem β-lactams.¹

Already a dozen years after the penicillin discovery,² Abraham and Chain³ described the production of “an enzyme able to destroy penicillin” by Escherichia coli. Indeed, acquired resistance to β-lactam antibiotics in Gram-negative bacteria is most frequently mediated by the production of β-lactamase (BLA) enzymes hydrolyzing the β-lactam ring. Unfortunately the original strain of Abraham and Chain³ has been lost and the actual identity of this enzyme remains unknown, although production of a BLA_AmpC is the most likely hypothesis.^4,5 The first confirmed description of a BLA_AmpC enzyme dates from 1973 and is produced by an E. coli K-12 harboring a chromosome-located bla_ampC gene.⁶ Today, it is known that most wild-type E. coli possesses such a bla_ampC gene expressed at a low basal level, conferring no clinical resistance to β-lactams. However, when this gene is overexpressed, clinical resistance to all penicillins, cephamycins, first-/second-generation cephalosporins and sometimes low-level resistance to third-generation cephalosporins occur.⁷ This BLA_AmpC enzyme is a member of the cephalosporinase families, also generically named AmpC rendering the nomenclature confusing.⁴ Indeed, other cephalosporinase families were described during the following years (CMY, DHA, LAT, MIR, ACT, FOX) with a broad hydrolysis capacity of different β-lactams⁴ that are encoded by chromosome or plasmid genes whose expression can be induced by the presence of β-lactams.⁸ Cephalosporinase enzymes classically confer resistance to narrow and large spectrum penicillins, all four generation cephalosporins (although not always the fourth ones), cephamycins, BLA inhibitors or oxapenams and monobactams, but not to carbapenems.⁴

During the following decades, Gram-negative bacteria produce a large set of non AmpC-like BLAs displaying extended inactivation spectra of β-lactams. Today, thousands of BLAs have been described⁹ and different classifications have been proposed based on molecular and/or functional criteria.^10,11 However, such molecular identification is not possible in routine diagnostic laboratories, especially in the veterinary field, being expensive and time consuming. The alternative is to use a phenotypic classification based on the inactivation spectra of β-lactams.

In addition to cephalosporinase families, three BLA families were described in the years 1960: BLA_TEM, BLA_SHV, and BLA_OXA. They confer resistance to narrow and large spectrum penicillins, to first and some second-generation cephalosporins and are encoded by acquired plasmid-located genes.⁷ In this article, they will be named “narrow-spectrum-β-lactamases” or “NSBLs.”

About 20 years later, BLAs conferring resistance to almost all penicillins, all four generation cephalosporins and monobactams but not to cephamycins, carbapenems, and oxapenams, were described for the first time and named “extended-spectrum-β-lactamases” (ESBLs).¹² Mutations in bla_TEM, bla_SHV, and bla_OXA genes,^13,14 or acquisition of genes encoding other BLA families, conferring ESBL resistance phenotypes, such as VEB,¹⁵ GES,¹⁶ and CTX-M¹⁷ that are the most prevalent in human enterobacteria,¹⁸ were progressively identified. Today, more than 240 BLA_TEM, 235 BLA_SHV, and 230 BLA_CTX-M variants have been described,⁹ as a consequence of the accumulation of mutations in the original bla genes.

The last resort β-lactam antibiotics against AmpC- and ESBL-producing Gram-negative bacteria are the carbapenems. However, a first carbapenemase (CP) enzyme, with serine in the active site, of enterobacteria was described in 1996 produced by a clinical isolate of Klebsiella pneumoniae.¹⁹ As for ESBLs, some CP derive from existing BLAs after accumulation of mutations in the encoding genes (bla_GES-2²⁰ or bla_OXA-48,²¹ for instance), whereas others correspond to newly acquired chromosome- or plasmid-encoded BLAs: KPC, IMI, VIM, and most recently NDM.²² CPs can show some particularity in their β-lactam inactivation profiles (narrow and large spectrum penicillins, all cephalosporin generations, cephamycins, oxapenams, monobactams, and carbapenems), due to their action mode: serine (KPC, IMI, OXA-48) or Zn⁺⁺ (metalloenzymes such as NDM and VIM) in their active sites.²³

Several studies have widely demonstrated the primary importance of monitoring antibiotic resistance among farm animals (e.g., cattle, small ruminants, and poultry), considered as reservoir for AMR. Yet, the direct impact on human health remains unclear.²⁴

Since raising of beef cattle is most widespread in Wallonia, >90% of bacteria tested at the regional laboratory (“Association Régionale de Santé et d'Identification Animales,” ARSIA) in Wallonia²⁵ for their antibiotic resistance profiles by the disk diffusion assay (DDA) are from bovine origin and ca. 70% of them are sampled from necropsied calves and calf feces.²⁶ In Belgium, since the publication of a royal decree in summer 2016, the use of carbapenems is prohibited and the use of third- and fourth-generation cephalosporins is strictly controlled in livestock.²⁷ Consequently, a progressive decrease of ESBL-producing pathogenic E. coli in cattle was observed already during the 2016–2017 calving season at ARSIA. Unfortunately, no identification of the resistance-encoding genes is routinely performed,²⁶ consequently the incidence of the different BLA families in these E. coli is unknown.

The purpose of this study was to (1) identify, by MA and whole genome sequencing (WGS), the different bla genes present in a set of bovine E. coli isolated from diarrheic and septicemic calves during the calving season 2017–2018 at ARSIA and (2) compare these bla genes with the β-lactam resistance phenotypes observed at the DDA. The results will help to develop PCRs targeting the identified bla genes to follow their evolution and to detect the emergence of other bla genes in pathogenic and commensal E. coli during the calving seasons in Belgium.

Materials and Methods

Bacterial isolates

A total of 94 bovine E. coli isolated at ARSIA (Ciney, Belgium) between November 2017 and February 2018 from feces, intestinal contents, or internal organs of <3-month-old calves suffering from diarrhea or septicemia were studied. The presence of β-lactam resistance was detected by DDA, following the EUCAST/CASFM guidelines with eight β-lactams: amoxicillin (AMX), amoxicillin + clavulanic acid (AMC), ceftiofur (XNL), cefquinome (CFQ), cefotaxime (CTX), cefotaxime + clavulanic acid (CTC), cefoxitin (FOX), and meropenem (MER).

A first group of 69 isolates, from feces, intestinal contents or internal organs, were chosen out of 433 bovine β-lactam-resistant E. coli to represent the NSBL, ESBL and AmpC spectra of resistance (Table 1). In this first set, isolates presenting a positive agglutination test (Biovac, Beaucouzé, France) after growth on Minca agar plates (Generon, San Prospero, Italy), for fimbriae F5 or F17 or for the CS31 antigen were, therefore, considered as pathogenic. The second group comprised 25 isolates that were randomly chosen out of 435 fecal or intestinal E. coli growing on MacConkey agar plates containing 1 mg/L of cefotaxime (McC-CTX) (Led Techno, Lienden, The Netherlands), selective for ESBL-, AmpC-, and CP-production, using the “RANDBETWEEN” function of Excel^® (Table 1). This second set contained E. coli with negative agglutination test, therefore classified as nonpathogenic. To avoid as much as possible clonal relationship between the isolates, one isolate per calf and herd was chosen.

Table 1.

Classification of Resistance Based on the Profile Observed at the Disk Diffusion Assay in Escherichia coli Isolated from Young Diarrheic or Septicemic Calves

Resistance profiles	Antibiotics								No. of isolates
Resistance profiles	AMX	AMC	XNL	CFQ	CTX	CTC	FOX	MER	Set1	Set2	Total
AmpC	R	R	R	R	R	R	R	S	8	3	11
AmpC-like	R	R	S	S	S	S	I/R	S	6	12	18
NSBL	R	S/I/R	S	S	S	S	S	S	15	1	16
ESBL	R	S/I	R	R	R	S/I	S	S	33	6	39
ESBL + NSBL	R	R	R	R	R	S/I	S	S	3	2	5
ESBL + AmpC-like	R	R	R	R	R	S/I	I/R	S	4	1	5
CP	R	R	R	R	R	R	R	R	0	0	0

R, resistant; I, intermediate; S, sensitive; AMX, amoxicillin; AMC, amoxicillin + clavulanic acid; XNL, ceftiofur; CFQ, cefquinome; CTX, cefotaxime; CTC, cefotaxime + clavulanic acid; FOX, cefoxitin; MER, meropenem; AmpC, cephalosporinase; NSBL, narrow-spectrum-β-lactamase; ESBL, extended-spectrum-β-lactamase; CP, carbapenemase; Set1, Escherichia coli isolates chosen for their resistance profile and their pathogenic classification; Set2, E. coli isolates randomly chosen for their growth on MacConkey-Cefotaxime selective agar plates and their nonpathogenic classification.

Since no bovine CP-producing E. coli was isolated, six CP-positive strains isolated from human were added as controls for the MA: 2 E. coli harboring the bla_NDM-1 or bla_OXA-48 genes, 2 K. pneumoniae harboring the bla_KPC-3 or bla_VIM-4 genes and 2 Pseudomonas aeruginosa harboring the bla_VIM-2 or bla_NDM-1 genes (kindly provided by Professor Y. Glupczynski, CHU Mont-Godinne, Catholic University of Louvain, Belgium).

Microarrays

The DNA of the 94 bovine E. coli and of the 6 CP-positive strains was extracted with the DNeasy Blood and Tissue kit^® (Qiagen, Venlo, The Netherlands) and stored at −20°C. Tests were performed by Check-Points BV (Wageningen, The Netherlands), following the MA protocol previously described.²⁸ The different bla genes and families that the Check-MDR CT-103XL^® microarrays kit (Check-Points BV, Wageningen, The Netherlands) can detect are listed in Table 2.

Table 2.

Targeted Families of bla Genes by the Check-MDR CT103XL Microarray Kit

Family	bla genes
NSBL	TEM-WT; SHV-WT
ESBL	BEL; CTX-M (Groups 1^a, 2, 9, and 8/25); GES-ESBL; PER; SHV-ESBL; TEM-ESBL; VEB
AmpC	ACC; ACT; CMY; DHA; FOX; MIR; MOX
CP	GES-Carba; GIM; IMP; KPC; NDM; OXA (groups 23-, 24-, 48-, and 58-like); SPM; VIM

Group 1 contains subgroup types 1-, 3-, 15-, and 32-like. Adapted from https://check-pointshealth.com/wp-content/uploads/2018/11/Check-MDR_CT103XL_IFU_10–0023_EN-v1.1–20170918.pdf

AmpC, cephalosporinases; CP, carbapenemases.

Whole genome sequencing

The genome of 29 isolates with discrepancy between the DDA and the MA results were sequenced as well as 15 isolates with matching DDA-MA results as controls. Sequencing was performed using NovaSeq 6000 Illumina technology (paired-end sequencing; Nextera XT DNA Sample Prep Kit, Illumina) by the Brain and Spine Institute (ICM Institute). Reads were cleaned and assembled with the shovill method (v1.0.4), with a minimum length of 200 bp and a minimum coverage of 2. Contigs were analyzed with ResFinder-3.1²⁹ to detect mutations and resistance genes and annotated with the RAST tool kit in PATRIC (Pathosystems Resource Integration Center).³⁰

Results

Identification of bla genes by MA and comparison of MA and DDA results

The following bla genes were detected in the 94 bovine E. coli isolates (Table 3): bla_TEM-WT in 67 isolates; bla_CTX-M-1, bla_CTX-M-2, and/or bla_CTX-M-9 in 50 isolates; and bla_CMY-2 or bla_DHA-1 in 6 isolates. The bla_TEM-WT genes were detected alone in 22 isolates and in association with bla_CTX-M or with bla_CMY-2/bla_DHA-1 genes in 40 and 5 isolates, respectively, whereas the bla_CTX-M genes were detected alone in 10 isolates. Four of the 5 bla_CMY-2 and the bla_DHA-1 genes were detected in association with bla_TEM-WT genes, whereas the remaining isolate was positive for the bla_CMY-2 gene alone. No CP-encoding genes were detected in the bovine isolates, but the expected ones were detected in the six CP-positive control bacteria.

Table 3.

Disk Diffusion Assays Phenotypes and Genes Involved in β-Lactam Resistance Detected by Microarrays in 94 Bovine Escherichia coli Isolates

DDA phenotype	Detected bla genes by MA coding for	No. of isolates	MA-DDA matching
AmpC (AMC^R)	CMY-2	1	Yes
	CMY-2 + TEM-WT	4	Yes
	CTX-M-1 + TEM-WT	1	No
	TEM-WT	4	No
AmpC-like (AMC^R)	TEM-WT	5	Partial
	DHA + TEM-WT	1	Partial
	None	13	—
NSBL (AMC^S/I/R)	TEM-WT	13	Yes
NSBL (AMC^S/I/R)	None	3	—
ESBL (AMC^S/I)	CTX-M-1 (type 1-like)	1	Yes
	CTX-M-1 (type 32-like)	4	Yes
	CTX-M-2	3	Yes
	CTX-M-9	1	Yes
	CTX-M-1 (type 1-like) + TEM-WT	8	Yes +
	CTX-M-1 (type 15-like) + TEM-WT	6	Yes +
	CTX-M-1 (type 32-like) + TEM-WT	5	Yes +
	CTX-M-1 (type 32-like) + CTX-M-2 + TEM-WT	1	Yes +
	CTX-M-2 + TEM-WT	6	Yes +
	CTX-M-9 + TEM-WT	4	Yes +
ESBL + NSBL (AMC^R)	CTX-M-1 (type 15-like)	1	Partial
	CTX-M-1 (type 1-like) + TEM-WT	1	Yes
	CTX-M-2 + TEM-WT	1	Yes
	CTX-M-9 + TEM-WT	2	Yes
ESBL + AmpC-like (AMC^R)	CTX-M-1 (type 15-like) + TEM-WT	2	Partial
	CTX-M-1 (type 32-like) +TEM-WT	1	Partial
	CTX-M-9 + TEM-WT	2	Partial

AMC, amoxicillin + clavulanic acid; R, resistant; I, intermediate; S, sensitive; AmpC, cephalosporinase; YES, matching results between DDA and MA (the detected genes explaining the observed phenotypes); YES +, matching results with additional genes detected by MA (some of the detected genes explain the observed phenotypes); PARTIAL, partially matching results between DDA and MA (the detected genes do not fully explain the observed phenotypes); NO, no matching results between DDA and MA/WGS (the detected genes do not explain the observed phenotypes); —, no gene. DDA, disk diffusion assay; MA, microarrays; WGS, whole genome sequencing.

The comparison of the MA and DDA results yields the following conclusions (Table 3): (1) a perfect match was observed for 61 isolates, that is, the detected genes are in agreement with the phenotypes. Nevertheless, for 30 of them, presenting an ESBL phenotype, additional genes were detected; (2) a partial match was observed for 12 isolates, that is, the detected genes were not in total agreement with the phenotype; (3) no match was observed for 5 isolates, that is, the detected genes were in total disagreement with the phenotypes; and (4) no gene was detected in 16 isolates, including 13 with an AmpC-like phenotype.

Detected genes were similar between the two sets of isolates, regardless of their origins (feces, intestinal contents, or organs) (Table 1).

Identification of bla genes by WGS and comparison with the DDA results

All genomic data related to this study are available through the NCBI BioProject PRJNA566319.

Forty-three isolates were genome sequenced: 16 isolates with no gene detected by the MA, 5 isolates with no MA-DDA matching results, 4 isolates with partial matching results, 3 isolates with matching results and additional genes detected, and 15 isolates with matching results, as controls (Table 4).

Table 4.

Detected Genes Involved in β-Lactam Resistance After Whole Genome Sequencing of 43 Isolates with Different Disk Diffusion Assays and Microarrays Matching Results

MA-DDA matching	DDA phenotype	Detected bla genes by MA coding for	Detected bla genes by WGS coding for	No. of isolates	WGS-DDA matching
—	AmpC-like (AMC^R)	None	MutAmpC (-1, -18, -42)	10	Yes
			MutAmpC (-1, -18, -42) + OXA-1	2
			MutAmpC (-1, -18, -42) + TEM-1B	1
	NSBL (AMC^S/I/R)	None	OXA-1	2	Yes
	NSBL (AMC^S/I/R)	None	OXA-1 + MutAmpC (-1, -18)	1	Yes
No	AmpC (AMC^R)	CTX-M-1 + TEM-WT	CTX-M-1 + TEM-35	1	Yes
No	AmpC (AMC^R)	TEM-WT	MutAmpC (-1, -18, -42) + TEM-1A + OXA-1	4	Yes
Partial	AmpC-like (AMC^R)	TEM-WT	MutAmpC (-1, -18, -42) + TEM-1A + OXA-1	1	Yes
		TEM-WT	MutAmpC (-1, -18, -42) + TEM-30	1	Yes
		DHA + TEM-WT	DHA + TEM-1B	1	Partial
	ESBL + AmpC-like (AMC^R)	CTX-M-9 + TEM-WT	CTX-M-14 + TEM-1B + MutAmpC (-1, -18)	1	Partial
Yes +	ESBL (AMC^S/I)	CTX-M-1 (type 1-like) + TEM-WT	CTX-M-1 + TEM-1B	1	Yes +
		CTX-M-1 (type 1-like) + TEM-WT	CTX-M-1 + TEM-1B + OXA-1	1
		CTX-M-1 (type 15-like) + TEM-WT	CTX-M-156^a + TEM-1B + OXA-1	1
Yes	AmpC (AMC^R)	CMY-2 + TEM-WT	CMY-2 + TEM-1A	2	Yes
	AmpC (AMC^R)	CMY-2 + TEM-WT	CMY-2 + TEM-1B	1	Yes
	NSBL (AMC^S/I/R)	TEM-WT	TEM-1A + OXA-1	1	Yes
			TEM-1B	4
			TEM-1B + OXA-1	1
			TEM-1B + MutAmpC (-1, -18)	1
	ESBL (AMC^S/I)	CTX-M-1 (type 32-like)	CTX-M-32	1	Yes
	ESBL (AMC^S/I)	CTX-M-2	CTX-M-2	2	Yes
	ESBL + NSBL (AMC^R)	CTX-M-1 (type 1-like) + TEM-WT	CTX-M-1 + TEM-1B + MutAmpC (-1, -18)	1	Yes +
	ESBL + NSBL (AMC^R)	CTX-M-2 + TEM-WT	CTX-M-2 + TEM-78 + MutAmpC (-1, -18)	1	Yes +

AMC, amoxicillin + clavulanic acid; R, resistant; I, intermediate; S, sensitive; AmpC, cephalosporinase; YES, matching results between DDA and MA/WGS (the detected genes explaining the observed phenotypes); YES +, matching results with additional genes detected by MA/WGS (some of the detected genes explain the observed phenotypes); PARTIAL, partially matching results between DDA and MA/WGS (the detected genes do not fully explain the observed phenotypes); NO, no matching results between DDA and MA/WGS (the detected genes do not explain the observed phenotypes); —, no gene detected by MA; MutAmpC, ampC gene with two or three mutations detected in the promoter, concerned position in bracket.

CTX-M 156 is a member of the CTX-M-1 group in the type 15-like subgroup.

The 16 isolates with no gene detected with the MA harbored the bla_ampC gene with 2 or 3 mutations in the promoter (14 isolates), the OXA-1-encoding gene (5 isolates) and/or the TEM-IB-encoding gene (1 isolate) (Table 4). The mutations in the promoter of the bla_ampC gene and the bla_OXA-1 gene cannot be detected by the MA (Table 2). The WGS of the remaining 27 isolates not only confirmed the presence of the genes detected by MA although the TEM-encoding genes were different from bla_TEM-1/2 in three isolates (bla_TEM-30, bla_TEM-35, and bla_TEM-78), but also identified the presence of the bla_ampC gene with mutations in the promoter and/or of the bla_OXA-1 gene in 14 of them.

A perfect match (yes and yes+ categories in Tables 3 and 4) was observed between the WGS and DDA results for 41 of the 43 genome-sequenced isolates versus only 18 of these 43 isolates for the MA and DDA results. In particular, 15 of the 16 AmpC-like isolates (Table 4) harbored a bla_ampC gene with 3 mutations in the promoter explaining this phenotype. The two isolates with partial matching between WGS and DDA results were one AmpC-like isolate harboring TEM-1B- and DHA-1-encoding genes and one ESBL+AmpC-like in which the bla_ampC gene carried only two mutations in the promoter (Table 4).

Of the 27 genome-sequenced isolates with genes detected by MA (Table 4), the results of the WGS matched the results of MA for 24 isolates, with additional genes detected by WGS for 13 of them and partially matched for the 3 isolates harboring the bla_TEM-30, bla_TEM-35, or bla_TEM-78 genes and also mutations in the promoter of the bla_ampC gene in two of them.

As for the DDA-MA comparison, the detected genes were similar in the two sets of isolates.

Discussion

Since the existing classifications of the BLA enzymes^10,11 cannot easily be applied in veterinary routine diagnostic laboratories in Belgium, the prevalence of the different BLA families and bla genes in bovine E. coli in Belgium is unknown. The purpose of this study was, therefore, to identify the bla genes present in E. coli isolated from young calves with diarrhea, enteritis, and/or septicemia during the 2017–2018 calving season with different resistance phenotypes to eight β-lactams at the DDA (Table 1). The results will help to develop appropriate PCR assays for future studies aiming at following the incidence of the bla genes over the years and at comparing bovine E. coli with human E. coli, focusing more especially on those with ESBL and AmpC resistance profiles.

E. coli isolates belonging to the ESBL, ESBL+NSBL, and ESBL+AmpC-like resistance phenotypes harbor gene(s) of the bla_CTX-M family, mostly the bla_CTX-M-1 and less frequently the bla_CTX-M-2 or bla_CTX-M-9 genes, which explain the ESBL resistance profile (Table 3 and 4). Such results are similar to those of E. coli in livestock especially cattle in Europe, with the bla_CTX-M-1 and bla_CTX-M-9 genes being the most frequent ESBL-encoding gene, compared with the bla_CTX-M-2 gene.^18,31–33 In several of the same isolates, other genes are detected whose presence, either cannot be suspected according to the DDA results (ESBL phenotype) or add resistance to other β-lactams, especially to oxapenams or to cephamycins (ESBL+NSBL and ESBL+AmpC-like phenotypes; Table 1): bla_TEM-1B, bla_OXA-1, and/or bla_ampC with two mutations in the promoter (n-18 G > A and n-1 C>T) (Tables 3 and 4).

The results of the isolates belonging to the AmpC and AmpC-like phenotypes are more heterogeneous. At first, a majority of AmpC and AmpC-like isolates, without DDA-MA matching, harbor a bla_ampC gene with 3 mutations in the promoter (n-42 C>T, n-18 G > A and n-1 C>T) (Tables 3 and 4). Chromosome-located bla_ampC genes are usually expressed at a very low level and do not confer clinical resistance to β-lactams.⁷ However the accumulation of mutations in their promoter may increase the gene expression and confer resistance to several β-lactams, but still not to latest generation cephalosporins and carbapenems.^34–36 These resistance phenotypes can, therefore, be confused with the NSBL phenotype in routine analysis (Table 1). The level of mRNA transcription was not measured, but the production of a BLA with the Diatabs test (Rosco, via International Medical Product, Brussels, Belgium) showed the hydrolysis of the ampicillin for 16 of these 19 isolates (data not shown). The increase of the bla_ampC gene expression is, therefore, the most likely explanation. Nevertheless, alternative explanations of the resistance to cephamycins exist: mutation within the genes coding for the penicillin-binding proteins, production of efflux pumps and/or a modification of the membrane permeability.³⁷ For instance, the resistance to cephamycins of the five ESBL+AmpC-like isolates may have similar alternative explanations,⁴ since (1) the two sequenced isolates harbor only two mutations in the bla_ampC gene promoter (n-18 and n-1) and (2) the same two mutations in two NSBL isolates and in two ESBL+NSBL isolates have no apparent consequence on their sensitivity to cephamycins. More isolates with (ESBL+) AmpC and AmpC-like resistance profiles should be genetically studied to more precisely understand the resistance profile and the actual role of the mutated bla_ampC gene.

In addition, other genes of the AmpC family are present that can explain the AmpC and AmpC-like phenotypes: the bla_CMY-2 gene in five AmpC isolates and the bla_DHA-1 gene in one AmpC-like isolate (Tables 3 and 4).^4,8 In Europe, the bla_CMY-2 gene is also frequent in livestock.^18,31–33 Finally the remaining AmpC isolate harbors a bla_CTX-M-1 gene and a bla_TEM-35 gene, which is an oxapenam-resistant variant and can, therefore, also explain the resistance to oxapenams,³⁷ in contrast to ESBL+NSBL isolates harboring bla_CTX-M and bla_TEM-WT genes (Tables 1, 3, and 4).

As for the isolates belonging to the NSBL phenotype, either or both bla_TEM-1B and bla_OXA-1 genes were detected. In addition, as discussed earlier, one bla_ampC gene with two mutations in the promoter was also identified in two of the genome-sequenced NSBL isolates along with the bla_TEM-1B and bla_OXA-1 genes (Tables 3 and 4), but with no consequence on the resistance phenotype. Very importantly, no bovine E. coli belong to the CP phenotype and no CP-encoding gene could be identified, neither by MA nor by WGS.

The comparison between the different resistance profiles observed at DDA of bovine E. coli isolates with their bla gene contents detected by MA and/or WGS was the second objective of this study. The results of the MA and DDA perfectly match (yes and yes+ categories in Table 3) for 61 of the 94 isolates studied. In addition, bla genes and/or mutations that could not be detected by the MA, such as the bla_OXA-1 gene and the bla_ampC gene with mutations in the promoter, were detected in the genome-sequenced of 29 of the 33 isolates with no perfect match between MA and DDA. Therefore, after WGS, only 2 of these genome-sequenced 29 isolates still have partial match between genetic and phenotypic results. MA and DDA mismatch results in this study are, therefore, explained by the presence of untargeted bla genes such as the bla_OXA-1 gene and of untargeted mutations in some bla genes, such as in the promoter of the bla_ampC gene. In contrast to these results on bovine E. coli, the MA kit Check-MDR CT103XL^® correctly detected several bla genes (Table 2) of different Enterobacteriaceae, Pseudomonaceae, and Acinetobacter species.^39,40 However, the probes of the MA were chosen according to the prevalence of the bla genes in humans and not in animals. Moreover, in those previous studies, the targeted bla genes were known before the MA test.

Conclusion

As a conclusion, the detected bla genes are in accordance with the different resistance phenotypes of the bovine E. coli in Belgium (Tables 1, 3, and 4) and with the bla genes contents of the bovine E. coli in other European countries.^18,31–33 Moreover, no difference was observed between intestinal and septicemic isolates, pathogenic or not. Therefore, the PCR for the bla_TEM-1/2, bla_OXA-1, bla_CTX-M-1, bla_CTX-M-2, bla_CTX-M-9, and bla_CMY-2 genes represent, currently, the most useful tools to follow the incidence of the most frequent bla genes identified in cattle from Belgium. Further studies are necessary to follow the evolution of the β-lactams resistance and to estimate the impact of the antibiotics use regulation, to have a better control of AmpC-/ESBL-producing E. coli in cattle.

Footnotes

Acknowledgments

The authors thank Professor Y. Glupscynski for providing the CP-positive bacterial strains isolates used as controls in this study, and the technicians from bacteriology laboratory of ARSIA for their technical help. The authors are also grateful to the ICM institute for its technical support in NovaSeq NGS sequencing. The authors thank Jean-Noël Duprez for his help in the realization of DDA and DNA extraction. The authors thank Check-Point for their excellent welcome in their laboratory, their detailed explanations, and their technical support during the MA tests. These results were presented, in part, during the 30th World Buiatrics Congress (Sapporo, Japan, August 2018) and during the 8th Symposium on Antimicrobial Resistance in Animals and the Environment (Tours, France, July 2019). The authors are grateful to two anonymous reviewers for their comments and suggestions that helped improved this article.

Authors' Contributions

V.G. performed the MA assay, analyzed the antibiotic sensitivity results and the genome sequences, and wrote the article. D.T. supervised the whole study, discussed and helped to analyze the phenotypic and genetic results, and wrote parts of the article. M.S. supervised the whole study, isolated and identified the bovine E. coli, performed the antibiotic sensitivity assay, and discussed the results. M.G. and J.M. supervised the whole study, discussed the results, and synthesized the different parts of the article. P.L. and Y.B. performed the genome sequencing and wrote parts of the article. F.C. and P.S.M. discussed the phenotypic and genetic results, and helped to write parts of the article.

Disclosure Statement

No competing financial interests exist.

Funding Information

This study was supported by the Belgian Federal Public Service Health, Food Chain Safety and Environment [Grant No. RF 17/6317 RU-BLA-ESBL-CPE].

References

Organisation for Economic Co-operation and Development (OECD). 2018. Stemming the Superbug Tide: Just A Few Dollars More, OECD Health Policy Studies. OECD Publication, Paris.

Fleming

1929. On the antibacterial action of cultures of a Penicillium, with special reference to their use in the isolation of B. influenzae. Br. J. Exp. Pathol. 10:226–236.

Abraham

E.P.

, Chain

. 1940. An enzyme from bacteria able to destroy penicillin. Nature, 3713:837.

Jacoby, G.A. 2009. AmpC Β-lactamases. Clin. Microbiol. Rev. 22:161–182.

Frère

J.-M.

, Sauvage

, and Kerff

. 2016. From “An enzyme able to destroy penicillin” to carbapenemases: 70 years of beta-lactamase misbehavior. Curr. Drug Targets, 17:974–982.

Burman

, Park

, Lindström

, and Boman

. 1973. Resistance of Escherichia coli to Penicillins: identification of the structural gene for the chromosomal penicillinase. J. Bacteriol. 116:123–130.

Livermore

1995. β-Lactamases in laboratory and clinical resistance. Clin. Microbiol. Rev. 8:557–584.

Pham

J.N.

, Chambers

, Poirel

, Nordmann

, and Bell

S.M.

. 2010. Detection of a plasmid-mediated inducible cephalosporinase DHA-1 from Escherichia coli. Pathology, 42:196–197.

Naas

, Oueslati

, Bonnin

R.A

, et al. 2017. Beta-lactamase database (BLDB)–structure and function. J. Enzyme Inhib. Med. Chem. 32:917–919.

10.

Ambler, R.P. 1980. The structure of beta-lactamases. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 289:321–331.

11.

Bush

, and Jacoby

G.A.

. 2010. Updated functional classification of β-lactamases. Antimicrob. Agents Chemother. 54:969–976.

12.

Ghafourian

, Sadeghifard

, Soheili

, and Sekawi

. 2015. Extended spectrum beta-lactamases: definition, classification and epidemiology. Curr. Issues Mol. Biol. 17:11–22.

13.

Kliebe

, Nies

B.A

, Meyer

J.F

, Tolxdorff-Neutzling

R.M

, and Wiedemann

. 1985. Evolution of plasmid-coded resistance to broad-spectrum cephalosporins. Antimicrob. Agents Chemother. 28:302–307.

14.

Paul

G.C.

, Gerbraud

, Bure

, Philippon

A.M

, Pangon

, and Courvalin

. 1989. TEM-4, a new plasmid-mediated β-lactamase that hydrolyzes broad-spectrum cephalosporins in a clinical isolate of Escherichia coli. Antimicrob. Agents Chemother. 33:1958–1963.

15.

Poirel

, Naas

, Guibert

, Chaibi

E.B

, Labia

, and Nordmann

. 1999. Molecular and biochemical characterization of VEB-1, a novel class A extended-spectrum β-lactamase encoded by an Escherichia coli integron gene. Antimicrob. Agents Chemother. 43:573–581.

16.

Poirel

, Le Thomas

., Naas

, Karim

, and Nordmann

. 2000. Biochemical sequence analyses of GES-1, a novel class A extended- spectrum β-lactamase, and the class 1 integron In52 from Klebsiella pneumoniae. Antimicrob. Agents Chemother. 44:622–632.

17.

Nordmann

, and Mammeri

. 2007. Extended-spectrum cephalosporinases: structure, detection and epidemiology. Future Microbiol. 2:297–307.

18.

Ewers

, Bethe

, Semmler

, Guenther

, and Wieler

L.H.

. 2012. Extended-spectrum β-lactamase-producing and AmpC-producing Escherichia coli from livestock and companion animals, and their putative impact on public health: a global perspective. Clin. Microbiol. Infect. 18:646–655.

19.

Yigit

, Queenan

A.M

, Anderson

G.J

, et al. 2001. Novel carbapenem-hydrolyzing β-lactamase, KPC-1, from a carbapenem-resistant strain of Klebsiella pneumoniae. Antimicrob. Agents Chemother. 45:1151–1161.

20.

Poirel

, Weldhagen

G.F

, Naas

, De Champs

, Dove

M.G

, and Nordmann

. 2001. GES-2, a class A β-lactamase from Pseudomonas aeruginosa with increased hydrolysis of imipenem. Antimicrob. Agents Chemother. 45:2598–2603.

21.

Poirel

, Héritier

, Tolün

, and Nordmann

. 2004. Emergence of oxacillinase-mediated resistance to imipenem in Klebsiella pneumoniae. Antimicrob. Agents Chemother. 48:15–22.

22.

Lauretti

, Riccio

M.L

, Mazzariol

, et al. 1999. Cloning and characterization of bla(VIM), a new integron-borne metallo- β-lactamase gene from a Pseudomonas aeruginosa clinical isolate. Antimicrob. Agents Chemother. 52:3589–3596.

23.

Nordmann

2014. Carbapenemase-producing Enterobacteriaceae: overview of a major public health challenge. Med. Mal. Infect. 44:51–56.

24.

Madec

J.-Y.

, Haenni

, Nordmann

, et al. 2017. Extended-spectrum β-lactamase/AmpC- and carbapenemase-producing Enterobacteriaceae in animals: a threat for humans? Clin. Microbiol. Infect. 23:826–833.

25.

Association Régionale de Santé et d'Identification Animales (ARSIA). 2018. Annual report (French version), 11–13. Available at https://www.arsia.be/wp-content/uploads/documents-telechargeables/RA-2018-web.pdf, (accessed February 7, 2021).

26.

Association Régionale de Santé et d'Identification Animales (ARSIA). 2019. Annual report (French version), 14. Available at https://www.arsia.be/wp-content/uploads/documents-telechargeables/RA-2019-web.pdf, (accessed February 7, 2021).

27.

Royal Decree of Belgium (French version). Moniteur Belge. 07-21-2016 updated 12-09-2020. Available at www.ejustice.just.fgov.be/eli/arrete/2016/07/21/2016024152/justel, (accessed February 7, 2021).

28.

Cohen Stuart

, Dierikx

, Al Naiemi

, et al. 2010. Rapid detection of TEM, SHV and CTX-M extended-spectrum β-lactamases in Enterobacteriaceae using ligation-mediated amplification with microarray analysis. J. Antimicrob. Chemother. 65:1377–1381.

29.

Zankari

, Hasman

, Cosentino

, et al. 2012. Identification of acquired antimicrobial resistance genes. J. Antimicrob. Chemother. 67:2640–2644.

30.

Brettin

, Davis

J.J

, Disz

, et al. 2015. RASTtk: a modular and extensible implementation of the RAST algorithm for building custom annotation pipelines and annotating batches of genomes. Sci. Rep. 5:8365

31.

Gonggrijp

M.A.

, I. Santman-Berends

M.G.A

, Heuvelink

A.E

, et al. 2016. Prevalence and risk factors for extended-spectrum β-lactamase- and AmpC-producing Escherichia coli in dairy farms. J. Dairy Sci. 99:9001–9013.

32.

Michael

G.B.

, Kaspar

, Siqueira

A.K

, et al. 2017. Extended-spectrum β-lactamase (ESBL)-producing Escherichia coli isolates collected from diseased food-producing animals in the GERM-Vet monitoring program 2008–2014. Vet. Microbiol. 200:142–150.

33.

Ghosh

K.K.

, Lebert

L.A

, McEwen

S.A

, et al. 2019. Extended-spectrum β-lactamase and ampc β-lactamase-producing Escherichia coli isolates from chickens raised in small flocks in Ontario, Canada. Microb. Drug Resist. 25:1250–1256.

34.

Caroff

, Espaze

, Gautreau

, Richet

, and Reynaud

. 2000. Analysis of the effects of -42 and -32 ampC promoter mutations in clinical isolates of Escherichia coli hyperproducing AmpC. J. Antimicrob. Chemother. 45:783–788.

35.

Mulvey

M.R.

, Bryce

, Boyd

D.A

, et al. 2005. Molecular characterization of cefoxitin-resistant Escherichia coli from Canadian hospitals. Antimicrob. Agents Chemother. 49:358–365.

36.

Tracz

D.M.

, Boyd

D.A

, Hizon

, et al. 2007. ampC gene expression in promoter mutants of cefoxitin-resistant Escherichia coli clinical isolates. FEMS Microbiol. Lett. 270:265–271.

37.

Hamed

R.B.

, Gomez-Castellanos

J.R

, Henry

, Ducho

, McDonough

M.A

, and Schofield

C.J.

. 2013. The enzymes of β-lactam biosynthesis. Nat. Prod. Rep. 30:21–107.

38.

Cantón

, Morosini

M.I

, Martin

, De La Maza

, and De La Pedrosa

E.G.G

. 2008. IRT and CMT β-lactamases and inhibitor resistance. Clin. Microbiol. Infect. 14 Suppl, 1:53–62.

39.

Cunningham

S.A.

, Vasoo

, and Patel

. 2016. Evaluation of the check-points check MDR CT103 and CT103 XL microarray kits by use of preparatory rapid cell lysis. J. Clin. Microbiol. 54:1368–1371.

40.

Powell

E.A.

, Haslam

, and Mortensen

J.E.

. 2017. Performance of the check-points check-MDR CT103XL assay utilizing the CDC/FDA antimicrobial resistance isolate bank. Diagn. Microbiol. Infect. Dis. 88:219–221.