Antibiotic Susceptibility and Virulence Factors in Escherichia coli from Sympatric Wildlife of the Apuan Alps Regional Park (Tuscany,Italy)

Abstract

Today a growing number of studies are focusing on antibiotic resistance in wildlife. This is due to the potential role of wild animals as reservoirs and spreaders of pathogenic and resistant bacteria. This study focused on isolating and identifying Escherichia coli from the feces of wild animals living in the Apuan Alps Regional Park (Tuscany, Italy) and evaluating some of their antibiotic resistance and pathogenicity traits. Eighty-five fecal samples from different species were studied. Seventy-one E. coli were identified by matrix assisted laser desorption ionization-time of flight mass spectrometry analysis, subjected to antibiograms and polymerase chain reaction for the detection of antibiotic resistance genes and pathogenicity factors. The highest resistance rates were found against cephalothin (39.4%) and ampicillin (33.8%), followed by amoxicillin/clavulanic acid (15.5%), streptomycin (12.7%), and tetracycline (5.6%). Regarding resistance genes, 39.4% of the isolates were negative for all tested genes. The remaining isolates were positive for bla_CMY_-2, sul2, strA-strB and aadA1, tet(B), and tet(A), encoding resistance to beta-lactams, trimethoprim/sulfamethoxazole, streptomycin, and tetracycline, respectively. With regard to virulence factors, 63.4% of the isolates were negative for all genes; 21.1% carried astA alone, which is associated with different pathotypes, 9.9% carried both escV and eaeA (aEPEC); single isolates (1.4%) harbored escV (aEPEC), escV associated with astA and eaeA (aEPEC), astA with stx2 and hlyA (EHEC) or astA and stx1, stx2, and hlyA (EHEC). These results show that wildlife from nonanthropized environments can be a reservoir for antibiotic-resistant microorganisms and suggest the need for a deeper knowledge on their origin and diffusion mechanisms through different ecological niches.

Introduction

Antimicrobial resistance (AMR) is currently one of the main concerns for human and animal health.¹ A wide body of knowledge demonstrates that the improper use of antimicrobials in human and especially veterinary medicine has dramatically affected the rise of antimicrobial-resistant microorganisms in the environment.^2,3

Over the last number of years, the number of studies focusing on AMR has increased along with studies focusing on resistant bacteria from wild animals.^4,5 Indeed, a deeper knowledge on the role of different compartments in the maintenance and dissemination of resistance genes seems to be crucial to tackle the issue.

Most of the studies on AMR in wildlife have focused on microorganisms relevant for human health, such as Escherichia coli,^6–10 Enterococcus spp.,^11,12 or Salmonella spp.¹³ According to Vittecoq et al.,⁵ the diversity of detected resistance mechanisms and the proportion of individual hosts carrying resistant bacteria increase with the proximity to anthropized settlements.

Currently, few studies have been carried out in Italy concerning AMR in wild animals, mammals,^13–15 birds,^13,16,17 or both. For this reason, the aim of the present study was to investigate AMR and pathogenicity factors in E. coli isolated from wildlife in the Apuan Alps Regional Park, located in north-western Tuscany (Italy). The studied area extends to over 20,600 hectares and is characterized by a heterogeneous environment with hills and mountain territory. In a natural reserve, the direct impact of humans is minimum, thus a high prevalence of antimicrobial-resistant bacteria could reflect the ability of these microorganisms to disseminate across environments. In this study, we considered different mammal species, especially canids, to gain a deeper knowledge on the distribution of pathogenic and nonpathogenic antibiotic-resistant E. coli in a natural ecosystem.

Materials and Methods

Fecal samples collection

From April 2017 to October 2017, 85 fecal samples from wild animals were collected in the Apuan Alps Regional Park located in Tuscany, Central Italy. Forty-two samples belonged to wolves (Canis lupus), 13 to foxes (Vulpes vulpes), 6 to badgers (Meles meles), 3 to wild boars (Sus scrofa), 3 to red deers (Cervus elaphus), 2 to roe deers (Capreolus capreolus), 9 to mouflons (Ovis aries musimon), 5 to wild goats (Capra hircus), and 2 to hares (Lepus europeus). Animals included in the study were not subjected to antibiotic treatment. No information on their health status was available.

E. coli isolation

For the isolation of E. coli, a nonselective pre-enrichment step was performed employing buffered peptone water (BioLife Italiana S.r.l.-Mascia Brunelli S.p.a., Milan, Italy) incubated at 37°C for 24 hr. After incubation, a loopful from each broth culture was seeded on Tryptone Bile X-Glucuronide agar plates (Thermo Fisher Scientific, Milan, Italy) and incubated at 44°C for 48 hr. From each plate, a variable number (from one to three) of colonies presumptively classified as glucuronidase positive, was selected for purification, streaking them on tryptone bile x-glucuronide (TBX) agar plates. Once purified, isolates were grown on Brain and Heart Infusion broth (Thermo Fisher Scientific), incubated at 37°C for 24 hr. Sterile glycerol (20%) was then added to each broth culture and isolates were stored at −80°C for further analysis.

E. coli identification through matrix assisted laser desorption ionization-time of flight mass spectrometry

Isolated bacteria were identified by matrix assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (MS) using a standard extraction method as previously described.¹⁸ Calibration was previously performed with a bacterial test standard (Bruker, Germany) containing extract of E. coli DH5 alpha. Measurements were performed with an UltrafleXtreme MALDI TOF mass spectrometer (Bruker, Germany) equipped with a 1,000 Hz Nd-YAG laser (neodymium-doped yttrium aluminium garnet).

The mass spectra obtained from each isolate was processed with the MALDI Biotyper 3.0 software package (Bruker, Germany) and results were shown as the top 10 identification matches along with confidence scores ranging from 0.00 to 3.00. According to the criteria recommended by the manufacturer, a log(score) below 1.70 does not allow reliable identification; a log(score) between 1.70 and 1.99 allows identification at the genus level; a log(score) between 2.00 and 2.29 means highly probable identification at the genus level and probable identification at the species level; and a log(score) higher than 2.30 (2.30–3.00) indicates highly probable identification at the species level. Analysis of each sample was performed in triplicate (three spots for each sample).

Antibiotic susceptibility testing

Antibiotic susceptibility testing was performed by agar disk diffusion method according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) recommendations (EUCAST Disk Diffusion Method for Antimicrobial Susceptibility Testing Version 6.0, January 2017, www.eucast.org). Isolates were revitalized on Tryptone Soy agar plates (Thermo Fisher Scientific). Bacterial suspensions with a turbidity equivalent to McFarland Standard 0.5 (approximately corresponding to 1–2 × 10⁸ colony forming unit/mL) were swabbed on Mueller Hinton agar plates (Thermo Fisher Scientific) with a sterile cotton swab. Antibiotic disks containing ampicillin (AM, 10 μg), amoxicillin/clavulanic acid (AMC, 20–10 μg), cefoxitin (FOX, 30 μg), cephalothin (KF, 30 μg), cefotaxime (CTX, 30 μg), chloramphenicol (C, 30 μg), tetracycline (TE, 30 μg), trimethoprim/sulfamethoxazole (SXT, 19:1; 25 μg), enrofloxacin (ENR, 10 μg), gentamicin (CN, 10 μg), streptomycin (S, 10 μg), imipenem (IPM, 10 μg), and aztreonam (ATM, 30 μg) (Kairosafe Srl, Trieste, Italy) were placed on the plates. Inhibition diameter zones, including the diameter of the 6-mm disks, were measured after incubation at 35°C for 16–20 hr. Isolates were classified as resistant, intermediate, and susceptible according to breakpoints provided by EUCAST or CLSI.^19,20

DNA extraction

E. coli isolates were cultured on Tryptic Soy Agar at 37°C for ∼20 hr. Template DNA was isolated using a GeneMATRIX Bacterial & Yeast Genomic DNA Purification Kit (Eurx, Poland) following the manufacturer's instructions.

Detection of antibiotic resistance genes

The presence of genes encoding resistance to β-lactam antibiotics (bla-_TEM; bla-_CMY2), streptomycin (aadA1; strA-strB), tetracycline (tet[A], tet[B], and tet[G]), and sulfonamide (sul1, sul2, sul3) was assessed by polymerase chain reaction (PCR) using gene-specific primers and cycling programs according to various authors (Table 1). The reaction mixture (25 μL) was composed as follows: 2.5 μL of Dream Taq PCR buffer, 0.12 μL of Dream Taq DNA polymerase (5 U/mL; Thermo Scientific), 1.25 μL of 8 mM deoxynucleoside triphosphates (Blirt, Poland), 0.8 μL of each primer (10 pmol/μL, Genomed, Poland), 1 μL of template DNA (20 ng), and 18.5 μL of water (Sigma, Poland).

Table 1.

Targeted Resistance Gene, Primers Sequences, Annealing Temperature, Expected Amplicon Size, and References

Gene	Primers (5′-3′)	Annealing (°C)	Amplicon size (bp)	Refs.
bla-_TEM	GTGGACAAAGGTACAACGAG	50	857	Maynard et al.³⁶ (2003)
bla-_TEM	CGGTAAAGTTCGTCACACAC	50	857	Maynard et al.³⁶ (2003)
bla-_CMY2	GACAGCCTCTTTCTCCACA	54	1,000	Kozak et al.³⁷ (2009)
bla-_CMY2	TGGACACGAAGGCTACGTA	54	1,000	Kozak et al.³⁷ (2009)
strA-strB	ATGGTGGACCCT AAAACTCT	62	891	Tamang et al.³⁸ (2007)
strA-strB	CGTCTAGGATCGAGACAAAG	62	891	Tamang et al.³⁸ (2007)
aadA1	GTGGATGGCGGCCTGAAGCC	65	525	Kozak et al.³⁷ (2009)
aadA1	AATGCCCAGTCGGCAGCG	65	525	Kozak et al.³⁷ (2009)
tet(A)	GCTACATCCTGCTTGCCTTC	64	210	Dahshan et al.³⁹ (2010)
tet(A)	CATAGATCGCCGTGAAGAGG	64	210	Dahshan et al.³⁹ (2010)
tet(B)	TTGGTTAGGGGCAAGTTTTG	64	659	Dahshan et al.³⁹ (2010)
tet(B)	GTAATGGGCCAATAACACCG	64	659	Dahshan et al.³⁹ (2010)
tet(G)	GCTCGGTGGTATCTCTGCTC	59	468	Dahshan et al.³⁹ (2010)
tet(G)	AGCAACAGAATCGGGAACAC	59	468	Dahshan et al.³⁹ (2010)
sul1	TGGTGACGGTGTTCGGCATTC	63	789	Costa et al.⁶
sul1	GCGAGGGTTTCCGAGAAGGTG	63	789	Costa et al.⁶
sul2	CGGCATCGTCAACATAACC	50	722	Costa et al.⁶
sul2	GTGTGCGGATGAAAGTCAG	50	722	Costa et al.⁶
sul3	GAGCAAGATTTTTGGAATCG	51	792	Costa et al.⁶
sul3	CATCTGCAGCTAACCTAGGGTTTGGA	51	792	Costa et al.⁶

Detection of virulence-associated genes

PCR was employed to detect the presence of 15 genes associated with E. coli virulence traits. A multiplex-PCR by Paton and Paton^21,22 was used for the detection of stx1, stx2, hylA, eaeA, and saa genes. The cycling conditions were as follows: initial denaturation at 95°C for 5 min, 10 cycles of 95°C for 45 sec, 65°C for 45 sec, 72°C for 75 sec, 20 cycles of 95°C for 45 sec, 62°C for 45 sec, 72°C for 75 sec, and a final extension step at 72°C for 8 min. Detection of ecsV, ent, bfpB, invE, astA, aggR, pic, elt, estIa, and estIb was performed according to Müller et al..²³ Thermocycling conditions were 95°C for 5 min, 30 cycles of 95°C for 45 sec, 62°C for 45 sec, 72°C for 75 sec, and a final extension step at 72°C for 8 min. Both multiplex PCR reactions were performed in an Eppendorf Mastercycler using Dream Taq polymerase (Thermo Scientific). Table 2 shows the primer sequences and concentrations in the final volume, expected amplicon sizes, and associated putative pathovars.

Table 2.

Targeted Virulence Factors Genes, Primers Sequences, Primers Concentrations, Expected Amplicon Sizes, Associated Pathovars, and References

Gene	Primers (5′-3′)	Concentration (μM)	Amplicon size (bp)	Associated pathovars	Refs.
escV	ATTCTGGCTCTCTTCTTCTTTATGGCTG	0.4	544	EHEC, EPEC	Müller et al.²³
escV	CGTCCCCTTTTACAAACTTCATCGC	0.4	544
ent	TGGGCTAAAAGAAGACACACTG	0.4	629
ent	CAAGCATCCTGATTATCTCACC	0.4	629
eaeA	GACCCGGCACAAGCATAAGC	0.3	384		Paton and Paton²¹
eaeA	CCACCTGCAGCAACAAGAGG	0.3	384		Paton and Paton²¹
bfpB	GACACCTCATTGCTGAAGTCG	0.1	910	Typical EPEC	Müller et al. ²³
bfpB	CCAGAACACCTCCGTTATGC	0.1	910	Typical EPEC	Müller et al. ²³
stx1	ATAAATCGCCATTCGTTGACTAC	0.3	180	EHEC	Paton and Paton²¹
stx1	AGAACGCCCACTGAGATCATC		180
stx2	GGCACTGTCTGAAACTGCTCC		255
stx2	TCGCCAGTTATCTGACATTCTG		255
hlyA	GCATCATCAAGCGTACGTTCC		534
hlyA	AATGAGCCAAGCTGGTTAAGCT		534
saa	CGTGATGAACAGGCTATTGC		119
saa	ATGGACATGCCTGTGGCAAC		119
invE	CGATAGATGGCGAGAAATTATATCCCG	0.2	766	EIEC	Müller et al.²³
invE	CGATCAAGAATCCCTAACAGAAGAATCAC	0.2	766	EIEC
astA	TGCCATCAACACAGTATATCCG	0.4	102	EAEC
astA	ACGGCTTTGTAGTCCTTCCAT	0.4	102
aggR	ACGCAGAGTTGCCTGATAAAG	0.2	400
aggR	AATACAGAATCGTCAGCATCAGC	0.2	400
pic	AGCCGTTTCCGCAGAAGCC	0.2	1,111
pic	AAATGTCAGTGAACCGACGATTGG	0.2	1,111
elt	GAACAGGAGGTTTCTGCGTTAGGTG	0.1	655	ETEC
elt	CTTTCAATGGCTTTTTTTTGGGAGTC	0.1	655
estIa	CCTCTTTTAGYCAGACARCTGAATCASTTG	0.4	157
estIa	CAGGCAGGATTACAACAAAGTTCACAG	0.4	157
estIb	TGTCTTTTTCACCTTTCGCTC	0.2	171
estIb	CGGTACAAGCAGGATTACAACAC	0.2	171

PCR products (8 μL) were separated by electrophoresis (100 V) on 2% agarose gels and visualized by ethidium bromide staining. PCR product sizes were determined by comparison with a M100-M 1000 bp DNA Ladder (Blirt, Poland).

Results

Isolation and identification of E. coli through MALDI-TOF MS

Presumptive E. coli colonies were successfully obtained from 67 out of 85 fecal samples (78.8%). Of the 96 isolates selected on TBX agar, 71 (74%) were identified as E. coli by MALDI-TOF-MS.

For 68 (95.8%) isolates, the identification value ranged from 2.000 to 2.299, while for 3 (4.2%) it ranged from 1.943 to 1.989. For all isolates and for all the three replicates, the first and the second-best matches generated by Biotyper 3.0 indicated the same species (E. coli) and therefore identification at the species level was considered as reliable. Sixty-three (74.1%) E. coli positive fecal samples were obtained.

Antibiotic susceptibility testing

Thirty-two out of 71 isolates (45.1%) were susceptible to all the tested molecules. The highest resistance rates were detected for KF (39.4%) and AM (33.8%), followed by AMC (15.5%) and S (12.7%). Low percentages of resistance were observed for TE (5.6%), SXT (2.8%), CN (1.4%), CTX (1.4%), and ENR (1.4%). All the isolates were susceptible to FOX, IPM, and C. As for ATM, one intermediate isolate was obtained (isolate 13), whereas the others were all susceptible.

Table 3 shows the resistance phenotypes observed among the 71 isolates. The most frequently expressed phenotypes were those of exclusive resistance against KF (16.9%) or KF associated with AM (12.7%). Eight isolates (11.3%) showed multiresistance, defined as resistance against three or more chemotherapeutic agents. In particular, isolate 66 from a wild goat and isolate 52 from a fox were resistant to 5 and 8 antibiotics, respectively.

Table 3.

Resistance Phenotypes (Number and Percentage of Detection) Observed Among the 71 Escherichia coli Isolates from Wildlife

Resistance phenotypes	No. of isolates	Percentage of isolates
Susceptible to all tested molecules	32	45.1
KF	12	16.9
AM-KF	9	12.7
AM-AMC	4	5.6
AM-S	2	2.8
AMC	2	2.8
AM-AMC-KF-S	1	1.4
AM-AMC-KF	1	1.4
AM	1	1.4
S	1	1.4
AM-AMC-TE	1	1.4
AM-KF-CTX-S	1	1.4
AM-KF-SXT-S	1	1.4
AM-KF-TE-S	1	1.4
AM-AMC-KF-TE-S	1	1.4
AM-AMC-KF-TE-SXT-ENR-CN-S	1	1.4
Resistant to all tested molecules	0	0

AM, amoxicillin; AMC, amoxicillin + clavulanic acid; KF, cephalothin; CTX, cefotaxime; TE, tetracycline; SXT, trimethoprim/sulfamethoxazole; ENR, enrofloxacin; CN, gentamicin; S, streptomycin.

Detection of antibiotic resistance genes

All E. coli isolates were subjected to endpoint PCR to detect some antibiotic resistance genes. Forty-three out of 71 isolates (60.6%) harbored at least one of the targeted genes. All the isolates were negative for bla_TEM. Although none of the observed resistance phenotypes suggested the possibility for extended spectrum beta-lactamases (ESBLs) production, bla_CMY_-2 gene was surprisingly detected in 54.9% (39/71) of the E. coli isolates. Among these, 53.8% (21/39) showed a resistance phenotype to different β-lactams (AM, KF, CTX, but never FOX), whereas 46.2% (18/39) did not show any resistance phenotype against β-lactams. A percentage of 23.9% (17/71) showed resistance against at least one of the tested β-lactams, but did not harbor bla_CMY_-2, the remaining 21.1% (15/71) did not show any phenotypic resistance and were negative for bla_CMY-2.

Resistance to SXT was observed in 2/71 isolates (2.8%), both positive to sul2, whereas sul1 and sul3 were not detected. All the SXT-susceptible isolates were negative for sul1, sul2, and sul3.

Out of 71 E. coli isolates, 4 (5.6%) showed a resistance phenotype against S and tested positive for strA-strB or aadA1; conversely, 7.0% isolates (5/71) were resistant to S, but no gene coding for this resistance was identified. Furthermore, 1.4% of the isolates (1/71) were susceptible to S, and harbored both aadA1 gene and strA-strB genes. Finally, 1.4% of the isolates (1/71) exhibited an intermediate phenotype against S and resulted in an aadA1 gene carrier.

With regard to TE, resistance was detected in 5.6% (4/71) of the isolates. No discrepancies were observed between phenotype and genotype, since tet(B) was detected in all resistant isolates, whereas tet(A) was found in 25% of the resistant isolates (1/4). Lastly, tet(G) was not detected.

Table 4 shows the resistance phenotype associated with the 43 isolates harboring at least one antibiotic resistance gene.

Table 4.

Resistance Phenotypes, Resistance Genotypes, Virulence Genes and Putative-Associated Pathovars in 52 Isolates Harboring at Least One Resistance or Virulence Gene

Isolate ID	Animal species	Resistance phenotype	Resistance genotype	Virulence associated genes	Putative pathovars
A1	Wolf	AM-AMC-KF	bla_CMY-2	—	—
A3	Wolf	AM-KF	bla_CMY-2	astA	VARIOUS
A5a	Wolf	AMC	bla_CMY-2	astA	VARIOUS
A5b	Wolf	KF	bla_CMY-2	astA	VARIOUS
A6b1	Wolf	AM-KF	bla_CMY-2	—
A6b2	Wolf	KF	bla_CMY-2	—
A9b	Wolf	AM-S	bla_CMY-2	astA	VARIOUS
A10b	Wolf	AM-KF	bla_CMY-2	—
1a	Wolf	AM-KF-CTX-S	bla_CMY-2	—
3	Badger	AM-AMC-KF-S	bla_CMY-2	—
4	Mouflon	—	bla_CMY-2	—
5a	Badger	—	bla_CMY-2	—
5b	Badger	—	bla_CMY-2	escV-eaeA	aEPEC
6a	Mouflon	KF	bla_CMY-2	escV-eaeA	aEPEC
10a	Fox	S	bla_CMY-2	—
11	Badger	—	bla_CMY-2	—
12a	Wolf	AM-AMC	bla_CMY-2	—
12b	Wolf	AM-KF	bla_CMY-2	—
12c	Wolf	KF	bla_CMY-2	—
13	Wolf	—	bla_CMY-2	—
14	Wolf	AM-AMC	bla_CMY-2	—
19	Wild boar	AM-AMC	bla_CMY-2	astA-escV-eaeA	aEPEC
20a	Badger	—	bla_CMY-2	—
20b	Badger	—	bla_CMY-2	—
21a	Badger	—	bla_CMY-2	astA	VARIOUS
22a	Wild boar	—	bla_CMY-2	astA-stx1-stx2-hlyA	EHEC
24a	Mouflon	—	bla_CMY-2	—
25a	Mouflon	—	bla_CMY-2	—
25c	Mouflon	AM-KF	bla_CMY-2	astA	VARIOUS
33a	Wolf	—	bla_CMY-2	—
34a	fox	AM-AMC-TE	bla_CMY-2-aadA1-tet(B)	—
35a	Wild boar	AM-S	bla_CMY-2	astA	VARIOUS
36a	Wolf	—	—-	escV-eaeA	aEPEC
37a	Wild goat	—	bla_CMY-2	—
40	Mouflon	—	bla_CMY-2	escV-eaeA	aEPEC
43	Wolf	KF	bla_CMY-2	escV	aEPEC
46	Wild goat	—	—	astA	VARIOUS
47	wild Wgoat	—	bla_CMY-2	astA	VARIOUS
49	Fox	—	bla_CMY-2-strA/strB-aadA1	astA	VARIOUS
50	Fox	AM-KF-SXT-S	strA/strB-sul2	astA	VARIOUS
52	Fox	AM-AMC-KF-TE-SXT-ENR-CN-S	strA/strB-tet(A)-tet(B)-sul2	—
54	Wolf	AM-KF-TE-S	strA/strB-tet(B)	—
58	Hare	—	—	astA	VARIOUS
59	Wolf	—	bla_CMY-2	escV-eaeA	aEPEC
61	Wolf	AMC	bla_CMY-2	—
63	Fox	—	—	escV-eaeA	aEPEC
64	Wolf	—	—	escV-eaeA	aEPEC
66	Wild goat	AM-AMC-KF-TE-S	strA/strB-tet(B)	—
68	Red deer	—	—	astA	VARIOUS
72	Hare	—	—	astA	VARIOUS
73	Roe deer	—	—	astA-stx2-hlyA	EHEC
74	Red deer	—	—	astA	VARIOUS

Detection of virulence-associated genes

The 71 E. coli isolates were subjected to multiplex-PCR for the detection of gene-encoding pathogenicity factors. Twenty-six out of 71 isolates (36.6%) were positive for at least one gene; 25.4% isolates (18/71) carried astA, 12.7% (9/71) escV, 11.3% (8/71) eaeA, 2.8% (2/71) hlyA and stx2, and 1.4% (1/71) carried stx1. Ent, bfpB, invE, aggR, pic, elt, estIa, estIb, and saa were not detected. Table 5 shows the different gene combinations detected together with their associated pathovars. Fifteen out of 71 isolates (21.1%) carried only astA (EAEC), whereas 9.9% of the isolates (7/71) presented with escV together with eaeA (aEPEC); single isolates (1.4%) were instead positive to escV (aEPEC), escV associated with astA and eaeA (aEPEC), astA associated with stx2 and hlyA (EHEC), or to astA associated with stx1, stx2, and hlyA (EHEC).

Table 5.

Virulence Gene Combinations (Number and Percentage of Isolates) and Associated Putative Pathovars Detected Among 71 Escherichia coli Isolates from Wildlife

Virulence gene combination	No. of isolates	Percentage of isolates	Putative pathovars
No genes detected	45	63.4	—
astA	15	21.1	VARIOUS
escV—eaeA	7	9.9	aEPEC
escV	1	1.4	aEPEC
astA—escV—eaeA	1	1.4	aEPEC
astA—stx2—hlyA	1	1.4	EHEC
astA—stx1—stx2—hlyA	1	1.4	EHEC

Discussion

As reviewed by Vittecoq et al.,⁵ several authors reported that depending on the diet, the incidence of resistant strains could vary, with carnivores more prone to be colonized by resistant bacteria than herbivores. The present study took into consideration antibiotic resistance and virulence factors in E. coli isolates from fecal samples from different animal species in the Apuan Alps Regional Park (Tuscany; Italy). Most of the collected samples (64.7%) and isolates (56.3%) belonged to Canidae (wolves, and foxes), whereas a lower number of samples belonged to other animal species. Since canids as predators occupy the top position on the trophic pyramid, they carry and accumulate microorganisms from all the lower levels. Thus, the rate of antibiotic resistance detected from these animals could be considered representative of the entire ecosystem.

Globally, high percentages of resistance against KF (39.4%) and AM (33.8%) were obtained, followed by AMC (15.5%), and S (12.7%). Resistance against TE, SXT, CN, CTX, and ENR was much lower (5.6%, 2.8%, 1.4%, 1.4%, and 1.4%, respectively). The same situation can be observed when considering the sole family of Canidae, with the highest rates of resistance against KF and AM (47.5% and 40%, respectively), followed by AMC (20%), S (15%), TE (7.5%), SXT (5%), CTX, ENR, and CN (all with 2.5%).

Studies by different authors reported quite different results and highlighted overall high rates of resistance against TE. Most of the European studies focusing on AMR in wild animals from natural parks were performed in Portugal. However, the comparison with other studies seems to be difficult, due to the variability of species included in the sampling plans.

In a study similar to ours, Gonçalves et al.²⁴ studied a Portuguese population of Iberian wolves and isolated E. coli from 195 fecal samples. In contrast to our observations, they detected the highest levels of resistance against TE (30%), AM (25%), S (25%) SXT (12%), and C (5%), whereas resistance to AMC, FOX, and CN was lower than 1%. Taking into account studies looking at top predators from a natural reserve, Gonçalves et al.²⁵ studied the AMR in E. coli from 30 fecal samples of Iberian lynx (Lynx pardinus), which is an endangered species living in Southern Spain. Among 18 recovered E. coli isolates, they detected high levels of resistance against TE (33%), S and nalidixic acid (28%), and SXT (22%). Taken together, these results suggest that in the Iberian natural parks there is a different distribution of resistance determinants compared to what was observed in the Apuan Alps Regional Park.

Costa et al.⁶ in accordance with Gonçalves et al.,^24,25 analyzing 112 E. coli from 72 fecal samples from a wide variety of different wild animals, also confirmed the highest resistance rate against TE (34.8%), which was followed by AM (22.3%), S (22.3%) and SXT (18.8%) in the Iberian Peninsula. Lower percentages of resistance were observed against AMC (7.1%), C (6.3%), CN (6.3%), CTX (1.8%), and ATM (1.8%). No resistance was detected against FOX and IPM. In Portugal, Dias et al.²⁶ analyzing 152 E. coli from 67 fecal samples from wild ungulates, which represent the main prey for wolves, detected the highest resistance against AMC (16.45%); followed by AM (9.87%), TE (8.55%), S (4.61%), and FOX (0.66%). All isolates were susceptible to ATM, CTX, C, and IPM. The reported AMC resistance rate was higher than that of AM. This is an unusual result, since resistance to AMC is usually equal to or lower than AM resistance. We also observed few E. coli isolates resistant to AMC and susceptible to AM. As suggested by Dias et al.,²⁶ this uncommon phenotype is not related to a particular resistance mechanism but is probably due to the use of EUCAST AMC cutoff value for interpretation of inhibition diameter zone (R < 19 mm), instead of that proposed by CLSI, which is lower (R < 14 mm).

In Italy, few studies on AMR in E. coli from wildlife are available, and none of them focus on wolf populations. In particular, Foti et al.^16,17 and more recently Botti et al.¹³ studied AMR in several Enterobacteriaceae isolated from wildlife. In both their works Foti et al.^16,17 focused on AMR from wild birds passing through the territory of Ustica island, Sicily (Italy), and Metaponto, Basilicata (Italy), respectively. In their first work, the authors isolated 183 strains belonging to 28 different species of the Enterobacteriaceae family, among which E. coli (53 strains) was identified. Despite the different animal species considered, similar to what we obtained, they reported the highest percentages of resistance against AM (42.6%), AMC (42.6%), and S (43.7%) and lower resistance rates against TE (6.6%), C (4.4%), FOX (1.2%), and SXT (0.6%). In their more recent work, Foti et al.¹⁷ analyzed 121 cloacal swabs from migratory birds, they isolated 122 strains belonging to 18 distinct species, none of them belonging to E. coli. They observed the highest resistance rates against amoxicillin (64.8%), AM (63.1%), rifampicin (61.5%), and AMC (54.1%), whereas lower rates were detected for IPM (25.4%) and meropenem (6.6%). Only, Botti et al.¹³ looked at the presence of antibiotic-resistant microorganisms from a wide collection of wild animal samples (2,713) collected during the period 2002–2010 in North Western Italy. The sampling plan included a substantial proportion of canids, all belonging to Vulpes vulpes (1,222), but also birds (1,101), mustelids (221), rodents (100), and ungulates (69). However, they focused on Salmonella and observed the highest resistance values against tetracycline class.

With respect to antibiotic resistance genes, none of our isolates demonstrated a typical ESBL-phenotype, which is characterized by resistance to penicillin; first-, second-, and third-generation cephalosporins and aztreonam. Thus, the high percentage of bla_CMY2-positive isolates (54.9% considering all isolates and 52.5% for canids) was unexpected. Indeed, bla_CMY2 encodes an AmpC type beta-lactamase (cephamycinase) and is the most widespread plasmid-borne β-lactamase detected from E. coli and Salmonella spp. of animal origin.²⁷ Our result was surprising since the presence of bla_CMY2 is associated with a broader spectrum of antimicrobial activity than those observed among the studied isolates. Indeed, isolates harboring bla_CMY2 were in almost all cases, AM-resistant and FOX-susceptible or even susceptible to all β-lactams tested. Hinthong et al.²⁸ detected isolates displaying the same characteristic among nonpathogenic E. coli from a water supply in Thailand. Davis et al.²⁹ demonstrated that in the absence of a selective pressure, which can be observed in natural settings, where antibiotic presence is limited, microorganisms are more prone to accumulate mutations leading to dysfunctional pseudogenes (also antibiotic resistance pseudogene), which lost their ability to be expressed. In our specific case, to confirm this hypothesis, sequencing analysis of amplification products would be needed.

Regarding S antibiotic resistance genes, we detected some discrepancies between phenotypes and genotypes, since five resistant isolates did not harbor any of the investigated genes, one intermediate isolate harbored aadA1, and one susceptible isolate harbored both aadA1 and strA-strB. The isolates not possessing any of these two genes but displaying a resistant phenotype may harbor other genes mediating S resistance, or resistance could be due to chromosomal mutations altering the ribosomal binding site of S. Sunde and Norström³⁰ observed that aadA genes are involved in low levels of S-resistance, whereas strA-strB genes probably confer high-level resistance to S, this could explain the intermediate phenotype. Davis et al.²⁹ observed in their study many isolates with inactive S-resistance gene, mostly strA-strB and aadA2 consistent with our observation of susceptible phenotype.

The SXT- and TE-resistant isolates which were in a lower percentage, all presented with a consistent genotype. In particular, sul2 was detected in the two SXT-resistant isolates from a fox, tet(B) was detected in all TE resistant isolates, whereas tet(A) was detected in one out of four (25%) TE-resistant isolates. Isolates carrying tet genes—with the exception of one isolate–were of canid origin. The presence of these genes among isolates from wild animals, especially predators, was also reported by Gonçalves et al.^31,24 and Costa et al.⁶ who considered different animal species.

Understanding the distribution of zoonotic pathogens in specific ecological niches is fundamental to evaluate the risk derived from the exposure to that specific environment. Thus, all the isolates were screened for the presence of some of the main virulence-associated genes. Most of the isolates (63.4%) did not present any of the tested genes, suggesting a prevalent circulation of nonpathogenic isolates. This percentage increases, if only Canidae samples are considered (72.5%). Moreover, 21.1% total isolates and 1.5% canid isolates presented only astA encoding for EAST1, a heat-stable enterotoxin. This gene was at first associated with EAEC, but was later detected in other pathotypes. The role of astA as marker for pathogenic E. coli is still unclear, as it was reported that astA contributes to pathogenicity only when in combination with other virulence-associated genes.³² Eight isolates (4 of them from canids) presented an EPEC putative pathotype, since they showed the presence of eaeA and lacked stx1 and stx2. In particular, these isolates could be ascribed to the atypical EPEC (aEPEC) group, due to the absence of bfp encoding the bundle-encoding pilus characteristic of typical EPEC (tEPEC). An additional isolate from a wolf presented with escV alone and could be attributed to the aEPEC group as well. This is in accordance with several authors reporting that while humans seem to be the only reservoir of tEPEC, aEPEC can be isolated from humans as well as from a wide variety of animals.^33,34 Some aEPEC strains from animals have been associated to human diseases, suggesting that animals could represent important reservoirs of zoonotic aEPEC.³⁵ Moreover, two putative EHEC isolates were detected from a wild boar and roe deer.

Conclusion

The present work highlights the prevalent presence of nonpathogenic E. coli with antibiotic-resistant traits in fecal samples from wild animals living in a natural park with a minimum exposure to a selective pressure. The highest percentages of resistance were observed against a first-generation cephalosporin (cephalothin) and ampicillin, whereas ESBL producers were not detected. Most of the analyzed samples belonged to canids, such as wolves and foxes, top predator carnivorous species, able to acquire antibiotic-resistant genes from diverse environments and disperse them across large distances. Indeed, due to the reduction of natural prey, farm animals become an important nutritional source, pushing predators toward anthropized settlements. In addition, considering their wide home range, such species may play a key role concerning AMR dynamics in natural ecosystems. On the other hand, species belonging to wild ungulates, especially wild boars, are frequently present near houses and farms, meaning that they could represent an important epidemiological link between domestic animals, humans, and wildlife.

Footnotes

Acknowledgment

The authors would like to thank Dr. Anna Paola Biagi for the English revision. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Disclosure Statement

No competing financial interests exist.

References

World Health Organization, 2014 Antimicrobial resistance global report on surveillance: 2014 summary. Available at www.who.int/drugresistance/documents/surveillancereport/en

Singer

R.S.

, Finch

, Wegener

H.C.

, Bywater

, Walters

, and Lipsitch

. 2003. Antibiotic resistance—the interplay between antibiotic use in animals and human beings. Lancet Infect. Dis. 3:47–51.

Tasho

R.P.

, and Cho

J.Y.

. 2017. Entry routes of veterinary antibiotics in the environment. In: M. Hashmi, V. Strezov, and A. Varma (eds.), Antibiotics and antibiotics resistance genes in soils. Soil biology. Vol 51. Springer,, Cham, pp. 55–70.

Radhouani

, Silva

, Poeta

, Torres

, Correia

, and Igrejas

. 2014. Potential impact of antimicrobial resistance in wildlife, environment and human health. Front. Microbiol. 5:23.

Vittecoq

, Godreuil

, Prugnolle

, Durand

, Brazier

, Renaud

, Arnal

, Aberkane

, Jean-Pierre

, Gauthier-Clerc

, Thomas

, and Renaud

. 2016. Antimicrobial resistance in wildlife. J. Appl. Ecol. 53:519–529.

Costa

, Poeta

, Saenz

, Vinué

, Cohelo

A.C.

, Matos

, Rojo-Bezares

, Rodrigues

, and Torres

. 2008. Mechanisms of antibiotic resistance in Escherichia coli isolates recovered from wild animals. Microb. Drug Resist. 14:71–77.

Poeta

, Radhouani

, Igrejas

, Gonçalves

, Carvalho

, Rodrigues

, Vinué

, Somalo

, and Torres

. 2008. Seagulls of the Berlengas natural reserve of Portugal as carriers of fecal Escherichia coli harboring CTX-M and TEM extended-spectrum beta-lactamases. Appl. Environ. Microbiol. 74:7439–7441.

Poeta

, Radhouani

, Pinto

, Martinho

, Rego

, Rodrigues

, Gonçalves

, Rodrigues

, Estepa

, Torres

, and Igrejas

. 2009. Wild boars as reservoirs of extended‐spectrum beta‐lactamase (ESBL) producing Escherichia coli of different phylogenetic groups. J. Basic Microbiol. 49:584–588.

Radhouani

, Pinto

, Coelho

, Gonçalves

, Sargo

, Torres

, Igejas

, and Poeta

. 2010. Detection of Escherichia coli harbouring extended-spectrum β-lactamases of the CTX-M classes in faecal samples of common buzzards (Buteo buteo). J. Antimicrob. Chemother. 65:171–173.

10.

Radhouani

, Igrejas

, Gonçalves

, Estepa

, Sargo

, Torres

, and Poeta

. 2013. Molecular characterization of extended-spectrum-beta-lactamase-producing Escherichia coli isolates from red foxes in Portugal. Arch. Microbiol. 195:141–144.

11.

Klibi

, Amor

I.B.

, Rahmouni

, Dziri

, Douja

, Said

L.B.

, Lozano

, Boudabous

, Slama

K.B.

, Mansouri

, and Torres

. 2015. Diversity of species and antibiotic resistance among fecal enterococci from wild birds in Tunisia. Detection of vanA-containing Enterococcus faecium isolates. Eur. J. Wildlife Res. 61:319–323.

12.

Lozano

, Gonzalez-Barrio

, Camacho

M.C.

, Lima-Barbero

J.F.

, de la Puente

, Höfle

, and Torres

. 2016. Characterization of fecal vancomycin-resistant enterococci with acquired and intrinsic resistance mechanisms in wild animals, Spain. Microb. Ecol. 72:813–820.

13.

Botti

, Navillod

F.V.

, Domenis

, Orusa

, Pepe

, Robetto

, and Guidetti

. 2013. Salmonella spp. and antibiotic-resistant strains in wild mammals and birds in north-western Italy from 2002 to 2010. Vet. Ital. 49:195–202.

14.

Caprioli

, Donelli

, Falbo

, Passi

, Pagano

, and Mantovani

. 1991. Antimicrobial resistance and production of toxins in Escherichia coli strains from wild ruminants and the alpine marmot. J. Wildl. Dis. 27:324–327.

15.

Zottola

, Montagnaro

, Magnapera

, Sasso

, De Martino

, Bragagnolo

, D'Amici

, Condoleo

, Pisanelli

, Iovane

, and Pagnini

. 2013. Prevalence and antimicrobial susceptibility of Salmonella in European wild boar (Sus scrofa); Latium Region–Italy. Comp. Immunol. Microbiol. Infect. Dis. 36:161–168.

16.

Foti

, Rinaldo

, Guercio

, Giacopello

, Aleo

, De Leo

, Fischella

, and Mammina

. 2011. Pathogenic microorganisms carried by migratory birds passing through the territory of the island of Ustica, Sicily (Italy). Avian Pathol. 40:405–409.

17.

Foti

, Mascetti

, Fisichella

, Fulco

, Orlandella

B.M.

, and Piccolo

F.L.

. 2017. Antibiotic resistance assessment in bacteria isolated in migratory Passeriformes transiting through the Metaponto territory (Basilicata, Italy). Avian Res. 8:26.

18.

Dudzic

, Urban-Chmiel

, Stępień-Pyśniak

, Dec

, Puchalski

, and Wernicki

. 2016. Isolation, identification and antibiotic resistance of Campylobacter strains isolated from domestic and free-living pigeons. Br. Poult. Sci. 57:172–178.

19.

The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 8.1, 2018. Available at www.eucast.org.

20.

CLSI. Performance Standards for Antimicrobial Susceptibility Testing; Twenty-Fifth Informational Supplement. 2015. CLSI document M100-S25. Clinical and Laboratory Standards Institute, Wayne, PA.

21.

Paton

A.W.

, and Paton

J.C.

. 1998. Detection and Characterization of Shiga toxigenic Escherichia coli by using multiplex PCR assays for stx 1, stx 2, eaeA, enterohemorrhagic E. coli hlyA, rfb O111, and rfb O157. J. Clin. Microbiol. 36:598–602.

22.

Paton

A.W.

, and Paton

J.C.

. 2002. Direct detection and characterization of Shiga toxigenic Escherichia coli by multiplex PCR for stx1, stx2, eae, ehxA, and saa. J. Clin. Microbiol. 40:271–274.

23.

Müller

, Greune

, Heusipp

, Karch

, Fruth

, Tschäpe

, and Schmidt

M.A.

. 2007. Identification of unconventional intestinal pathogenic Escherichia coli isolates expressing intermediate virulence factor profiles by using a novel single-step multiplex PCR. Appl. Environ. Microbiol. 73:3380–3390.

24.

Gonçalves

, Igrejas

, Radhouani

, Correia

, Pacheco

, Santos

, Monteiro

, Guerra

, Petrucci-Fonseca

, Brito

, Torres

, and Poeta

. 2013. Antimicrobial resistance in faecal enterococci and Escherichia coli isolates recovered from Iberian wolf. Lett. Appl. Microbiol. 56:268–274.

25.

Gonçalves

, Igrejas

, Radhouani

, Santos

, Monteiro

, Pacheco

, Alcaide

, Zorrilla

, Serra

, Torres

, and Poeta

. 2013. Detection of antibiotic resistant enterococci and Escherichia coli in free range Iberian Lynx (Lynx pardinus). Sci. Total Environ. 456:115–119.

26.

Dias

, Torres

R.T.

, Kronvall

, Fonseca

, Mendo

, and Caetano

. 2015. Assessment of antibiotic resistance of Escherichia coli isolates and screening of Salmonella spp. in wild ungulates from Portugal. Res. Microbiol. 166:584–593.

27.

X.Z.

, Mehrotra

, Ghimire

, and Adewoye

. 2007. β-Lactam resistance and β-lactamases in bacteria of animal origin. Vet. Microbiol. 121:197–214.

28.

Hinthong

, Pumipuntu

, Santajit

, Kulpeanprasit

, Buranasinsup

, Sookrung

, Chaicumpa

, Aiumurai

, and Indrawattana

. 2017. Detection and drug resistance profile of Escherichia coli from subclinical mastitis cows and water supply in dairy farms in Saraburi Province, Thailand. PeerJ, 6:e3431.

29.

Davis

M.A.

, Besser

T.E.

, Orfe

L.H.

, Baker

K.N.

, Lanier

A.S.

, Broschat

S.L.

, New

, and Call

D.R.

. 2011. Genotypic-phenotypic discrepancies between antibiotic resistance characteristics of Escherichia coli from calves in high and low antibiotic use management settings. Appl. Environ. Microbiol. 77:3293–3299.

30.

Sunde

, and Norström

. 2005. The genetic background for streptomycin resistance in Escherichia coli influences the distribution of MICs. J. Antimicrob. Chemother. 56:87–90.

31.

Gonçalves

, Igrejas

, Radhouani

, Estepa

, Pacheco

, Monteiro

, Brito

, Guerra

, Petrucci-Fonseca

, Torres

, and Poeta

. 2012. Iberian wolf as a reservoir of extended-spectrum β-lactamase-producing Escherichia coli of the TEM, SHV, and CTX-M groups. Microb. Drug Resist. 18:215–219.

32.

Ngeleka

, Pritchard

, Appleyard

, Middleton

D.M.

, and Fairbrother

J.M.

. 2003. Isolation and association of Escherichia coli AIDA-I/STb, rather than EAST1 pathotype, with diarrhea in piglets and antibiotic sensitivity of isolates. J. Vet. Diagn. Invest. 15:242–252.

33.

Blanco

, Schumacher

, Tasara

, Zweifel

, Blanco

J.E.

, Dahbi

, Blanco

, and Stephan

. 2005. Serotypes, intimin variants and other virulence factors of eae positive Escherichia coli strains isolated from healthy cattle in Switzerland. Identification of a new intimin variant gene (eae-η2). BMC Microbiol. 5:23.

34.

Chandran

, and Mazumder

. 2013. Prevalence of diarrhea-associated virulence genes and genetic diversity in Escherichia coli isolates from fecal material of various animal hosts. Appl. Environ. Microbiol. 79:7371–7380.

35.

Hernandes

R.T.

, Elias

W.P.

, Vieira

M.A.

, and Gomes

T.A.

. 2009. An overview of atypical enteropathogenic Escherichia coli. FEMS Microbiol. Lett. 297:137–149.

36.

Maynard

, Fairbrother

J.M.

, Bekal

, Sanschagrin

, Levesque

R.C.

, Brousseau

, Masson

, Larivière

, and Harel

. 2003. Antimicrobial resistance genes in enterotoxigenic Escherichia coli O149: K91 isolates obtained over a 23-year period from pigs. Antimicrob. Agents Chemother. 47:3214–3221.

37.

Kozak

G.K.

, Boerlin

, Janecko

, Reid-Smith

R.J.

, and Jardine

, 2009. Antimicrobial resistance in Escherichia coli isolates from swine and wild small mammals in the proximity of swine farms and in natural environments in Ontario, Canada. Appl. Environ. Microbiol. 75:559–566.

38.

Tamang

M.D.

, Oh

J.Y.

, Seol

S.Y.

, Kang

H.Y.

, Lee

J.C.

, Lee

Y.C.

, Cho

D.T

, and Kim

. (2007). Emergence of multidrug-resistant Salmonella enterica serovar Typhi associated with a class 1 integron carrying the dfrA7 gene cassette in Nepal. Int. J. Antimicrob. Agents. 30:330–335.

39.

Dahshan

, Shahada

, Chuma

, Moriki

, and Okamoto

. 2010. Genetic analysis of multidrug-resistant Salmonella enterica serovars Stanley and Typhimurium from cattle. Vet. Microbiol. 145:76–83.