OUP user menu

Broad-spectrum β-lactamases among Enterobacteriaceae of animal origin: molecular aspects, mobility and impact on public health

Annemieke Smet, An Martel, Davy Persoons, Jeroen Dewulf, Marc Heyndrickx, Lieve Herman, Freddy Haesebrouck, Patrick Butaye
DOI: http://dx.doi.org/10.1111/j.1574-6976.2009.00198.x 295-316 First published online: 1 May 2010


Broad-spectrum β-lactamase genes (coding for extended-spectrum β-lactamases and AmpC β-lactamases) have been frequently demonstrated in the microbiota of food-producing animals. This may pose a human health hazard as these genes may be present in zoonotic bacteria, which would cause a direct problem. They can also be present in commensals, which may act as a reservoir of resistance genes for pathogens causing disease both in humans and in animals. Broad-spectrum β-lactamase genes are frequently located on mobile genetic elements, such as plasmids, transposons and integrons, which often also carry additional resistance genes. This could limit treatment options for infections caused by broad-spectrum β-lactam-resistant microorganisms. This review addresses the growing burden of broad-spectrum β-lactam resistance among Enterobacteriaceae isolated from food, companion and wild animals worldwide. To explore the human health hazard, the diversity of broad-spectrum β-lactamases among Enterobacteriaceae derived from animals is compared with respect to their presence in human bacteria. Furthermore, the possibilities of the exchange of genes encoding broad-spectrum β-lactamases – including the exchange of the transposons and plasmids that serve as vehicles for these genes – between different ecosystems (human and animal) are discussed.

  • mobile genetic elements
  • human health hazard
  • β-lactamases


Resistance to β-lactams in Enterobacteriaceae is mainly due to the production of β-lactamases, which may be encoded either chromosomally or on plasmids (Bradford, 2005). Resistance to extended-spectrum β-lactams has been associated with the production of broad-spectrum β-lactamases such as extended-spectrum β-lactamases (ESBLs), AmpC β-lactamases and metallo-β-lactamases (MBLs) (Batchelor et al., 2005b). ESBLs confer resistance to most β-lactam antibiotics, but are not active against cephamycins and carbapenems and are inactivated by β-lactamase inhibitors such as clavulanic acid. This is in contrast to AmpC β-lactamases, which are not inhibited by clavulanic acid and usually confer resistance to all β-lactams, with the exception of dipolar ionic methoxy-imino-cephalosporins, such as cefepime and the carbapenems (Bradford, 2005). MBLs can hydrolyse all clinical β-lactam substrates, with the exception of aztreonam.

Broad-spectrum β-lactamase-producing Enterobacteriaceae have increasingly been detected in humans since the early 1990s and in animals since 2000 (Paterson & Bonomo, 2005; Arlet et al., 2006; Bertrand et al., 2006; Cloeckaert et al., 2007; Li et al., 2007; Carattoli, 2008; Pitout & Laupland, 2008; Smet et al., 2008, 2009; Walsh, 2008; Jacoby et al., 2009). Resistance in bacteria of animals and its impact on human health have drawn considerable attention worldwide (Phillips et al., 2004; Aarestrup et al., 2006). The first pathway of resistance transfer is the direct transfer of a pathogen from animals to humans. This is the case for zoonotic agents such as Salmonella, where resistance against β-lactams, including extended-spectrum cephalosporins, has been demonstrated (Bertrand et al., 2006; Cloeckaert et al., 2007). Animals may also harbour resistance genes in their residing commensal flora. Commensal Escherichia coli isolates have been implicated in the transmission of genetic resistance traits (Kruse & Sorum, 1994) because their resistance genes may jump from one bacterium to another, mainly by means of mobile genetic elements (MGEs) such as transposons and plasmids.

The present review analyses the growing burden of β-lactam resistance among Enterobacteriaceae of animals, focusing on the use of β-lactams in veterinary medicine and food-animal production, and on the β-lactamase genes and their ability to be transferred. The diversity of broad-spectrum β-lactamases among Enterobacteriaceae from food-producing, companion and wild animals, as well as from humans, is compared and the possible movement across bacterial populations from different hosts is discussed.

Use of β-lactam antimicrobials in veterinary medicine and food animal production

One of the most important group of antimicrobial agents in veterinary medicine and food animal production are the β-lactams. These antimicrobial agents can be divided into different groups, three of which are used in veterinary medicine: the penicillins, the first- to fourth-generation cephalosporins and the β-lactamase inhibitors. An overview of the antimicrobial activity of the different β-lactam groups and their use in veterinary and human medicine is given in Table 1. The recommended use of β-lactams for the treatment of disease in swine, cattle, horses, poultry, cats and dogs is presented in Table 2.

View this table:
Table 1

β-Lactams used in veterinary and human medicine (Hornish & Kotarski, 2002; Guardabassi et al., 2008; Hammerum & Heuer, 2009)

β-LactamsSpectrum of activityVeterinary medicineHuman medicine
PenicillinsMainly active against Gram-positive bacteria. Aminopenicillins are also active against Gram-negative bacteria.Ampicillin, amoxicillin, benzylpenicillin, cloxacillin, hetacillinPenicillin, ampicillin, amoxicillin
Penicillin-β-lactamase inhibitor combinationsExhibit negligible antimicrobial activity. Their sole purpose is to prevent the inactivation of β-lactam antibiotics and, as such, they are coadministered mostly with penicillinsAmoxicillin-clavulanateAmoxicillin-clavulanate, piperacillin-tazobactam
First-generation cephalosporinsare moderate-spectrum agents. They are effective alternatives for treating staphylococcal and streptococcal infections.Cepadroxil, cefapirin, cephalexinCefalozin
Second-generation cephalosporinsHave a greater Gram-negative spectrum while retaining some activity against Gram-positive bacteria.Cefaclor, cefamandole, cefonicid, ceforanide, cefuroximeCefuroxime, cefoxitin
Third-generation cephalosporinsHave a broad spectrum of activity and further increased activity against Gram-negative organisms.Cefovecin, cefpodoxim, ceftiofurCeftriaxone, cefotaxime, ceftazidime
Fourth-generation cephalosporinsHave the broadest activity both against Gram-negative and Gram-positive bacteria.CefquinomeCefepime
MonobactamsHave a strong activity against susceptible Gram-negative bacteria, but no useful activity against Gram-positive bacteria or anaerobes.Not in useAztreonam
CarbapenemsHave a broad spectrum of activity against aerobic and anaerobic Gram-positive and Gram-negative bacteria.Imipenem, MeropenemImipenem, Meropenem
  • * Use only in the case of life-threatening infections and when susceptibility tests have shown resistance to all other antimicrobials except carbapenems.

View this table:
Table 2

Use of β-lactams as the first or the second choice for the treatment of diseases in swine, cattle, poultry, horses and cats and dogs (modified from Guardabassi et al., 2008)

AnimalClinical indicationβ-Lactam use (first choice)β-Lactam use (second choice)
SwineNecrotic enteritisPenicillins
Respiratory/systemic diseasePenicillins, ceftiofur
CattleNeonatal septicaemiaeThird (ceftiofur)- or fourth (cefquinome)-generation cephalosporins
Calf diarrhoeaAmpicillin, amoxicillin, amoxicillin-clavulanic acid
Septic arthritisAmpicillin, amoxicillinThird- or fourth-generation cephalosporins
Foot rotAmpicillin
MetritisAmpicillin, penicillin
MastitisPenicillinThird- or fourth-generation cephalosporins
CollibacilosisAmpicillin, amoxicillin
Ornithobacterium rhinotracheale infectionAmpicillin, amoxicillin
Fowl choleraAmpicillin, amoxicillin
Riemerella anatipestifer infectionsAmpicillin, amoxicillin
HorsesClostridial myostitisPenicillins
Pigeon feverPenicillins
Septic arthritisPenicillinCeftiofur
CystitisPenicillin, ampicillinceftiofur
Endocarditis, pericarditisPenicillinCeftiofur
Bacterial meningitisPenicillinCeftiofur
ListeriosisPenicillin, ampicillinCeftiofur
Neonatal septicaemiaePenicillinThird-generation cephalosporins
Cats and dogsReccurrent pyodermaAmoxicillin-clavulanic acid, first-generation cephalosporins
Skin woundsCefotaxime, cephalexin
CystitisAminopenicillinsAmoxicillin–clavulanic acid, cephalosporins
Acute peritonitisCefoxitin, cefotetan
PneumoniaeAmoxicillin-clavulanic acid, cephalosporins
OsteomyelitisAmoxicillin-clavulanic acid, cephalosporins
LeptospirosisPenicillin G, amoxicillin
  • * In combination with gentamicin.

  • Ceftiofur is used as the last choice.

The different surveillance programmes around the world have made it possible to obtain good knowledge of β-lactam usage in food and non-food animals (Schwarz & Chaslus-Dancla, 2001; Schwarz et al., 2001; Guardabassi et al., 2008; CDC, 2004; Batchelor et al., 2005a, b; Li et al., 2005; CIPARS, 2006; DANMAP, 2006; McEwan & Fedorka-Cray, 2006; Lloyd et al., 2007; Hammerum & Heuer, 2009). By way of summary, ampicillin and amoxicillin are regarded as the drugs of choice in avian medicine in many continents. These drugs are used in most European countries, with the exception of Finland, Denmark and Sweden. In Spain, amoxicillin–clavulanic acid is also allowed for use (Schwarz & Chaslus-Dancla, 2001). Third-generation cephalosporins are rarely used in poultry and only under very limited conditions for treatment of valuable poultry stocks (Guardabassi et al., 2008). However, ceftiofur, licensed for veterinary use in the United States since 1988, has been given to 1-day-old chickens to prevent early mortality in the United States (Batchelor et al., 2005a, b). In Europe, cephalosporins are not allowed for use in poultry (Schwarz & Chaslus-Dancla, 2001; Smet et al., 2008), although extra-label use occurs.

In cattle, pigs, horses, cats and dogs, aminopenicillins are often used. Third- and fourth-generation cephalosporins have been approved as the second-choice treatment of different disease conditions such as metritis, foot rot and mastitis in cattle, respiratory disease in ruminants, horses and swine and skin diseases in cats and dogs (Bradford et al., 1999; Mason & Kietzmann, 1999; Watson & Rosin, 2000; Batchelor et al., 2005a, b; Guardabassi et al., 2008).

Besides their use in clinical therapy, β-lactams, especially penicillins, have been used as feed additives to improve growth. They were phased out in Europe after the appearance of the ‘Swann Report’ in 1975, in which concerns were raised that the use of antimicrobials for growth promotion could lead to increasing resistance in bacteria of human and animal origin (Anonymous, 1968). However, in the United States, penicillins are still used at subtherapeutic levels for growth promotion (Aarestrup, 2006).

β-Lactam resistance mechanisms

All β-lactam agents have a four-membered β-lactam ring in their structure. A β-lactam ring is a cyclic amide with a heteroatomic ring structure consisting of three carbon atoms and one nitrogen atom. These antimicrobial agents prevent the bacterial cell wall from forming by interfering with the final stage of peptidoglycan synthesis through acting on penicillin-binding proteins (PBPs). The number of PBPs varies between bacterial species and these PBPs are found as both membrane-bound and cytoplasmic proteins. The peptidoglycan layer maintains the cell shape and protects the bacterium against osmotic forces. In Gram-positive bacteria, peptidoglycan forms a thick layer on the cytoplasmic membrane, whereas in Gram-negative bacteria and mycobacteria, peptidoglycan is a thin layer sandwiched between the outer membrane and the cytoplasmic membrane (Greenwood, 2000; Poole, 2004; Giguère et al., 2006).

Bacterial resistance to β-lactams can be due to at least three mechanisms. The first mechanism consists of mutations in genes encoding PBPs, the acquisition of alternative PBPs or the creation of mosaic PBPs. All these altered PBPs have a reduced affinity for β-lactams and as such can retain their function in maintaining the cell wall. The second mechanism consists of a change in the permeability of the cell wall. This may be due to alterations in the expression of porins or active efflux. The third mechanism, and by far the most common one, is the inactivation of the drug by β-lactamases (Massova & Mobashery, 1998; Poole et al., 2004; Shah et al., 2004; Batchelor et al., 2005a, b). In this review, the focus will be on the resistance mediated by β-lactamases, because this is the predominant mechanism of β-lactam resistance in Gram-negative bacteria, especially in Enterobacteriaceae. β-Lactamases inactivate β-lactams by hydrolysing their four-membered β-lactam ring. They break a bond in the β-lactam ring to disable the molecule (Shah et al., 2004).

β-Lactamases: an overview

Until now, >400 β-lactamases have been reported and new β-lactamases continue to emerge worldwide (Jacoby & Medeiros, 1991; Bradford et al., 2005; Gupta et al., 2007; Walsh et al., 2008; Jacoby et al., 2009). The main characteristics of the different families of β-lactamases, as presented in the brief overview below, are well documented on the following website: http://www.lahey.org/Studies/


TEM-type β-lactamases

More than 150 TEM-type β-lactamases have been found, and all of them are derivatives of TEM-1 or TEM-2 by point mutations. TEM-1 was first demonstrated in 1965 in an E. coli isolate from a patient in Athens, Greece, named Temoneira (designation TEM) (Datta & Kontomichalou, 1965). In contrast to the majority of TEM β-lactamases, TEM-1, TEM-2 and TEM-13 are not ESBLs and are only able to hydrolyse penicillins. Some TEM derivatives have been found to have a reduced affinity for β-lactamase inhibitors and are called inhibitor-resistant TEM. These enzymes have negligible activity against extended-spectrum cephalosporins and are not considered to be ESBLs (Chaibi et al., 1999). However, mutants of the TEM derivatives (called CMT-1, CMT-2, CMT-3 and CMT-4) have been identified that have the ability to hydrolyse both third-generation cephalosporins and β-lactamase inhibitors (Fiett et al., 2000; Neuwirth et al., 2001).

Sulphydryl variable (SHV)-type β-lactamases

Another family of β-lactamases are the SHV enzymes. The progenitor of the SHV enzymes, SHV-1, was first described in Klebsiella pneumoniae. SHV-1 confers resistance to broad-spectrum penicillins. In 1983, a Klebsiella ozaenae strain was isolated in Germany possessing an SHV-2 enzyme that efficiently hydrolysed cefotaxime and, to a lesser extent, ceftazidime. To date, >50 SHV derivatives are known, all being derivatives of SHV-1 or SHV-2. Like the TEM-type enzymes, the majority of the SHV enzymes are ESBLs (Paterson & Bonomo, 2005; Gupta et al., 2007).

CTX-M-type β-lactamases

A third family consists of the CTX-M enzymes, which are also ESBLs, and were first isolated in Munich. The designation CTX-M reflects the hydrolytic activity of these β-lactamases against cefotaxime. It appears that the CTX-M-type β-lactamases are closely related to β-lactamases of Kluyvera spp. CTX-M enzymes have 40% or less identity with TEM and SHV-type ESBLs. So far, >70 CTX-M enzymes have been isolated. They are divided into five clusters on the basis of the amino acid sequence: CTX-M-1, CTX-M-2, CTX-M-8, CTX-M-9 and CTX-M-25 (Paterson & Bonomo, 2005; Gupta et al., 2007).

OXA-type β-lactamases

Most OXA-type β-lactamases, so named because of their oxacillin-hydrolysing capabilities, do not hydrolyse extended-spectrum cephalosporins and are not regarded as ESBLs. The exceptions to this rule are OXA-10 and OXA-13 to OXA-19 (Toleman et al., 2003). The evolution of ESBL OXA-type β-lactamases from parent enzymes with narrow spectra has many parallels with the evolution of TEM- and SHV-type ESBLs.

Other examples of ESBLs

PER (Pseudomonas extended-resistant), VEB (Vietnam ESBL), BES (Brazil extended-spectrum), GES (Guiana extended-spectrum), TLA (named after Tlahuicas Indians), SFO (Serratia fonticola) and IBC (integron-borne cephalosporinase) are other examples of ESBLs that have been discovered (Paterson & Bonomo, 2005; Jacoby et al., 2006; Gupta et al., 2007). These enzymes are not so common among Enterobacteriaceae as the ESBLs described above.

AmpC β-lactamases

Another large group of broad-spectrum β-lactamases are the AmpC enzymes, which are typically encoded on the chromosome of many Gram-negative bacteria, including Citrobacter, Serratia and Enterobacter species, where its expression is usually inducible (Jacoby et al., 2009). AmpC-type β-lactamases may be carried β on plasmids of bacterial species lacking the chromosomal ampC gene, such as Klebsiella spp. Plasmid-mediated AmpC enzymes have been found in E. coli, although this species can also increase the production of its normally weakly expressed chromosomal AmpC enzyme by gene duplication or mutation in the ampC promoter or attenuator with consequent enhanced gene expression (Caroff et al., 2000). Plasmid-mediated AmpC enzymes have been named – with an inconsistency typical of β-lactamase nomenclature – according to the resistance produced to cephamycins (CMY, 43 varieties), cefoxitin (FOX, seven varieties), moxalactam (MOX, three varieties) or latamoxef (LAT, four varieties), according to the type of enzyme [ACC (Ambler class C), four varieties or ACT (AmpC type), three varieties] or according to the site of discovery, such as the Miriam Hospital in Providence (MIR-1) or the Dhahran Hospital in Saudi Arabia (DHA, two varieties). BIL-1 was named after the patient (Bilal) who provided the original sample (Philippon et al., 2002).


The MBLs are a molecularly diverse group of broad-spectrum β-lactamases. They are widespread in Pseudomonas aeruginosa, Acinetobacter species [such as VIM (Verona integron-encoded MBLs) and OXA types] and have been more recently detected in Enterobacteriaceae [such as VIM, KPC (K. pneumoniae carbapenamases) and GES types] (Jacoby et al., 2006; Walsh et al., 2008). So far, MBLs have not been detected in bacteria of animals, and so this group of β-lactamases will not be reviewed further in this paper.

Classification of broad-spectrum β-lactamases: a source of confusion

Because of the diversity of the enzymatic characteristics of the β-lactamases, many attempts have been made to categorize these enzymes using their biochemical attributes. Several different classification schemes for bacterial β-lactamases have been described. The two most often used of these systems will be discussed in this chapter.

The first classification system, devised by Bush et al. (1995), is based on the activity of the β-lactamases against different β-lactam antimicrobials (substrate specificity). It contains a wide variety of subgroups. Three major groups of enzymes can be defined: (1) cephalosporinases that are not considerably inhibited by clavulanic acid, (2) penicilinases, cephalosporinases and broad-spectrum β-lactamases that are inhibited by β-lactamase inhibitors and (3) MBLs (carbapenamases) that hydrolyse penicillins, cephalosporins and carbapenems, with the exception of aztreonam (Bush et al., 1995).

Recently, Giske et al. (2009) enlarged the ESBL definition with a view to achieving a balance between scientific and clinical needs. Several points of the Bush classification system were reconsidered, as described below.

First, Bush et al. (1995) described ESBLs as enzymes that can be differentiated from their parental enzymes, which do not hydrolyse extended-spectrum β-lactams.

Secondly, several β-lactamases (e.g. GES-5) that possess carbapenamase activity are still reported in the literature as ESBL.

Third, the Bush classification system does not make a distinction between AmpC β-lactamases of chromosomal origin and those of plasmid origin. However, plasmid-mediated AmpC enzymes have greater clinical importance from an infection control perspective than do their chromosomal counterparts (i.e. the added risk of transferable resistance genes).

Finally, Bush et al. (1995) wrote that the classification scheme does not take into account the actual importance of phenotypic tests for β-lactamases.

Thus, if scientific researchers adopt these points of the Bush classification system, there is little hope for the clinicians being properly informed about the appropriate treatment. Thus, by broadening the ESBL definition, the communication with the clinicians and infection control practitioners will be largely enhanced (Giske et al., 2009).

The second – and most widely used – classification scheme for β-lactamases is the Ambler system, which divides β-lactamases into four classes (A, B, C and D) on the basis of their amino acid sequences. At first, Ambler described two classes: class A β-lactamases (TEM, SHV and CTX-M enzymes), which have their active site at a serine residue, and class B enzymes (MBLs), which utilize a bivalent metal ion (zinc ion) to attack the β-lactam ring (Ambler et al., 1980). Later on, when new insights were acquired, a novel class of serine β-lactamases, class C (AmpC β-lactamases), was defined. These showed little sequence similarity to the class A enzymes. And finally, another new class of serine β-lactamases, known as the OXA β-lactamases (class D), was identified (Ambler et al., 1980).

A revision of the Ambler classification system by Hall & Barlow (2005) presupposed the existence of differences among the MBLs. They proposed to divide the class B enzymes into three subgroups: B1, B2 and B3. Members of subgroups B1 and B2 share many sequence similarities, but differ largely from the members of subgroup B3 (Hall & Barlow, 2005).

These different classification schemes have been and will continue to be updated according to the growing knowledge of the β-lactamases. It is better, however, not to abandon the existing classification systems, because they are widely used and well known, but rather to adapt them to the growing knowledge.

Mobility of broad-spectrum β-lactamase genes among Enterobacteriaceae

β-Lactamase (bla) genes encoding broad-spectrum β-lactamases have mainly been reported in Enterobacteriaceae and are associated with MGEs, such as insertion sequences (ISs), integrons, transposons, plasmids and phage-related elements. It has to be noted that ISs and transposons can hop from the chromosome to a plasmid or between plasmids or back to the chromosome within a bacterial cell, but require a conjugative element (plasmid or conjugative transposon) or phage to be mobilized between bacterial cells. The diversity of these genetic vehicles enhances the spread of broad-spectrum β-lactamases. What follows is an overview of these genetic elements.

MGEs described as carriers of ESBL genes

MGEs for blaTEM genes

All blaTEM genes are carried by three of the earliest described bacterial transposons, namely Tn1, Tn2 and Tn3. These transposons contain the transposase and resolvase genes, tnpA and tnpR, as well as a res resolution site. They are about 99% identical to each other, with most of the differences confined to a region close to the res site (Partridge & Hall, 2005). Most of the structures surrounding the blaTEM genes are Tn3-like transposons possessing 38-bp inverted repeats (Heffron et al., 1979; Partridge & Hall, 2005). Table 3 shows a summary of all known blaTEM gene–transposon associations. Some examples of blaTEM-carrying transposons are given in Fig. 1a and b.

View this table:
Table 3

Known blaTEM gene–transposon associations

bla geneTransposon-like structureReference
bla TEM-1aTn3Partridge & Hall (2005)
bla TEM-1bTn2Partridge & Hall (2005)
bla TEM-2Tn1Partridge & Hall (2005)
bla TEM-3Tn1Mabilat et al. (1992)
bla TEM-10Tn2Bradford et al. (1994)
bla TEM-12Tn2Bradford et al. (1994)
bla TEM-15Tn801 (Tn3-like)Chouchani et al. (2007)
bla TEM-20Tn3Sunde et al. (2009)
bla TEM-21Tn801 (Tn3-like)Dubois et al. (2005)
bla TEM-52Tn3Cloeckaert et al. (2007)
bla TEM-67Tn1Naas et al. (2003)
bla TEM-121Tn3Dubois et al. (2002)
bla TEM-144Tn2Vignoli et al. (2006)
Figure 1

Modular schematic structure of the backbones of MGEs containing ESBL genes. (a) A blaTEM-52-carrying Tn3-like transposon (inverted repeats, vertical black rectangles) (Cloeckaert et al., 2007). (b) A Tn2-like transposon harbouring the blaTEM-144 gene (Rom protein gene, horizontal black rectangles) (Vignoli et al., 2006). (c) A blashv-5 gene, which was found together with seven other colinear genes (ΔigBM, fucA, ygbK, ygbJ and ygbL) originating from Klebsiella pneumoniae and flanked by two IS26 elements inserted in the same orientation (Miriagou et al., 2005). (d) An ISEcp1-like element in association with the blaCTX-M-15 gene and the Tn3-like transposon harbouring a blaTEM-1 gene (Canton & Coque, 2006). (e) A complex class 1 integron, comprising the class 1 integron and its gene cassettes (dfrA12, orfF and aadA12) with 59-be (black ovals), intI1 (integrase gene) and attI (recombination site, black circle), the blaCTX-M-14 gene associated with the ISCR1 element and a duplication of the qacEΔ1/sul1 tandem (Bae et al., 2007). (f) The blaCTX-M-10 gene located downstream of phage-related elements (orf2, orf3, orf4 and DNA invertase) (Oliver et al., 2005; Riaño et al., 2009). (g) A class 1 integron harbouring the gene cassettes (blaVEB-1blaOXA-10-like and arr-2-like gene cassettes) with 59-be (black ovals), intI1 (integrase gene) and attI (recombination site, black circle) (Girlich et al., 2001; Riaño et al., 2009). (h) An In52 class 1 integron structure harbouring the blaGES-1, aac(6′)lb′, dfrXVB, clmA4 aadA2 gene cassettes [59-be (black ovals), intI1 (integrase gene) and attI (recombination site, black circle)] (Poirel et al., 2000). (i) A class 3 integron harbouring the gene cassettes (blaGES-1 gene cassette and a fusion event between blaOXA-10-type and aac(6′)-lb gene cassettes) with 59-be (black ovals), intI3 (integrase gene) and attI (recombination site, black circle) (Correia et al., 2003).

MGEs for blaSHV genes

Few studies have investigated the genetic support of blaSHV-like genes. These genes originate from the chromosome of K. pneumoniae and are spread together with other K. pneumoniae chromosomal DNA fragments from the chromosome to a plasmid within a bacterial cell, following IS26-dependent mobilization (Fig. 1c). Their presence on a conjugative element allows them to spread between Gram-negative bacteria (Teshager et al., 2000; Carattoli et al., 2005; Gangoué-Piéboji et al., 2005; Miriagou et al., 2005; Rankin et al., 2005; Romero et al., 2005; Riaño et al., 2006; Yu et al., 2006; Jouini et al., 2007; Chiaretto et al., 2008; Poirel et al., 2008; Rodriguez-Baño et al., 2008; Tian et al., 2009a, b). So far, the blaTEM and blaSHV genes have never been described in integron structures.

MGEs for blaCTX-M genes

Like the blaSHV genes, blaCTX-M genes are also mobilized from the chromosome of a naturally resistant bacterium. In this case, Kluyvera species are involved (Rodríguez et al., 2004; Olson et al., 2005). Several genetic elements such as the ISEcp1 and ISCR1 and phage-related elements, which have been found to be involved in the mobility of blaCTX-M genes, are discussed below (Oliver et al., 2005; Canton & Coque, 2006; Bae et al., 2007; Poirel et al., 2008).

ISEcp1-like elements belong to the IS1380 family of ISs and have been identified in association with genes belonging to the blaCTX-M-1, blaCTX-M-2, blaCTX-M-25 and blaCTX-M-9 ESBL gene clusters (Bae et al., 2006a, b; Eeckert et al., 2006; Fernandez et al., 2007; Liu et al., 2007; Navon-Venezia et al., 2008; Shen et al., 2008). Extensive analysis has shown that ISEcp1 is responsible for the mobility of a transposition unit including itself and a blaCTX-M gene. This IS element is located upstream of a blaCTX-M gene. A schematic representation of this ISEcp1-like element in association with a blaCTX-M gene is given in Fig. 1d (Poirel et al., 2005a; Canton & Coque, 2006).

Another, rather specific IS element found in association with blaCTX-M genes is the IS common region (ISCR1), formerly known as orf 513 (Stokes et al., 1993; Partridge & Hall, 2003). The function of this IS element remained unknown until comparative analysis linked these so-called common region elements to a group of IS91-like ISs (Toleman et al., 2006a). The IS91-like ISs are a family of unusual IS elements that differ from most other IS elements, both in the structure and in the mode of transposition. They can perform rolling-circle (RC) transposition, in which a single IS element can mobilize the sequences to which it is attached (Tavakoli et al., 2000; del Pilar Garcillan-Barcia et al., 2001). It has been proposed that ISCR1 may mobilize the nearby sequence and a truncated 3′ conserved sequence (CS) from one integron to the 3′ CS of another integron through RC transposition, thus facilitating the formation of complex class 1 integrons associated with ISCR1 (Toleman et al., 2006a, b), as shown in Fig. 1e (Garcia et al., 2005; Canton & Coque, 2006; Novais et al., 2006; Bae et al., 2007; Riaño et al., 2009).

A third type of genetic elements associated with blaCTX-M genes are the phage-related elements (Fig. 1f) (Oliver et al., 2005). This highlights the fact that phages may serve as tools for natural genetic engineering processes that eventually lead to the evolution and spread of antimicrobial resistance (Muniesa et al., 2004; Oliver et al., 2005; Riaño et al., 2009).

MGEs for other ESBL genes

In contrast to the blaCTX-M genes, the blaVEB-1 gene was identified as a gene cassette in class 1 integrons (Fig. 1g) (Poirel et al., 1999, 2005b; Girlich et al., 2001). Moreover, the blaGES-1 gene was also found to be a gene cassette in a class 1 integron structure (Fig. 1h) (Poirel et al., 2000). Another study described the location of the blaGES-1 gene cassette on a class 3 integron (Fig. 1i) (Correia et al., 2003). These findings underlined the fact that integron-located ESBL genes are not only part of the Ambler class D (OXA-10) (Girlich et al., 2001) or class B (IMP-1, active on imipenem) (Miriagou et al., 2005) but may also be part of class A. This is of interest because most of the plasmid-mediated ESBL that are spreading worldwide are of class A and their integron location may confer them additional potential for spreading.

MGEs described as carriers of ampCβ-lactamase genes

Many genetic elements have been found to be involved in the mobilization of ampC genes onto plasmids within a bacterium (Jacoby & Tran, 1999; Queenan et al., 2001; Winokur et al., 2001; Raskine et al., 2002; D'Andrea et al., 2006), the most important of which are discussed below.

ISEcp1 has been identified in association with many blaCMY genes, including the blaCMY-2 gene (Fig. 2a) (Poirel et al., 1999, 2000; Wu et al., 1999; Winokur et al., 2001; Hossain et al., 2004; Literacka et al., 2004; D'Andrea et al., 2006; Hopkins et al., 2006b; Toleman et al., 2006b; Nakano et al., 2007) and the blaACC-1 and blaACC-4 genes (Papagiannitsis et al., 2007; Partridge et al., 2007). ISEcp1 plays a dual role. It is not only responsible for the mobility of a transposition unit including itself and a resistance determinant, as was also described for blaCTX-M genes, but it can also supply an efficient promoter for the high-level expression of neighbouring genes (Hossain et al., 2004).

Figure 2

Modular schematic structure of the backbones of MGEs containing plasmidic ampC genes. (a) An ISEcp1-like element in association with the blaCMY-2/5 gene. The blaCMY-2 gene is the most common ampCβ-lactamase gene among Enterobacteriaceae worldwide, being related to the ampC genes of Citrobacter freundii (Winokur et al., 2001; Giles et al., 2004). The blc and sugE genes showed 96% sequence identity with the two genes just found downstream of the C. freundii chromosomal ampC gene (Giles et al., 2004). (b) blaDHA-1, blaCMY-10 and blaCMY-11 in association with ISCR1 in a complex class 1 integron (59-be: black ovals and attI recombination site: black circle) with aadA2 as the gene cassette (Verdet et al., 2000; Lee et al., 2004). The blaCMY-10 and blaCMY-11 genes, related to chromosomal ampC genes of Aeromonas spp., are evolved from the blaCMY-1 gene. A sequence identical to the ISCR1 was found upstream from the blaCMY-10 and blaCMY-11 genes (Lee et al., 2004; Jacoby et al., 2009). The genetic organization of the gene coding for DHA-1 was mobilized from the Morganella morganii chromosome and inserted into a sulI-type integron (Verdet et al., 2000). (c) Schematic representation of the 4252-bp C. freundii-derived sequence containing the blaCMY-13 gene (Miriagou et al., 2004). The blaCMY-13-ampR region was bound by two directly repeated IS26 elements.

Other ampC genes, such as blaCMY-1, blaCMY-8, blaCMY-9, blaCMY-10, blaCMY-11, blaCMY-19blaMOX-1 and blaDHA-1, have been found downstream of an ISCR1 involved in gene mobilization in complex class 1 integrons (Toleman et al., 2006b). This MGE has also been associated with blaCTX-M genes, as mentioned above. The schematic representation of blaDHA-1, blaCMY-10 and blaCMY-11 genes in association with ISCR1 in a complex class 1 integron is given in Fig. 2b.

A Citrobacter freundii-derived sequence of 4252 bp, which included a blaCMY-13 gene and was bound by two directly repeated IS26 elements, was found on an IncN plasmid from E. coli (Fig. 2c), indicating that this gene may have spread, from the chromosome to a plasmid within a bacterium, following IS26-dependent mobilization as was also described for blaSHV genes (Miriagou et al., 2004).

Impact of MGEs as carriers of broad-spectrum β-lactamases on epidemiology and coresistance selection

Genes encoding broad-spectrum β-lactamases can be carried by a variety of MGEs such as ISs, transposons and plasmids. These MGEs may have different potentials in the dissemination of resistances, such as ISEcp1, which seems to mobilize and transport genes onto plasmids quite easy and efficiently (Bae et al., 2006a, b; Eeckert et al., 2006; Fernandez et al., 2007; Liu et al., 2007; Navon-Venezia et al., 2008; Shen et al., 2008).

The success of the spread of specific β-lactamases will thus be determined mainly by the MGEs in terms of the selection and dispersion of these enzymes.

The most striking examples of the impact of MGEs on the epidemiology and coresistance selection are discussed below.

MGE as a carrier for the blaTEM-52 gene

The blaTEM-52 gene is located on a Tn3-like transposon that is found on a large stable conjugative IncI1 plasmid. This plasmid is found mainly in Enterobacteriaceae isolated from poultry and humans (Hasman et al., 2005; Cloeckaert et al., 2007; Smet et al., 2008, 2009). Surprisingly, until now, no other resistance genes have been associated with this plasmid. This is most probably due to its broad host spectrum and the possibility of its spreading internationally, although it has spread mainly in Europe (Weill et al., 2004a, b; Hasman et al., 2005; Cloeckaert et al., 2007; Smet et al., 2008).

MGE as a carrier for the blaCTX-M-15 gene

Another plasmid-mediated gene is blaCTX-M-15. This gene is associated with ISEcp1 and is mainly described in Enterobacteriaceae of human origin worldwide (Boyd et al., 2004; Hopkins et al., 2006c; Karisik et al., 2006; Lavollay et al., 2006; Nicolas-Chanoine et al., 2008). This gene has been found on a multitude of plasmids of different sizes and with different incompatibility (Inc) groups, such as IncI1 and IncFII. The latter Inc group is frequently coassociated with replication genes originating from IncFIA or IncFIB (Hopkins et al., 2006c; Karisik et al., 2006). This indicates that the association with ISEcp1 is important in the spread of this gene among different plasmids. Besides the horizontal spread of a blaCTX-M-15-carrying plasmid from one bacterium to another, the worldwide clonal spread of certain CTX-M-15-positive E. coli isolates has also been shown to be important in the dissemination of this resistance gene (Lavollay et al., 2006; Nicolas-Chanoine et al., 2008).

MGE as a carrier for the blaCTX-M-2 and blaCTX-M-9 genes

Other blaCTX-M genes, such as the blaCTX-M-2 and blaCTX-M-9 genes, are located on plasmids, similar to blaCTX-M-15, although they are associated with different Inc groups and are associated with ISCR1 as part of a complex class 1 integron, as already mentioned (Garcia et al., 2005; Canton & Coque, 2006; Novais et al., 2006; Bae et al., 2007; Riaño et al., 2009). The blaCTX-M-2 and blaCTX-M-9 genes are mainly found on IncHI2 plasmids and may also be located on IncP, IncA/C or IncFI plasmids. The blaCTX-M-9 gene is also found on the chromosome (Hopkins et al., 2006c; Novais et al., 2006; Fernandez et al., 2007). These genes are found in Enterobacteriaceae from food-producing animals and humans. An important difference between blaCTX-M-9 and blaCTX-M-15 is the coresistance caused by non-β-lactam resistance gene cassettes associated with these class 1 integrons. As such, these blaCTX-M-9 genes may be easily selected using other non-β-lactam antimicrobial agents (Sabaté et al., 2002; Novais et al., 2006; Fernandez et al., 2007; Riaño et al., 2009). Similarly for blaCTX-M-2, however, a variety of non-β-lactam resistance gene cassettes have been found in these complex class 1 integrons, which makes them perfect discrimination tools (Arduino et al., 2002; Power et al., 2005; Valverde et al., 2006).

MGE as a carrier for the blaCMY-2 gene

The blaCMY-2 gene has been found to be located on an IncI1 or an IncA/C plasmid (Hopkins et al., 2006c). These plasmids seem to be very promiscuous, having been found all over the United States, mainly associated with Salmonella (Winokur et al., 2000; Zhao et al., 2008). They are also emerging in other continents, although less in association with Salmonella (Pai et al., 2004; Briñas et al., 2005; Su et al., 2005; Li et al., 2008). This gene is like the blaCTX-M-15 gene associated with an ISEcp1-like element, which makes the spread of this bla gene between plasmids within a bacterium possible. Therefore, the dissemination of this gene could not be explained in terms of the clonal spread of CMY-2-producing bacteria, but rather in terms of different plasmids carrying a blaCMY-2 gene (Winokur et al., 2001; Naseer et al., 2009).

Broad-spectrum β-lactamases among Enterobacteriaceae from animals

Data on the presence of ESBL- and AmpC β-lactamase-producing Enterobacteriaceae isolated from healthy and sick animals are discussed in this section. Per group (food-producing animals, companion animals and wild animals), ESBL-producing bacteria are handled first, followed by the AmpC β-lactamase-producing bacteria.

Food-producing animals

Healthy animals

Between 2002 and 2009, the number of publications reporting commensal broad-spectrum cephalosporin-resistant Enterobacteriaceae isolated from food-producing animals has increased drastically. A summary of the published data of ESBL- or AmpC β-lactamase-producing bacteria isolated from food-producing animals is listed in Supporting Information, Table S1.

The diversity among the ESBL-encoding genes in Enterobacteriaceae from food-producing animals is by far larger than what is seen for the AmpC-encoding genes. So far, the presence of ESBLs among commensal Enterobacteriaceae has been found to range from 0.2% to 40.7% (Table S1). Some ESBLs seem to be confined to specific individual countries, such as TEM-106 in Belgium, CTX-M-8 and SHV-5 in Tunisia and several CTX-M enzymes in China (Duan et al., 2006; Jouini et al., 2007; Smet et al., 2008; Tian et al., 2009a, b). Other ESBLs have been found to be more widely distributed. So far, TEM-52- and SHV-12-producing Enterobacteriaceae, isolated especially from poultry, have only been described on the European continent (Table S1) (Briñas et al., 2003; Hasman et al., 2005; Riaño et al., 2006; Cloeckaert et al., 2007; Chiaretto et al., 2008; Machado et al., 2008; Smet et al., 2008; Costa et al., 2009). ESBLs such as CTX-M-1, CTX-M-2 and CTX-M-14 have been found in many European countries, being associated with E. coli mainly from poultry (Table S1) (Briñas et al., 2003; Shiraki et al., 2004; Kojima et al., 2005; Girlich et al., 2007; Jouini et al., 2007; Machado et al., 2008; Smet et al., 2008; Costa et al., 2009). The CTX-M-15 enzyme, the most widely diffused enzyme among human Enterobacteriaceae, was only recently detected among E. coli from poultry and pigs (Smet et al., 2008; Tian et al., 2009a, b).

The presence of AmpC β-lactamase-mediated resistance in commensal Enterobacteriaceae ranged from 0.01% to 88.5% (Table S1). CMY-2 is the most common enzyme identified among these isolates. On a dairy farm, the overwhelming presence of CMY-2-producing E. coli (88.5% of the isolated strains) could be linked to the use of ceftiofur to treat respiratory infections in calves (Donaldson et al., 2006). There is a striking difference in the presence of CMY-2 between E. coli and Salmonella isolates from poultry, cattle and pigs in Japan and Canada (Table S1). This may indicate that there is somehow a different epidemiology of CMY-2-producing Enterobacteriaceae in those countries among different animal species.

Sick animals

To date, only a few studies have been published reporting ESBL or AmpC-producing Enterobacteriaceae isolated from sick pigs and cattle. The presence of ESBL- or AmpC-producing bacteria among diseased poultry has so far not been described. An overview is given in the second part of Table S1.

Data on the presence of ESBL-producing Enterobacteriaceae among diseased cattle and pigs have so far only been described in a Korean and a French report. TEM and SHV ESBLs have been described in Korea, whereas different members of the CTX-M family are predominantly present in France (Table S1) (Madec et al., 2008; Rayamajhi et al., 2008).

AmpC enzymes have been detected among clinical bovine and porcine Enterobacteriaceae. The prevalence of these AmpC-producing animal pathogens varied from 0.3% to 77%. In most reports, CMY-2 enzymes and mutations in the promoter and attenuator regions of the chromosomal AmpC enzyme were found, but in one report DHA-1 enzymes were also found (Table S1).

Companion animals

Healthy animals

A few studies describing the presence (7–20%) of ESBL- or AmpC β-lactamase-producing E. coli isolated from the faecal microbial community of healthy pets have been published (Table S2).

ESBLs – predominantly CTX-M-1 – have not only found their way into commensal E. coli from dogs in Europe, but have also recently been found among commensal E. coli from cats and dogs in Latin America (Costa et al., 2004; Carattoli et al., 2005; Moreno et al., 2008).

So far, CMY-2-producing E. coli have only been isolated from the faeces of healthy dogs in Italy (Carattoli et al., 2005).

Sick animals

Five recent prevalence studies are available on the presence of broad-spectrum β-lactamases in Enterobacteriaceae from sick companion animals with urinary tract infections. The presence of ESBL- or AmpC β-lactamase-producing E. coli from diseased dogs ranged from 1.4% to 19.4% (Table S2). ESBLs of the CTX-M-1 cluster were the most predominant enzymes found. In four reports, CMY enzymes – mainly CMY-2 – were found, although CMY-7 enzymes were found in one study (Table S2). This is a situation similar to what is found in food-producing animals. However, the limited data available on broad-spectrum β-lactamases in pet animals are not sufficient to provide a good idea as to which broad-spectrum β-lactamases are spreading. Further epidemiological surveillance should be performed in pet animals to estimate the burden of this resistance.

Wild animals

Healthy animals

In Portugal, national surveillance in birds of prey and seagulls shows the emergence of broad-spectrum β-lactamases among Enterobacteriaceae from wild animals (Costa et al., 2006; Poeta et al., 2008). To date, no AmpC β-lactamases have been identified. The occurrence of ESBLs was as high as 19% in E. coli isolates from faecal samples, mainly of birds of prey and seagulls, with a predominance of CTX-M-1, CTX-M-14 and TEM-52 enzymes (Table S3). A recent French study describes a 9.4% prevalence of commensal ESBL-producing E. coli from wild yellow-legged gulls. The most predominant ESBL was CTX-M-1 (Table S3). These enzymes were also found in the faecal flora of food-producing animals and companion animals (Miro et al., 2005; Ho et al., 2008). The animals may have been contaminated by eating the leftovers from human food (Costa et al., 2006; Poeta et al., 2008). More studies should be performed in order to trace the origins of this contamination, including analysis of the prevalence of this type of resistance in different ecosystems.

Broad-spectrum β-lactamases in animal-associated bacteria: significance for human health

Presence of broad-spectrum β-lactamase-producing Enterobacteriaceae from humans

In order to be able to compare the differences and similarities between the situation in animal and human Enterobacteriaceae, the diversity of broad-spectrum β-lactamase-producing commensals and pathogens present in humans will first be discussed in this chapter. The ESBL-producing bacteria will first be discussed, followed by the AmpC β-lactamase-producing bacteria.

Healthy humans

Since 2000, the number of publications reporting the faecal carriage of broad-spectrum β-lactamase-producing commensal Enterobacteriaceae from humans has been increasing (Valverde et al., 2004; Kaneko et al., 2005; Miro et al., 2005; Moubareck et al., 2005; Kader et al., 2007; Pallechi et al., 2007; Ho et al., 2008; Leflon-Guibout et al., 2008; Pongpech et al., 2008; Rodriguez-Baño et al., 2008; Vinué et al., 2009). An overview of commensal broad-spectrum β-lactamase-producing bacteria, mainly E. coli, is given in Table S4. Most available publications describe the faecal carriage of ESBL-producing Enterobacteriaceae with a presence ranging from 0.6% as a lower limit in France (Leflon-Guibout et al., 2008) to the upper limit of 68% in Spain (Rodriguez-Baño et al., 2008) (Table S4). The most predominant family of ESBLs reported among commensal E. coli is the CTX-M family, with the CTX-M-9 cluster being the most common cluster worldwide (Table S4). This may indicate that commensal E. coli of humans may constitute a reservoir of blaCTX-M genes.

So far, information about the presence of AmpC β-lactamase-producing strains from faecal samples of healthy humans remains limited. In Japan, a prevalence of commensal CMY-2-producing E. coli of 3.2% has been reported (Table S4).

Sick humans

Most studies dealing with the presence of broad-spectrum β-lactamases in bacteria from humans refer to pathogens. The occurrence of broad-spectrum β-lactamase-producing pathogens has already been extensively reviewed (Paterson & Bonomo, 2005; Pitout et al., 2005a, b; Arlet et al., 2006; Livermore et al., 2007; Jacoby et al., 2009). Infections caused by Enterobacteriaceae complicate therapy and limit treatment options. Population studies of human pathogens producing ESBLs or AmpC β-lactamases in hospitals and the community worldwide are summarized in Table S5.

ESBL-producing human pathogens have been reported worldwide, with a presence of 0.3–91% in Europe, 0.8–5.6% in North America and 12–31% in Africa and the Middle East. In the Asia and Pacific region, the presence of ESBL-producing pathogens ranges from 11.3% to 38.6% (Table S5).

ESBLs, such as TEM-52 and SHV-12, among human pathogens were first reported in Europe (Knothe et al., 1983; Arlet et al., 1994; Babini & Livermore, 2000; De Gheldre et al., 2001; Vahaboglu et al., 2001; Saladin et al., 2002; Arpin et al., 2003; Weill et al., 2004a, b; Politi et al., 2005; Livermore et al., 2007). Later on, TEM-type ESBLs were described in the United States (Jacoby et al., 1988; Rice et al., 1996). Since then, CTX-M ESBLs have become dominant, with a much greater penetration into E. coli strains worldwide.

There are considerable geographical differences in the occurrence of ESBLs, especially of the CTX-M enzymes. Many different CTX-M enzymes are widely distributed, mostly among E. coli, causing urinary tract infections (Brenwald et al., 2003; Paterson et al., 2005; Pitout et al., 2005a, b; Doi, 2007; Livermore et al., 2007). Some CTX-M enzymes seem to be dominant in specific European countries, such as CTX-M-14 and CTX-M-9 in Spain and Portugal (Garcia et al., 2005; Romero et al., 2005; Mendonça et al., 2007; Rodriguez-Baño et al., 2009), CTX-M-1 and CTX-M-15 in Italy and France (Brigante et al., 2005; Livermore & Hawkey, 2005; Carattoli et al., 2008) and CTX-M-15 in the United Kingdom (Livermore et al., 2007). In the United States, the most common CTX-M-type ESBL is CTX-M-15, followed by CTX-M-16, CTX-M-8 and CTX-M-14 (Lewis et al., 2007; Hanson et al., 2008; Sjölund et al., 2008). Different CTX-M enzymes, with the predominance of CTX-M-15 and CTX-M-14, have also been reported among nosocomial and community-acquired E. coli isolates causing urinary tract infections in the African and Asian continents (Cao et al., 2002; Chmelnitsky et al., 2005; Gangoué-Piéboji et al., 2005; Ho et al., 2005; Soge et al., 2006; Bae et al., 2007; Song et al., 2009).

AmpC β-lactamases have been found less frequently than ESBLs among Enterobacteriaceae. CMY-2 is the AmpC enzyme with the broadest geographic spread, being an important cause of β-lactam resistance in nontyphoid Salmonella strains in many countries. This type of resistance is increasing worldwide (Dunne et al., 2000; Winokur et al., 2000; Kruger et al., 2004; Batchelor et al., 2005a; Li et al., 2005; Su et al., 2005; Whichard et al., 2005).

The AmpC phenotype in E. coli is more often due to the increased production of the chromosomal AmpC β-lactamase, as reported in a few prevalence studies (Bergström & Normark, 1979; Mulvey et al., 2005; Potz et al., 2006; da Silva Dias et al., 2008; Mammeri et al., 2008) (Table S5). However, CMY-producing E. coli strains, mainly CMY-2, have also been demonstrated (Table S5) (Gazouli et al., 1998; Winokur et al., 2001; Moland et al., 2002; Mulvey et al., 2005; Hopkins et al., 2006a, b; Pitout et al., 2007; Adler et al., 2008; Li et al., 2008; Pai et al., 2004).

It seems that broad-spectrum β-lactamases have been evolving and spreading at a rapid rate among humans worldwide over the last 20 years.

Differences and similarities between the situation in animal and human Enterobacteriaceae

β-Lactamases were first detected in the early 1980s in humans, and their presence and diversity have been increasing ever since. The first time cephalosporin resistance was noted in animals was in early 2000. Compared with what is known in humans, the knowledge of the epidemiology of broad-spectrum β-lactamase-producing bacteria in animals is rather limited. As the spread of these β-lactamases in animals started to increase only recently, it is possible that these genes may be of human origin (Hernadez et al., 2005). However, β-lactamases in humans can also be of animal origin, as has been shown for the zoonotic Salmonella Infantis and Virchow isolates (Bertrand et al., 2006; Cloeckaert et al., 2007), in which cases the infecting cephalosporin-resistant bacterium was directly derived from the animal.

The diversity of broad-spectrum β-lactamases in human Enterobacteriaceae is much higher than the situation in animal bacteria (Fig. 3).

Figure 3

Schematic representation of the number of studies reporting the presence of the most predominant ESBLs among Enterobacteriaceae isolated from healthy (a) and sick animals (b) and healthy (c) and sick humans (d) (modified from Tables S1, S2, S3, S4 and S5).

In animals, there is a predominance of TEM-52, CTX-M-1, CTX-M-14 and CMY-2-producing Enterobacteriaceae, with the predominance of CMY-2 in North-America, and of CTX-M-1, CTX-M-14 and TEM-52 enzymes in Europe (Winokur et al., 2000; Allen & Poppe, 2002; Costa et al., 2004, 2006, 2009; Carattoli et al., 2005; Hasman et al., 2005; Donaldson et al., 2006; Cloeckaert et al., 2007; Machado et al., 2008; Smet et al., 2008; Zhao et al., 2008).

These enzymes, together with CTX-M-9 and CTX-M-15, are also predominantly present in human bacteria. Some enzymes in human bacteria are even limited to specific countries such as CTX-M-39 in Israël, CTX-M-13 in China, CTX-M-40 in Thailand and TEM-63 and TEM-131 in South Africa (Kruger et al., 2004; Chmelnitsky et al., 2005; Ho et al., 2005; Kiratisin et al., 2008). These enzymes have not been detected in animal Enterobacteriaceae. However, it must be said that the presence of these broad-spectrum β-lactamases in animal Enterobacteriaceae has not yet been investigated in these countries.

The most prevalent enzymes in commensal and pathogenic E. coli from both humans and animals are CTX-M-9, SHV-12 and CTX-M-14 in Spain, CTX-M-14 and CTX-M-32 in Portugal, CTX-M-1 in France and Italy, CTX-M-2 in Japan and, finally, CMY-2 in Canada and the United States. This may indicate that there is somehow a similar epidemiology among animal and human bacteria. Comparison of the genetic relatedness of Enterobacteriaceae recovered from different countries and origins and harbouring the same ESBL or AmpC enzyme may help to explain this hypothesis.

Impact on public health of animal-derived Enterobacteriaceae producing broad-spectrum β-lactamases

The presence of broad-spectrum β-lactamase-producing bacteria in animals is increasing, and it is not unrealistic to expect that this will have an impact on human health. Resistance may be transferred in two ways. Because of close contact or consumption of animal meat, a β-lactam-resistant zoonotic strain, in most cases Salmonella spp., may be transferred directly from animals to humans, thus possibly causing infection, as has been demonstrated in a number of reports (Fey et al., 2000; Espié et al., 2005; Bertrand et al., 2006; Cloeckaert et al., 2007; Riaño et al., 2009). As for direct transfer of resistance, the use of antimicrobial agents, selecting resistant bacteria, may be the most important factor. However, the exact role of different antimicrobials in resistance development or dissemination remains unknown, although it could possibly be assessed by means of the pharmacokinetics and pharmacodynamics of these antimicrobials. In vivo studies may also likely provide insights into the role of antimicrobial agents.

Moreover, resistance may possibly be acquired indirectly, through the transfer of resistance genes from bacteria of animal origin to bacteria infecting humans. Studies pointing out the possibilities of indirect transfer of resistance genes remain limited.

Cloeckaert et al. (2007) emphasized that TEM-52-producing Salmonella sp. are not only spreading between poultry and humans through direct transfer, but that the stable plasmid carrying this gene (Smet et al., 2009) may also be spreading between different Salmonella serotypes, thus indicating a possibility for indirect resistance transfer.

Another example of indirect transfer of resistance is the dispersion of CMY-2-producing E. coli from cattle and pigs to humans, or vice versa, due to the association of this gene with ISEcp1 (Winokur et al., 2001; Hopkins et al., 2006a, b, c; Naseer et al., 2009). Again, this highlights the importance of MGEs in the spread of resistance genes.

Therefore, both the selective effect of the antibiotics and the MGEs carrying these bla genes could be important factors in indirect resistance transfer. Little is known about the influence of these MGEs on the spread of the bla genes. This lack of knowledge could make it difficult to predict the possibilities of spread, thus underlining the need for further investigations.

Concluding remarks

The widespread use of extended-spectrum cephalosporins creates a reservoir of resistant bacteria and resistance genes that may add to the burden of β-lactam resistance in human medicine and reduce the time that these valuable antimicrobial compounds will be available for the effective treatment of infections. Moreover, multiresistance frequently associated with strains carrying ESBLs and AmpCs is worrisome because these strains can be selected by a manifold of different antimicrobial agents, which could drastically reduce the treatment options.

Resistance against β-lactams is increasingly being reported and is on the rise in Enterobacteriaceae from both humans and animals. This coincides with the increasing numbers of β-lactamase variants. It is also interesting to note that there is no specific β-lactamase associated with animals because most enzymes are also predominantly present in bacteria found in humans.

Some ESBLs and AmpC β-lactamases seem to be dedicated to a specific geographical region, while others are more widely spread. This appears to be without obvious reason. However, the world wide detection of certain broad-spectrum β-lactamases does not seem to be linked to the expansion of bacterial clones, but rather to plasmid-mediated horizontal gene transmission. Epidemiological studies dealing with the dissemination of β-lactam resistance may help to explain these findings.

The understanding of why some MGEs seem to be very successful in spreading is of considerable importance. Moreover, these MGEs frequently carry other resistance genes that are cotransferred and cause coresistance selection.

The clinical and commercial pressure to use β-lactams as well as the global mobility of humans, animals and food products guarantee that the spread of β-lactamase genes will continue. β-Lactam antibiotics may enter the environment, such as water sources, having been excreted in the faeces and/or the urine of treated animals. Water may therefore also be a potential source of selective pressure.

More studies are needed to perform a more accurate risk assessment concerning the spread of antimicrobial resistance, as well as on the mechanisms of linkage and transferability of β-lactam resistance determinants in natural environments. Therefore, the evaluation of the possible impact of this resistance in animals for human health studies should not be limited to pathogenic bacteria, but must also include commensals, because they may be a major reservoir of resistance genes, as has already been shown to be the case in poultry (Smet et al., 2008).


F.H. and P.B. shared senior authorship.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Table S1. Presence of Enterobacteriaceae producing broad-spectrum β-lactamases in food-producing animals (per animal group, ESBL-producing Enterobacteriaceae are listed first followed by the AmpC β-lactamase-producing Enterobacteriaceae).

Table S2. Presence of Enterobacteriaceae producing broad-spectrum β-lactamases in companion animals (per animal group, ESBL-producing Enterobacteriaceae are listed first followed by the AmpC β-lactamase-producing Enterobacteriaceae).

Table S3. Presence of Escherichia coli producing broad-spectrum β-lactamases in wild animals in Portugal and France (Costa et al., 2006; Poeta et al., 2008; Bonnedahl et al., 2009).

Table S4. Occurrence of faecal carriage of broad-spectrum β-lactamase-producing Enterobacteriaceae in healthy humans worldwide.

Table S5. Population studies of broad-spectrum β-lactamase-producing human pathogens worldwide.


  • Editor: Eva Top


View Abstract