Antibiotic resistance in E. coli

E. coli is a commensal bacteria in the digestive tract of animals and humans (Milanov et al., Reference MILANOV, PRUNIĆ, VELHNER, TODOROVIĆ and POLAČEK2015) which has developed resistance to antibiotics used in both human and in veterinary medicine. The antibiotic resistance in E. coli represents a severe clinical threat to humans. It is well documented that E. coli as commensal bacteria can be reservoirs of resistance genes for potentially pathogenic microorganisms (Wray and Gnanou, Reference WRAY and GNANOU2000; Velhner et al., Reference VELHNER, PETROVIĆ, STOJANOV, RATAJAC and STOJANOVIĆ2010). It can be transferred between humans and animals (Velhner and Milanov, Reference VELHNER and MILANOV2015) and can be spread through water, food or soil. Sørum and Sunde (Reference SØRUM and SUNDE2001) suggested that antibiotic resistance in isolates of E. coli from faecal samples of animals may be anthropogenic in origin. Moreover, E. coli is an important faecal indicator bacteria and it is commonly used in order to determine contaminated water. The isolates of the E. coli are often examined to detect antibiotic resistance in populations of wild animals, including gulls. Antibiotic resistant E. coli have been described in livestock (Guerra et al., Reference GUERRA, JUNKER, SCHROETER, MALORNY, LEHMANN and HELMUTH2003; Lanz et al., Reference LANZ, KUHNERT and BOERLIN2003; Milanov et al., Reference MILANOV, PRUNIĆ, VELHNER, TODOROVIĆ and POLAČEK2015), chickens and in humans (Kang et al., Reference KANG, JEONG, OH, TAE, CHOI, MOON, LEE, LEE, SEOL, CHO and LEE2005) and can spread to the environment, infecting gulls which becoming new host reservoir of resistant E. coli in the environment. Recently, the emergence and dissemination of E. coli strains resistant to broad spectrum cephalosporins as well as E. coli strains resistant to fluoroquinolones are of great public health concerns and represent a serious problem for the therapy of human infections (Paterson and Bonomo, Reference PATERSON and BONOMO2005; Strahilevitz et al., Reference STRAHILEVITZ, JACOBY, HOOPER and ROBICSEK2009). Furthermore, these strains are disseminated into the environments and can colonize wild animals, particularly omnivorous gulls (Gionechetti et al., Reference GIONECHETTI, ZUCCA, GOMBAC, MONTI-BRAGADIN, LAGATOLLA, TONIN, EDALUCCI, VITALI and DOLZANI2008; Poeta et al., Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008; Bonnedahl et al., Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009; Dolejska et al., Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009; Literak et al., Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010).

The epidemiological significance of gulls in the spreading of antibiotic resistance

The number of gulls have increased in many countries because they live in proximity to humans. They eat fish, bivalves, eggs and other food which they can find near the coastline, but they also eat leftover food from garbage (Poeta et al., Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008). Moreover, they can fly long distance so they can disseminate resistant strains in different regions (Gionechetti et al., Reference GIONECHETTI, ZUCCA, GOMBAC, MONTI-BRAGADIN, LAGATOLLA, TONIN, EDALUCCI, VITALI and DOLZANI2008). Dolejska et al. (Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009) stated that gulls feeding on garbage dumps and in urbanised areas could be important reservoirs and vectors of antibiotic resistant strains of E. coli in the environment. Gulls often feed at garbage dumps, sewage plants or in agricultural areas in vicinity of poultry and livestock. Consequently, the antibiotic resistant E. coli can be high in gull populations associated with humans (Dolejska et al., Reference DOLEJSKA, CIZEK and LITERAK2007; Bonnedahl et al., Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009; Simões et al., Reference SIMÕES, POIREL, DA COSTA and NORDMANN2010). Furthermore, gulls have a great potential and documented significance as a source of faecal contamination of water and swimming areas. Their faeces is considered a serious threat to human health (Lévesque et al., Reference LÉVESQUE, BROUSSEAU, BERNIER, DEWAILLY and JOLY2000). Gulls do not normally come into contact with antibiotics, but they can be colonised by resistant strains of E. coli and can be an important reservoir of resistance genes and the occurrence of the resistance genes in E. coli in gulls indicate that (Literak et al., Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010).

The shedding time for E. coli in gulls is still unknown and represents an important question. Le Devendec et al. (Reference LE DEVENDEC, BOUDER, DHEILLY, HELLARD and KEMPF2011) observed that chickens can carry resistant E. coli for several weeks. Furthermore, Hernandez et al. (Reference HERNANDEZ, BONNEDAHL, ELIASSON, WALLENSTEN, COMSTEDT, JOHANSSON, GRANHOLM, MELHUS, OLSEN and DROBNI2010) suggested that gulls bring antibiotic resistant bacteria to remote areas, indicating longer shedding time.

All the above mentioned facts suggested that the gulls have many characteristics which make them suitable for resistance dissemination studies (Dolejska et al., Reference DOLEJSKA, CIZEK and LITERAK2007; Bonnedahl et al., Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009; Bonnedahl et al., Reference BONNEDAHL, DROBNI, JOHANSSON, HERNANDEZ, MELHUS, STEDT, OLSEN and DROBNI2010; Simões et al., Reference SIMÕES, POIREL, DA COSTA and NORDMANN2010), making them a potential threat to free range poultry.

The relationship between the level of contact of gulls with humans and the level of antibiotic resistance in E. coli

Gulls are common in environments in human vicinity and they often feed on human and animal waste. Gulls can be a reservoir of antibiotic resistant E. coli because they can travel long distances and inhabit a wide variety of environment (Poeta et al., Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008; Literak et al., Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) so they can spread resistance genes. A higher occurrence of antibiotic resistant strains of E. coli in gulls which live near humans than those which live in areas without humans indicates a transfer of E. coli from humans to gulls (Gordon and Cowling, Reference GORDON and COWLING2003). Österblad et al. (Reference ÖSTERBLAD, NORRDAHL, KORPIMÄKI and HUOVINEN2001) reported that birds which lived isolated from humans possessed lower levels of resistant bacteria, but now there is evidence for the presence of antibiotic resistant E. coli in wild gulls in isolated areas. Furthermore, extended-spectrum-beta-lactamases (ESBL)-producing E. coli harbouring blaCTX-M-15 were found in gulls sampled at the Commander Islands in Russia (Hernandez et al., Reference HERNANDEZ, BONNEDAHL, ELIASSON, WALLENSTEN, COMSTEDT, JOHANSSON, GRANHOLM, MELHUS, OLSEN and DROBNI2010) and 37% ESBL harbouring isolates in gulls lived in remote areas in Barrow, Alaska (Bonnedahl et al., Reference BONNEDAHL, HERNANDEZ, STEDT, WALDENSTRÖM, OLSEN and DROBNI2014). Moreover, Hernandez et al. (Reference HERNANDEZ, BONNEDAHL, ELIASSON, WALLENSTEN, COMSTEDT, JOHANSSON, GRANHOLM, MELHUS, OLSEN and DROBNI2010) found E. coli belonging to the globally disseminated human pathogen O25b- ST131 (CTX-M-15) clone in gulls into an environment lacking antibiotic pressure. That could indicate that the antibiotic resistance could be disseminated worldwide. They explained their findings by the fact that the gulls probably acquired antibiotic resistant genes from human environments or from other birds.

Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007) suggested that young Black-headed Gulls are an important host reservoir of resistant E. coli strains and that this reflects the presence of resistant strains in their sources of food and water. They noted that gulls which nested near waste or agricultural water tend to harbour more antibiotic resistant E. coli than those which lived near unpolluted water. Gulls excrete faecal coliforms which can affect drinking and recreational water (Lévesque et al., Reference LÉVESQUE, BROUSSEAU, BERNIER, DEWAILLY and JOLY2000; Fogarty et al., Reference FOGARTY, HAACK, WOLCOTT and WHITMAN2003). Edge and Hill (Reference EDGE and HILL2005) examined the occurrence of the antibiotic resistant strains of E. coli from wastewater treatment plants and bird faeces in the inland Hamilton Harbour on Lake Ontario Canada and they suggested that faeces of wild ducks, geese and gulls could be a more important contributor of resistant strains of E. coli found on a beach near Hamilton than municipal wastewater sources. Besides that, Skurnik et al. (Reference SKURNIK, RUIMY, ANDREMONT, AMORIN, ROUQUET, PICARD and DENAMUR2006) and Ljubojević et al. (Reference LJUBOJEVIĆ, PUVAČA, PELIĆ, TODOROVIĆ, PAJIĆ, MILANOV and VELHNER2016) suggested that the level of antibiotic resistance in faecal bacteria from animals can be in correlation with the level of contact of these animals with humans.

The existence of direct exchange of resistant determinants between animals and human pathogens is not completely clarified as well as the extent to which the use of antibiotics in veterinary practice participate in selective pressure which could be the important contributor in spreading of antibiotic resistance in human isolates (Carattoli, Reference CARATTOLI2008). The results obtained by Moulin-Schouleur et al. (Reference MOULIN-SCHOULEUR, RÉPÉRANT, LAURENT, BRÉE, MIGNON-GRASTEAU, GERMON, RASSCHAERT and SCHOULER2007) and Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) support the theory that E. coli with or without antibiotic resistance are transmitted between humans and birds. Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) stated that the high resistance level in gulls in the south of France was a result of long exposure to high antibiotic pressure due to human activities. Radhouani et al. (Reference RADHOUANI, POETA, IGREJAS, GONÇALVES, VINUE and TORRES2009) explained the high level of antibiotic resistance in the faecal isolates of E. coli from gulls from the fact that they usually fly to the coast where they can eat the remains of human food. Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) stated that high level of the usage of antibiotics in human and veterinary medicine in the Mediterranean region influenced comparable resistance levels in gulls with different feeding habits. The resistance to tetracycline, streptomycin and ampicillin is common in farm and wild animals which live in human vicinity (Sørum and Sunde, Reference SØRUM and SUNDE2001; Costa et al., Reference COSTA, POETA, SÁENZ, VINUÉ, ROJO-BEZARES, JOUINI, ZARAZAGA, RODRIGUES and TORRES2006). Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007, Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009), Radhouani et al. (Reference RADHOUANI, POETA, IGREJAS, GONÇALVES, VINUE and TORRES2009) and Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) found the similar resistance patterns in gulls.

Although gull populations are not directly influenced by antibiotic therapy, Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) found the antibiotic resistant E. coli strains, including the ESBL-producing strains and isolates with the qnrS genes, in the faeces of Herring gulls on the coast of the Baltic Sea in Poland. E. coli resistant strains isolated from faeces have been found at various levels in gull populations. The gulls which live in vicinity to people and high density of livestock, including chickens, were colonised with antibiotic-resistant strains of E. coli (Dolejska et al., Reference DOLEJSKA, CIZEK and LITERAK2007; Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009; Radhouani et al., Reference RADHOUANI, POETA, IGREJAS, GONÇALVES, VINUE and TORRES2009; Bonnedahl et al., Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009; Literak et al., Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010). Results obtained by Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) suggested that dense gull populations could be a hot spot for creating of new E. coli variants carrying genes of antibiotic resistance and also they could serve as effective samplers of genetically diverse E. coli from various sources. Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, WALDENSTRÖM, MCMAHON, TOLF, OLSEN and DROBNI2015) noted the widespread occurrence of resistant bacteria in gulls across Europe. Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) found that faecal samples from gull populations contained a high prevalence of antibiotic resistant E. coli and that ESBL producing isolates of E. coli showed high genetic diversity.

Various strains, including E. coli O157, can colonise gulls (Makino et al., Reference MAKINO, KOBORI, ASAKURA, WATARAI, SHIRAHATA, IKEDA, TAKESHI and TSUKAMOTO2000; Nielsen et al., Reference NIELSEN, SKOV, MADSEN, LODAL, JESPERSEN and BAGGESEN2004). Makino et al. (Reference MAKINO, KOBORI, ASAKURA, WATARAI, SHIRAHATA, IKEDA, TAKESHI and TSUKAMOTO2000) suggested that seagulls from Japan may be important carriers and distributers of Shiga toxin producing E. coli (STEC).

The prevalence of antibiotic resistant E. coli in gulls

Antibiotic-resistant E. coli have been reported in Black-headed gulls (L. ridibundus) in the Czech Republic (Dolejska et al., Reference DOLEJSKA, CIZEK and LITERAK2007; Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009), in seagulls of different species in Portugal (Poeta et al., Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008), Mediterranean herring gulls (Larus cachinans) in Italy (Gionechetti et al., Reference GIONECHETTI, ZUCCA, GOMBAC, MONTI-BRAGADIN, LAGATOLLA, TONIN, EDALUCCI, VITALI and DOLZANI2008), Yellow-legged gulls (Larus michahellis) in France (Bonnedahl et al., Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) and Black-headed gull (Larus ridibundus) in Sweden (Bonnedahl et al., Reference BONNEDAHL, DROBNI, JOHANSSON, HERNANDEZ, MELHUS, STEDT, OLSEN and DROBNI2010).

Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007) studied the occurrence of antibiotic resistance in E. coli from young Black-headed Gulls (Larus ridibundus) in three nesting colonies and reported that 75 (29.9%) of 257 isolates were resistant to one or more antibiotics. They found that 42 of 75 isolates were resistant to only one antibiotic, while 33 isolates were resistant to more than one, and the dominant type of resistance was to tetracycline, detected in 49 (19.1%) isolates. They detected resistance to ampicillin in 30 (11.7%), cephalothin in 11 (4.3%), streptomycin in 24 (9.3%), sulphonamides in 20 (7.8%) and chloramphenicol in 5 (1.9%) isolates. None of the isolates from the young Black-headed Gulls was resistant to the broad-spectrum cephalosporin antibiotic ceftazidime.

Poeta et al. (Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008) and Simões et al. (Reference SIMÕES, POIREL, DA COSTA and NORDMANN2010) reported that 32% and 19.3% of gulls from Portugal carried ESBL-producing E. coli. Poeta et al. (Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008) tested fifty-seven fresh faecal samples obtained in different areas of the Berlengas Island for the presence of ESBL-producing E. coli isolates. They detected E. coli isolates in 11 (19.3%) of the samples, all of which were intermediate or resistant to cefotaxime and/or ceftazidime. One of the 11 E. coli isolates was resistant to amoxicillin plus clavulanic acid. Besides that, Poeta et al. (Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008) reported that all 11 E. coli isolates were resistant to tetracycline, 10 isolates were resistant to nalidixic acid, seven isolates were resistant to ciprofloxacin, seven isolates were resistant to streptomycin, five isolates were resistant to trimethoprim-sulphamethosazol and four isolates were resistant to chloramphenicol.

Dolejska et al. (Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009) examined young Black-headed gulls (Larus ridibundus) for the presence of antibiotic-resistant E. coli and found 213 (99%) of gull samples were positive, with antibiotic resistance in 28%. Twelve (6%) were resistant to only one antibiotic and 48 to more than one antibiotic. They reported that resistance to tetracycline was predominant, followed by resistance to ampicillin, streptomycin, sulphonamides and trimethoprim-sulphamethosazol. Moreover, they found E. coli isolates producing ESBL in seven (3%) samples from gulls.

Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) examined the presence of antibiotic resistance in Yellow-legged gulls (Larus michahellis) in two populations living around a dense human population in Camargue, southern France, but differing in the availability of associated food sources. One population fed mainly at Marseille city dump and the other near the coast. They noted that 47.1% (72/153 isolates) of the E. coli isolates were resistant to at least one antibiotic. Furthermore, there was no difference in resistance rates between two examined populations of gulls. In about one third of the isolates was detected resistance to two or more antibiotics whereby resistance to tetracycline, ampicillin and streptomycin was the most frequently detected. They further reported that 9.4% of the sampled Yellow-legged gulls in France harboured ESBL-producing E. coli and found resistance to cefadroxil which is indication of potential ESBL carriage in four strains. Furthermore, Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) detected high levels of ESBL-producing bacteria in gulls showing genotypes highly related with human isolates. Bonnedahl et al. (Reference BONNEDAHL, DROBNI, JOHANSSON, HERNANDEZ, MELHUS, STEDT, OLSEN and DROBNI2010) found antibiotic resistance in 13% E. coli isolates from Black-headed gulls (Larus ridibundus) sampled in Sweden.

Radhouani et al. (Reference RADHOUANI, POETA, IGREJAS, GONÇALVES, VINUE and TORRES2009) collected 53 faecal samples from Yellow-legged gulls (Larus cachinnans) at the Berlengas nature reserve in Portugal. The percentage of resistant isolates were 43.7% for ampicillin, 39.6% for tetracycline, 34% for nalidixic acid, 32.1% for streptomycin, 26.4% for trimethoprim-sulphamethosazol, 18.9% for ciprofloxacin, 18.9% for chloramphenicol, 7.5% for gentamicin, 7.5% for tobramycin, 5.7% for amikacin and 1.9% for amoxicillin-clavulanic acid. Furthermore, all examined isolates were susceptible to cefoxitin, ceftazidim, cefotaxime, aztreonam and imipenem.

Hernandez et al. (Reference HERNANDEZ, BONNEDAHL, ELIASSON, WALLENSTEN, COMSTEDT, JOHANSSON, GRANHOLM, MELHUS, OLSEN and DROBNI2010) investigated antibiotic resistance levels in gulls in the Commander Islands and Kamchatka, Russia. They found resistance in four isolates from Black-headed gulls in the Petropavlovsk river delta, and two from the more remote Commander Islands. Also, they detected four ESBL-producing E. coli and ESBL-producing E. coli from gulls shared phylogenetic diversity with human E. coli.

Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, MCMAHON, HASAN, OLSEN, DROBNI and WALDENSTRÖM2014) collected 3158 faecal samples from the closely related Herring gull (Larus argentatus) and Lesser Black-backed gull (Larus fuscus) from Northern and Western Europe, and the Yellow-legged gull (Larus michahellis) from Southern and Southwestern Europe. Samples were obtained from nine European countries and the authors tested 2210 randomly isolated E. coli for resistance against 10 antibiotics. They reported that 31.5% of the 2210 tested E. coli isolates were resistant to at least one antibiotic while 19% of the isolates were resistant to two or more antibiotics. They noted considerable variation between countries and found highest level of resistance in Spain (61.2%), England (44.1%) and Latvia (38.7%), while the lowest levels were observed in Denmark (8.3%) and Ireland (11.6%). They observed that tetracycline resistance was the most common (19%), followed by ampicillin resistance (18.1%) and that none of the isolates were resistant to tigecycline. Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007, Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009) and Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009, Reference BONNEDAHL, DROBNI, JOHANSSON, HERNANDEZ, MELHUS, STEDT, OLSEN and DROBNI2010) reported most frequent resistance to tetracycline and ampicillin. That could be a consequence of the fact that these antibiotics have been commonly used in both human and animals. It is interesting that Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, MCMAHON, HASAN, OLSEN, DROBNI and WALDENSTRÖM2014) found mecillinam resistance in isolates, which is still rare in clinical studies (Naber et al., Reference NABER, SCHITO, BOTTO, PALOU and MAZZEI2008), of E. coli from gulls in six countries at low levels. Furthermore, Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, MCMAHON, HASAN, OLSEN, DROBNI and WALDENSTRÖM2014) detected high rates of resistance to nalidixic acid in samples from Spanish gulls (30.8-36.9%). Bonnedahl et al. (Reference BONNEDAHL, DROBNI, JOHANSSON, HERNANDEZ, MELHUS, STEDT, OLSEN and DROBNI2010) found the nalidixic acid resistance in E. coli isolated from Yellow legged gulls in Sweden. Moreover, isolates resistant to fluoroquinolones have emerged recently in different gull species in Italy, Portugal, Greenland and in Czech Republic (Gionechetti et al., Reference GIONECHETTI, ZUCCA, GOMBAC, MONTI-BRAGADIN, LAGATOLLA, TONIN, EDALUCCI, VITALI and DOLZANI2008; Sjölund et al., Reference SJÖLUND, BONNEDAHL, HERNANDEZ, BENGTSSON, CEDERBRANT, PINHASSI, KAHLMETER and OLSEN2008; Poeta et al., Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008; Dolejska et al., Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009; Camarda et al., Reference CAMARDA, CIRCELLA, PENNELLI, MADIO, BRUNI, LAGRASTA, MARZANO, MALLIA and CAMPAGNARI2010).

Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) obtained cloacal smears from 27 Herring gulls (Larus argentatus) which they caught during winter 2008/2009 (adults) and during nesting period (chicks) in 2009 and found that the prevalence antibiotic-resistant E. coli was 11% (3/27). One isolate was resistant to ciprofloxacin.

Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, WALDENSTRÖM, MCMAHON, TOLF, OLSEN and DROBNI2015) used three gull species Herring gulls (Larus argentatus) and Lesser-Black backed gulls (Larus fuscus) in northern Europe, and Yellow-legged gulls (Larus michahellis) in southern Europe and collected more than 3000 faecal samples during breeding in nine European countries. They detected ESBL-producing bacteria in 906 of 3158 samples (28.7%) and noted significant variation in prevalence between countries. They found low ESBL prevalence in Denmark (0%), Poland (0.7%) and Ireland (4.5%), intermediate in Portugal (12.7%), Latvia (17.4%) and Sweden (20.7%), and the highest in England (27.1%), The Netherlands (37.8%), and Spain (74.8%). They compared the results with human clinical and veterinary data and found that a high level of acquired ESBL resistance exists. The results obtained by Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, WALDENSTRÖM, MCMAHON, TOLF, OLSEN and DROBNI2015) suggested that antibiotic resistance genes could be spread via gulls from environmental dissemination of ESBL, which is high in human sources.

The resistant phenotypes

Hernandez et al. (Reference HERNANDEZ, BONNEDAHL, ELIASSON, WALLENSTEN, COMSTEDT, JOHANSSON, GRANHOLM, MELHUS, OLSEN and DROBNI2010) found that the phenotypic resistance in E. coli from a Glaucous-winged gull on the remote Commander Islands was ampicillin, cefadroxil, tobramycin, streptomycin, tetracycline, ciprofloxacin, sulphamethosazol-trimethoprim, trimethoprim, and all isolates from Black-headed gull were multi-drug resistant. The resistance phenotypes from black headed gulls included one with resistance to ampicillin, cephadroxil, gentamicin, tobramycin, tetracycline, ciprofloxacin, sulphamethosazol-trimethoprim, trimethoprim and chloramphenicol; one resistant to ampicillin, cephadroxil, gentamicin, tobramycin, tetracycline, ciprofloxacin, sulphamethosazol-trimethoprim, trimethoprim and chloramphenicol, and finally one isolate resistant to ampicillin and cephadroxil.

Results obtained by Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) showed one isolate from Herring gulls to be resistant to nalidixic acid, one with resistance to ampicillin, nalidixic acid and tetracycline and one isolate resistant to ampicillin, ciprofloxacin, nalidixic acid, sulphamethosazol-trimethoprim and tetracycline. Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) found three ESBL-producing E. coli isolates and all isolates were multi-resistant. One was resistant to ampicillin, cephalothin and sulphonamides, one to ampicillin, cephalothin, gentamicin, sulphonamides and sulphamethosazol-trimethoprim and one to ampicillin, cephalothin, streptomycin, sulphonamides, sulphamethosazol-trimethoprim and tetracycline. Besides that, Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) characterised three quinolone-resistant isolates with the qnrS genes whereby one isolate was resistant to ampicillin, one was resistant to ampicillin, nalidixic acid and tetracycline and one to ampicillin, nalidixic acid, streptomycin, sulphonamides and tetracycline.

Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, MCMAHON, HASAN, OLSEN, DROBNI and WALDENSTRÖM2014) found 59 different phenotypes with resistance to three or more antibiotics and observed the large diversity in resistance phenotypes both within and between sites, so they stated that more than one clone per country was involved. They reported that the highest proportion (6.5%) of resistance to six or more antibiotics, were observed in Spain and that no isolates with resistance to six or more antibiotics were found in gull samples from Denmark, Poland and Ireland.

Mechanisms of antibiotic resistance

Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007) detected specific genes of antibacterial resistance and class 1 integrons in resistant E. coli isolates. They detected the following resistance genes in E. coli isolates: tetA, tetB (strains resistant to tetracycline), blaTEM (beta-lactams), strA, aadA (streptomycin), sul1, sul2 (sulphonamide) and cat (chloramphenicol). They detected the tetA gene in 27 and the gene tetB in 22 out of 49 tetracycline-resistant strains and did not find isolates carrying both genes tetA and tetB. Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007) identified gene blaTEM in 29 of 30 isolates resistant to beta-lactam antibiotics, and did not detect blaSHV and blaOXA genes. Furthermore, they detected the gene strA in 16 of 24 streptomycin-resistant isolates, the aadA gene in five and both genes strA and aadA in three streptomycin-resistant isolates. In 19 of 20 sulphonamide–resistant E. coli Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007) found at least one sul-type determinant, in 14 isolates they found a single sul-type resistance determinant (sul1 in three and sul2 in 11 isolates), and in five isolates they observed both sul1 and sul2, while one sulphonamide-resistant E. coli isolate was negative for both sul-type determinants. They found the cat gene in all five chloramphenicol-resistant isolates but did not find the cmlA gene in chloramphenicol-resistant isolates.

Dolejska et al. (Reference DOLEJSKA, CIZEK and LITERAK2007) detected nine isolates carrying integrons whereby they found the int1 gene in nine (about 12%) of 75 antibiotic-resistant isolates and eight of them carried class 1 integrons. They noted that one resistant isolate carrying an int1 gene was negative for the presence of integron and gene cassettes, and observed integrons with amplicons of 1.6-kb (five isolates) and 1-kb (three isolates). All the 1-kb integrons carried the aadA gene cassette, while the 1.6-kb integrons carried two types of gene cassettes: dfr1-aadA (two isolates) and dfrA12–aadA (one isolate). They did not detect gene cassettes in two isolates. The aadA gene was present in three integron-positive isolates which were susceptible to streptomycin.

Poeta et al. (Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008) detected eight blaTEM-52, one blaCTX-M-1 plus blaOXA-1, one blaCTX-M-14a and one blaCTX-M-32 beta-lactamase genes in E. coli isolates. Of these, 73% of the ESBL-positive E. coli isolates form gulls harboured the blaTEM-52 gene and that 27% of the E. coli isolates harboured the blaCTX-M gene. Moreover, they found different genetic environments surrounding the blaCTX-M genes. The genetic environment of the blaCTX-M1 included IS26/ISEcp1 upstream of the bla gene and orf477 downstream, the genetic environment of the blaCTX-M-14a gene included ISEcp1 upstream of the bla gene and IS903 downstream and in the case of blaCTX-M-32 gene ISEcp1/IS5 were found upstream and orf477 downstream. They also detected different resistance genes, such as cmlA, tetA, tetB, aadA1, sul1, sul2 and sul3 among ESBL-producing E. coli isolates.

Results obtained by Poeta et al. (Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008) further showed that five isolates harboured class 1 integrons with the different gene cassettes in their variable regions. In two isolates they found dfrA1 plus aadA1, in one isolate sat plus psp plus aadA2, in one isolate sat plus aadA1 and in one isolate blaOXA-1 plus aadA1. Their results further showed that five isolates harboured class 2 integrons and the gene cassette dfrA1 plus sat plus aadA1 was found in all of them. One E. coli isolate harboured both class 1 and class 2 integrons.

Dolejska et al. (Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009) identified the following genes in E. coli isolates: blaTEM, blaOXA-1-like (resistance to ampicillin), tetA, tetB (tetracycline), strA, aadA1, aadA2, aadA5 (streptomycin), sul1, sul2, sul3 (sulphonamides) and cat, cmlA, floR (chloramphenicol). They noted that the gene tetA was predominant in tetracycline-resistant isolates from gulls, and isolates with both genes (tetA and tetB) were frequent, while one tetracycline-resistant isolate from a gull was negative for all test determinants.

Dolejska et al. (Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009) found class 1 integrons and gene cassettes in variable regions of the class 1 integron in nine (15%) of 60 antibiotic-resistant E. coli isolates from gulls and class 2 integrons and ESBL-producing E. coli isolates (with blaCTX-M-1, blaCTX-M-15-like, blaSHV-2 and blaSHV-12) in 13% and 3% of gull isolates, respectively. They found three types of class 2 integrons in gull isolates: a class 2 integron with a variable region of 2.5 kb in size and the estX-sat-aadA1 gene cassette in three E. coli isolates; an integron 2 kb in size with the dhfr1-sat-aadA1 cassette in three isolates and a 1.5 kb integrin with the sat-aadA1 cassette in one isolate.

Dolejska et al. (Reference DOLEJSKA, BIEROŠOVÁ, KOHOUTOVA, LITERAK and ČÍŽEK2009) found the following genes in ESBL-producing E. coli isolates: blaCTX-M-1 in one isolate, blaCTX-M-15-like in two isolates, blaSHV-2 in one isolate and blaSHV-12 in two isolates, while no bla genes was found in one ESBL-producing isolate. They observed the same antibiotic resistance profile (resistance to ampicillin and cephalothin) in all blaSHV-positive isolates exhibited. Further, they noted that all blaCTX-M-positive isolates were resistant to between four and seven antibiotics, and two carried a class 1 integron of 1.7 kb with the dhfr17-aadA5 gene cassette, and one contained a 1.5 kb class 1 integron with dhfr1-aadA1.

Radhouani et al. (Reference RADHOUANI, POETA, IGREJAS, GONÇALVES, VINUE and TORRES2009) detected the blaTEM resistance gene in 17 of 23 ampicillin–resistant isolates, although they did not find blaSHV. They found tetA and/or tetB resistance genes, associated with an active efflux system, in 18 of 21 tetracycline resistant isolates. Moreover, they detected aadA genes, encoding an aminoglycoside adenyl transferase that modifies streptomycin in 12 of 17 streptomycin-resistant isolates They found the cmlA gene in all chloramphenicol–resistant isolates; the aac(3)-II with or without aac(3)-IV genes, encoding an aminoglycoside acetyl transferase that modifies gentamicin and tobramycin, in all four gentamycin- resistant isolates and sul1 and/or sul2 and/or sul3 resistant genes in all 14 trimetoprim-sulphamethosazol resistant isolates. Radhouani et al. (Reference RADHOUANI, POETA, IGREJAS, GONÇALVES, VINUE and TORRES2009) detected intI I gene in 10 of 14 trimethoprim-sulphamethosazol-resistant isolates wherein three of them contained class 2 integrons from Yellow-legged gulls (Larus cachinnans) at the Berlengas nature reserve in Portugal. They identified four different gene cassette arrangements among class 1 integrons (aadA, dfrAI+aadAI, dfrA12+orfF+aadA2 and sat+psp+aadA2) and one among class 2 integrons (dfrAI+sat+aadAI).

Radhouani et al. (Reference RADHOUANI, POETA, IGREJAS, GONÇALVES, VINUE and TORRES2009) reported that the TEM β-lactamase was the most frequent mechanism of ampicillin resistance among the isolates (73.9%) and has been detected in resistant isolates from black-headed gulls in the Czech Republic (Dolejska et al., Reference DOLEJSKA, CIZEK and LITERAK2007). Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) found that the presence of ESBL-producing E. coli was high (9.4%) as well as the proportion of the CTX-M-1 group (6%). They detected the presence of blaCTX-M, blaTEM and blaSHV and found that ten isolates harboured CTX-M-1 group enzymes. They defined genotypes to eight isolates with blaCTX-M-1 and one with blaCTX-M-15. The CTX-M-1 group is the major type in human clinical isolates in France (Galas et al., Reference GALAS, DECOUSSER, BRETON, GODARD, ALLOUCH and PINA2008). Bonnedahl et al. (Reference BONNEDAHL, DROBNI, GAUTHIER-CLERC, HERNANDEZ, GRANHOLM, KAYSER, MELHUS, KAHLMETER, WALDENSTROM, JOHANSSON and OLSEN2009) detected blaTEM in combination with blaCTX-M in two isolates and with blaSHV in two isolates and separately in five isolates and isolates carrying class 1 and class 2 integrons. Gionechetti et al. (Reference GIONECHETTI, ZUCCA, GOMBAC, MONTI-BRAGADIN, LAGATOLLA, TONIN, EDALUCCI, VITALI and DOLZANI2008) also found the antibiotic resistant isolates of E. coli carrying class 1 and class 2 integrons in Mediterranean herring gulls (Larus cachinans) in Italy.

Hernandez et al. (Reference HERNANDEZ, BONNEDAHL, ELIASSON, WALLENSTEN, COMSTEDT, JOHANSSON, GRANHOLM, MELHUS, OLSEN and DROBNI2010) found that two ESBL-producing E. coli collected from Black-headed gull at Petropavlovsk river delta harboured blaCTX-M-14 and one harboured blaCTX-M-15, while one isolate from a Glaucous-winged gull on the remote Commander Islands harboured CTX-M-15. They hypothesised that multi-resistant CTX-M-14 and CTX-M-15 harbouring E. coli are a result of clonal dissemination of successful clones with clearly human origin. They further concluded that the relatively susceptible CTX-M-15 is an example of horizontal transfer of a more native CTX-M harbouring plasmid to an E. coli unassociated with clonal dissemination.

Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) found isolates of E. coli with class 1 integrons and gene cassettes in Herring gulls on the Polish coast of the Baltic Sea and stated that the integrons are commonly detected in E. coli isolates from gulls living in Europe in areas inhabited by humans and that the gene cassettes found within these integrons are variable notwithstanding that genes aadA1 and dhfr1 are detected regularly. Results obtained by Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) showed that one isolate from Herring gulls was resistant to nalidixic acid and antibiotic resistant genes were not detected, one isolate showed resistance to ampicillin, nalidixic acid and tetracycline and blaTEM and tetA antibiotic resistance genes were found. One other isolate was resistant to ampicillin, ciprofloxacin, nalidixic acid, sulphamethosazol-trimethoprim and tetracycline and the antibiotic genes blaTEM, tetA, tetB, int1; I1, 1.5 kb: dhfr1-aadA1 were detected. They found that the plasmids containing the bla genes for ESBL production were transferable by conjugation to E. coli and Salmonella in two isolates. The qnrS plasmids were transferable by conjugation to E. coli and Salmonella in two isolates.

Stedt et al. (Reference STEDT, BONNEDAHL, HERNANDEZ, WALDENSTRÖM, MCMAHON, TOLF, OLSEN and DROBNI2015) reported that CTX-M type ESBLs were common in the faeces microbiota from gulls across Europe. They identified 13 different variants of CTX-M by genotyping and reported that blaCTX-M-1 and blaCTX-M-14 were the most frequently detected, whereas blaCTX-M-14 was the most abundant in England, Spain and Portugal, blaCTX-M-15 most abundant in Sweden and blaCTX-M-1 in the rest of the countries (Denmark, Ireland, Latvia, the Netherlands, and Poland). They found blaCTX-M-2 only in three countries, Latvia, the Netherlands and England, and one E. coli carried a blaCTX-M-8 in Spain.

The phylogenetic groups

Poeta et al. (Reference POETA, RADHOUANI, IGREJAS, GONÇALVES, CARVALHO, RODRIGUES, VINUÉ, SOMALO and TORRES2008) found that eight of the ESBL-producing E. coli isolates from seagulls corresponded to the A and B1 phylogenic group and one blaTEM-52 isolate corresponded to the B2 phylogenetic group, which is associated with extra intestinal infections. All examined isolates were isolated from faeces, which explained the low percentage of B2 phylogenetic group.

Hernandez et al. (Reference HERNANDEZ, BONNEDAHL, ELIASSON, WALLENSTEN, COMSTEDT, JOHANSSON, GRANHOLM, MELHUS, OLSEN and DROBNI2010) isolated an ESBL-producing E. coli of phylogenetic group B2 from a Glaucous-winged gull on the remote Commander Islands, while all ESBL-producing isolates from Black-headed gulls belonged to group D. Literak et al. (Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010) found that the ESBL producing E. coli belonged to A, B2 and D phylogenetic groups. They characterised three quinolone-resistant isolates with qnrS genes, and two belonged to A and one to B1 phylogenetic groups.

The preventive measures

It is very important to limit food supply for gulls and keep them away from swimming areas or free range poultry farms. The education of owners of beaches and restaurants is important, as well making councils put efforts into keeping beaches clean (Simões et al., Reference SIMÕES, POIREL, DA COSTA and NORDMANN2010). The prohibition of the feeding of birds is another measure for the prevention of dissemination of resistant E. coli from gulls (Literak et al., Reference LITERAK, DOLEJSKA, JANOSZOWSKA, HRUSAKOVA, MEISSNER, RZYSKA, BZOMA and CIZEK2010). Actions such as monitoring and reporting incidents in the future as well as improving established regulations could contribute to better control of distribution and dissemination of resistant E. coli from gulls.

Acknowledgments

This work was financially supported by a grant from the Ministry of Education, Science and Technological Development, Republic of Serbia, Project number TR 31071.