Characterization of antimicrobial resistance of recent Salmonella enterica serovar Gallinarum isolates from chickens in South Korea

Abstract

Salmonella enterica serovar Gallinarum isolates (n=105) from chickens in South Korea between 2002 and 2007 were tested for antimicrobial susceptibility by determining minimum inhibitory concentrations of 16 antimicrobials, and their predominant resistance profiles were genetically characterized. Most isolates (99/105; 94.3%) were resistant to nalidixic acid and resistant/intermediately resistant to fluoroquinolones, and 63.8% (67/105) of the isolates were resistant to three or more antimicrobials. Forty-two quinolone-resistant isolates, of which the quinolone resistance-determining regions of the gyrA genes were sequenced, contained a substitution of a Ser to a Phe or Tyr at position 83 (71.4%), or a substitution of an Asp to an Asn, Gly, or Tyr at position 87 (28.6%). Fifty-seven sulphamethoxazole-resistant isolates were tested for the presence of class 1 integrons by polymerase chain reaction, and their resistance gene cassettes were analysed by sequencing. Three different class 1 integrons containing the resistance-gene insert aadA (52.6%; n=30), aadB (12.3%; n=7), or aadB-aadA (12.3%; n=7) were identified. Most isolates harbouring the integron containing aadB-aadA displayed resistance to all three aminoglycosides tested and also showed increased resistance to fluoroquinolones. These findings suggest that fluoroquinolone resistance may be epidemiologically linked to multiple aminoglycoside resistance.

Introduction

Salmonella enterica serovar Gallinarum causes the severe systemic disease of fowl typhoid in chickens and other poultry (Shivaprasad, 2000, 2003). Fowl typhoid is highly host adapted and can be transmitted through the egg by transovarian infection (Shivaprasad, 2000, 2003). Therefore, the main control strategy is to establish breeding flocks free of the causative organism, and to hatch and rear their progeny under conditions that preclude direct or indirect contact with infected birds (Shivaprasad, 2000, 2003). Although strict control measures have been successful in eradicating fowl typhoid from commercial poultry in North America, Western Europe, and other developed countries such as Australia and Japan, the disease still causes considerable economic losses in Africa, Asia, and Central and South America (Jones et al., 2001; Shivaprasad, 2003).

In South Korea, fowl typhoid was first reported in 1992 when a commercial layer farm was implicated in an outbreak (Kim et al., 1995). Subsequently, the disease rapidly spread to other chicken farms including farms of different chicken breeds, and has since had a significant economic impact on the commercial chicken industry in South Korea (Lee et al., 2003b). A national control programme that was implemented has been focused on elimination of infected birds from breeder flocks until the flocks are free of the disease. In spite of these efforts, fowl typhoid continues to be a significant problem in commercial chicken farms in South Korea, which has prompted various efforts including vaccination and antimicrobial therapy to diminish economic losses. Antimicrobial drugs such as β-lactam, aminoglycoside, and fluoroquinolone antibiotics have been widely used to reduce mortality from fowl typhoid, even though the drugs have not been found capable of eliminating infection from a treated flock (Lee et al., 2003b, 2004; Shivaprasad, 2003).

In 2001, a live vaccine based on an attenuated S. enterica serovar Gallinarum 9R (SG 9R) strain (Silva et al., 1981) was introduced to commercial layer farms in South Korea. Subsequently, the number of reported cases of fowl typhoid decreased substantially. Nevertheless, the use of antimicrobial drugs has remained the first disease control option.

In the present study, we examined and genetically characterized antimicrobial resistance of recent isolates of S. enterica serovar Gallinarum from chickens in South Korea that might be related to current control measures.

Materials and Methods

Bacterial isolates

One hundred and five isolates of S. enterica serovar Gallinarum collected from clinical cases of chickens submitted to the National Veterinary Research and Quarantine Service (NVRQS) in South Korea between 2002 and 2007 were tested. S. enterica serovar Gallinarum was conventionally identified based on lack of motility, carbohydrate fermentation, amino acid decarboxylation, and agglutination tests using Salmonella antisera (Becton Dickinson, Sparks, Maryland, USA), as previously described (Ewing, 1986; Grimont & Weill, 2007).

Antimicrobial susceptibility testing

Isolates were tested for antimicrobial susceptibility by determining minimum inhibitory concentrations (MICs) of 16 antimicrobials (all from Sigma-Aldrich, St Louis, Missouri, USA) including ampicillin, amoxicillin, ceftiofur, cephalothin, chloramphenicol, florfenicol, ciprofloxacin, enrofloxacin, nalidixic acid, apramycin, neomycin, streptomycin, gentamicin, tetracycline, trimethoprim, and sulphamethoxazole. MICs were measured by agar dilution in Mueller–Hinton agar plates (Becton Dickinson) according to the guidelines of the Clinical and Laboratory Standards Institute (2006). MIC data were expressed as the MIC range and the concentrations at which 50% and 90% of isolates were inhibited (MIC₅₀ and MIC₉₀, respectively). The MIC breakpoint values of ampicillin (≥32 µg/ml), amoxicillin (≥32 µg/ml), cephalothin (≥32 µg/ml), chloramphenicol (≥32 µg/ml), ciprofloxacin (≥4 µg/ml), nalidixic acid (≥32 µg/ml), tetracycline (≥16 µg/ml), trimethoprim (≥16 µg/ml), sulphamethoxazole (≥512 µg/ml), enrofloxacin (≥2 µg/ml), and gentamicin (≥8 µg/ml) were the breakpoints defined by the Clinical and Laboratory Standards Institute (2002, 2006). The breakpoints for apramycin (≥32 µg/ml), ceftiofur (≥8 µg/ml), florfenicol (≥32 µg/ml), neomycin (≥16 µg/ml), and streptomycin (≥ 32 µg/ml) were used as defined by the Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (2006).

Genomic DNA preparation

Obtained cells of each isolate were lysed by boiling as previously described (Gussow & Clackson, 1989) with some modifications. Briefly, several colonies, all originating from one single colony, were picked from the agar medium with a sterile toothpick and suspended in 100 µl Tris–EDTA (TE) buffer in a microcentrifuge tube. The cell suspension was boiled in a water bath for 5 min and then briefly centrifuged to pellet cell debris. The supernatant was transferred to a new tube and was stored at −20°C until used as a polymerase chain reaction (PCR) template.

PCR amplication of gyrA

The genetic basis of quinolone resistance of isolates was examined by performing PCR amplification and sequencing of the gyrA fragments. A 347-nucleotide gyrA gene fragment was amplified from a genomic DNA template using primers P1 (5′-TGTCCGAGATGGCCTGAAGC-3′) and P2 (5′-TACCGTCATAGTTATCCACG-3′) as previously described (Griggs et al., 1996; Lee et al., 2004).

Detection of class 1 integrons

Genetic determinants for resistance to multiple aminoglycosides were identified by detecting class I integrons and their resistance gene cassettes. PCR amplication of class I integrons was performed with primers 5≠-CS (5≠-GGCATC CAAGCAGCAAG-3≠) and 3≠-CS (5≠-AAGCAGACTTGACCTGA-3≠) as previously described (Levesque et al., 1995). An isolate of S. enterica serovar Gallinarum (SG03030), which had the class 1 integron containing both aadA and aadB in the initial PCR and sequencing analysis, was used as the positive control strain in the subsequent PCR tests.

Sequence analysis

PCR products were revealed by electrophoresis on a 1% agarose gel. Two independent PCR products for each isolate were sequenced on both strands (Macrogen, Seoul, South Korea). The sequences were edited and analysed using Vector NTI software (Invitrogen, Carlsbad, California, USA). The gyrA sequence was aligned and analysed with that previously determined for S. enterica serovar Gallinarum 9501 (GenBank accession number AY183436) to detect mutations in the quinolone resistance-determining region (QRDR) of the gene using Vector NTI software (Invitrogen). The presence of class 1 integrons and resistance gene cassettes was examined by determining the identities of the sequences using the BLAST program of the National Center for Biotechnology Information (Altschul et al., 1997).

Nucleotide sequence accession numbers

Relevant gene sequences determined in this study have been deposited in GenBank as accession numbers GU330239 to GU330246.

Results

A total of 105 isolates of S. enterica serovar Gallinarum obtained from chickens between 2002 and 2007 showed 30 different antimicrobial resistance patterns to the 16 antimicrobials tested. Of these 105 isolates, 103 (98.1%) were resistant to nalidixic acid and 99 (94.3%) were resistant (11 isolates) or intermediately resistant (88 isolates) to the fluoroquinolone enrofloxacin (Table 1). In particular, eight isolates (7.6%) showed intermediate resistance to another fluoroquinolone, ciprofloxacin, as well as resistance to enrofloxacin (data not shown). Also, 67 isolates (63.8%) were resistant to three or more antimicrobials in different combinations; of these, 54 isolates were resistant to at least one aminoglycoside (Table 1).

Table 1.  Antimicrobial resistance patterns in S. enterica serovar Gallinarum isolates (n =105) from chickens, 2002 to 2007.

Resistance phenotype^a	Number of isolates (%)	Year
AMP	1 (1)	2002
NAL	27 (25.7)	2002, 2003, 2004
AMP-NAL	1 (1)	2002
AMP-STR	1 (1)	2002
ENR-NAL	1 (1)	2003
NAL-STR	5 (4.8)	2003, 2004, 2007
NAL-TET	1 (1)	2005
NAL-TMP	1 (1)	2006
AMP-AMX-NAL	2 (1.9)	2002, 2003
AMP-NAL-SMX	9 (8.6)	2002
AMP-NAL-STR	1 (1)	2002
NAL-NEO-STR	1 (1)	2003
NAL-STR-SMX	2 (1.9)	2002, 2006
AMP-NAL-STR-GEN	2 (1.9)	2002
NAL-STR-GEN-SMX	13 (12.4)	2002, 2003, 2005, 2006, 2007
NAL-STR-TET-SMX	1 (1)	2002
AMP-AMX-NAL-GEN-SMX	4 (3.8)	2003, 2005
AMP-AMX-NAL-TET-TMP	1 (1)	2002
AMP-NAL-STR-GEN-SMX	2 (1.9)	2002
AMP-NAL-STR-GEN-TET	1 (1)	2002
ENR-NAL-STR-GEN-SMX	1 (1)	2004
NAL-NEO-STR-GEN-SMX	1 (1)	2004
NAL-STR-GEN-TET-SMX	1 (1)	2002
AMP-AMX-NAL-STR-GEN-SMX	3 (2.9)	2002, 2004, 2005
AMP-AMX-NAL-STR-GEN-TET	2 (1.9)	2002
ENR-NAL-NEO-STR-GEN-SMX	6 (5.7)	2003, 2004
AMP-AMX-ENR-NAL-STR-GEN-SMX	2 (1.9)	2003, 2004
AMP-AMX-NAL-APR-STR-GEN-SMX	1 (1)	2005
AMP-AMX-NAL-STR-GEN-TET-SMX	10 (9.5)	2002, 2003, 2005, 2007
AMP-AMX-ENR-NAL-STR-GEN-TET-SMX	1 (1)	2003
^aResistance phenotypes were determined by measuring MICs using agar dilution in Mueller-Hinton agar plates (Clinical and Laboratory Standards Institute, 2006). The resistance breakpoints were ≥32 µg/ml for ampicillin (AMP), ≥32 µg/ml for amoxicillin (AMX), ≥2 µg/ml for enrofloxacin (ENR), ≥32 µg/ml for nalidixic acid (NAL), ≥32 µg/ml for apramycin (APR), ≥16 µg/ml for neomycin (NEO), ≥32 µg/ml for streptomycin (STR), ≥8 µg/ml for gentamicin (GEN), ≥16 µg/ml for tetracycline (TET), and ≥512 µg/ml for sulphamethoxazole (SMX) (Clinical and Laboratory Standards Institute, 2002, 2006; Danish Integrated Antimicrobial Resistance Monitoring and Research Programme, 2006). Eleven isolates were resistant to ENR and 88 isolates were intermediately resistant to ENR (although the intermediate resistance phenotype is not shown).

Forty-two nalidixic acid-resistant isolates were selected and analysed for base substitutions in the QRDRs of gyrA. Thirty isolates (71.4%) showed a substitution of Ser to Phe or Tyr at position 83, while 12 isolates (28.6%) had a substitution of Asp to Asn, Gly, or Tyr at position 87 in the QRDRs (Table 2). MIC₉₀ values showed that most isolates with a base substitution at position 83 were resistant to enrofloxacin (≥2) and had relatively higher MICs than those with a base substitution at position 87 (Table 2).

Table 2.  Mutations in gyrA genes of 42 nalidixic acid-resistant S. enterica serovar Gallinarum isolates from chickens, 2002 to 2007.

			MIC (µg/ml)^a
Substitution			NAL	Enrofloxacin			Ciprofloxacin
Position 83	Position 87	Number of isolates (%)		Range	MIC50	MIC90	Range	MIC50	MIC90
Ser → Phe	Wild type	15 (35.7)	256	1 to 2	1	2	0.25 to 2	0.5	1
Ser → Tyr	Wild type	15 (35.7)	256	0.5 to 2	1	2	0.25 to 2	1	2
Wild type	Asp → Asn	1 (2.4)	256	1	–	–	0.5	–	–
Wild type	Asp → Gly	5 (11.9)	256	0.5 to 1	0.5	1	0.25 to 0.5	0.25	0.5
Wild type	Asp → Tyr	6 (14.3)	256	0.5 to 1	1	1	0.25 to 1	1	1
Ser, serine; Phe, phenylalanine; Tyr, tyrosine; Asp, aspartic acid; Asn, asparagine; Gly, glycine; NAL, nalidixic acid. ^aMIC50 and MIC90 are the minimum concentrations of the antimicrobial agents required to inhibit growth of 50% and 90% of the isolates tested, respectively. The resistance breakpoints were ≥2 µg/ml for enrofloxacin and ≥4 µg/ml for ciprofloxacin, respectively (Clinical and Laboratory Standards Institute, 2002, 2006).

Fifty-seven sulphonamide-resistant isolates were analysed for the presence of class 1 integrons. The isolates revealed three different class 1 integrons. Thirty isolates (52.6%) had an integron containing an aminoglycoside resistance gene, aadA; seven isolates (12.3%) had an integron containing another aminoglycoside resistance gene, aadB; and another seven isolates (12.3%) had the largest integron containing both aadA and aadB (Table 3). Six of seven isolates harbouring the integron containing both aadA and aadB had resistance to all three aminoglycosides tested, including streptomycin, gentamicin, and neomycin, and also showed increased resistance to fluoroquinolones (Table 3).

Table 3.  Gene cassettes present in integrons identified in 57 sulphamethoxazole-resistant S. enterica serovar Gallinarum isolates from chickens, 2002 to 2007

Cassette	Resistance phenotype^a	Number of isolates (%)
aadA	NAL-STR-SMX	1 (1.8)
	NAL-STR-GEN-SMX	12 (21.1)
	AMP-NAL-STR-GEN-SMX	2 (3.5)
	ENR-NAL-STR-GEN-SMX	1 (1.8)
	NAL-STR-GEN-TET-SMX	1 (1.8)
	AMP-AMX-NAL-STR-GEN-SMX	2 (3.5)
	ENR-NAL-NEO-STR-GEN-SMX	1 (1.8)
	AMP-AMX-NAL-APR-STR-GEN-SMX	1 (1.8)
	AMP-AMX-NAL-STR-GEN-TET-SMX	8(14.0)
	AMP-AMX-ENR-NAL-STR-GEN-TET-SMX	1 (1.8)
aadB	AMP-AMX-NAL-GEN-SMX	4 (7.0)
	AMP-AMX-NAL-STR-GEN-SMX	1 (1.8)
	AMP-AMX-ENR-NAL-STR-GEN-SMX	1 (1.8)
	AMP-AMX-NAL-STR-GEN-TET-SMX	1 (1.8)
aadB-aadA	NAL-NEO-STR-GEN-SMX	1 (1.8)
	ENR-NAL-NEO-STR-GEN-SMX	5 (8.0)
	AMP-AMX-ENR-NAL-STR-GEN-SMX	1 (1.8)
^aResistance phenotypes were determined by measuring MICs using agar dilution in Mueller-Hinton agar plates (Clinical and Laboratory Standards Institute, 2006). The resistance breakpoints were ≥32 µg/ml for ampicillin (AMP), ≥32 µg/ml for amoxicillin (AMX), ≥2 µg/ml for enrofloxacin (ENR), ≥32 µg/ml for nalidixic acid (NAL), ≥32 µg/ml for apramycin (APR), ≥16 µg/ml for neomycin (NEO), ≥32 µg/ml for streptomycin (STR), ≥8 µg/ml for gentamicin (GEN), ≥16 µg/ml for tetracycline (TET), and ≥512 µg/ml for sulphamethoxazole (SMX) (Clinical and Laboratory Standards Institute, 2002, 2006; Danish Integrated Antimicrobial Resistance Monitoring and Research Programme, 2006).

Discussion

The first outbreak of fowl typhoid in South Korea occurred in commercial layer farms in 1992. The affected flocks displayed severe clinical signs and high mortality. Subsequently, the disease rapidly spread nationwide to other chicken breeds and became one of the most important diseases in the poultry industry in the late 1990s (Lee et al., 2003a). Fowl typhoid accounted for 10.3% of the total cases of avian infectious diseases diagnosed at NVRQS in South Korea between 1995 and 2001, and 71.4% of the cases occurred in commercial layers with significant economical losses (Lee et al., 2003a). S. enterica serovar Gallinarum isolates from chickens in South Korea in 1995 were susceptible to most of the conventional antimicrobial agents, whereas many isolates from chickens in 2001 were resistant or intermediately resistant to ampicillin, gentamicin, kanamycin, enrofloxacin, and ciprofloxacin (Lee et al., 2003b). The resistance profile noted in 2001 has apparently continued, since the present examination of isolates acquired from 2002 to 2007 revealed resistance mainly to quinolones and aminoglycosides, and reduced susceptibility to fluoroquinolones.

Lee et al. (2004) reported that fluoroquinolone-resistant S. enterica serovar Gallinarum isolates from chickens in South Korea were not detected in 1995, whereas a number of isolates from 2001 showed reduced susceptibilities to fluoroquinolones showing the MIC range of 2 to 8 µg/ml. They also revealed that fluoroquinolone-resistant isolates had an amino acid substitution at position 83 or 87 in GyrA, explaining their reduced susceptibilities to fluoroquinolones, with the substitution at position 83 predominating. Single mutations in the gyrA QRDR are associated with resistance to quinolones, and the reduced susceptibility to fluoroquinolones in various microorganisms including S. enterica and double mutations in the gene may confer higher-level resistance to fluoroquinolones (Griggs et al., 1996; Eaves et al., 2004). The current examination also revealed amino acid substitutions at positions 83 (71.4%) or 87 (28.6%) in GyrA of the tested quinolone resistant isolates. Although double mutations in the gene have not been found, the present study is consistent with a constant association between gyrA mutations and fluoroquinolone resistance of S. enterica serovar Gallinarum isolates since the year 2000. Further analyses of other genes associated with fluoroquinolone resistance, including gyrB, parC, parE, qnrA, and qnrS (Eaves et al., 2004; Lee et al., 2004), will provide better insights into the mechanisms of fluoroquinolone resistance in S. enterica serovar Gallinarum.

The present study also confirms that nalidixic acid-resistant and multidrug-resistant (MDR) S. enterica serovar Gallinarum are common in chickens in South Korea, and that the resistance to multiple aminoglycosides is mainly mediated by integrons containing the aminoglycoside resistance gene cassette aadA (52.6%), aadB (12.3%), or aadB-aadA (12.3%). Class 1 integrons with a variety of gene cassettes conferring resistance to antimicrobials such as aminoglycosides (aadA1, aadA2, aacA4, and aadB), trimethoprim (dfrA1, dfrA7, and dfrA12), β-lactams (blaPSE1 and blaOXA30), and chloramphenicol (cmlA and cmlA5) have been reported in MDR S. enterica isolates; clonal expansion and horizontal gene transfer may have contributed to the spread of antimicrobial resistance integrons in these organisms (Levesque et al., 1995; Krauland et al., 2009). In a previous study, examination of S. enterica serovar Gallinarum isolates acquired between 1992 and 2001 in South Korea most frequently detected class 1 integrons containing one (aadA), two (aadB-aadA), or three resistance genes (dhfrXII-orfF-aadA) and the integron containing the aadA gene cassette (91%) (Kwon et al., 2002). Although new resistance genes other than aminoglycoside resistance genes were not evident in the integrons of the S. enterica serovar Gallinarum isolates acquired since 2002, the frequency of integrons containing two resistance genes, aadB-aadA, has increased since 2002.

Antimicrobials including aminoglycosides and fluoroquinolones are commonly used in commercial chicken farms to prevent or treat fowl typhoid in South Korea. In particular, fluoroquinolones have been widely used due to the advantage of oral administration and high potency against many Gram-negative organisms (Lee et al., 2004). Therefore, the high resistance of the isolates to quinolones and aminoglycosides, and the reduced susceptibility to fluoroquinolones, can probably be directly attributed to this frequent use of the antimicrobials in chicken farms. Consequently, the high frequency of MDR isolates indicates the consistent use of antimicrobials in chickens despite the introduction of the live vaccine. Current findings also suggest that the increased resistance to fluoroquinolone occurred in parallel with multiple aminoglycoside resistance.

Acknowledgements

The authors thank Mi-Gyeong Nam and Young-Seon Yu for technical assistance. The present work was supported by a grant from the NVRQS, Ministry for Food, Agriculture, Forestry and Fisheries, South Korea.