gyrA Polymorphism in Campylobacter jejuni: Detection of gyrA Mutations in 162 C. jejuni Isolates by Single-Strand Conformation Polymorphism and DNA Sequencing

Abstract

Mutations in the quinolone resistance-determining region of the gyrA gene from 138 ciprofloxacin-resistant (MIC, ≥4 μg/ml) and 24 ciprofloxacin-susceptible (MIC, ≤1 μg/ml) clinical Campylobacter jejuni isolates were subjected to single-strand conformation polymorphism analysis and sequencing. All of the isolates could be assigned to three genotypic clusters based on silent mutations. All resistant isolates had a point mutation at codon 86.

Fluoroquinolones are the antimicrobials most commonly used for treatment of adults with Campylobacter infections. During the 1990s, however, resistance to this antimicrobial group has increased worldwide among Campylobacter spp. (3, 8, 9, 12-15, 18, 19, 21, 22). The primary target of the fluoroquinolones in Campylobacter jejuni has been shown to be DNA gyrase, a type II topoisomerase that is an essential enzyme for DNA replication (2, 20). DNA gyrase is composed of two A subunits and two B subunits encoded by the gyrA and gyrB genes, respectively. A point mutation in the quinolone resistance-determining region (QRDR) of the gyrA gene at codon 86 (ACA to ATA), substituting isoleucine for threonine, is the most common cause of class-wide fluoroquinolone resistance among C. jejuni isolates (17, 24, 27). There have also been anecdotal reports of mutations leading to additional amino acid changes, as well as silent nucleotide mutations (1, 24, 27).

The purpose of the present study was to explore the molecular epidemiology of the QRDR of gyrA among C. jejuni isolates. To this end, we analyzed the QRDR (codons 69 to 120) of gyrA from 138 ciprofloxacin-resistant and 24 ciprofloxacin-susceptible clinical C. jejuni isolates by single-strand conformation polymorphism (SSCP) analysis and DNA sequencing.

A total of 162 clinical C. jejuni isolates collected in Finland between 1995 and 2000 were included in this study. All isolates were from stool samples, and the majority (n = 158) were collected from Finnish travelers returning from 22 different countries. Four isolates were from Finnish patients without any travel history. One hundred thirty-eight of the isolates were resistant to ciprofloxacin (MIC, ≥4 μg/ml), and 24 were ciprofloxacin susceptible (MIC, ≤1 μg/ml). The ciprofloxacin MICs were 8 μg/ml for 10 of the ciprofloxacin-resistant isolates and ≥16 μg/ml for 128 of the ciprofloxacin-resistant isolates. The resistant isolates were from travelers to 20 different countries, the majority being from travelers to Spain (34%) and Thailand (21%). The susceptible isolates were from travelers to 12 countries, the number of isolates varying between one and three from each country.

Chromosomal DNA from each strain was prepared by boiling for 10 min and proteinase K (Promega, Madison, Wis.) digestion for 30 min. DNA was amplified by PCR for the gyrA gene as described by Wang et al. (24). One microliter of [α-³²P]dCTP (activity, 10 μCi/μl) was added to each PCR mixture for SSCP. SSCP was run in accordance with the manufacturer's instructions on nondenaturing MDE Gel (BioWhittaker Molecular Applications, Rockland, Maine) at 4°C. Autoradiograms were incubated overnight. The QRDR of the gyrA gene was sequenced in both directions with PCR primers from one or more representative isolates belonging to a different SSCP pattern (Table 1). Sequencing of a nonradioactive PCR product was performed with an ABI Prism BigDye Terminator Kit (Applied Biosystems, Foster City, Calif.). All of the gyrA sequences of the different SSCP patterns were confirmed by PCR and sequencing with two additional oligonucleotide primers, Cj-gyrA-393 (5′-CTTTGCCTGACGCAAGAG-3′) and Cj-gyrA-759r (5′-TCGCTTTCTGAACCATCA-3′). One representative isolate of every different SSCP pattern was confirmed to be C. jejuni by 16S rRNA gene sequencing (6). C. jejuni RH3583 was used as a control strain.

TABLE 1.

SSCP patterns and QRDRs of gyrA of 138 ciprofloxacin-resistant and 24 ciprofloxacin-susceptible C. jejuni isolates

SSCP pattern	No. (%) of isolates analyzeda		Ciprofloxacin MIC range (μg/ml)	Nucleic acid codons and corresponding amino acids of C. jejuni QRDR of gyrAb
	CIP-S	CIP-R		Codon	Amino acid	Codon	Amino acid	Codon	Amino acid	Codon	Amino acid	Codon	Amino acid	Codon	Amino acid	Codon	Amino acid	Codon	Amino acid
ICIP-S	7 (29)		0.064-0.5	GGT	Gly-78	CAC	His-81	ACA	Thr-86	GAT	Asp-90	CCA	Pro-104	GGC	Gly-110	AGT	Ser-119	GCC	Ala-120
ICIP-R1		21 (15)	16->64	---	-	---	-	-T-	Ile-86	---	-	---	-	---	-	---	-	---	-
ICIP-R2		1 (0.7)	>64	---	-	---	-	-T-	Ile-86	A--	Asn-90	---	-	---	-	---	-	---	-
ICIP-R3		1 (0.7)	16	---	-	---	-	-T-	Ile-86	---	-	T--	Ser-104	---	-	---	-	---	-
IICIP-S	2 (8)		0.25-0.5	--C	-	---	-	---	-	---	-	---	-	---	-	--C	-	---	-
IIICIP-S	15 (63)		0.125-1	---	-	--T	-	---	-	---	-	---	-	---	-	--C	-	--T	-
IIICIP-R1		113 (82)	8->64	---	-	--T	-	-T-	Ile-86	---	-	---	-	---	-	--C	-	--T	-
IIICIP-R2		2 (1.4)	32	---	-	--T	-	-T-	Ile-86	---	-	---	-	--T	-	--C	-	--T	-

^aNumber of isolates sequenced per SSCP pattern: I_CIP-S, 4; I_CIP-R1, 2; I_CIP-R2, 1; I_CIP-R3, 1; II_CIP-S, 2; III_CIP-S, 9; III_CIP-R1, 12; III_CIP-R2, 2. CIP-S, ciprofloxacin susceptible (MIC, ≤1 μg/ml); CIP-R, ciprofloxacin resistant (MIC, ≥4 μg/ml).

^b-, no change.

On the basis of the SSCP patterns of the QRDR of gyrA, all 138 ciprofloxacin-resistant C. jejuni isolates could be distinguished from the 24 susceptible isolates. Eight different SSCP patterns were found: five among the ciprofloxacin-resistant isolates and three among the ciprofloxacin-susceptible isolates. These patterns were designated as follows. Roman numerals I to III were used to define the three susceptible SSCP patterns. The subscript CIP-S indicates ciprofloxacin susceptibility, and the subscript CIP-R indicates ciprofloxacin resistance. The numerals after the subscript CIP-R define resistant variants within each cluster.

The most common susceptible SSCP pattern was III_CIP-S, with 15 (63%) isolates, followed by patterns I_CIP-S and II_CIP-S, with 7 (29%) and 2 (8%) isolates, respectively (Fig. 1). The most common resistant SSCP pattern, III_CIP-R1, was exhibited by 113 (82%) resistant isolates. Pattern I_CIP-R1 was exhibited by 21 (15%) resistant isolates, pattern III_CIP-R2 was exhibited by 2 (1.4%) resistant isolates, and patterns I_CIP-R2 and I_CIP-R3 were exhibited by 1 resistant isolate each.

FIG. 1.

Autoradiogram of SSCP patterns of the QRDR of gyrA of C. jejuni isolates. SSCP patterns are designated as explained in the text and Table 1. Lanes: 1 and 16, 100-bp DNA ladder (Gibco BRL, Life Technologies, Inc., Gaithersburg, Md.) labeled with [γ-³²P]ATP by using T4 polynucleotide kinase; 2 and 3, ciprofloxacin-susceptible isolates showing pattern III_CIP-S; 4 and 5, ciprofloxacin-resistant isolates showing pattern III_CIP-R1; 6 and 7, ciprofloxacin-resistant isolates showing pattern III _CIP-R2; 8 and 9, ciprofloxacin-susceptible isolates showing pattern I_CIP-S; 10 and 11, ciprofloxacin-resistant isolates showing pattern I_CIP-R1; 12, ciprofloxacin-resistant isolate showing pattern I_CIP-R2; 13, ciprofloxacin-resistant isolate showing pattern I_CIP-R3; 14 and 15, ciprofloxacin-susceptible isolates showing pattern II_CIP-S.

We sequenced the gyrA QRDRs of 1 to 12 representative C. jejuni isolates of all eight defined SSCP patterns (Table 1). Compared to ciprofloxacin-susceptible C. jejuni UA580 (24) and C. jejuni NCTC 11168 (11), all of the resistant isolates sequenced had a point mutation at codon 86 (ACA to ATA), substituting isoleucine for threonine. Two resistant isolates harbored one additional mutation leading to an amino acid change each. In addition, silent mutations were identified among isolates with SSCP patterns III_CIP-R1 and III_CIP-R2 (Table 1).

Among members of the ciprofloxacin-susceptible C. jejuni population, the QRDRs of gyrA from the representative isolates (C. jejuni IH 111169 and C. jejuni IH 111607; Fig. 1 and Table 1) showing SSCP pattern I_CIP-S were identical to those of C. jejuni UA580 (24) and C. jejuni NCTC 11168 (11). None of the susceptible C. jejuni isolates had mutations leading to amino acid changes. However, silent nucleic acid mutations were observed. The QRDR of gyrA from the representative isolates (C. jejuni IH 41670 and C. jejuni IH 111677; Fig. 1 and Table 1) showing SSCP pattern II_CIP-S and from those showing SSCP pattern III_CIP-S (C. jejuni RH 3583 and C. jejuni IH 41957; Fig. 1 and Table 1) had mutations in two and three codons, respectively.

Genetic variation of the QRDR of gyrA in C. jejuni was demonstrated here by the finding of eight SSCP patterns among our whole study population and, in particular, by identification of three defined SSCP patterns among members of the ciprofloxacin-susceptible population. Considering the relatively small number of susceptible isolates studied, it seems likely that additional variants of C. jejuni do exist. This polymorphism is surprising because bacterial topoisomerase genes are generally highly conserved even across the borders of the bacterial genera (4, 5, 7, 10, 16, 23, 24). However, the present findings are in line with the results of the C. jejuni genome sequencing project reported by Parkhill et al. (11), who found hypervariability in the sequence of C. jejuni NCTC 11168. They suspected that C. jejuni might be lacking in DNA repair functions. They also assumed that it may not be possible to produce a single definitive sequence for the C. jejuni genome (11).

Since fluoroquinolones are currently recommended as the first-choice empirical therapy for adult patients with suspected bacterial gastroenteritis, recognition of C. jejuni as the causative agent of a patient's disease is important, as is its susceptibility to antimicrobial agents. Many studies have shown that the Thr-86-Ile mutation in the QRDR of gyrA is the most common mechanism causing ciprofloxacin resistance in C. jejuni (17, 24, 27). Consequently, several molecular techniques have been developed to detect this particular mutation (25-27). If specific primers or probes are used, the nucleic acid variation on the areas described in this report or in previous reports (1, 24, 27) might confuse the primer-target interaction and cause false results. This should be considered in clinical laboratories if traditional susceptibility testing methods are replaced with modern PCR- or hybridization-based methods to detect the mutation at codon 86 (Thr to Ile).

A total of 130 ciprofloxacin-resistant and 23 ciprofloxacin-susceptible C. jejuni isolates from patients with an identified travel history were evaluated for their SSCP pattern profiles (Table 2). Of the five different resistant genotypic variants, two SSCP patterns were dominant while the other three were exhibited by merely one or two resistant isolates. The main pattern, defined as III_CIP-R1, which was presented by 113 (82%) resistant isolates, has also prevailed in earlier studies (27). This pattern seems to have spread all over the world, whereas pattern I_CIP-R1 was found mainly in Europe. Unexpectedly, pattern I_CIP-S, which is identical to the C. jejuni wild-type sequence described by Wang et al. (24), was not our leading susceptible variant. Pattern III_CIP-S exceeded this wild-type pattern in frequency, with a percentages of 63 versus 29%.

TABLE 2.

Geographical distribution of the SSCP patterns of gyrA of 130 ciprofloxacin-resistant and 23 ciprofloxacin-susceptible C. jejuni isolates

Continent (no. of countries)	No. of isolatesa		No. of isolates with SSCP pattern:b
	CIP-S	CIP-R	ICIP-S	ICIP-R1	ICIP-R2	ICIP-R3	IICIP-S	IIICIP-S	IIICIP-R1	IIICIP-R2
Asia (10)	7	59	1	3		1		6	54	1
Africa (3)	4	5						4	5
Central America (1)	0	1							1
Europe (9)	12	65	6	16	1		2	4	48

^aCIP-S, ciprofloxacin susceptible (MIC, ≤1 mg/ml); CIP-R, ciprofloxacin-resistant (MIC, ≥4 mg/ml).

^bSSCP patterns are designated as explained in the text and Table 1.

In conclusion, consistent with earlier findings, all of the resistant C. jejuni isolates sequenced here had the Thr-86-Ile mutation in the QRDR of gyrA. Only this mutation distinguished the two main resistant genotypic variants from the two main susceptible variants, suggesting a common origin. Considering that three different SSCP patterns were identified among the relatively small number of susceptible isolates studied, it seems likely that additional variants of C. jejuni exist. Polymorphism of the QRDR of gyrA should be considered when any PCR- or hybridization-based method is used to detect the Thr-86-Ile mutation in gyrA as an indication of C. jejuni fluoroquinolone resistance.