Identification of Thermotolerant Campylobacter Species by Fluorescence In Situ Hybridization

Abstract

Thermotolerant Campylobacter spp. (Campylobacter jejuni, C. coli, C. lari, and C. upsaliensis) are leading causes of food-borne diarrhea in humans. In this study, the usefulness of fluorescence in situ hybridization (FISH) for the identification of Campylobacter isolates was investigated. A hierarchical FISH probe set that included six group-, genus-, and species-specific probes was developed and evaluated with 12 reference strains and 94 clinical isolates of Campylobacter, Arcobacter, and Helicobacter. FISH correctly identified all isolates to the genus level and detected all thermotolerant Campylobacter isolates. The assay showed high degrees of sensitivity for the identification of C. jejuni (90%), C. coli (97%), C. lari (81%), and C. upsaliensis (100%) to the species level.

Thermotolerant Campylobacter spp. (Campylobacter jejuni, C. coli, C. lari, and C. upsaliensis) are leading causes of food-borne human gastroenteritis and the corresponding late-onset complications, such as reactive arthritis and Guillain-Barré syndrome (5). The phenotypic identification of Campylobacter spp. is complicated and of limited reliability (2, 5, 8, 11, 12). The identification of C. lari and C. upsaliensis to the species level and the discrimination of the close relative Arcobacter from Campylobacter are especially problematic, leading to uncertainty about the true clinical relevance of these organisms (2, 5, 8, 11, 12, 24). Various molecular methods have therefore been proposed as alternative diagnostic methods (5-7, 11-13, 20, 27). Among these, fluorescence in situ hybridization (FISH) has been described for the identification of Campylobacter (15, 25) and its relatives, Helicobacter and Arcobacter (3, 15, 16, 26), in environmental samples and chicken products. FISH is a microscopic method that uses fluorescently labeled oligonucleotide DNA probes that bind specifically to unique target sites on ribosomal RNA (10, 18, 23). The advantages of FISH are its simple methodology, high speed, low cost, and minimal equipment requirements (only a fluorescent microscope is needed) (10, 18, 23). The aim of this study was to establish and evaluate a FISH assay for the identification of thermotolerant Campylobacter in a clinical setting. A hierarchical set of six FISH probes (Table 1) was designed with ARB software (http://www.arb-home.de). One probe covers all Campylobacter and its relatives, Arcobacter and Helicobacter (the HelCArc probe). One group-specific probe targets the four thermotolerant Campylobacter spp. (the Catherm probe). Species-specific probes were designed for C. jejuni (the Cajej probe), C. upsaliensis (the Cup probe), and C. lari (the Clar1 and Clar2 probes). A combination of two probes was implemented for C. lari, since it was not possible to cover this heterogeneous species with a single probe. We did not succeed in designing a probe for C. coli with sufficient sensitivity.

TABLE 1.

Oligonucleotide probes

Probe	Target (position)	Target organism(s)	Sequence (5′-3′)	Reference or source
HelCArc	23S rRNA (1760)	Campylobacter spp., Helicobacter spp., and Arcobacter spp.	AAC AGT CGG GAG GGA CTC	This study
Catherm	23S rRNA (1419)	Thermotolerant Campylobacter spp	GCC CTA AGC GTC CTT CCA	This study
Cajej	23S rRNA (1693)	C. jejuni	AGC TAA CCA CAC CTT ATA CCG	This study
Clar 1a	16S rRNA (622)	C. lari	TCC CAA GCA GTT CAA CGG T	This study
Clar 2a	16S rRNA (1126)	C. lari	GAA GTG TTA GCA ACT AAA T	This study
Cup	16S rRNA (1695)	C. upsaliensis	CTC TAC AGA ATT TGT TGG AT	This study
EUB	16S rRNA (338)	Bacterial kingdom	GCT GCC TCC CGT AGG AGT	1

^aThe two C. lari-specific probes are used simultaneously.

Probes were directly 5′ labeled with the fluorescent dye Cy3 (red) or 6-carboxyfluorescein (FAM; green) (Thermo, Ulm, Germany). Hybridization was performed as described previously with formamide at a concentration of 30% (10, 21, 23). A FAM-labeled eubacterial probe was always implemented as a control (1). Suspensions of bacteria were prepared in 0.9% saline from overnight cultures on agar plates. Ten microliters of the suspension was applied to glass slides. The slides were air dried and fixed for 20 min in 2% paraformaldehyde. Each slide was hybridized with one specific Cy3-labeled probe in combination with the FAM-labeled eubacterial probe (Fig. 1).

FIG. 1.

C. jejuni reference strain stained with FISH probes. The results obtained with C. jejuni ATCC 33560 are shown. Single slides were each stained simultaneously with the fluorescein isothiocyanate-labeled eubacterial probe (EUB) (green; upper row) and the Cy3-labeled Campylobacter-specific probes (red; lower row). Bars, 5 μm.

The assay was first evaluated with 12 bacterial reference strains (C. jejuni subsp. jejuni ATCC 33560, C. coli ATCC 33559, C. lari ATCC 35221, C. upsaliensis ATCC 43954, C. sputorum ATCC 35980, C. concisus ATCC 33237, C. fetus ATCC 27374, Arcobacter butzleri ATCC 49616, Arcobacter cryaerophilus ATCC 43158, Arcobacter nitrofigilis ATCC 33309, Helicobacter pylori ATCC 49396, and Helicobacter pylori DSMZ 4867). All probes correctly stained the corresponding target reference strains without any cross-reaction with nontarget reference strains.

The probes were further evaluated by using 94 isolates cultured from specimens from humans and animals (Table 2) from the Federal Institute for Risk Assessment, Berlin, Germany; the Animal Sciences Group, Wageningen-Lelystad, The Netherlands; and the Institute of Medical Microbiology, University of Ulm, Ulm, Germany. The isolates were phenotypically characterized by phase-contrast microscopy (characteristic morphology and motility) and Gram staining and by examination of catalase and oxidase production, growth at 25°C and 43°C, indoxyl acetate hydrolysis (22), hippurate hydrolysis (19), and susceptibility to nalidixic acid and cephalothin (5, 14). The identities of three C. coli isolates were confirmed by a previously published PCR approach (27).

TABLE 2.

Number and percentage of positive FISH results obtained with the isolates tested

Species	No. (%) of the following species (probes):
	Total	Helicobacter, Campylobacter, Arcobacter (HelCArc)	Thermotolerant Campylobacter (Catherm)	C. jejuni (Cajej)	C. upsaliensis (Cup)	C. lari (Clar1 and Clar2)
C. jejuni	29	29 (100)	29 (100)	26 (90)	0 (0)	0 (0)
C. coli	32	32 (100)	32 (100)	1 (3)	0 (0)	0 (0)
C. upsaliensis	10	10 (100)	10 (100)	0 (0)	10 (100)	0 (0)
C. lari	11	11 (100)	11 (100)	0 (0)	0 (0)	9 (81)
Helicobacter pylori	12	12 (100)	0 (0)	0 (0)	0 (0)	0 (0)

The corresponding group-specific probe correctly detected all thermotolerant Campylobacter isolates without any cross-reaction (Table 2). For the species-specific identification of C. jejuni, the assay showed a sensitivity of 90% (26/29 isolates) (Table 2). The C. jejuni-specific probe showed one false-positive reaction (1/55 isolates) with a C. coli isolate and thus reached a specificity of 98%. The sensitivities of the C. lari-specific probe and the C. upsaliensis-specific probe were 81% (9/10 isolates) and 100% (11/11 isolates), respectively. The specificity of the C. lari- and C. upsaliensis-specific probes was 100% (Table 2).

The most striking capacity of the assay was the 100% reliable recognition of thermotolerant Campylobacter within less than 2 h with limited effort. From a clinical point of view, the identification of a Campylobacter as thermotolerant and, thus, pathogenic is critical in order to initiate adequate therapy and infection control measures. Our results extend a recent report of the successful application of a similar FISH probe for the detection of thermotolerant Campylobacter spp. in poultry (25).

We suggest the use of a two-step FISH procedure for the further differentiation of Campylobacter. In the first step, the C. jejuni-specific probe may be used in combination with the probe specific for thermotolerant Campylobacter (the Catherm probe) and with the probe specific for Campylobacter, Arcobacter, and Helicobacter (the HelCArc probe). This step identifies the most frequent isolate, C. jejuni, with minimal effort (Table 3). Strains that are negative with the C. jejuni-specific probe but positive with all other probes represent thermotolerant Campylobacter spp. other than C. jejuni. The corresponding strains may be further characterized with considerable reliability in a second step by using the species-specific probes (Table 2). Strains that are negative with the three available species-specific probes may be considered C. coli, with a sensitivity of 97% and a specificity of 92% according to the data obtained with our sample collection.

TABLE 3.

Algorithm for interpretation of FISH results

Result obtained with the following probe (probe specificity):				Interpretation
EUB (all bacteria)	HelCArc (Helicobacter, Campylobacter, Arcobacter)	Catherm (thermotolerant Campylobacter)	Cajej (C. jejuni)
+	+	+	+	C. jejuni
+	+	+	−	Thermotolerant Campylobacter other than C. jejuni
+	+	−	−	Arcobacter, Helicobacter, or nonthermotolerant Campylobacter
+	−	−	−	Some bacteria other than Arcobacter, Helicobacter, or Campylobacter
−	−	−	−	No result (the FISH procedure did not work)
+	−	+	−	No result (contradictory binding pattern)

Strains that stain negative with the probe specific for thermotolerant Campylobacter spp. (the Catherm probe) but positive with the probe specific for Campylobacter and its relatives (the HelCArc probe) represent Arcobacter, Helicobacter, or nonthermotolerant Campylobacter spp. Recognition of these strains provides a considerable advantage, because Arcobacter in particular but also nonthermotolerant Campylobacter spp. may be confused with thermotolerant Campylobacter by biochemical methods (4, 9, 24). The corresponding strains may be further analyzed biochemically or by FISH with previously published probes specific for Campylobacter (15, 17), Arcobacter (15), and Helicobacter (3).

In summary, FISH is suitable for the rapid identification of cultured isolates of thermotolerant Campylobacter spp. in a routine laboratory.