ABSTRACT
Survival of Campylobacter jejuni at 4 and 20°C was investigated by using cellular integrity, respiratory activity, two-dimensional (2D) protein profile, and intact DNA content as indicators of potential viability of nonculturable cells. Intact DNA content after 116 days, along with cellular integrity and respiring cells, was detected for up to 7 months at 4°C by pulsed-field gel electrophoresis. Most changes in 2D protein profiles involved up- or down-regulation.
Campylobacter jejuni, a common food-borne enteropathogen in developed countries, is unable to multiply in foods, but it clearly survives in numbers sufficient to cause human disease. It is generally accepted that low temperatures enhance survival of campylobacters (3), whereas high temperatures provoke quick transformation from culturable spiral-shaped to nonculturable coccoid forms. Given the importance of foods as vehicles of C. jejuni along with the extensive use of low temperatures for preservation, studies examining the survival ofC. jejuni are required. Furthermore, the survival of bacterial cells in cold environments could be for longer than that detected by culturability if viability is maintained in the absence of culturability, i.e., if cells are in the so-called viable-but-nonculturable (VBNC) state. This state has been proposed previously for C. jejuni (19, 23), but resuscitation of putative VBNC cells in laboratory animals (22, 23) has not always been reproducible (17). A number of methods based on maintenance of cellular structures (10), metabolic activity (1, 14, 18, 23), and/or the presence of nucleic acid (21, 25, 27) have been proposed to assess the viability of nonculturable cells, but at present, none has been agreed upon as being suitable overall. So, more than one criterion must be taken into account for considering the viability of nonculturable cells (16). In the present study, change in total cell protein profile is also included to test the viability of C. jejuninonculturable cells after exposure to adverse conditions, mainly nutrient depletion and low temperature. The concurrence of spiral and/or coccoid forms in such nonculturable cells is also discussed. TwoC. jejuni strains were used, a human isolate from the Hospital of Txagorritxu, Vitoria-Gasteiz, Spain, designated C-1, and its derivative, C-1RR, obtained after passage twice through the mouse intestine.
Culture conditions and bacterial counts.For culture and long-term incubation purposes, strains were grown on campylobacter agar base (Oxoid) supplemented with 5% lysed horse blood (Oxoid) for 24 h at 42°C, under a microaerobic atmosphere (7% CO2, 8% O2, and 85% N2). For survival experiments, strains were grown in nutrient broth no. 2 (Oxoid) for an additional 24 h, harvested by centrifugation, suspended in 500 ml of phosphate-buffered saline (PBS) (pH 7.3) at a final density of 109 cells ml−1, and then incubated without shaking in the dark at 4 and 20°C. At the time of inoculation and at regular intervals, culturability was assessed by standard plate counting and epifluorescence direct counts. Total bacterial counts were microscopically performed by the standard acridine orange direct procedure (10). Metabolic activity was determined by tetrazolium salt reduction as an indication of an active electron transport chain (18), and the number of respiring cells was determined by staining with 5-cyano-2,3-ditolyl tetrazolium chloride according to the method of Cappelier et al. (4). Counts were the means of at least three determinations. Morphological changes and average dimensions of the C. jejuni cells were monitored by computerized image analysis with PC-Image (Foster Findlay Assoc. Ltd.) with an Olympus epifluorescence microscope (BX40) equipped with a Sony DXC-950-P video camera.
Survival curves.Direct cell counts determined in parallel with respiratory activity and culturability showed that the cellular integrity and respiratory activity were maintained much longer than culturability. In fact, survival continued for up to 7 months based on signs of viability other than culturability. Changes in cell morphology from spiral to coccoid forms were also detected (Fig.1). At the beginning of the incubation period, C. jejuni cells from late log phase were mainly spiral cells with a stable average (95% confidence interval) length of ca. 1.4859 (1.4069 to 1.5649) μm. At the end of the incubation period, this average (95% confidence interval) length significantly decreased to ca. 1.2409 (1.1627 to 1.3191) μm at 4°C and ca. 1.2925 (1.2253 to 1.3597) μm at 20°C. Electron microscopy (Fig.2) revealed typical spiral rods with a single polar (or bipolar) flagellum and a relatively smooth surface and a few spheroid cells with or without flagella.
Survival curves of C. jejuni C-1 during incubation in PBS at 4°C (A) and 20°C (B) and of its derivative C-1RR at 4°C (C) and 20°C (D). Initial numbers of respiring cells in experiments at 4°C were 9.77 × 108 and 1.23 × 109 ml−1 and numbers of culturable cells were 3.14 × 108 and 2.3 × 109 CFU ml−1 for strains C-1 and C-1RR, respectively. At 20°C, initial numbers of respiring cells were 1.22 × 108 and 1.02 × 109 ml−1 and the numbers of culturable cells were 1.85 × 108 and 1.05 × 109 CFU ml−1 for strains C-1 and C-1RR, respectively. Points are mean values from triplicate determinations.
Transmission electron micrographs showing the different morphologies of C. jejuni cells at several times during incubation at 4°C. Bacterial cells were viewed on 300-mesh grids by transmission electron microscopy (Philips CM10 microscope) following negative staining with 2% (wt/vol) potassium phosphotungstate. (A) Typical spiral cell; (B and C) coccoid and concomitant spiral and coccoid forms showing the presence of the flagella; (D and E) cells showing bleb formation (arrows).
Unlike findings for other curved bacilli such as Vibrio parahaemolyticus (13), C. jejuni and other campylobacter (11) cells became spheroid more quickly when kept at room temperature. In previous studies, the transition to nonculturable cells was assumed to be associated with a morphological change from spiral to coccal shape (2, 17). However, our results do not support this assumption. The transition to coccoid form was not always related to the decrease of culturability, as loss of culturability occurred when only a third of the cells were coccoid forms. Furthermore, within the first 30 days of starvation at 4°C, the percentage of respiring cells was higher than those of either spiral or culturable cells, features which were also noted with the C-1 strain after 5 days of incubation at 20°C. Such results could be explained if the spiral forms of C. jejuni cells are in either the culturable or the nonculturable physiological state and if cells other than culturable spiral forms are also active respiring cells. Some of these results are in agreement with those reported for the closely related species Helicobacter pylori(15). There is, however, a major difference; while Kusters et al. (15) propose the loss of culturability as a loss of viability, we observe the viability of these nonculturable cells on the basis of their potential for respiration as well as cellular integrity (and other characteristics, discussed below) in spite of their spiral or coccoid morphology. Therefore, our data indicates that two forms of nonculturable C. jejuni cells might exist: viable and nonviable, which might not correspond with spiral and coccoid forms, respectively.
Strain influence was found mainly on the basis of the highest percentages of spiral cells detected for C-1RR long after nonculturability was reached during incubation at 20°C. It seems likely that the origin of the strain may influence the rate of morphological change. There is evidence to suggest that the adaptation to the intestinal tract enhances the ability of strains to colonize (5) or to be virulent (20). Our results suggest that adaptation to the mouse intestine enhances the maintenance of the spiral shape during starvation at room temperature.
Bleb-like membrane vesicles were also observed around the cells incubated at 4°C. Formation of these visible excrescences, possibly formed by pieces of cell envelope, is known to occur in other curved bacilli such as Vibrio cholerae (12), V. parahaemolyticus (13), and H. pylori(15) during starvation. We agree with the explanation of Jiang and Chai (13) for these features. Instead of degenerative forms as proposed by Kusters et al. (15), these vesicles could be due to a process of cell volume adjustment by bleb formation representing a survival strategy for minimizing cell maintenance requirements and enhancing substrate uptake due to a high surface-volume ratio. It was proposed by McDougald et al. (16) that changes in cell walls and cell membranes allow for long-term stability and survival of the bacterial cells. Our data supports these assumptions because we found blebs and statistically significant (P ≤ 0.0001) cell size reduction as well as maintenance of spiral morphology during incubation of C. jejuni cells at low temperature.
Two-dimensional gel electrophoresis.To further characterize the nonculturable cells, protein profiles from crude cell extracts were analyzed by two-dimensional gel electrophoresis (Fig.3C and D) and compared with profiles at day 0 (Fig. 3A and B) for both strains, C-1 and C-1RR. In controls, five major characteristic bands were readily identified and labeled as proteins 1, 2, 3, 4, and 5. Several other protein bands (arrows in Fig. 3) were usually observed in both strains, but their visualization was silver stain dependent, and that made comparisons difficult. It is noteworthy that all of the C. jejunistrains assayed (not all shown here) had remarkable similarity in two-dimensional profiles. The major outer membrane protein was clearly identified as the protein of about 40 kDa (protein 4). It was predominant in all cases, but its relative amount decreased with increased time of incubation at either temperature. The 62-kDa protein (protein 2) was identified as flagellin, and the 14-kDa protein (protein 5) could be the same as that reported by Wu et al. (26).
Two-dimensional, silver-stained protein profiles ofC. jejuni cells. Profiles are shown for culturable cells of strain C-1 (A) and strain C-1RR (B) from late log phase at day 0 (control) and for nonculturable cells of strain C-1 after 196 days of incubation in PBS at 4°C (C) and 20°C (D). Isoelectric focusing was performed with a gel containing 4.1% (vol/vol) acrylamide and by sodium dodecyl sulfate-polyacrylamide gel electrophoresis with a 12.5% (wt/vol) polyacrylamide gel (Mini Protean II two-dimensional system; Bio-Rad). A low-range sodium dodecyl sulfate-polyacrylamide gel electrophoresis standard (14.4 to 97.4 kDa; Bio-Rad) was run in all the gels. Individual proteins that were usually observed in both strains are labeled with arrows. Numbers 1, 2, 3, 4, 5, and 6 show major characteristic protein bands. Number 7 denotes an up-regulated protein after 196 days at 20°C.
Differences in the protein profiles of culturable and nonculturable cells of C. jejuni were also detected at both temperatures. Most of them could be explained by up- or down-regulation mechanisms except for protein 7 (60 kDa), which was always detected after long-term incubation (196 days) at 20°C, without apparent reduction of expression of the 40-kDa and 62-kDa major proteins. Similar events have been described previously for Vibrio anguillarum(7) in response to growth at adverse temperatures. Culture conditions were quite different, and those authors reported a great decrease of the 40-kDa major outer membrane protein with an increase of the 60-kDa protein. Another possible explanation is that some of the main protein components were degenerating and modifying their isoelectric point. If so, a great reduction of expression of at least one such protein should have been detected, but this was not the case. Therefore, in the absence of more detailed information, this 60-kDa protein must have a different origin and requires future investigation. It must be remembered that de novo protein synthesis below the minimal growth temperature has been observed previously for C. jejuni (9).
DNA maintenance.Restriction endonuclease digestion and pulsed-field gel electrophoresis (PFGE) were carried out to determine whether the DNA content was maintained intact over the long-term incubation at both temperatures. The experiments were carried out with cells of strain C-1, incubated in PBS at 4 and 20°C. Survival sampling was done at the beginning of incubation (as control) and long after cells had become nonculturable. PFGE patterns (Fig.4) with SalI andSmaI were determined because these restriction endonucleases were used by Chang and Taylor (6). The number of fragments we obtained was almost identical with those reported by Chang and Taylor (6).
PFGE of DNA extracted from C. jejuni and digested with SalI (lanes a to e and k) and SmaI (lanes g to i and l). The restriction fragments were separated in a 1.2% agarose gel by electrophoresis for 24 h at 170 V and 14°C, with ramped pulse times from 10 to 35 s (contour-clamped homogeneous electric field DR II PFGE apparatus; Bio-Rad). After electrophoresis, the gels were stained with ethidium bromide. Lanes: a to c, strain C-1 at day 0 of starvation at 1 × 109, 5 × 109, and 1 × 1010 cells ml−1, respectively; d, strain C-1 at day 20 of starvation at 20°C; e, strain C-1 at day 116 of starvation at 4°C; f, λ DNA ladder ranging from 48.5 to 485 kb; g, strain C-1 at day 0 of starvation at 5 × 109 cells ml−1; h, strain C-1 at day 20 of starvation at 20°C, i, strain C-1 at day 116 of starvation at 4°C; j, strain C-1 at day 116 of starvation at 4°C and not digested; k and l, strain C-1 at day 61 of starvation at 20°C.
Digested DNA from controls displayed the same PFGE profile in each run, in spite of the initial number of cells, indicating that the PFGE protocol generated reproducible results (Fig. 4, lanes a to c). Because of the difference in detection levels of the bands, 5 × 109 spiral cells ml−1 were required for DNA extraction, and even-higher numbers were required when cells became coccoid. PFGE profiles from incubations at 4°C were easily compared, but those from incubations at 20°C were difficult to resolve despite the higher number of cells lysed. Decreases in DNA detectability may be attributable not only to the progressive loss and/or degradation of DNA but also to the cell envelope alteration and/or lower efficiency of cell lysis. Cells which enter the VBNC state undergo changes which allow them to survive in the environment for extended periods (16). These changes, observed for several organisms including C. jejuni (19), involve alterations of the composition of the cell wall and cell membrane (and thereby of function) but do not affect cellular or DNA integrity.
Nonculturable cells after 116 days in PBS at 4°C (Fig. 4, lanes e and i) maintained intact chromosomal DNA and yielded easily recognizable bands in agarose gels producing the same profiles as fresh culturable cells from day 0 (Fig. 4, lanes a to c and g). During incubation at 20°C, bands in agarose gels became nearly undetectable. Even so, in profiles of cells from 20 and 61 days at 20°C (Fig. 4, lanes d, h, k, and l) PFGE banding patterns were the same as those from control cells as well as those from cells kept at 4°C. Although the PFGE patterns are relatively stable, bacteria with small genomes, such as C. jejuni, may undergo genetic variation to increase their potential to adapt to new environments (8, 24), and so the variation in PFGE genotype may be attributable to genomic variation. It has been reported previously for Vibrio vulnificus (25) and Legionella pneumophila (27) that prolonged exposure of cells to cold leads to a gradual degradation of DNA and RNA in an increasing fraction of the population, while a small subpopulation that maintains intact nucleic acids may retain viability. Because no changes in the recognition sequence were detected, PFGE genotypes from our strain were stable after 4 and 2 months of incubation at 4 and 20°C, respectively (more than 1 month after becoming nonculturable). These findings may be compatible with viability in such cells. Therefore, we agree with authors who consider these nonculturable cells maintaining intact DNA to be VBNC cells. This conversion to VBNC forms and the transition to coccoid forms are two different but related phenomena. Coccoid forms could correspond to the second phase of conversion proposed by Weichart et al. (25) in the formation of VBNC cells, and thus the real VBNC forms ofC. jejuni could be found among the spiral nonculturable cells that maintain cellular integrity along with intact DNA. However, the inability to isolate exclusively coccoid forms makes this difficult to prove.
ACKNOWLEDGMENTS
This work was supported by Education, University and Investigation Department grant PI95/40 from the Basque Government.
We thank Lourdes Michaus from the Txagorritxu Hospital for providing some of the strains. We thank Susan Barrow for technical assistance with the English language.
FOOTNOTES
- Received 7 June 1999.
- Accepted 10 July 1999.
- Copyright © 1999 American Society for Microbiology