Gram-negative bacteria are known to enter a viable but non-culturable (VBNC) state, in which they no longer grow on conventional media, but remain intact and retain viability. This phenomenon has been regarded, in many ways, to be analogous to sporulation in Gram-positive bacteria. We have investigated the presence of spoO-like genes in Vibrio cholerae and Escherichia coli by polymerase chain reaction using primers based on conserved regions of Bacillus subtilis spoOA and spoOF genes. We did not identify regions of V. cholerae or E. coli that exhibited complete homology to spoO genes from B. subtilis and, therefore, concentrated on an alternate approach of transposon mutagenesis. Over 2,500 transposon mutants of V. cholerae were screened under low nutrient conditions in artificial sea water for an altered VBNC response, compared to the wild type. Mutant JR09H1 entered the VBNC state more rapidly than the wild type at both 25^oC and 4^oC.

Physiological investigations, done in parallel with the genetic analysis of control of the VBNC state, focused on two aspects of the VBNC response. Metabolic activity of V. cholerae and E. coli cells at the point of entry into the VBNC state was investigated in substrate uptake experiments, in which uptake of ³H-labeled thymidine and ¹⁴C-labeled glucose and acetate were measured. Uptake of substrate decreased as cells entered the VBNC state and dramatically increased with temperature upshift from 4^oC to 30^oC, with recovery of culturable cells of V. cholerae. The increased uptake was not observed with E. coli, which did not recover culturability. Transition from the VBNC state, after temperature upshift from 4^oC to 30^oC, was studied. Results indicated that the recovery observed for V. cholerae during temperature upshift was due to regrowth of a few culturable cells rather than true transition of VBNC cells to the culturable state. However, we do not exclude the possibility that VBNC cells converted to culturable cells as a first step in the process. Elucidation of the mechanism(s) of genetic control of VBNC state will be required before this phenomenon can be fully understood.

Fresh water and marine bacteria are known to enter a viable but nonculturable state, in which they no longer grow on conventional culture media, but remain intact and retain metabolic activity (Xu et al., 1982; Roszak et al., 1984; Rollins and Colwell, 1986; Nilsson et al., 1991). It has been demonstrated that Vibrio cholerae, the gram-negative bacterium responsible for the acute diarrheal disease, asiatic cholera, and the enteric bacterium, Escherichia coli, can enter the viable but nonculturable state in response to adverse environmental conditions (Xu et al., 1982; Colwell et al., 1985; Roszak and Colwell, 1987). In this state, cells are reduced in size, become ovoid and, in contrast to starved cells, cannot grow at all on standard laboratory media (Grimes et al., 1986). However, the method of Kogure et al. (1979) can be used to demonstrate that these cells retain viability. Moreover, it has been shown that viable but nonculturable cells remain potentially pathogenic (Colwell et al., 1985).

We have studied the molecular genetics and physiology of the viable but nonculturable response in V. cholerae and E. coli. To investigate the genetic control of this state, mutants were obtained that demonstrated an altered viable but nonculturable response, compared to the wild type. The use of transposon mutagenesis has proven to be a powerful method for obtaining mutants. Thus, we used a system described by Taylor et al. (1989) to generate over 2,500 mutants. These mutants were screened for altered viable but nonculturable responses.

Physiological investigations focused on two aspects of the viable but nonculturable response. Metabolic activity of V. cholerae and E. coli cells on entry into the viable but nonculturable state was investigated in substrate uptake experiments, in which uptake of tritiated thymidine and ¹⁴C-labeled glucose and acetate were measured.

Preparation of microcosms. Viable but nonculturable cells were obtained by prolonged incubation of microcosm flasks containing nutrient-free medium. Microcosms were prepared in 1 liter, acid-washed conical flasks and contained 500 ml of autoclaved 10 ppt Instant Ocean (Aquarium Systems, Mentor, Ohio) filtered through Sterivex-GS (0.2 m pore size) membranes (Millipore Corp., Bedford, MA). Microcosms were maintained in the dark. Strains of V. cholerae 569B and E. coli ATCC 25922 were used in this study. Microcosms were inoculated with washed logarithmic-growth-phase cells to a final concentration of ca. 10⁵ cells/ml.

Bacterial counts. Total cell numbers were determined by direct microscopic count of formalin-fixed cells strained with 0.1% (w/v) acridine orange solution, as described elsewhere (Hobbie et al., 1977). The number of culturable cells was determined by plate counts on LB agar. The number of viable cells was determined by the direct viable count (DVC) method of Kogure et al. (1979). Cells responsive to the addition of yeast extract in the presence of nalidixic acid (10 g/ml), after overnight incubation of 25^oC, were counted as viable. The difference between plate counts and DVCs is estimated to comprise the number of cells in the viable but non-culturable state.

Transposon mutagenesis. Transposon mutagenesis was performed as described by Taylor et al. (1989). Donor strain of E. coliSM10 pir carrying the vector pRT733 was mated with a streptomycin-resistant derivative of V. cholerae 569B by cross-streaking and incubated at 37^oC for 18 hrs. Selection for kanamycin and streptomycin resistance encoded by TnphoA and V. cholerae 569B, respectively, resulted in transposition of TnPhoA into the chromosome of V. cholerae 569B, since E. coli SM10 pir is streptomycin sensitive and plasmid pRT733 could not be maintained in V. cholerae 569B which lacked the pir gene.

Mutants were screened by a two step screening procedure. The first screening was done in LB medium (Maniatis et al., 1982). Each mutant was grown in a single well of microtiter plates in 250 l LB at 30^oC. Master plates in which LB was supplemented with 30% glycerol were frozen at -70^oC. Mutants were plated every 5 days for 25 days. This screening was designed to detect mutants with an altered response compared to the wild type on prolonged growth in rich medium, on the assumption that these mutants may also have an altered viable but nonculturable response. Eleven mutants were selected for further screening. All these mutants demonstrated a reduced ability to grow on LB agar plates, compared to the wild type, after incubation for ca. 15 days in LB liquid medium.

In the second screening step, the viable but nonculturable responses of the eleven mutants that exhibited an altered response were investigated in microcosm experiments. Samples were monitored by plate counts, AODC, and DVCs, as described above.

Substrate uptake experiments. The thymidine uptake experiment was performed following a cold TCA precipitation procedure (Fuhrman and Azam, 1980). Uptake of ¹⁴C-glucose and ¹⁴C-acetate was determined by pulsing samples (3 ml) with 0.3 Ci of ¹⁴C-glucose (320 mCi/mmol) or 0.3 Ci of ¹⁴C-acetate (55 mCi/mmol) (Dupont, Wilmington, DE). Two ml of formalin-fixed negative controls were pulsed with 0.2 Ci of radioactive substrate. Samples were incubated at room temperature for 2 hrs. Following TCA precipitation, uptake of glucose and acetate were determined by liquid scintillation counting.

Temperature upshift experiments. Microcosms of V. cholerae were prepared according to the method described above. Plate counts and AODC were monitored periodically. Subsamples were removed to sterile polypropylene 15 ml tubes for incubation at 30^oC in the dark without shaking. Plate counts were obtained to monitor culturable bacteria in these subsamples. Dilution experiments were performed to investigate whether increased plate counts observed in V. cholerae microcosms after temperature upshift resulted from true resuscitation or from growth. Microcosms were monitored until plate counts decreased to below 2 CFU/ml. Samples from microcosms were diluted in ASW to give 1/10 and 1/100 dilutions. Incubation temperature of 1/10 and 1/100 dilutions and a 5 ml undiluted control was shifted to 30^oC. Plate counts and total cell counts were monitored as described above.

Transposon mutant with an altered viable but nonculturable response. Over 2,500 transposon mutants were examined for an altered starvation response in rich medium in an initial screening step. Eleven mutants demonstrated a reduced ability to grow on LB agar plates, compared to the wild type, after incubation for ca. 15 days in LB liquid medium. To demonstrate the logarithmic growth in rich medium was not affected by transposon insertion in these mutants, all mutants were grown in 300 ml LB at 30^oC with shaking, and absorbance (OD_{600 nm}) was monitored. None of these mutants presented patterns of logarithmic growth different from the wild type.

In the second screening step, viable but nonculturable responses of the eleven mutants that exhibited an altered response in rich medium were investigated in microcosm experiments. Mutant JR09H1 was tested in duplicate at 4^oC and 25^oC (Fig. 1A and B). This mutant consistently demonstrated more rapid entry into the viable but nonculturable state, compared to the wild type. All plate counts obtained for mutant JR09H1 were lower than corresponding counts for the wild type, at both 4^oC and 25^oC. This difference in culturability was particularly pronounced at 4^oC. At 4^oC, mutant JR09H1 was nonculturable after 13 days, whereas the wild type remained culturable for 27 days. Temperature of incubation has previously been shown to be an important factor influencing the rate of entry of bacteria into the viable but nonculturable state (Wolf and Oliver, 1992).

Total cell counts and direct viable counts of the mutant JR09H1 and wild type were also monitored in the experiments. No differences were observed in these parameters. In addition, no morphological differences between mutant and wild type cells were detected during microscopic examinations.

Mutant JR09H1 is alkaline phosphatase-negative, indicating that the genomic region disrupted by the insertion event in this mutant does not code for an exported or membrane associated protein capable of forming a hybrid protein which displays phoA activity.

Substrate uptake by viable but nonculturable cells. In substrate uptake experiments, V. cholerae and E. coli cells entered the viable but nonculturable state at 4^oC in ca. 30 days (Fig. 2 and 3). At this time there were, at most, 1 plateable cell/10 ml in the microcosms. The number of active cells (determined by DVC) had decreased slightly by day 30 but over 90% of cells were still active when there were less than 1 plateable cells/10 ml. When cells in microcosms had entered the viable but nonculturable state, microcosms were monitored at 4^oC for a further 24 hrs and then subjected to a temperature upshift to 30^oC to observe the effect of release of temperature stress. V. cholerae cells in all three microcosms became plateable after 24 hrs and the plateable cell count reached almost the same level as in the original cultures after 48 hrs (Fig. 2). Viable counts determined by DVC also increased after temperature upshift (Fig. 2). Microcosms were held at 30^oC and plateable cell counts remained steady for ca. 2 days and gradually decreased ca. 100 fold over 6 weeks. Temperature upshift to 30^oC did not result in recovery of culturable E. coli cells (Fig. 3).

Uptake of thymidine by V. cholerae decreased as cells entered the viable but nonculturable state (Fig. 4A). In some cases, uptake decreased to a level below detection when cells had entered the VBNC state. E. coli microcosms exhibited similar patterns of decrease in uptake with loss of culturability (Fig. 5A). Uptake of thymidine by V. cholerae cells increased after temperature upshift to 30^oC (Fig. 4A). No increase in thymidine uptake with temperature upshift occurred in samples containing E. coli cells (Fig. 5A).

Uptake of glucose and acetate decreased as the number of plateable cells in microcosms decreased for V. cholerae (Figs. 2 and 4) and E. coli (Figs. 3 and 5). Cells showed very low levels of glucose and acetate uptake upon entry into the viable but nonculturable state. After temperature upshift, V. cholerae cells exhibited an increase in glucose and acetate uptake (Fig. 4). Substrate uptake levels in E. coli cells were below detection limits before and after temperature upshifts.

We propose that the recovery of culturable V. cholerae cells in these substrate uptake experiments was due to growth of residual culturable cells and that no recovery of culturable E. coli cells was observed because there were no remaining culturable cells to respond to the temperature increase.

The primary aim of these substrate uptake experiments was to determine the metabolic activity of V. cholerae and E. coli cells on entry into the viable but nonculturable state. It was previously shown that cells which lost the ability to form colonies on standard agar media also did not incorporate substrates such as glucose (Barcina et al., 1990). Our results confirm that substrate uptake decreases on the entry of cells into the viable but nonculturable state but does not cease. This indicates that, with loss of culturability, microorganisms maintain viability in a state of greatly reduced substrate uptake and metabolic activity.

Dilution and temperature upshift experiments. In order to investigate whether recovery was due to true resuscitation or to growth of a few culturable cells which remained in microcosms, dilution experiments were done. The initial number of culturable cells in microcosms was 1.7 culturable cells/ml microcosms (Table 1). If true resuscitation occurred, plate counts in 1/10 and 1/100 diluted samples would be expected to recover to 1/10 and 1/100 the level in the undiluted sample, respectively. However, growth from a small number of cells that had remained culturable would give an increase in plate counts in all three samples to approximately the same concentration, since nutrients available in each sample would be similar. Results from this experiment are shown in Table 1. After 67 hrs, plate counts in undiluted 1/10- and 1/100-diluted samples were all at ca. 2.2 x 10⁵ CFU/ml. The increase in the total cell count was similar for all three samples after 91 hrs at 30^oC (Table 2). Dilutions were performed in quadruplicate. One of the 1/100-diluted samples did not recover culturability and those data were not included in Table 1.

Resuscitation of V. vulnificus from the viable but nonculturable state on temperature upshift was reported by Nilsson et al. (1991). However, it was subsequently found that recovery of culturable V. vulnificus cells on temperature upshift was probably due to regrowth rather than resuscitation (Weichart et al., 1992). We have confirmed that true resuscitation from the viable but nonculturable state does not occur in E. coli or V. cholerae following temperature upshift. However, our results clearly indicate that temperature changes are important in the recovery of V. cholerae from the viable but nonculturable state and this phenomenon may be involved in the sporadic nature of cholera outbreaks.

Entry of Gram-negative bacteria into the viable but nonculturable state has important implications for detection of pathogenic bacteria and monitoring of genetically engineered microorganisms. However, genetic and physiological information is needed before an understanding of the regulation of VBNC is developed. Isolation of a transposon mutant with an altered viable but nonculturable response is a significant first step in genetic analysis of the viable but nonculturable state.

Genetic analysis of the V. cholerae transposon mutant JR09H1 is in progress. This mutant consistently demonstrates more rapid loss of culturability than the wild type and may be disabled in a gene important in controlling rate of entry into the viable but nonculturable state. Cloning and sequencing the region of DNA disrupted by the transposition event in this mutant is in progress. Isolation of additional transposon mutants with altered viable but nonculturable responses will also be helpful. The molecular basis of the viable but nonculturable state is likely to be complex (analogous to sporulation in Gram-positives such as Bacillus subtilis) and genetic analysis of a suite of viable but nonculturable transposon mutants is a helpful approach to elucidation of the molecular basis of the viable but nonculturable state. Other approaches, such as analysis of regulons and genetic factors controlling growth and cell division in the so-called "decline" phase of bacterial growth would be useful, as would screening for factors associated with pre-sporulation stages in Gram-positive bacteria. Clearly, much remains to be done to elucidate those stages of the growth curve, other than the lag and log phases, which have been the major focus of bacterial physiologists and geneticists to date.

This work was supported by Cooperative Agreement No. CR-817791-01-0 between the U.S. Environmental Protection Agency and the University of Maryland.

Barcina, I., J. M. Gonzalez, J. Iriberri, and L. Egea. 1990. Survival strategy of Escherichia coli and Enterococcus faecalis in illuminated fresh and marine systems. J. Appl. Bacteriol. 68:189-198.

Colwell, R. R., P. R. Brayton, D. J. Grimes, D. R. Roszak, S. A. Huq, and L. M. Palmer. 1985. Viable, but non-culturable, Vibrio cholerae and related pathogens in the environment: implications for release of genetically engineered microorganisms. Biotechnol. 3:817-820.

Fuhrman, J., and F. Azam. 1980. Bacterioplankton secondary production estimates for coastal waters of British Columbia, Antarctica and California. Appl. Environ. Microbiol. 39:1085-1096.

Grimes, D. J., R. W. Atwell, P. R. Brayton, L. M. Palmer, D. M. Rollins, D. B. Roszak, F. L. Singleton, M. L. Tamplin, and R. R. Colwell. 1986. The fate of enteric pathogenic bacteria in estuarine and marine environments. Microbiol. Sci. 3:324-329.

Hobbie, J. E., R. J. Daley, and S. Jasper. 1977. Use of nucleopore filters for counting bacteria by fluorescence microscopy. Appl. Environ. Microbiol. 33:1225-8.

Kogure, K., U. Simidu, and N. Taga. 1979. A tentative direct microscopic method for counting living marine bacteria. Can. J. Microbiol. 25:415-420.

Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor.

Nilsson, L., J. D. Oliver, and S. Kjelleberg. 1991. Resuscitation of Vibrio cholerae from the viable but nonculturable state. J. Bacteriol. 173:5054-5059.

Rollins, D. M., and R. R. Colwell. 1986. Viable but nonculturable stage of Camylobacter jejuni and its role in survival in the natural aquatic environment. Appl. Environ. Microbiol. 52:531-538.

Roszak, D. B., and R. R. Colwell. 1987. Survival strategies of bacteria in the natural environments. Am. J. Pub. Health 51:365-379.

Roszak, D. B., D. J. Grimes, and R. R. Colwell. 1984. Viable but nonrecoverable stage of Salmonella enteritidis in aquatic systems. Can. J. Microbiol. 30:334-338.

Taylor, R. K., C. Manoil, and J. J. Mekalanos. 1989. Broad-host-range vectors for delivery of TnphoA: Use in genetic analysis of secreted virulence determinants of Vibrio cholerae. J. Bacteriol. 171:1870-1878.

Weichart, D., J. D. Oliver, and S. Kjelleberg. 1992. Low temperature induced non-culturability and killing of Vibrio vulnificus. FEMS Microbiol. Lett. 100:205-210.

Wolf, P. W., and J. D. Oliver. 1992. Temperature effects on the viable but non-culturable state of Vibrio vulnificus. FEMS Microbiol. Ecol. 101:33-39.

Xu, H. S., N. Roberts, F. L. Singleton, R. W. Attwell, D. J. Grimes, and R. R. Colwell. 1982. Survival and viability of nonculturable E. coli and V. cholerae in the estuarine and marine environment. Microb. Ecol. 8:313-323.

Chowdhury Table 1

chowdhury Fig. 1.

Chowdhury fig 2

Chowdhury fig 3

Chowdhury fig 4

Chowdhury fig 5

1. Corresponding author: Dr. Rita R. Colwell, University of Maryland Biotechnology Institute, 4321 Hartwick Road, Suite 550, College Park, MD 20740, Tel: (301) 403-0501, Fax: (301) 454-8123