Density-Dependent Sorting of Physiologically Different Cells of Vibrio parahaemolyticus

A pure bacterial culture is composed of clonal cells in differentphysiological states. Separation of those subpopulations iscritical for further characterization and for understandingvarious processes in the cultured cells. We used density-dependentcell sorting with Percoll to separate subpopulations from culturesof a marine bacterium, Vibrio parahaemolyticus. Cells from culturesin the exponential and stationary phases were fractionated accordingto their buoyant density, and their culturability and abilityto maintain culturability under low-temperature and low-nutrientstress (stress resistance) were determined. The buoyant densityof the major portion of the cells decreased with culture age.The culturability of stationary-phase cells increased with increasingbuoyant density, but that of exponential-phase cells did not.Stress resistance decreased with increasing buoyant densityregardless of the growth phase. The results indicate that density-dependentcell sorting is useful for separating subpopulations of differentculturabilities and stress resistances. We expect that thismethod will be a powerful tool for analyzing cells in variousphysiological states, such as the viable but nonculturable state.

INTRODUCTION

Top

Introduction

A bacterial culture started from one colony is a clonal population.This, however, does not mean that all the cells behave identically(20, 23). The presence of subpopulations with different physiologicaland morphological characteristics is known, especially amongcells in the stationary phase of culture (16). Recent studiesindicate that the cell's physiological states are controlledby a set of genes specifically expressed in various growth phases(16, 17, 25, 35).

The presence of subpopulations is also expected for cells inthe viable but nonculturable state (10, 39). It is commonlyobserved that when bacterial cells are transferred from richto poor medium, some cells start losing culturability, althoughthey still show certain metabolic activities (19, 26). Entryinto the viable but nonculturable state is dependent on thecellular growth phase or physiological state. In general, cellsin stationary phase or those experiencing stress conditionstend to retain culturability longer (18, 27). In a culture enteringthe viable but nonculturable state, there may be at least threesubpopulations, that is, cells retaining culturability, cellslacking culturability but retaining metabolic activity (i.e.,viable but nonculturable state), and cells showing no detectablebiological activity. Because they coexist in one batch culture,the separation of cells in the viable but nonculturable statefrom other subpopulations is critical for further detailed investigations.However, currently no simple, reliable method is available forthis purpose.

Density-gradient centrifugation is a useful technique for separatingcells or organelles of different densities. In bacteriology,this technique has been applied to separation of cells in differentstage of the cell cycle (6, 13, 15, 31), separation of competentcells for transformation into Bacillus subtilis (9, 14), differentiationof type 1 and type 2 methanotrophic bacteria (33), separationof bacteria from freshwater sediments (12), and separation ofsubpopulations with different gene expression patterns (24).However, few attempts have been made to separate cells of differentviability or activity. Whiteley et al. (37) combined the tetrazoliumsalt 2-p-iodophenyl-3-p-nitrophenyl-5-phenyltetrazolium chloridemethod and Percoll density centrifugation to separate respiringbacterial cells from mixed populations in natural seawater.Because respiring cells deposit insoluble formazan crystals,the active cells increase in density and are readily separated.Siegele et al. (34) separated subpopulations of Escherichiacoli having high culturability by Regnografin-76 density gradientcentrifugation.

Cells in some of these states might be scarce but physiologicallydistinct from other populations. We therefore examined density-dependentcell sorting with Percoll as a means of separating subpopulationsfrom cultures of a marine bacterium, Vibrio parahaemolyticus,that were in the exponential and stationary growth phases. Wefound that cells with different densities had different culturabilitiesand stress resistances.

MATERIALS AND METHODS

Top

Materials and Methods

Bacterium and culture conditions.
The marine bacterium Vibrio parahaemolyticus (ATCC 17802) wascultured at 25°C in 1/5x ZoBell 2216E medium prepared withsynthetic seawater (21) and 1/5x ZoBell 2216E medium containingpolypeptone (1 g) (Japan Pharmaceutical Company, Tokyo, Japan)and yeast extract (0.2 g) (Difco Laboratories, Detroit, Mich.)per liter of synthetic seawater. Cells from overnight cultureswere inoculated into fresh medium and grown to the exponentialphase (3-h cultures) or stationary phase (48-h cultures). Afterwe had harvested the cells by centrifugation at 6,000 x g for15 min at 20°C, they were resuspended in an appropriateamount of synthetic seawater.

Density-dependent cell sorting.
Percoll was used because of its nontoxicity, low viscosity,and ease of preparation at the desired osmolarity and pH (29).Unless otherwise stated, the cell suspension (1 ml) was layeredon top of 9 ml of a Percoll gradient working solution, whichcontained 6.3 ml of Percoll (Amersham Biosciences, Uppsala,Sweden), 0.9 ml of 4 M NaCl, 0.9 ml of 10x phosphate-bufferedsaline (pH 7.0), and 0.9 ml of distilled water. Density markerbeads (Amersham Biosciences) were added when necessary. Themixture was centrifuged for 20 min at 20,000 rpm at 20°Cwith a Beckman SW40Ti rotor in a Beckman Optima XL-90 centrifuge.Under these conditions, we determined the buoyant density ofthe cells.

Fractionation of samples.
After centrifugation, the Percoll gradient sample was fractionatedwith a Piston gradient fractionator (Biocomp, Fredericton, Canada).Aliquots of 200 µl from top to bottom were transferredinto the wells of a 96-well microtiter plate. The optical densityat 450 nm was measured in a microplate reader (Thermo Max; MolecularDevices, Sunnyvale, Calif.).

Determination of culturability.
Fractionated cell suspensions were diluted with synthetic seawaterto give the same levels of cell concentration for determinationof culturability and stress resistance. The total numbers ofcells in the diluted cell suspensions were kept at approximately10⁶ cells ml^-1. Total counts were obtained by epifluorescencemicroscopy (BH-2; Olympus, Tokyo, Japan) after staining with4',6'-diamidino-2-phenylindole (DAPI) (Wako, Osaka, Japan) (32).Plate counts were determined as CFU on 1/5x ZoBell 2216E mediumagar plates. Culturability was expressed as (CFU/total counts)x 100.

Resistance to low temperature and nutrient deficiency.
Fractionated cells were diluted to 10⁶ cells ml^-1. The cellswere suspended in synthetic seawater and left at 4°C. Assynthetic seawater does not contain any organic substrates (carbon,nitrogen, or phosphorus sources), this condition induced multiplenutrient deficiency. Plate counts were determined after 3 daysof incubation. The results were expressed as stress resistance,which was calculated as (CFU on day 3/CFU on day 0) x 100.

RESULTS

Top

Results

Major changes in buoyant densities of cells in exponential and stationary phases.
Cells in the exponential phase of culture and those in the stationaryphase were subjected to density-dependent cell sorting. Foreach culture phase, one major band with a distinct buoyant densityappeared (Fig. 1). Most of the exponential-phase cells accumulatedat the bottom of the tube and had a buoyant density of approximately1.095 g cm^-3, whereas most of the stationary-phase cells stayednear the top and had a buoyant density of 1.073 g cm^-3. In otherwords, the buoyant density of the cells decreased with cultureage; in addition, the band became slightly broader.

View larger version (60K):
[in this window]
[in a new window]
FIG. 1. Percoll density gradient centrifugation of Vibrio parahaemolyticus. The left and right tubes contain cells from exponential- and stationary-phase cultures, respectively. The densities of marker beads were: blue, 1.018 g cm^-3; orange, 1.035 g cm^-3; green, 1.051 g cm^-3; red, 1.064 g cm^-3; blue, 1.074 g cm^-3; orange, 1.087 g cm^-3; green, 1.102 g cm^-3; and red, 1.119 g cm^-3.

Sorting of cell populations.
After centrifugation, the sample was divided among the wellsof microtiter plates with a Piston gradient fractionator. Thecell concentration in each well was measured according to theoptical density at 450 nm (Fig. 2). Buoyant density differencesbetween cells from the exponential and stationary phases ofculture were again apparent. Cells from exponential-phase cultureswere concentrated near the bottom of the centrifuge tube (peakat fraction 39, some cells in fractions 30 to 40). Cells fromstationary-phase cultures were concentrated near the top ofthe tube (fraction 7); a second small peak of these cells occurredin fraction 39; and many cells were evenly distributed in almostall fractions, suggesting the presence of multiple cell subpopulationsin stationary-phase cultures.

View larger version (24K):
[in this window]
[in a new window]
FIG. 2. Relative concentration of cells in each fraction. The cell concentration was measured as the optical density at 450 nm (OD₄₅₀) after Percoll density gradient centrifugation and fractionation, from the top (fraction number 1) to the bottom of the centrifuge tube. Cells were from cultures in the (•) exponential phase and ( ${triangleup}$ ) stationary phase.

Determination of culturability of fractionated cell populations.
Culturability is one of the most popular, simple indicatorsof the physiological state of cells. The culturability of cellsfrom the exponential-phase culture was greater than 60% in allthe fractions tested, with no clear change in cell buoyant density(Fig. 3). However, the culturability of cells from the stationary-phaseculture ranged from 11.5 to 96.2%; culturability increased sharplywith increasing buoyant density from fractions 4 to 10 and thengradually increased with further increases in buoyant density.Even though many cells from the stationary-phase culture werepresent in fractions 6 to 8, their culturability was low; fewercells were present in each fraction of higher buoyant density,but their culturability was high.

View larger version (17K):
[in this window]
[in a new window]
FIG. 3. Culturability of cells in each fraction. Lower fraction numbers indicate lower densities. Cells were from cultures in the (•) exponential phase and ( ${triangleup}$ ) stationary phase.

Stress resistance of fractionated cells.
We also examined the ability of cells to maintain culturabilityunder conditions of stress. They were suspended in syntheticseawater and held at 4°C for 3 days. The percent changein CFU in samples taken on days 0 and 3 was used as an indicatorof stress resistance. In cells from both growth phases, low-buoyant-densitycells showed higher resistance to stress (Fig. 4). The resistancedecreased with increasing density up to fraction 10 or 20 andremained constant at higher buoyant densities.

View larger version (17K):
[in this window]
[in a new window]
FIG. 4. Stress resistance of cells in each fraction. Lower fraction numbers indicate lower densities. Cells were from cultures in the (•) exponential phase and ( ${triangleup}$ ) stationary phase.

The stress resistance of cells from the exponential-phase culturewas higher than that of cells from the stationary phase in allfractions (Fig. 4). In contrast, the stress resistance of unfractionatedcells from the original culture was lower in the exponential(0.80) than in the stationary (3.89) phase. Therefore, as awhole population, cells in the stationary phase showed higheroverall stress resistance because a larger fraction of cellswere in a low-buoyant-density, stress-resistant state.

DISCUSSION

Top

Discussion

A pure bacterial culture is composed of subpopulations of cellsin different physiological states (20). From the ecologicalpoint of view, the coexistence of these subpopulations can beregarded as a strategy to ensure broad adaptability to environmentalconditions (7, 40). However, those subpopulations are usuallysmall and have been hard to separate from other major populations,mainly due to the lack of an appropriate method. This has beenan obstacle to further characterization of the subpopulationsat the molecular or genetic level.

In this work, density-dependent cell sorting was applied toa pure culture of a marine bacterium, Vibrio parahaemolyticus.Cells from cultures in the exponential and stationary phaseswere fractionated according to their buoyant densities, anddifferences in culturability and stress resistance were foundamong the separated subpopulations.

Cells from the exponential-phase culture with the lowest buoyantdensities showed the highest culturability and stress resistance,although their relative number in the culture was small (Fig.2 to 4). Low-density cells from the stationary-phase culturehad the highest stress resistance but lowest culturability,indicating that the few cells that retained culturability hada high ability to maintain it. It is noteworthy that cells fromstationary-phase cultures or those undergoing stresses are generallyconsidered to retain culturability longer than cells from exponential-phasecultures or those not subjected to stress (18, 27). This generalizationis not supported if the presence of subpopulations with differentphysiological states is considered (Fig. 4).

Many reports on factors that lead to differences in cell densityhave appeared. These factors include differences in bacterialspecies (36), bacterial strains (22, 30), type of intracytoplasmicmembranes in methanotrophic bacteria (33), composition (22)and osmolarity (2, 3, 4) of the growth medium, hyperosmoticshock (5), mutation causing overproduction of proline (1), antibiotic(ß-lactam) resistance mutation (8), presence of capsule(28), growth rate (38), degradation of DNA and RNA (30), cellsize (31), and the phase of the cell division cycle (11). Furtherelaborate biochemical work will be required to clarify whichfactors are involved in the change of buoyant density underthe present experimental conditions. Recently, Makinoshima etal. (24) showed that cells in stages that express differentgenes could be separated by density gradient centrifugation.Our results indicate that this method can be used to obtaincells of different culturabilities and stress resistances, althoughwe do not yet understand how buoyant density is related to theseproperties.

We believe that density-dependent cell sorting is an effectivetool for separating cells in the viable but nonculturable statefrom coexisting subpopulations. Flow cytometry can be also atool for physiological sorting of bacterial cells (41). Priorto sorting, however, a specific staining step is inevitable.This leads to an alteration of the physiological states thatconsiderably limits the subsequent application of experimentalprocedures. In contrast, the cells obtained by density-dependentcell sorting can be used for various works with minimum physiologicaldisturbance. It is noteworthy that this technique has alreadybeen successfully applied to Vibrio alginolyticus and Escherichiacoli for separating cells with different leucine incorporationrates. Therefore, density-dependent cell sorting has great possibilitiesfor separating various types of physiologically different subpopulationsfrom many types of bacteria.

In conclusion, we found that density-dependent cell sortingis useful for separating subpopulations of cultured bacteria.The separated cell subpopulations showed different culturabilitiesand stress resistances. Additional studies, in combination withmolecular approaches (e.g., gene expression and gene regulation),are currently under way in our laboratory.

ACKNOWLEDGMENTS

We thank A. Ishihama, National Institute of Genetics, for kindlyproviding us with the basic idea for this method.

This work was supported by Grant-in-Aid for Creative Basic Research12NP0201 (DOBIS) and by Grant-in-Aid for Basic Research 12490009from MEXT, Japan.

FOOTNOTES

* Corresponding author. Mailing address: Ocean Research Institute, The University of Tokyo, 1-15-1, Minamidai, Nakano, Tokyo 164-8639, Japan. Phone: 81 3 5351 6485. Fax: 81 3 5351 6482. E-mail: nishino{at}ori.u-tokyo.ac.jp.

REFERENCES

Top