Ion channels in volume regulation of clonal kidney cells

Abstract

Objectives:  Clonal kidney cells (Vero cells) are extensively utilized in the manufacture of biological preparations for disease diagnostics and therapeutics and also in preparation of vaccines. In all cells, regulation of volume is an essential function coupled to a variety of physiological processes and is a topic of interest. The objective here was to investigate involvement of ion channels in the process of volume regulation of Vero cells.

Methods:  Involvement of ion channels in cell volume regulation was studied using video‐microscopy and flow cytometry. Pharmacologically unaltered cells of different sizes, which are presumably at different phases of the cell cycle, were used.

Results:  Ion transport inhibitors altered all phases of regulatory volume decrease (RVD) of Vero cells, rate of initial cell swelling, V _max and volume recovery. Effects were dependent on type of inhibitor and on cell size (cell cycle phase). Participation of aquaporins in RVD was suggested. Inhibitors decelerated growth, arresting Vero cells at the G₀/G₁ phase boundary. Electrophysiological study confirmed presence of volume‐activated Cl⁻ channels and K⁺ channels in plasmatic membranes of the cells.

Conclusion:  Vero cells of all sizes maintained the ability to recover from osmotic swelling. Activity of ion channels was one of the key factors that controlled volume regulation and proliferation of the cells.

Introduction

Cell proliferation is a fundamental property of tissue growth and cell reproduction; central in cell proliferation is the cell cycle. This cycle in eukaryotic cells consists of four phases: quiescent G₀ phase, where a cell may remain over time before re‐entering the cycle; G₁, where the cell increases in size and prepares itself for DNA synthesis; S, where DNA synthesis takes place and replicates the total set of chromosomes; G₂, in where the cell prepares for division; and mitosis, or M phase, in which actual cell division takes place.

There are growing numbers of observations that show that progression through the cell cycle is linked to ion permeability of the plasma membrane and that pharmacological blockage of ion channels may lead to inhibition of cell proliferation (1, 2, 3, 4, 5, 6, 7). Data indicate that the effect may be caused by disruption of cell volume control. The relationship between ion permeability and cell cycle progression is controversial. For example, there are cells in which Cl⁻ current is high in G1, down‐regulated in S and increased once more in M (8). In other types of cells the current is down‐regulated in G₀/G₁ and increased in S [for example, in human cervical cancer cells (9) and Ehrlich Lettre ascites cells (5)] or is independent of the cell cycle phase (10). Participation of ion channels in cell volume regulation is most prominent in recovery of volume by swollen cells (11, 12). During this process, called regulatory volume decrease (RVD), swollen cells lose water by expelling intracellular solutes by separate volume‐sensitive Cl⁻ and K⁺ channels. Therefore, this is the process which is most likely to be affected by cell cycle‐related changes in ion channel activities mentioned above.

Cell size is a fundamental attribute impacting cell design, fitness and function. Although cell size control is highly complex in nature, biosynthetic activity also drives increases in cell size; thus, it is reasonable to view cell size as the sum of previous cell growth. In metazoans, cell cycles can be actively coupled to increase in dimension. At its most basic level, cell size homeostasis in proliferating cells requires coordination of growth with cell division so that on average, each cell division is accompanied by doubling in cell mass [see (13, 14, 15)]. Correlation between the cell cycle and cell size is not a prerogative of metazoans. Impressive results have recently been obtained from fission‐yeast cells, total length of which reflects both cell‐cycle stage and cell size (16). Therefore, sorting cells of untreated, proliferating cultures based on their pre‐swelling size could result in groups of cells at different phases of the cell cycle. However, in the majority of reported studies, experiments have been conducted with cells synchronized with respect to progression through the cell cycle by serum starvation or by cell cycle inhibitors. Such an approach leads to considerable metabolic change and uncouples the synthesis of DNA, proteins and lipids usually occur synchronously during cell cycle progression (17, 18). This means that the physiological state of pharmacologically synchronized cells is considerably different from the natural state (19). Thus, we have chosen to work with pharmacologically unaltered cells sorted simply by size. Throughout this study the smallest, intermediate sized and largest cells were assigned as cells A, B and C groups; it was assumed that cells of these groups would be at the beginning of the cell cycle (G₀/G₁ phase), in S phase and at the end of the cycle (G2/M) respectively. To our knowledge, there is only one previous publication that has opted for this approach (10).

The aim of this study was to contribute to understanding of relationships between ion channel activities, RVD and cell cycle progression, by investigating (i) pattern of RVD for differently sized cells, which presumably represents cells at different phases of the cell cycle and (ii) effects of ion channel inhibitors on amplitude of cell swelling under osmotic challenger, RVD, and on cell growth and distribution between phases of the cycle.

Here, we have used clonal kidney cells (Vero cells), which are extensively utilized in manufacture of biological preparations for diagnostics, therapeutics and for development of vaccines (20, 21, 22, 23, 24). We have demonstrated that all differently sized groups of the cells were able to undergo RVD and found that 5‐Nitro‐2‐(3‐phenylpropylamino) benzoic acid (NPPB) and 4,4‐diisothiocyanostilbene‐2,2‐disulfonic acid (DIDS) (Cl⁻ channel blockers) and tetraethylammonium chloride (TEA) (K⁺‐channel blocker) and glibenclamide (GB) (which blocks ATP‐sensitive K⁺ channels, volume‐activated anion channels and CFTR), inhibited RVD and cells remained swollen over the entire recording period (30 min). Peak cell volume in hypotonic solution, V _max, was found to be higher for the largest, category C, cells. The rate of initial cell swelling and V_max was influenced by the inhibitors in a cell‐size‐dependent manner. Participation of aquaporins in rate of swelling of the cells was suggested. All inhibitors used were able to decelerate cell size increase, arresting it in G₀/G₁ phase. Cumulative data propose a link among cell cycle progression, membrane permeability and volume regulation and permit the suggestion of ion channel inhibitors as additives to be incorporated into the design of more effective anti‐cancer treatment strategies.

Materials and methods

Reagents

Inhibitors of cell permeability DIDS, NPPB, TEA, GB and NaCl, KCl, CaCl₂, MgCl₂, MgSO₄, ATP, EGTA, DMSO, Tris and HEPES were purchased from Sigma (St Louis, MO, USA). DIDS, NPPB and GB were dissolved in DMSO. Final concentrations of DMSO were always <0.1%. DIDS, NPPB and GB were used at concentration of 100 μm. Final concentration of TEA was 10 mm. At used concentrations, these inhibitors did not affect cell viability.

Cell cultures

The Vero cell line is derived from kidney of the African green monkey (Cercopithecus aetiops). Cells were grown as monolayer cultures in plastic tissue culture flasks containing modified Eagle’s medium (MEM; Biosystems, Curitiba, Paraná, Brazil) supplemented with 10% foetal calf serum (FCS; Invitrogen Brasil Ltda, Sao Paulo, SP, Brazil), 100 IU/ml penicillin and 100 μg/ml streptomycin (Sigma) at 37 °C, in a humidified atmosphere of 5% CO₂. Cells were routinely seeded at density of 1 × 10⁴ cells/cm². Osmolality of culture medium was 300 mOsm/kg of H₂O. Cells were subcultured every 3–4 days, and those to be studied were seeded 1–5 days before recording.

Preparation of cells for volume measurements and current recordings

Cells cultured in flasks were trypsinized, centrifuged and re‐suspended in fresh medium. Cell viability measured by trypan blue exclusion assay was >90%; cell suspensions were kept at 25 °C. For electrical and optical recordings, an aliquot of suspension was transferred to a 2‐ml recording chamber mounted on the stage of an inverted microscope (Leica DMIL; Leica Microsystems GmbH, Bensheim, Germany) for 15–20 min. After that time, cells attached to the bottom of the chamber were subsequently superfused with solution containing (mmol/l): 130 NaCl, 2 CaCl₂, 2 MgCl₂, 2.8 KCl and 10 HEPES, with the pH adjusted to 7.4 with Tris‐OH and osmolality adjusted to 300 mOsm/l with mannitol.

Generally, there are two typical procedures to conduct RVD experiments. In the first, cells would be maintained in normal Ringer solution (with all potential determining ions at standard, close to normal extracellular concentration) before application of osmotic challenges. The latter is achieved either by dilution of solution (10, 25) or by partial removal of one of the salts, mainly NaCl, from the Ringer [see (26, 27)]. In rare cases, when these two approaches should be compared, similar swelling‐sensitive current is invariably activated in cell response to either simple standard external solution dilution by addition of distilled water to the bath or by ion removal (28). In the second procedure, cells would initially be placed in considerably modified solution where approximately half normal solution osmolality is built with mannitol. Ion concentrations of such solutions are far from optimal for cells. On the other hand in this case, hypotonic solutions would be achieved by omitting mannitol or by decreasing its (initially high) concentration (3, 29, 30) keeping ion composition constant.

Each procedure has both advantages and disadvantages. We chose to use the first of these, which keeps cells at conditions most approximate to their natural habitat before osmotic stress application. So, hyposmotic solution used in RVD experiments was 1:1 mixture of control bath solution and water. To prepare hyperosmotic solution, osmolality of the standard solution was adjusted with mannitol. Lightly hypotonic solution used in electrophysiological experiments was 5:1 mixture of control bath solution and water. In all cases, perfusion was accomplished using a conventional gravity‐fed flow system (1 ml/min).

The aim of electrophysiological experiments was to qualitatively confirm presence of different types of cation channels in the cells in isotonic solution and considerable increase in anionic conduction under lightly hypotonic stress to support RVD experiments.

Cell volume measurements

Cell volume in RVD experiments was measured using a video imaging system consisting of a CCD video camera (Moticam 2000; Quimis, Diadema, SP, Brazil) attached to the Leica DMIL inverted microscope (Leica Microsystems GmbH). Cell images were collected once per minute during 30‐min long recording periods and were stored directly on to the computer. Each image was then analysed off‐line using freeware image analysis program (ImageJ, NIH, Bethesda, Maryland, USA). Cross‐sectional area of single cells before and after hyposmotic challenge was measured and volume approximated assuming spherical geometry. Cell volume was calculated using equations

where S = cell area (μm²). Peak volume in hypotonic solution was assigned as V _max.

In separate experiments, flow cytometry (FACSCalibur with CellQuest software; BD Biosciences, San Jose, CA, USA) was used to estimate change in volume of the population of large cells under osmotic stress. In this assay, the cell suspension was rapidly mixed with an appropriate volume of water or 900 mm mannitol, then submitted for measurement; ‘blind time’ was around one minute. Cell concentration of the suspension was chosen to be 1 × 10⁶ cells/ml, to shorten time needed to collect information of 20 000 cells, of ∼20 s. Mean value of forward light scattering was assumed to represent mean cell size. All experiments were conducted at room temperature (24 ± 2 °C).

Electrophysiology

Patch‐clamp recordings were carried out at room temperature, using whole‐cell configuration. Recordings were performed using an HEKA EPC8 amplifier (HEKA Elektronik, Lambrecht, Germany) under control of WCP software developed by Dr J. Dempster (University of Strathclyde, Glasgow, Scotland, UK). Patch pipettes were of thin‐walled borosilicate glass (Sutter Instrument, Novato, CA, USA) and had resistances between 2 and 5 MΩ when filled with pipette solution (mmol/l): 127 K‐aspartate, 16 KCl, 2 MgSO₄, 2 ATP, 1 EGTA, 20 HEPES, pH 7.2 with Tris‐OH (285 mOsm/l). Osmolality of solutions was measured using an osmometer (Fiske^R Mark3; Fiske Associates, Norwood, MA, USA). Recorded signals were digitized at 4–10 kHz, filtered at 1/3 sampling rate and stored on a computer. Recordings were not adjusted for electrode junction potentials. Reference Ag/AgCl electrodes were connected to the bath by a 2% agarose/150 mm NaCl salt bridge, assembled within standard 200 μl pipette tips; no series resistance compensation was employed. Membrane conductance, defined as slope of current–voltage (I–V) characteristics, was measured every 30 s, by ramping membrane voltage from −60 to +60 mV (over 2.5 s) relative to holding potential −40 mV. Voltage ramp speed was ∼50 mV/s and capacitive current was <0.5 pA, that is, insignificant in comparison to recorded current of hundreds pA. Increase in cell volume was induced by perfusion of solution with lower osmolality (250 mOsm).

Resting membrane potential of cells measured as zero‐current potential in whole‐cell current clamp recordings was not large (−26 ± 7 mV, N = 5), this is in accordance with previously published data (31).

Presence of outward currents in the cells at isotonic conditions were examined in whole‐cell patch clamp study using depolarization pulse‐protocol. Based on qualitative similarity of our records and previously published data, presence of rectifier K⁺ current, transient outward K⁺ current and Ca²⁺‐activated K⁺‐current were noted. At first glance, we could not record Ca²⁺‐activated K⁺ current due to presence of 1 mm EGTA in our pipette solution. However, detailed analysis of EGTA diffusion through pipette tip into the cells demonstrated that it was small (∼3.4 × 10⁻²²  m/h) and unable to significantly change intracellular Ca²⁺ concentration in our 15‐min experiment (for details, please see Fig. S1). Experiments were conducted at room temperature (24 ± 2 °C).

Evaluation of cell proliferation

Cell proliferation measuring was achieved using the video imaging system consisting of CCD video camera (Moticam 2000; Quimis) attached to the Leica DMIL inverted microscope (Leica Microsystems GmbH). Cells were cultured in plastic tissue culture flasks (seeded at density of 2 × 10³ cells/cm²) at 37 °C in an atmosphere of 5% CO₂ in air. After 36 h culturing, ion transport inhibitors (TEA, GB, DIDS or NPPB) or appropriate amounts of DMSO (solvent control) were added culture media for the following 48 h. Cell images at five fixed locations of each culture flask were collected twice a day and stored directly onto the computer. Each image was then analysed off‐line using a freeware image analysis program (ImageJ, NIH, USA). For each group, the experiment was repeated at least three times.

Cell cycle monitoring by flow cytometry

In this study, proportions of cells in G₀/G₁, S and G₂/M phases was obtained by analysing cells for their different DNA contents based on propidium iodide (PI) staining using Guava Cell Cycle Assay Kit (Guava Technologies Inc., Hayward, CA, USA). Cells were collected in accordance with the manufacturer’s protocol; apparent clumped cells were disaggregated using Guava ViaCount CDR Reagent (Guava Technologies Inc.). Cells were than washed twice in phosphate‐buffered saline (PBS) and fixed in ice‐cold 70% ethanol for at least 60 min. Before analysis, cells were washed twice in PBS, re‐suspended in PBS to ∼1 × 10⁶ cells/ml, mixed with Guava Cell Cycle reagent and incubated at room temperature for 30 min, shielded from light. Stained cells were then analysed by flow cytometry with system software (CytoSoft 2.1.4; Guava Technology Inc., Hayward, CA, USA); for each cell population, 10 000 cells were analysed. Data were deconvoluted mathematically using the flow cytometer system software and percentage of cells in each cell cycle phase was quantified. Flow cytometric analysis was performed in at least three independent experiments for each group.

Statistical analysis

Unless otherwise indicated, data are presented as mean ± SEM of N experiments and, where appropriate, have been analysed using Student’s t‐test or one‐way ANOVA followed by Tukey test. A P‐value of <0.05 was considered statistically significant.

Results and discussion

RVD by flow cytometry

It is known that cells can function as miniature osmometers because amplitude of the first rapid phase of cell volume change in response to alteration of bath solution osmolality is proportional to amplitude of its deviation. To discover whether this is true for Vero cells, we analysed changes in volume of a large population of the cells under osmotic challenges using flow cytometry (FACSCalibur with CellQuest software; BD Biosciences). Mean value of forward light scattering obtained for cells in isotonic solution was taken as 100% volume. Examples of records are presented in Fig. 1 and cumulative results are shown in Fig. 2a.

Figure 1.

  Forward versus right‐angle light‐scatter plots and their respective histograms of a representative Vero cell population under different osmolality levels of bath solution. (a, d) 300 mOsm; (b, e) 150 mOsm and (c, f) 450 mOsm. (b, e and c, f) Parameters recorded within 1–2 min after osmotic shift.

Figure 2.

  Dependence of apparent cell volume on bath solution osmolality (a) and regulatory volume decrease of the cells (b) obtained with flow cytometer analysis. (a) Change in osmolality of bath solution from 300 mOsm to desired level was achieved by mixing cell suspension with appropriate volume of water or 900 mm of mannitol solution. Dashed line demonstrates expected cell volume estimated using modified van’t Hoff law: , where π _i is osmolality of the isotonic solution; V _i is volume of the cells placed in solution with osmolality π _i; and b is osmotically inactive cell volume equal to 16 ± 5% of the total volume of Vero cells that is in the range known for other cells (53). (b) Change in osmolality of bath solution from 300 to 150 mOsm was achieved by a mixture of cell suspension with an equal volume of pure water or 200 μm NPPB solution. Flow cytometer, FACSCalibur, with CellQuest software was employed. Mean value of right‐angle scattering of 20 000 cells measured once per minute. Results presented as mean ± SE of three independent experiments.

It was seen that changes in cell volume were fewer than predicted under the influence of hypo‐osmotic solutions, but corresponded to those for hyper‐osmotic stress. There are several reasons for the clear disagreement seen in the hypo‐osmotic zone. It may indicate greater activity of the RVD process than that of regulatory volume increase (RVI) in Vero cells, on action of a compensatory mechanism or abnormally low light scattering by bleb‐containing cells, the population of which grows with decreasing osmolality of the bath solution (data not shown). This result demonstrates limitations in the use of flow cytometry in RVD studies of Vero cells with solution osmolality below 200 mOsm, although kinetics of volume changes could be recorded (Fig. 2b). It could be seen that the cells in absence of any pharmacological agents undergo RVD, although amplitude of the estimated change in volume did not exceed 20%. The potent chloride channel blocker, NPPB, demonstrated the ability to completely block restoration of cell volume increased by hyposmotic shock.

RVD of individual cells

To be free of flow cytometry limitations, the following examination of RVD was performed with single cells with simple geometry, which did not develop blebs after a hyposmotic challenge. To characterize RVD, we used relative units to determine maximal swollen volumes, V _max (a few minutes after osmotic shift), rates of cell shrinkage and final volumes attained by the end of the measuring period (30 min after the osmotic shift).

General view

First, RVD of individual cells of untreated proliferating culture was measured and summarized to obtain a general view of the process and to compare it with results from flow cytometry. Our results (Fig. 3) indicated that individual cells under hypotonic challenge achieved much larger volumes than indicated by flow cytometry (Fig. 2). Cells initially became swollen, up to one and a half times pre‐swelling volume, then gradually recovered their volume to 110 ± 3%. Characteristic time for this process was 15 ± 2 min.

Figure 3.

  RVD in proliferating Vero cells in control conditions and in presence of specific inhibitor of cellular anion (a) and cation (b) permeability. DIDS, NPPB and GB used at concentration of 100 μm. Final concentration of TEA was 10 mm. N = number of cells in experimental groups. Osmolality of the extracellular solution was switched from 300 to 150 mOsm/l after a few minutes of recording in an iso‐osmotic condition. The inhibitors were present in the hypoosmotic solution. Note the changes in RVD induced by pharmacological blockade of the cell membrane permeability. The initial swelling part of RVD was fitted well with a single exponential function: , where V _i is the cell volume at the time, t ₁; V ₀ is the pre‐swelling volume; t ₁ is the time since the application of a hypotonic solution; is the adjustable parameter analogous to maximal cell volume under osmotic challenge measured experimentally, .τ _swell represents the time required for a cell to increase its volume by up to approximately 63% of the maximum possible value, in other words, when . The decreasing part of the volume recovered process in the absence of cell permeability inhibitors was fitted well with a single‐exponential function: , where t ₂ is the time since the process of the volume decreasing started; V _i is the cell volume at the time, t ₂; V _∞ is an attainable restored cell volume at the end of the RVD process; V _max is the maximal cell volume under osmotic challenge. The characteristic time, τ _rec, of the process is equal to the time required to restore approximately 63% of the altered volume, in other words, when (V _i − V _∞)/(V _max − V _∞) = e⁻¹ ≈ 0.3679. The data are presented as mean ± SE.

RVD was largely inhibited by extracellular application of NPPB, DIDS, GB and TEA (Fig. 3; Table 1). Thus, in the presence of inhibitors cells became swollen, but did not undergo regulatory volume decrease. No volume restoration was established if TEA, NPPB or GB were introduced into the bath solution, and only very slow processes (∼0.5%/min) were observed in the presence of DIDS after 20 min of the osmotic challenge (Fig. 3). Results indicate presence of (and participation in RVD) HCO₃ ⁻/Cl⁻ transporter, Cl⁻ and K⁺ channels in plasma membranes of the cells. Interestingly, rate of swelling, τ_swell, and peak volume in hypotonic solution, V _max, were dependent on type of inhibitor, only two of them, GB and TEA, lowered V _max significantly and only GB significantly altered τ_swell (Table 1). Thus, inhibitors with the ability to block cation channels appeared to alter RVD parameters more effectively than the others. The established change in τ_swell suspects the alteration in water influx under influence of ion transport inhibitors. This supposition is consistent with observation that GB and TEA can affect aquaporin function (32, 33, 34, 35). Moreover, GB, as with practically all other ion channel blockers, can affect different types of ion channels including ATP‐dependent K⁺‐channels, CFTR, Ca²⁺‐activated and swelling‐activated Cl⁻ channels (36, 37, 38, 39, 40).

Table 1.

 RVD characteristics of Vero cells at different phases of cell cycle and under the influence of ion transport inhibitors

	Cell size, μm2	N	, min	V max, %	, %	, min
Control total	160 ± 40	104	1.28 ± 0.3	154 ± 4	162 ± 4.2	15 ± 2
Control A	110 ± 12	41	1.36 ± 0.16	153.7 ± 3.8	160 ± 2.6	12 ± 2
Control B	146 ± 10c	38	1.14 ± 0.17	155.6 ± 3.5	161 ± 3.4	20 ± 5c
Control C	182 ± 12c	25	1.32 ± 0.12	159.5 ± 4.2	165 ± 2.5	15 ± 3c
DIDS total	149 ± 30	25	1.34 ± 0.26	148.6 ± 3.6	150 ± 3.9	No
DIDS A	116 ± 9	8	1.14 ± 0.14	158.7 ± 2.7	161 ± 6.5	No
DIDS B	144 ± 8c	8	1.74 ± 0.2ĉ	150.4 ± 2.2ĉ	154 ± 2.6	No
DIDS C	182 ± 18c	9	1.66 ± 0.3	140.6 ± 2.2cê	139 ± 2ce	No
NPPB total	149 ± 32	60	0.98 ± 0.19	148.6 ± 3.2	148 ± 4.4â	No
NPPB A	113 ± 10	15	1.38 ± 0.16	157.2 ± 4.1	161 ± 4.4	No
NPPB B	153 ± 11c	23	0.95 ± 0.1ĉ	142.9 ± 3.5cê	139 ± 2ce	No
NPPB C	180 ± 11c	22	0.83 ± 0.12ce	149.5 ± 4.6	146 ± 3.2ce	No
TEA total	149 ± 32	58	1.47 ± 0.4	139.3 ± 3.1â	141 ± 5.5a	No
TEA A	112 ± 10	14	1.48 ± 0.1	157.8 ± 4.5	158 ± 2	No
TEA B	149 ± 13c	18	1.41 ± 0.35	139.3 ± 3.3ce	141 ± 3.7ce	No
TEA C	186 ± 12c	26	1.43 ± 0.12	130.2 ± 2.3ce	130 ± 2ce	No
GB total	154 ± 32	49	3.57 ± 0.72a	137.7 ± 2.2a	138 ± 5.2a	No
GB A	118 ± 9	6	4.37 ± 0.42e	129.8 ± 3.7ê	127 ± 2.2e	No
GB B	152 ± 13c	24	3.95 ± 0.37e	139.3 ± 3.2e	139 ± 2.8ĉe	No
GB C	189 ± 16c	19	2.81 ± 0.19ce	139 ± 2.2ĉê	138 ± 2.8ĉ	No

Statistical analysis of the difference among A‐, B‐ and C‐groups of cells was made with one‐way ANOVA followed by Tukey test. The results indicate that the difference among A‐, B‐ and C‐groups is significant, whereas the difference among different A‐groups (control and treated by inhibitors), B‐groups and C‐groups is not significant at the 0.05 level.

Values V _max and are presented as mean ± SE. All other values are as mean ± SD.

N– is the number of cells; is the characteristic time of initial cell swelling established from the fitting of the initial swelling part of RVD with the equation, where V _i is the cell volume at the time, t ₁; V ₀ is the pre‐swelling volume; t ₁ is the time since the application of a hypotonic solution; is the adjustable parameter analogous to maximal cell volume under osmotic challenge measured experimentally, V _max; represents the time required for the cell to increase its volume up to approximately 63% of the maximum possible value, in other words, when . is the characteristic time of volume recovery established from the fitting of the decreasing part of the volume recovered process with the equation: ; where is an attainable restored cell volume at the end of the RVD process; t ₂ is the time since the process of the volume decreasing started. , represents the time required to restore approximately 63% of the altered volume, in other words, when .

^a and ^â The difference between the RVD parameters of the whole populations of cells treated by inhibitors and the control is significant at ≤0.01 and ≤0.05 levels, respectively.

^c and ^ĉ The difference between the RVD parameters of A‐ and B‐ or C‐cell populations, which are presumably at different phases of cell cycle, within groups is significant at ≤0.01 and ≤0.05 levels, respectively.

^e and ^ê The difference between RVD parameter of A‐, B‐ or C‐cell populations treated by inhibitors and its control counterpart is significant at ≤0.01 and ≤0.05 levels, respectively.

There are at least two suppositions which may explain such broad effects of inhibitors. It may be a result of presence of similar binding sites of those different protein structures, that is low specificity, or that plasma membrane proteins are not independent, but assembled in a functional platform where alteration of one component (by highly specific inhibitors) might modify function of the others by possible long‐distance allosteric mechanisms. This question is currently under debate (41, 42, 43, 44, 45, 46).

RVD in cells of different sizes

A major problem with any cell cycle‐related study based on measurements of individual cells, is lack of a reliable means of identifying cycle status of any individual living cell. At any given time, an unsynchronized proliferating culture contains cells which are in all possible phases of the cell cycle. The advantage of using such cells is in no use of any chemical agents, preserving their normal physiological conditions. The challenge is in identification of cycle status of any individual living cell. In this connection, it is important to be reminded of the correlation between size of cells proliferating in a culture and their position in the cell cycle; this has been well documented more than a decade ago (13, 17, 47). As a result, differently sized cells represent those at different phases of cell cycle. Nevertheless, we have conducted two special experiments to verify whether this is true for Vero cells. First, we measured cross‐sectional area of cells (a directly measured parameter of cell size) of the hundreds cells in isotonic conditions (dimensions only of trypan blue excluding cells were collected), and second, distribution of nuclear DNA content within a cell population as determined by flow cytometry. Cumulative results are shown in Fig. 4.

Figure 4.

  Distribution of Vero cells by their size (a and c), DNA content (b) and RVD in proliferating cells of different sizes (d) in the absence of ion channel inhibitors. (a) The size distribution of the trypan blue excluded Vero cells from proliferating cultures in isotonic solution. (b) To measure DNA content, the cells were plated at a density of 2.5 × 10⁵ per 25 cm² culture flask and were cultured for 3 days. Flow cytometric analysis was used to measure the distribution of cells in different cell cycle phases. The first peak in the histograms represents cells in the G₀/G₁ phases. The second peak represents the cells in the G₂ and M phases. Between are cells in the S phase. (c) The size distribution of Vero cells (in isotonic solution), which did not produce the blebs under hypotonic challenger. (d) Each curve represents proliferating cells of different sizes (A < B < C), which presumably correspond to cells at the G₀/G₁; S and G₂/M phases of the cell cycle respectively. Other conditions are as in the legend to Fig. 3. The correlation between the characteristic time of swelling, τ _swell, and volume recovery, τ _rec, is shown in inset. The data are presented as mean ± SE.

It could be seen that Vero cell size distribution (Fig. 4a ) resembled cell distribution among the cell cycle phases (Fig. 4b). Both show decrease in number of cells with increase in size or DNA content, indicating presence of the expected cell size–cell cycle phase relationship. Quantitative differences between these distributions might be the result of selection criterion used (trypan blue exclusion) for sampling of individual cells and a relatively small sample was used to construct the size distribution histogram. Despite the difference, results allowed us to assume that differently sized Vero cells were indeed cells in the different cell cycle phases.

Average size of cells of proliferating cultures, which did not produce blebs under hypotonic challenge [RVD as described in the preceding section (Fig. 3)], was 160 ± 40 μm² (n = 104). Size distribution histogram of these cells is shown in Fig. 4c. This histogram is considerably different from histograms obtained at isotonic conditions (Fig. 4a) as a result of one more selection criterion applied. However, both distributions were well fitted by sum of three Gaussian plots and cells were divided accordingly into three differently sized groups: A (110 ± 12 μm²), B (146 ± 10 μm²) and C (180 ± 12 μm²), which may indeed respectively match cells of G₀/G₁, S and G₂/M phases of the cell cycle. Differences in average sizes for neighbouring groups are significant (P < 0.01; ANOVA followed by post‐tests). Average RVD for each of the three cell groups is shown in Fig. 4d and quantitative characteristics of the process and statistics are shown in Table 1.

It appears that parameters of RVD of Vero cells depend, albeit not strongly, on their size (Fig. 4d, Table 1). We discovered that rates of swelling and volume restoration changed reciprocally (correlation coefficient, r = −0.96; Fig. 4d, inset). Volume restoration was more complete (end volume of ∼104%) for B‐cells than for A‐ (∼110%) and C‐cells (∼110%) (Fig. 4d) and peak volume, V _max slightly increased with cell size. Established regularities indicate that possible cell cycle‐dependent changes in transport processes were involved in RVD.

Effectiveness of inhibitors to alter RVD parameters of the cells depended strongly on their size (Fig. 5; Table 1). Thus, under the influence of DIDS, NPPB or TEA, V _max of A‐cells was somewhat higher in comparison to its control counterpart, whereas in the presence of GB, V _max was significantly lower (Fig. 5e). All inhibitors used significantly reduced V _max of larger cells (B and C groups), with exception of B‐group in DIDS and C‐group in NPPB, where the effect was not significant. As a result, cells were ranked according to alteration in their V _max by inhibitors:

Figure 5.

  RVD in proliferating cells of different sizes in the presence of ion channel inhibitors. Each curve represents proliferating cells of different sizes (groups A, B and C, see the text) and is labelled correspondingly. Conditions are as in the legend to Fig. 3. The data are presented as mean ± SE. Note the changes in RVD induced by pharmacological blockade of Cl⁻ channels (DIDS and NPPB) and K⁺‐channels (GB and TEA) in comparison with control behaviour (Fig. 4d).

• 1(DIDS) C > B > A
• 2(NPPB) B > C > A
• 3(TEA) C > B > A
• 4(GB) A > B ∼ C.

Also as expected, swelling rate was affected by inhibitors in a cell size‐dependent manner. In comparison to control counterparts, DIDS significantly slowed down the process in B‐cells. The decelerating influence of TEA on swelling rate was not significant. Strong deceleration was established under the influence of GB, where cells were ranked according to swelling alteration as follows, A > B > C. In contrast to other inhibitors, NPPB exerted a positive significant influence on rate of swelling of cells of all size groups, C > B > A (Fig. 5; Table 1).

The process of volume restoration of all groups of cells was largely blocked under the influence of any of the inhibitors used. Only a very slow restoration rate was observed with the A‐cells in DIDS and TEA groups (∼0.6 and 0.35%/min, respectively; Fig. 5b,c).

Analysis of the data demonstrates that the ability of inhibitors to affect initial cell swelling and V _max depended on cell size. It is plausible that differences in cell sensitivity may be a result of cell cycle‐dependent change in ion channels and aquaporin expression as a priory, each type of ion channel/aquaporin possesses different pharmacology.

Arrest of cell cycle progression by ion channel inhibitors

To discover whether ion channel inhibitors, which altered RVD, would be able to arrest cell cycle progression of Vero cells, cell distribution in the various cell cycle phases was investigated. Data show that in the control group (no additives), 64.0 ± 7.6% (n = 7) of the cells were at G₀/G₁ phase. However, the G₀/G₁ cell population increased significantly when cells were treated with ion channel inhibitors (Fig. 6 inset). Effects of NPPB and DIDS were stronger than TEA and GB. Results are consistent with decrease in rate of cell proliferation under the influence of all tested inhibitors (Fig. 6).

Figure 6.

  Influence of ion transport inhibitors at Vero cell growth. Cells were cultured in plastic tissue culture flasks (seeded at a density of 2 × 10³ cells/cm²) at 37 °C in an atmosphere of 5% CO₂ in air. After 36 h of culturing, the ion transport inhibitors (TEA, GB, DIDS or NPPB) or an appropriate amount of DMSO (solvent control) was added to culture media for the following 48 h as indicated. DIDS, NPPB and GB were used in concentration of 100 μm. The final concentration of TEA was 10 mm. Each point is expressed as the mean ± SE of three determinations. Phase distribution of Vero cells are shown in Inset. Vero cells were plated at a density of 2.5 × 10⁵ per 25 cm² culture flask and were cultured for 24 h. Cells were then incubated in medium without (control) or with channel blockers, NPPB (100 μm), DIDS (100 μm), GB (100 μm) or TEA (10 mm). Flow cytometric analysis was carried out at 48 h after treatments. Data are shown as percentage of total number of cells analysed and are the mean ± SE of at least five experiments. *P < 0.05; **P < 0.01 (versus control).

Therefore, RVD capacity and progression of Vero cells through the cell cycle changed significantly under the influence of ion channel inhibitors. This observation suggests that cation‐ and anion channels play important roles, not only in volume regulation, but also in modulation or control of cell cycle progression of these cells.

Electrophysiological evidence of Cl⁻ channel participation in RVD of Vero cells

Volume‐sensitive ion channels were activated by hyposmotic swelling of patched cells and their progress was measured by applying voltage ramps (current spikes in Fig. 7a). Activation of volume‐sensitive channels took approximately one minute. Increase in membrane conductance could be seen as increase in slope of IV relationship (Fig. 7b). Nevertheless, pharmacology of volume‐sensitive Cl⁻ channels in various cell types is somewhat different (48); Cl⁻ channel inhibitor, NPPB, effectively blocked volume‐sensitive conductance of the cells (Fig. 7). As shown in 3, 5, the cells were not able to undergo RVD in presence of 100 μm NPPB and remained swollen for the entire recording period. This observation suggests that magnitude of changes in volume‐sensitive Cl⁻ conductance is one of the factors responsible for rate of cell shrinkage during RVD. Swelling‐activated K⁺ current did not appear in the recording demonstrated in Fig. 7, probably since the lightly hypotonic stress used was not sufficient to activate K⁺ channels. Our RVD data (obtained under strong hypotonic stress) indicate involvement of K⁺ channels in the process.

Figure 7.

  The inhibition of I _Cl,vol by extracellular applied NPPB in a representative cell. (a) The current was induced by gentle hypotonic stress as described in the Methods. The current spikes were caused by ±60 mV ramps (2.5 s long) applied every 30 s from the holding potential of −23 mV. The voltage ramp speed was ∼50 mV/s. The capacitive current was <0.5 pA, i.e. it was insignificant in comparison with the recorded current of hundreds pA. The numbers above the current traces show the voltage ramps that were used to generate the averaged IV characteristics shown in panel (b). Dashed line indicates zero current. (b) Averaged IV characteristics (means ± SEM) for control condition (300 mOsm/l) and for hypotonic stress (250 mOsm/l; when the current was fully activated and the reversal potential were about −30 mV) prior to and during NPPB application.

As one might expect, we observed a low level of depolarization under the influence of TEA and glibenclamide applied at isotonic conditions (data not shown). However, there were no considerable changes in resting potential under hypotonic challenge, probably due to two simultaneously occurring processes: (i) shifting of equilibrium potential for potential‐determining ions (K⁺, Na⁺, and Cl⁻) and (ii) activation of Cl⁻ channels (that increased relative permeability of ion Cl⁻) that appeared to compensate for each other (results of simulation are presented in Fig. S1). Mean values of reversal potential obtained from I–V relationships, as shown in Fig. 7 were found to be very close to one another: −28 mV; −30 mV and −27 mV for isotonic, hypotonic and hypotonic + NPPB solutions respectively. The moderate shift in reversal potential towards hyperpolarization under hypotonic stress was expected (please see Fig. S1). However, shift in V _rev towards depolarization under combined hypotonic stress + NPPB influence was unforeseen. Theoretical analysis shows that this is only possible if NPPB blocked Cl⁻ channels as well as K⁺ channels. Such a conclusion is supported by recently demonstrated ability of NPPB to block Ca²⁺ activated K⁺ currents in human leukaemic HL‐60 and glioblastoma GL‐15 cell lines (49).

Cation‐selective channels in Vero cells

Figure 8 illustrates families of membrane currents in representative Vero cells at isotonic conditions. Currents were recorded with the voltage protocol as shown in inset of Fig. 8a; two components of current can be seen in this. One component shows gradual activating current at potentials between −20 and +30 mV (that is, delayed rectifier K⁺ current) and the other a rapidly activating current with noisy oscillation at +40 to +60 mV, similar to voltage‐activated and Ca²⁺‐activated K⁺ current, reported recently by Heubach et al. (50, 51). These results suggest that two types of the channels are co‐expressed in that cell. Figure 8b displays current traces recorded in a typical experiment, showing outward current as a delayed rectifier K⁺ current, whereas Fig. 8c indicates transient outward current recorded in a further cell, similar to transient outward K⁺ current observed in cardiac and neuronal cells (52). Figure 8d demonstrates records of a significant noisy oscillation such as Ca²⁺‐activated K⁺ current in a representative Vero cell and influence of TEA on it. It can be seen that TEA (10 mm) considerably blocked current. This effect was reversible and current could be completely recovered by washing out TEA with fresh solution. Current–voltage relationships constructed with data from these experiments demonstrate extent and voltage dependence of the block; the data prove that K⁺‐channels are responsible for outward current recorded in these cells.

Figure 8.

  Families of membrane currents recorded in Vero cells. The experiments were conducted in whole‐cell configuration at isotonic condition. (a) Two components of outward currents are present, one is gradual activated current such as rectifier K⁺ current, at potentials from −20 to +30 mV, and another with noisy oscillation, such as Ca²⁺‐activated K⁺‐current. (b) Gradual activated current such as rectifier K⁺ current. (c) Transient outward K⁺ current. (d) Influence of TEA at current–voltage relationship of Vero cell. Insets show the records of membrane currents in Vero cell used to build the IV‐curves. Rapidly activating current with noisy oscillation like Ca²⁺‐activated K⁺‐current (Inset – Control) is largely inhibited by 10 mm TEA (Inset – TEA) and could be completely recovered by washing it out by fresh solution (Inset – Wash Out). a, b and c – Membrane current was activated by 300‐ms (a, b and c) or 200‐ms (d) steps between −60 to +60 mV from −40 mV at 0.2 Hz. Current and time scales in (d) are 0.15 nA and 50 ms respectively.

Thus, the electrophysiological study confirms presence of K⁺ and Cl⁻‐channels in plasmatic membranes of Vero cells, as established in our RVD experiments.

Conclusion

In summary, we have found that the ability of swollen Vero cells to restore volume (RVD) was preserved throughout cell growth (Fig. 4a). The importance of Cl⁻ and K⁺ channels in determining RVD has been confirmed by showing that its pharmacological suppression (with four widely known inhibitors) resulted in practically complete inhibition of RVD (4, 5). Rate of initial cell swelling and V _max was influenced by the inhibitors in a cell size‐dependent manner, suggesting that cell growth is accompanied by change in ion channels and aquaporins present in the plasma membrane. Inhibition of these channels decelerated proliferation, arresting cells in G₀/G₁ (Fig. 6). These results are consistent with a growing mass of evidence supporting the hypothesis that harmonized function of ion channels plays a vital role in cell volume regulation and cell cycle progression. Ion channel inhibitors may be considered for incorporation into the design of anti‐cancer treatment strategies.

Supporting information

Fig. S1 Simulation of the resting potential change under influence of hypotonic stress and K‐channel blockers.

Supporting info item