A volume-sensitive Cl− conductance in a mouse neuroblastoma × rat dorsal root ganglion cell line (F11)

A volume-sensitive Cl− conductance in a mouse neuroblastoma × rat dorsal root ganglion cell line (F11)

178 614 (1993) 178-184 © 1993 Elsevier Science Publishers B.V. All rights reserved 0006-8993/93/$06.00 BRES 18919 A volume-sensitive C1- conductanc...

704KB Sizes 0 Downloads 111 Views

178

614 (1993) 178-184 © 1993 Elsevier Science Publishers B.V. All rights reserved 0006-8993/93/$06.00

BRES 18919

A volume-sensitive C1- conductance in a mouse neuroblastoma × rat dorsal root ganglion cell line ( F l l ) Christopher E. Pollard Department of Pharmacology, Fisons plc, Loughborough, Leicestershire ( UK) (Accepted 12 January 1993)

Key words: Volume-sensitivity; CI

conductance; Cell swelling; F l l cell; Hybrid cell line; Patch clamp

Whole cell currents were recorded in F11 cells, a mouse neuroblastoma (NG18TG2) x rat D R G hybrid cell line, using pipette and bath solutions intended to isolate any chloride conductance pathways. W h e n recording with a pipette solution which was 40 mmol- k g - i hypotonic to the bath solution, all cells showed a transient rise in input conductance which peaked 5.3 + 0.4 min after breaking into the cell and returned to the basal state 11.7 +_ 1.2 min later. At the peak of the effect, cell conductance had increased approximately sixfold. The use of short (300 ms) duration voltage steps at the peak of the conductance increase evoked whole-cell currents which were time-independent and had an outwardly rectifying current/voltage relationship. Ion substitution experiments showed that the whole-cell currents were carried by chloride ions and that the anion selectivity sequence of the conductance was I > Br > CI > F > acetate. The stilbene derivative 4,4'-diisothiocyanostilbene-2,2'-disulphonic acid (DIDS) caused a reversible, 51% inhibition of the chloride currents. In cells which had already undergone this transient rise in conductance, whole-cell currents with identical properties could be activated by changing to a very hypotonic bath solution. Coincident with current activation, this manoeuvre caused a visible swelling of the cell. The increase in conductance and the cell swelling were both reversed by returning to the normal bath solution. In contrast, when a very hypotonic pipette solution was used, little or no increase in cell conductance was observed. These data suggest that the F l l cell line possesses a volume-activated chloride conductance which can be controlled by manipulating the relative osmolarity of the bath and pipette solutions.

INTRODUCTION

Cell types from diverse biological systems have been shown to possess ion transport mechanisms which operate to return cell volume to normal after swelling induced by a hypotonic extracelluar medium; so-called regulatory volume decrease (RVD) 6. One of the mechanisms mediating RVD is chloride channel activation following an increase in cell volume, leading to effiux of chloride ions and osmotically obligated water. Although this type of volume-activated chloride conductance has been described in several cell types (intestinal epithelial cells 14, Erlich ascites cells 15, T-lymphocytes 5, airway epithelial cells t9'29, colonic cells 32, liver cells lz, renal cells 27 and ciliary epithelial cells34), there is no clear evidence for its existence in neuronal tissue. In this report, the whole-cell patch clamp technique was utilized to record from F l l cells using solutions intended to isolate any chloride conductance. This cell

line is derived by fusion of rat dorsal root ganglion cells and mouse N18TG2 neuroblastoma cells and has been reported to possess many of the differentiated features of substance-P-containing D R G neurones 21'9. The data, which have already been presented in preliminary form z2, suggest the presence of a volume-activated chloride conductance in these cells. MATERIALS

AND METHODS

Maintenance of F11 cell cultures Growing stocks of early passage F11 cells were obtained from Dr M.C. Fishman (Harvard) and designated passage 1. They were grown in 25-cm 2 tissue flasks in a 'growth medium' consisting of H a m ' s F-12 (Gibco, UK) supplemented with 15-20% fetal calf serum (FCS; Northumbria Biologicals, UK), 50 I U - m l - 1 penicillin, 50 / x g . m l - r streptomycin and 100 # M hypoxanthine, 0.4 /~M aminopterin, 16 /zM thymidine (HAT). Cells were maintained in a humidified atmosphere of 5% CO 2, 95% air at 37°C and fed every other day. For electrophysiological recording, cells were grown on uncoated glass coverslips by taking a 100 /zl aliquot of cells maintained in 'growth medium' (passage 4-67) and adding it to a well of a tissue culture

Correspondence." C.E. Pollard, Department of Pharmacology, Fisons plc, Pharmaceutical Division, Bakewell Road, Loughborough, Leicestershire, L E l l 0RH, UK. Fax: (44) (509) 210450.

179 plate containing the coverslip. The well was then filled with 1 ml of 'differentiation medium' consisting of Ham's F12 supplemented with 1% FCS, HAT, 0.5 mM dibutyryl cyclic AMP, 10/xM 3-isobutyl-1methylxanthine and 50 n g - m l - l / m o u s e salivary gland nerve growth factor (2.5 S) (Sigma). The cells were recorded from 24-48 h after plating. HAT was always present in both types of media to maintain selective pressure for the hybridoma over the neuroblastoma parent. This is based on the fact that the aminopterin content of HAT blocks de novo synthesis of purines and pyrimidines forcing the cells to use alternative nucleotide biosynthetic pathways involving hypoxanthine. The parent neuroblastoma cells lack the enzyme hypoxanthine phosphoribosyltransferasezl and cannot utilize this salvage pathway to survive in the presence of aminopterin, while in the hybridoma this ability is provided by the relevant enzyme originating from the DRG cell.

mmol'kg -1 and contained (in mM): TEACI, 140; HEPES, 10 and ethylene glycol-bis(/3-aminoethyl ether) N,N,N',N'-tetraacetic acid (EGTA), 1 (pH 7.30-7.35 with TEA-OH). On some occasions a hypotonic pipette solution with an osmotic pressure of 210 mmol. kg- l was used, in which the TEACI concentration was decreased to 100 mM. Bath solutions containing DIDS (Aldrich) were prepared by first making a 100 mM stock solution in dimethyl sulphoxide (DMSO) and diluting this 1,000-fold to give a final DIDS concentration of 100 ~M and a final DMSO concentration of 0.1% (14 mM). This concentration of DMSO alone had no effect on the electrophysiological properties of cells (n = 4). Unless stated, all compounds were obtained from commercial sources and were of the highest purity available. RESULTS

Electrophysiology Coverslips were placed in a gravity fed Perspex chamber (volume, ~ 1 ml; flow rate, ~ 10 ml.min -1) mounted on the stage of an inverted, phase-contrast microscope (Leica, UK). Recordings were made at 24-26°C using the conventional whole-cell patch clamp technique 13. Recording pipettes were constructed from borosilicate glass (GC120F15, Clarke Electromedical, UK) using a Flaming-Brown electrode puller (P87, Sutter Instruments, USA). After fire polishing using a microforge (MF-83, Narishige Scientific Instruments, Japan) they had resistances of 3-5 MI2 and were positioned using a hydraulic manipulator (MW-3, Narishige Scientific Instruments, Japan). Whole-cell currents were amplified using an Axopatch-lD patch-clamp system (Axon Instruments, USA). A Compaq 386/25 computer (Compaq Computers, USA) running pClamp software (Axon Instruments, USA) was used to control voltage clamp commands and record subsequent current responses which had been low pass filtered at 20 kHz using a 4-pole Bessel filter. Although it was always possible to compensate for around 80% of the series resistance (5-17 MI2), the voltage error resulting from the residual series resistance was taken into account when constructing current-voltage (1/V) relationships. Junction potentials were corrected for using a flowing 3 M KCI electrode 16 and the appropriate corrections made to the data. The current reversal potential (Erev) was obtained by fitting a third-order polynomial to the rectifying I / V plots using least squares regression analysis and the same fit was used to derive chord conductance values between Erev and +50 mV. Anion-permeability ratios for the chloride conductance reported were calculated from the value of Erev following ion substitution, using the Hodgkin-Katz modification of the Goldman equation as follows:

ex Pc,

[a ]o10 E'-/59'- [Cl- ], [x],-[X]o.10Er~/59.'

Development and decay o f whole-cell currents Fig. 1 A s h o w s a t y p i c a l r e c o r d i n g o f w h o l e - c e l l c u r r e n t s in a n F l l

cell u s i n g a p i p e t t e s o l u t i o n w i t h a n

o s m o t i c p r e s s u r e o f 290 m m o l • k g - 1 a n d a b a t h s o l u t i o n w i t h a n o s m o t i c p r e s s u r e o f 330 m m o l . k g - t .

In

the rest of the text, e x p e r i m e n t s using t h e s e solutions will b e d e s c r i b e d as b e i n g p e r f o r m e d

under standard

conditions. Two min following membrane

r u p t u r e to establish

t h e w h o l e - c e l l c o n f i g u r a t i o n , c u r r e n t r e s p o n s e s to volt-

A

04hA L ~

4

,

,,

:j

k

-

" ~ --~

~ ' - - " -~

O8nA

B

_

_

100 ms I

ii c

i

2 c 1

where P is permeability; [CI] and [X] are the molar concentrations of chloride and the replacement anion, respectively, i and o denote the inside and outside of the cell, respectively and Erev is the reversal potential in inV. Results are presented as means±S.E.M. followed by the number of observations (n).

Solutions and chemicals The osmotic pressure of solutions was measured on the basis of freezing point depression (Osmomat 030, Gonotec, UK). The standard bath solution had an osmotic pressure of 330 mmol-kg-I and contained (in raM): tetraethylammonium chloride (TEACI), 140; N-2-hydroxyethylpiperazine-N'-2-ethanesulphonicacid (HEPES), 10; glucose, 20; NiC12.6H20 , 1; MgSO4.7H20, 2 and CaCI 2 2 (pH 7.30-7.35 with TEA-OH). In some experiments this solution was exchanged for hypotonic bath solutions with osmotic pressures of 196 or 253 mmol.kg-I in which the concentration of TEACI was decreased to 70 mM or 100 raM, respectively. In experiments assessing anion selectivity, all the TEACI in the standard bath solution was substituted with TEAX, where X = Br, I, F or acetate. The standard pipette solution had an osmotic pressure of 290

¢_

-1 40(

-50 0 50 Voltage (nA)

I(D0

Fig. 1. Properties of whole-cell currents in an F11 cell using a pipette solution 40 mmol.kg -1 hypotonic to the bath solution. A: the development and decay of whole-cell currents resulting from 1.2-s voltage pulses +40 mV and - 4 0 mV from 0 mV. B: current responses recorded immediately after the whole-cell configuration was established (left), at the peak of the current development (middle) and after a return to the initial current level (right). Each set of records was evoked by a series of 300 ms voltage pulses ± 90 mV from 0 mV. C shows the I / V relationships of the currents illustrated in B; e, immediately after the whole-cell configuration was established, D, at the peak of the current development; ,x, after a return to the initial current level. Pipette and bath solution osmolarities were 290 mmol. kg- 1 and 330 mmol. kg- 1, respectively.

180 age pulses +_40 mV from 0 mV are small, indicating a relatively low input conductance. However, the cell then undergoes a transient rise in input conductance as demonstrated by the gradual increase and then decrease in whole-cell current. The rate at which these changes took place was relatively slow; it took 6 min to reach the maximum whole-cell current amplitude and 16 rain for current levels to return to a stable level. In order to investigate these whole-cell currents in more detail, the voltage protocol used to generate Fig. 1A was interrupted at key points and substituted for one in which short (300 ms) voltage pulses were used to generate a full I / V relationship over the voltage range +_90 inV. From the resulting responses (Fig. 1B,C) it is apparent that in the basal state and when the current development has peaked, the whole-cell currents are not time dependent and that the I / V relationship derived from the data shows outward rectification. Experiments of the sort shown in Fig. 1 were repeated to quantify the basic characteristics of this effect. The time to reach maximum conductance was 5.3 + 0.4 min (n = 28) and from this point, the return to a stable conductance level took 11.7 _+ 1.2 min (n = 28). The chord conductance of cells, measured 50 mV positive to E .... increased from 4.1 + 0.8 nS (n = 22) around 2 min after breaking into the cell, to 26.1 _+ 2.7 nS (n = 31) at the point of maximum conductance and decreased back to 4.1 _+ 0.7 nS (n = 26) when the input conductance re-stabilized. However, the conductance value quoted soon after breaking in to the cell is probably an over-estimate, since during the time lag between membrane rupture and adjustment of the patch clamp amplifier prior to recording an I / V relationship, the increase in conductance may already have begun. At the point of maximum conductance, Erev was 0.7 + 0.3 mV (n = 20).

Ionic basis for whole-cell currents Fast action potentials have been recorded from F l l ceils, implying the presence of time-dependent sodium and potassium currents 2a and two types of calcium current have also been recorded 3. The time-independent nature of the whole-cell currents argues against any contribution from these sources and the composition of bath and pipette solutions will limit current flow through these or any time-independent current pathways permeable to sodium, potassium or calcium (1 mM N i - blocks both transient and sustained calcium currents in these cells3). Since these observations clearly imply that the whole-cell currents are carried by chloride ions, the effect of substituting all the TEAC1 in the bath solution with TEA-acetate was tested at the peak of the conductance increase (Fig. 2A). Under these conditions, Erev was 26.4 + 2.4 mV (n = 11), indicating that the permeability of the conductance to acetate, relative to chloride (eacetate/ecl), is around 0.32. In addition, it was found that the chloride-channel blocker DIDS (100/zM), applied at the peak of the conductance increase, caused a reversible decrease in whole-cell current (Fig. 3). Since, as shown in Fig. 1A, whole-cell current steadily declines after reaching a maximum, the effect of DIDS was always superimposed on a falling baseline. This made the magnitude of the DIDS effect difficult to quantify. However, to get an approximation, a line was drawn between the the peak of the + 40 mV current response immediately before addition of DIDS and that after washout. The whole-cell current at + 40 mV at the maximum effect of DIDS was then compared with the extrapolated current level at the same time point. On this basis, DIDS was found to inhibit the whole-cell current at +40 mV by 51 + 2 % ( n = 6 ) . This effect and the change in Erev following substitution of bath chloride

B

A

2 ¸

3 ¸

2 ¸ c

c

I

L

0

J


-]~oo

-5o

o

5o

Voltage (mY)

~6o

-1,

-~oo

-5o

o Voltage

so

~ob

(mY)

Fig. 2. Anion selectivity of the F11 cell chloride conductance. A: I / V relationships following the total replacement of bath TEACI with TEA1 (o) or TEA-acetate ( o ) . B: I / V relationships following the total replacement of bath TEACI with T E A B r (e) or T E A F (O). Each I~ V relationship came from a different cell and was recorded at the peak of the current activation. Pipette and bath solution osmolarities were 290 mmol. k g - 1 and 330 mmol. kg - 1, respectively.

181

2oA I

li ~

C

_

2nAI lOOms

~

6, < c 2.

u _ 2. -4,

-loo -5o

o

5o

760

Voltage (mV) Fig. 3. Effect of chloride channel blocker DIDS on F l l cell chloride currents. A: the effect of 100 /zM DIDS (solid bar) on whole-cell currents evoked by 1.2-s voltage pulses + 40 mV and - 40 m V from 0 mV. B shows current responses of the same cell immediately before (left) in the presence of DIDS (right). Each set of records was evoked by a series of 300 ms voltage pulses over the voltage range _+90 mV. C: the I / V relationship of the currents illustrated in B under control conditions (o) and in the presence of DIDS (O). Pipette and bath solution osmolarities were 290 m m o l . k g - L and 330 mmol. kg - 1, respectively.

with acetate, supports the conclusion that the whole-cell currents are carried by chloride.

Selectivity of the chloride conductance In order to compare this chloride conductance with those reported elsewhere, halide selectivity was determined using the bromide, fluoride and iodide salts of TEA. In all experiments of this kind, standard conditions were used until the whole-cell currents evoked by + 40 mV pulses from 0 mV reached a peak, at which point the bath solution was changed to TEA-halide. The value of Erev determined from I / V relationships in the presence of different halides (Fig. 2) was: I, - 8 . 2 + 0.5 mV (n = 7); Br, - 4 . 2 + 0.7 mV (n = 4) and F, 24.4 + 1.0 mV (n = 7). The calculated selectivity of the conductance for these anions, relative to chloride, indicates the following halide selectivity sequence (Px/Po values are shown in brackets): I (1.33)> Br (1.13) > CI (1.00) > F (0.35).

Basis for current development and decay There is a striking similarity between the time course of chloride current development and decay in F l l cells and that reported in, for example, human airway epithelial cells m. In airway cells it was found that current

development could be prevented by decreasing the osmolarity of the pipette solution and activated by changing to a relatively hypotonic bath solution. With this analogy in mind, eight cells were recorded from using a pipette solution with an osmotic pressure of 210 m m o l - k g - 1 , thus making the pipette solution 120 m m o l . k g -1 hypotonic to the bath solution. Under these circumstances, there was no conductance increase in three cells and in the five cells where there was a conductance increase, the effect was relatively small (see Fig. 5) and short-lived; peak conductance occurred after 1.3 + 0.1 min and from this point the recovery took 3.3 + 0.3 min. Furthermore, in three of the cells which displayed this brief response, the input conductance was allowed to decay to normal and the bath solution was then changed to one with an osmotic pressure of 253 mmol • kg-1, thus imposing an osmotic gradient similar to that which exists under standard conditions. This manoeuvre did not change the input conductance of any of the cells. The effect of reversing the osmotic gradient (i.e., bath hypotonic to pipette) was then observed in cells which had already undergone the conductance develo p m e n t / d e c a y sequence under standard conditions. This was achieved by changing to a bath solution which had an osmotic pressure of 196 mmol. k g - L As illustrated in Fig. 4A, this caused an initial decrease, followed by a gradual rise in conductance. However, it is important to note that the size of the conductance increase in the hypotonic bath solution will be underestimated because of the much reduced concentration of the permeating anion and that the true extent of the effect is only realised when the standard bath solution is re-introduced. The whole cell currents gradually decayed on return to the standard bath solution (Fig. 4A). The whole-cell currents activated by this hypotonic bath solution were DIDS-sensitive (47 + 4% decrease (n = 3)) and the form and I / V relationship of the currents (Fig. 4B,C) were identical to those activated under standard conditions (cf. Fig. 1B,C). Using the experimental design illustrated in Fig. 4, it was possible to quantify the peak conductance evoked when using standard conditions and then repeat this measurement when the hypotonic bath solution had re-activated the chloride conductance in the same cell. To make these two measurements comparable, both the peak conductance recorded under standard conditions and that evoked by the hypotonic bath solution were referenced to the conductance of the cell immediately prior to introduction of the hypotonic bath solution. The results of this analysis are shown for seven cells in Fig. 5, which demonstrates that the conductance increase evoked when the bath solution is 94

182

04 n~'l__ 2rain

ft ,' '! ....

,

II B

C

/

3 ~c 2

mmol • kg- l hypotonic to the pipette solution, is significantly greater than under standard conditions ( P = 0.0004; paired, two-tailed t-test). The graph also indicates the reduced level of conductance activation in eight different cells using the very hypotonic pipette solution (see above). In addition to the conductance increase induced by the hypotonic bath solution, there was also a coincident swelling of the cell and a return of the cell to its normal size on re-introduction of the standard bath solution.

o 2 nA

L~

DISCUSSION

100 ms

-50 0 Voltage

50 (mV)

~)0

Fig. 4. Re-activation of chloride currents by a hypotonic bath solution in an F l l cell. A shows whole-cell chloride currents evoked by 1.2-s voltage pulses + 4 0 mV and - 4 0 mV from 0 mV. After the initial development and decay sequence using a pipette solution 40 m m o l . k g - ~ hypotonic to the bath solution (pipette and bath solution osmolarities were 290 m m o l . k g - 1 and 330 mmol . k g - i , respectively), the bath was transiently perfused with a solution 94 m m o l . k g - t hypotonic to the pipette solution (solid bar). B illustrates the wholecell currents activated by the hypotonic bath solution, recorded 2 min after a return to the standard bath solution. The current responses were evoked by a series of 300 ms voltage pulses over the voltage range + 90 mV. C: the I / V relationship of the currents illustrated in B.

2o] t

C O 10" XJ U

"6

O,

120

40

-94

A Osmolar'lty (bath-pipette) ( m m o l kg -1)

Fig. 5. The effect of varying the difference between pipette and bath solution osmolarity on the relative increase in input conductance in F l l cells. The data shown for A osmolarity (bath-pipette) 40 and - 9 4 m m o l . k g -1 were collected from the same cells (n = 7) using the experimental design exemplified by Fig. 4A. In these experiments, the change in conductance under each condition was calculated relative to the conductance of the cell immediately before changing to the hypotonic bath solution. In order for the two conductance values to be comparable, the conductance increase evoked by the hypotonic bath solution was actually measured when the standard bath solution was re-introduced. Eight different cells were used to show the change in conductance with A osmolarity (bath-pipette) 120 m m o l . k g 1.

The main conclusion of this paper is that F l l cells possess a chloride conductance activated by cell swelling. Whether this property is derived from the D R G cell parent or from the N18TG2 neuroblastoma cell parent is unclear; anion channels have been linked with RVD in N1E115 neuroblastoma cells 8 but their activation appeared to be delayed relative to cell swelling and given their large conductance, a distinct increase in current noise levels would have been expected if such channels had been active in F l l cell recordings.

Basis for actication of chloride conductance The most obvious explanation for the development and decay of the chloride conductance under standard conditions (e.g., Fig. 1A) is related to the fact that the total osmotic pressure of the cytoplasm is the sum of the contribution from the small ionic solutes and that from the relatively large macromolecules ~s. Following membrane rupture to achieve the whole-cell recording configuration, there will be relatively rapid mixing of the very mobile, small ionic solutes such that the intracellular small solute osmolarity will rapidly approximate to that of the pipette solution. The larger macromolecules will take time to dialyze from the cell so there will be a time period during which cellular osmolarity will equal the sum of the contribution from the macromolecules and that from the ions in the pipette solution. Since the osmolarity of the pipette solution is likely to be significantly greater than the osmolarity of the small ionic solutes inside the cell prior to 'break-in', the total osmolarity will transiently exceed that of the bath solution, so water will enter the cell and cause swelling. As the macromolecules leave the cell, the osmotic gradient will gradually approximate t o that expected from the known osmolarity of pipette and bath solutions (i.e., pipette 40 mmol. kg-1 hypotonic to bath) and the cell will loose water and shrink.

183 This interpretation is supported by the fact that if the osmolarity of the pipette solution is made very hypotonic (210 m m o l . kg -1) relative to the bath solution, then following 'break-in' there is either a very brief increase in chloride conductance or no increase at all. Presumably, under these conditions, even when the macromolecular component of the cell contents have yet to leave the cell, the total intracellular osmolarity never, or only briefly, exceeds that of the bath solution because of the much reduced osmotic contribution from the pipette solution. Furthermore, after several minutes recording with this very hypotonic pipette solution, during which time the large molecules will have had time to dialyse from the cell, introduction of a bath solution with an osmolarity of 253 mmol • k g does not produce conductance activation, despite the fact that the resulting osmotic gradient is similar to that under standard conditions, which always evoked a conductance increase. While visual inspection of a cell undergoing a transient increase in chloride conductance under standard conditions did not reveal overt signs of cell swelling, there was unmistakable cell swelling, and a much larger increase in chloride conductance, when a very hypotonic bath solution was introduced (see Figs. 4 and 5). The characteristics of the chloride conductance activated by a hypotonic bath solution were indistinguishable from those recorded under standard conditions, suggesting a continuum of chloride conductance activation dependent on the extent of cell swelling. Although there have been many examples of calcium-activated chloride conductances since their initial observation in spinal neurones 2° and rises in intracellular calcium have been reported during cell swelling 7'25, it is unlikely that the chloride conductance in F l l cells is activated by calcium given the presence of E G T A in the pipette solution.

Comparison with other volume-activated chloride conductances The slow time course of conductance activation and decay in F l l ceils and the ability to control the extent of activation by manipulation of bath and pipette solution osmolarity are characteristic of volume-activated chloride conductances in several other cell types 5'7'19'27'32'34. Where tested, the conductance has also been found to be inhibited by stilbene derivatives v'14'27 and the halide selectivity reported here (1 > Br > CI > F) corresponds with sequences determined using either patch clamp techniques 32 or less direct means 25. This sequence does not, however, coincide with the size of these halides or their mobility in aqueous solution, suggesting factors such as interaction

of the permeating ions with sites inside the chloride channels are important in determining selectivity. Indeed, the selectivity sequence corresponds to sequence 1 of Wright and Diamond 33, which may indicate a relatively weak electrostatic interaction of the permeating anions with cationic sites inside the channel, with the dehydration energy largely determining the change in free energy on anion binding 33. Although the outwardly rectifying I / V relationship of volume-activated F l l cell chloride currents is a common feature of this category of current, their timeindependent nature contrasts with that of the volumeactivated currents in other cell types, all of which display marked time dependence 19'27'29'32'34.

Comparison with spontaneously active chloride channels in neurones There are several recent reports of chloride conductances in nerve cells not activated by extracelluar ligands or intracellular calcium e'4'8'1°'17'23'24'26'28. However, most of the data are from single-channel recording, making it difficult to relate their characteristics to the macroscopic currents reported here. In view of this, the only valid comparison relates to the fact that none of the single-channel data indicate channel activation or inactivation with time. Such channel activity would be manifested as a time-independent current in a wholecell recording. Also, the only detailed single-channel study revealed a halide selectivity sequence identical to the one described here 1°. The macroscopic chloride currents recorded in squid axons 17 are analogous to those in F l l ceils; they are time independent, have an outwardly rectifying I / V relationship and are blocked by stilbene derivatives. Although there is evidence of a role for second messengers in activation of volume-sensitive conductances 7, the most obvious mechanism for channel activation is membrane stretch resulting from an increase in cell volume. With this in mind and the possible membrane distorting effects of the patch clamp technique, it is worth considering the possibility that recordings of spontaneously active chloride channels in neurones could result from the artificial activation channels sensitive to changes in cell volume. Possible functions of a volume-activated chloride conductance In order to isolate the chloride conductance in F l l cells, it was necessary to use non-physiological recording conditions. Despite this, it is apparent from its current/voltage relationship that the chloride conductance would be active at the resting membrane potentials which exist in neurones.

184 By analogy with epithelial cells, the obvious role for the conductance described here is that of a route for outward ion movement during RVD, a process which has been described in N 1 E l l 5 neuroblastoma cells 8. Although this pre-supposes an outward electrochemical gradient for chloride ions under physiological conditions, such a gradient has been reported in nerve cellsL~L While it is difficult to envisage circumstances when a neurone would be exposed to a hypotonic extraceUuar solution in vivo, it is well established that neurones do swell under certain conditions even in the absence of an osmotic gradient 3°'3~, implying a need for processes mediating RVD. The opening of chloride channels in an osmotically compromised cell would also have the effect of lowering excitability by decreasing input resistance. F l l cells are undergoing a differentiation process which involves dendrite outgrowth, the mechanical effects of which could result in membrane stretch and local chloride conductance activation. With this in mind and the fact that the F l l cells are partly derived from a neuronal carcinoma cell line, another possible function may be related to cell growth. Acknowledgements. 1 thank Dr. Mark Fishman and Margaret Boulos (Harvard) for supplying the F11 cell line, Dr. J. Schmidt and Miss Mary Leeson for assistance with tissue-culture techniques and Dr. Dale Jackson for his encouragement.

REFERENCES 1 Adams, P.R. and Brown, D.A., Actions of y-aminobutyric acid on sympathetic ganglion cells, J. Physiol., 250 (1975) 85-120. 2 Blatz, A.L., Properties of single fast chloride channels from rat cerebral cortex neurones, J. Physiol., 441 (1991) 1-21. 3 Boland, L.M. and Dingledine, R., Multiple components of both transient and sustained barium currents in a rat dorsal root ganglion cell line, J. Physiol., 420 (1990) 223-245. 4 Bolotina, V., Boreck~, J., Vlachovfi, V., Baudy~ovfi, M. and Vysko~il, F., Voltage-dependent chloride channels with several substates in excised patches from mouse neuroblastoma cells, Neurosci. Lett., 77 (1987) 298-302. 5 Cahalan, M.D. and Lewis, R.S., Role of potassium and chloride channels in volume regulation by T-lymphocytes. In R.B. Gunn and J.C. Parker (Eds.), Cell Physiology of Blood, Rockefeller University Press, New York, 1988, pp. 282-301. 6 Chamberlin, M.E. and Strange, K., Anisosmotic volume regulation: a comparative view, Am. J. Physiol., 257 (1989) C159-C173. 7 Doroshenko, P. and Neher, E., Volume-sensitive chloride conductance in bovine chromaffin cell membrane, J. Physiol., 449 (1992) 197-218. 8 Falke, LC. and Misler, S., Activity of ion channels during volume regulation by clonal NIE115 neuroblastoma cells, Proc. Natl. Acad. Sci. USA, 86 (1989) 3919-3923. 9 Francel, P.C., Harris, K., Smith, M., Fishman, M.C., Dawson, G. and Miller, R.J., Neurochemical characteristics of a novel dorsal root ganglion x neuroblastoma hybrid cells line, F l l , J. Neurochem., 48 (1987) 1624-1631. 10 Franciolini, F. and Nonner, W., Anion and cation permeability of a chloride channel in rat hippocampal neurones, J. Gen. Physiol., 90 (1987) 453-478. ll Gallagher, J.P., Higashi, H. and Nishi, S., Characterization and

ionic basis of GABA-induced depolarizations recorded in citro from cat primary afferent neurones, J. Physiol., 275 (1978) 263282. 12 Haddad, P., Beck, J.S., Boyer, J.L. and Graf, J., Role of chloride ions in liver cell volume regulation, Am. Z Physiol., 261 (1991) G340-G348. 13 Hamill, O.P., Marty, A., Neher, E., Sakmann, B. and Sigworth, F.J., Improved patch-clamp technique for high-resolution current recording from cells and cell-free membrane patches, Pfliigers Arch., 391 (1981) 85-100. 14 Hazama, A. and Okada, Y., Ca 2+ sensitivity of volume-regulatory K + and C1- channels in cultured human epithelial cells, Z Physiol., 402 (1988) 687-702. 15 Hudson, R.L. and Schultz, S.G., Sodium-coupled glycine uptake by Ehrlich ascites tumor cells results in an increase in cell volume and plasma membrane channel activities, Proc. Natl. Acad. Sci. USA, 85 (1988) 279-283. 16 Hughes, D., McBurney, R.N., Smith, S.M. and Zorec, R., Caesium ions activate chloride channels in rat cultured spinal neurones, J. Physiol., 392 (1987) 231-251. 17 lnoue, I., Voltage-dependent chloride conductance of the squid axon membrane and its blockade by some stilbene derivatives, J. Gen. Physiol., 85 (1985) 519-537. 18 Macknight, A.D.C., Principles of cell volume regulation, Renal Physiol. Biochem., 3 (1988) 114-141. 19 McCann, J.D., Li, M. and Welsh, M.J., Identification and regulation of whole-cell chloride currents in airway epithelium, J. Gen. Physiol., 94 (1989) 1015-1036. 20 Owen, D.G., Segal, M. and Barker, J.L., A calcium-dependent chloride conductance in cultured mouse spinal neurones, Nature 311 (1984) 567-570. 21 Platika, D., Boulos, M.H., Baizer, L. and Fishman, M.C., Neuronal traits of clonal cell lines derived by fusion of dorsal root ganglia neurons with neuroblastoma cells, Proc. Natl. Acad. Sci. USA, 82 (1985) 3499-3503. 22 Pollard, C.E. and Jackson, D.M., Volume-sensitive chloride currents in a rat dorsal root ganglion (DRG) cell line (Fll), J. Physiol., 459 (1993) 243P. 23 Proebstle, T., Rothmund, J., Rudel, R. and Fromherz, P., Single chloride channels in neurites of cultured leech neurones, J. Physiol., 438 (1991) 327P. 24 Rabasseda, X., Valmier, J., Larmet, Y. and Simonneau, M., Large unit conductance voltage chloride channels are expressed in mouse neural crest cells and embryonic dorsal root ganglion cells, Dec. Brain Res., 51 (1990) 283-286, 25 Rothstein, A. and Bear, C., Cell volume changes and the activity of the chloride conductance path, Ann. N.Y. Acad. Sci., 574 (1989) 294-308. 26 Shukla, S. and Pockett, S., A chloride channel in excised patches from cultured rat hippocampal neurons, Neurosci. Lett. 112 (1990) 229-233. 27 Strange, K., Volume regulatory Cl- loss after Na + pump inhibition in CCT principal cells, Am. J. Physiol. 260 (1991) F225-F234. 28 Strupp, M. and Grafe, P., A chloride channel in rat and human axons, Neurosci. Lett., 133 (1991) 237-240. 29 Solc, C.K. and Wine, J.J., Swelling-induced and depolarizationinduced chloride channels in normal and cystic fibrosis epithelial cells, Am. J. Physiol. 261 (1991) C658-C674. 30 Sykova, E. (1991)., Ionic and volume changes in neuronal microenvironment, Physiol. Res. 40 (1991) 213-222. 31 Tomita, M. and Gotoh, F., Cascade of cell swelling: thermodynamic potential discharge of brain cells after membrane injury, Am. J. Physiol. 262 (1992) H603-H610. 32 Worrel, R.T., Butt, A.G., Cliff, W.H. and Frizzel, R.A., A volume-sensitive chloride conductance in human colonic cell line T84, Am. J. Physiol., 256 (1989) Cl111-Cl119. 33 Wright, E.M. and Diamond, J.M., Anion selectivity in biological systems, Physiol. Rec. 57 (1977) 109-156. 34 Yantorno, R.E., Carre, D.A., Coco-Prados, M., Krupin, T. and Civan, M.M., Whole cell patch clamping of ciliary epithelial cells during anisosmotic swelling, Am. J. Physiol. 262 (1992) C501C509.