Calcium homeostasis in guinea pig type-I vestibular hair cell: possible involvement of an Na+Ca2+ exchanger

III ELSEVIER

IIKIt

Hearing Research 89 (1995) 101-108

Calcium homeostasis in guinea pig type-I vestibular hair cell: possible involvement of an Na+-Ca a÷ exchanger Christian Chabbert a,l,,, Yvan Canitrot b, Alain Sans a, Jacques Lehouelleur

a

a INSERM U.432, Laboratoire de Neurophysiologie Sensorielle et CeUulaire, 34095 Montpellier, France b Laboratoire de Chimie Physique, Universit~ de Perpignan, 66860 Perpignan, France

Received 14 January 1995; revised 23 May 1995; accepted 27 May 1995

Abstract

In type-I vestibular hair cells (VHCs), the mechanisms involved in intracellular calcium homeostasis have not yet been established. In order to investigate the involvement of an Na+-dependent ionic exchanger in the regulation of cytosolic free calcium concentration, we analyzed the effect of the removal of external sodium on the cytosolic concentration of calcium ions ([Ca2+ ]i), sodium ions ([Na + ]i), and protons (pHi). These concentrations were measured in type-I VHCs isolated from guinea pig labyrinth, using Fura-2, sodium benzofuran isophtalate (SBFI), and 1,4 diacetoxy-2,3 dicyanobenzol (ADB) respectively. Complete replacement of Na + in the supeffusion solution with N-methyl-D-glucamine (NMDG +), reversibly increased [Ca2+ ]i by 276 ___89% (n = 46) and decreased [Na + ]i by 23 _ 6% (n = 14). Both responses were prevented by removing external Ca 2÷ or chelating internal Ca 2÷. This suggests the presence of coupled Ca 2+ and Na + transport. The [Ca2÷ ]i increase evoked by Na +-free solution was reduced by about 55 % with the application of amiloride derivatives and was totally abolished in the presence of high [Mg 2+ ]o. No pH i variation was detected during [Na + ]o reduction. In the absence of external K +, the Na +-free solution failed to induce [Ca2+ ]i increase; the readmission of external K + restored the [Ca2÷ ]i response. These results are consistent with a Na+-Ca 2+ exchanger operating in reverse mode. An K + dependence of this exchange is also suggested. Keywords: Ca2+ homeostasis; Vestibule; Na+-Ca 2+ exchange; Guinea pig; Fura-2; Sodium benzofuran isophtalate

1. I n t r o d u c t i o n

In vestibular sensory cells, the hair bundle deflection towards the kinocilium induces the opening of cationselective transmembrane channels, the transduction channels which are situated near the tips of the stereocilia. The large electrochemical gradient across the apical surface of hair cells causes K + and Ca 2+ to enter the cell through the transduction channels (for review see Roberts et al., 1988). The recording of transduction currents in voltage-clamped vestibular hair cells (VHCs) revealed the presence of an adaptive process which induces the closing of the transduction channels upon sustained hair bundle displacement (Eatock et al., 1987). It has recently been proposed that Ca 2+ entering through the transduction channels mediates

* Corresponding author: Laboratory of Sensory Neuroscience, The Rockefeller University, 1230 York Avenue, Box 314, New York, NY 10021-6399. USA. Tel.: (212) 327-7354; Fax: (212) 327-7935. 1 Present address: Howard Hughes Medical Institute, Center for Basic Neuroscience Research, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75235-9117, USA. 0378-5955/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 3 7 8 - 5 9 5 5 ( 9 5 ) 0 0 1 2 6 - 3

this mechanosensory adaptation by acting on myosin-I motors that regulate the tension of the transduction channels' 'gating springs' (Hudspeth and Gillespie, 1994). The recording o f [Ca2+]i variations, in VHCs isolated from chick (Ohmori, 1988) and type-I VHCs isolated from guinea pig (Chabbert et al., 1994), has revealed transitory [Ca2+ ]i increases restricted to the apical part of the cells. These results suggest the presence of a strong Ca 2+ clearance mechanism in the hair bundle of VHCs, which controis [Ca 2+ ]i in the apical part of the cell. Certain properties of afferent synaptic transmission in VHCs are also strongly dependent on the presence of [Ca 2+ ]i-regulatory mechanisms. In VHCs isolated from the bullfrog's sacculus, depolarization induces Ca 2+ entry into the cell through voltage-activated Ca 2+ channels clustered at presynaptic active zones (Issa and Hudspeth, 1994). This Ca 2+ entry, expected to promote neurotransmitter release, also opens Ca2+-activated K + channels closely associated with the voltage-activated Ca 2+ channels (Roberts et al., 1990). The interaction between the Ca 2+ channels and K + channels, together with the presence of cytosolic mobile Ca 2+ buffer (Roberts, 1994), allows the

102

C. Chabbert et al. / H e a r i n g Research 89 (1995) 101-108

cell to undergo membrane potential oscillations and to generate fast presynaptic Ca 2÷ signals. In type-I VHCs, the presence of voltage-dependent Ca 2÷ channels has not been clearly established (Rennie and Ashmore, 1991; Eatock and Hutzler, 1992; Griguer et al., 1993). Nevertheless, because the Ca 2÷ dependence of neurotransmitter release had been widely demonstrated, other mechanisms may mediate [Ca 2÷ ]i elevation in this cell type. Calcium ions are also implicated in the efferent modulation of hair cell activity. Application of acetylcholine, a putative neurotransmitter released by the efferent fibers, has been reported to increase [Ca2+] i and hyperpolarize hair cells by opening Ca2+-activated K ÷ channels (Shigemoto and Ohmori, 1991; Doi and Ohmori, 1993). The cellular mechanisms involved in Ca 2÷ homeostasis play a key role in these sensory cells, because adaptation, afferent synaptic transmission, and efferent modulation all require local Ca 2÷ entry and fast [Ca 2÷ ]i regulation. It has been suggested that the presence of mobile cytoplasmic Ca 2÷ buffers, Ca 2+ pumps, or exchangers, is required to limit the spread of Ca 2÷ in the cytosol and the crosstalk between the different signalling pathways (for review see Lenzi and Roberts, 1994). In VHCs, however, the cellular mechanisms involved in Ca 2÷ homeostasis have not yet been established. In guinea pig outer hair cells (Ikeda et al., 1992), the presence of N a + - C a 2+ exchangers was demonstrated using microfluorimetry techniques. Such an Na+-dependent membrane exchanger has been studied in numerous cell types (for review see Allen et al., 1989). It has been clearly established that at a high external Na + concentration ([Na+]o) and negative membrane potential, the activation of a N a + - C a 2÷ exchanger mediates Ca 2÷ efflux and Na ÷ influx. The inversion of the Na ÷ electrochemical gradient drives the N a + - C a 2+ exchanger in reversed mode and induces Ca 2÷ influx and Na ÷ efflux. We investigated whether this type of membrane exchanger is present in type-I VHCs isolated from guinea pig, by analyzing the effect of [Na+]o reduction on [Ca 2÷ ]i, [Na+]i, and pH i using respectively Fura-2, sodium benzofuran isophtalate (SBFI), and 1,4 diacetoxy-2,3 dicyanobenzol (ADB). We report coupled [Ca2+] i and [Na÷]i variations triggered by a reduction of [Na+]o, independent of cytosolic acidification, and sensitive to ionic exchanger inhibitors. These results are consistent with the presence of a N a + - C a 2÷ exchanger in type-I VHCs. In regard to the K ÷ requirement for the Ca 2+ response to external Na + removal, the K+-dependence of the N a + - C a 2÷ exchange is also discussed.

2. Materials and methods

2.1. Solutions Experiments were performed in standard Hank's balanced-salt solution (HBSS) containing 140 mM NaC1, 6

mM KC1, 0.9 mM MgC12, 1.3 mM CaCI2, 5 mM glucose, 10 mM HEPES. An Na+-free solution was prepared by replacing Na + by N-methyl-D-glucamine (NMDG +, Sigma, St. Louis, MO). CaZ+-free and K+-free solutions were made by omission of respectively Ca 2+ and K + from the Na+-free solution. For the experiments with various Na + concentrations, Na + was replaced with equimolar amounts of NMDG +. Amiloride hydrochloride (100 /xM, Sigma) and amiloride 5-(N,N-dimethyl) hydrochloride (100 ~M, Sigma) were used as inhibitors of the N a + - C a 2+ exchanger. All the physiological solutions were adjusted to pH 7.3 with NaOH (or CsOH in the Na+-free solutions) and to an osmolarity of 300 m O s m / k g with sucrose. Superfusion allowed us to apply standard HBSS or the test solutions close to the studied cells using a peristaltic pump. In order to avoid mechanical stimulation of tested cells, solutions were applied at a low rate (0.5 m l / m i n ) from the bottom of the cells. In order to avoid osmotic changes due to evaporation, the recording chamber was continuously perfused with standard HBSS.

2.2. Isolation of type-I vestibular hair cells Vestibular hair cells were enzymatically dissociated from the macula utriculi and crista ampullaris of young pigmented guinea pigs (150-200 g) using a modified form of a previously described procedure (Valat et al., 1989). The animals were anesthetized deeply with ether and rapidly decapitated. In order to weaken the junctions between the cells, sensory epithelia were bathed for 5 rain in low-calcium HBSS, containing 0.1 mM CaC12. Sensory epithelia were incubated for 3 min in 0.5 m g / m l protease type XXV (Sigma) in intermediate-Ca 2+ HBSS containing 0.5 mM CaC12, then for 5 min in 0.25 m g / m l collagenase solution (Worthington, Interchim, Montlu~on, France). Epithelia were rinsed in standard HBSS and then mechanically dissociated. Following this procedure, we isolated 50 healthy isolated type-I cells. These cells were easily distinguished from type-II cells due to their characteristic shape (Scarfone et al., 1991). Some of these isolated cells which lost their hair bundle during the dissociation protocol responded in similar manner according to the experimental protocol. Cell isolation and experiments were performed at constant temperature (28-30°C).

2.3. Dye loading Type-I VHCs were loaded with the membrane-permeant acetoxymethyl ester form of either the Ca2+-indicator Fura-2 (Molecular Probes, Eugene, OR) or the Na+-indica tor benzofuran isophtalate (SBFI, Sigma). The indicators were dissolved in pluronic acid-DMSO solution (DMSO, dimethyl sulphoxide, 0.05%; pluronic F-127, 0.04%; Sigma). The cells were loaded using separate incubation protocols: 2 /xM Fura-2-AM for 25 min at 37°C or 5 /xM SBFI-AM for 50 min at 37°C. During the incubation

C. Chabbert et al. /Hearing Research 89 (1995) 101-108 periods, the dye-loaded cells settled onto a glass coverslip treated with concanavalin A (1 m g / m l , Sigma). The loaded cells were subsequently rinsed several times in standard HBSS and mounted on a plexiglas adapter compatible with the microscope stage.

A

103

140

[Na+]

140

--1

(raM) o

o

t

' 2

' 3

1.5

2.4. Fluorescence measurements f o r [Ca 2 +]i and [Na +]i

0.5

1

4

5

6

t i m e (rain)

The instrumentation used for microfluorimetric analysis, which was the same for both [Ca2+] i and [Na+]i measurements, has been described in detail (Chabbert et al., 1994). Cell body fluorescence-intensity ratios were recorded at 1280 ms intervals and are expressed as mean + SEM where ( n ) indicates the number of experiments. In all the figures, microfluorimetry traces are single records. Because of the uncertainty in estimating absolute [Ca 2+ ]i, we chose to present fluorescence ratio variations rather than absolute [Ca2+] i variations. Fura-2 fluorescence was calibrated using an in vitro method (Grynkiewicz et al., 1985). The fluorescence of 30 /xM Fura-2 pentapotassium salt (Calbiochem, La Jolla, CA) was determined in Ca z+free ( E G T A - K O H 10 mM; R m i n = 0 . 3 5 ) and Ca 2÷saturated (CaC12 10 mM; Rmax = 6.05) calibration solutions, with K = 7.75. K o. Assuming an K o of 224 nM, the [Ca 2÷ ]i can be estimated by: [ Ca2+ ]i = g ( R

-

Rmin ) / / ( Rmax - R)

Relative [Na+]i was estimated using an in situ method (Harootunian et al., 1989). Fluorescence intensity ratios were calibrated by exposing the isolated SBFI-loaded VHCs for 10 min to various mixtures of high-[Na ÷] and

A 0

1.2 [ /

2 FM Grarnicidin 5 pM Monensin 5 IJM Nigericin

11 0.8

o It.

e

0.6t

I

] 0

I 20

I 50

I J 100 140

{Na+]°(rnMl

1.2

1.0 ii 0 0.8 co

u. o.6

0

B

140

[Na+]

140

I

(raM) o 0 cO ¢0 LL

1

U.

0.9

o

;

o

r

~

3

~

~

t i m e (min)

Fig. 2. Effect of complete removal of external Na+ ions on [ C a 2+ ]i and [Na+ ]i' A: superfusionof a Fura-2-1oadedtype-I VHC with a solution in which Na+ have been completely replaced by NMDG+ induced a F340/F380 ratio increase from 0.65 to 1.43 correspondingto a [Ca2+ ]i increase of 313% (from 98 to 405 nM in this cell) in about 30 s. The increase in [ C a 2+ ]i was followed by a progressive decrease until the readmission of standard HBSS (140 mM Na + ). At this point, the cells recovered their initial resting [ C a 2+ ]i in less than 1 rain. A representative response of the 46 tested cells is shown. B: superfusionof a SBFI-loaded cell with Na+-free solution for 2 min induced a F340/F380 ratio decrease from 0.97 to 0.90 corresponding to a 27% decrease in [Na+ ]i (from 89 to 65 mM in this cell) in about 1 min. Upon readmissionof Na+-containingsolution, the cell recovered its initial [Na+ ]i in about 2 min. Similar results were obtained in 14 cells.

-[K ÷ ] solutions, with [Na ÷ ] + [K ÷ ] = 140 mM and 0, 20, 50, 100 and 140 mM [Na+]. Cells were pretreated for 10 min with the ionophores gramicidin (2 /xM), monensin (5 /xM), and nigericin (5 /xM); the ionophores were then washed away with 0 m M Na ÷ solution (Fig. 1A). The standard 140 m M Na ÷ solution contained 110 mM Na ÷gluconate, 30 m M NaC1, 1.3 m M CaCI2, 0.9 mM MgC12, 10 m M HEPES. The F 3 4 0 / F 3 8 0 ratio (means 4-SEM) obtained from 15 type-I VHCs were plotted against [Na ÷ ]o, and [Na+]i was estimated by linear regression analysis according to the equation y = (2.62x + 736)10 -3 with a correlation coefficient of 0.978 (Fig. 1B).

2.5. Microspectrofluorimetric measurement o f p H i 20

50

100

1~-0

[Na+L(mM)

Fig. 1. Calibrationof [Na+ ]i using SBFI in isolated type-I vestibularhair cells. A: illustration of the time course of a complete experiment of [Na+ ]i calibration in an isolated type-I vestibular hair cell. F340/F380 ratio variations were recorded at 1 min intervals in the cell body of an SBFI-loaded cell. The cell was clamped at 0, 20, 50, 100 and 140 mM [Na+ ]i after a 10 min treatmentwith 2 /xM gramicidin, 5 p.M monensin and 5 #M nigericin (thick line). B: plot of F340/F380 ratio (means+ SEM) against [Na÷ ]o obtained from 15 type-I VHCs. [Na+ ]i was estimated by linear-regressionanalysis according to the equation y = (2.62- x + 736)10-3 (coefficientof correlation, r = 0.978).

For estimation of intracellular pH (pH i), isolated VHCs were loaded with 20 /zM A D B (1,4 diacetoxy-2,3 dicyanobenzol; Faesel, Frankfurt, Germany) for 5 min at room temperature. The colorless ADB penetrated into cells and was enzymatically hydrolyzed into colored dicyanohydroquinone (DCH). The fluorescence spectrum of this resulting component was then recorded. The microspectrofluorimeter consisted of an inverted microscope (Leitz) connected to an optical multichannel analyser (OMA, Princeton Applied Research) equipped with a silicon-in-

C. Chabbert et al. / Hearing Research 89 (1995) 101-108

104

tensified-target (SIT) detector. The excitation wavelength was 358 nm and the recording time of each spectrum was 12 s. The complex fluorescence spectrum of DCH was resolved into its components (Lahmy et al., 1989) using a library of characteristic fluorescence spectra for the semiprotonated form and the deprotonated form previously recorded. The fit between the experimental fluorescence spectrum and the calculated spectrum was monitored by a method previously described in detail (Salmon et al., 1988).

A

[Na*]

(raM) o

S

[-L£_r

140

1.5L

'] Ol amiloride

I0[

1

LL 0.5

0

6

12

18

24

30

36

42

time (min)

B

[Mg 2+]

[

0.9

(raM) o

I 4 I

o I

[Na+]o

I

o9 o

I

3. Results (mM)~CO ~'OO3 1.5[ 1

After a resting period of 15 min from the end of the incubation, measurement of basal [Ca 2+ ]i and [Na+]i were performed on isolated type-I VHCs. In Fura-2-1oaded cells the mean fluorescence intensity ratio was 0.71 -t- 0.13 (n = 62), corresponding to [Ca2+ ]i of 117 ___46 nM (n = 62). In SBFI-loaded cells the mean fluorescence intensity ratio was 0.96___0.02 (n = 27), corresponding to [Na+]i of 85 ___11 mM (n = 27). Resting [Ca2+] i and [Na+]i values

A

[Ca 2+]

[

(rnM) o [Na +] (raM) o

ior

I--I

[ca'+]

o I

4

L

(raM) o [Na+] (raM) o [--I 0 I

Iol--

8 12 time (min)

140

4

~

1'0

1~

1~

[Mg2+]

[

09

I

[Na+]o

~

o I

I

(mM) o

(raM)

'

I

0.9

o

r

I o

[

I

I

0

8 12 time (rain)

2

4

6

8

10

12

, 14

time (rain)

16

I O I

~. 0.9

0

C

0

o ~t e3

B

~

time (min)

~

o co 0o

0.5

~

~ 0.9

1.3 140

Ho

IJ-o. 5 0

16

Fig. 3. Effect of simultaneous removal of external Ca 2+ and Na + on [Ca2+ ]i and [Na + ]i- A: superfusion of Fura-2-1oaded cells with Na÷-free solution induced a F340/F380 ratio increase from 0.65 to 1.55 corresponding to 253% increase in [Ca 2+ ]i (from 131 to 463 nM in this cell). Superfusion of Na +- and Ca2+-free solution prevented [Ca 2+ ]i increase. In contrast, we noted a 25% decrease in [Ca 2+ ]i (from 116 to 86 nM in this cell). This decrease was reversed upon the readmission of standard HBSS, and the Ca 2÷ response recovered during superfusion of Ca 2+containing Na+-free solution. A trace representative of 7 experiments is shown. B: control superfusion of SBFI-loaded cells with Na+-free solution induced a F340/F380 ratio decrease from 0.98 to 0.91 corresponding to a 29% decrease in [Na ÷ ]i from 93 to 66 mM in this cell). Superfusion of a Na ÷- and Cae+-free solution prevented the [Na + ]i decrease. A trace representative of 8 experiments is shown.

Fig. 4. Effect of Na + - C a 2+ exchanger inhibitors on [Ca 2+ ]i and [Na + ]i variations induced by Na+-free solution. A: supeffusion with Na+-free solution induced a F340/F380 ratio increase from 0.66 to 1.42 corresponding to a 313% increase in [Ca 2+ ]i (from 98 to 405 nM in this cell). Pretreatment with amiloride 5-(N,N-dimethyl) hydrochloride (100 /xM, Sigma) caused a F340/F380 ratio increase from 0.70 to 1.05 corresponding to 115% increase in [Ca 2+ ]i (from 113 to 243 nM in this cell) upon superfusion of Na+-free solution. The standard response was inhibited by 63%. The amiloride derivative solution by itself did not induce [Ca 2+ ]i variation. Washing the cell for 8 min with standard HBSS allowed the cell to recover the standard [Ca 2+ ]i increase during application of Na+-free solution. A representative of 5 experiments is shown. B,C: superfusion with Na+-free solution induced a F340/F380 ratio increase from 0.62 to 1.40 corresponding to a 355% increase in [Ca 2÷ ]i (from 86 to 392 nM in this cell, B) and a F340/F380 ratio decrease from 0.95 to 0.90 corresponding to a 20% decrease in [Na ÷ ]i (from 81 to 65 mM in this cell, C). Increasing [Mg 2-- ]o from 0.9 to 4 mM, prevented both the [Ca2÷ ]i increase and the [Na + ]i decrease induced by Na÷-free solution. Subsequent reduction of [Mg 2+ ]o to 0.9 mM triggered both [Ca 2÷ ]i and [Na÷ ]i variations. Representative traces of respectively 7 and 5 experiments are shown.

remained stable for up to 2 h under superfusion of standard HBSS. We investigated the effect on [Ca2+] i and [Na+]i of complete replacement of external Na ÷ by N M D G ÷. Superfusion of isolated type-I VHCs with 140 mM N M D G + induced an increase in [Ca2+] i and a decrease in [Na+]i (Fig. 2). In Fura-2 loaded VHCs, the fluorescence ratio increased from 0.69 + 0.06 to 1.44 + 0.15 (n = 46), corresponding to an increase in [Ca2+] i by 2 7 6 + 89% (Fig.

C. Chabbert et aL /Hearing Research 89 (1995) 101-108

A. f

I

0 Na +

2

4

2A). [Ca2+] i rose to its peak value in the first minute following application of Na+-free solution, then decreased slowly till the end of the 2 min superfusion. Readmission of Na ÷ ions to the external medium induced fast [Ca 2+ ]i regulation: the cell recovered its resting [ C a 2 + ] i w i t h i n 1 min. Cellular superfusion with solutions of [Na+]o (100 mM, 50 mM, and 0 mM) induced increases in the resting [Ca2+ ]i by 43 + 6%, 160 _ 14%, and 306 ___31%, respectively (n = 5, results not shown). In SBFI-loaded cells, a similar application of a Na+-free solution induced a decrease in fluorescence ratio from 0.96 ___0.02 to 0.90 ___0.02 (n = 14), corresponding to a decrease of the resting [Na+]i by 23 + 6% (Fig. 2B). Significant variations in the fluorescence appeared after a delay of 30 s and the lower ratio value was maintained until the addition of Na+-containing standard medium. Both a [Ca2+] i increase and a [Na+]i decrease are therefore triggered by [Na+]o reduction. To determine if the variations in [Ca 2÷ ]i and [Na+]i are coupled, we investigated the effect on [Ca2+]i and [Na+]i of simultaneous removal of external Na ÷ and Ca e+. Complete removal of the Ca 2+ ions from the Na+-free superfusion solution reversibly prevented both the [Ca2+] i increase (Fig. 3A) and the [Na+]i decrease (Fig. 3B). In Fura-2-1oaded VHCs, the simultaneous absence of external Ca 2+ and Na 2+ ions induced a decrease in resting [Ca 2+ ]i by 25 ___4% (n = 7). Similar [Ca2+]i decreases were observed during VHC superfusion with Ca2+-free HBSS.

I

pH i ;'.5 7

. 0

I ..,+

B

6

8

IC

6

8

IC

6

8

~c

I

7.5 PHi7 0

2

C

4

[ oo2 I

pH i

75f "z

-

-

~

6.5

o

2

;.

time (rain)

Fig. 5. Effect of complete removal of external Na + on pH i. A: superfusion with Na+-free solution (0 Na + ) did not evoke any significant pH i variation in the 14 ADB-loaded type I VHCs. B: cell superfusion with 25 mM NH4C1 for 3 min induced a pH i rise from 7.25 to 7.92. Before the end of the NH4CI superfusion, pH i decreased towards its initial values. The readmission of standard HBSS allowed the cell to recover its basal pH i. C: cell superfusion with HBSS equilibrated with 5% CO 2 and 95% 0 2 induced pH i decrease from 7.24 to 6.97. Upon the readmission of the standard HBSS, pH i increased slowly towards its initial value.

At.a'l. t- -I

o

,

2

;

.o

(raM) v

105

B

lll~8~l rL= (mM) o

t= 140 ,,

0

J

o

2

3

140

I1.

°'o

,

;

;

i

,

time (rain) C

[Na*]. (raM) v o o0 e~ $L

I,

140

I

o

4

s

e

lime (rain)

140

I

D [Na.~],, O M)"

[

140

140

I

o

I

0.9 I

,

~

i

z

,

J

0

I

2

3

4

5

6

0.5

,¢ I1.

U.

,

i

time (min)

;

time (rain)

Fig. 6. Effect of the chelation of internal Ca 2+ ions on [Ca 2+ ]i and [Na+]i variations induced by Na+-free solution. A,B: effects of control superfusions of Na+-free solution on [Ca 2+ ]i (A) and [Na+]i (B). C,D: after cell pretreatment with 100 /xM BAPTA-AM, both the [Ca 2÷ ]i (C) and the [Na+]i (D) variations were irreversibly abolished. Representative traces of respectively 3 and 5 experiments are shown. The resting [Ca 2+ ]i decreased from 107 nM to 12 nM upon treatment of this cell with BAPTA.

106

C. Chabbert et al. / Hearing Research 89 (1995) 101-108

These results indicate that the [Ca 2+ ]i increase evoked by Na+-free solution corresponds to a Ca 2÷ influx into the cell, and that the [Na+]i decrease, probably corresponding to Na ÷ efflux from the cell, is strictly dependent on the presence of external Ca 2÷ ions. These results are consistent with coupled Na ÷ and Ca 2+ transport across the plasma membrane. To test for the possible involvement of an ionic exchanger in the coupled [Ca 2÷ ]i and [Na+]i variations, we studied the effects on [Ca 2÷ ]i and [Na+]i variations of two inhibitors of Na÷-coupled transporters (Kleyman and Cragoe, 1988). Pretreatment with 100 /~M amiloride hydrochloride reduced the increase in [Ca 2÷ ]i induced by Na ÷free solution by 55 + 4% (from 259 ___37% to 116 + 21%; n = 4). Pretreatment with 100 /zM amiloride 5-(N,N-dimethyl) hydrochloride reduced the increase in [Ca 2÷ ]i by 57 + 7% (285 + 45% to 122 ___28%; n = 5, Fig. 4). These effects were reversed by washing with standard HBSS. These results indicate that the [Ca 2+ ]~ increase evoked by the removal of external Na ÷ is mediated by an amiloridesensitive mechanism. We also analyzed the effect of external magnesium concentration ([Mg2+] o) o n [Ca2+]i increase and [Na+]i decrease. Mg 2÷ ions have been reported to compete with Ca 2÷ ions for the external binding site of Na+-coupled membrane exchangers (Smith et al., 1987). Elevation of [Mg 2÷ ]o to 4 mM prevented both the [Ca 2÷ ]i increase (Fig. 4B) and the [Na+]i decrease (Fig. 4C). Subsequent reduction of the [Mg 2÷ ]o to 0.9 mM resulted in the recovery of both [Ca 2+ ]i and [Na÷]i variations. This amiloride and Mg 2÷ sensitivity suggests that the coupled variations of [Ca 2÷ ]i and [Na+]i triggered by the removal of external Na ÷ involves a Na÷-coupled ionic exchanger. N a + - H ÷ and N a + - C a 2÷ exchangers are candidates to mediate such coupled ionic variations. Indeed, [Ca2÷]i increase and [Na+]i decrease may be induced by cytosolic acidification through the activity of an N a + - H + exchanger coupled with an H + - C a 2+ exchanger (Moody, 1984). We therefore measured pH i at rest, and upon removal of external Na ÷, in ADB-loaded isolated type-I VHCs. The mean resting pH i value was 7.28 + 0.11 (n = 11). The application of Na+-free solution evoked no pH~ variation in 14 cells (Fig. 5A). As a control, cells were exposed to 25 mM NH~- (Fig. 5B) or to HBSS equilibrated with 5% CO2-95% 02 (Fig. 5C). These exposures induced pH i changes from 7.28 ___0.11 (n = 11) to respectively 7.95 + 0.18 (n = 5) or 6.92 + 0.13 (n = 6). The effect of [Na+]o reduction on [Ca2+] i and [Na+]i thus does not seem to involve cytosolic acidification. In order to investigate the involvement of an N a + - C a 2+ exchanger in the coupled [Ca2+]i a n d [Na+]i variations, we tested the effect of chelating internal Ca 2÷ on both responses. A requirement of internal Ca e+ for the reverse operation of the N a + - C a 2+ exchanger has been reported for a variety of cells (Dipolo and Beauge, 1986). Pretreatment of cells for 45 min with 100 /zM BAPTA-AM (1,2-bis (2-aminophenoxy) ethane-tetraacetic acid tetra-

EK+]o (mM) [ Na+]

(mM) o

L

6

Io t 14.0

[--1 o I

I o ]

I o I ~

S "

~ 16 li time (rain)

iX

1~

Ig

Fig. 7. Effect of [K + ]o variations on [Ca 2+ ]i increase induced by Na+-free solution. Superfusion with Na+-free solution induced a F340/F380 ratio increase from 0.70 to 1.45 corresponding to a 267% increase in [Ca 2+ ]i (from 113 to 415 nM in this cell). Complete removal of K* from the Na+-free solution prevented the [Ca 2+ ]i increase induced by Na*-free solution and induced a slight decrease in [Ca z÷ ]i (from 115 to 104 nM in this cell). Upon readmission of K + ions, [Ca 2÷ ]i variation induced by Na+-free solution recovered. A representative trace of 7 experiments is shown.

acetoxymethyl ester; Sigma) completely abolished both the [Ca2+] i increase (Fig. 6 A - C ) and the [Na+]i decrease (Fig. 6 B - D ) evoked by the removal of external Na ÷ ions. BAPTA pretreatment induced a marked decrease in the resting [Ca 2÷ ]i by 90 ___6% (n = 12), while no difference in resting [Na+]i was detected in the 11 SBFI-loaded, pretreated cells compared to untreated cells. These results indicate that the coupled variations o f [Ca2+]i and [ N a + ] i triggered by the removal of external Na ÷ require the presence of Ca 2+ on the internal face of the plasma membrane, and provide additional evidence for the presence of a N a + - C a 2+ exchanger in type-I VHCs. In cardiac ventricular myocytes, the N a + - C a 2÷ exchanger functions in the absence of external K + (Yasui and Kimura, 1990). In squid axon (Allen and Baker, 1986), and in rod outer segment (Schnetkamp et al., 1988), however, N a + - C a 2+ exchange requires the presence of K + ions on the same membrane surface as Ca e÷. In type-I VHCs, the complete removal of the K ÷ from the Na+-free superfusion solution reversibly prevented t h e [Ca2+] i increase in the 7 tested cells, and in fact induced a slight decrease of the resting [Ca2+]i (Fig. 7). These results indicate that the [Ca 2+ ]i increase evoked by the removal of external Na ÷ requires the presence of K ÷ on the external face of the plasma membrane and that K + activates the Ca2+-transport mechanism.

4. Discussion Several results suggest that the [Ca 2+ ]i increase and the [Na+]i decrease triggered by the removal of external Na ÷ are mediated by coupled Ca 2÷ and Na ÷ transport across the plasma membrane of type-I VHCs. First, in our experimental conditions, voltage-dependent Ca 2÷ channels do not seem to be involved in [Ca 2÷ ]i increase. These calcium

C. Chabbert et al. / Hearing Research 89 (1995) 101-108

channels only open during membrane depolarization, and it was shown in outer hair cells that the complete replacement of external Na + by NMDG + induced cell hyperpolarization (Sunose et al., 1991). Moreover, the presence of voltage-dependent Ca 2+ channels has not been clearly established in type-I VHCs (Rennie and Ashmore, 1991; Eatock and Hutzler, 1992; Griguer et al., 1993). Second, the [Ca 2+]i increase upon removal of external Na + does not seem to be directly associated with a release of Ca 2+ from intracellular stores, since this Ca 2+ response depends strongly on the presence of external Ca 2+. Third, the decrease in [Na+]i upon removal of external Na + cannot be interpreted as an effiux of Na + ions through transduction channels localized in the hair-bundle membrane or through unselective channels for it was abolished in the absence of external Ca 2+. The complete replacement of external Na + by NMDG + as been reported to lead to a loss of cell sodium, potassium and chloride, after periods of time longer than 3 min. Such effects have been interpreted by the possible passage of NMDG across the cell membrane, and were accompanied by acidification of the cells (Mroz and Lechene, 1993). In our studies external Na + was replaced by NMDG + for periods of 2 min and no variation of pH i was detected. However, a possible permeation of NMDG could not be ruled out. The high [Na+]i found in resting isolated type-I VHCs was previously reported in isolated outer hair cells (Ikeda et al., 1992). Such [Na+]i has been suggested to be pathological (Mroz et al., 1993). However, we recently reported that in patch-clamped type I VHCs, isolated using similar enzymatic procedure, the resting [Ca2+]i was 102 _ 42 nM (n = 50), and the mean resting potential was - 6 1 . 7 _ 7.2 mV (n = 15), suggesting a normal physiologic state (Griguer et al., 1995). Moreover we never saw any differences in the responses to the removal of external [Na +] in cells with different [Na+]i . In our experimental conditions, the high [Na+]i was attributed to a slow Na + influx into the cell through transduction channels. Na+-dependent membrane transporters can mediate coupled [Ca2+]i and [Na+]i variations upon removal of external Na +. While an N a + - H + exchanger coupled with a H + - C a 2+ exchanger would be a good candidate for such an effect (Moody, 1984), the absence of pH i variations during the application of a Na +-free solution argues against this hypothesis. An N a + - C a 2+ exchanger could also mediate [Ca 2+ ]i increase and [Na+]i decrease upon removal of external Na +. Such an exchanger can function in a bidirectional manner depending on the electrochemical gradient for Na + and Ca 2+ across the cell membrane (Allen et al., 1989). In our experiments, [Ca2+] i increase and [Na+]i decrease during removal of external Na + could be due to reverse operation of this exchanger, while the return toward the resting levels could be attributed to forward operation. Since the only specific inhibitor of the N a + - C a 2+ exchanger, the XIP peptide (Li et al., 1991), acts at an

107

internal site, it is inappropriate in our experimental conditions. Therefore, we used amiloride derivatives that despite incomplete block of the N a + - C a 2+ exchanger activity, are useful tools to slow such coupled transport (Kleyman and Cragoe, 1988). In our experimental conditions, we reported such a blocking effect on the [Ca 2+ ]i increase induced by Na+-free solution. Several studies show that Mg 2+ ions compete with Ca 2+ for the divalent binding site of the N a + - C a 2+ exchanger, without being transported (Smith et al., 1987; Supplisson et al., 1991). The complete inhibition of both the [Ca2+]i increase and [Na+]i decrease by high [Mg2+]o, further suggests the involvement of such an exchanger in the coupled [Ca2+] i and [Na+]i variations. The inhibition of both the [Ca2+]i increase and the [Na+]i decrease induced by Na+-free solution in cells pretreated with BAPTA-AM suggests that internal Ca 2+ is required for this coupled ionic transport. Such a requirement has been reported for the reverse operation of the N a + - C a 2+ exchanger in variety of cells, since Ca 2+ ions act on an internal regulatory site of the N a + - C a 2+ exchanger (Dipolo and Beauge, 1986). K + ions have been reported to modulate N a + - C a 2+ exchanger activity in squid axon (Baker et al., 1969) and in rod outer segment (Cervetto et al., 1989; Schnetkamp et al., 1989). In these cells, a high [K+]o increased [CaZ+]odependent Na + effiux, and high [K+]i increased [Na+]o dependent Ca 2+ effiux. Moreover, in the rod outer segment N a + - C a 2+ exchange is not only dependent on the presence of K +, but requires the co-transport of K +, acting as a 4Na+:lCa 2+, 1K + exchanger (Cervetto et al., 1989; Schnetkamp et al., 1989). In contrast, the cardiac Na + Ca 2+ exchanger, acting as a conventional 3Na+:lCa 2+ exchanger, operates in the absence of external K + (Yasui and Kimura, 1990). Because the [Ca 2+ ]i increase induced in type-I VHCs by Na+-free solution was observed only when external K + was present, a regulatory effect of K + on the N a + - C a 2+ exchanger is likely. Comparing the amplitude of [Ca2+]i and [Na+]i variations during removal of external Na + reveals that the decrease in [Na+]i exceeds that expected for conventional, 3Na+:lCa 2+ exchange. However, the detailed study of the ionic exchange stoichiometry would require recording exchange currents using the patch-clamp technique. Moreover, the different binding properties of each dye did not permit a comparative analysis of the time course of [Ca 2+ ]i and [Na+]i changes. It must be noted, however, that at rest, the high internal Na + level would normally preclude any form of Ca 2+ extrusion through a conventional 3Na + :lCa 2+ exchanger, and would instead mediate Ca 2+ loading of the cell. This was never observed in our experiments. Thus, if the carrier is to act as a Ca 2+ extruder at rest, it must be a non-conventional exchanger, such as the K+-dependent exchanger described in rod outer segments. Despite an unresolved problem of stoichiometry, the present results argue for the presence in type-I VHCs of an N a + - C a 2+ exchanger, which in our experimental condi-

108

C. Chabbert et al. / Hearing Research 89 (1995) 101-108

tions acts in reverse mode. In vivo, the operation of this Na +-dependent membrane exchanger in the forward mode could contribute to [ C a 2 + ] i regulation during vestibular information transmission. The specific role of K + in the exchanger function remains to be established.

Acknowledgements The authors are grateful to Drs. D.W. Hilgeman and A.J. Hudspeth and Messrs. N.P. Issa and R.G. Walker for helpful comments on the manuscript. We thank Drs. J.F. Ashmore, F. Morel and G. Dayanithi for their critical comments on this work, Dr. J.M. Salmon for helpful contribution on the pH i measurement, and J. Boyer for editorial assistance. This work was supported in part by the Direction des Recherches et Etudes Techniques (DRET, Grant 91.103), Janssen France, and the Corporate Office of Science and Technology (COSAT).

References Alien, T.J.A. and Baker, P.F. (1986) Comparison of the effects of potassium and membrane potential on the calcium-dependent sodium efflux in squid axons. J. Physiol. 378, 53-76. Allen, T.J.A., Noble, D. and Reuter, H. (1989) Sodium-Calcium Exchange. Oxford University Press, Oxford, UK, 332 pp. Baker, P.F., Blaustein, M.P., Hodgkin, A.L. and Steinhardt, R.A. (1969) The influence of calcium on sodium efflux in squid axons. J. Physiol. 200, 431-458. Cervetto, L., Lagnado, L., Perry, R.J., Robinson, D.W. and McNaughton, P.A. (1989) Extrusion of calcium from rod outer segment is driven by both sodium and potassium gradients. Nature 337, 740-743. Chabbert, C., Geleoc, G., Lehouelleur, J. and Sans, A. (1994) Intracellular calcium variations evoked by mechanical stimulation of mammalian isolated vestibular type I hair cells. Pfliigers Arch. 427, 162-168. Dipolo, R. and Beauge, L. (1986) In squid axons reverse Na-Ca exchange requires internal Ca and/or ATP. Biochim. Biophys. Acta 854, 298-306. Doi, T. and Ohmori, H. (1993) Acetylcholine increases intracellular Ca 2÷ concentration and hyperpolarizes the guinea pig outer hair cell. Hear. Res. 67, 179-188. Eatock, R.A., Corey, D.P. and Hudspeth, A.J. (1987) Adaptation and mechanoelectrical transduction in hair cells of the bullfrog's sacculus. J. Neurosci. 9, 2821-2836. Eatock, R.A. and Hutzler, M.J. (1992) Ionic currents of mammalian vestibular hair ceils. In: B. Cohen, D.L. Tomko, and F. Guedry (Eds.), Ann. NY Acad. Sci., 59-74. Griguer, C., Kros, C.J., Sans, A. and Lehouelleur, J. (1993) Potassium currents in type II vestibular hair cells isolated from the guinea pig's crista ampullaris. Pfliigers Arch. 425, 344-352. Griguer, C., Chabbert, C., Sans, A. and Lehouelleur, J. (1995) Transient increase in cytosolic free calcium evoked by repolarization in type I vestibular hair cells of rats. Cell Calcium 17, 327-334. Grynkiewicz, G., Poenie, M. and Tsien, R.Y. (1985) A new generation of calcium indicators with fluorescence properties. J. Biol. Chem. 260, 3440-3450. Harootunian, A., Kao, J.P.Y., Eckert, B.K. and Tsien, R.Y. (1989) Fluorescent ratio imaging of cytosolic free Na in individual fibroblasts and lymphocytes. J. Biol. Chem. 264, 19458-19467. Hudspeth, A.J. and Gillespie, P.J. (1994) Pulling springs to tune transduction: Adaptation by hair cells. Neuron 12, 1-9.

Ikeda, K., Saito, Y., Nishiyama, A. and Takasaka, T. (1992) Na + -Ca 2+ exchange in the isolated cochlear outer hair cells of the guinea-pig studied by fluorescence image microscopy. Pfliigers Arch. 420, 493499. Issa, N.P. and Hudspeth, A.J. (1994) Clustering of Ca 2+ channels and Ca2+-activated K + channels at fluorescently labeled presynaptic active zones of hair cells. Proc. Natl. Acad. Sci. 91, 7578-7582. Kleyman, T.R. and Cragoe, E.J. (1988) Amiloride and its analogs as tools in the study of ion transport. J. Membrane Biol. 105, 1-21. Lahmy, S., Salmon, J.M., Vigo, J. and Viallet, P. (1989) pH i and DNA content modifications after ADR treatment in 3T3 fibroblasts. A microfluorimetric approach. Anticancer Res. 9, 929-936. Lenzi, D. and Roberts, W.M. (1994) Calcium signaling in hair cells: multiple roles in a compact cell. Curr. Opin. Neurobiol. 4, 496-502. Li, Z., Nicoll, D.A., Collins, A., Hilgemann, D.W., Filoteo, A.G., Penniston, J.T., Weiss, J.N., Tomich, J.M. and Philipson, K.D. (1991) Identification of a peptide inhibitor of the cardiac sarcolemnal Na-Ca exchanger. J. Biol. Chem. 266, 1014-1020. Moody, W.J. (1984) Effects of intracellular H-- on the electrical properties of excitable ceils. Ann. Rev. Neurosci. 7, 257-278. Mroz, E.A. and Lechene, C. (1993) Extracellular N-methyl-D-glucamine leads to loss of hair-cell sodium, potassium, and chloride. Hear. Res. 70, 146-150. Mroz, E.A., Nissim, K.R. and Lechene, C. (1993) Electron-probe analysis of isolated goldfish hair cells: implications for preparing healthy cells. Hear. Res. 70, 9-21. Ohmori, H. (1988) Mechanical stimulation and fura-2 fluorescence in the hair bundle of dissociated hair cells of the chick. J. Physiol. 399, 115-137. Rennie, K. and Ashmore, J.F. (1991) Ionic currents in isolated vestibular hair cells from the guinea pig cristae ampullaris. Hear. Res. 51, 279-292. Roberts, W.M. (1994) Localization of calcium signals by a mobile calcium buffer in frog saccular hair cells. J. Neurosci. 14, 3246-3262. Roberts, W.M., Howard, J. and Hudspeth, A.J. (1988) Mechanoelectrical transduction, frequency tuning and synaptic transmission by hair cells. Ann. Rev. Cell Biol. 4, 63-92. Roberts, W.M., Jacobs, R.A. and Hudspeth, A.J. (1990) Colocalization of ion channels involved in frequency selectivity and synaptic transmission at presynaptic active zone of hair cells. J. Neurosci. 10, 36643684. Salmon, J.M., Vigo, J. and Viallet, P. (1988) Resolution of complex fluorescence spectra recorded on single unpigmented living cells using a computerised method. Cytometry 9, 25-32. Scarfone, E., Ulfendahl, M., L6fstrand, P. and Flock, A. (1991) Lightand electron microscopy of isolated vestibular hair cells from the guinea pig. Cell Tiss. Res. 266, 51-58. Schnetkamp, P.P.M., Szerencsei, R.T. and Basu, D.K. (1988) Na-Ca exchange in bovine rod outer segments requires potassium. Biophys. J. 53, 389a. Schnetkamp, P.P.M., Basu, D.K. and Szerencsei, R.T. (1989) Na-Ca exchange in the outer segment of bovine rod photoreceptors requires and transports potassium. Am. J. Physiol. 257, C153-C157. Shigemoto, T. and Ohmori, H. (1991) Muscarinic receptor hyperpolarizes cochlear hair cells of chick by activating Ca2+-activated K + channels. J. Physiol. (Lond.) 442, 669-690. Smith, J.B., Cragoe, E.J. and Smith, L. (1987) Sodium-calcium antiport in cultural arterial smooth muscle cells. J. Biol. Chem. 262, 1198811994. Sunose, H., Ikeda, K., Nishiyama, A., Saito, Y. and Takasaka, T. (1991) Membrane potential measurement in isolated outer hair cells of the guinea pig. BSTR 14th ARO Midwinter Meeting, p 14. Supplisson, S., Kado, R.T. and Bergman, C. (1991) A possible Na/Ca exchange in the follicle cells of Xenopus oocyte. Dev. Biol. 145, 231-240. Valat, J., Devau, G., Dulon, D. and Sans, A. (1989) Vestibular hair cells isolated from guinea pig labyrinth. Hear. Res. 40, 255-260. Yasui, K. and Kimura, J. (1990) Is potassium co-transported by the cardiac Na-Ca exchange?. Pfliigers Arch. 415,513-515.