ce//ca/dum(la9o) 11, 353-360 0LongmanGrwpUKLtdl~O

Modulation of free Ca oscillations in single hepatocytes by changes in extracellular K+, Na+ and Ca2+ N.M. WOODS, C.J. DIXON, K.S.R. CUTHBERTSON and P.H. COBBOLD The Llepartment of Human Anatomy and Cell Biology, The University of Liverpool, Liverpool, UK Abstract - Single rat hepatocytes, microinjected with the calcium-sensitive photoprotein aequorin, when stimulated with either phenyfephrine or arg8-vasopressin exhibit agonist-specific oscillations in cytosolic free calcium levels (free Ca). In the majority of the cells examined adding excess potassium chloride, sodium chloride or choline chloride abolished transient behaviour. However, in cells that continued to oscillate the transient parameters were subtly modified by these treatments. In experiments using phenylephrine as the agonist, adding excess potassium chloride to the superfusate significantly reduced transient length, increased the rate of transient rise and reduced the smoothed peak free Ca level without significantly altering the intertransient resting free Ca level or the falling time constant. The possible mechanisms by which these alterations may occur are discussed.

In single rat hepatocytes, calcium mobilising agonists induce changes in free Ca that are complex, consisting of free Ca oscillations, the characteristics of which, depend closely on agonist species and concentration [l-3]. These oscillatory changes in free Ca probably arise as a result of the action of phosphoiuositidase C (PIG) on a plasmalemmal phospholipid, phosphatidylinositol 4,5 bisphosphate (PIP;?), the link between agonist-receptor interaction and PIP2 breakdown being provided by guanine nucleotide binding proteins (G-proteins). The products of this transduction pathway, inositol 1,4,5 trisphosphate (IP3) and diacylglycerol (DAG), established roles in releasing have well intracellularly sequestered calcium and activating protein kinase C, respectively [4-71, but the mechanism of transient generation is subject to debate [S, 91.

An important question that remains to be answered is the relative contributions of external calcium influx and intracellular calcium release to the free Ca change in a single rat hepatocyte during transient generation. A role for inositol tetrakisphosphate (IP4) in promoting calcium influx has been suggested [ 10, 111, although whether this has a role in the oscillatory free Ca changes seen in a single liver cell remains unclear. We have examined, in one of the first cell types to have been shown to exhibit agonist-induced free Ca oscillations, the dependence of the transient behaviour on the external cation environment. The results suggest that alterations of some of the transient parameters can be elicited by increasing the level of external monovalent cations (IS+ and Na+) and that reducing the extracellular level of calcium could affect transient generation. 353

354

CELL CALCIUM

Vasou

Vasop + K’

3

60C

I

I

I

r‘l

-1

Gmin

I

I

3 min

Fig. 1 (a) The reversibly

inhibitory

effect of 40 mM KCI (K’) on free Ca transients

induced by 1 @I phenylephrine

(Phe) in a single rat

hepatocyte (b) The effect 40 mM KC1 (K’) on transients

induced by 0.35 nM rug’-vasopressin

(Vasop). Time constant

for transients

1 s and for

resting levels 15 s. The results shown are typical of eight separate experiments

Materials and Methods Single hepatocytes were prepared as described before [2]. The sources of chemicals, the microinjection procedure, experimental conditions and data collection have been described previously [ 1, 21. Aequorin was dialysed using a modification of the procedure described before [12], in that ditbiothreitol (1 mM) and an osmotic buffer, polyvinylpyrrolidone (0.7 mM, av. mol.’ wt 4x104 were included in the dialysis fluid The experimental medium consisted of Williams Medium E to which agonists (phenylephrine or arg-vasopressin) and excess solute (choline chloride, potassium chloride or sodium chloride) were added to give the final desired concentration (40 mM). Experiments were carried out using sucrose (80 mM) added to Williams Medium E as a hypertonic control. In experiments involving reduced extracellular free Ca levels the experimental medium was Williams medium E {where calcium had been omitted at the

preparation stage (Flow Labs.)}, to which 0.5 m.M EGTA and agonist (phenylephrine or arg’-vasopressin) were added. The apparent association constants used to calculate the free Ca concentration were Ca-EGTA 2.75 x lo7 M and Mg-EGTA 2.88 x lo2 M at pH 7.40. Off line analysis of the data was performed to determine the following transient parameters; length, period, rate of rise, smoothed peak free Ca, intertransient resting free Ca and falling time constant. Statistical examination of the data was undertaken using tire ‘Students’ t-test, significance was assumed when P < 0.05.

Results Single rat hepatocytes when super-fused with low concentrations of either arg8-vasopressin or phenylephrine, sufficient to activate phosphorylase

MODULATION

OF FREE Ca2+ OSCILLATIONS

IN HEPATOCYTElS

[13], exhibit a series of repetitive transient rises in free Ca [ 1, 2, 141. Adding 40 mM excess KC1 to the superfusate abolished or reduced the frequency of the free Ca transients induced by both thesFeagonists (Fig. .la, lb). Previous studies, using chloride uptake experiments [15], have revealed that increasing the external KC1 concentration by this amount did not result in significant membrane potential depolarisation, in good agreement with earlier direct microelectrode measurements in perfused and isolated rat livers [16, 171. Increasing the KC1 concentration above this value elicited a reduction in the membrane potential [ 15-171. In all of the cells tested the transients were either abolished (10/18) or the frequency was reduced (8/18) and in 98 cells the reduction in transient frequency was accompanied by a reversible reduction in peak free Ca (Fig. 2). The reduction in frequency was prompt, usually occurring within one to two transient periods after KC1 application and _ l-1 ’ ” ” ” ‘4

l-

’

I

I

6 min Fig. 2 The effect of 40 mM excess KCI (K+) on transient Ca evoked

by

aequorin-microinjected

c

Phe

1

Phe t -he

Ffg. 3 The effect of 40 mM choline chloride (ChCl) on free Ca transients hepatocyte.

induced

by 2 phi phenylephrine

Time constant for transients

@‘he) in a single rat

1 s and for resting levels

20 s. The results shown are typical of five separate experiments. Time scale bar 4 min

PhetK+

Phe

L

free

355

1 pM phenylephrine hepatocyte.

(Phe)

peak

in a single

Time constant as for Figure 1.

The results shown are typical of five separate experiments

was readily reversible (see Fig. la). Where a reduction in peak free Ca was observed the first few transients were smaller and then gradually n%rned to normal free Ca levels over the course of 3 to 8 transients (Fig. 2). Comparing the four transients immediately prior to KCl application with the first four transients following KCl application revealed that in 4 out of 8 cells raised KC1 significantly reduced the peak free Ca level reached (P < 0.05). In contrast to the recovery seen with peak free Ca levels, the reduction in transient frequency persisted throughout the period of KC1 application. A summary of the transient parameters from phenylephrine and phenylephrine + 40 mM KC1 treated cells can be seen in Table 1. Similar inhibitory results to those described for raised potassium levels were obtained using either raised sodium chloride levels (2 cells) (results not shown) or raised choline chloride concentrations (5 cells) (Fig. 3), both at 40 mM. The inhibitory effects of raised KC1 were not due to increases in the exfracellular Clconcentration since anionic substitution of 40 mM NOs- for Cl- [ 181 proved to be equally effective in inhibiting the free Ca transients (data not shown).

356

CELL CALCIUM

Table 1 The effect of raised extraeellular potassium levels on ke Ca transient parameters in single rat hepatoeytes. Single rat hepatoeytes were stimulated by the al adrenoeeptor agonist phenylephrine and then 40 mM excess KC1

was added to the superfusate in combination with the agonist. The data were sampled at 20 ms intervals and the transient parameters shown were computed. The parameters calculated for the phenylephrine-only treated cells refers to the control series of transients elicited before adding raised KC1 (n = 74), for phenylephrine t KC1 treated cells the parameters shown were calculated from all the transients seen throughout the period of agonist + KC1 application, (n = 60). * indicates statistical significance at the P < 0.05 level using the ‘Students’ t-test Treatment

Phe

Length

Rate of rise

Smoothed peak

Intertransient

Falling time

(e)

(nM s-l)

free Ca (nM)

resting Ca (nM)

constant

6.02 f 0.1

388 f 18

830 f 18

210f6

3.2 f 0.1

4.9ItO.l’

513 f 27’

734 f u3*

192f6

3.0 fO.l

Phe + 4Omh4 KC1

The inhibitory effects of incrcascd extracellular potassium ion levels could only be partly ascribed to increases in tonicity of the bathing medium since control experiments, where 80 mM sucrose was included, proved only fractionally as effective in

reducing transient frequency (15 + 10.6 %, n = 2) (Fig. 4). The inhibitory effects of 40 mM KC1 could not be reversed by increasing the external calcium concentration from normal (1.8 mM) to 10.8 mM (data not shown). In 3/4 cells increasing calcium concentration proved to

800-

Pas

Phe + Sucmee

the external be further

Phe

inhibitory, whereas in the remaining cell the transient frequency returned to normal but the peak free Ca level attained was reduced (data not shown). Exposure of cells to 40 mM KC1 alone had no effect on the resting free Ca level. However, subsequent stimulation with phenylephrine, in the continued presence of raised KCl, evoked a few transients, typically two to four, that were quickly inhibited (data not shown). Experiments where the extracellular calcium concentration was reduced to c. 100 p.M or below also resulted in inhibition or reduction of transient frequency (Fig. 5a-d), reminiscent of some of the changes in transient parameters resulting from KC1 addition. The effects of increasing the extracellular calcium concentration were less conclusive. In 6/11 cells increasing the calcium concentration up to 9 mM resulted in slight increases in transient frequency, whereas in 3/11 cells similar increases had no effect on transient frequency and in 2/11 cells such increases proved to be inhibitory (data not shown)

Discussion

J

L

4min Fig. 4

single

The effect of 80 mM sucrose on free Ca transients tat hepatocyte

Time constants

induced

by 2 @VI phenylephrine

as for Figure 3. Time scale bar 4 min

in a (Phe).

The data obtained indicate that raising the external concentration of monovalent cations or reducing the superflrsate calcium concentration can reversibly inhibit free Ca transient generation in single rat hepatocytes. Closer examination of the results revealed that subtle alterations in the transient

MODULATION

OF FREE Ca2+ OSCILLATIONS

357

IN HEPATOCYTES

1000

8ooF

b Phe Phe+Cafree

Phe

,

800 -

600

800 -

400

3 1’1.)

‘+ (u

_OLd

-

I-O”

-

200

200 u

’

” p

Phe + 1OOuM Ca2+

I-

Vp+lOOuM

Ca2+

I

800

1 3

600

Fig. 5 The inhibitory (a) Inhibition

effect of calcium-free

of transients

and reduced extracellular

induced by 1 @I phenylephrine

calcium media on free Ca transients in single rat hepatocytes

(Phe) in calcium-free

(0.5 mM EGTA) medium. Time constant for transients

0.3 s and for resting levels 15 s. Time scale bar 2 min (b) The reversibly

inhibitory

effect of calcium-free

(Phe) in a single rat hepatocyte.

(c) The effect of reduced extracellular Time constant for transients (d) The effect of reduced

Time constant

for transients

medium

Time constant for transients free calcium

(0.5 mM EGTA) on free Ca transients

induce-d by 2 j&l phenylephrine

1 s and for resting levels 30 s. Time scale bar 8 min

levels (100 pM Ca”)

on free Ca transients

induced by 2 pIk4 phenylephrine

(Phe).

induced by 0.4 nM arg’-vasopressin

(VP).

1 s and for resting levels 30 s. Time scale bar 4 min extracellular

calcium

levels (100 ph4 Ca2’) on free Ca transients

1 s and for resting levels 30 s. Time scale bar 8 min

parameters could be elicited by such treatments. The principal effect of raised potassium levels (40 mh4) was on transient frequency which, in lo/18 cells, was reduced to zero, while the remaining cells exhibited a greater than two fold decrease in transient frequency. In addition, increased potassium resulted in an approximate 21% decrease in transient duration, a 30% increase in the rate of transient rise, a 12% decrease in the mean peak free Ca level

attained during stimulation and a 8% reduction in the mean intertransient resting free Ca level. Earlier studies [19] revealed that, in whole perfused livers, depoltising levels of potassium (80 mM) were excitatory but this was almost certainly due to depolarising effects on cells other than parenchymal hepatocytes since the effect could be antagonised with indomethacin or bromophenacyl bromide. Direct microelectrode measurements [16, 171 have

358

demonstrated that the liver is particularly, perhaps uniquely, unresponsive, at least in terms of depolarisation, to moderate increases in extracellular potassium ion levels of up to 40 mM, and even up to 56 mM [16]. The physiological basis for this apparent ‘step’ in sensitivity to increases in [Kt10 is unknown. More recent studies I:151 provide evidence that significant hepatocyte depolarisation occurs only when [Kt10 equals or is greater than 60 mM and that increases in [Kt10 to this level were found to inhibit arg- vasopressin induced increases in 45Ca uptake [15]. It therefore seems unlikely that the inhibitory effects of raised potassium levels reported here can be explained on the basis of changes in membrane potential. Further evidence in support of this conclusion arises from studies on the effects of similar increases in extracellular choline chloride (Fig. 3) and sodium chloride levels where both were as effective as raised potassium in abolishing transient generation. Therefore cessation of transient behaviour is unlikely to be caused by a reduced electrochemical gradient for calcium entry or by voltage-dependent effects on plasmalemmal ion channels. It is interesting therefore to compare the sensitivity of agonist-induced free Ca changes in rat inhibition by depolarising hepatocytes to concentrations of potassium with studies on other cell types, for example human endothelial cells [20], where free Ca changes can still occur even in the presence of 140 mM external potassium [20]. Whether or not these differences indicate fundamental variations in the basic properties of the cellular oscillator remains to be determined. The data presented in Table 1 make it unlikely that, in cells that continued to oscillate, IP3 binding to the endoplasmic reticulum was depressed since the rate of rise of free Ca actually increased with raised potassium. How then might these inhibitory effects be explained? It is possible that the membrane signalling events during stimulation may be susceptible to changes in ionic composition or tonicity of the bathing medium. A reduction of transient frequency alone would be expected if agonist- receptor interaction were depressed [ 1, 21, so an extracellular action of raised KC1 (and other ions) may be to interfere with normal receptor function at or around the agonist binding site. We have previously shown that activation of protein

CELL CALCIUM

kinase C by phorbol esters results in a reversible reduction of transient frequency, but it is difficult to see how modification of the extracellular ionic milieu could activate PKC. In many biological membranes the presence of negatively charged phospholipids leads to the development of a negative surface potential, making interactions with oppositely charged ions in the bathing medium likely. Furthermore, alterations in the ionic milieu could potentially affect the charge distribution on the membrane (and therefore around the receptor site) and have effects on agonist-receptor interaction. Studies on the fibronectin receptor 1211 have revealed that changes in receptor affinity and specificity could be prompted by altering the ionic composition of the bathing medium. Moreover, there is additional evidence obtained from studies on opiate receptor subtypes [22] that increases in both mono and divalent cations could suppress agonist binding. in While caution should be expressed extrapolating data obtained from other cell and receptor types to the results reported here, they at least raise the possibility that some of these results may be explicable, at least in part, by changes in the ionic environment affecting agonist-receptor interaction. The observation that in hepatocytes, transient frequency can be reduced in lowered extracellular calcium levels, but where an appreciable electrochemical gradient still exists, suggests that effects on receptor properties rather than on intracellular calcium stores may offer an adequate explanation of these effects. In some cells however the effect of raised potassium was not totally inhibitory. In cells that continued to show free Ca oscillations, at reduced frequency in the presence of raised potassium, the rate of rise increased by approximately 30%, the peak free Ca level attained was reduced as were the length of the transients and the mean intertransient resting free Ca level. A change in extracellular osmolarity might, perhaps, affect cytoplasmic processes, but the failure of extracellular sucrose (80 mM) to appreciably affect transient frequency argues against such a mechanism. A further factor to be assessed when interpreting these data is the effect of increasing the tonicity of the bathing medium on the signalling pathway. That hepatic osmoreceptors

MODULATION

OF FKEE Ca2+ OSCILLATIONS

359

IN HEPATOCYTES

do exist is well documented [23,24] although they are located primarily in the hepatic portal vein area and not on parenchymal cells. The control experiments using sucrose revealed a much smaller increase in transient period and it therefore seems likely that the inhibitory effects reported here could only partly be ascribed to increases in tonicity of the bathing medium. There is evidence in some cell types [25], including hepatocytes [26], that the initial event following agonist binding is a stimulation of extracelhuar calcium influx, measurable before and therefore calcium intracellular release presumably before PI hydrolysis. The role of this initial calcium influx is unclear, however, the suggestion that G-protein activation of PIC may be enhanced by increased free Ca levels [27] leaves open the possibility that the early calcium entry may have a role in the oscillator mechanism [28], especially if a localised sub-plasmalemmal zone of raised free Ca occurred. Recently, evidence for a pacemaker like rise in free Ca preceding each calcium transient has been observed in Furaloaded endothelial cells [20], which could represent such a zone. In hepatocytes it has been suggested that the early calcium influx may occur via an inositol lipid-independent mechanism possibly mediated by a G-protein activated calcium channel [26]. If this early calcium entry were inhibited by raised potassium levels this would have the effect of inhibiting PIC activation and hence intracellular calcium release. In cells that continued to oscillate the initial calcium entry may only be reduced, not totally blocked, so that the time taken to generate sufficient IP3 to cause intracellular calcium release may be increased, hence prolonging the intertransient interval. Interestingly, procedures designed to reduce agonist mediated calcium entry such as reduced extraccllular calcium levels (to c. 100 p.M or below, see Fig. 3) or the use of low concentrations of organic calcium channel antagonists (verapamil at concentrations of l-5 u.lvl or nifedipine at concentrations of 2-5 pM results not shown), induced exactly the same changes in transient frequency and transient parameters as those seen with raised potassium where the cells continued to oscillate.

Acknowledgements Thisworkwasfunded

by The Wellcome Trust

References 1. Woods NM. duthbertson KSR. Cobbold PH. (1986) Repetitive transient rises in cytoplasmic free calcium in hormone-stimulated hepetocytes. Natum, 319.600-602. 2. Woods NM Cuthbertson KSR Cobbold PH. (1987) Agonist-induced oscillations in cytoplasm& free calcium concentration in single rat hepatocytes. Cell Calcium, 8, 79-100. 3. Cobbold PH. Woods NM. Wainwright J. Cua KSR (1988) Single cell measurements on calcium mobilising purinoceptors. J. Rec. I&s., 8.481491. 4. Streb H. Irvine RF. Berridge MJ. Schulz 1. (1983) Release of calcium from a non-mitochondtial store in pancmatic acinar cells by inositol trisphosphate. Nature, 306.67-69. 5. Joseph SK. Thomas AP. Williams IU. Irvine RF. Williamson JR (1984) myo-inositol 1,4,%risphosphate. J. Biol. Chem., 259, 3077-3081. 6. Benidge MJ. (1984) Inositol ttisphosphate and diacylglycerul as second messengers. Biochem. J., 220, 345-360. I. Nishizuka Y. (1986) Studies and perspectives of protein kinase C. Science, 233,305-312. 8. Benidge UT. Galione A. (1988) Cytosolic caIcium oscillatota. FASEB J., 2, 30743082. 9. Cobbold PH. Dixon cf. Sanchez-Bueno A. Woods NM. Daly MJ. Cuthbertson KSR (1989) Receptor control of calcium transients. In: Nahxski S. ed. Ttansmembrane signalling, Intracellular messengers and Implications for drug development. England Wiley. 10. Putney Jr JW. (1986) A model for receptor-regulated calcium entry. Cell Calcium, 7, 1-12. 11. Irvine RF. Moor KM. (1987) Inositol(1,3,4,5) tetmlcisphosphate-induced activation of sea urchin eggs requires the presence of inositol trisphcephate. Biochem. Biophys. Res. Commur~, 146,284290. 12. Cobbold PH. Cuthbertson KSR Goyns MI% Rice VR (1983) Aequorin measurements of free calcium in single mammalian cells. J. Cell. Sci., 61. 123-136. 13. Hue L. Feliu JE. Hers H-G. (1978) Control of gluconeogenesis and of enzymes of glycogen metabolism in isolated rat hepatccytes. B&hem. J., 176.791-797. 14. Woods NM. Cuthbertson KSR. Cobbold PH. (1987) Phcrbolester-induced alteration of free calcium ion transients in single rat hepatocytes. Bicchem. J., 246, 619-623. 15. Savage AL. BifTen M. Martin BR. (1989) Vasopressin stimulated calcium influx‘ in rat hepatocytes is inhibited in high K’ medium B&hem. J., 260, 821-827. 16. Claret M. Coraboeuf E. (1970) Membrane potential of perfmed and isolated rat liver. J. Physiol., 210. 137P. 17. Williams JA. Withrow CD. Woodbury DM. (1971) Eikts of nephrectomy and KCl on transmembrane potentials,

360

18.

19.

20.

21.

22.

23.

24.

CELL CALCIUM intracellular electrolytes, and cell pH of rat muscle and liver in vivo. I. Physiol, 212, 117-128. Casteels R. (1971) The distribution of chloride ions in the smooth muscle cells of the gtdnea pigs taenia coli. I. Physiol., 214,225-243. Altin JG. B&n TJ. Karjalainen A. Bygrave FL. (1988) Exposure to depohuising concentrations of potassium inhibits hormonally-induced calcium influx in tat liver. Biochem. Biophys. Res. Commun., 153,1282-1289. Jacobs R. Merrit JE. Hallam TJ. Rink TJ. (1988) Repetitive spikes in cytosohc calcium evoked by histamine in human endothehal cells. Natum, 335,40-45. Gailit I. Ruoslahti E. (1988) Regulation of the tibronectin receptor affinity by divslent cations. I. Biol. Chem., 263, 12927-12932. Sargent DF. Bean JW. Kosterlitz HW. Schwyzer R. (1988) Cation dependence of opiate receptor binding supports theory of membrane mediated receptor selectivity. Biochemistry, 27,4974-4977. Ada&i A. Niijima A. Jacobs HL. (1976) An hepatic osmoreceptor mechanism in the rat: Electrophysiological and behaviorial studies. Am. I. Physiol., 231, 1043-1049 Baertschi AI. Vallet PG. (1981) Osmosensitivity of the hepatic portal vein ama and vasopressin release in rats. I.

Physiol., 315.217-230. 25. Sage SO. Rink TJ. (1987) The kinetics of change in intmcellular calcium in Fun&&loaded human platelets. I. Biol. Chem., 262.1636416369. 26. Blackmore PF. (1988) Hormonal-stimulation of Ca” influx in hcpatocytes by a process not involving inositol lipid breakdown: Possible direct involvement of a G protein. FASEB I., Al343 27. Meyer T. Stryer L. (1988) Molecular model for receptor-stimulated calcium spiking. Pmt. Natl. Acad. Sci. USA., 85, 5051-5055. 28. Cobbold PH. Cuthbertson KSR. Woods NM. (1988) The generation of repetitive free calcium transients in a hormone-stimulated hcpatocyte. In: Nunez I. Dumont JE. Carafoh E. eds. Hormones and Cell Regulation No 12. JNSERMlJohn Libbery Eurotext, pp 135-145. Please send reprint requests to : Dr Niall M. Woods, The Department of Human Anatomy and Cell Biology, The University of Liverpool, P.O. Box 147, Liverpool L69 3BX, UK. Received Accepted

: 6 September 1989 : 20 December 1989