JOURNAL OF CELLULAR PHYSIOLOGY 153:332339 (1992)

Hepatocyte Growth Factor Induces Calcium Mobilization and lnositol Phosphate Production in Rat Hepatocytes G Y O R G Y BAFFY, L I J U N YANG, GEORGE K. M I C H A L O P O U L O S , AND JOHNR. W I L L I A M S O N * Department of Riochcmistry and Biophysics, University of Pennsylvania, Philadelphia, Pennsylvania 19 704-ha89 (G.B., L.Y., ).R. W.), and Department of Pathology, Univeoity of Pittsburgh, Pittsburgh, Pennsylvania 1526 I IC.K.M.)

The effects of hepatocyte growth factor (HGF) on intracellular CaL+ mobilization were studied using fura-2-loaded single rat hepatocytes. Hepatocytes microperfused with different amounts of HGF responded with a rapid concentration-dependent rise in the cytosolic free Ca2+ concentration with a maximum increase of 142% at 80 ng/ml of HCF. The lag period of the Ca" response was decreased with increasing HGF concentrations, being 64 2 12 s, 42 i 6 s, and 14 5 2 s, respectively, with 8, 20, and 80 ngiml of HGF. The detailcd pattern of Ca2+ transients, however, was variable. Out of 16 cells tested using 20 ng/ml ot t iGF, 68% showed sustained oscillatory responses, whereas other cells showed a sustained increase in the cytosolic-frce Ca2 ' upon expowre to HGF, which was dependent on the presence of extracellular Caz+. H C F also induced CaL+ entry across the plasma membrane. Mobilization of Ca'+ by H L F was accompanied by a rapid accumulation of inositol 1,4,5-trisphosphate (Ins 1,4,5-P,). The effects of HGF and epidermal growth factor (EGF) were comparable and partly additive for Ins 1,4,5-P, production and for the sustained phase of Ca" mobilization. Preincubation of cells with 10 pM of genistein to inhibit protein tyrosine kinases abolished thc HCF-induced Ca2 '~response and also inhibited HGF-induced Ins 1,4,5-P, production in rat liver cells. These data indicate that early events in the 5ignal transduction pathways mediated by HGF and EGF have in common the requirements for tyrosine kinase activity, Ins 1,4,5-P, production, and CaL+ mobilization. 8 1992 WiI+ iv,, Inc.

Hepatocyte growth factor (HGF) is a potent mitogen first purified from rat platelets and human and rabbit plasma (Nakamura et al., 1987; Gohda et al., 1988; Zarnegar and Michalopoulos, 1989). HGF is a heterodimeric protein consisting of a and p subunits with apparent molecular weights of 65 and 35 kDa, respectively (Nakamura et al., 1987, 1989; Tashiro et al., 1990). The primary structure of rat HGF has been elucidated by cDNA cloning and shown to have a 38% identity with plasminogen. The a subunit contains four kringle domains, whereas the p subunit shows a n extensive homology with the serine protease domain of plasmin (Tashiro et al., 1990).HGF promotes liver regeneration after partial hepatectomy or toxic liver injury (Gohda et al., 1990; Higuchi and Nakamura, 1991; Ichihara, 1991). Although HGF was originally described as specific for parenchymal hepatocytes, a mitogenic effect of HGF has been also reported in a number of other cell types including nonparenchymal liver cells, melanocytes, keratinocytes, kidney tubular epithelial cells, and mammary gland cells (Igawa et al., 1991; Kan et al., 1991; Matsumoto e t al., 1991; Rubin et al., 1991). Moreover, the mRNA of HGF is expressed in various tissues of rat, including the liver, kidney, lung, and brain (Tashiro et al., 1990). However, the fact that 0 1992 WILEY-LISS, INC

HGF mRNA could be detected only in nonparenchymal cells of liver even under conditions of increased liver regeneration suggests that HGF exerts its effects on hepatocytes in a paracrine fashion (Kinoshita et al., 1989; Michalopoulos, 1990). Rapid tyrosine phosphorylation of target cells stimulated by HGF suggests t h a t the mitogenic signal is mediated through a tyrosine kinase receptor (Rubin et al., 1991). Recent work suggests that the product of the c-met proto-oncogene can serve as a specific receptor for HGF (Bottaro et al., 1991;Naldini et al., 1991).Mitogenesis initiated by different growth factors such as epidermal growth factor (EGF) or platelet-derived growth factor suggests that tyrosine kinase receptors are usually accompanied by cell signaling events associated with enhanced inositol phospholipid metabolism resulting in the formation of Ins 1,4,5-P,, diacylglycerol and inositol phospholipids containing phosphate in the D-3 position of the inositol ring (Carpenter and Cohn, 1990; Ullrich and Schlessinger, 1990; Cantley et al., 1991).

Received February 20.1992; accepted May 25,1992. *:To whom reprint requestsicorrespondence should be addressed

333

HEPATOCYTE GROWTH FACTOR AND (>a"- MOBILIZATION

In hepatocytes, Ins 1,4,5-P, elicits the release of Ca2+ from intracellular vesicular stores (Joseph et al., 1984) and also appears to promote C a Z t entry into the cell (Hansen et al., 19911, whereas diacylglycerol is the well-known activator of protein kinase C (Nishizuka, 1988).These changes, together with those of the 3-phosphorylated phospholipids, have been implicated in the overall mitogenic signal exerted by certain classes of growth factors (Cantley et al., 1991). At present little is known about the early signaling events exerted by HGF. In this report we show that HGF induces Ins 1,4,5-P, production and Ca2+ mobilization in freshly isolated rat hepatocytes and that these responses can be inhibited by pretreatment of the cells with the tyrosine kinase inhibitor genistein.

MATERIALS AND METHODS Isolation of rat hepatocytes Hepatocytes were prepared from the livers of 150250 g male Sprague-Dawley rats by collagenase digestion, according to previously published procedures (Meijer et al., 1975). Cell viability was 90-95%, and freshly isolated cells were resuspended in modified L-15 medium, pH 7.4, supplemented with 15 mM sodium HEPES and 5.5 mM glucose and stored on ice until use (Monck et al., 1988). Measurement of cytosolic free Ca"+ concentration in single hepatocytes Parenchymal hepatocytes were plated onto po1y-Llysine coated round glass coverslips (25 mm diameter) at a density of 2-4 x lo5 cells/cm2 and incubated for 30 min a t 37°C before loading. Fura-2 loading of cells was achieved by incubating the coverslips with 5 p M fura2iAM for 45 min at 37°C. Coverslips with loaded cells were placed in a temperature-controlled cell chamber containing modified Hanks solution (pH 7.4; NaC1,137 mM; KC1, 5.4 mM, NaHCO,, 4.2 mM; KH,PO,, 0.44 mM; Na,HPO,, 0.33 mM; sodium HEPES, 20 mM; MgSO,, 1 mM; CaCl,, 1.3 mM; glucose, 10 mM). Ca2+ was omitted and 1mM EGTA added when a nominally Ca2+-freeextracellular medium was required. The cell chamber was mounted onto a n inverted Nikon microscope equipped for epifluorescence. The fluorescence at 340 and 380 nm excitation was measured, ratioed, digitized, and analyzed a s previously published (Monck et al., 1988). Cytosolic free Ca2+was calculated using the method of Grynkiewicz et al. (1985). Stocks of HGF and other Ca2+-mobilizingagonists were diluted in incubation buffer containing 0.2% bovine serum albumin and expelled with gentle pressure from 10 p m diameter micropipettes attached to a multichannel General Valve picospritzer located near the selected cell.

A bbreuiations

AM acetoxymethyl AngII angiotensin I1 Ins 1,4,5-P, inositol 1,4,5-trisphosphate EGF epidermal growth factor HEPES N-2-hydroxyethylpiperazine-N'-2-ethanesuIfonic acid HGF hepatocyte growth factor THG thapsigargin

TABLE 1. Effect of HGF on Ca2+ mobilization of rat hepatocytes' HGF addition 8 20 80

Cell responsiveness

Lag period

82

64 2 12%

90

42

91

6" 16 2 2 ?

hCa' 99 ? 18 111 ? 13 145 t 28

'Freshly isolaled ral hepalocytes were plaled and loaded with fura-BAM. Basal cytosolic-free Ca2 of unstimulated cells was 102 f 10 nM. Individual cells were microperfused with diffcrcnt conccntrations of IIGF up to 120 s. The time periods elapsed between the start of the stimulation and (he onset of the Ca" response (lay period) and the net changes in the cytosolic free Ca2+ 1 ACa2-, I are summarized. Data show the mean 2 SEM of measurements on a t least 15 individual cells from three scparatc cxperiments 'Significantly differs from rcsults obtained with 80 ng:ml HGF tP .< 0.01).

'

Measurement of Ins 1,4,5-P3production Isolated hepatocytes were preincubated a t a density of 5 x 10" cellsiml a t 37°C in L-15 medium for 30 min. Aliquots (5 x lo6) of the hepatocyte suspension were treated with HGF (20 ngiml) or EGF (60 ng/ml) for different times, followed by addition of 10 ml of ice-cold phosphate-buffered saline (PBS). The cells were centrifuged and extracted with ice-cold perchloric acid (10% wiv). The supernatants were centrifuged and neutralized with 1.5 M KOHIGO mM HEPES to pH 7.0. Ins 1,4,5-P, was assayed using a 13Hl radioreceptor assay kit (NEN-DuPont). Materials Rat HGF was purified as previously described (Zarnegar and Michalopoulos, 1989). Fura-2iAM was obtained from Molecular Probes (Eugene, OR) and stored as 5 mM stock in dimethyl sulfoxide. Genistein was purchased from UBI (Lake Placid, NY) and stored as 50 mM stock in dimethyl sulfoxide. All other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO). RESULTS Effects of HGF on CaZ+mobilization in rat hepatocytes In order to measure changes of cytosolic free Ca2' induced by HGF, the hormone was microperfused for 60-100s onto single fura-2 loaded r a t hepatocytes at concentrations of 8, 20, or 80 ngiml. Typically, there was a lag period after exposure of the cell to HGF that was followed by an abrupt increase in the cytosolic free CaZ+to a peak value within 1 s. With HGF concentrations of 8 ngiml and above, between 80% and 90% of the cells tested responded with a Ca2+ transient. At HGF concentrat,ions of 2 ng/ml or lower, consistent effects on the cytosolic free Ca2' were not observed since less than 50% of cells tested gave a response. In cells that did respond, there was a n extensive lag period and the maximum increase of the cytosolic free Ca2' was less than that observed with 8 ngiml of HGF. Table 1 summarizes results obtained with 8 ng/ml of HGF and above. The peak CaZ+increased slightly with increasing HGF concentrations, but the greatest effect was on the latency period, measured as the time elapsed between the start of microperfusion and the onset of the Ca2+ rise, which decreased from 64 s to 16 s with a tenfold increase of HGF from 8 ngiml (Table 1).

334

BAFFY

m-AL

7 0

70

140

2 i0

1

HGF

280

Time (sec)

0

60

120

180

240

Time (sec)

"'1

HGF [without Ca-1 I

n l 0

100

200 Time (sec)

300

403

t

Ca+' -

8

0 300

1

70

2 10

140

280

Time (sec) Fig. 2. Sustained phase of Ca2+transients of rat hepatocytes stimulated by HGF depends on extracellular Ca" . A. Two representative traces obtained by HGF stimulation (80 ngiml) with 1.3 mM of extracellular-free Ca2+ concentration or in a medium nominally free of Ca2+(Ca2+, < 1 KM).Experiments typical of 5. B. Cells were allowed to replenish with Ca2+ by incubation in Ca2+-containing Hanks' buffer for 10 min. An abbreviated Ca2+transient was obtained when cells were microperfused with HGF in Ca2--free medium for 3 min. When microperfusion was stopped so that the cells were re-exposed t o normal Ca2-, the cytosolic free Ca" increased to its previous peak value suggesting a Ca" influx due to previous HGF stimulation. Experiment typical of 4. +

loo

10

70

140

2 10

280

Time (sec) Fig. 1. Various Ca2+ transient patterns of single fura-2-loaded rat hepatocytes stimulated with HGF. Representative traces obtained by 60 s microperfusions of 20 ng/ml HGF selected from more than 40. A. Brief burst of Ca2+ spikes followed by a return of the cytosolic free Caz+ to the baseline. B. Series of rapid Ca2 ' spikes superimposed in CaZ+oscillations with a periodicity of 1to 2 min. C. Rapid increase of the cytosolic free Ca" ' to a plateau level maintained for a t least 5 min.

When HGF (20 ngiml) was microperfused onto single hepatocytes for a standard time of 60 s, there was a marked heterogeneity of the individual Ca2+ responses. As illustrated in Figure 1,three characteristic types of response were observed. The majority of the cells tested (11of 16) responded as in Figure 1B with a series of rapid Ca2+ spikes, about 10 s apart, superimposed on Ca2+ oscillations with a periodicity of 1 to 2

min. Other cells (3 of 16) responded with a brief burst of Ca2+ spikes followed by a return of the cytosolic-free Ca2+to the baseline (Fig. lA), whereas in a third group (2 of 16 cells) Ca2+ increased rapidly to a plateau level, which was maintained for at least 5 min (Fig. 1C). The sustained phase of the HGF-induced Ca2+ plateau was dependent on the presence of extracellular Ca2+,as indicated in Figure 2A, which shows that the Ca2+ transient decreased to resting Ca2+ levels within 90 s of its onset when HGF (20 ng/ml) was microperfused onto cells in nominally Ca2+-freemedium. Likewise, a n abbreviated Ca2+transient was obtained when HGF was microperfused from a pipette containing Ca2+-freemedium onto the cells that were, however,

HEPATOCYTE GROWTH FACTOR AND Ca2 ' MOBILIZATION

O! 0

90

iao

270

0

360

80

500

320

240

160

D

335

HG F

1

I

I

300 J

0

60

120

180

0 240

Time (sec)

0

7c

140

2 :c

286

Time jsec;

Fig. 3. HGF induces incomplete release of Ca" from intracellular pools of rat hepatocytes. A. In a nominally Ca' ' -free medium, THG (2 pM) depleted intracellular Ca2+in a few minutes. B. Addition of 2 pM THG aftcr a prior stimulation with 80 ngiml HGF produced a second Ca2 ' transient, which was, however, abbreviated. C. Prior stimula-

tion with 10 nM angiotensin I1 (AngIIi prevented THG from eliciting another increase in cytosolic free Ca2 ' . D. HGF (80 ng/ml) caused essentially no further Ca2' mohilization if microperfused onto cells aftcr a prior addition of2 pM THG. Experiments typical of 4.

incubated with normal Ca2+(Fig. 2B). When microperfusion with the Ca' ' -free HGF medium was stopped so that the cell was re-exposed to normal Ca2+,the cytosolic free Ca2+ increased to its previous peak value. These studies show that HGF induced both a mobilization of intracellular Ca" and a n increased Ca2+ entry into the cell. In order to determine the extent of depletion of intracellular Ca2+ stores produced by HGF, hepatocytes were placed in nominally Ca2' -free medium and stimulated first with HGF and then with different C a 2 ' mobilizing agents. Thapsigargin (THG), which is known to be effective in depleting intracellular Ca2+ stores, was added to the cells at a concentration of 2 FM after a previous HGF-induced Ca2+ transient had returned to the original baseline. Figure 3A shows that microperfusion with THG alone induced a rapid increase of cytosolic-free Ca2+followed by a short plateau and return to the baseline before microperfusion was terminated. Addition of THG after a prior stimulation with 80 ngiml of HGF produced a second Ca" transient (Fig. 3B), which was, however, smaller than that in Figure 3A. However, prior stimulation of cells with 10 nM angiotensin I1 prevented THG from eliciting another increase in cytosolic free Ca2+ (Fig. 3C). When HGF (Fig. 3D) or angiotensin I1 (not shown) were added

after THG, there was essentially no further Ca2+mobilization. These results show that under the conditions of these experiments, whereas angiotensin I1 and thapsigargin caused a complete depletion of the intracellular hormone-sensitive Ca2+ pool and mutually excluded the effect of each other on the rise of the cytosolic free Ca", HGF caused only a partial depletion of the pool.

Inhibitory effects of genistein on HGF-induced changes of Ins 1,4,5-P, production and Ca2+ mobilization in rat hepatocytes In order to determine whether HGF-induced Ca2' mobilization, like that induced by EGF, was caused by a receptor-mediated activation of phospholipase C and production of Ins 1,4,5-P,, the time course of Ins 1,4,5-P3production was measured in hepatocytes after addition of HGF (Fig. 4A). A threefold increase of Ins 1,4,5-P, mass was observed within 30 s of HGF addition and was maintained for at least 5 min. After 30 min, however, Ins 1,4,5-P3 levels had fallen to control values. The HGF-induced increase of Ins 1,4,5-P, was dependent on tyrosine kinase activity since inhibition of tyrosine kinases by pretreatment of the hepatocytes for 30 min with 10 pM genistein caused a complete abrogation of the HGF-induced increase of Ins 1,4,5-P, (Fig.

BAFFY ET AL

mobilization induced by angiotensin I1 (Fig. 5). Further experiments showed that when control cells were incubated for 30 min with 0.2% DMSO, angiotensin I1 produced a n increase in the cytosolic free Ca2 ' of 249 t 21 nM (n = 81, whereas after incubation for 30 min with 10 FM enistein, angiotensin I1 increased the cytosolic free Ca by 297 17 nM (n = 51, a difference that was not statistically significant.

B

A control genistein-treated

f+

1

0

30"

1'

5'

t'

30'

1'

ECF ECF

HCF

+ HCF

Fig. 4. HGF-induced Ins 1,4,5,-P,formation in rat hepatocytes: Time course, inhibition by genistein, and additive effect of EGF. A. Cells were stimulated with 20 ng/ml HGF and Ins 1,4,5-P, production was measured by I3H1 radioreceptor assay at different times. Results are expressed a s the mean ? SEM of three experiments (striped bars). For inhibition studies, cells were pretreated for 30 rnin with 10 +M genistein (filled bars). B. Control cells (striped bars) and cells pretreated for 30 rnin with 10 pM genistein (filled bars) were stimulated either with 60 ngiml EGF alone or simultaneously with 20 ng!ml HGF and Ins 1,4,5-P, production was measured by ['HI radioreceptor assay. Results obtained after 1 min of stimulation are expressed as the mean 2 SEM of three experiments.

/ 10G

-S

c

*

Comparison of effects of HGF and EGF on Ca2+ mobilization and Ins 1,4,5-P, production Epidermal growth factor (EGF) is widely known to act through its specific tyrosine kinase receptor and the early biochemical changes initiated by EGF in many responsive cells include a rise of the cytosolic free Ca2 ' produced by the mobilization from Ins 1,4,5-P3-sensitive stores and sustained by a n increased influx of extracellular Ca2+(Yarden and Ullrich, 1988; Carpenter and Cohn, 1990).The effects of HGF and EGF on DNA synthesis of primary cultured rat hepatocytes were found to be additive (Nakamura et al., 1989). Possible additive effects of HGF and EGF on Ins 1,4,5-P, production and Ca2+ mobilization were studied, therefore, in isolated rat hepatocytes. Figure 4B shows that Ins 1,4,5-P, production was comparable 1min after EGF to that seen 1min after HGF addition and that when both growth factors were added together, Ins 1,4,5-P, production was 55-75% greater than when either was added separately. Genistein treatment completely inhibited Ins 1,4,5-P, production induced by EGF alone or when EGF and HGF were added together (Fig. 4B). Figure 6 shows representative effects of EGF and HGF, added separately or together at maximally effective concentrations, on Ca2t mobilization using fura-2 loaded cells. Microperfusion of EGF onto hepatocytes incubated in normal Ca2+ medium caused a somewhat higher Ca2+transient than HGF, but the major effect of combined EGF and HGF microperfusion was to elevate the plateau phase of the Ca2' transient (Fig. 6A). When cells were incubated in nominally Ca2+-freemedium, it was apparent that HGF mobilized less Ca2+than EGF alone or EGF with HGF (Fig. 6B). Repetit,ion of the experiments with cells in Ca2+-containing medium showed that the addition of HGF together with EGF produced a n increase of cytosolic-free Ca2+ compared with either HGF or EGF alone (247 ? 16 nM, n = 6, for H G F + EGF vs. 199 6 nM, n = 15, for EGF and 145 -+ 28 nM, n = 10, for HGF). In order to characterize further possible differential effects of HGF and EGF on Ca2' influx, a different protocol was adopted. When a plateau of cytosolic free Ca2+ transient was reached by stimulation with maximally effective concentration of HGF, microperfusion was continued from another pipett,e containing both EGF and HGF. As shown in Figure 7, if EGF was applied at the plateau of a n HGF-induced Ca2+transient, a further increase in cytosolic free Ca2' was observed, suggesting that EGF was more effective than HGF in enhancing Ca2+entry into hepatocytes.

*

90

:a0

27C

360

Time (sec) Fig. 5. Genistein inhibits HGF-induced Ca2+mobilization in rat hepatocytes. Cells were stimulated with 20 ng/ml HGF for a t least 4 min after 30 min preincubation with 10 FM genistein at 37°C. HGF-induced Ca2 ' response was completely abolished. However, 10 nM angiotensin I1 (AngII) induced large Ca2 transient after the ineffective stimulation with HGF. Experiment typical of 18. +

4A). Genistein also exerted a n inhibitory effect on DISCUSSION HGF-induced Ca2+ response, a s shown in Figure 5. Different growth factor receptors possessing tyrosine Prior incubation of cells with 10 p.M genistein almost completely prevent,ed the Ca2+ mobilizing effect of 80 kinase activity transduce early signaling events when ng/ml HGF. Genistein, however, had no effect on Ca2+ cells are suitably stimulated (Yarden and Ullrich,

HEPATOCYTE GROWTH FACT(3R AND Ca2' MOBILIZATION

A

4007

337

EGF+ HGF

300 -

z

v

200-

s ._

V

0

0

m #vwJt!

100

100

0

40

80

120

160

.

50

100

I

50 _.

100

150 I

200

1

Fig. 7. Effects of HGF and EGF on the Ca2+influx into rat hepatocytes. Cells were stimulated first with 80 ngiml HGF. On the plateau of the HGF-induced CaZ' transient, perfusion was continued from another micropipette containing both HGF (80 ngiml) and EGF (60 ng!ml). Experiments typical of 4.

I

0 0

01 0

Time (sec)

Time (sec)

4001

p

150

200

Time (sec) Fig. 6 . Effects of HGF and EGF on the Ca2+ mobilization in rat hepatocytes. Cells were stimulated with either 80 n g h l HGF or 60 ng/ml EGF, or both. Ca2+ transients are compared in the presence (Ca"', = 1.3 mM) (A) or in the absence of extracellular Ca2(Ca2 c: 1 *MI (B). Experiments typical of 6-15. +

1988; Carpenter and Cohn, 1990). These early events include Ca2+mobilization from intracellular stores and influx of external Ca2+ into the cell. HGF is a very potent hepatotrophic mitogen that does not share any sequence homology with the various other known growth factors. It appears instead to be related to the family of serine proteases (Nakamura et al., 1989; Rubin et al., 1991). However, the recently described interaction of HGF with the c-met tyrosine kinase receptor suggests a signal transduction process similar to that of EGF (Bottaro et al., 1991; Naldini et al., 1991). Studies reported here show that stimulation of freshly isolated rat hepatocytes by HGF leads t o Ca"+ mobilization. HGF-induced CaZ+ transients occurred within 1 min of hormone addition and were seen a t concentrations of HGF from 8 to 80 ngiml. Similarly, the purified rat HGF preparation used for these experiments had a maximum mitogenic effect between 40 and 80 ngiml (G.K. Michalopoulos, unpublished observa-

tion). The nonoscillatory Ca2+ response of rat hepatocytes to HGF observed in Figure 1C is a typical biphasic response similar to that. observed in a variety of cell types with many hormones and neurotransmitters (Williamson and Monck, 1989; Hughes et al., 1990). The initial abrupt increase of Ca2+ results from the release of Caa+ from intracellular stores, which is followed by Ca2+entry from the extracellular space. The mechanism of intracellular Ca2+ release induced by HGF appears to be the same as that reported for many other agonists, namely, by means of the Ca2+ mobilizing second messenger Ins 1,4,5-P,, which is produced by a phospholipase C-mediated breakdown of phosphatidylinositol4,5-bisphosphate(Berridge, 1987; Berridge and Irvine, 1989). The present study shows that HGF stimulation of rat hepatocytes resulted in a rapid and transient accumulation of Ins 1,4,5-P,. Since Ins 1,4,5-P3 production was measured in hepatocyte suspensions, whereas Ca2+ mobilization was measured in single cells, a direct comparison of the kinetics or concentration dependency of the changes in these parameters induced by HGF cannot be made, but the changes are compatible with a cause and effect relationship between Ins 1,4,5-P accumulation and Ca2' mobilization. Entry of Ca2'is required for sustained Ca2+oscillations or a n elevated plateau level of Ca2+, but the mechanism of HGF-induced Ca2+ entry into hepatocytes is presently unknown. Recent studies with hepatocytes have provided evidence in favor of both a n Ins 1,4,5-P3-gated Ca2+ entry mechanism (Hansen et al., 1991) and a capacitative mechanism (Zhang et al., 1991); the latter model relating to the proposal that Ca2+ entry into cells is secondary to depletion of the hormone-sensitive Ca2+pool (Putney, 1986, 1990). The pattern of Ca2+ transients varied greatly from cell to cell, but was characterized by different types of oscillatory responses (Figs. 1,2). A wide range of Ca2 ' oscillatory behavior has been observed with a number

BAFFY

338

of agonists in different cell types, including those characterized by spikes superimposed on oscillatory waves a s in Figures 1 and 2 (Berridge and Galione, 1988; Dixon et a]., 1990). The chemical basis for oscillations of a component in a pathway is the necessity for positive feedback in the system. Superimposed on this requirement there may be negative feedback or further secondary positive feedback influences. Several mathematical models have been presented that are able to represent the different kinds of observed agonist-dependent fluctuations of cytosolic free Ca2+ in single cells (Meyer and Stryer, 1988; Chay, 1990; Goldbeter et al., 1990; Swillens and Mercan, 1990; Cuthbertson and Chay, 1991). These, however, are more useful in demonstrating what type of interactions may or may not occur in order to account for Ca2+ oscillations than in simulating known biochemical mechanisms. Fundamentally, the models illustrate that nonsinusoidal Ca2&spiking, which is a characteristic of many observed Ca2+ transients, can be generated by positive feedback effects at the level of G-protein or phospholipase C interactions or by a positive interaction of Ca2' itself, as in some form of Ca"-induced Ca2+release or Ca2+influx. The former type involves Ins 1,4,5-P3a s a n oscillatory component, whereas the Cast pool oscillators act downstream from Ins 1,4,5-P3 generation and can account for induction of Ca2' oscillations by a nonhydrolyzable analogue of Ins 1,4,5-P3 (Wakui et al., 1989). A combination of these two types of oscillation generators may be present in the same cell, thereby accounting for the special type of Ca2+spikes with a 1s frequency superimposed on Caz+oscillations with a frequency of about 100 s (Fig. 1).The physiological relevance and purpose of Ca2+ oscillations are still being debated, but on the assumption that individual cells in the intact tissue can exhibit synchronous Ca2+ oscillations, the suggest,ion that the frequency of Ca2' oscillations effectively regulates the average phosphorylation state of proteins (Goldbeter et al., 1990) is particularly appealing for transmission of effects mediated by growth hormones. The present study shows that both Ins 1,4,5-P, formation and Ca2 mobilization induced by HGF in hepatocytes were inhibited by genistein. Genistein is a specific inhibitor of tyrosine kinases acting within minutes of addition (Akiyama et al., 1987). Its inhibitory action was previously described on receptors of EGF and platelet-derived growth factor (Dean et al., 1989). Genistein thus appears to have a n inhibitory action on c-met receptor tyrosine kinase serving as a specific receptor for HGF. Similar inhibitory effects of genistein have been observed in hepatocytes treated with EGF (Yang et al., 1991b). There is in fact a striking similarity between EGF and HGF in these early signaling events, but it remains to be determined whether tyrosine phosphorylation of PLC-y is involved in the initiation of phosphatidylinositol lipid breakdown by HGF, a s it appears t o be with EGF (Yang et al., 1991a). Of further interest is the finding that the effects of HGF and EGF on Ins 1,4,5-P3 production and Ca2' mobilization were comparable but not quite additive. This suggests that the two receptors may be couPled t o PIP,-breakdown by the Sam; signal& pathway(s). EGF coupling to phospholipase c in hepatocytes +

m AL. requires the participation of a Gi-protein (Liang and Garrison, 1991; Yang et al., 1991a), whereas the details of the HGF coupling pathway are currently unknown. Alternatively, the effects of HGF and EGF on Ca2+ mobilization as well as on DNA synthesis (Nakamura et al., 1989) may be limited by receptor number. Thapsigargin, a tumor promoting sesquiterpene lactone, specifically inhibits the Ca2+-ATPaseof the endoplasmic reticulum and discharges Ca2' fi-om the Ins 1,4,5-P3-sensitive intracellular stores (Thastrup et al., 1990; Lytton et al., 1991). A similar mechanism of action has been observed in hepatocytes (Llopis et al., 1991). Compared with angiotensin 11, which prevented thapsigargin from inducing a second Ca2+transient in Ca"-free medium, HGF appears to be unable to deplete the intracellular Ins 1,4,5-P3-sensitive Ca2+ stores completely. Again, this may reflect a relatively low density of HGF receptors in freshly isolated hepatocytes or differences in the ability of the two hormones t o resequester CaZ+into intracellular stores. Taken together, our dat,a show that HGF has much in common with EGF in being able to induce a typical biphasic Ca" ' response in freshly isolated hepatocytes along with the accumulation of Ins(1,4,5)P,. Furthermore, these early HGF-induced signaling events can be inhibited by genistein, supporting the role of a tyrosine kinase-dependent mechanism in HGF signal transduction.

ACKNOWLEDGMENTS This work was supported by National Institutes of Health Grant DK 15120. G.B. is a recipient of a n NIH Fogarty International Center Fellowship (F05 TW04545). LITERATURE CITED Akiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Watanabe, S., Itoh, N., Shibuya, M., and Fukami, Y. (1987) Genistein, a specific inhibitor of tyrosine-specific protein kinases. J. Biol. Chem., 262r5592-5595. Berridge, M.J. (1987) Inositol trisphosphate and diacylglycerol: Two interacting second messengers. Annu. Rev. Biochem., 56:159-193. Berridge, M.J., and Galione, A. 11 988) Cytosolic calcium oscillators. FASEB J.,2r307P3082. Berridge, M.J., and Irvine, R.F. (1989) Inositol phosphates and cell signalling. Nature, 341.197-205. Bottaro, D.P., Rubin, J.S., Faletto, D.L., Chan,A.M.L., Kmiecik, T.E., Vande Woude, G.F., and Aaronson, S.A. (1991) Identification ofthe hepatocyte gruwth factor receptor as the c-met proto-oncogene product. Science, 251t802-804. Cantley, L.C., Auger, K.R., Carpenter, C., Duckworth, B., Graziani, A., Kapeller, R., and Soltoff, S. (1991) Oncogenes and signal transduction. Cell, 64:281302. Carpenter, G I and Cohn, S. (1990) Epidermal growth factor. J. Biol. Chem., 265t7709-7712. Chay, T. (1990) Effect of compartmentalized Ca"' ions on electrical bursting activity of pancreatic p-cells. Am. J . Physiol., 258C955C965. Cuthbertson, K.S.R., and Chay, T.R. (1991) Modelling receptor-controlled intracellular calcium oscillators. Cell Calcium, 12:97-109. Dean, N.M., Kanemitsu, M., and Boynton, A.L. (1989) Effects of the tyrosine-kinase inhibitor genistein on DNA synthesis and phospholipid-derived second messenger generation in mouse 10T112 fibroblasts and rat liver T51B cells. Biochem. Biophys. Res. Commun., 1135:795-801. Dixon, C.J., Woods, N.M., Cuthbertson, K.S.R., and Cobbold, P.H. (1990) Evidence for two Ca' ' -mobilizing purinoceptors on rat hepatocvtes. Biochem. .J.. 269:499-502. Gohda, E., Tsubouchi, H., Nakayama, H., Hirono, S., Sakiyama, O., Takahashi. K., Miyazaki, H., Hashimoto, S., and Daikuhara, Y.

HEPATOCYTE GROWTH FACTOR AND Ca'- MOBILIZATION

339

(1988) Purification and partial characterization of hepatocyte Michalopoulos, G.K. (1990) Liver regeneration: molecular mechanisms of growth control. FASEB J., 4: 176-187. growth factor from plasma of a patient with fulminant hepatic failure. J. Clin. Invest., 81t414-419. Monck, J.R., Reynolds, E.E., Thomas, A.P., and Williamson, J.R. Gohda, E., Hayashi, Y., Kawaida, A., Tsubouchi, H., and Yamamoto, I. (1988) Novel kinetics of single cell Ca2+ transients in stimulated (1990) Hepatotrophic growth factor in blood of mice treated with hepatocytes and A10 cells measured using fura-2 and fluorescent carbon tetrachloride. Life Sci., 46r1801-1808. videomicroscopy. J. Bid. Chem., 263:45694575. Goldbeter, A,, Dupont, G., and Berridge, M.J. (1990) Minimal model Nakamura, T.. Nawa? K., Ichihara, A., Kaise, N., and Nishino, T. for signal-induced Ca2 ' oscillations and for their frequency encod(1987) Purification and subunit structure of hepatocyte growth factor from rat platelets. FEBS Lett., 224:311-316. ing through protein phosphorylation. Proc. Natl. Acad. Sci. USA, Nakamura, T., Nishizawa, T., Hagiya, M., Seki, T., Shimonishi, M., 87t1461-1465. Sugimura,A,, Tashiro, K., and Shimizu, S. (1989) Molecular cloning Grvnkiewicz. G.. Poenie, M.. and Tsien. R.Y. (1985) A new generation and expression of human hepatocyte growth factor. Nature, of Ca2- indicators with greatly improved fluorescence properties. J. 342,440443. Biol. Chem., 260t3440-3450. Hansen. C.A.. Y a w . L.. and Williamson. J.R. (1991) Mechanisms of Naldini, I+ Vigna, E., Narsimhan, R.P., Gaudino, G., Zarnegar, R., Michalopoulos, G.K., and Comoglio, P.M. (1991) Hepatocyte growth receptor-mediate;l.Ca"+ signaling in rat hepatocytes. J. Biol. Chem., factor (HGF) stimulates the tyrosine kinase activity of the receptor 266: 18573-18579. encoded by the proto-oncogene c-MET. Oncogene, 6.501-504. Hieuchi, 0..andNakamura. T. (1991)Identification and change in the Nishizuka, Y. (1988) The molecular heterogeneity of protein kinase C receptor for hepatocyte growth factor in rat liver after partial hepaand its implications for cellular regulation. Nature, 334:661465. tectomy or induced hepatitis. Biochem. Biophys. Res. Commun., Putney, J.W., J r . (1986)A model for receptor-regulated calcium entry. 176:59%607. Cell Calcium, 7:l-12. Hughes, A.R., Horstman, D.A., Takemura, H., and Putney, J.W., J r . Putnev. J.W.. Jr. (1990) Canacitative calcium entrv revisited. Cell (19901 Inositol phosphate metabolism and signal transduction. Am. Cal&m, i i t 6 i i - m 4 . Rev. Respir. Dis., 141ts115-sl18. Rubin, J.S., Chan, A.M.L., Bottaro, D.P., Burgess, W.H.,Taylor, W.G., Ichihara, A. (1991)Mechanisms controlling growth of hepatocytes in Cech. A.C.. Hirschfield. D.W.. Wow. J.. Miki. T.. Finch. P.W.. and primary culture. Dig. Dis. Sci., 36t489-493. Aaronson, 'S.A. (1991)'A broad-spectru'm human lung'fibroblastIgawa, T., Kanda, S., Kanetake, H., Saitoh, Y., Ichihara, A,, Tomita, derived mitogen is a variant ofhepatocyte growth factor. Proc. Natl. Y., and Nakamura, T. (1991) Hepatocyte growth factor is a potent Acad. Sci. USA, 88:415419. mitogen for culture rabbit renal tubular epithelial cells. Biochem. Swillens, S., and Mercan, D. (1990)Computer simulation of acytosolic Biophys. Res. Commun., 174t831-838. calcium oscillator. Biochem. J.,27lt83.5-838. Joseph, S.K., Thomas, A.P., Williams, R.J., Irvine, R.F., and Williamson, J.R. (1984) myo-Inositol 1,4,5-trisphosphate: a second messen- Tashiro, K., Hagiya, M., Nishizawa, T., Seki. T., Shimonishi, M., Shimizu, S., and Nakamura, T. (1990)Deduced primary structure of ger for the hormonal mobilization of intracellular Ca2+ in liver. J . rat hepatocyte growth factor and expression of the mRNA in rat Biol. Chem.,259:3077-3081. tissues. Proc. Natl. Acad. Sci. USA, 87t3200-3204. Kan, M., Zhang, G.H., Zarnegar, R., Michalopoulos, G.K., Myoken, G., Thastrup, O., Cullen, P.J., Drohak, B.K., Hanky, M.R., and Dawson, McKeehan, W.L., and Stevens, J.I. (1991)Hepatocyte growth factor! A.P. (1990) Thapsigargin, a tumor promoter, discharges intracelluhepatopoietin A stimulates the growth of rat kidney proximal tular Ca2- stores by specific inhibition of the endoplasmic reticulum bule epithelial cells (RPTE),rat nonparenchymal liver cells, human Ca2' -ATPase. Proc. Natl. Acad. Sci. USA, 872466-2470. melanoma cells, mouse keratinocytes and stimulates anchorageindenendent m-owth of SV-40 transformed RPTE. Biochem. Bio- Ullrich, A,, and Schlessinger, J. (1990) Signal transduction by receptors with tyrosine kinase activity. Cell, 61.203-212. phys'. Res. CoLmun., I74:331-337. Kinoshita, T., Tashiro, K., and Nakamura, T. (1989) Marked increase Wakui. M., Potter, B.V.L., and Pctersen, O.H. (1989) Pulsatile intracellular calcium release does not depend on fluctuations in inositol of HGF mRNA in non-parenchymal liver cells of rats treated with trisphosphate concentration. Nature, 3.?9:317320. hepatotoxins. Biochem. Biophys. Res. Commun., 165t1229-1234. Williamson, J.R., and Monck, J.R. (1989) Hormone effects on cellular Liang, M., and Garrison, J.C. (1991) The epidermal growth factor calcium fluxes. Annu. Rev. Physiol., 5Ir107-124. receptor is coupled t o a pertussis toxin sensitive guanine nucleotide Yang, L., Baffy, G., Rhee, S.G., Manning, D., Hansen, C.A., and regulatory protein in rat hepatocytes. J. Biol. Chem., 266:13342Williamson, J.R. (1991a) Pertussis toxin-sensitive Gi protein in13349. volvement in epidermal growth factorinduced activation of phosLlopis, J., Chow, S.B., Kass, G.E., Gahm, A., and Orrenius. S. (1991) pholipase c-gamma in rat hepatocytes. J . Biol. Chem., 266t22451Comparison between the effects of the microsomal Ca"'-translocase inhibitors thapsigargin and 2,5-di(t-butyl)-l,4-benzohydroquinone 22458. Yang, L., Camoratto, A.M., and Williamson, J.R. (1991b)EGF-depenon cellular calcium fluxes. Biochem. J.,277553-556. dent association of Gi-alpha with EGF-R and PLC-gamma in priLytton, J.,Westlin, M., and Hanky, M.H. (1991) Thapsigargin inhibmary culture hepatocytes. J . Cell Biol., 115r25a. its the sarcoplasmic or endoplasmic reticulum Ca-ATPase family of Yarden, Y., and Ullrich, A. (1988) Growth factor receptor tyrosine calcium pumps. J . Biol. Chem., 266t17067-17071. kinases. Annu. Rev. Biochem., 57t443-478. Matsumoto, I

Hepatocyte growth factor induces calcium mobilization and inositol phosphate production in rat hepatocytes.

The effects of hepatocyte growth factor (HGF) on intracellular Ca2+ mobilization were studied using fura-2-loaded single rat hepatocytes. Hepatocytes ...
846KB Sizes 0 Downloads 0 Views