Guanine nucleotide-dependent pertussis-toxin ... - Europe PMC

141

Biochem. J. (1986) 239, 141-146 (Printed in Great Britain)

Guanine nucleotide-dependent pertussis-toxin-insensitive stimulation of inositol phosphate formation by carbachol in membrane preparation from human astrocytoma cells

a

John R. HEPLER and T. Kendall HARDEN* Program in Neurobiology and Department of Pharmacology, University of North Carolina School of Medicine, Chapel Hill, NC 27514, U.S.A.

The efficacy of muscarinic-receptor agonists for stimulation of inositol phosphate formation and Ca2+ mobilization in intact 1321N1 human astrocytoma cells is correlated with their capacity for formation of a GTP-sensitive high-affinity binding complex in membranes from these cells [Evans, Hepler, Masters, Brown & Harden (1985) Biochem. J. 232, 751-757]. These observations prompted the proposal that a guanine nucleotide regulatory protein serves to couple muscarinic receptors to the phospholipase C involved in phosphoinositide hydrolysis in 1321N1 cells. Inositol phosphate (InsP) formation was measured in a cell-free preparation from 1321N1 cells to provide direct support for this idea. The formation of InsP3, InsP2 and InsP, was increased in a concentration-dependent manner (Ko.5 - 5 /iM) by guanosine 5'-[ythio]triphosphate (GTP[S]) in washed membranes prepared from myo-[3H]inositol-prelabelled 132IN1 cells. Both GTP[S] and guanosine 5'-[Ify-imido]triphosphate (p[NH]ppG) stimulated InsP formation by 2-3-fold over control; GTP, GDP and GMP were much less efficacious. Millimolar concentrations of NaF also stimulated the formation of inositol phosphates in membrane preparations from 1321N1 cells. In the presence of 10,1M-GTP[S], the muscarinic cholinergic-receptor agonist carbachol stimulated (KO5 10 ,uM) the formation of InsP above that achieved with GTP[S] alone. The effect of carbachol was completely blocked by atropine. The order of potency of nucleotides for stimulation of InsP formation in the presence of 500 /M-carbachol was GTP[S] > p[NH]ppG > GTP = GDP. Pertussis toxin, at concentrations that fully ADP-ribosylate and functionally inactivate Gi (the inhibitory guanine nucleotide regulatory protein), had no effect on InsP formation in the presence of GTP[S] or GTP[S] plus carbachol. These data are consistent with the idea that a guanine nucleotide regulatory protein that is not Gi is involved in receptor-mediated stimulation of InsP formation in 1321N1 human astrocytoma cells.

INTRODUCTION The mechanism whereby activation of hormone receptors results in an increase in the conversion of phosphatidylinositol bisphosphate into InsP3 has not been defined. However, this process probably involves a guanine nucleotide regulatory protein. GTP or its analogues have been shown to promote exocytotic release of substances from mast cells (Gomperts, 1983) and platelets (Haslam & Davidson, 1984). Pertussis toxin, which ADP-ribosylates and inactivates Gi, also has been shown to modify the phosphoinositide response to hormones in several cells of bone-marrow origin (Brandt et al., 1985; Ohta et al., 1985). Finally, several laboratories have demonstrated guanine nucleotidemediated stimulation of phosphoinositide hydrolysis in cell-free preparations (Litosch et al., 1985, 1986; Cockroft & Gomperts, 1985; Smith et al., 1985; Wallace & Fain, 1985; Uhing et al., 1985; Gonzales & Crews, 1985). Activation of muscarinic cholinergic receptors on 1321N1 human astrocytoma cells results in hydrolysis of phosphoinositides (Masters et al., 1984), mobilization of Ca2+ (Masters et al., 1984) and activation of a

Ca2+-calmodulin-regulated phosphodiesterase (Meeker

& Harden, 1982; Tanner et al., 1986). These receptors do not interact with adenylate cyclase (Meeker & Harden, 1982; Hughes et al., 1984) or Gi (Hughes et al., 1984; Evans et al., 1985a). Nonetheless, the muscarinic receptors of 1321N1 cells apparently interact with a guanine nucleotide regulatory protein, as evidenced by marked effects of guanine nucleotides on the binding of muscarinic-receptor agonists in washed membranes from these cells (Evans et al., 1985a,b). Circumstantial evidence of a role for this putative protein in coupling muscarinic receptors to the phospholipase C involved in phosphoinositide hydrolysis has been provided by the demonstration that the capacity of agonists to form a GTP-sensitive high-affinity binding complex in membranes from these cells is correlated with their efficacy for stimulation of inositol phosphate formation and Ca2+ mobilization (Evans et al., 1985b). We now provide direct evidence for the involvement of a guanine nucleotide regulatory protein by demonstrating that carbachol-stimulated inositol phosphate formation in membranes from 1321N1 cells is guanine nucleotidedependent. Furthermore, fluoride also is shown to stimulate inositol phosphate formation in membrane preparations from these cells. Consistent with previous results with this cell line (Masters et al., 1985; Martin

Abbreviations used: InsP., InsP2, InsP1, inositol tris-, bis- and mono-phosphate; GTP[S], guanosine 5'-[y-thioltriphosphate; p[NH]ppG, guanosine 5'-[f8y-imido]triphosphate; Gi, the inhibitory guanine nucleotide regulatory protein. * To whom reprint requests should be sent.

Vol. 239

142

J. R. Hepler and T. K. Harden

et al., 1985; Nakahata et al., 1986), data are presented that indicate that this putative guanine nucleotide regulatory- protein is not Gi. MATERIALS AND METHODS 1321NI human astrocytoma cells were cultured as previously described (Meeker & Harden, 1982). Cells were grown to a confluent monolayer in 140 mm-diam. dishes and then labelled for 48 h with 5 1tCi of myo-[3H]inositol/ml in inositol-free Dulbecco's minimal essential medium containing 500 fetal-calf serum. Approx. 15-30 assays were carried out per dish of confluent cells (1 50-300 ,ug of protein/assay). To prepare 132 IN1 membranes, the labelling medium was removed and replaced by cold hypo-osmotic lysis buffer (10 mmHepes/2 mm EGTA, pH 7.0). Cells were incubated in lysis buffer at 4 °C for 15-20 min, the lysis buffer was removed by aspiration, the dishes were scraped with a rubber policeman, and tissue was collected in conical micro-centrifuge tubes. The lysates were resuspended in a final volume of 1.5 ml of cold lysis buffer and pelleted twice by centrifugation at 15000 g for 1 min in an Eppendorf micro-centrifuge. The washed membrane pellet was resuspended in lysis buffer for assay. Under these conditions, no intact cells were observed in the membrane preparation, as determined by Trypan Blue exclusion under microscopic examination. Membranes were routinely prepared fresh, warmed to 37 °C, and used immediately. Assays were initiated by the addition of 100 l1 of membranes from myo-[3H]inositol-prelabelled cells to 100 #1 of prewarmed (37 °C) assay buffer. Final concentrations in the reaction mixture were 110mMKCl, 10 mM-NaCl, 1 mM-KH2PO4, 20 mM-Hepes, 4 mM-MgCl2, 1 mM-EGTA, 10 mM-LiCl, 3 mm-Na2ATP, 8 mM-phosphocreatine and 12 units of creatine kinase/ml (pH 7.0). All drugs, including agonists, antagonists and guanine nucleotides, were mixed fresh in assay buffer and

included in the assay mixture described above. The assays were terminated by the addition of 1.Oml of ice-cold 5 % (v/v) trichloroacetic acid to each assay tube. The precipitate was removed by centrifugation (4000 g) and trichloroacetic acid was extracted from each sample by washing with 3 x 2 ml of diethyl etner. Residual ether was removed by aspiration and exposure to N2 gas. [3H]Inositol phosphates in the aqueous sample were determined essentially as described by Masters et al. (1984). The sugar products were verified by h.p.l.c. analysis (N. Nakahata & T. K. Harden, unpublished work) as described by Irvine et al. (1985). In general, results were qualitatively reproducible across experiments. However, the total amounts of inositol phosphates measured varied according to the amount of membranes used per assay (150-300 ,ug of protein) and the time and efficiency of labelling of the membrane phosphoinositide pool. In addition, the percentage increase in inositol phosphate formation observed in the presence of agonist plus guanine nucleotide and guanine nucleotide alone over that observed with no additions (i.e. controls) varied somewhat from experiment to experiment. The reasons for this are as yet unclear. Experiments with pertussis toxin were carried out after incubating cells overnight with toxin. Toxin was prepared as previously described (Hughes et al., 1984). Assays to measure GTP-mediated inhibition of forskolinstimulated adenylate cyclase activity in cell-free preparations and pertussis-toxin-catalysed [32P]ADP ribosylation of 1 321N1 membranes were performed as described previously (Hughes et al., 1984; Evans et al., 1985a). myo-[3H]Inositol was supplied by New England Nuclear or American Radiolabelled Chemicals. Carbachol and atropine were supplied by Sigma. Guanine nucleotides were supplied by Boehringer-Mannheim; AG1-X8 Dowex resin was obtained from Bio-Rad. Statistical analysis of the data was performed by Student's t test.

Table 1. Guanine nucleotide specificity for inositol phosphate formation in 1321N1 membranes

Membranes were prepared from 1321N1 cells prelabelled with myo-[3H]inositol as described in the Materials and methods section. Membranes were incubated at 37 °C for 10 min in the absence or presence of the indicated concentration of guanine nucleotide, in the absence or presence of 500 ,M-carbachol. The data are presented as means + S.E.M. of total inositol phosphate formation and as percentages of control values, where control refers to inositol phosphate formation in the absence of either guanine nucleotide or carbachol. The data are representative of three experiments. * Increase in inositol phosphate formation by guanine nucleotide significantly greater than control (P < 0.05) ** increase in inositol phosphate formation by hormone plus guanine nucleotide significantly greater than increase by guanine nucleotide alone (P < 0.05). Inositol phosphate formation No carbachol

+carbachol

Nucleotide

(C.p.m.)

(% of control)

(c.p.m.)

None

203 + 2 554+ 16 335+6 411+ 10 216+2 213 + 7 240+12 246 + 20

100 273*

224+ 13 660+6 379+9 511 +24 283 + 11 271+3 322+18 255+4

GTP[S] (10 /SM) p[NH]ppG: (316 /LM)

(1000 ,tM)

GTP: (316 pM)

(1000 /M)

GDP (1000 1UM) GMP (1000 #M)

165* 202* 106* 105 118 121

(% of control) 110 325** 1879:* 252** 139** 133** 159** 126

1986

Guanine nucleotide-stimulated inositol phosphate formation

9

+ GTP[S]

0

8

7

_

E

= Ckc

t

6

-,

+ GTP [SI + atropine

X 0

6 5

4 10

0

1---L.4 .--17 00

I

Il

I

6 5 4 -log {[Carbachol] (M)}

IL 3

Fig. 1. Carbachol-stimulated inositol phosphate formation in washed 1321N1 membranes Membranes were prepared from 1 321N1 cells prelabelled with myo-[3H]inositol as described in the Materials and methods section. Incubations were for 10 min at 37 °C in assay buffer containing the indicated concentrations of carbachol in the absence of additional drug (0) or in the presence of 10 ,#M-GTP[S] (M) or GTP[S] + 10 /ZM-atropine (A). The data are means of triplicate determinations and are representative of results from three experiments.

RESULTS

Addition of 10 I,M-GTP[S] to a washed membrane preparation from 1321N1 cells previously labelled with [3H]inositol resulted in an increase in inositol phosphate formation (Table 1; Fig 1). In the absence of GTP[S], carbachol had no effect on inositol phosphate formation at all concentrations tested (Fig. 1). In contrast, in the presence of GTP[S], carbachol increased inositol phosphate formation in a concentration-dependent manner (Fig. 1). The Ko 5 for carbachol (10 ,SM) was similar to that previously observed in studies of carbacholstimulated inositol phosphate formation in intact 1321N1 cells (Masters et al., 1984). The effect of carbachol was apparently through a muscarinic cholinergic receptor, since, in the presence of a saturating concentration of atropine, carbachol was without effect

(Fig. 1).

As illustrated in Fig. 2, GTP[S] alone or in combination with 500 uM-carbachol stimulated the formation of InsPl, InsP2 and InsP3. Accumulation of

Vol. 239

143

InsP3 reached a maximum within 1 min [preliminary studies using intact 1321N1 cells indicate that carbachol stimulates the rapid formation of the two isomers of InsP3, Ins(1,3,4)P3 and Ins(1,4,5)P3, as well as Ins(1,3,4,5)P4; however, in washed membrane preparations from 1321N1 cells, initial work suggests that carbachol selectively stimulates the formation of Ins(l,4,5)P3], accumulation of InsP2 was linear for at least 5 min, and, after a lag of approx. 1 min, InsPj accumulation was linear for at least 5 min (results not shown). The K0,5 of GTP[S] for stimulation of accumulation ofinositol phosphates under each condition was 3-10/,M. Although GTP[S] was the most potent nucleotide studied, inositol phosphate formation also was increased by p[NH]ppG. In some experiments GTP, GDP and GMP marginally increased (by 10-20%) inositol phosphate formation. Carbachol also increased inositol phosphate formation in the presence of p[NH]ppG, GTP and GDP, but not in the presence of GMP (Table 1). NaF is a well-known activator of guanine nucleotide regulatory proteins (Gilman, 1984). As illustrated in Fig. 3, NaF at millimolar concentrations also stimulated inositol phosphate formation in 1321N1 membranes in amounts comparable with those stimulated by a maximally effective concentration of GTP[S]. An important unresolved question concerning receptormediated regulation of inositol phosphate formation centres on the identity of the putative regulatory protein that subserves a trans-membrane signalling function in this system. Although experiments with several bonemarrow-derived cells (Brandt et al., 1985; Ohta et al., 1985) suggest that Gi or a similar pertussis-toxin substrate is involved in regulation of phosphoinositide hydrolysis, experiments with other mammalian tissues do not support this conclusion (Masters et al., 1985; Schlegel et al., 1985). Information accumulated to date for 1321N1 cells would support the latter group of studies, since pertussis toxin has no effect on muscarinicreceptor-mediated activation of cyclic AMP phosphodiesterase (Hughes et al., 1984; Nakahata et al., 1986), mobilization of Ca2+ (Masters et al., 1985), or hydrolysis of phosphoinositides (Masters et al., 1985; Nakahata et al., 1986). The availability of a cell-free preparation in which phosphoinositide hydrolysis can be measured provides another approach for examining the possible role of Gi or a similar pertussis-toxin substrate in Ca2+ signalling. As illustrated in Fig. 4, pertussis toxin reversed in a concentration-dependent manner the capacity of GTP to inhibit forskolin-stimulated adenylate cyclase activity (Fig. 4a). The same results were obtained if p[NH]ppG, rather than GTP, was used as the nucleotide for inhibition (results not shown). As discussed previously by Seamon & Daly (1982) and supported by work from other laboratories (Hildebrandt et al., 1982; Kurose et al., 1983), these data are consistent with a functional inactivation of Gi. The Ko.5 for action of toxin was 10-20 ng/ml; the effects of GTP were completely reversed at 46-100 ng/ml. Under these conditions pertussis-toxin-catalysed incorporation of [32P]ADP-ribose into the oc-subunit of Gi was decreased by over 95%O (Fig. 4b, and Evans et al., 1985a). As illustrated in Fig. 4(b), incubation of 1321N1 cells in the presence of concentrations of toxin that were 10-100 times higher than those necessary to ADP-ribosylate and functionally inactivate Gi had no effect on inositol phosphate formation in the presence of carbachol plus

144

J. R. Hepler and T. K. Harden 4.5 4.5

2.0 I. L -_

ci E

3.5

1.5 1-

2.5 k

-o 0 r-

1.5

2.5

1.0 I_

0 00

7

I

7

00

5

I

I

I

5

3

O

k L -1-i ,a I

00

I 7

a

I 5

I

I 3

-log { [GTP[S]] (M) }

Fig. 2. InsP1, InsP2 and InsP3 formation in the presence of GTPISI and carbachol 132INI membranes were prepared as described in the Materials and methods section. Incubations were for 10 min at 37 °C in assay buffer containing the indicated concentrations of GTP[S] in the absence (A) or presence (0) of carbachol (CARB; 500 /,tM). The data are means of triplicate determinations and are representative of results from three experiments.

GTP[S]. Also, inositol phosphate formation in the presence of GTP[S] alone was not affected by toxin (results not shown).

6

4

E L

C

I-

X 0

Il,

2

0

0

1

10

100

[F-] (mM) Fig. 3. NaF-stimulated inositol phosphate formation in washed 1321N1 membranes Membranes were prepared from 1321N1 cells prelabelled with myo-[3H]inositol as described in the Materials and methods section. Incubations were for 10 min at 37 °C in assay buffer containing no further additions (CONT), 100 /sM-GTP[S], 500 /M-carbachol (CARB)+GTP[S], or various concentrations of NaF (M). The data are means +S.E.M. of triplicate determinations and are representative of results from two experiments.

DISCUSSION The present work indicates that guanine nucleotides stimulate the formation of inositol phosphates in membranes from 132INI cells. In the absence of added guanine nucleotide, carbachol had no effect, whereas in the presence of guanine nucleotide, carbachol stimulated inositol phosphate formation over that achieved with guanine nucleotide alone. In addition, fluoride, which is an activator of known guanine nucleotide regulatory proteins, stimulates inositol phosphate formation in washed 1321N1 membranes. Thus these studies directly support our previous proposal of a role for a guanine nucleotide regulatory protein in the muscarinic-receptor/phosphoinositide response of 1321N1 cells (Evans et al., 1985a). Although the work does not identify this putative G-protein, it does confirm with a cell-free response system the conclusions made previously with intact tissue (Masters et al., 1985; Nakahata et al., 1986) or N-ethylmaleimide-pretreated 1321N1 membranes (Martin et al., 1985) that Gi or a similar pertussis-toxin substrate is not involved in cholinergic action in these cells. The availability of such a cell-free system should aid in the identification of this functionally important trans-membrane coupling protein. It also should promote comparisons of mechanism between the type of response typified by muscarinic receptors of 132 INI cells and that typified by the fMet-Leu-Phe receptor of bone-marrowderived cells (Brandt et al., 1985; Ohta et al., 1985), which apparently regulates phosphoinositide hydrolysis through a protein that is a substrate for pertussis toxin. The guanine nucleotide binding protein, GO, purified 1986

Guanine nucleotide-stimulated inositol phosphate formation 100 _

(a) .

4J1 02 r_

0 0

.

90 .

. .

L)

80 1co

cl)

70 -

0

1

10

1000

100

[PT] (ng/mi)

(b) 7.0 F-

6.0[C-E

5.0

1-

66 -

0~

x

45-

4.01-

3.0 1-

1-9+ z H

o mJ C

31-

~~~~C ~~a

fof iof--

l

1

b

l

10

d

C |

100

1000

[PT] (ng/ml)

Fig. 4. Differential effects of pertussis toxin (PT) on GTPmediated inhibition of forskolin-stimulated adenylate cyclase activity and muscarinic-receptor-mediated stimulation of inositol phosphate formation in 1321N1 membranes

(a) Cells were incubated overnight with the indicated concentrations of pertussis toxin. Membranes were prepared and adenylate cyclase assays carried out in the presence of 100 /sM-forskolin (control) or 100 ,sM-forskolin + 316 4uM-GTP. The results are plotted as the percentage of forskolin-stimulated adenylate cyclase activity in the absence of GTP, and are means of quadruplicate determinations, representative of results from three experiments. (b) Cells were labelled with myo-[3H]inositol as well as being pretreated overnight with the indicated concentrations of pertussis toxin. Membranes were prepared and incubated in the absence (control; CONT) or presence of 500 /tM-carbachol (CARB)+ 10 ,uM-GTP[S]. The data are means +S.E.M. of quadruplicate determinations and are representative of two experiments. Inset: plasma membranes were prepared from cells treated overnight with either vehicle or 500 ng of pertussis toxin/ml. The membranes were then incubated with pertussis toxin and [32P]NAD+, solubilized with SDS and Vol. 239

145

from bovine brain by Sternweis & Robishaw (1984) and Neer et al. (1984), is a potential candidate for the regulatory protein involved in regulation of phosphoinositide hydrolysis. However, although G. has been shown to interact with muscarinic receptors in a model phospholipid vesicle system (Florio & Sternweis, 1985), Go may not be ubiquitous, and to date we have been unable to obtain evidence for its presence in 132 IN 1 cells (G. L. Waldo, M. W. Martin, A. R. Hughes, T. Evans, J. K. Northup & T. K. Harden, unpublished work). Furthermore, Go is a substrate for pertussis toxin. Assuming that the toxin modifies Go in a manner similar to Gi, involvement of Go in the phosphoinositide response of 1321N1 cells would result in pertussis-toxininduced prevention of inositol phosphate formation in these cells. This is clearly not the case (Fig. 4, and Masters et al., 1985; Nakahata et al., 1986). The product of the ras gene is a membrane-bound guanine nucleotide binding protein that has been implicated in phosphoinositide metabolism (Berridge & Irvine, 1984). However, to date no compelling evidence, aside from circumstantial, has appeared to assign a defined role for the ras protein in this pathway. Evans et al. (1986) have purified a heretofore undescribed guanine nucleotide binding protein from human placenta. Perhaps this protein, which is composed of a GTP-binding a-subunit of approx. Mr 25 000, is involved in regulation of phosphoinositide hydrolysis in 1321 NI cells. Our prejudice is that the putative guanine nucleotide regulatory protein that we propose to couple muscarinic receptors to phospholipase C in 132 IN 1 cells subserves the same function for many other hormones in other tissues. Both histamine (Nakahata et al., 1986) and bradykinin (J. R. Hepler & T. K. Harden, unpublished work) which increase inositol phosphate formation in intact 132IN1 cells, have been shown to increase inositol phosphate formation in a guanine nucleotide-dependent manner in membranes from these cells (J. R. Hepler & T. K. Harden, unpublished work). The availability of a cell-free preparation for analysis of this hormone signalling system should be of considerable value in identification of the coupling protein that is of crucial importance in the action of many hormones. This work was supported by a grant-in-aid (GM29536) from the US Public Health Service and the American Heart Association. J. R. H. was supported by a graduate student fellowship from Lilly Research Laboratories. T.K.H. is an Established Investigator of the American Heart Association. We thank Margaret Tapp for her excellent work in preparing the manuscript and Dr. Norimichi Nakahata and Dr. Marco Conti for their valuable help and discussion. J. R. H. also thanks electrophoresed on 11% -polyacrylamide gels. An autoradiogram showing Mr values ( x 10-3) of protein standards is presented. Labelling conditions were as follows: lane a, control cells, i.e. cells not treated overnight with PT, and [32P]ADP-ribosylation reaction carried out in the absence of toxin; lane b, control cells, and [32P]ADPribosylation reaction carried out in the presence of toxin; lane c, pertussis-toxin-pretreated cells, and [32P]ADPribosylation reaction carried out in the absence of toxin; lane d, pertussis-toxin-pretreated cells, and [32P]ADPribosylation reaction carried out in the presence of toxin. Under these last conditions, i.e. lane d, over 95% of the ADP-ribosylation was blocked by toxin pretreatment, as determined by densitometric measurement.

146

also thanks Dr. Irene Litosch and Dr. Marvin Gershengorn for their patience and their generous and valuable discussions. Finally, we extend special thanks to L. A. Petch for the always stimulating and rigorous discussions and her timely aid and ministrations.

REFERENCES Berridge, M. J. & Irvine, R. F. (1984) Nature (London) 312, 315-321 Brandt, S. W., Daugherty, R. W., Lapetina, E. G. & Niedel, J. E. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 3277-3280 Cockcroft, S. & Gomperts, B. D. (1985) Nature (London) 314, 534-536 Evans, T., Martin, M. W., Hughes, A. R. & Harden, T. K. (1985a) Mol. Pharmacol. 27, 32-37 Evans, T., Hepler, J. R., Masters, S. B., Brown, J. H. & Harden, T. K. (1985b) Biochem. J. 232, 751-757 Evans, T., Brown, M. L., Fraser, E. D. & Northup, J. K. (1986) J. Biol. Chem. 261, 7052-7059 Florio, V. A. & Sternweis, P. C. (1985) J. Biol. Chem. 260, 3477-3483 Gilman, A. G. (1984) J. Clin. Invest. 73, 1-4 Gomperts, B. D. (1983) Nature (London) 306, 64-66 Gonzales, R. A. & Crews, F. T. (1985) Biochem. J. 232, 799-804 Haslam, R. J. & Davidson, M. M. L. (1984) FEBS Lett. 174, 90-95 Hildebrandt, J. D., Hanoune, J. & Birnbaumer, L. (1982) J. Biol. Chem. 257, 14723-14725 Hughes, A. R., Martin, M. W. & Harden, T. K. (1984) Proc. Natl. Acad. Sci. U.S.A. 81, 5680-5684 Irvine, R. G., Anggard, E. E., Letcher, A. J. & Downes, C. P. (1985) Biochem. J. 229, 505-511

J. R. Hepler and T. K. Harden Kurose, H., Katada, T., Amano, T. & Ui, M. (1983) J. Biol. Chem. 258, 4870-4875 Litosch, I., Wallis, C. & Fain, J. N. (1985) J. Biol. Chem. 260, 5464-5471 Litosch, I., Calista, C., Wallis, C. & Fain, J. (1986) J. Biol. Chem. 261, 638-643 Martin, M. W., Evans, T. & Harden, T. K. (1985) Biochem. J. 229, 539-544 Masters, S. B., Harden, T. K. & Brown, J. H. (1984) Mol. Pharmacol. 26, 149-155 Masters, S. B., Martin, M. W., Harden, T. K. & Brown, J. H. (1985) Biochem. J. 227, 933-937 Meeker, R. B. & Harden, T. K. (1982) Mol. Pharmacol. 22, 310-319 Nakahata, N., Martin, M. W., Hughes, A. R., Hepler, J. R. & Harden, T. K. (1986) Mol. Pharmacol. 291, 188-195 Neer, E. S., Lok, J. M. & Wolf, L. G. (1984) J. Biol. Chem. 259, 14222-14229 Ohta, H., Fumikazu, 0. & Ui, M. (1985) J. Biol. Chem. 260, 15771-15780 Schlegel, W., Wuarin, F., Zbaren, C., Wollheim, C. B. & Zahnd, G. R. (1985) FEBS Lett. 189, 27-32 Seamon, K. B. & Daly, J. W. (1982) J. Biol. Chem. 257, 11591-11596 Smith, C. D., Lane, B. C., Kusaka, I., Verghese, M. W. & Snyderman, R. (1985) J. Biol. Chem. 260, 5875-5878 Sternweis, P. C. & Robishaw, J. D. (1984) J. Biol. Chem. 259, 13806-13813 Tanner, L. I., Harden, T. K., Wells, J. N. & Martin, M. W. (1986) Mol. Pharmacol. 29, 455-460 Uhing, R. J., Jiang, H., Prpic, V. & Exton, J. H. (1985) FEBS Lett. 188, 317-320 Wallace, M. A. & Fain, J. N. (1985) J. Biol. Chem. 260, 9527-9530

Received 18 February 1986/11 June 1986; accepted 19 June 1986

1986