Specific guanine nucleotide binding by membranes from cucurbita ...

Vol. 155, No. 3, 1988

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 1478-1484

September 30, 1988

SPECIFIC GUANINE NUCLEOTIDE BINDING BY MEMBRANES FROM CUCURBITAPEPO SEEDLINGS

M.Jacobs1, M.P. Thelen, R.W. Farndale,M.C. Astle2 and P.H. Rubery

Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge CB2 1QW, England Received August 26, 1988 A microsomal membrane preparation from hypocotyls of dark-grown Cucurbita peDo L. (zucchini) seedlings contains specific high-affinity binding sites for the non-hydrolyzable GTP analog guanosine 5'-[7-thio]tdphosphate (GTP-7-S). Both the binding affinity and the pattern of binding specificity for GTP and guanine nucleoside triphosphate analogs are shared with the more thoroughly characterized animal G-proteins that are known to be involved in signal transduction. The sensitivity of GTP-7-S binding to Mg+2 ions and temperature was similar to that reported for rabbit liver G-protein, although the plant complex dissociated more readily. GTP-7-S could be recovered unchanged from the binding complex. Proteins (Mr 33 and 50 kDa) present in zucchini membrane preparations were revealed by immunoblotting with antiserum specific for the c~subunit of platelet Gs. These may be homologous to animal G-proteins. © 1988Aoademio Press,

Inc.

A family of membrane-associated, guanine nucleotide-binding proteins (G-proteins) that can function in transmembrane signalling in animal cells has been firmly established (1). A major role of G-proteins is to couple membrane-associated receptors to effectors that are either enzymes or ion channels. Activation of the receptor by agonist binding leads to the release of GDP by the G-protein, and to the binding of GTP (see 2). While GTP remains bound, the Gprotein can regulate the activity of an effector; that regulation ceases when GTP is hydrolyzed to GDP. The G-proteins are heterotrimers; the ~ subunit binds GTP and appears to determine the specificity of the proteins for their receptors and effectors (3). In plants, there is no direct evidence yet for G-protein function; nevertheless, several reports have recently appeared describing GTP binding to membranes and membrane proteins. One group (4,5) examined the aquatic flowering plant Lemna Daucicostata and the dark-grown epicotyls of Alaska pea (Pisum sativumL They separated fractions from detergent-treated membranes by gel filtration, measured [35S]GTP-~S binding rates and estimated "Km values" ranging from 1 to 50 nM; Kd'S for binding were not explicitly measured. Another author (6)

1permanent Address: Department of Biology, Swarthmore College, Swarthmore, PA 19081 2perman'ent Address: Imperial Chemical Industries plc, Jealott's Hill Research Station, Bracknell, Berkshire, RG2 6EY, England Gpp(NH)p = guanylylimidodiphosphate; GTP-7-S = guanosine 5'-[7thio]triphosphate; PEI = polyethyleneimine; TS = test solution. 0006-291X/88 $1.50 Copyright © 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.

1478

Vol. 155, No. 3, 1988

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

described a light-increased binding of GTP-~-S and GTP to a spinach thylakoid membrane preparation but presented no binding curve and checked single high concentrations of TTP, UTP and CTP as a test of the effects of other nucleotides on the binding. A third group (7) used the technique of direct binding of guanine nucleotides to nitrocellulose blots of proteins resolved by SDS-PAGE (8) to identify four proteins from zucchini hypocotyl membranes that resembled animal Gn proteins (7). The effects of single concentrations of ATP, GDP and GTP-'/-S on binding of radiolabeled GTP were reported. We report here guanine nucleotide binding to hypocotyl cell membranes from zucchini seedlings (Cucurbita Depo L.) that shares characteristics with well-defined G-proteins from animals, including similar binding affinity and specificity as well as recognition of membrane proteins from the plant by an animal G-protein-specificantibody. MATERIALS AND METHODS Chemicals and Radiochemicals. The following nucleotides were obtained as sodium salts from the Sigma Chemical Company: ATP (prepared by phosphorylation of adenosine), GMP, GDP, GTP, guanylylimidophosphate (Gpp(NH)p), ITP, UTP. CTP was obtained from British Drug Houses and Li4-guanosine-5'[~-thio]triphosphate (GTP--/-S) from Fluka. [35S]guanosine-5'[~,thio]triphosphate (1262 Ci/mmol) was obtained from New England Nuclear. Plant Matedal and Membrane Preparation. Zucchini seedlings (Cucurbita peoo L. var. Green Bush) were grown at 25 C in darkness for 6 days in moist vermiculite in a humid chamber. The hypocotyls were excised and used for membrane preparation. Microsomes were routinely prepared as described earlier (9). The 90,000 x g pellets were resuspended by homogenizing in Test Solution (TS: 250 mM sucrose, 5 mM MgSO4, 10 mM HEPES/NaOH pH 7.0) to yield 1.7 g original fresh weight per ml of resuspended particles. This preparation was used for binding assays and for one portion (Fig. 3, gel B) of the SDS PAGE and immunoblotting experiments. The alternate preparation for SDS PAGE and immunoblotting (Fig. 3, gel A) was a polyethylene glycol-supported sedimentation procedure (exposure to 6% (w/v) PEG-4000 followed by bench centrifugation: see ref. 10) used after filtration of the initial homogenate (9), prepared on this occasion in the presence of 0.5 mM phenylmethylsulphonyl fluoride and 50 pM leupeptin (Sigma).

[35S!GTP-~$ Binding Assays. Portions (1 ml) of the membrane preparation were added to centrifuge tubes containing 1 ml of TS supplemented with 0.2 nM [35S]GTP-'/-S with varying concentrations of unlabeled GTP-7-S or other nucleotides. The tubes were vortexed and then centrifuged at 90,000 x g for 45 min at 0 C. Pellets were counted as described earlier (9). As in earlier reports (11,12), the term "specific binding" is used to describe the difference [cpm of 35SGTP-y-S pelleted without unlabeled GTP-~(-S present] - [cpm of 35S- GTP--f-S pelleted in the presence of saturating (10-50 I~M) unlabeled GTP-),-S]. Unbound [35S]GTP-~-S was measured in supernatant aliquots. In some experiments (Table 1), the high speed pellet was resuspended in TS from which MgSO4 had been omitted. The effect of temperature on [35S]GTP-~'-S association was tested by incubating the membrane preparation with[35S]GTP-7-S + unlabeled GTP-7-S, with and without Mg+2, at either 0 C or 30 C before centrifugation at 90,000 x g for 45 min. Temperature and Mg+2 effects on dissociation of [35S]GTP-~-S already bound to sites in the presence of Mg+2 at 0 C were investigated by resuspending radioactive pellets in 4 ml each of TS + 5 mM MgSO4, supplemented with 10 ~M unlabeled GTP-7-S. These samples were then incubated at 0 C or 30 C for 30 min before recentrifugation at 90,000 x g for 45 min. Samples of the supernatant were both counted and analyzed by PEI-cellulose thin layer chromatography to verify the identity of the radioactive material displaced from the membrane preparations by the unlabeled GTP-7-S. A marker of [35S]GTP-7--S was run in parallel. The solvent was 0.5 M K2HPO4 + 10 mM I'J-mercaptoethanol. The plates were dissected into 1 cm sections and counted in 1 ml scintillant (diphenyloxazole: toluene, 3.5 g/I) after standing in darkness at 10 C for 2-3 hours. 1479

Vol. 155, No. 3, 1988


SDS PAGE and Immunoblottina. Zucchini hypocotyl membrane pellets, prepared using either of the two alternative methods described above, were resuspended at 5 g fresh weight per ml by homogenizing in TS diluted 1:1 (v/v) with glycerol. Human platelet membranes (13), purified rabbit Gscz (the gift of Professor E. Ross, prepared as described [14]), and 14C-methylated protein markers (Amersham International) were included as standards. Proteins were separated by SDS PAGE in a 10% polyacrylamide gel (13), then transferred to a nitrocellulose membrane filter in electroblotting buffer (150 mM glycine, 25 mM Tds-HCI, pH 8.3, 20% methanol) at 1.1 to 1.5 A for 90 min. G-proteins were detected on the nitrocellulose as described by Mumby et al (15), using a 1:2,000 dilution of a rabbit antiserum (No. 584, the kind gift of Drs. S. M. Mumby and A.G. Gilman) raised against the Gsc~ subunit synthetic peptide antigen of sequence CKQLQDKQVYRATHR (15), followed by incubation with 1251-labeled goat anti-rabbit serum (specific activity ca. 5 Ci/g)(16).

RESULTS [-3-~]GTP-~-S Bindina Substantial binding of [35S]GTP-7-S to hypocotyl membranes was obtained (Fig. 1A; representative of several repeats), with a very low level (ca. 5%) of non-specific binding found in the presence of saturating concentrations (50 p.M) of unlabeled GTP-y-S. Scatchard analysis (Fig. 1B) shows a shallow curve difficult to distinguish from a straight line, and systematic purification and characterization of the binding component(s) rather than statistical fitting procedures will be needed to resolve the kinetic complexity of the population as a whole. The data (Figs. 1A,B) allow the conclusion that high affinity [35S]GTP-~-S binding occurs, with 50% inhibition by about 300 nM nonradioactive GTP-7-S. The tissue level of sites is at least 77 pmol per g fresh wt. From the known multiplicity of animal G proteins (1,2), it seems unlikely that the zucchini microsomes will contain only a single type of binding site. Even though the standard binding assays were conducted at 0 C and in the presence of 5 mM Mg +2 - conditions that strongly inhibit dissociation of [35S]GTP-7--S from membrane sites in animal tissue (17) - we found GTP-7-S binding to zucchini membranes under these conditions to be at least partially reversible. Use of 10 pM unlabeled GTP to wash pelleted membranes to which [35S]GTP-?-S

250

I

A

mw 0.6~

o---II

X

~

o~-

200 E t~ ~_ 0.4t3

U3

I

150

imo" 0.3-

0 oo

100

oo

0.2-

50

oc:O

I

E

0

r---I1 _~ --oo

. -8

.

. -7

. -6

.

' ~ ' o.o -5

-4

log [O?-~-s] (M)

O

,

,

,

3

6

9

0 I

0

12

[GTP-?'-S] specifically bound, M x 10-8

Figure 1. (A) Binding of [35S]GTP-y-S to zucchini membranes as a function of increasing concentration of unlabeled GTP-7-S. Means of triplicates are shown, with bar indicating SEM where it is largerthan the indicatingsymbol (B) A Scatchardplot of the data from A, with each of the triplicate samplesfor each concentration of unlabeled GTP-7-S shown. 1480

15

Vol. 155, No. 3, 1988


had bound in the absence of competitors caused the release of an amount of [35S]GTP-T-S equivalent to 38% of the saturable binding (32% of total binding) measured when 10 I~M GTP was present throughout a standard binding assay. Thin layer chromatography on PEI-cellulose plates showed that all of the 35S above background levels that was present in the supernatant from the wash centrifugation was associated with the zone of migration of authentic [35S]GTP-y-S.

SDecificitv of Nucleotide Bindina The specificity of the guanine nucleotide binding protein from zucchini hypocotyl membranes was tested by measuring [35S]GTP-y-S binding in the presence of increasing concentrations of unlabeled nucleotides (Fig. 2). The best competitor was the non-radioactive GTP-7-S, while GTP and GppNHp (which, like GTP-~S, is an activating ligand for G-proteins [1]) were almost as effective at 10 ~M. GDP and ITP were intermediate; the other nucleoside triphosphates tested (ATP, UTP, CTP) and GMP had no significant effect. Similar specificity has been reported for animal G-proteins (17), although GDP appeared less effective than in the solubilized rabbit liver system (17). TemDerature and Ma--L2-Effects on [-~-S1GTP-7--S Bindina A preliminary test of the effects of Mg +2 and temperature on [35S]GTP-~-S binding to zucchini membranes (Table 1) indicates that 5 mM Mg +2 increases binding considerably. Higher temperature (30 C) can also increase G T P - ~ S binding over the level found at 0 C in the absence of Mg +2, but when Mg +2 is present nucleotide binding is not further increased by raised temperature. Northup et al (17) reported that [35S]GTP-y-S did not dissociate from rabbit liver G protein at ice temperature or at 30 C when millimolar Mg +2 was present. Without Mg +2, dissociation was 60-70% complete after 30 min. We investigated release of [35S]GTP-7-S from preloaded zucchini pellets (0 C, 5 mM Mg +2) that were resuspended in nonradioactive medium + 5 mM Mg +2 at 0 C and 30 C for 30 min before recentrifugation. At ice temperature and in the presence

~-, i 200 -

~'~×c~ o GlaP

X

E 150-

~,UW

"o c

vArP

100 -

$ GDP

I

offP

i

+

50 -"

oJr--II --oo

GppNHp

o GTP m GTP-'~-S I

I

I

I

-8

-7

-6

-5

log [nucleofide](M) Figure 2. Effect of different nucleotides on [35S]GTP-y-S binding to zucchini membranes. Points shown are means of duplicate samples. The average range, as a percent of the duplicate mean, was 1.83 + 0.38 (SEM; n = 18 duplicates).

1481

Vol. 155, No. 3, 1988


Effects of Mg +2 and temperature on -[35S]GTP-~-S binding. Procedures as described in Materials and Methods. Results are means of duplicate determinations ~ range, in cpm. Each determination represents 1.45 g fresh weight of tissue.

Table i.

Assay conditions

A B [35S]G--TP-~-S [35S]G~P-~-S alone + i0 #M GTP-~-S

Specific Binding (A - B)

_ Mg+2 0 C

i 0 0 , i 0 7 ~ 1,166

11,552~ 144

30 C

134,282~ 2,610

19,103~ 1,430

115,179

220,741~ 3,592

ii,866~ 262

208,875

220,989+_402

22,620~ 636

198,369

+ 5 mMMg

88,555

+2

0 C 30 C

of Mg +2, 37% of the originally pelleted radioactivity was released, in contrast to (17); when Mg +2 was omitted this value increased to 59%. The higher temperature increased dissociation either in the absence of Mg +2 (69%) or at 5 mM Mg +2 (64%). The released radioactivity migrated with marker GTP-7-S on PEI cellulose plates. Thus, at 30 C and without Mg +2, [35S]GTP-7-S release from zucchini membrane binding sites was increased, as in the rabbit liver system (17), but the dissociation appeared faster under the tested conditions. ImmunoN0tting with antiserum to a G.~LDe0tide Immunoblots of SDS PAGE-separated proteins from both the routine zucchini membrane preparation used for the binding studies and material prepared more rapidly in the presence of PEG and protease inhibitors were obtained. They each contained a 50 kDa component that was recognized by the Gsc~ antiserum (Fig. 3, gel A and gel B, track 1). The routine membrane preparation also showed a band of 33 kDa on the immunoblot. Control experiments showed that no bands were produced unless the Gs~ antiserum was present. DISCUSSION The characteristics of [35S]GTP-7-S binding to the zucchini membranes show several similarities and some differences in comparison with animal G-proteins. The binding affinity (i50 = 300 nM) is stronger than reported by Gilman's group for the solubilized rabbit liver Gs (Kd = 700 nM). However, the plasma membrane-bound rabbit liver Gs gave a Kd of 70 nM, so in terms of binding strengths the plant and animal systems are within the same order of magnitude. The nucleotide specificities are similar, with ITP (inosine = 6-hydroxypurine riboside) occupying an intermediate position between GTP (G = 2-amino-6-hydroxypurine) and ATP (A = 6aminopurine)(17,18). The binding of [35S]GTP-7-S in the crude zucchini membrane preparation seems more readily reversible than for animal G-proteins: the effects of Mg +2 and temperature are 1482

Vol. 155, No. 3, 1988


1

2

3

4

46

30 A

B

Figure 3. Recognition of zucchini membrane proteins and reference G-protein preparations by antiserum to a Gs¢ subunit peptide, assessed by immunoblotting. Autoradiograph portions show radiolabeled antibody bound to proteins, and [14C].protein standards (Mr in kDa shown on right of Figure). Gel A: zucchini membranes (50 I~g protein, prepared using PEG and protease inhibitors). Gel B: track 1, zucchini membranes (90,000 x g pellet, 100 p.g protein); track 2, purified rabbit Gso~ (about 5 ng); track 3, human platelet membranes (40 p.g protein); track 4, Mr standards. Arrowheads show proteins of 50 and 33 kDa in the zucchini samples.

qualitatively similar but, in zucchini, association and dissociation of [35S]GTP--f-S seem less Mg +2- and temperature-dependent (Table 1 and results). Hasunuma and co-workers have studied the association of radioactivity from [35S]GTP-,fS with material present in crude extracts of Lemna oaucicostata and Pisum sativum (pea) epicotyls (4,5). For the pea preparation, seven sequential samples from a gel filtration column contained [35S]GTP-7-S binding activity (5). In all seven of the fractions, ATP was an effective competitor of [35S]GTP-7-S, and in some cases ATP apparently had the higher affinity. The [35S]GTP-7-S binding proteins in pea were claimed to have the general characteristics of animal GTP-binding proteins "as they show specific affinities to ATP". However, ATP is not an activator of G-protein coupled effectors (1,2), and failed to inhibit [~-32p]GTP binding to platelet membrane proteins (18). Drobak et al (7) reported guanine nucleotide binding to proteins transferred to nitrocellulose after SDS-PAGE of a zucchini hypocotyl membrane preparation similar to the one we have used. The Mr's of the four proteins identified by Drobak et al as binding radiolabeled GTP ranged from 23.4 to 28.5 kDa. These proteins strongly resemble the "Gn proteins" reported by Bhullar and Haslam when they used the nitrocellulose 32p-GTP binding technique on rabbit and human platelet membrane proteins (8). The Gn proteins do not correspond in Mr to the subunits of any of the known G proteins and are not r a s gene products. Their function remains unknown. The two proteins detected in our zucchini preparations by the anti-Gsec antibody are 33 and 50 kDa. The M r values of the o~subunits of various animal G-proteins range from 39 kDa to 46 kDa (1), with one from placenta reported as 21 kDa (19). No 33 kDa Gsc~ subunit has been reported, but an Mr from SDS PAGE of 52 kDa was reported for Gsc~ (15). The ability of the Gs~specific antibody preferentially to detect particular components of the zucchini membrane preparation suggests that plant G-proteins with homologies to animal polypeptides are present.

1483

Vol, 155, No. 3, 1988


The possibility that the 33 kDa band is a proteolytic fragment of the 50 kDa component needs further investigation. The level of guanine nucleotide binding sites we find (>77 pmol/g fresh weight) is comparable to that of auxin binding proteins (20). It has recently been reported that auxin treatment can rapidly increase inositol (1,4,5)trisphosphate levels in cultured plant cells (21), a result reminiscent of the G-protein-mediated effects of several animal hormones (1,2). Clearly, guanine nucleotide binding proteins could play a significant role in plants; further work is needed to characterize them and to test their physiological relevance. ACKNOWLEDGEMENTS We are grateful to Dr. L.M. Allan who prepared and donated 1251 secondary antibody and to Dr. B.R. Martin for helpful discussions.

Dr. A.A. Newton kindly gave samples of some

nucleotides. The work was supported in part by National Science Foundation grant PCM 8314844 to MJ, who is also grateful to the John Simon Guggenheim Foundation for a Fellowship. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

Gilman,A.G. (1987) Annu. Rev. Biochem. 56, 615-649. Stryer, L. (1986) Annu. Rev. Neurosci. 9, 87-119. Cerbai,E., Klockner, U., and Isenberg, G. (1988) Science 240, 1782-1783. Hasunuma,K. and Funadera, K. (1987) Biochem. Biophys. Res. Comm. 143,908-912. Hasunuma,K., Furukawa, K., Tomita, K., Mukai, C. and Nakamura, T. (1987) Biochem. Biophys. Res. Comm. 148, 133-139. Millner, P.A. (1987) FEBS Letters 226, 155-160. Drobak,B.K., Allan, E.F., Comerford, J.G., Roberts, K. and Dawson, A.P. (1988) Biochem. Biophys. Res. Comm. 150, 899-903. Bhullar,R.P. and Haslam, R.J. (1987) Biochem. J. 245, 617-620. Sabater,M. and Rubery, P.H. (1987) Planta 171,501-506. Michalke,W. and Schmieder, B. (1979) Planta 145, 129-135. Ray, P.M., Dohrmann, U. and Hertel, R. (1977) Plant Physiol. 59, 357-364. Jacobs,M. and Rubery, P.H. (1988) Science 241,346-349. Farndale,R.W., Wong, S.K.F. and Martin, B.R. (1987) Biochem. J. 242, 637-643. Sternweis,P.C., Northup, J.K., Smigel, M.D. and Gilman, A.G. (1981) J. Biol. Chem. 256, 11517-11526. Mumby,S.M., Kahn, R.A., Manning, D.R. and Gilman, A.G. (1986) Proc. Natl. Acad. Sci. USA 83, 265-269. Hunter,W.M. and Greenwood, F.C. (1962) Nature 194, 495-496. Northup,J.K., Smigel, M.D. and Gilman, A.G. (1982) J. Biol. Chem. 257, 11416-11423. Lapetina,E.G. and Reep, B.R. (1987) Proc. Natl. Acad. Aci. USA 84, 2261-2265. Evans,T., Brown, M.L., Fraser, E.D. and Northup, J.K. (1986) J. Biol. Chem. 261,70527059. Venis, M. (1985) Hormone Binding Sites in Plants. Longman, New York. Ettlinger,C. and Lehle, L. (1988) Nature 331,176-178.

1484