THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol . 267, No. 8, Issue of March 15, pp. 5085-5089,1992 Printed in U.S.A.

A GTP-binding Protein in Rat Liver Nuclei Serving as the Specific Substrate of Pertussis Toxin-catalyzed ADP-ribosylation*

(Received for publication, July 19, 1991)

Yoshinori Takei, Hiroshi Kurosu, Katsunobu Takahashi, and Toshiaki KatadaS From the Department of Life Science, Tokyo Institute of Technology, Yokohama 227, Japan

The ADP-ribosyl moiety of NAD was transferred to a 40-kDa protein when rat liver nuclei were incubated with pertussis toxin. The 40-kDa substrate in the nu- clei displayed unique properties as follows, some of which were apparently distinct from those observed with the toxin-substrate GTP-binding protein (Gi) in the liver plasma membranes. 1) The nuclear 40-kDa protein was recognized with antibodies reacting with the a-subunits (ai-1 and ai-2) of Gi, but not with anti- Go-a-subunit antibody. 2) The nuclear protein had a higher mobility than a-subunit of the plasma mem- brane-bound Gi upon electrophoresis with a urealso- dium dodecyl sulfate-containing polyacrylamide gel. 3) The nuclear protein was not extracted from the nuclei with 1% Triton X-100, whereas Gi was easily solubilized from the plasma membranes. 4) There was a by-subunit-like activity in the nuclei, which was assayed by an ability to support pertussis toxin-cata- lyzed ADP-ribosylation of a purified a-subunit of Gi. Moreover, a 36-kDa protein in the nuclei was recog- nized with antibody raised against purified &subunits of Gi. 6) Pertussis toxin-induced ADP-ribosylation of the nuclear protein was selectively inhibited by the addition of a nonhydrolyzable GTP analogue, and its inhibitory action was competitively blocked by the simultaneous addition of GDP or its analogues, as had been observed with plasma membrane-bound Gi. It thus appeared that a novel form of aby-trimeric GTP-bind- ing protein serving as the substrate of pertussis toxin was present in rat liver nuclei. In order to examine a possible role of the nuclear GTP-binding protein, rats were injected with carbon tetrachloride, a necrosis inducer of hepatocytes. There was a marked increase in the nuclear substrate activity from 3-6 days after the injection, without a significant change in the activ- ity of Gi in the plasma membranes. The time course of the increase corresponded with a recovering stage from the hepatocyte necrosis. These results suggested that the nuclear GTP-binding protein found in the present study might be involved at some stages in the hepatocyte growth.

A family of structurally and functionally homologous mem-

* This work was supported by research grants from the Scientific Research Fund of the Ministry of Education, Science, and Culture of Japan and the Human Frontier Science Program. The costs of pub- lication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertise-

this fact. rnent” in accordance with 18 U.S.C. Section 1734 solely to indicate

$ TO whom correspondence should be addressed. Tel.: 81-45-922- 1111 (ext. 2225); Fax: 81-45-923-0367; Telex: 3823553 TITNAG J.

brane-associated GTP-binding proteins (G proteins)’ are present in a variety of vertebrate cells (see Refs. 1 and 2 for review). G proteins, which have been characterized as a com- mon heterotrimeric structure consisting of an a-, a p-, and a y-subunit, act as signal transducers from plasma membrane- bound receptors to effectors such as enzymes or ion channels. Binding of GTP to their a-subunits induces activation of G proteins and consequently regulates the activity of appropri- ate effectors. Hydrolysis of GTP initiates the deactivation of G proteins. Effector systems then generate intracellular sig- nals that are responsible for a set of early biological reactions. In addition, some of the intracellular signals are transmitted to nuclei, resulting in production of slow cell responses, in- cluding DNA replication.

An additional feature of G proteins is their susceptibility to bacterial toxins. Pertussis toxin (IAP), catalyzes the trans- fer of an ADP-ribosyl moiety of NAD to the a-subunits of Gi, Go, and G,. The site to be modified by IAP was a cysteine that was 4 amino acid residues away from the carboxyl ter- minus of the a-subunits (3, 4). Since the carboxyl terminus region containing the cysteine residue was responsible for the interaction with receptors (5, 6), the ADP-ribosylation re- sulted in an uncoupling of the G proteins from receptors. Thus, IAP is used as a specific and sustained modifier of recepter-mediated signal transductions via plasma mem- branes in many cell types.

It has been also reported that slow cell responses, such as induction of mitosis (7) or differentiation (8) by certain fac- tors are inhibited by prior exposure of cells with IAP. This implies that an IAP-sensitive G protein(s) is involved at some stages in initiating cell division or differentiation. Although the role of apy-trimeric G proteins is assumed to be in the early events of receptor-mediated signal transductions, it can not be excluded that the proteins may play a postreceptor role in such slow cell responses. In this regard, several papers have recently reported evidence for the existence of GTP-binding proteins in nuclei (9-12). However, these reports did not clearly address whether the proteins were substrates for IAP- catalyzed ADP-ribosylation or not. As a probe to search a nuclear G protein possibly involving in the nuclear signal- transducing systems, we again employed IAP with rat liver nuclei. This approach allowed us to find an apy-trimeric G protein serving as the substrate of IAP in the nuclear fraction. The present paper deals with some characteristics of a newly discovered G protein in rat liver nuclei that is different from plasma membrane-bound Gi in terms of their associated

The abbreviations used are: G protein, GTP-binding protein; Gi, the G protein that mediates the inhibition of adenylate cyclase; G,, the G protein from rod outer segments; Go, a similar G protein of unknown function purified from brain; IAP, islet-activating protein (pertussis toxin); GTPyS, guanosine 5’-(y-thio)triphosphate; GDPPS, guanosine 5’-(p-thio)diphosphate; App(NH)p, adenosine 5’- (P,y-imid0)triphosphate; SDS, sodium dodecyl sulfate.

5085

5086 A Nuclear GTP-binding Protein in Rat Liver

forms. A possible role of the nuclear G protein is also inves- tigated in the hepatocyte growth from their necrosis by meas- uring the IAP substrate activity in the nuclei.

EXPERIMENTAL PROCEDURES

Preparations of Various Fractions from Rat Liuers-Liver nuclei were isolated from male Donryu rats (280-300 g, body weight) based on the method reported by Kaufmann et al. (13) with a slight modification. Isolated livers from 10 rats were homogenized with 4 volumes of buffer A consisting of 15 mM Tris-HC1 (pH 7.41, 60 mM KC1, 15 mM NaCI, 0.15 mM sperumin, 0.5 mM sperumidine, 0.25 M sucrose, 1 mM EDTA, 0.5 mM [ethylenebis(oxo-nitri1o)ltetraacetic acid, and 1 p~ GDP and centrifuged a t 800 X g for 10 min. The supernatant was subjected to the purifications of cytosol and micro- some-rich fractions, as described by Carey and Hirschberg (14). The pellet was washed twice with 4 volumes of buffer A and mixed with buffer A containing 2.1 M sucrose to give a final concentration of sucrose at 1.7 M. A 15-ml aliquot of the mixture was layered over 7 ml of buffer A containing 2.1 M sucrose, and each 7 ml of buffer A containing 1.4 M and 0.25 M sucrose was further layered over the mixture. After centrifugation at 70,000 X g for 60 min in a Beckman SW-28 rotor, plasma membrane fraction was harvested from the interface between 0.25-1.4 M sucrose and washed with 50 mM Tris- HCl (pH 7.4), 1 mM EDTA, and 100 kallikrein inhibitory units/ml of aprotinin. The nuclear fraction, after being harvested from the bottom layer, was resuspended in buffer A containing 5 mM MgClz and digested with 250 pg/ml of deoxyribonuclease 1 and 250 pg/ml of ribonuclease A (13). The purified nuclei and plasma membranes thus obtained were resuspended in 50 mM Tris-HC1 (pH 7.4), 1 mM EDTA, and 100 kallikrein inhibitory units/ml of aprotinin and used in most experiments. For the removal of outer nuclear envelopes, the purified nuclei were washed with 1% citric acid, as described by Rubins et al. (9). The nuclear matrix was obtained from the treatment of the purified nuclei with 1% Triton X-100, based on the procedure re- ported by Berezner and Coffey (15).

In some experiments, IAP substrate protein in the purified nuclei was solubilized with 4 volumes of 50 mM Tris-HC1 (pH 7.4), 2 M NaCl, 250 mM dithiothreitol, and 1% sodium cholate a t 4 "C for 30 min and partially purified through a Sephacryl-S300 HR column by the method described previously (16). Plasma membrane-bound Gi was also obtained from rat livers by means of three sequential column chromatographies of DEAE-Sephacel, Sephacryl-S300 HR, and phenyl-Sepharose CL-4B (16).

Assay of ZAP-catalyzed ADP-ribosylation-Prior to ADP-ribosyla- tion, IAP (1 mg/ml in 2 M urea and 100 mM Napi) was activated by incubation a t 37 "C for 10 min with 4 volumes of 50 mM Tris-HC1 (pH 7.5) containing 100 mM dithiothreitol and 0.1 mM ATP. Various fractions from rat livers (10-20 pg of protein) were incubated a t 30 "C for 30 min with the activated IAP (0.5 pg) in 25 p1 of a reaction mixture containing 100 mM Tris-HC1 (pH 7.5), 2 p~ [a-"PINAD (5,000-10,000 cpm/pmol), 2 mM MgCl, 1 mM EDTA, 40 p~ GDPBS, 0.1 mM NADP, 10 mM thymidine, 1 mM ADP-ribose, and 20 mM nicotinamide. The ADP-ribosylation was terminated by the addition of an equal volume of 2-fold concentrated Laemmli buffer followed by boiling at 90 "C for 3 min. The aliquots were then subject to SDS- polyacrylamide gel (12%) electrophoresis (17, 18). The gels were stained with Coomassie Brilliant Blue R-250 and exposed to Kodak X-Omat AR film for 24-48 h with an intensifying screen a t -80 "C. In some experiments (Figs. 4 and 5), the intensity levels on the x-ray film were quantified using a Beckman DU-65 spectrophotometer with the gel scan accessory as described previously (19).

Immunoblot Procedure-Samples were separated by SDS-poly- acrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane a t a constant voltage of 50 V for 1 h (15). The membranes, after being blocked with 2% bovine serum albumin for 1 h, were incubated with polyclonal antibodies at 30 "C overnight and then treated with horseradish peroxidase-conjugated anti-rabbit IgG. The detection was carried out with a peroxidase immunostain kit (Wako Pure Chemical Industries, Tokyo).

In Vivo Pretreatment of Rats with Carbon Tetrachloride-Male donryu rats were intraperitoneally injected once with 200 p1 of 50% CC1, in olive oil and sacrificed at the indicated days after the injection. For the control experiments, 200 p1 of sterilized phosphate-buffered saline were used instead of the necrosis inducer of hepatocytes. Isolated livers were subjected to the preparations of nuclei, plasma membranes, and cytosol as described above. L-Tryptophan 2,3-diox- ygenase activity in the cytosol was assayed as described previously

(20). Sera were obtained from the rats, and the activity of glutamate- oxaloacetate transaminase was also measured (21).

Miscellaneous-The concentration of proteins was determined by the method of Lowry et al. (22) with bovine serum albumin as a standard. Electrophoresis on SDS-polyacrylamide gels was performed as described previously (19). Urea/SDS-polyacrylamide gel electro- phoresis was carried out as described by Ribeiro and Rodbell (23). The separating gel (8.5% acrylamide) contained 4 M urea instead of 8 M, and the electrophoresis was performed at a constant voltage of 125 V. Polyclonal antibodies, EC2 and AS7, were obtained from Du Pont-New England Nuclear. AP3 and @subunit antibodies were affinity-purified rabbit IgGs raised against the synthetic peptide GAGESGKSTIVKQMK (24, 25) and By-subunits purified from rat brain (26), respectively.

RESULTS

An IAP Substrate in Rat Liver Nuclei-Nuclear fraction was prepared from rat livers as described under "Experimental Procedures." Based on marker enzymic activities, such as 5'- nucleotidase, the nuclear fraction purified in the present study had a low contamination with plasma membranes or micro- somes (data not shown). When the nuclear fraction was incubated with IAP in the presence of [a-"PINAD and ana- lyzed by SDS-polyacrylamide gel electrophoresis, the radio- activity of 32P was incorporated into a 40-kDa protein (Fig. LA, lane 1) . The radiolabeling of the 40 kDa band in the nuclei was not observed unless IAP was added in the reaction mix- ture (lane 2). There was no labeling of the 40-kDa protein when the nuclear fraction that had been heated at 90 "C for 5 min or treated with trypsin was incubated with IAP and [32P]NAD (data not shown).

In addition to the nuclei, fractions of plasma membranes, microsomes, and cytosol were also prepared from rat livers and subjected to the assay of IAP-induced [32P]ADP-ribosyl- ation. The reaction mixtures that contained the same protein amounts were subjected to SDS-polyacrylamide gel electro- phoresis and followed by autoradiography (Fig. 1B). As ex- pected, the radioactivity of "P was incorporated into a 40- kDa protein in the plasma membranes (lane 3 ) but not in the cytoplasmic fraction (lane 1 ). The radiolabeled protein is very likely to be the a-subunit of Gi (mainly ai.*) present in the plasma membranes. A visible incorporation of the radioactiv- ity of 32P was also evident in the microsome-rich fraction

A B 1 2 3 4 1 ~- . 2 ""...T.*"i"-\ . ." ,

40 kDa --C

c40 kDa

FIG. 1. ADP-ribosylation of a 40-kDa protein by IAP in rat liver nuclei. Panel A, the nuclei purified from rat liver were incu- bated with 2 pM ["'PINAD in the presence (lane I ) or absence (lane 2) of 20 pg/ml of IAP. The radiolabeled proteins were then analyzed by SDS-polyacrylamide gel (12%) electrophoresis and autoradiogra- phy as described under "Experimental Procedures." Each lane con- tained approximately 15 pg of protein. Panel B, various fractions from rat livers (30 pg of protein) were also incubated with [cY-~*P] NAD plus IAP and subjected to SDS-polyacrylamide gel (10%) elec- trophoresis and autoradiography. Lanes 1-4, cytosol, microsome-rich fraction, plasma membranes, and nuclei, respectively.

A Nuclear GTP-binding Protein in Rat Liver 5087

(lane 2). The amount of radiolabeled 40-kDa protein in the nuclei (lane 4 ) appeared to be rather abundant in a compari- son with the microsomal fraction and was approximately a half of G,-n in the plasma membranes. These results suggested that the 40-kDa polypeptide serving as the substrate of IAP- catalyzed ADP-ribosylation might be abundantly present in rat liver nuclei in addition to the plasma membranes.

Comparison of IAP Suhstrate Proteins between Plasma Membranes and Nuclei-In order to investigate that the IAP substrate found here is really associated with the nuclei, the following experiments were carried out. The two IAP sub- strates in the plasma membranes and nuclei were first sub- jected to immunoblot analyses. As shown in Fig. 2, antibody AP3, which was raised against a common amino acid sequence of many a-subunits (24, 25), reacted with 40-kDa proteins in the plasma membranes (lane I ) and nuclei (lane 2). The two 40-kDa proteins appeared to be recognized by AS7, an ant i - body specific for n,-1 and ai-2 ( lanes 5 and 6). EC2, which was rather selective to n,-3 or e,,, reacted with a 40-kDn protein in the plasma membranes (lane 3 ) . There was, how- ever, no apparent band reacting with EC2 in the nuclei (lane 4 ) . An anti-n,, antibody did not react with any protein in the two fractions (data not shown). These results suggested that n,-1 and/or n,-2 were, but a,-3 was not, present in the nuclear fraction, though the plasma membranes contained tr,-3, as well as the two 0,-subunits.

[‘“PIADP-ribosylated 40-kDa proteins in the plasma mem- branes and nuclei were also analyzed by electrophoresis on a urea/SDS-polyacrylamide gel. As shown in Fig. 3A, the two radiolabeled proteins displayed different mobilities from each other; the latter mobilized faster than the former. The follow- ing differences were also observed between the two IAP sub- strates. The two fractions, after being treated with 1% Triton X-100 in the absence of high concentrations of NaCl or dithiothreitol, were centrifuged, and the resultant pellets were subjected to IAP-catalyzed [‘“PIADP-ribosylation. As shown i n Fig. 3R, the radiolabeling of the 40 kDa band was still observed in the treated nuclei (lane 2) but not in the plasma membranes (lane 1). When the nuclear 40-kDa protein, which had been initially [:’2P]ADP-ribosylated by IAP, was washed with 1% citric acid and followed by the extraction with 1% Triton X-100, the radioactivity in the 40-kDa protein was still retained in the nuclear fractions (Fig. 3C, lanes 1-3) .

fly-Suhunit-like Activity in Rat Liver Nuclei-By-subunits were essentially required for IAP-catalyzed ADP-ribosylation of the n-subunits of G proteins (17), and such an example is illustrated in Fig. 4A. The n-subunit of Gi.S purified from bovine brain membranes was not [‘“PIADP-ribosylated by IAP unless purified By-subunits were simultaneously added (compare lanes 3 with I ). Nuclei isolated from rat livers were

1 2 3 4 5 6

AP3 EC2 AS7 FIG. 2. Immunoblot analyses of the a-subunits of G proteins

in rat liver plasma membranes and nuclei. M I ’ suhstrate-rich fractions (approximately 4 0 pg) that had heen partially purified from t.he plasma memhranes (Innes I, 3, and 5 ) and nuclei (lanes 2, 4 , and A ) were electrophoresed and suhjected to immunohlot analyses as descrihed under “Experimental Procedures.” Antihodies used were A1’3 (lanes I and 2 ) , EC2 (lanes 9 and 4 ) . and AS7 (Innes 5 and 6 ) . respectively.

A 1 2 R 1 7

c 1 2 3

A 1 2 3 4 s B 1 2

40 kDa -c e36 kDa

first treated with IAP plus nonradioactive NAD in order t o modify the endogenous nuclear 40-kDa protein. There wns thus no [‘“P]AD~’-ribosylation of the 40-kDa band in the treated nuclear fraction ( I m p 4 ) . When the treated nuclei were mixed with the purified tr,-2 and then incubated with [‘“PINAD, the a-subunit was readily [“~‘~’]AI)I’-ribosvlat~rl by IAP (lane 5), as had been observed with the purified t i ? -

subunits. These results suggested that there were also I f?- subunits (or the subunit-like activity) in rat liver nuclei. The more direct evidence is shown in Fig. 4H, where nn imrnuno- blot analysis was performed. An antibody raised ngainst the P-subunit of G proteins reacted with not only the &sr~t~c~nit of plasma membrane G , (lnnc 2 ) but nlso n :If-kDa protein in the nuclei ( l a n ~ 1 ).

Properties of the ADF’-rihos~lntion of Nuclrnr [A[’ Suh- strate-We next studied properties of the ADP-rihosylation of nuclear IAP substrate. The substrnte activity o f all I A P - sensitive G proteins so fnr known was profnrlndly inhibited

5088 A Nuclear GTP-binding Protein in Rat Liver

by the addition of a nonhydrolyzable GTP analogue, GTPyS, since its binding to G proteins resulted in a dissociation into the nucleotide-bound a- and By-subunits (17). Thus, IAP- catalyzed ADP-ribosylation of the nuclear and plasma mem- brane substrates was carried out in the presence of various guanine nucleotides. As shown in Fig. 5A, Gi in the plasma membranes was less ADP-ribosylated by IAP upon the addi- tion of GTPyS. There was, however, no marked change in the ADP-ribosylation when GTP, GDP, or GDPBS was added. Essentially the same results were observed with the IAP substrate in the nuclei (Fig. 5B); only GTPyS exerted its inhibitory effect on the ADP-ribosylation of the nuclear 40- kDa protein. To check the nucleotide specificity for the nu- clear substrate, a competitive experiment was performed for the GTPyS-induced inhibition (Fig. 5C). The inhibitory ac- tion of GTPyS was blocked by the simultaneous addition of GDP or GDPBS, but not by ADP, ATP, or App(NH)p. Such characteristics were consistent with those observed with plasma membrane-bound G proteins. These results, together with the evidence of a By-subunit-like activity in the nuclei (see Fig. 4), suggested that there was a novel form of apy- trimeric G protein in rat liver nuclei, of which CY was approx- imately 40 kDa and ADP-ribosylated by IAP.

Effect of I n Vivo Pretreatment of Rats with a Necrosis Inducer of Hepatocytes on In Vitro Activity of the Nuclear I A P Substrate-In order to examine a possible role of the nuclear G protein in cell growths, rats were intraperitoneally injected once with a necrosis inducer of hepatocytes, CC1,. After the injection, livers were isolated to assay the IAP-substrate ac- tivity in the nuclei and plasma membranes at the indicated days. Glutamate-oxaloacetate transaminase activity in sera obtained from the pretreated rats was also measured as the indicator of hepatocyte necrosis (27,28). There was a marked increase in the glutamate-oxaloacetate transaminase activity at 1 day after the injection (Fig. 6 A ) . The elevated glutamate- oxaloacetate transaminase activity was returned to a normal level of the control rats within 3 days. In contrast, there was a gradual decrease in the L-tryptophan 2,3-dioxygenase activ- ity of the liver cytosol until 3 days, which was measured as a marker of the maturation and proliferation of hepatocytes (29), following by its elevation thereafter. This implied that

A 1 2 5cn 100

/

Control G T P 6

FIG. 5. Effects of guanine nucleotides on ADP-ribosylation of the nuclear IAP substrate. Panels A and B, the plasma mem- branes (panel A ) and nuclei (panel B ) were ["2P]ADP-ribosylated by IAP in the presence or absence of the indicated guanine nucleotides (80 PM). Panel C, the nuclei were ["PIADP-ribosylated by IAP in the presence of 25 PM GTPyS plus the indicated guanine nucleotides (100 PM). The samples were then subjected to SDS-polyacrylamide gel electrophoresis and autoradiography. In panel C, intensities of the 40-kDa protein on the autoradiogram are illustrated as percentages of the amount of ['"PIADP-ribosylation without any addition of nucleotides (Control).

: I 00 I".".Tl 0:j

Time afler CC14 pretreatment (days)

Nxkr

I 40 kDa

T C T C T C T C 0 7 3 6 16 days

FIG. 6. Effects of in vivo pretreatment of rats with carbon tetrachloride on in vitro activity of the nuclear IAP substrate. Rats were intraperitoneally injected once with CCl, and sacrificed at the indicated days after the injection (closed symbols with solid lines or T in panel C). For the control experiments, sterilized phosphate- buffered saline were used instead of CC1, (open symbols with dotted lines or C in panel C). At the indicated days, four rats were sacrificed, and nuclei, plasma membranes, and cytosol were isolated from the livers. Sera were also obtained from the rats. Panel A , glutamate- oxaloacetate transaminase (GOT) activity in the sera (0, 0) and L- tryptophan 2,3-dioxygenase (TO) activity in the cytosol (A, A) were measured as described under "Experimental Procedures." Panels B and C; the nuclei (0, 0) and plasma membranes (A, A) were ["'PI ADP-ribosylated by IAP and subjected to SDS-polyacrylamide gel electrophoresis and autoradiography (panel C) . The intensities of the 40-kDa protein are also illustrated as percentages of the amount of ["PIADP-ribosylation at day 0 (panel B ) .

the pretreatment with CCl, induced a partial necrosis of hepatocytes within 1 day and that the maturation or prolif- eration of new hepatocytes proceeded at 3-6 days.

Under these circumstances, there was a sustained increase in the nuclear IAP-substrate activity from 3 to 6 days after the CCl, treatment (Fig. 6B). The elevated activity was again returned to a normal level of the control rats at 16 days. There was, however, no apparent change in the IAP substrate activity of Gi in the plasma membranes. Similar results were obtained from three independent experiments. The apparent changes in the nuclear IAP substrate observed in the present studies appeared to be resulted from the real alterations of the substrate amount, since the activity was directly propor- tional to the amounts of nuclear protein and since the IAP substrate activity of purified G proteins was not modified in the presence of the various nuclear fractions (data not shown). I t is thus tempting to speculate that an increase in the IAP substrate activity of the nuclear G protein is responsible for hepatocyte growth.

DISCUSSION

In the present study, we found that a 40-kDa protein was specifically ADP-ribosylated by IAP in nuclear fraction pu- rified from rat livers (see Fig. 1). For preparation of the nuclei, we employed a homogenizing buffer containing no MgC12 or CaC12, which were usually added as a stabilizer for nuclear structure. Instead of such divalent cations, polyamines were added to the homogenizing buffer (29). This replacement was

A Nuclear GTP-binding Protein in Rat Liver 5089

very effective in increasing the recovery of the nuclear 40- kDa substrate from rat livers, probably due to an elimination from the divalent cation-induced inactivation of the substrate activity. A further addition of GDP to the above buffer (or a general M$+-containing buffer) was also useful for the sta- bilization of the activity. These improvements allowed us to first detect IAP substrate activity in the purified nuclei.

As shown in Fig. lB, IAP substrates appeared to be abun- dantly present in the nuclei in addition to the plasma mem- branes of rat livers. The IAP substrate in the nuclei, however, exerted its unique property apparently distinct from Gi in the plasma membranes as follows. 1) The nuclear 40-kDa protein had a higher mobility than the a-subunit of Gi upon electro- phoresis with a urea/SDS-containing polyacrylamide gel (see Fig. 3A), though it appeared to be recognized with an anti w- 1/2 antibody (see Fig. 2). 2) The nuclear protein was not extracted from the nuclei with 1% Triton X-100, whereas Gi was easily solubilized from the plasma membranes (see Fig. 3B). It is thus unlikely that the 40-kDa IAP substrate ob- served in the nuclei is simply due to a contamination with plasma membrane-bound Gi. The washing of the nuclear fraction with 1% citric acid and 1% Triton X-100 is reported to be effective in removing the outer nuclear envelopes (9) and both the outer and some parts of inner nuclear mem- branes (9, 15), respectively. The 40-kDa IAP substrate, how- ever, was not extracted from the nuclei by means of such treatments (see Fig. 3C). It is, therefore, likely that the nuclear IAP substrate is somehow associated with the nuclear matrix rather than outer envelopes, though there are still other possibilities such as the 40-kDa protein binding to cytoskel- eta1 systems associating with nuclei.

There was also a By-subunit-like activity in the nuclei (see Fig. 4A). A 36-kDa protein in the nuclei was, indeed, recog- nized by an antibody raised against the @-subunit of G pro- teins (Fig. 4B). Moreover, an apparent activity of the nuclear 40-kDa protein as the substrate of IAP-catalyzed ADP-ribo- sylation was profoundly affected by the presence of guanine nucleotides, as had been observed with plasma membrane- bound G proteins (see Fig. 5). These results suggested that an soy-trimeric G protein serving as the substrate of IAP was also present in rat liver nuclei.

Using a photoaffinity labeling technique, several papers concerning nuclear GTP-binding proteins have been previ- ously published. There were many proteins photoreactively labeled with [cx-~'P]GTP and/or [cx-~'P]ATP in the nuclei of Swiss 3T3 cells (10) and rat liver nuclear envelopes (9), of which 23-26-kDa proteins were rather specific for the guanine nucleotide. It was also reported that a 28-kDa GDP/GTP- binding protein was specifically localized in the nuclear en- velopes of rat livers (12). However, these reports did not address how the proteins were related to IAP substrates or the a-subunits of G proteins.

In the present studies, we also investigated a possible role of the nuclear G protein in hepatocyte growth (see Fig. 6). When rats were once injected with CC14, a necrosis inducer of hepatocytes, there was a marked increase in the nuclear substrate activity from 3 to 6 days without a significant change in the plasma membrane substrate activity. The time course of the increment corresponded with a recovering stage from the necrosis. Quite recently, Crouch (11) reported that the

immunoreactive a-subunit of Gi might be translocated from the plasma membranes to nuclear sites in response to insulin- or epidermal growth factor-induced mitosis in Balb/c 3T3 cells. It is thus likely that that the 40-kDa IAP substrate found in the present study might have a relation to the appearance of nuclear Gi-a after the mitosis of 3T3 cells, though our data suggest that the properties of the two G proteins in the nuclei and plasma membranes ( i e . Gi) are not identical with each other as discussed above. Even if such a transfer of GI-a from plasma membranes to nuclei occurs, a modification or processing of the molecule may be required for its translocation. The present data suggested that the nuclear G protein was located in inner nuclear membranes or the nuclear matrix. These sites are much concerned with various nuclear functions such as DNA replication, RNA synthesis, and RNA processing (30-32). Thus, it is tempting to speculate that the nuclear G protein is involved at some stages in the signal-transduction pathway of cell growth or differentiation. The purification and more characterization of the nuclear G protein would be of help in clarifying its nature and function.

REFERENCES 1. Gilman, A. G. (1987) Annu. Reu. Biochem. 56,615-649 2. Bourne, H. R., Sanders, D. A,, and McCormick, F. (1990) Nature 348 ,

3. West, R. E., Jr., Moss, J., Vaughan, M., Liu, T., and Liu, T.-Y. (1985) J. 125-132

4. Hoshino, S., Kikkawa, S., Takahashi, K., Itoh, H., Kaziro, Y., Kawasaki, Biol. Chern. 260,14428-14430

5. Masters, S. B., Sullivan, K. A., Miller, R. T., Beiderman, B., Lopez, N. G., H., Suzuki, K., Katada, T., and Ui, M. (1990) FEBS Lett. 2 7 6 , 227-231

6. Sullivan, K. A., Miller, R. T., Masters, S. B., Beiderman, B., Heideman, Ramachandran, J., and Bourne, H. R. (1988) Science 241,448-451

7. Fujinaga, Y., Morizumi, N., Sato, K., Tokumitau, K., Kondoh, Y., Ui, M., W., and Bourne, H. R. (1987) Nature 330 , 758-760

8. Tohkin, M., Iiri, T., Ui, M., and Katada, T. (1989) FEBS Lett. 255 , 187- and Okajima, F. (1989) FEBS Lett. 245,117-121

1 nn 9. Rubins, J. B., Benditt, J. O., Dickey, B. F., and Riedel, N. (1990) Proc.

10. Takeda, S., Sugiyama, H., Natori, S., and Sekimizu, K. (1989) FEBS Lett. Natl. Acad. Sci. U. S. A . 87, 7080-7084

244.469-472 11. Crouch, M. F. (1991) FASEB J. 5,200-206 12. Seydel, U., and Gerace, L. (1991) J. Biol. Chem. 266,7602-7608 13. Kaufmann, S. H., Gibson, W., and Shaper, J. H. (1983) J. Biol. Chem. 2 5 8 ,

271 0-271 9 14. Carey, D. J., and Hirschherg, C. B. (1980) J. Biol. Chem. 255,4348-4354 15. Berezney, R., and Coffey, D. S. (1977) J. Cell. Biol. 73,616-637 16. Kobayashi, I., Shibasaki, H., Takahashi, K., Tohyama, K., Kurachi, Y., Ito,

17. Katada, T., Oinuma, M., and Ui, M. (1986) J. Biol. Chem. 261,8182-8191 H., Ui, M., and Katada, T. (1990) Eur. J . Biochem. 191,499-506

18. Sternweis, P. C., and Robishaw, J. D. (1984) J. Biol. Chem. 259 , 13806-

- . - - . -.

1 RR1 R 19. IirL,"?.;Tohkin, M., Morishima, N., Ohoka, Y., Ui, M., and Katada, T.

20. Knox, W. E., Yip, A., and Reshef, L. (1975) Methods Enzymol. 17A, 415- (1989) J. Biol. Chem. 264,21394-21400

21. Bergmeyer, H. U., and Bernt, E. (1974) in Methods of Enzymatic Amlysis 419

(Bergmeyer, H. U., ed) 2nd Ed., Vol. 2, pp. 727-733, Academic Press, Nnur Vnrk

22. Lowry, 0. H., Rosehrough, N. J., Farr, A. L., and Randall, R. J. (1951) J.

23. Ribeiro, N. F. A. P., and Rodbell, M. (1989) Proc. Natl. Acad. Sei. U. S. A.

-.-.. "_.. Biol. Chem. 193,265-275

86.2577-2581 24. Mumby, S. M., Kahn, R. A., Manning, D. R., and Gilman, A. G. (1986)

25. Shibasaki, H., Kozasa, T., Takahashi, K., Inanobe, A., Kaziro, Y., Ui, M.,

26. Oinuma, M., Katada, T., and Ui, M. (1987) J. Biol. Chem. 262,8347-8353 27. Nakamura, T., Aoyama, K., and Tomomura, A. (1983) Biochem. Biophys.

Acta 678,91-97 28. Noda, C. , Tomomura, M., Nakamura, T., and Ichihara, A. (1984) J.

Bwchem. 95,37-45 29. Burgoyne, L. A,, Waqar, M. A,, and Atkinson, M. R. (1970) Biochem.

Biophys. Res. Commun. 39,254-259 30. Bodnar, J. W. (1988) J. Theor. Biol. 132,479-507 31. Razin, S. V. (1987) Bioessays 6,19-23 32. Verheijen, R., Van Venrooij, W., and Ramaekers, F. (1988) J. Cell Sci. 9 0 ,

Proc. Natl. Acad. Sci. U. S. A. 8 3 , 265-269

and Katada, T. (1991) FEBS Lett. 285,268-270

11-36