- Email: [email protected]
Requirement of ADP-ribosylation for the pertussis toxin-induced alteration in electrophoretic mobility of G-proteins
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages ] 227- ] 232
November 14, 1991
REQUIREMENT OF ADP-RIBOSYLATION FOR THE PERTUSSIS TOXIN-INDUCED ALTERATION IN ELECTROPHORETIC MOBILITY OF G-PROTEINS Sandra C. Roerig, Horace H. Loh and P.Y. Law Department of Pharmacology University of Minnesota Minneapolis, MN 55455 Received
September
9, 1991
S U M M A R Y : Pertussis toxin (PTX) catalyzes the ADP-ribosylation of the tx-subunit of GTPbinding proteins (G-proteins) in the presence of NAD +. Pertussis toxin also decreases the electrophoretic mobility of the tx-subunit on urea SDS PAGE. This effect of PTX has been suggested to be a property of the toxin different from its ability to catalyze ADP-ribosylation. However, the present report provides evidence to the contrary; ie, this mobility shift required the ADP-ribosylation of ct-subunits. This conclusion was based on: (1) in the presence of increasing concentrations of NAD + (0.026-1.3 ktM), there was a linear increase in the formation of the slower migrating tx-subunit as measured by immunoblotting with selective antisera, (2) addition of NADase to the incubation mixture completely eliminated the formation of this protein, and (3) increasing concentrations of nicotinamide (50-250 mM), which inhibits ADP-ribosylation, decreased the amount of the slower migrating ct-subunit. Thus, in addition to PTX, NAD + was required for the mobility shift and the slower migrating o~-subunit is likely the ADP-ribosylated form.
~
1991 Academic
Press,
Inc.
The involvement of certain GTP-binding proteins (G-proteins) in transduction of hormoneinduced signals is implicated if a hormone receptor-mediated event is attenuated by pertussis toxin (PTX) treatment. These membrane receptor-coupled G-proteins are heterotrimers consisting of ctl]? subunits and certain o~-subunits are substrates for PTX. Pertussis toxin catalyzes the transfer of ADP-ribose from NAD + to the tx-subunits, thus preventing the hormone receptor activation of the G-protein. The site for ADP-ribosylation is a cysteine residue four amino acid residues from the carboxy terminus (1) and a number of G-proteins have been identified which contain this cysteine (2). One method of detecting G-proteins which are present in a particular membrane involves utilization of [~x-32p]NAD+ and PTX to radiolabel the tx-subunits which can then be separated by SDS PAGE, visualized by autoradiography and subsequently identified with various antisera (3). However, a difficulty arises when attempting to identify a PTX substrate with antisera to ct-subunit because of a change in the electrophoretic mobility of substrate proteins after PTX treatment (4, 5,13). This change in electrophoretic mobility was proposed to be independent of PTX ADPribosylation activity, but due to some other activity of the toxin (4). In the process of identifying the G-protein tx-subunits in neuroblastoma x glioma NG10815 hybrid cell membranes, we also observed the shift in electrophoretic mobility induced by PTX. 0006-291X/91 $1.50 1227
Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
However, in contrast to conclusions from a previous report (4), this shift was dependent on the presence of NAD +, thus was likely due to the ADP-ribosylation of the t~-subunits.
METHODS
AND MATERIALS
Neuroblastoma x glioma NG108-15 hybrid cells were cultured as previously described (6). Briefly, cells were cultured in Dulbecco's modified Eagle's medium containing 0.1 mM hypoxanthine, 10 I~M aminopterin and 17 ~tM thymidine plus 5% fetal calf serum in a humidified atmosphere of 10% CO2 and 90% air. Membranes were prepared by homogenizing the ceils in 0.32 M sucrose, 2 mM EGTA, 20 mM HEPES pH 7.7 using a Dounce homogenizer (7). Homogenates were centrifuged at 1000 x g for 10 min to remove nuclei and supernatants were recentrifuged 60 re_in at 100,000 x g. Pellets (P2P3) were resuspended in 25 mM HEPES pH 7.7 to give a final protein concentration of approximately 10 mg/ml. Protein concentrations were determined by an adaptation of the method of Lowry et al. (8). Aliquots of membrane preparations were stored at -70oc. Membrane proteins were ADP-ribosylated with pertussis toxin (PTX) in an assay mixture containing 100 ktg protein, 0.25% Lubrol PX, 50 mM Tris-HC1 pH 8.0, 20 mM thymidine, 1 mM A T E 5 ~tM GTP, 20 mM arginine, 50 mM NaC1, 4 ktM MgC12, 100 mM dithiothreitol (DTr), 1 Ixg PTX and 3-5 ktCi [ct-32p]NAD + in a final volume of 0.1 ml. In some experiments, membranes were incubated with the above agents except the D T r , Lubrol and PTX but including 0.2 u/ml NADase (EC 3.2.2.5, Sigma Chemical Co.) for 30 min before the addition of these compounds and subsequent incubation for 2 hr. Following incubation for 2 hr at 30°C, 1 ml ice cold acetone was added to the mixtures and proteins were pelletted by centrifugation at 13,000 x g in a microfuge for 10 min at 4°C. Protein pellets were resuspended in 0.065 mM Tris-HC1 pH 6.8, 2% SDS, 10% glycerol, 5% 2-mercaptoethanol, heated at 95°C for 2 min and subjected to urea gradient SDS-PAGE (9% acrylamide gel in a linear gradient of 4-8 M urea (9, 10). Proteins in the gels were transferred to Immobilon-P membranes for exposure to X-ray film and subsequent immunoblotting. Intensity of bands on autoradiographs was quantified using the Lynx Molecular Biology Workstation from Applied Imaging Corp. which produced a linear response for cpm versus band intensity. Immunoblotting was performed as previously described (11) with overnight incubation in the primary antisera in a 1:1000 dilution. Antisera used were AS/7, which is selective for carboxy terminal sequences of Gilct,Gi2cz (12), (E.I. DuPont de Nemours & Co) and the antiserum selective for an internal amino acid sequence in Gi2ct (antiserum J-883), a gift from Dr. Susanne Mumby (13). The secondary antibody used was coupled to alkaline phosphatase (goat anti-rabbit IgG alkaline phosphatase conjugate, Bio-Rad Laboratories) and color was developed using BCIP/NBT Phosphatase Substrate System (Kirkegaard & Perry Laboratories, Inc.). The Image software program on a Macintosh II was used to quantify band intensity on the immunoblots. To facilitate identification of the [32p]ADP-ribosylated proteins, autoradiograph markers were attached to the Immobilon P transfers before exposure to X-ray film. After film exposure and development, the transfers were subjected to immonublotting with the various antisera and immunoreactive proteins were visualized colormetrically, with the alkaline-phosphatase coupled secondary antibody. The X-ray film could then be aligned exactly with the color developed transfer so that the bands could be unequivically identified.
RESULTS Membrane proteins from NG 108-15 cells were ADP-ribosylated in the presence of PTX and increasing concentrations of [ct-32p]NAD+ (0.26-1.3 [tM) and separated with SDS PAGE in a urea gradient. The autoradiogram in Figure la shows the five separate proteins between 39-41 kD tool wt which were PTX substrates in these membranes. These substrates were assigned numbers 1-5 in order from the top of the gel to the bottom. The greatest amount of [32p] radioactivity was
incorporated into bands #1 and 4, which showed approximately equal incorporation. Bands #2, 3, and 5 incorporated less radioactivity than did protein bands #1 and 4. 1228
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
a 46---~ i~~
PTX [32p]NAD+
1.3
+ 0
+ 0.026
+ 0.13
+ 1.3 I.tM
PTX [32p]NAD
ii~
1,3
,il
+ 0
+ 0.026
~ii
+ 0.13
+ 0.26
+ 1.3 p.M
Figure 1. Panel a. Autoradiogram of urea gradient SDS PAGE gel showing PTX substrates in the presence of increasing concenlrations (0-1.3 IXM)of [32p]NAD+. Panel b. Immunoblot of proteins transferred to Immobilon P using ASH antiserum from experiment similar to that in panel a.
Figure la also shows that increasing amounts of radioactivity were incorporated into all five bands with increasing concentrations of [tz-32p]NAD +. When immunoblotting with ASH antiserum (selective for Gil and 2ct) was performed, results typified by those shown in Figure lb were obtained. In the absence of PTX, a single immunopositive band was observed (lane 1). In the presence of added [o~-32p]NAD+, a second band with slower mobility in this SDS PAGE system, appeared and intensified with increasing concentrations of [cz-32p]NAD + (lanes 3-5). This band was likely an altered form of the faster migrating band as observed previously (4,5,13), and was possibly the ADP-ribosylated form (13). The slower migrating band was also observed even in the absence of added [tx-32p]NAD+ (lane 2). Thus, if this band was the ADP-ribosylated form, the observation of this band without added [c~-32p]NAD+ was possibly due to the presence of endogenous NAD +. To ascertain that endogenous NAD + could account for the slower migrating band observed in Figure lb, NADase was added to other incubation mixtures. Immunoblotting of Immobilon P transfers from these experiments with AS/7 antiserum gave results shown in Figure 2. In the presence of added NADase, the slower migrating band was not observed (lanes 4-6), even when [(x-32p]NAD + was added (lane 6). These results suggested that NAD + and PTX were both required for the formation of the slower migrating protein band, likely the ADP-ribosylated form of the ct-subunit. The mobility shift of proteins was further demonstrated to be dependent on ADPribosylation by the use of nicotinamide. Nicotinamide has been shown to block the ADPribosylating activity (5) as well as the glycohydrolase activity of PTX. However, Ribeiro-Neto and Rodbell (4) have reported that nicotinamide (up to 50 mM) did not block the PTX-induced mobility shift. Similar experiments were carried out in the present studies and results are shown in figure 3. The autoradiogram in figure 3a shows the effect of increasing concentrations of nicotimamide on the PTX-induced ADP-ribosylation using 1.3 I.tM [tz-32p]NAD +. In the absence of PTX no radioactivity was incorporated into proteins (lanes 1 and 3). Addition of PTX produced the normal five band pattern of radioactive incorporation (lane 2). When increasing concentrations of nicotinamide (5-250 mM) were added, a gradual decline in [32p] incorporation into all five protein bands occurred (lanes 4-7). Thus, as reported earlier (4, 5), nicotinamide attenuated the PTX-induced ADP-ribosylation Gct subunits. 1229
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
PTX [32PlNAD NADase
1.3
÷ 0
+ 1.3
+ 0 +
1.3 +
+ 1.3 IAM +
Figure 2. Immunoblot of proteins transferred to Immobilon P (from urea gradient SDS PAGE gel) ushag ASH antiserum. NG108-15 membranes were incubated in the absence (lanes 1-3) or presence (lanes 4-6) of NADase. PTX was added to the samples for lanes 2,3,5 and 6. 1.3 gM [32p]NAD+ was included in samples for lanes 1,3,4, and 6.
Immunoanalysis using the J-883 antiserum (selective for Gi2(x) produced the immunoblots represented by the one shown in figure 3b. In the absence of PTX, a single immunopositive protein band was observed (lanes 1 and 3). Addition of PTX resulted in the appearance of the slower migrating protein (lane 2). However, in the presence of increasing concentrations of nicotinamide, the slower migrating band gradually disappeared (lanes 4-7). Thus, in contrast to the previous report, (4) nicotinamide reversed the decrease in the mobility shift produced in the presence by PTX and N A D +. There was a direct correlation between [32p] incorporation and the mobility shift. With no detectable [32p]ADP-ribose incorporation at 250 mM nicotinamide, only one protein band was recognized by the J-883 antiserum (lane 7, fig.3a and b). Identical results were obtained using the AS/7 antiserum (data not shown).
a
PTX Nic
+
0
0
;0
+
+
5
s0
+
+
PTX
250
Nic
+
0
o
50
+
+
5
50
+
100
[] G 12 et ADP-rlbose
140 120 100 80
!
60 40 20 0 -20 1 O0
200
300
mM Nlcotlnamide
Figure 3. Panel a. Autoradiogram of urea gradient SDS PAGE gel. NG 108-15 membranes were incubated with [32p]NAD+ in the absence (lanes 1 and 3) or presence (lanes 2, 4,5,6, and 7) of PTX. Nicotinamide was added in increasing concentrations (0-250 raM) as indicated. Arrow shows the position of the 46 kDa standard protein. PaneI b. Immunoblot of proteins transferred to Immobilon P using J-883 antiserum from experiment similar to that m panel a. Panel c. Densitometric analysis of immunoblots from four different experiments, calculated as a percent of lane 3. Mean values + SEM are plotted. 1230
+
250ram
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
The immumoblots from four separate experiments (using different membrane preparations and both J-883 and ASP antisera) were digitized and the mean density values were normalized to the density of the band in lane 3 (0 PTX plus nicotinamide). Results expressed as per cent of density of lane 3 are shown in Figure 3c. The intensity of the faster migrating band (unribosylated Gi2a) was not significantally changed in the presence of increasing amounts of nicotinamide. The slower migrating band (ADP-ribosylated Gi2c0 was gradually eliminated by the nicotinamide. The intensity of the Gi2 band might be expected to increase when the intensity of the ribosylated Gi2 band decreased since the total protein amount remained the same. increase of the Gi2 band was observed (fig 3c).
A slight trend toward an
DISCUSSION When identifying the G-proteins which are expressed in selected membranes, one commonly used technique is to ADP-ribosylate the proteins with PTX or cholera toxin in the presence of [c~-32p]NAD+ (3). These toxins have specific sites of action in the ot-subunits and incorporation of the [32p| radioisotope provides useful information about the nature of the protein. Persussis toxin acts on the heterotrimeric form of transducin, Gi, and Go (2) by catalyzing the ADP-ribosylation of the cysteine residue near the carboxy terminus of these cx-subunits (tool wt 39-41 kDa). Cholera toxin catalyzes ADP-ribosylation of an internal arginine residue of Gs proteins (mol wt 45 kDa) which are not PTX substrates (14). Thus, this method provides a sensitive indication of the identity of the G-protein when the various the radiolabeled subunits are separated by SDS PAGE. More specific identification of specific et-subunits can be provided using antisera developed against known peptide sequences of the cloned proteins (15). However, PTX treatment changes the electrophoretic mobility of ¢x-subunits (4, 13) so that the radiolabeled band on an SDS PAGE gel may appear to have a different mobility from an immunopositive protein band. This apparent discrepancy becomes more obvious when the urea gradient SDS PAGE technique is used since greater separation of the various ~-subunits is obtained than with conventional SDS PAGE (10). In addition, if there are a number of PTX substrates in a given membrane, identification of each band becomes more difficult. The amount of NAD + used in radiolabeling experiments is usually not sufficient to completely ADP-ribosylate all the ot-subunit protein in a sample. The antisera interact with both ribosylated and non-ribosylated proteins. Thus, immunoanalysis of radiolabeUing experiments indicates the presence of both ribosylated and non-ribosylated forms of the protein. Direct comparison the mobilities of radiolabeled protein bands with immunoreactive protein bands becomes confusing since there are potentially two immunopositive bands for each radiolabeled band. In the present study, the shift in electrophoretic mobility was shown to require the presence of NAD +. Thus, ADP-ribosylation was responsible for the decreased mobility, a conclusion opposite from that obtained in a previous report (4). In order to unequivicaily identify a particular protein which incorporates [32p] from [ot-32p]NAD +, a separate reaction which includes PTX and a saturating amount of NAD + is required. This reaction will completely ribosylate the protein and shift the mobility of the immunoreactive band to the slower migrating form so that it can be 1231
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
compared to the radiolabeled protein band on the gel. This procedure will allow for identification of multiple PTX substrates in a selected membrane.
ACKNOWLEDGMENTS: The authors would like to thank Dr. Tim Walseth for his many valuable suggestions during the course of this study. Also, we are grateful for the J-883 antiserum which was kindly supplied by Dr. Suzanne Mumby. This work was supported by DA05370 (S.C.R.), DA00564 (H.H.L.) and DA07339 (P.Y.L.).
REFERENCES 1. West, R. E., Moss, J., Vaughan, M., Liu, T., Liu, T. Y. (1985) J. Biol. Chem. 260, 1442814430. 2. Birnbaumer, L., Perez-Reyes, E., Bertrand, P., Gudermann, T., Wei, X.-Y., Kim, H.-Y. , Castellano, A.-Y., Codina, J.-Y.. (1991) Biol. of Reprod. 44, 207-224. 3. Kopf, G. S., Woolkalis, M. J. (1991) Methods in Enzymology 195, 257-266. 4. Ribeiro-Neto, F. A. P., Rodbell, M. (1989) Proc. Natl. Acad. Sci. 86, 2577-2581. 5. Katada, T., Ui, M. (1982) Proc. Nad. Acad. Sci. 79, 3129-3133. 6. Law, P. Y., Koehler, J. E., Loh, H. H. (1982) Mol. Pharmacol. 21,483-491. 7. Law, P. Y., Wu, J., Koehler, J., Loh, H. H. (1981) J. Neurochem. 36, 1834-1846. 8. Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 9. Scherer, N. M., Toro, M.-J., Entman, M. L., Birnbaumer, L. (1987) Arch. Biochem. Biophys. 259, 431-440. 10. Codina, J., Grenet, D., Chang, K.-J., Birnbaumer, L.-J.. (1991) J. Receptor Res. 11,587601. 11. Harris, B. A., Robinshaw, J. D., Mumby, S. M., Gilman, A. G. (1985) Science 229, 12741277. 12. Goldsmith, P., Rossiter, K., Carter, A., Simonds, W., Unson, C. G., Vinitsky, R., Spiegel, A. M. (1988) J. Biol. Chem. 263, 6476-6479. 13. Mumby, S., Pang, I.-H., Gilman, A. G., Sternweis, P. C. (1988) J. Biol. Chem. 263, 2020-2026. 14. VanDop, C., Tsubokawa, M., Bourne, H. R., Ramachandran, J. (1984) J. Biol. Chem. 259, 696-699. 15. Mumby, S. M., Gilman, A. G. (1991) Methods in Enzymology 195, 215-233.
1232
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages ] 227- ] 232
November 14, 1991
REQUIREMENT OF ADP-RIBOSYLATION FOR THE PERTUSSIS TOXIN-INDUCED ALTERATION IN ELECTROPHORETIC MOBILITY OF G-PROTEINS Sandra C. Roerig, Horace H. Loh and P.Y. Law Department of Pharmacology University of Minnesota Minneapolis, MN 55455 Received
September
9, 1991
S U M M A R Y : Pertussis toxin (PTX) catalyzes the ADP-ribosylation of the tx-subunit of GTPbinding proteins (G-proteins) in the presence of NAD +. Pertussis toxin also decreases the electrophoretic mobility of the tx-subunit on urea SDS PAGE. This effect of PTX has been suggested to be a property of the toxin different from its ability to catalyze ADP-ribosylation. However, the present report provides evidence to the contrary; ie, this mobility shift required the ADP-ribosylation of ct-subunits. This conclusion was based on: (1) in the presence of increasing concentrations of NAD + (0.026-1.3 ktM), there was a linear increase in the formation of the slower migrating tx-subunit as measured by immunoblotting with selective antisera, (2) addition of NADase to the incubation mixture completely eliminated the formation of this protein, and (3) increasing concentrations of nicotinamide (50-250 mM), which inhibits ADP-ribosylation, decreased the amount of the slower migrating ct-subunit. Thus, in addition to PTX, NAD + was required for the mobility shift and the slower migrating o~-subunit is likely the ADP-ribosylated form.
~
1991 Academic
Press,
Inc.
The involvement of certain GTP-binding proteins (G-proteins) in transduction of hormoneinduced signals is implicated if a hormone receptor-mediated event is attenuated by pertussis toxin (PTX) treatment. These membrane receptor-coupled G-proteins are heterotrimers consisting of ctl]? subunits and certain o~-subunits are substrates for PTX. Pertussis toxin catalyzes the transfer of ADP-ribose from NAD + to the tx-subunits, thus preventing the hormone receptor activation of the G-protein. The site for ADP-ribosylation is a cysteine residue four amino acid residues from the carboxy terminus (1) and a number of G-proteins have been identified which contain this cysteine (2). One method of detecting G-proteins which are present in a particular membrane involves utilization of [~x-32p]NAD+ and PTX to radiolabel the tx-subunits which can then be separated by SDS PAGE, visualized by autoradiography and subsequently identified with various antisera (3). However, a difficulty arises when attempting to identify a PTX substrate with antisera to ct-subunit because of a change in the electrophoretic mobility of substrate proteins after PTX treatment (4, 5,13). This change in electrophoretic mobility was proposed to be independent of PTX ADPribosylation activity, but due to some other activity of the toxin (4). In the process of identifying the G-protein tx-subunits in neuroblastoma x glioma NG10815 hybrid cell membranes, we also observed the shift in electrophoretic mobility induced by PTX. 0006-291X/91 $1.50 1227
Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
However, in contrast to conclusions from a previous report (4), this shift was dependent on the presence of NAD +, thus was likely due to the ADP-ribosylation of the t~-subunits.
METHODS
AND MATERIALS
Neuroblastoma x glioma NG108-15 hybrid cells were cultured as previously described (6). Briefly, cells were cultured in Dulbecco's modified Eagle's medium containing 0.1 mM hypoxanthine, 10 I~M aminopterin and 17 ~tM thymidine plus 5% fetal calf serum in a humidified atmosphere of 10% CO2 and 90% air. Membranes were prepared by homogenizing the ceils in 0.32 M sucrose, 2 mM EGTA, 20 mM HEPES pH 7.7 using a Dounce homogenizer (7). Homogenates were centrifuged at 1000 x g for 10 min to remove nuclei and supernatants were recentrifuged 60 re_in at 100,000 x g. Pellets (P2P3) were resuspended in 25 mM HEPES pH 7.7 to give a final protein concentration of approximately 10 mg/ml. Protein concentrations were determined by an adaptation of the method of Lowry et al. (8). Aliquots of membrane preparations were stored at -70oc. Membrane proteins were ADP-ribosylated with pertussis toxin (PTX) in an assay mixture containing 100 ktg protein, 0.25% Lubrol PX, 50 mM Tris-HC1 pH 8.0, 20 mM thymidine, 1 mM A T E 5 ~tM GTP, 20 mM arginine, 50 mM NaC1, 4 ktM MgC12, 100 mM dithiothreitol (DTr), 1 Ixg PTX and 3-5 ktCi [ct-32p]NAD + in a final volume of 0.1 ml. In some experiments, membranes were incubated with the above agents except the D T r , Lubrol and PTX but including 0.2 u/ml NADase (EC 3.2.2.5, Sigma Chemical Co.) for 30 min before the addition of these compounds and subsequent incubation for 2 hr. Following incubation for 2 hr at 30°C, 1 ml ice cold acetone was added to the mixtures and proteins were pelletted by centrifugation at 13,000 x g in a microfuge for 10 min at 4°C. Protein pellets were resuspended in 0.065 mM Tris-HC1 pH 6.8, 2% SDS, 10% glycerol, 5% 2-mercaptoethanol, heated at 95°C for 2 min and subjected to urea gradient SDS-PAGE (9% acrylamide gel in a linear gradient of 4-8 M urea (9, 10). Proteins in the gels were transferred to Immobilon-P membranes for exposure to X-ray film and subsequent immunoblotting. Intensity of bands on autoradiographs was quantified using the Lynx Molecular Biology Workstation from Applied Imaging Corp. which produced a linear response for cpm versus band intensity. Immunoblotting was performed as previously described (11) with overnight incubation in the primary antisera in a 1:1000 dilution. Antisera used were AS/7, which is selective for carboxy terminal sequences of Gilct,Gi2cz (12), (E.I. DuPont de Nemours & Co) and the antiserum selective for an internal amino acid sequence in Gi2ct (antiserum J-883), a gift from Dr. Susanne Mumby (13). The secondary antibody used was coupled to alkaline phosphatase (goat anti-rabbit IgG alkaline phosphatase conjugate, Bio-Rad Laboratories) and color was developed using BCIP/NBT Phosphatase Substrate System (Kirkegaard & Perry Laboratories, Inc.). The Image software program on a Macintosh II was used to quantify band intensity on the immunoblots. To facilitate identification of the [32p]ADP-ribosylated proteins, autoradiograph markers were attached to the Immobilon P transfers before exposure to X-ray film. After film exposure and development, the transfers were subjected to immonublotting with the various antisera and immunoreactive proteins were visualized colormetrically, with the alkaline-phosphatase coupled secondary antibody. The X-ray film could then be aligned exactly with the color developed transfer so that the bands could be unequivically identified.
RESULTS Membrane proteins from NG 108-15 cells were ADP-ribosylated in the presence of PTX and increasing concentrations of [ct-32p]NAD+ (0.26-1.3 [tM) and separated with SDS PAGE in a urea gradient. The autoradiogram in Figure la shows the five separate proteins between 39-41 kD tool wt which were PTX substrates in these membranes. These substrates were assigned numbers 1-5 in order from the top of the gel to the bottom. The greatest amount of [32p] radioactivity was
incorporated into bands #1 and 4, which showed approximately equal incorporation. Bands #2, 3, and 5 incorporated less radioactivity than did protein bands #1 and 4. 1228
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
a 46---~ i~~
PTX [32p]NAD+
1.3
+ 0
+ 0.026
+ 0.13
+ 1.3 I.tM
PTX [32p]NAD
ii~
1,3
,il
+ 0
+ 0.026
~ii
+ 0.13
+ 0.26
+ 1.3 p.M
Figure 1. Panel a. Autoradiogram of urea gradient SDS PAGE gel showing PTX substrates in the presence of increasing concenlrations (0-1.3 IXM)of [32p]NAD+. Panel b. Immunoblot of proteins transferred to Immobilon P using ASH antiserum from experiment similar to that in panel a.
Figure la also shows that increasing amounts of radioactivity were incorporated into all five bands with increasing concentrations of [tz-32p]NAD +. When immunoblotting with ASH antiserum (selective for Gil and 2ct) was performed, results typified by those shown in Figure lb were obtained. In the absence of PTX, a single immunopositive band was observed (lane 1). In the presence of added [o~-32p]NAD+, a second band with slower mobility in this SDS PAGE system, appeared and intensified with increasing concentrations of [cz-32p]NAD + (lanes 3-5). This band was likely an altered form of the faster migrating band as observed previously (4,5,13), and was possibly the ADP-ribosylated form (13). The slower migrating band was also observed even in the absence of added [tx-32p]NAD+ (lane 2). Thus, if this band was the ADP-ribosylated form, the observation of this band without added [c~-32p]NAD+ was possibly due to the presence of endogenous NAD +. To ascertain that endogenous NAD + could account for the slower migrating band observed in Figure lb, NADase was added to other incubation mixtures. Immunoblotting of Immobilon P transfers from these experiments with AS/7 antiserum gave results shown in Figure 2. In the presence of added NADase, the slower migrating band was not observed (lanes 4-6), even when [(x-32p]NAD + was added (lane 6). These results suggested that NAD + and PTX were both required for the formation of the slower migrating protein band, likely the ADP-ribosylated form of the ct-subunit. The mobility shift of proteins was further demonstrated to be dependent on ADPribosylation by the use of nicotinamide. Nicotinamide has been shown to block the ADPribosylating activity (5) as well as the glycohydrolase activity of PTX. However, Ribeiro-Neto and Rodbell (4) have reported that nicotinamide (up to 50 mM) did not block the PTX-induced mobility shift. Similar experiments were carried out in the present studies and results are shown in figure 3. The autoradiogram in figure 3a shows the effect of increasing concentrations of nicotimamide on the PTX-induced ADP-ribosylation using 1.3 I.tM [tz-32p]NAD +. In the absence of PTX no radioactivity was incorporated into proteins (lanes 1 and 3). Addition of PTX produced the normal five band pattern of radioactive incorporation (lane 2). When increasing concentrations of nicotinamide (5-250 mM) were added, a gradual decline in [32p] incorporation into all five protein bands occurred (lanes 4-7). Thus, as reported earlier (4, 5), nicotinamide attenuated the PTX-induced ADP-ribosylation Gct subunits. 1229
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
PTX [32PlNAD NADase
1.3
÷ 0
+ 1.3
+ 0 +
1.3 +
+ 1.3 IAM +
Figure 2. Immunoblot of proteins transferred to Immobilon P (from urea gradient SDS PAGE gel) ushag ASH antiserum. NG108-15 membranes were incubated in the absence (lanes 1-3) or presence (lanes 4-6) of NADase. PTX was added to the samples for lanes 2,3,5 and 6. 1.3 gM [32p]NAD+ was included in samples for lanes 1,3,4, and 6.
Immunoanalysis using the J-883 antiserum (selective for Gi2(x) produced the immunoblots represented by the one shown in figure 3b. In the absence of PTX, a single immunopositive protein band was observed (lanes 1 and 3). Addition of PTX resulted in the appearance of the slower migrating protein (lane 2). However, in the presence of increasing concentrations of nicotinamide, the slower migrating band gradually disappeared (lanes 4-7). Thus, in contrast to the previous report, (4) nicotinamide reversed the decrease in the mobility shift produced in the presence by PTX and N A D +. There was a direct correlation between [32p] incorporation and the mobility shift. With no detectable [32p]ADP-ribose incorporation at 250 mM nicotinamide, only one protein band was recognized by the J-883 antiserum (lane 7, fig.3a and b). Identical results were obtained using the AS/7 antiserum (data not shown).
a
PTX Nic
+
0
0
;0
+
+
5
s0
+
+
PTX
250
Nic
+
0
o
50
+
+
5
50
+
100
[] G 12 et ADP-rlbose
140 120 100 80
!
60 40 20 0 -20 1 O0
200
300
mM Nlcotlnamide
Figure 3. Panel a. Autoradiogram of urea gradient SDS PAGE gel. NG 108-15 membranes were incubated with [32p]NAD+ in the absence (lanes 1 and 3) or presence (lanes 2, 4,5,6, and 7) of PTX. Nicotinamide was added in increasing concentrations (0-250 raM) as indicated. Arrow shows the position of the 46 kDa standard protein. PaneI b. Immunoblot of proteins transferred to Immobilon P using J-883 antiserum from experiment similar to that m panel a. Panel c. Densitometric analysis of immunoblots from four different experiments, calculated as a percent of lane 3. Mean values + SEM are plotted. 1230
+
250ram
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
The immumoblots from four separate experiments (using different membrane preparations and both J-883 and ASP antisera) were digitized and the mean density values were normalized to the density of the band in lane 3 (0 PTX plus nicotinamide). Results expressed as per cent of density of lane 3 are shown in Figure 3c. The intensity of the faster migrating band (unribosylated Gi2a) was not significantally changed in the presence of increasing amounts of nicotinamide. The slower migrating band (ADP-ribosylated Gi2c0 was gradually eliminated by the nicotinamide. The intensity of the Gi2 band might be expected to increase when the intensity of the ribosylated Gi2 band decreased since the total protein amount remained the same. increase of the Gi2 band was observed (fig 3c).
A slight trend toward an
DISCUSSION When identifying the G-proteins which are expressed in selected membranes, one commonly used technique is to ADP-ribosylate the proteins with PTX or cholera toxin in the presence of [c~-32p]NAD+ (3). These toxins have specific sites of action in the ot-subunits and incorporation of the [32p| radioisotope provides useful information about the nature of the protein. Persussis toxin acts on the heterotrimeric form of transducin, Gi, and Go (2) by catalyzing the ADP-ribosylation of the cysteine residue near the carboxy terminus of these cx-subunits (tool wt 39-41 kDa). Cholera toxin catalyzes ADP-ribosylation of an internal arginine residue of Gs proteins (mol wt 45 kDa) which are not PTX substrates (14). Thus, this method provides a sensitive indication of the identity of the G-protein when the various the radiolabeled subunits are separated by SDS PAGE. More specific identification of specific et-subunits can be provided using antisera developed against known peptide sequences of the cloned proteins (15). However, PTX treatment changes the electrophoretic mobility of ¢x-subunits (4, 13) so that the radiolabeled band on an SDS PAGE gel may appear to have a different mobility from an immunopositive protein band. This apparent discrepancy becomes more obvious when the urea gradient SDS PAGE technique is used since greater separation of the various ~-subunits is obtained than with conventional SDS PAGE (10). In addition, if there are a number of PTX substrates in a given membrane, identification of each band becomes more difficult. The amount of NAD + used in radiolabeling experiments is usually not sufficient to completely ADP-ribosylate all the ot-subunit protein in a sample. The antisera interact with both ribosylated and non-ribosylated proteins. Thus, immunoanalysis of radiolabeUing experiments indicates the presence of both ribosylated and non-ribosylated forms of the protein. Direct comparison the mobilities of radiolabeled protein bands with immunoreactive protein bands becomes confusing since there are potentially two immunopositive bands for each radiolabeled band. In the present study, the shift in electrophoretic mobility was shown to require the presence of NAD +. Thus, ADP-ribosylation was responsible for the decreased mobility, a conclusion opposite from that obtained in a previous report (4). In order to unequivicaily identify a particular protein which incorporates [32p] from [ot-32p]NAD +, a separate reaction which includes PTX and a saturating amount of NAD + is required. This reaction will completely ribosylate the protein and shift the mobility of the immunoreactive band to the slower migrating form so that it can be 1231
Vol. 180, No. 3, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
compared to the radiolabeled protein band on the gel. This procedure will allow for identification of multiple PTX substrates in a selected membrane.
ACKNOWLEDGMENTS: The authors would like to thank Dr. Tim Walseth for his many valuable suggestions during the course of this study. Also, we are grateful for the J-883 antiserum which was kindly supplied by Dr. Suzanne Mumby. This work was supported by DA05370 (S.C.R.), DA00564 (H.H.L.) and DA07339 (P.Y.L.).
REFERENCES 1. West, R. E., Moss, J., Vaughan, M., Liu, T., Liu, T. Y. (1985) J. Biol. Chem. 260, 1442814430. 2. Birnbaumer, L., Perez-Reyes, E., Bertrand, P., Gudermann, T., Wei, X.-Y., Kim, H.-Y. , Castellano, A.-Y., Codina, J.-Y.. (1991) Biol. of Reprod. 44, 207-224. 3. Kopf, G. S., Woolkalis, M. J. (1991) Methods in Enzymology 195, 257-266. 4. Ribeiro-Neto, F. A. P., Rodbell, M. (1989) Proc. Natl. Acad. Sci. 86, 2577-2581. 5. Katada, T., Ui, M. (1982) Proc. Nad. Acad. Sci. 79, 3129-3133. 6. Law, P. Y., Koehler, J. E., Loh, H. H. (1982) Mol. Pharmacol. 21,483-491. 7. Law, P. Y., Wu, J., Koehler, J., Loh, H. H. (1981) J. Neurochem. 36, 1834-1846. 8. Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 9. Scherer, N. M., Toro, M.-J., Entman, M. L., Birnbaumer, L. (1987) Arch. Biochem. Biophys. 259, 431-440. 10. Codina, J., Grenet, D., Chang, K.-J., Birnbaumer, L.-J.. (1991) J. Receptor Res. 11,587601. 11. Harris, B. A., Robinshaw, J. D., Mumby, S. M., Gilman, A. G. (1985) Science 229, 12741277. 12. Goldsmith, P., Rossiter, K., Carter, A., Simonds, W., Unson, C. G., Vinitsky, R., Spiegel, A. M. (1988) J. Biol. Chem. 263, 6476-6479. 13. Mumby, S., Pang, I.-H., Gilman, A. G., Sternweis, P. C. (1988) J. Biol. Chem. 263, 2020-2026. 14. VanDop, C., Tsubokawa, M., Bourne, H. R., Ramachandran, J. (1984) J. Biol. Chem. 259, 696-699. 15. Mumby, S. M., Gilman, A. G. (1991) Methods in Enzymology 195, 215-233.
1232