Evidence for the Participation of Cytosolic Protein Kinases in Membrane Phosphorylation in Intact Erythrocytes

Transcription

1 ~ ~~ ~ Eur J. Blochem X (197X) Evidence for the Participation of Cytosolic Protein Kinases in Membrane Phosphorylation in Intact Erythrocytes David A. PLUT. M. Marlene HOSEY. and Mariano TAO Department of Biological Chemistry. University of Illinois at the Medical Center. Chicago (Received Julj ) The effects of adenosine 3' : 5'monophosphate (cyclic AMP) on the phosphorylation of membrane proteins in intact rabbit and human erythrocytes were investigated. The addition of cyclic AMP to intact human or rabbit erythrocytes results in an increase in the incorporation of ~rtho[~~p]phospliate into several membrane protein components which are known to serve as substrates for the cyclicampdependent protein kinases. Thus this increase in protein phosphorylation is probably due to the activation of either soluble or membranebound cyclicampdependent protein kinases. Incubation of human erythrocytes in the presence of ~rtho[~'p]phosphate and cyclic AMP also leads to the phosphorylation of a membrane protein component, band 7, which has not been previously detected in the autophosphorylation of isolated ghosts. Since rabbit erythrocyte membranes do not contain any cyclicampdependent protein kinase, the results suggest that cytoplasmic kinases also play a role in the phosphorylation of membrane proteins in intact cells. Previous studies from several laboratories have established that the niembranes of human [I 51 and rabbit [3 51 erythrocytes possess multiple protein kinase and phosphoryl acceptor activities. Both cyclic AMPdependent and cyclicampindependent protein kinases are found in human erythrocyte membranes [2 51. In contrast. rabbit erythrocyte membranes contain only cyclicampindependent enzymes [3~ 51. This represents ii major difference between the human and the rabbit erythrocyte membrane autophosphorylation systems. The cyclicampdependent protein kinase in the human erythrocyte membranes catalyzes the phosphorylation of at least five membrane polypeptides, band and 4.8. Studies [4] of membrane phosphorylation in the presence of partially purified cytoplasmic protein kinases indicate that rabbit erythrocyte ghosts contain protein substrates which can be phosphorylated by the cyclicampdependent protein kinases isolated from the rabbit erythrocyte cytosol [6]. In order to further evaluate the role of cyclic AMP and inembratiebound and cytoplasmic protein kinases.4hhrcvitriiou. Cyclic AMP. adenosine 3': j'monophospiiate. Err:ww. Protein kinase [ t C ). in membane phospliorylation, the phosphorylation of membrane proteins was investigated in intact erythrocytes. The membrane phosphorylation patterns of intact cells were compared to the autophosphorylation patterns of isolated ghosts. The results indicate that both cyclic AMP and cytoplasmic factors or enzymes can regulate membrane phosphorylation in intact erythrocytes. EXPERIMENTAL PROCEDURE Mu terids Ortho[32P]phosphate and [Y~~P]ATP were purchased from Amersham/Searle. Cyclic AMP was obtained from Sigma Chemical Co. Reagents for gel electrophoresis were supplied by BioRad Laboratories. Methotly Erythrocytes were separated from freshly drawn human or rabbit blood by centrifugation. The packed erythrocytes were washed three times at 4 C in 10 volumes of a buffer (buffer A) containing 118 mm NaCI, 5 mm KCI, 1.7 mm MgC inm Na2HPO4,

2 334 Meiiibr,ine Phosphorylat~onin Intact Frythrocytes and 35 mm Tris; the entire solution was adjusted to ph 7.4 with HCI. The washed cells were incubated for 24 h at 37 "C with gentle agitation in a reaction mixture containing 0.1 ml of washed cells, Buffer B (same as buffer A but containing 10 mm glucose and no Na2HP04), i 200 pci 32P, (carrierfree) and? 0.1 mm cyclic AMP (made in buffer B). The volume was adjusted to 0.5 ml with buffer B. Three types of experiments were performed on the incubated cells. In the first type. ghosts were prepared according to Dodge p t (11. [7] from cells which had been incubated with 32P, f cyclic AMP. The phosphorylated membranes (about 25 pg) were analyzed using sodium dodecyl sulfate gel electrophoresis (4.5 O,, polyacrylamide. 0.3 sodium dodecyl sulfate) according to the method of Fairbanks ct ~ 1. [ 8and ] radioautography as previously described [3.4]. In the second type, ghosts were prepared from cells which had been incubated in the absence of 32P,but in the presence or absence of cyclic AMP. These ghosts were then autophosphorylated in the presence of [ Y ~ ~ P ] ATP cyclic AMP as previously described [3,4] and the phosphorylation analyzed as described above. In the third type, red cells were incubated in the presence and absence of cyclic AMP. At the end of the incubafied previously as substrates for the cytosolic cycliction, the cells were washed quickly three times (within AMPdependent protein kinase [4]. 15 min) in 10 volumes of buffer A at 0 C. 1 ml cold Since rabbit erythrocytes do not contain a 10 trichloroacetic acid was added with rapid agitamembranebound cyclicampdependent protein kition to the washed, packed cells. The resulting supernase [3.4], the results indicate that direct interactions between the cytoplasmic cyclicampdependent pronatant was processed and assayed for cyclic AMP according to the method of Gilman [91 as modified by tein kinases and the membrane substrates occur in the Murad 1'1 (21. [lo]. intact cells. The addition of cyclic AMP to the incubation medium leads to an activation of the cytoplasmic cyclicampdependent protein kinases and RESULTS AND DISCUSSION hence to an increase in the phosphorylation of specific membrane proteins. The data shown in Fig. 1 represent densitometric A siinilar study has been conducted in human tracings of radioautograms of gels of rabbit erythroerythrocytes (Fig.?). In the absence of cyclic AMP cyte membrane phosphopeptides obtained from intact cells which had been incubated for 2 ti with 32Pin the (Fig. 3A) the labelling of membrane proteins in the absence (Fig. 1 A) and presence (Fig. 1 B) of 0.1 inm intact cells incubated with 32P is qualitatively very similar to that obtained from autophosphorylation cyclic AMP. The phosphoprotein profiles shown iii Fig. 1 are diflerent from the autophosphorylation of isolated human erythrocyte ghosts in the presence profile of rabbit erythrocyte ghosts incubated in the of [y3'p]atp and cyclic AMP [3]. However, the presence of [:q32p]atp, Mg?', and f cyclic AMP 131. labelling in the band 3 region is somewhat depressed The major differences are in the phosphorylation of whereas the labelling of bands and 4.8 is slightly the minor protein components, 2.1, , 4.8 and enhanced in the intact cells compared to the auto5.5. In the autophosphorylation of isolated ghosts. phosphorylated ghosts [3]. That phosphorylation of the phosphorylation of these components occurs band 2.9 IS greater than band 3 in the intact cells either to a lesser extent or not at a11 However, the resembles that found in isolated ghosts incubnted with labelling pattern in intact cells appears to resemble the cytoplasmic casein kinases of rabbit erythrocytes qualitatively that obtained from the phosphorylation [ll]. Bands 4.5 and 4.8 are both substrates ofthecyclicof isolated ghosts in the presence of [.p3'p]atp, cyclic AMPdependent protein kinases and are phosphorylated in cells incubated in the absence of added cyclic AMP, and cytosolic cyclicampdependent protein kinase [4]. AMP (Fig. 2A) suggesting that the cyclicampdependent protein kinases are partially activated in the The labelling of erythrocyte membrane polypepintact cells. On the other hand, the phosphorylation tides 2.1, 4.1 and 4.8 In the intact cells was enhanced of proteins migrating in the regions of4.5 and 4.8 could by cyclic AMP. These polypeptides have been Identi'I,,

3 D. A. Plut. M. M. Hosey, and M. Tao 335 A Rabbit 2 21 Relative mobility Fig. 2..~ot/ililll t/ork~i;l'/.\li//ii/i, /~~J/!'~ie':,'/tltllitk' gc/ P/c~ctr(J/l/lorc,sisi.S of' liumun i~r.vtlirocj~tc, twtnhruiw phosphoprotcins isolared from intact cclls luhrlled with "Pi in the uhsrnce (A) or presence (B) qf'0.l mm cjdic AMP. The intact cells were incubated for 2 h in the presence of 200 pci of carrierfree 32Pi. Other experimental details were as described in Fig. 1 and in the text Fig. 3. Rucliouuiogrum depictitlg /he UutoJ~l/io.~~l/lo~J~/uti~Jil IJtrlter~7s c~f' rabbit erythrocyte membranes isolatd ji.m i*ells incubated nitll or without cyclic AMP. The cells were incubated for 2 h in the presence (B, D) or absence (A, C) of 0.1 mm cyclic AMP. Erythrocyte membranes were prepared from these incubated cells and autophosphorylated in the presence of [i.32p]atp(250 counts x min' x pmoli) with (C, D) or without (A, B) 2 pm of cyclic AMP. The radioautogram was exposed for three days also result from the action of other red cell cytoplasmic protein kinases. In the presence of exogenous cyclic AMP (Fig. 2 B) increased 32P incorporation is apparent in bands 2.1, 2.7, 4.1, 4.5, 4.8 and 7. The cyclicampstimulated phosphorylation of bands 2.1, 4.1, 4.5 and 4.8 is quantitatively similar to the autophosphorylation of ghosts [3] and thus most likely represents the activity of the membranebound cyclicampdependent protein kinase. Band *2.7 appears to be a membrane protein whose cyclicampdependent phosphorylation in ghosts is normally masked by band 2.93 phosphorylation (unpublished observation). That the cyclic AMPdependent phosphorylation of band 7 is only observed in studies utilizing whole cells suggests that relationships in situ among the various cellular components must be maintained for this reaction to occur. Phosphoprotein 7 may be a membrane protein which is phosphorylated in the intact cell by a cytosolic protein kinase. Alternatively, band 7 may be a cytosolic protein which becomes associated with the membrane upon phosphorylation and remains there during the hypotonic lysis of the cells. The effects of externally added cyclic AMP on the incorporation of 32P into membrane proteins appear to be specific and are not duplicated by either ATP or AMP (data not shown). In view of the effects of exogenous cyclic AMP on membrane phosphorylation in intact cells, it was of interest to study autophosphorylation of ghosts prepared from cells which had been incubated with cyclic AMP. Fig.3 is a radioautogram depicting the results of a representative experiment using rabbit erythrocytes. Each pair of gels represents duplicate treatments. The autophosphorylation pattern of ghosts derived from cyclicamptreated cells (Fig. 3 B and D) is different from that of control cells (Fig. 3A and C) but neither are altered by the presence of cyclic AMP in the autophosphorylation reaction. It is clear from the radioautograms that band 4.8 and to a lesser extent, band 4.5, of cyclicamptreated cells are autophosphorylated to a greater degree than those of untreated cells. Bands 4.5 and 4.8 have previously been identified as substrates for the soluble cyclicampdependent protein kinases [4]. The data suggest that incubation of intact cells with cyclic AMP causes a dissociation of the cytoplasmic cyclicampdependent protein kinases. The free catalytic subunits become attached to the membranes and remain associated with the membranes during hypotonic lysis. In support of this contention we found that the membranebound catalytic subunits could be removed from the ghosts by washing with 0.15 M KCI. The KCIwashed ghosts exhibited an autophosphorylation pattern similar to ghosts derived from cells incubated in the absence of cyclic AMP (unpublished observation). Fig. 4 illustrates the results of a similar experiment on human erythrocyte ghosts. Incubation of intact cells with cyclic AMP (Fig. 4 B) did not alter membrane autophosphorylation with ATP in the absence of cyclic AMP (Fig.4A versus B). However, a difference was observed in the autophosphorylation of ghosts

4 336 Membrane Phosphorylation in Intact Erythrocytes Human *Q \ 2 93\ 4 1 I 4 548r w i t h A T P plus cyclic A M P. As shown in big.4, bands 4.5 and 4.8 were phosphorylated to a lesser extent in membranes of cyclicamptreated cells (Fig. 4 D) than in membranes of untreated cells (Fig.4C). The reduction in phosphorylation in gel D might be attributed to cyclicampstimulated partial phosphorylation of certain membrane proteins o r the loss of the cyclicampdependent protein kinase catalytic subunits from membranes during the incubation of cells with cyclic A M P. The latter possibility is supported by preliminary findings of Owens and Haley [12] which indicate that cyclic A M P iiiay mediate the release of the c'italytic subunit froin the cell membranes Since exogenous cyclic A M P differentially stimulated the phosphorylation of membrane proteins in intact cells. it was of interest to determine whether this was due to a n alteration of the cyclic A M P concentration in erythrocytes. Table I shows the cyclic A M P levels in control cells and in cells which have been incubated for two hours with cyclic A M P The middle value represents a control to determine the effectiveness of the washing procedure in removing exogenous cyclic A M P. In this control experiment, the cells were incubated for two hours; at the end of the incubation period, 0.5 inm cyclic A M P was added and the cells immediately washed a s described in Fxperiniental procedure. This experiment indicates that most but not all of the exogenous added cyclic A M P was removed by the three quick washes. hence the determination of the concentration of cyclic A M P in the treated cells included a correction for this residual amount of cyclic A M P. The values were also corrected for recovery of cyclic A M P added to the extiaction and assay procedure. The results indicate that the control Table 1. C'i.c./ic~.4 M P ~ w i / ~ ow/ /h ~ r / m r t ierl./hroc,j*tc.r Human erythrocytes were incubated ill. buffer B i cyclic AMP (0.S m M ) for 3 h at 37 C. Cells were washed three times in 5 ml of buffer A within 15 min at 0 C. I nil of lo", trichloroacetic acid was added to the washed cells and the resulting supernatant was processed and assayed for cyclic AMP according to Murad P I ul. [lo]. Calculation of cyclic AMP concentration in cells is based on 2 x lo9 cellsimg protein [I61 and an internal volume of 100 pm3 [7]. The number in parentheses indicates the number of determinations Incubation conditions Cyclic AMP pmol,mg protein Control Control with cyclic A M P in first wash i 0.5 mm cyclic AMP nm (4) (3) (4) cells contain a small amount of cyclic A M P and that exogenous cyclic A M P causes a 2 3fold increase in these levels. Although the increase in cyclic A M P concentration (from 8.5 n M to 20 n M ) is modest, it is within the range of activation of cyclicampdependent protein kinases [ The data presented above indicate that exogenous cyclic A M P can differentially stimulate the phosphorylation of membrane proteins in erythrocytes presumably by activating membranebound and/or cytosolic cyclicampdependent protein kinases. That cytosolic protein kinases play a role in meinbrane phosphorylation seems evident in rabbit erythrocytes. Since rabbit erythrocyte membranes d o not contain a cyclicampdependent protein kinase [3], the stimulation of intact cell membrane phosphorylation by exogenous cyclic A M P probably results from the activation of the cytosolic cyclicampdependent protein kinases. This study was supported in part by grants liom the Chicago Heart Association and the American Cancer Society (BC6SC). M. M. Hosey is :I Nationnl Institute of Health National Research Fellow. M. Tao is a n Iistiiblished Investigator of the American Heart Asociation REFERENCES 1. Avruch. J. & Fairbanks. G. (1974) B ; ~ j [, / i [, f i t i, \ / r j, Fairbanks. G. & Avruch. J. (1974) Biol./ii,t,ii,sr,.~' Hosey. M. M. & Tao, M. (1976) Bi[jc,/i~,tiij.\/,.j.,IS % Hosey, M. M. & Tao, M. (1977) J. Biol. C / t c w i. 232, Hosey, M. M. & l'ao. M. (1977) B i d i b, i. B i o p l i ~ '..4c,rc7. ~. 182, TLIO. M. & Hackett. P. (1973) J. Biol. C'lti'tii Dodge. J. T., Mitchell. C. & Hannhan. 13. C. (1963) A r d i. Biol~hc.lll.Bio/,hj..s I Fairbanks. G., Stech. T. L. Sr Wnllach. [I.F. H. (1971) & ( I c,/icwii.s/rj, Gilman. A. G. (1970) Pro(,. h ' ~ ~ / A c t i.%i. il. C h

5 D. A. Plut. M. M. Hosey, and M. Tao Murad. F.. Manganiello, V. & Vaughn, M. (1971) Plot,. NNrl 14. Garren, L. D.. Gill. G. N. & Walton. G. M. (1971) Am. A. Y. Amti, Sci. U.S.A Acud. Sci Kumar. R. & Tao. M. (1975) Biocliim. Biopbjls. Acttr. 410, 15. Rubin, C. S.. Erlichman, J. & Rosen. 0. M. (1972) J. Bid C/7etn Owens. J. & Haley. B. (1977) J. Sirproniol. S/ruc.l. 6,.suppl. I, 16. Lepke, S., Fasold, H.. Pring, M. & Passow, H.( 1976)~. Mrrh Biol. 29% Walsh. D. A,. Perkins, J. P., Brostrom. C. 0.. Ho. E. S. & Krebs. E. G. (1971) J. Biol. Chern D. A. Plut. M. M. Closey. and M. Tao. Department of Biologicnl Chemistry, University of Chicago at the Medicnl Center. 835 South Wolcotl Avenue, Chicago. Illinois, U.S.A