Human brain protein phosphorylation in vitro: Cyclic AMP stimulation of electrophoretically-separated substrates

Brain Research, 222 (1981) 323-333

323

Elsevier/North-Holland Biomedical Press

HUMAN BRAIN PROTEIN PHOSPHORYLATION IN VITRO: CYCLIC AMP S T I M U L A T I O N OF E L E C T R O P H O R E T I C A L L Y - S E P A R A T E D SUBSTRATES

ARYEH ROUTTENBERG, DAVID G. MORGAN*, RICHARD G. CONWAY**, MICHAEL J. SCHMIDT and BERNARDINO GHETTI Cresap Neuroscienee Laboratory, Northwestern University, Evanston, Ill. 60201, (M.J.S.) Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Ind. 46206 and (B.G.) Department of PathoIogy, Indiana University School of Medicine, hzdianapolis, Ind. 46202 (U.S.A.)

(Accepted March 26th, 1981) Key words: human brain - - protein phosphorylation - - cyclic AMP - - pyruvate dehydrogenase - -

electrophoresis

SUMMARY In vitro phosphorylation of electrophoretically-separated brain proteins was studied in human frontal cortex obtained 3-16 h post-mortem from 13 patients ages 3 days-82 years with extensive, mild or no neuropathological involvement. In 12 of the 13 cases, cyclic A M P increased incorporation of phosphate into acid-precipitable protein. Analysis of the autoradiographic profiles of separate proteins indicated that phosphorylation of a doublet of molecular weight 86-80,000 was stimulated by cyclic A M P in certain samples. This doublet corresponded to the cyclic A M P stimulated doublet from rat frontal cortex we have termed band D-l,2 (proteins Ia and Ib of Ueda and Greengard°"°). Of special interest was the fact that, while co-migration was observed in the other phosphoprotein bands studied, band D-l,2 of humans consistently migrated slightly less than rat protein band D-l,2. This difference was not a function of post-mortem time, subcellular fraction or buffer used in the reaction phosphorylation assay. The use of post-mortem tissue was not a contributing factor as the retardation in band D-l,2 migration was still observed when post-mortem rat brain was used for comparison. In two human post-mortem samples, there was no measureable band D-l,2 phosphorylation even in the presence of cyclic AMP. This was the case in both h o m o g e nate and crude synaptosome/mitochondrial preparations. Band F-1 (tool. wt. = 47,000) was not observed in any of the h u m a n samples studied. This is consistent with prior studies in rat which show that band F-1 phospho*Andrus Gerontology Institute, University of Southern California, Los Angeles, Calif. 90007, U.S.A. **University of Illinois Chicago Medical Center, 1737 W. Polk, Chicago, Ill., U.S.A.

324 rylation is not detected in post-mortem brain. Band F-2 (mol. wt. 41,000) recently identified as pyruvate dehydrogenase 11, was lightly phosphorylated under the reaction conditions used in this study.

INTRODUCTION Study of in vitro phosphorylation of brain protein in experimental animals has suggested that this regulatory mechanism might play an important role in brain cell function and behavior7, la. To assess the relation between in vitro phosphorylation and the in vivo brain state 5, we studied the effects of method of sacrifice in relation to post-mortem time. It was found that cyclic AMP-dependent protein kinase activity of rat cerebral cortex was unchanged 16 h after death in rats 16. Moreover, even a 24 h post-mortem interval did not alter phosphorylation of most electrophoretically-separated proteins, nor influence the cyclic AMP stimulation of band D-l,2 ~ (protein Ia, Ib20). These findings in rodent brain made it feasible to study certain brain phosphoproteins in post-mortem human brain. We have recently shown, in fact, that protein kinase activity of human brain can be detected 19. Moreover, stimulation of kinase activity by cyclic AMP can be demonstrated in such tissue samples. The aim of the present studies was to characterize the electrophoretic profiles of phosphorylated proteins in detail beyond that presently available 3,6 and to use tissue samples which had been subjected to a thorough neuropathological analysis from a defined brain region. Such findings would permit assessment of the relation of in vitro phosphorylation to identifiable neurological disorders. MATERIALS AND METHODS Human brain tissue was obtained post-mortem from patients ranging from 2 days to 82 years of age who had been hospitalized at Indiana University Medical Center prior to death. No tissue was used in which the delay between death and autopsy was greater than 16 h. Neuropathological examination was performed on all cases. At autopsy a section was taken from the frontal pole of one of the cerebral hemispheres. Gray matter was dissected from white matter and stored at --70 °C until shipped in dry-ice to Cresap Neuroscience Laboratory, Northwestern University. Table I presents the list of patients, age at time of death, sex, time interval between death and autopsy, primary anatomical and neuropathological diagnoses. Adenosine-5'-triphosphate disodium salt (ATP), adenosine 3' :5'-monophosphoric acid (cyclic AMP), 2-mercaptoethanol, bromphenol blue, ethyleneglycol-bis-(betaaminoethyl ether) N,N'-tetraacetic acid (EGTA), phosphorylase, bovine catalase, ovalbumin, bovine serum albumin, and carbonic anhydrase were obtained from Sigma Chemicals. Enzyme grade sucrose from Schwarz-Mann was used for the centrifugation media. Ultra pure electrophoresis grade acrylamide, N,N'-methylene-bis-acrylamide,

325

i

~.o o;>

.~ ~

~ o ~o .~

~ . o~ .o~~_ ~ b

"00

"~ 0

. 0~-

.~

0

~

<

0
E

0~n ~0

-~

"~

8"~

=~

~

==

~ '-".== = = : o "

~._~ 0 o = . ->

~ = = . ~ o. =o ~ ~

~= -~ - ~ o o

o~ °

b

6

• ~,

~

"~

~.,-

B

~

~

,=-~

b~

0 ~:~0~

~ o ~'n 0

0

~

o

o'~

"0

~-d

~= ~ -~ o

~,-~

,.~ -~

•-

~

0

o'~

' ~

~,

-~

~

.n ~

}.o

.=

o0

-O

~

=~,B

-~

o

o~

•= .~

"~

~ ~ ..

o~'E

~ o ~

~

z

o~

.~

z

dn

.=_

.N

"2.

•E . ~ +

'~"

~.

t'--

~

t",l

r,~

~'~

0

°~ e~

0

326 and Tris-(hydroxymethyl) aminomethane (Tris) were purchased from Polysciences and ultragrade HEPES was obtained from Calbiochem. Sequanol grade sodium dodecyl sulfate (SDS) was from Pierce Chemical. N,N,N',N'-tetramethyl ethylene diamine (TEMED), and (ethylene dinitrilo) tetraacetic acid disodium salt (EDTA) were from Eastman Organic Chemicals. 3a70B Liquid scintillation cocktail was purchased through Research Products International. The (gamma-(32)P)ATP was supplied by ICN at a specific activity of 100-200 Ci/mmol.

Endogenous phosphorylation of electrophoretically-separated proteins Two series of studies were performed. In both series, tissue samples of human cerebral cortex obtained at autopsy, and of adult male rat cerebral cortex (Holtzman, Madison, Wisc.) from decapitated preparations dissected in the cold room (2 °C) were used. In the first series, tissue samples were weighed, slightly thawed in a solution of 0.32 M sucrose and were immediately homogenized in cold unbuffered 0.32 M sucrose, using 10-12 up and down strokes of a hand-held glass Dounce homogenizer with the small clearance pestle 'B'. The 10,000 >,~ g (20 min) P-2 fraction of Whittaker was prepared as beforeS, 15 and used as the crude intact synaptosome/mitochondrial preparation. Protein concentration was determined using bovine serum albumin as standard. In this first series of studies the endogenous in vitro phosphorylation assay was conducted in an acetate buffer. The assay, SDS-polyacrylamide gel electrophoresis, autoradiography of the gels and quantitative densitometry were performed as described previouslyS, 14. The phosphorylation assay contained 50 mM sodium acetate, 5/zM-(gamma-(32)P)ATP, which had been diluted with non-radioactive Tris.ATP to yield a final specific activity of 2-3 × 104 CPM/pmol, 5 # M cyclic AMP when present, and 84 #g protein in a final volume of 60 #1 (final protein concentration of 1.4 mg/ml). Samples were pre-incubated for 5 min at 30 °C, and the reaction initiated by the addition of (32)P-ATP. A 10 #1 aliquot of the reaction mixture was removed and the TCA filter paper precipitation procedure of Reimann, et al. 12 was performed to determine the total phosphate incorporation (see Table I) and to estimate autoradiographic exposure time. The reaction was stopped after 30 sec or 2 min by the addition of 'SDS stop solution' such that the final concentrations were 3 ~ SDS, 2 ~ 2-mercaptoethanol and 6 ~ sucrose. Samples were boiled for 2 rain and subjected to SDS-polyacrylamide gel electrophoresis on 1 0 ~ slab gels prepared by a modification of the method of Laemmli 9. A 15 ~ zone of acrylamide was layered at the bottom of the gel to resolve the low molecular weight H-bands (see Results). The 1.5 mm thick resolving gels were polymerized in a Hoefer SE500 unit containing 10 or 15 %oacrylamide, 0.27 ~o bisacrylamide, 0.375 M Tris.HC1 (pH 8.8), 0.1 ~ SDS, 0.25 ~ tetramethylethylenediamine, with 0.05 ~ ammonium persulfate as catalyst. The electrophoresis buffer contained 0.05 M Tris.HC1, 0.357 M glycine and 0.1 ~ SDS (final pH -- 8.3) in both the upper and lower buffer chambers; stacking gel was at pH 6.7. Using a constant voltage source, the gels were run at room temperature at 70 V for 30 rain, and then 155 V for 3~, h. Gels were stained with Coomassie blue and vacuum dried for autoradiography, using

327 K o d a k No-Screen X-ray film. Quantitative analysis of a u t o r a d i o g r a p h s followed procedures of prior reports 4'17 a n d were based on the high positive correlation ( + 0.87) between densitometric analysis and scintillation c o u n t i n g of gel slices. I n the second series of studies, reactions were carried out in H E P E S buffer rather t h a n acetate buffer. Additionally, h o m o g e n a t e a n d P-2 tissue preparations were c o m p a r e d a n d the gel was a u n i f o r m l0 ~ concentration. The h o m o g e n a t e assay was r u n at 0.6 m g / m l a n d the P-2 at 1.0 mg/ml. The assay contained a 50 m M H E P E S

®

(9

®

®

M

II H

D2 3

....

~

~

]

I

1

1 2

-

+

-

+

-

-t-

-

+

Fig. 1. In vitro phosphorylation of frontal cortex from human post-mortem tissue (cases 1 and 2) compared to freshly prepared (case 3) and 4 h post-mortem rat brain (case 4) in the absence ( I ) and presence (÷) of cyclic AMP. Sample 1 is case H-2, sample 2 is H-7. Note of band D-I, 2 in case 1. Note in case 2 slight retardation of migration of band D-l,2 relative to cases 3 and 4. Band nomenclature based on prior studies10a,ll with rat. Note absence of band E-1 and F-1 in human samples. Reaction, using 50 mM HEPES buffer (pH -- 7), and 10 mM MgC12, was initiated with 5 ttM (32)PATP. See Methods for further details.

i

328 buffer (pH = 7.1, adjusted with NaOH), 10 mM MgC12 and 5/zM (32)P-ATP. In all other respects, the methods used in the second series of studies were similar to those used in the first. RESULTS Fig. 1 demonstrates that it is possible to detect in vitro phosphorylation of human post-mortem brain proteins using electrophoretic methods. Phosphorylation of certain electrophoretically-separated protein substrates was increased in the presence of cyclic AMP. These substrates, band D-l,2, have been shown 13 to have similar properties as protein Ia and Ib of Ueda and Greengard 20. The failure of previous reports'~, 6 to demonstrate this cyclic AMP stimulation of band D-l,2 may have been due to the individual human post-mortem samples obtained (see Discussion).

Influence of post-mortem time and chronological age on total endogenous phosphorplation We examined the effect of the time interval between death and tissue preparation (until tissue was frozen) on incorporation of (32)P-ATP into acid-precipitable material in the presence of cyclic AMP (Table I). No systematic relation was found between endogenous phosphate turnover and time to autopsy (r -- 0.18, P < 0.50). Although rate of temperature rise post-mortem is different in rat and human, the present results are consistent with recent reports in experimental animals that a delay of up to 16 h has minimal effect on total phosphorylation of brain proteins in vitro 5 or brain protein kinase activity. The relation between endogenous phosphate turnover and chronological age at the time of death was also evaluated. Though a trend for samples from younger subjects to show higher incorporation was present, this was not statistically significant (r = --0.64, n = 13, P < 0.10).

Electrophoretically-separated human phosphoprotein reaction products Certain phosphoproteins in human brain co-migrate in the electropherogram with proteins previously identified in rat synaptosomes4,5, TM~,H. As shown in Fig. 1, band D-3, band E-2, E-3 and band F-2 in rat brain are seen to co-migrate with phosphorylated bands in human cerebral cortex. A cyclic-AMP stimulated doublet of a slightly higher molecular weight then band D-l,2 was observed in human tissue samples. Based on its clear capacity (Table II) to be stimulated by cyclic AMP, its doublet configuration and the similar molecular weight, the tentative assumption is that this substrate is band D-l,2 (see Discussion). The slight retardation in migration of band D-l,2 in human compared to rat was observed consistently in both homogenate and P-2 preparations when either acetate or HEPES buffer was used. In comparing samples from human and from rat in Fig. 1, differences in migration were not present in the other phosphoprotein bands. Note that in sample 4 which was from a post-mortem rat preparation, D-l,2 co-migrates with the 'fresh D- 1,2' of sample 3. Moreover, in Fig. 2 different subcellular preparations

329

-H 44~" 'A. q

-H

-H~t

b

&-,

a,O

q4q 'A.~.

N~ ~d

A' -Hq

5

s~ -H~

d

~r, 5

.,._; 0

V

~5

I+

I

II 0 o

N <

e

II

330

j® I ® I ® I ® l S

D E

'I 2

~"~ "~ .... ;g

3

~"~;.~

2

+"~ " : ~

1 2

i

--

+

-

+

-

+

-

+

Fig. 2. In vitro phosphorylation of frontal cortex from h u m a n case H-5 as a function of reaction duration (30 sec: 1-2; 2 min : 3-4) and subcellular preparation (homogenate: 1, 3 ; P-2: 2, 4). See Fig. 1 and Methods for further details.

or different reaction times did not alter migration distance. In sum, based on the tentative assumption that band D-I,2 of rat and human are similar, the possibility exists that human D-l,2 and rat D-l,2 differ slightly in molecular weight. The presence of cyclic AMP stimulation of band D-l,2 was clearly observed in certain samples (H-3, H-7). In other samples (H-2, H-15) phosphorylated substrates in this position were not observed. This absence is especially noteworthy because band D-1,2 phosphorylation, and stimulation by cyclic AMP were resistant to post-mortem manipulations5. Moreover, the absence of phosphorylation of band D-l,2 is still observed in the homogenate and cannot, therefore, be related to a difference in subcellular fractionation. The in vitro phosphorylation of band D-l,2 was observed in 11 of 13 samples of

331 human cortex P-2 fraction in the acetate buffer. The cyclic A M P stimulation ratio (i.e. the ratio of phosphorylation of band D-l,2 in the presence of 5 # M cyclic AMP to that in its absence) ranged from 1.0 to 3.0 and was greater than 1.5 in 10 of the samples. The mean value for the cyclic AMP stimulation ratio of band D-l,2 phosphorylation was 1.75 (see Table II). Thus, 5/~M cyclic AMP significantly (P < 0.01) increased phosphate incorporation into band D-l,2 in the P-2 fraction of human cortical tissue obtained at autopsy. This significant cyclic AMP stimulation was not observed in any other phosphoprotein band in the first series. In the second series, the improved separation of the E-bands indicates that band E-2 phosphorylation was, in fact, stimulated by cyclic AMP. Band D-3 phosphorylation was readily detected in human brain tissue. The lowest incorporation of radioactive phosphate into band D-3 occurred in samples from the two youngest individuals (H-4 and H-15). The effects of cyclic AMP on phosphorylation of band D-3 (Table II) were inconsistent, i.e. in two cases cyclic AMP stimulated (H-12, H-5), while in others (H-10, H-14, H-17 and H-3), cyclic A M P decreased phosphorylation. As shown in Table II, the highest levels of in vitro phosphorylation in human brain occurred in bands D-3, E (probably E-2 predominantly) and H-2. Band D-3 showed low phosphorylation in young subjects, but band E was, in general highly labeled in tissues from young, middle-aged and old subjects. Fig. 1 demonstrates an increase in cyclic AMP stimulation of band E-2, but no increase by such stimulation of band E-3. Similar differential results can be seen in both freshly prepared and post-mortem rat cerebral cortex. Thus, the difference between band E-2 and E-3 does not result from post-mortem effects. Special interest was given to band F-1 and band F-2 because of recent work indicating reactivity of these substrates to functional manipulations 11a, 13,14. AS shown in Fig. 1, band F-1 phosphorylation in rat is no longer detectable following a 4 h postmortem interval, confirming our earlier report 5. In line with this observation, then, we have also not observed band F-1 phosphorylation in human tissue. Of potential relevance is a faint band seen above the band F- l location in certain human samples (Figs. 1 and 2) and in the post-mortem rat (Fig. 1, case 4, -b cyclic AMP). This band was analyzed densitometrically in Table II and designated band *F-1. Band F-2 has recently been identified in our laboratory as pyruvate dehydrogenase11,11 a. A co-migrating band was observed in all human samples studied, and is best seen in homogenate, rather than P-2 material (Fig. 2). Its level of phosphorylation is not altered appreciably by changing buffer systems. The identification of 3 low molecular weight phosphoproteins in human brain as bands H-l, 2, 3 of the rat was made on the basis of electrophoretic co-migration on SDS-acrylamide gels using a discontinuous 10/15~ gel system (data not shown). Although phosphorylation of the H bands in the human was highly variable among samples, phosphorylation similar to that in the rat was seen in most cases. As shown in Table II, band H-3 was most highly phosphorylated, while band H-2 incorporated an intermediate amount of phosphate and band H-1 was the least phosphorylated of these three bands. These relative levels of labeling correspond to those observed in rat.

332 DISCUSSION This is the first report to demonstrate, in human brain, cyclic AMP stimulation of two closely migrating bands of 80,000 approximate molecular weight. Because this doublet migrates close to band D-1 and D-2, which is the well-characterized rat brain protein Ia and Ib of Ueda and Greengard z0, it is tentatively identified as a similar protein species with a slightly higher molecular weight. Prior reports studying human brain phosphorylation have either not observed 3 or not reported 6 on band D-l, 2 and its stimulation by cyclic AMP. As we have also not observed this band in several cases from human frontal cortex, it is conceivable that the individual samples obtained in these prior studies had the same properties as in the present instances where band D-I, 2 were not observed. Perhaps those cases lacking band D-l, 2 have neurological significance. While further accumulation of samples may answer this question, consideration may be given to explanations based on procedural factors. For example, biochemical1, 21 and immunocytochemical 2 evidence indicates that a considerable amount of band D-l, 2 is localized to presynaptic vesicles. Moreover, these proteins appeared to be associated with the external surface of the vesicle membrane. But the ultrastructure of a 10 min post-mortem brain tissue shows a rather severe clumping of synaptic vesicles la. It is possible, therefore, that this clumping or aggregation prevents the adequate solubilization of band D-l, 2 and thus proper entry into the gel matrix. While this explanation leaves unresolved the issue of why band D-l, 2 is absent in only certain cases, it points to possible alternative interpretions. The present results do hold out the possibility of evaluating subcellular specific alterations in a well-characterized brain phosphoprotein z0 as a function of neurological or neuropathological condition. It should be noted, in this regard, that when death was associated with brain pathology (described in Table I), endogenous phosphorylation and its stimulation by cyclic AMP was reduced relative to other samples in the study. Of special interest is the fact that the 4 cases diagnosed as 'normal brain', showed the highest levels of phosphorylation and cyclic AMP stimulation. Although these cases are from the younger subjects, case H-10, also a younger subject, had low phosphorylation and detectable neuropathological alterations. Moreover, no significant correlation was observed between in vitro phosphorylation and chronological age. Thus, the possible relation between brain pathology and in vitro phosphorylation merits further exploration. ACKNOWLEDGEMENTS Portions of this work were supported by NSF (BNS19388) and N I M H (MH25281) to A.R. We especially appreciate the efforts of Mr. Randall D. Morgan and Mr. Guy J. Hansen who were instrumental in helping obtain human brains for study. These experiments have been approved by the Human Subjects Committees both at the University of Indiana Medical School and at Northwestern University. Gratitude is expressed to Betty Wells for preparation of the manuscript.

333 REFERENCES 1 Berzins, K., Cohen, R., Blomberg, F. and Siekevitz, P., Specific occurrence in post-synaptic density and synaptic vesicles of two proteins phosphorylated by c-AMP dependent protein kinase, J. Cell. Biol., Abstr., 79, CN504 (1978). 2 Bloom, F. E., Ueda, T., Battenberg, E. and Greengard, P., Immunocytochemical localization, in synapses, of protein I, an endogenous substrate for protein kinases in mammalian brain, Proe. nat. Acad. Sei. (Wash.), 76, No. 11 (1979) 5982-5986. 3 Boehme, D. H., Kosecki, R. and Marks, N., Protein phosphorylation in human synaptosomal membranes: evidence for the presence of substrates for cyclic nucleotide guanosine 3'-5'-monophosphate dependent protein kinases, Brain Res. Bull., 3 (1979) 697-700. 4 Conway, R. G. and Routtenberg, A., Endogenous phosphorylation in vitro: selective effects of sacrifice methods on specific brain proteins, Brain Research, 139 (1978) 366-373. 5 Conway, R. G. and Routtenberg, A., Endogenous phosphorylation in vitro: differential effects of brain state (anesthesia, post-mortem) on electrophoretically separated brain proteins, Brain Research, 170 (1979) 313-324. 6 DeLorenzo, R. J., Antagonistic action of diphenylhydantoin and calcium on the level of phosphorylation of particular rat and human brain proteins, Brain Research, 134 (1977) 125-138. 7 Greengard, P., Possible role for cyclic nucleotides and phosphorylated membrane proteins in postsynaptic actions of neurotransmitters, Nature (Lond.), 260 (1976) 101 108. 8 Greengard, P., Phosphorylated proteins as physiological effectors, Science, 199 (1978) 146-152. 9 Laemmli, U. K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature (Lond.), 227 (1970) 680-685. 10 Lowry, O. H., Rosebrough, N. J., Farr, R. L. and Randall, R. J., Protein measurement with the Folin phenol reagent, J. biol. Chem., 193 (1951) 265-275. 10a Mitrius, J. C., Morgan, D. G. and Routtenberg, A., In vivo phosphorylation following z2p. orthophosphate injection into neostriatum or hippocampus: Selective and rapid labeling of electrophoretically separated brain proteins, Brain Research, 212 (1981) 67-81. I1 Morgan, D. G. and Routtenberg, A., Evidence that a 41,000 dalton brain phosphoprotein is pyruvate dehydrogenase, Biochem. biophys. Res. Commun., 95 (1980) 569-576. l l a Morgan, D. G. and Routtenberg, A., Brain pyruvate dehydrogenase: Phosphorylation and enzyme activity altered by a training experience, Science, in press. 12 Reimann, E. M., Walsh, D. A. and Krebs, E. G., Purification and properties of rabbit skeletal muscle adenosine 3'-5'-monophosphate-dependent protein kinase, J. cell Biol., 246 (1971) 19861995. 13 Routtenberg, A., Anatomical localization of phosphoprotein and glycoprotein substrates of memory, Progr. Neurobiol., 12 (1979) 85-113. 14 Routtenberg, A. and Benson, G., In vitro phosphorylation of a 41,000-MW protein band is selectively increased 24 h after footshock or learning, Behav. Neural. Biol., 29(2) (1980) 168-175. 15 Routtenberg, A. and Ehrlich, Y. H., Endogenous phosphorylation of four cerebral cortical membrane proteins: role of cyclic nucleotides, ATP and divalent cations, Brain Research, 92 (1975) 415430. 16 Routtenberg, A. and Tarrant, S., Synaptic morphology and related densities: Immediate post-mortem effects, Tiss. Cell, 6 (1974) 777-788. 17 Schmidt, M. J., Truex, L. L. and Thornberry, J. F., Cyclic nucleotides and protein kinase activity in the rat brain post-mortem, J. Neurochem., 31 (1978) 427-431. 18 Schmidt, M. J., Truex, L. L., Conway, R. G. and Routtenberg, A., Cyclic AMP-dependent protein kinase activity and synaptosomal protein phosphorylation in the brains of aged rats, J. Neurochem., 32 (1979) 335-344. 19 Schmidt, M. J., Truex, L. L., Ghetti, B. and Routtenberg, A., Cyclic AMP-dependent protein kinase activity in human brain across age, J. Neurochem., 35 (1980) 261-265. 20 Ueda, T. and Greengard, P., Adenosine 3'-5'-monophosphate-regulated phosphoprotein system of neuronal membranes. I. Solubilization, purification, and some properties of an endogenous phosphoprotein, J. cell Biol., 252 (1977) 5155-5163. 21 Ueda, T., Greengard, P., Berzins, K., Cohen, R. S., Blomberg, F., Grab, D. J. and Siekevitz, P., Subcellular distribution in cerebral cortex of two proteins phosphorylated by a cAMP-dependent protein kinase, J. cell BioL, 83 (1979) 308-319.