Inhibition of Protein Phosphatase 1 Stimulates Secretion of Alzheimer Amyloid Precursor Protein

Edgar F. da Cruz e Silva, Odete A. B. da Cruz e Silva, Cassia Thais B. V. Zaia, and Paul Greengard The Rockefeller University, Laboratory of Molecular and Cellular Neuroscience, New York, New York, U.S.A.

ABSTRACT Background: Aberrant metabolism of the Alzheimer Results: The availability of protein phosphatase inhibi- amyloid precursor protein (APP) or its amyloidogenic A,B tors with different relative potencies against the different fragment is thought to be centrally involved in Alzhei- types of serine/threonine-specific protein phosphatase mer's disease. Nonamyloidogenic processing of APP in- allowed us to examine which of the four known types of volves its cleavage within the AP3 domain by a protease, protein phosphatase might be involved in the regulation termed a-secretase, and release of the large extracellular of APP secretion. Both okadaic acid and calyculin A domain, termed APPs. Secretion of ApPs can be stimu- stimulated the secretion of APP from COS-1 cells in a lated by phorbol esters, activators of protein kinase C, dose-dependent manner. The half-maximal dose for with concurrent inhibition of AP production. While the stimulation of APP secretion was approximately 100-fold role of protein kinases on APP metabolism has been higher with okadaic acid than with calyculin A. investigated, considerably less effort has been devoted to Conclusions: The nearly 100-fold difference in the ob- elucidating the role played by protein phosphatases. served IC50 values for okadaic acid and calyculin A im- Okadaic acid, a protein phosphatase inhibitor, has been plicates a type 1 protein phosphatase in the control of shown to stimulate secretion of APPs, but the identity of APPs production. Protein phosphatase 1 (PP1) is known the phosphatase involved has not been investigated. to be highly expressed in adult mammalian brain, both Materials and Methods: The secretion of APP5 from in neurons and glia. The identification of a specific phos- COS-1 cells was measured in the absence or presence of phatase type in the control of APP secretion opens new various doses of serine/threonine-specific phosphatase avenues to the development of rational therapeutic in- inhibitors. Quantitation of the derived IC50 values was tervention strategies aimed at the prevention and/or used to determine the identity of the phosphatase in- treatment of Alzheimer's Disease. volved in the control of APP secretion.

INTRODUCTION genic pathway by a protease (a-secretase) that Alzheimer's disease (AD) is the most common of cleaves APP within the AP3 domain and releases the neurodegenerative diseases, probably affect- the large extracellular domain (APPS). ing more than 50% of individuals over the age of Previous work has shown that treatment of a 80. Central to the pathology of the disease is the variety of cells with phorbol esters stimulates the formation of senile plaques in the brains of af- release of APPs, presumably via protein kinase fected individuals. The amyloid core of these C-mediated phosphorylation of a target protein plaques is comprised largely of a small 39-43 (2-5). A similar effect can be obtained using amino acid peptide (1), termed AP3, derived from okadaic acid (an inhibitor of serine/threonine- the Alzheimer amyloid precursor protein (APP). specific protein phosphatases), possibly via in- APP can also be processed via a nonamyloido- creased phosphorylation of the same target protein due to phosphatase inhibition. In addition, Address correspondence and reprint requests to: Edgar F. protein kinase C (PKC) was shown to phosphor- da Cruz e Silva, The Rockefeller University, Box 296, 1230 York Avenue, New York, NY 10021, U.S.A. ylate the cytoplasmic domain of APP (6,7) at a

Molecular Medicine, Volume 1, Number 5, July 199 5 5 3 5 - 541 536 Molecular Medicine, Volume 1, Number 5, July 1995 site that is also phosphorylated in vivo (M. Oishi (DMEM), and fetal bovine serum were obtained and P. Greengard, unpublished observations). from Life Technologies Inc. (Gaithersburg, MD, However, APPs secretion can still be stimulated U.S.A.). Cantharidin was purchased from Sigma by phorbol esters or okadaic acid in the absence Chemical Co. (St. Louis, MO, U.S.A.). Murine of the cytoplasmic domain of APP (4,5). Thus, monoclonal antibody 22C 11, prepared against although the identity of the target protein has the amino terminus of APP (12), was purchased not been elucidated, it is unlikely to be the APP from Boehringer Mannheim Corporation (Indi- protein itself. Phorbol ester stimulation of APPs anapolis, IN, U.S.A.). Murine monoclonal anti- secretion implicates PKC in APP processing. Like- body 6E10, whose epitope is located within res- wise, the stimulation of APPs secretion by oka- idues 1-17 of AP3, was obtained from Drs. H. M. daic acid implicates a phosphatase. However, the Wisniewski and K. S. Kim (New York State In- question of which type of serine/threonine-spe- stitute for Basic Research in Developmental cific protein phosphatase is involved in APP se- Disabilities). cretion has not been addressed previously. Four major types of serine/threonine-specific protein phosphatase have been identified in Cell Culture, Drug Treatment, and eukaryotic cells (termed 1, 2A, 2B, and 2C), Detection of Secreted APP which can be distinguished by a variety of bio- COS- 1 cells were grown in DMEM supplemented chemical and enzymatic properties (reviewed in with 10% fetal bovine serum at 370C and 5% Ref. 8). Of the four types of protein phosphata- CO2. The cells were subcultured in triplicate into ses, PP1 and PP2A exhibit broad and overlapping six-well plates and grown to confluency. After substrate specificities in vitro, accounting for al- two washes with phosphate buffered saline, the most all measurable activity towards a variety of cells were incubated in the absence or presence phosphoproteins regulating cellular processes. of varying concentrations of either PDBu (for 60 PP2B (also known as calcineurin) is active on a min) or the indicated phosphatase inhibitors (for much more restricted range of substrates, and 90 min). When a combination of drugs was used, PP2C is structurally unrelated to the other types. the time of exposure was kept to 60 min. All Recent recombinant DNA approaches identified drugs were prepared by diluting the stock solu- the existence of several isoforms for each phos- tions into serum-free DMEM. Following incuba- phatase type and revealed that types 1, 2A, and tion as described above, the conditioned medium 2B belong to a gene family distinct from that of was collected into tubes containing SDS (1% phosphatase 2C. A number of "novel" phos- final concentration), and boiled for 5 min. The phatases have also been identified in a variety of samples, normalized for protein content (13), organisms, the first of which was termed protein were subjected to SDS-polyacrylamide gel elec- phosphatase X (9). trophoresis on 7.5% gels (14). The separated A variety of compounds is now available ca- proteins were then electrophoretically trans- pable of specifically inhibiting serine/threonine- ferred onto nitrocellulose and probed with anti- specific protein phosphatases. Two of the best bodies 22C 11 or 6E10, as described in the figure characterized are okadaic acid and calyculin A legends. The immunoblots were washed and (10). Since they readily enter intact cells, they then incubated successively with rabbit anti- can be used to determine which phosphatase is mouse IgG (Cappel, Durham, NC) and 125I-pro- active in a particular process by virtue of their tein A (Amersham, Arlington Heights, IL, different relative potencies against the various U.S.A.). Immunoreactive bands were detected by phosphatase types. In this study we have used autoradiography and quantitated using a Phos- them to identify the protein phosphatase in- phorlmager (Molecular Dynamics, Sunnyvale, volved in the control of APPs production. The CA, U.S.A.). data presented here were the subject of a prelim- inary report (1 1). RESULTS MATERIALS AND METHODS Several previous studies showed that APP secretion can be stimulated in a variety of cells using Materials phorbol esters. Indeed, in a previous report we Calyculin A, okadaic acid, PDBu (phorbol 12,13- demonstrated that PDBu (phorbol 12,13-dibu- dibutyrate), Dulbecco's modified Eagle's medium tyrate) could stimulate the secretion of both en- E. F. da Cruz e Silva et al.: PP1 Inhibition Stimulates APP Secretion 537

TABLE 1. Concentration of test substance PDBu C -9 -8 -7 -6 required to achieve half-maximal secretion of APPS EC50 (nM) Test Substance 22C11 6E10

Okadaic Acid 80 110 Calyculin A 1 1 Cantharidin 500 ND FIG. 1. Stimulation of APPs production from PDBu 54 ND COS-1 cells in response to PDBu treatment Cells were treated with the PDBu concentrations in- The range of concentrations used for each test substance dicated (10-9_10-6 M), and the APPs released into was as follows: 10-10_10-6 M okadaic acid, 10-1"-10-7 M 5 the medium was quantitated by immunoblotting calyculin A, 10-9_10 M cantharidin, and 10-10_10-6 M PDBu. The amount of APP secreted into the medium was with antibody 22C 11. C, control without PDBu. expressed as a percentage of the maximal amount, as ex- emplified in Fig. 2B. Immunoreactive bands were quantitated by immunoblotting using a Phosphorlmager. The drug concentration required to achieve half-maximal secretion of APP from COS-1 cells (EC50) was calculated from dogenous APP and transfected wild-type and the average results obtained in several experiments, each performed in triplicate, and are tabulated above. ND, not mutant APP expressed in COS-1 cells (4). In the determined. present study, we have further characterized the control of secretion of endogenous APP from COS- 1 cells in response to agents capable of modulating the state of phosphorylation of intracellular proteins. PDBu treatment of COS- 1 cells was shown to stimulate secretion of endogenous production of APPs from COS-1 cells in a dose- APP in a dose-dependent manner (Fig. 1). The dependent manner (Fig. 2A). Quantitation of the concentration of PDBu required for half-maxi- results from several dose-response experiments mal effect (EC50) was approximately 50 nM using antibody 22Cl yielded average EC50 val- (Table 1). We have previously reported the EC50 ues of 80 and 1 nM for okadaic acid and calyculin value for PDBu-stimulated processing of APP to A, respectively (Fig. 2B, Table 1). Since antibody be 17 nM in PC 12 cells, as measured by following 22C 11 has been reported to cross-react with the disappearance of mature intracellular APP APP-related proteins (16,17) and does not differ- using an antibody whose epitope lies in the entiate between a-secretase and f3-secretase C-terminal intracellular domain of APP ( 15). cleaved APPS, we sought to confirm and extend Thus, there is reasonable agreement between the our results using a more specific antibody. Thus, two experimental systems, even though different using antibody 6E10, which specifically recog- parameters were measured and different cell nizes APPs produced by the a-secretase pathway, types were used. we obtained EC50 values of 110 and 1 nM for APP secretion can be stimulated not only by okadaic acid and calyculin A, respectively (Fig. 3, activators of PKC, such as PDBu, but also by Table 1). Given the approximately 100-fold dif- inhibitors of serine/threonine-specific protein ference in the calculated EC50 values for okadaic phosphatases. Thus, okadaic acid was shown to acid and calyculin A, these results identify pro- stimulate APP secretion in PC12 cells (2) and in tein phosphatase 1 (PP1) as being the target of COS-1 cells (4). However, the identity of the the phosphatase inhibitors used (see Discussion phosphatase targeted by such inhibitors was not below), and therefore implicate it in the control elucidated. The availability of other naturally oc- of APP cleavage and secretion. curring compounds that are cell permeable and Cantharidin, isolated from blister beetles, is a act as specific inhibitors of certain of the serine/ recently described phosphatase inhibitor (18,19) threonine-specific protein phosphatases has al- thought to be the active ingredient in the pur- lowed us to investigate the nature of the phos- ported aphrodisiac, "Spanish fly". Cantharidin, phatase involved in the control of APP secretion. like okadaic acid and calyculin A, was also capa- Both okadaic acid and calyculin A stimulated the ble of stimulating APPs production from COS-1 538 Molecular Medicine, Volume 1, Number 5, July 1995

I OA C -9 -8 -7 -6 A] B] FIG. 2. Stimulation of APPs production from COS-1 cells, in response to 1 00 |+ OA220C11 treatment with phosphatase inhibi- A- C*22C11 1 tors, measured with antibody 22Cl 80-j (A) Immunoblot analysis of COS- 1 cells 60 A- treated with the indicated concentrations CA C -10 -9 -8 -7 401 / / of OA (10-9-10-6 M okadaic acid) or CA (10-10_10-7 M calyculin A; higher 20 j-4 jj concentrations of calyculin A were tox- ic). C, control without drug treatment. 10-12 1I0-10 110-8 10o-6 (B) Quantitation of the dose response from several experiments. The values .. plotted are the mean ± SEM.

cells in a dose-dependent manner (Fig. 4A). In result in further stimulation of APPs production this case, the calculated EC50 was approximately beyond the levels achieved with either drug 500 nM (Table 1). The higher EC50 value ob- alone (Fig. 4B and data not shown). These results tained for cantharidin in comparison with oka- are consistent with phorbol esters and phos- daic acid and calyculin A is in agreement with phatase inhibitors affecting the state of phos- their relative in vitro phosphatase inhibitory po- phorylation of the same target protein via the tencies (Ref. 19 and E. F. da Cruz e Silva et al., activation of PKC or the inhibition of PP1. manuscript in preparation). Using submaximal The evidence presented above indicating doses of PDBu (10-8 M) and cantharidin (10-7), that PP1 regulates APPs production in COS-1 an additive effect was measured (Fig. 4B); the use of maximally effective doses of both did not

1CT C -8 -7 -6 -5 1

OA C -9 -8 -7 -6

AL PDBu - -8 - -8 -6 -6 CT - - -7 -7 -5 - CA C -1O -9 -8 - 7

[1 I FIG. 4. Cantharidin stimulation of APPS pro- FIG. 3. Stimulation of APPs production from duction from COS-1 cells as measured with an- COS-1 cells, in response to treatment with tibody 22CII phosphatase inhibitors, measured with anti- (A) Immunoblot analysis of the dose response using body 6E10 the indicated concentrations of cantharidin (10-8- Cells were treated with the indicated concentrations 10-5 M). C, control without cantharidin treatment. of OA (10-9_10-6 M okadaic acid) or CA (10- 10- (B) Immunoblot analysis of APPS production in re- 10-7 M calyculin A). C, control without drug treat- sponse to the indicated concentrations of PDBu and ment. cantharidin (CT). -, absence of drug. E. F. da Cruz e Silva et al.: PP1 Inhibition Stimulates APP Secretion 539 cells led us to examine these cells for the expres- immunoblotting, immunocytochemistry, and in sion of PP1. The mammalian genome is known situ hybridization (22,23). to contain at least three different genes encoding Several neurotransmitters affect the phys- PP1 catalytic subunits, namely, PPla, PP1IB, and iological properties of neurons by regulating PPIly (20,21). PPIy is known to undergo tissue- the phosphorylation state of PPI inhibitor pro- specific alternative splicing to generate two proteins (28-30), hence affecting PP1 activity. For teins differing solely at their extreme carboxyl example, dopamine, by increasing intracellular termini. One of these, PPIyl, is ubiquitously ex- cAMP levels and activating protein kinase A, pressed in most tissues, whereas the other iso- causes the phosphorylation and activation of form, PP 1 y2, is thought to be testis specific the PPL inhibitor DARPP-32 (dopamine- and (20,22). Previously described isoform-specific cyclic AMP-regulated phosphoprotein, Mr antibodies (22,23) detected both PP1 a and PP1y 32,000 daltons) in the medium-sized spiny in COS-1 cells (data not shown). neurons of the neostriatum (28,31). Con- versely, glutamate, by increasing intracellular calcium levels and activating PP2B, causes the dephosphorylation and inactivation of DISCUSSION DARPP-32 (32). The convergence of major neurotransmitter pathways (e.g., dopaminer- The availability of several phosphatase inhibitors gic, glutamatergic, and GABAergic) on PP1 with different potencies against the various types suggests a critical role for this phosphatase in of phosphatase allowed us to investigate the mediating the actions of these neurotransmit- identity of the phosphatase involved in the con- ters. trol of APP secretion. Both okadaic acid and ca- Defects in signal transduction-dependent lyculin A are potent inhibitors of protein phos- regulation of APP cleavage may play a critical phatase types 1 and 2A, whereas PP2B is only role in the pathogenesis of AD. For instance, inhibited at much higher concentrations and deficits in both neurotransmission (33,34) and PP2C is essentially resistant to these compounds. PKC activity (35-37) have been reported to be However, while calyculin A is equally potent in associated with AD. Consistent with this no- vitro against purified PP1 and PP2A (IC50 = 1-2 tion, the stimulation of APPs release in re- nM), okadaic acid inhibits PP1 with an IC50 of sponse to neurotransmitters and other first nM 10-100 and PP2A with an IC50 of 0.5-1 nM messengers known to act via PKC has been (10,24,25). Thus, okadaic acid and calyculin A reported in several studies (e.g., Refs. 38 and can be used to characterize which type of phos- 39). The possibility now arises that regulation phatase is involved in a particular physiological of APP metabolism, APPs release process (e.g., Refs. 26 and 27). In our including and experi- formation, may also be modulated ments, calyculin A was 100-fold AP3 by neu- approximately rotransmitters or first more potent than okadaic acid at stimulating messengers whose re- APPs production (Figs. 2 and 3, Table 1). Thus, ceptors are linked to modulation of PP1 activ- our data implicate a type 1 phosphatase in the ity. The characterization of the protein kinases control of APPs production. Furthermore, we and phosphatases involved in the control of have shown by immunoblotting that COS- 1 cells APP metabolism should facilitate the develop- express at least two of the known isoforms of PP1 ment of novel rational therapeutic interven- (namely, PPla and PPIyj), which exhibit very tion strategies for AD. similar sensitivity profiles to okadaic acid, calyculin A and cantharidin (E. F. da Cruz e Silva et al., manuscript in preparation). Given the dy- namic nature of the protein phosphorylation process, and given that PP1 has been detected in ACKNOWLEDGMENTS all eukaryotic cells examined to- date (including We are grateful to Drs. J. D. Buxbaum, S. E. neurons and glia), it is likely that PP1 also regu- Gandy, and A. C. Nairn for critical reading of the lates APPs production in other cell types. PP1 has manuscript. This work was supported by U. S. been shown to be particularly enriched in brain Public Health Service Grant AG-09464. CTBVZ compared with peripheral tissues, and the occur- was supported by a postdoctoral fellowship from rence and distribution of known PP1 isoforms CNPq and by PICD from Universidade Estadual has been described in mammalian brain both by de Londrina (Brazil). 540 Molecular Medicine, Volume 1, Number 5, July 1995

REFERENCES ization of precursors of Alzheimer's disease 1. Glenner GG, Wong CW. (1984) Alzheimer's A4 amyloid protein. Cell 57: 115-126. disease: Initial report of the purification and 13. Smith PK, Krohn RI, Hermanson GT, et al. characterization of a novel cerebrovascular (1985) Measurement of protein using bicin- amyloid protein. Biochem. Biophys. Res. Com- choninic acid. Anal. Biochem. 150: 76-85. mun. 120: 885-890. 14. Laemmli UK. (1970) Cleavage of structural 2. Caporaso GL, Gandy SE, Buxbaum JD, Ra- proteins during the assembly of the head of mabhadran TV, Greengard P. (1992) Protein bacteriophage T4. Nature 227: 680-685. phosphorylation regulates secretion of Alz- 15. Buxbaum JD, Gandy SE, Cicchetti P, et al. heimer ,3/A4 amyloid precursor protein. (1990) Processing of Alzheimer ,3/A4 amy- Proc. Natl. Acad. Sci. U.S.A. 89: 3055-3059. loid precursor protein: Modulation by agents that regulate protein phosphorylation. Proc. 3. Gillespie SL, Golde TE, Younkin SG. (1992) Natl. Acad. Sci. U.S.A. 87: 6003-6006. Secretory processing of the Alzheimer amy- 16. Wasco W, Bupp K, Magendantz M, Gusella loid ,B/A4 protein precursor is increased by Biophys. JF, Tanzi RE, Solomon F. (1992) Identifica- protein phosphorylation. Biochem. tion of a mouse brain cDNA that encodes a Res. Commun. 187: 1285-1290. protein related to the Alzheimer disease-as- 4. da Cruz e Silva OAB, Iverfeldt K, Oltersdorf sociated amyloid beta protein precursor. T, et al. (1993) Regulated cleavage of Alzhei- Proc. Natl. Acad. Sci. U.S.A. 89: 10758-10762. mer f3-amyloid precursor protein in the ab- 17. Wasco W, Gurubhagavatula S, Paradis MD, sence of the cytoplasmic tail. Neuroscience 57: et al. (1993) Isolation and characterization of 873-877. APLP2 encoding a homologue of the Alzhe- 5. Hung AY, Selkoe DJ. (1994) Selective imer's associated amyloid beta protein pre- ectodomain phosphorylation and regulated cursor. Nature Genet. 5: 95-100. cleavage of f3-amyloid precursor protein. 18. Li YM, Casida JE. (1992) Cantharidin-bind- E.M.B.O. J. 13: 534-542. ing protein: Identification as protein phos- 6. Gandy SE, Czernik AJ, Greengard P. (1988) phatase 2A. Proc. Natl. Acad. Sci. U.S.A. 89: Phosphorylation of Alzheimer disease amy- 11867-11870. loid precursor protein peptide by protein ki- 19. Li YM, Mackintosh C, Casida JE. (1993) Pro- nase C and Ca2+/calmodulin-dependent tein phosphatase 2A and its [3H] canthari- protein kinase II. Proc. Natl. Acad. Sci. U.S.A. din/ [3H] endothall thioanhydride binding 85: 62 18-6221. site. Inhibitor specificity of cantharidin and 7. Suzuki T, Nairn AC, Gandy SE, Greengard P. ATP analogues. Biochem. Pharmacol. 46: (1992) Phosphorylation of Alzheimer amy- 1435-1443. loid precursor protein by protein kinase C. 20. Sasaki K, Shima H, Kitagawa Y, Irino S, Sug- Neuroscience 48: 755-761. imura T, Nagao M. (1990) Identification of 8. Shenolikar S, Nairn AC. (1991) Protein members of the protein phosphatase 1 gene phosphatases: Recent progress. Adv. Sec. Mess. family in the rat and enhanced expression of Phosphoprot. Res. 23: 1-123. protein phosphatase 1a gene in rat hepato- cellular carcinomas. Jpn. J. Cancer Res. 81: 9. da Cruz e Silva OB, da Cruz e Silva EF, 1272-1280. PTW. Identification of a novel Cohen (1988) 21. da Cruz e Silva EF, Greengard P. (1995) protein phosphatase catalytic subunit by of neuronal isoforms of F.E.B.S. Lett. 242: 106-110. Cloning protein cDNA cloning. phosphatase 1 by low-stringency screening 10. Ishihara H, Martin BL, Brautigan DL, et al. of cDNA libraries. Neuroprotocols 6: 2-10. (1989) Calyculin A and okadaic acid: Inhib- 22. da Cruz e Silva EF, Fox CA, Ouimet CC, itors of protein phosphatase activity. Bio- Gustafson E, Watson SJ, Greengard P. chem. Biophys. Res. Commun. 159: 871-877. (1995) Differential expression of protein 11. da Cruz e Silva OAB, da Cruz e Silva EF, phosphatase 1 isoforms in mammalian Greengard P. (1994) Role of protein phos- brain. J. Neuroscience 15: 3375-3389. phatase 1 in the metabolism of Alzheimer 23. Ouimet CC, da Cruz e Silva EF, Greengard P. amyloid precursor protein. Soc. Neurosci. Abs. (1995) The a and yl isoforms of protein 20: 440. phosphatase 1 are highly and specifically 12. Weidemann A, Konig G, Bunke D, et al. concentrated in dendritic spines. Proc. Natl. (1989) Identification, biogenesis and local- Acad. Sci. U.S.A. 92: 3396-3400. E. F. da Cruz e Silva et al.: PP1 Inhibition Stimulates APP Secretion 541

24. Hescheler J, Mieskes G, Ruegg JC, Takai A, 32. Halpain S, Girault J-A, Greengard P. (1990) Trautwein W. (1988) Effects of a protein Activation of NMDA receptors induces de- phosphatase inhibitor, okadaic acid, on phosphorylation of DARPP-32 in rat striatal membrane currents of isolated guinea-pig slices. Nature 343: 369-372. cardiac myocytes. Pfluegers Arch. 412: 248- 33. Davies P, Maloney AJF. (1976) Selective loss 252. of central cholinergic neurons in Alzhei- 25. Bialojan C, Takai A. (1988) Inhibitory effect mer's disease. Lancet 2: 1403. of a marine-sponge toxin, okadaic acid, on protein phosphatases. Specificity and kinet- 34. Tomlinson BE, Corsellis JAN. (1984) Ageing ics. Biochem J. 256: 283-290. and the dementias. In: Adams JH, Corsellis 26. Starke LC, Jennings ML. (1993) K-Cl co- JAN, Duchen LW (eds). Greenfield's Neuropa- transport in rabbit red cells: Further evi- thology. 4th Ed. John Wiley & Sons, New dence for regulation by protein phosphatase York, pp. 951-1025. type 1. Am. J. Physiol. 264: C118-C124. 35. Cole G, Dobkins KR, Hansen LA, Terry RD, 27. Murata K, Sakon M, Kambayashi J, et al. Saitoh T. (1988) Decreased levels of protein (1993) The possible involvement of protein kinase C in Alzheimer brain. Brain Res. 452: phosphatase 1 in thrombin-induced Ca21 in- 165-174. flux of human platelets. J. Cell. Biochem. 51: 36. Van Huynh T, Cole G, Katzman R, Huang 442-445. K-P, Saitoh T. (1989) Reduced protein ki- 28. Walaas SI, Aswad DW, Greengard P. (1983) nase C immunoreactivity and altered protein A dopamine- and cyclic AMP-regulated phosphorylation in Alzheimer's disease fi- phosphoprotein enriched in dopamine-in- broblasts. Arch. Neurol. 46: 1195-1199. nervated brain regions. Nature 301: 69-71. 37. Masliah E, Cole GM, Hansen LA, et al. 29. Snyder GL, Girault J-A, Chen JY, et al. (1991) Protein kinase C alteration is an early (1992) Phosphorylation of DARPP-32 and biochemical marker in Alzheimer's disease. protein phosphatase inhibitor-I in rat cho- J. Neurosci. 11: 2759-2767. roid plexus: Regulation by factors other than dopamine. J. Neurosci. 12: 3071-3083. 38. Buxbaum JD, Oishi M, Chen HI, et al. 30. Snyder G, Fisone G, Greengard P. (1994) (1992) Cholinergic agonists and interleukin Phosphorylation of DARPP-32 is regulated 1 regulate processing and secretion of the by GABA in rat striatum and substantia Alzheimer ,/3A4 amyloid precursor protein. nigra. J. Neurochem. 63: 1766-1771. Proc. Natl. Acad. Sci. U.S.A. 89: 10075-10078. 31. Hemmings Jr HC, Greengard P, Tung HYL, 39. Nitsch RM, Slack BE, Wurtman RJ, Grow- Cohen P. (1984) DARPP-32, a dopamine- don JH. (1992) Release of Alzheimer amy- regulated neuronal phosphoprotein, is a po- loid precursor derivatives stimulated by acti- tent inhibitor of protein phosphatase-1. Na- vation of muscarinic acetylcholine receptors. ture 310: 503-505. Science 258: 304-307.

Contributed by P. Greengard on April 21, 1995.