Regulatory subunits of cyclic AMP-dependent protein kinase: presence in

granules and secretion by exocrine and endocrine cells

A. R. HAND*

National Institute of Dental Research, National Institutes of Health, Building 10, Room 1AI3, Bethesda, Maryland 20892, USA

and M. I. MEDNIEKS

Department of Pediatrics, University of Chicago, Wyler Childrens Hospital, Chicago, Illinois 60637, USA

•Author for correspondence

Summary

Cyclic AMP-dependent protein kinase (cAPK) is theintracellular mediator of signal transduction eventsinvolving the adenylate cyclase-cyclic AMP sys-tem. A monoclonal antibody (MAb BB1) to the typeII regulatory subunit (RII) of cAPK was used in apost-embedding immunogold-labeling procedureto determine the ultrastructural localization of RIIin several different secretory cells of the rat. Labelwas present in nuclei, especially over the hetero-chromatin, and in the cytoplasm, particularly inareas containing rough endoplasmic reticulum.Immunolabeled RII -was also present in secretorygranules of the parotid gland, exocrine and endo-crine pancreas, seminal vesicle, anterior and inter-mediate pituitary, and intestinal endocrine cells.Photoaffinity labeling of parotid saliva, pancreaticand seminal fluids with the cyclic AMP analogue,32P-labeled-8-azido-cyclic AMP, revealed the pres-

ence of cyclic AMP-binding proteins •with electro-phoretic mobilities similar to those of authenticcAPK regulatory subunits. These results confirmour previous observations on the localization ofcAPK regulatory subunits in the rat parotid usingpolyclonal antibodies, and extend them to a num-ber of other exocrine and endocrine cells. Theapparent widespread occurrence of cAPK subunitsin secretory granules and secretory fluids suggeststhat cAPK may be involved in specific intragranularregulatory and/or phosphorylation events, or that ithas an unidentified extracellular function.

Key words: cyclic AMP-dependent protein kinase,immunocytochemistry, protein A-gold, rat, exocrine andendocrine cells, secretion.

Introduction

Cyclic AMP-dependent protein kinase (cAPK) partici-pates in the regulation of various cellular activities (seereviews by Flockhart & Corbin, 1982; Lohman & Walter,1984). These include both nuclear and cytoplasmicevents; among the latter, in many glandular tissues, is theprocess of exocytosis. The precise mechanisms involvedin determining the specificity of cAPK actions are notwell known. Intracellular redistribution of the holo-enzyme or of its subunits may be critically important incyclic AMP-mediated responses of cells to hormonalstimulation. Immunocytochemical experiments and sub-cellular fractionation have shown that the catalytic (C)and regulatory (R) subunits are present in the nuclei andcytoplasm of many cell types (for references see Kuettelet al. 1985; Mednieks et al. 1987), and occasionally areassociated with other organelles such as the Golgi appar-

Journal of Cell Science 93, 675-681 (1989)Printed in Great Britain © The Company of Biologists Limited 1989

atus (Nigg et al. 1985; De Camilli et al. 1986), cytoskel-etal elements (Vallee et al. 1981; Browne et al. 1982; DeCamilli et al. 1986) and the acrosome (Pariset et al.1986).

The results of our previous immunocytochemicalstudies of the distribution of cAPK R subunits in ratparotid acinar cells showed that in addition to theirexpected localization in the nucleus and in the cytoplasm,both the type I (RI) and type II (RII) regulatorysubunits were present in the contents of the secretorygranules (Mednieks et al. 1987). Earlier, we demon-strated that cyclic AMP-binding proteins secreted inresponse to /J-adrenergic stimulation of the parotid glandcan be detected in ductal saliva by photoaffinity labelingwith 32P-labeled-8-azido-cyclic AMP (Mednieks & Hand,1984). In this report, we demonstrate that immuno-reactive cAPK R subunits are present in the secretorygranules of several other rat exocrine and endocrine cells,

675

and that the secretory fluids of the exocrine pancreas andthe seminal vesicle also contain cyclic AMP-bindingproteins.

Materials and methods

Immunological reagentsA mouse monoclonal antibody (MAb) to RII from the ratparotid gland was prepared as described (Siraganian et al. 1983;Mednieks et al. 1989). The 40-70K (K = \&MT) chromato-graphic peak of a secretory granule preparation was used as theoriginal antigen and the hybrid cells were screened against typeII cAPK from bovine heart (Sigma Chemical Co., St Louis,MO). The IgM antibodies secreted by the clone BBl IIB1(MAb BBl) were inhibited in competitive enzyme-linkedimmunosorbent assays (ELISA) by preincubation with either

the original antigen or with bovine heart cAPK. MAb BBlimmunoprecipitated labeled proteins with the same electro-phoretic mobility as RII from I-labeled total soluble proteinsof rat and liver homogenates (Hand & Jungmann, 1989), as wellas RII from type II cAPK (bovine heart; Sigma) and from ratand human parotid saliva that had been photoaffinity labeledwith 32P-labeled-8-azido-cyclic AMP (Mednieks et al. 1989).This MAb reacted with protein bands of 52 and 54K (electro-phoretically distinct RII isoforms) on immunoblots of parotidgland extracts and parotid saliva (Mednieks et al. 1989).Antigen-absorbed antibody was prepared by incubating thereconstituted ascites fluid (diluted 1: 30) overnight at 4°C with acalculated tenfold excess of bovine heart cAPK or with ratparotid saliva, then diluting to the final concentration beforeimmunostaining. Unlabeled secondary antibody (affinity-puri-fied goat anti-mouse IgM) was obtained from Cappel (WestChester, PA), and 5 and 15 nm protein A-gold solutions were

m

. ksg

K

• • * • -

> . .

^ 1

. ' • ' • • ^

' .it

^-aa^

^

w1"

4;-

*

Fig. 1. Inimunogold labeling of thin sections of exocrine cells with MAb BBl. A. Nucleus (n) and secretory granules (sg) ofparotid acinar cell labeled with MAb BBl. Few gold particles are seen over euchromatin; mitochondria (in) are unlabeled. Theelectron-lucent appearance of the granules is due to omission of OsC>4 post-fixation. X27000. B. Zymogen granules of pancreaticacinar cell labeled with MAb BBl. X22000. C. Apical portion of seminal vesicle cell labeled with MAb BBl. Dense cores ofsecretory granules and luminal content (/) are heavily labeled. X390O0. Bars, 0-5 /*m.

676 A. R. Hand and M. I. Mednieks

purchased from Janssen Life Sciences Products (Beerse, Bel-gium).

Tissue preparation and immunogold labelingTissues were obtained from adult male Sprague-Dawley ratsafter fixation of anesthetized (chloral hydrate, 400mgkg~ ,i.p.) animals by vascular perfusion with 1% phosphate-buffered glutaraldehyde. Parotid glands, pancreas, seminalvesicle, duodenum and pituitary glands were fixed for 1 h,rinsed in 0-1 M-phosphate buffer, pH7-4, containing 5 % su-crose, dehydrated in ethanol, substituted with acetone orpropylene oxide, and embedded in Epon. Thin sections werecollected on nickel grids and immunostained by a modifiedprotein A-gold method as described (Mednieks et al. 1989).Briefly, the sections were etched with sodium metaperiodate,treated with 1 % bovine serum albumin (BSA) in Tris-bufferedsaline, incubated for 16-24 h at 4°C with reconstituted MAbBB1 ascites diluted in 0-02M-Tris-HCl, p H 7 4 , containing0-5 M-NaCl and 0-1 % Tween 20, incubated for 1 h at 25°C withunlabeled, affinity-purified goat anti-mouse IgM, and finallyincubated for 30min at 25°C with protein A-gold. The sectionswere rinsed with Tris-buffered saline after each incubation, andwith distilled water prior to staining with uranyl acetate andlead citrate. The sections were observed and photographed in aZeiss EM 10A transmission electron microscope. Quantitativeanalysis of relative labeling intensities was performed as de-scribed (Mednieks et al. 1987, 1989) on micrographs of cellslabeled with 15 nm gold particles using a ZIDAS digitizingsystem (Carl Zeiss, Inc., Thornwood, NY) interfaced with amicrocomputer. Statistical analyses (<-test) were performedusing 'Blossom', a program developed by the Division ofComputer Research and Technology, NIH, for use with Lotus1-2-3.

Results

Incubation of thin sections of parotid acinar cells withMAb BB1, followed by goat anti-mouse IgM and proteinA-gold (Fig. 1A), showed labeling principally of nuclearheterochromatin, regions of cytoplasm containing roughendoplasmic reticulum (RER), and the contents of se-cretory granules. The labeling pattern of parotid glandcells with MAb BB1 has been described in detail (Med-nieks & Hand, 1987; Mednieks et al. 1989), and anexample is included here for comparison with the otherexocrine and endocrine tissues.

In exocrine pancreatic acinar cells (Fig. IB) the distri-bution of gold particles was similar to that found forparotid acinar cells. Label was present over nuclei, thebasal cytoplasm and zymogen granules, although thelabeling intensities of these compartments were less thanin the parotid gland (Fig. 2A,B). Epithelial cells of theseminal vesicle contain large granules in their apicalcytoplasm characterized by a dense core surrounded byan electron-lucent matrix (Fig. 1C). The dense cores ofthe granules were strongly labeled by MAb BB1(Fig. 2C), as was the dense luminal content, whichconsists of released secretory product. Nuclear andcytoplasmic labeling also occurred in the seminal vesiclecells, but these compartments were less intensely labeledthan in the parotid gland and pancreas.

The secretory granules of somatotrophs (Figs 2D and3A) and other cells of the anterior pituitary gland

Collection of secretory fluids and tissuesRats were anesthetized with chloral hydrate (400mgkg~1, i.p.)and tracheotomized. Parotid secretion was stimulated by iso-proterenol (30mgkg~ , i.p.) and the saliva was collected byextraoral cannulation of the main excretory ducts with blunted30 gauge needles attached to polyethylene tubing (PE-10).Pancreatic fluid was collected after pilocarpine stimulation(5mgkg~', s.c.) by cannulation of the bile duct near itsjunction with the duodenum and ligation of the duct proximalto the entrance of the pancreatic ducts. Seminal vesicle contentwas collected by removing the seminal vesicles and flushing thelumen with phosphate-buffered saline. All secretory fluids werecollected in tubes containing protease inhibitors (10/iM-benza-midine and phenylmethylsulfonyl fluoride) and kept on ice.

Parotid glands were excised from rats killed by asphyxiationwith CO2 and exsanguination. The glands were trimmed free offat, connective tissue and lymph nodes, minced with scissorsand homogenized with a Dounce homogenizer in 0-32M-sucrose, 0-05 M-Tris-HCl, p H 7 5 , l/^M-MgCl^, containingprotease inhibitors, then centrifuged at 600^ for lOmin.

Photoaffinity labeling and electrophoresisCyclic AMP-binding proteins in the secretory fluids and parotid600 £ supernatant fraction were photoaffinity labeled with thecyclic AMP analogue, 3ZP-labeled-8-azido-cyclic AMP (ICN,Irvine, CA), as described (Mednieks & Hand, 1984; Medniekset al. 1987). The labeled proteins were separated by sodiumdodecyl sulfate-polyacrylamide gel electrophoresis (Laemmli,1970) and transferred to nitrocellulose membranes (Towbin etal. 1979). The membranes were stained with Coomassie Blue,dried and exposed to X-Omat film (Kodak, Rochester, NY) toidentify radioactive bands by autoradiography.

50

40

•3 30

a- 20

O 10

0

20

15

•a 10

a1 5o

INuclei Cytoplasm Granules

B

Nudei Cytoplasm Granules

50

40

30

20-

10-

n

C

(77)

1—'

1

(8)-

120-

15-

10

5-

0-

Nudei Cytoplasm Granulei Lumen

—. D ^S 3;

8

IifPituitary Pancresi Pancreas

somatotToph A cell* B cellj

Fig. 2. Labeling intensities of rat secretory cell organelleswith MAb BB1. A. Parotid acinar cells; B, pancreatic acinarcells; C, seminal vesicle cells; D, granules of endocrine cells.Filled bars, MAb BB1; open bars, preabsorbed with type IIcAPK. Error bars indicate S.E.M; number of organelles orareas counted are given in parentheses. Background overempty plastic ranged from 027 to 0'76 particles /im~z.

Secretion of cAPK regulatory subunits 611

s*.!*

•

• ^

* •

. . , r , . . ^

*

^ - ^

V , - ' : '''' •'•% • • - , : . • , ,

r?

Fig. 3. Immunogold labeling of thin sections of endocrine cells with MAb BBl. A. Somatotroph of anterior pituitary labeledwith MAb BBl. Gold particles are present over secretory granules and nucleus («). X26000. B. Pituitary intermediate lobe celllabeled with MAb BBl. Nucleus (n) and many secretory granules are labeled. X27 0O0. C,D. Cells of endocrine pancreaslabeled with MAb BBl. C. Secretory granules of an A cell are labeled with gold; a few particles are present over the nucleus(«). X29000. D. B cell granules are labeled with 5 nm gold particles. X75 0O0. E. Intestinal endocrine cell labeled with MAbBBl. Gold particles are present over nucleus (n) and secretory granules. X38 0O0. Bars: A-C,E, 0-5//m; D, 0-1 fun.

678 A. R. Hand and M. I. Mednieks

contained RII, as demonstrated by labeling with MAbBBl. Similarly, the granules of cells of the intermediatelobe of the pituitary were also labeled (Fig. 3B). In theislets of Langerhans in the pancreas, the granules of Acells and B cells were labeled with gold particles(Fig. 3C,D). The labeling intensity of the A cell granuleswas considerably higher than that of the B cell granules(Fig. 2D). Finally, the granules of intestinal endocrinecells also showed reactivity with MAb BBl (Fig. 3E).

To test the specificity of the immunolabeling, preab-sorption of MAb BBl was carried out using either bovineheart type II cAPK or rat parotid saliva. On ELISA,with antibody dilutions of ^1:1000, complete inhibitionwas obtained after preabsorption with a calculated ten-fold excess of bovine heart type II cAPK (data notshown). When thin sections were incubated with preab-sorbed antibody, the labeling of the nuclei, cytoplasmand granules of the various cell types was markedlyreduced (Fig. 2). The percentage reduction varied, how-ever, among the different cell types and subcellularcompartments. The labeling intensity of the nuclei ofparotid and pancreatic acinar cells and pituitary somato-trophs, for example, was reduced by 84-85%, whereasthat of seminal vesicle cell nuclei was reduced by only62 %. The reduction of secretory granule labeling showedeven greater variability, ranging from 84% in parotidacinar cells to 35 % in A cells of pancreatic islets. Whenthin sections of parotid acinar cells were incubated withMAb BBl that had been preabsorbed with rat parotidsaliva, labeling of secretory granules was reduced by over92%. The reduction of label over the nuclei, however,was only 68 %. The variability of these results may reflectdifferent affinities of antibody binding to the antigenfrom the two different sources, as well as the presence ofvarious concentrations of RII isoforms in the different

R 1 2 3 4 5 6

I —

fr-

subcellular compartments of the tissues. Labeling wasabolished when either the MAb or the secondary anti-body (anti-mouse IgM) was omitted from the stainingsequence.

Fig. 4 shows Coomassie Blue-stained protein bandingpatterns and autoradiograms of rat parotid gland tissueextract, rat parotid saliva, pancreatic fluid and seminalvesicle content photoaffinity-labeled with the cyclic AMPanalogue, 32P-labeled-8-azido-cyclic AMP. Radioactivelylabeled bands with the mobilities of RI and RII werepresent in all of the samples. Generally, the bandcorresponding to RII was strongly labeled, whereas thatcorresponding to RI was labeled less intensely. Labeledbands of higher mobility were also present, particularlyin seminal fluid obtained from the lumen of the seminalvesicle. Labeled bands of similar mobility have beenobserved in other studies (Mednieks & Hand, 1984;Mednieks et al. 1987, 1989) and are probably due toproteolysis of the R subunits. Specificity of the photoaffi-nity labeling was shown by including excess unlabeledcyclic AMP in the reaction mixture, which markedlyreduced labeling of the R subunits and the faster movingcomponents with the analogue.

Discussion

The cytoplasmic and nuclear distribution of R subunitsseen in the various cell types is consistent with the knownand suspected roles of cAPK in regulating a variety of cellprocesses (Flockhart & Corbin, 1982; Lohmann &Walter, 1984). The most striking observation is that Rsubunits also appear to be packaged into secretorygranules by a number of different cells, and at least in theparotid, pancreas and seminal vesicle, are released to the

10 11

- +12

m92

[67

[43

[30

Fig. 4. Photoaffinity labeling of exocrine tissues and secretory fluids. Lanes 1-3, parotid 600 £ supernatant; lanes 4—6, parotidsaliva; lanes 7-9, pancreatic fluid; lanes 10-12, seminal vesicle contents. Lanes 1,4,7,10, autoradiograms of samplesphotoaffinity-labeled with 32P-labeled-8-azido-cyclic AMP; lanes 2,5,8,11, photoaffinity labeling in the presence ( + ) of excessunlabeled cyclic AMP; lanes 3,6,9,12, corresponding Coomassie Blue-stained protein banding patterns. Relative mobilities(Mr X 10~3) of standard proteins indicated on the right and of RII, RI, and proteolytic fragments (fr) shown on the left.

Secretion of cAPK regulatory subunits 679

extracellular environment. The presence of R subunits inparotid acinar cell granules has also been shown byJoachim & Schwoch (1988). A similar localization of RIIapparently occurs in the acrosome of ram spermatozoa(Pariset et al. 1986). The acrosome is produced by theGolgi apparatus and is analogous to a secretory granule(Bennett, 1984).

The mechanism by which the R subunits are incorpor-ated into the granules is not known. Are they synthesizedand transported in a manner similar to other secretoryproteins, or do they gain access to the granules post-translationally? Significant labeling of the RER cisternalspace and the Golgi apparatus was observed in thinsections of parotid acinar cells incubated with polyclonalantibodies to RI and RII (Mednieks et al. 1987).Localization of RII to the Golgi apparatus in culturedfibroblasts and epithelial cells (Nigg et al. 1985) andneurons and other brain cells (De Camilli et al. 1986) hasbeen described. Alternatively, a possible mechanism fordirect translocation into granules may involve organelle-specific amino acid targeting sequences (Verner &Schatz, 1988).

The function of the R subunits in the granules and/orthe secretory fluids is unknown. One possibility is that,either through phosphorylation by cAPK C subunits, ifthey are also present in the granules, or by directinteraction with granule proteins, the R subunits mayregulate the activity of other granule components. Forexample, Khatraef al. (1985) reported that in addition totheir established function of binding to C subunits,physiological concentrations of purified RII inhibit phos-phoprotein phosphatase activity. Recently, atrial natriur-etic peptides (Rittenhouse et al. 1986) and mast cellgranule components (Kurosawa & Parker, 1987) havebeen shown to be substrates for cAPK. In the formercase, the biological activity of the phosphorylated atrialpeptides was significantly enhanced. Also, a proenkepha-lin-processing enzyme in synaptosomes derived from therat striatum appears to be activated by cAPK (Ueda et al.1987). Rat parotid secretory granules contain a similarcarboxypeptidase B-like activity as well as other post-translational processing components (von Zastrow et al.1986). Although the natural substrates for these enzymesare at present unknown, they may constitute a system,regulated by cAPK, for intragranular processing ofexocrine secretory products. It should be noted, how-ever, that we were unable to detect phosphorylation byendogenous C subunits in parotid saliva (Mednieks &Hand, 1984), and Joachim & Schwoch (1988), alsoemploying immunogold labeling, found no evidence forthe presence of C subunits in parotid granules. Finally,due to their cyclic AMP binding properties, the Rsubunits may function to transport cyclic AMP to theextracellular fluid, or they may interact directly with cellsor extracellular molecules after their release from thesecretory granules.

The technical assistance of Mrs Betty Ho is gratefullyacknowledged. We thank Ms Olevia Ambrose for help inpreparing the photographs.

References

BENNETT, G. (1984). Role of the Golgi complex in the secretoryprocess. In Cell Biology of the Secretory Process (ed. M. Cantin),pp. 102-147. Basel: Karger.

BROWNE, C. L., LOCKWOOD, A. H. & STEINER, A. (1982).

Localization of the regulatory aubunit of type II cyclic AMP-dependent protein kinase on the cytoplasmic nucrotubule networkof cultured cells. Cell Biol. Int. Rep. 6, 19-28.

D E CAMILLI, P., MORETTI, M., DONINI, S. D., WALTER, U. &

LOHMANN, S. M. (1986). Heterogeneous distribution of the cAMPreceptor protein RII in the nervous system: evidence for itsintracellular accumulation on microtubules, microtubule-organizingcenters, and in the area of the Golgi complex. J. Cell Biol. 103,189-203.

FLOCKHART, D. A. & CORBIN, J. D. (1982). Regulatory mechanismsin the control of protein kinases. CRC Cnt. Rev. Biochem. 12,133-186.

HAND, A. R. & JUNGMANN, R. A. (1989). Localization of cellularregulator)' proteins using postembedding immunogold labeling.Am. J. Aitat (in press).

JOACHIM, S. & SCHWOCH, G. (1988). Immunoelectron microscopiclocalization of catalytic and regulatory subunits of cAMP-dependent protein kinases in the parotid gland. Eur.J. Cell Biol.46, 491-498.

KHATRA, B. S., PRINTZ, R., COBB, C. E. & CORBIN, J. D. (1985).

Regulatory subunit of cAMP-dependent protein kinase inhibitsphosphoprotein phosphatase. Biochem. biophvs. Res. Commun. 130,567-573.

KUETTEL, M. R., SQUINTO, S. P., KWAST-WELFELD, J., SCHWOCH,

G., SCHWEPPE, J. S. & JUNGMANN, R. A. (1985). Localization ofnuclear subunits of cyclic AMP-dependent protein kinase by theimmunocolloidal gold method. 7. Cell Biol. 101, 965-975.

KUROSAWA, M. & PARKER, C. W. (1987). Protein phosphorylation inrat mast cell granules. Cyclic AMP dependent phosphorylation of a44K protein associated with broken granules. Biochem. Phamtac.36, 131-140

LAEMMU, U. K. (1970). Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature, Loud. TIT,680-685.

LOHMANN, S. M. & WALTER, U. (1984). Regulation of the cellularand subcellular concentrations and distribution of cyclicnucleotide-dependent protein kinases. Adv. Cyclic Nucl. Prot.Phosphorylation Res. 18, 63-117.

MEDNIEKS, M. 1. & HAND, A. R. (1984). Cyclic AMP-bindingproteins in saliva. Expenentia 40, 945-947.

MEDNIEKS, M. I. & HAND, A. R. (1987). Cyclic AMP-dependentprotein kinase compartmentalization in stimulated exocrine cells.Ann. N.Y. Acad. Sci. 494, 135-137.

MEDNIEKS, M. I., JUNGMANN, R. A., FISCHLER, C. & HAND, A. R.

(1989). Immunogold localization of the type II regulatory subunitof cyclic AMP-dependent protein kinase. Monoclonal antibodycharacterization and RII distribution in rat parotid acinar cells. J.Histochem. Cytochem. 37, 339-346.

MEDNIEKS, M. 1., JUNGMANN, R. A. & HAND, A. R. (1987).Ultrastructural localization of cyclic AMP-dependent protein kinaseregulatory subunits in rat parotid acinar cells. Eur. J. Cell Biol 44,308-317.

NIGG, E. A., SCHAFER, G., HILZ, H. & EPPENBERGER, H. M. (1985).

Cyclic AMP-dependent protein kinase type II is associated withthe Golgi complex and with centrosomes. Cell 41, 1039-1051.

PARISET, C. C , FEINBERG, J. M. F., DACHEUX, J. L. & WEINMAN,

S. J. (1986). Protein kinase A subunits localization by immunoelectron microscopy during epididymal maturation of ramspermatozoa. J . Cell Biol. 103, 242a.

RITTENHOUSE, J., MOBERLY, L., O'DONNELL, M. E., OWEN, N. E. &

MARCUS, F. (1986). Phosphorylation of atrial natriuretic peptidesby cyclic AMP-dependent protein kinase. J. biol. Chem. 261,7607-7610.

SIRAGANIAN, R. P., Fox, P. C. & BERENSTEIN, E. H. (1983).

Methods of enhancing the frequency of antigen-specifichybridoma8. Meth. Enzym. 92, 17-26.

TOWBIN, H., STAEHEUN, T. & GORDON, J. (1979). Electrophoretic

transfer of proteins from polyacrylamide gels to nitrocellulose

680 A. R. Hand and M. I. Mednieks

sheets: Procedure and some applications. Proc. natn. Acad. Sci. VERNER, K. & SCHATZ, G. (1988). Protein translocation acrossU.SA. 76, 43S0-4354. membranes. Science 241, 1307-1313.

UEDA, H., FUKUSHIMA, N. & TAKACI, H. (1987). A novel VON ZASTROW, M., TRnroN, T. R. & CASTLE, J. D. (1986).proenkephalin processing carboxypeptidaae and its activation by Exocrine secretion granules contain peptide amidation activity.cyclic AMP dependent protein kinase. Biochem. biophys. Res. Proc. natn. Acad. Sci. U.SA. 83, 3297-3301.Commun. 142, 595-602.

VALLEE, R. B., DIBARTOLOMEIS, M. J. & THEURKAUF, W. E. (1981).

A protein kinase bound to the projection portion of MAP 2(microtubule-asnociated protein 2). J. Cell Biol. 90, 568-576. (Received 30 January 1989 - Accepted, in revised form, 2 May 1989)

Secretion of cAPK regulatory subunits 681