Vel. 269, No. 35, Issue of September 2, pp. 21929-21932, 1994 THE JOURNAL OF BIOLOGICAL CHEMISTRY

Printed in U.S.A.

Functional Identification of a Vesicular Acetylcholine Transporter and Its Expression from a “Cholinergic” Gene Locus*

(Received for publication, June 9, 1994, and in revised form, July 14, 1994)

Jeffrey D. EricksonSl, HBlhne Varoquilll, Martin K.-H. Schafer**, William Modi$*, Marie-Franqoise Dieblerll, Eberhard Weihe**, James Rand§§, Lee E. EidenSQ, Tom I. Bonnerllli, and Ted B. Usdinm From the $Section on Molecular Neuroscience, llflLaboratory of Cell Biology, National Institute of Mental Health, Bethesda, Maryland 20892, the IDdpartement de Neurochimie, Laboratoire de Neurobiologie Cellulaire, CNRS, 91190 Gif-sur-Yvette Cedex, France, the **Department of Anatomy, Johannes-Gutenberg University, Mainz, Federal Republic of Germany, the $$Frederick Cancer Research and Development Facility, NCI, National Institutes of Health, Frederick, Maryland 20797, and the §§Program in Molecular and Cell Biology, Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma 73104

The vesicular acetylcholine transporter (VAChT) has been identified and characterized based on the acquisi- tion of high affinity vesamicol binding and proton-de- pendent, vesamicol-sensitive acetylcholine accumula- tion by a fibroblast cell line transfected with a clone from a rat pheochromocytoma cDNA library encoding this protein. The distribution of VAChT mRNA coincides with that reported for choline acetyltransferase (ChAT), the enzyme required for acetylcholine biosynthesis, in the peripheral and central cholinergic nervous systems. A human VAChT cDNA was used to localize the VAChT gene to chromosome 10q11.2, which is also the location of the ChAT gene. The entire sequence of the human VAChT cDNAis contained uninterrupted within the first intron of the ChAT gene locus. Transcription of VAChT and ChAT mRNA from the same or contiguous promot- ers within a single regulatory locus provides a previ- ously undescribed genetic mechanism for coordinate regulation of two proteins whose expression is required to establish a mammalian neuronal phenotype.

Cholinergic neurotransmission requires uptake of extracel- lular choline, biosynthesis of acetylcholine (ACh)’ from choline

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

to the GenBankTMIEMBL Data Bank with accession number(s) U09210, The nucleotide sequence(s) reported in this paper has been submitted

U09211, and U10554. Q To whom correspondence should be addressed. I( Guest scientist, Laboratory of Cell Biology, NIMH. ‘The abbreviations used are: ACh, acetylcholine; CUT, choline

acetyltransferase; VAChT, vesicular acetylcholine transporter; VMAT, vesicular monoamine transporter; kb, kilobase(s).

and acetyl coenzyme A, accumulation of ACh into synaptic vesicles driven by proton antiport, and quantal release of ACh from synaptic vesicles triggered by electrical depolarization of the cholinergic neuron (1). The protein machinery for genera- ting transmembrane ion gradients important in neurotrans- mitter uptake and storage and that for vesicular exocytosis appear to be common to all synaptic vesicle-containing neurons regardless of their neurotransmitter content (2, 3). Thus, the cholinergic neuronal phenotype is determined by the expres- sion of choline acetyltransferase (ChAT), an ACh-specific vesic- ular transporter, and possibly also by a plasma membrane so- dium-dependent choline transporter. Mammalian ChAT and its gene locus have been studied extensively (4-8). The identity, structure, and genetic regulation of the choline plasma mem- brane and ACh vesicular transporters are unknown. The Cae- norhabditis elegans UNC-17 gene product has been tentatively identified as a vesicular ACh transporter, but direct demonstra- tion of ACh-specific transport function is lacking (9). Recently we have identified a homolog of UNC-17 expressed i n the elec- tric organ of the marine ray Torpedo that binds vesamicol(101, a neuromuscular blocking agent that disrupts cholinergic neuro- transmission by inhibiting vesicular accumulation of ACh (11).

Here we identify a rat protein homologous to UNC-17 and the Torpedo vesamicol-binding protein that is a functional ac- etylcholine transporter, based on reconstitution of ACh trans- port in vitro. The mRNA encoding this protein appears to be expressed exclusively within cholinergic neurons in the central and peripheral nervous systems and is transcribed from the same genetic locus as the ACh biosynthetic enzyme ChAT.

EXPERIMENTAL PROCEDURES Preparation of cDNA Libraries and Cloning of cDNAs-Oligo(dT)-

primed, size-selected cDNA libraries from rat PC12 cells and human SK-N-SH cells were prepared in the plasmid expression vector CDM7I amp (12) and screened with a PCR-amplified and random-primed 32P- labeled Torpedo ocellata coding sequence probe (bases 154-925) at low stringency as described previously (10). All DNA sequencing reported here, including that of the human VAChT gene locus, was performed on each strand of the DNA, following subcloning of overlapping fragments into pUC18 as described (10).

Uptake of ACh-CV-1 fibroblasts were plated in choline-free Dulbec- CO’S modified Eagle’s medium. Functional expression of cDNAs was performed using the vaccinia virushacteriophage T7 hybrid system (10, 13). Cells were collected in phosphate-buffered saline containing the esterase inhibitor ecothiopate (5 x M) and incubated (30 “C) with shaking for various lengths of time in medium containing 100 mM NaCI, 50 mM glucose, 50 mM NaHPO,, 5 mM KCI, 5 x M ecothiopate, M paraoxon (pH 7.4) with [3HlACh (400 PM of [3H]acetylcholine iodide (~cetyl-~H-labeled), 90 mCi/mmol, DuPont NEN). Uptake was termi- nated by filtration on GFK filters followed by a 5-ml wash with medium containing 500 mM KCl, 10 mM Tris, 2 mM ACh (pH 7.4). Cells were pretreated with transporter-specific drugs (45 min), H+-ATPase inhibi- tors, H+-ionophore, or hemicholinium-3 (5 min). [3HlVesamicol binding was performed as described (10).

In Situ Hybridization Histochemistry-Rat brain frozen sections were hybridized (14) to rat VAChT cRNA probes containing predomi- nantly 3”untranslated sequences (bases 1790-2881). Sections were washed at 0.2 x SSC at 60 “C and exposed to emulsion for 14 days (14).

Northern Blot Hybridization-Poly(A+) RNA was purified by the gua- nidinium isothiocyanateicesium trifluoroacetic acid method followed by a single selection with oligo(dT)-cellulose (15). Northern blots were washed at 55 “C in 0.2 x SSC, 0.5% SDS followed by autoradiography for 12 h at -70 “C with an enhancing screen.

Preparation ofHuman Cosmid VAChT Gene Clone-Ahuman cosmid library in pWE.15 (Stratagene) was screened using polymerase chain reaction-amplified human VAChT coding sequences (bases 417-2114 of

21929

21930

FIG. 1. Predicted rat VAChT amino acid sequence, species-conserved features of the protein, and its pro- posed structure within the synaptic vesicle membrane. Twelve putative transmembrane domains (I-XZZ) and po- tential sites for N-linked glycosylation (three-pronged branches) and phosphoryl- ation by protein kinase C (P in circle) are indicated. Red indicates amino acids con- served among all known vesicular trans- porters, and blue indicates amino acids unique to all VAChTs. Solid black indi- cates intramembrane aspartic acid resi- dues common to all known vesicular transporters, and dark blue (in trans- membrane domain N) indicates a unique intramembrane aspartic acid common in all VAChTs.

the cDNA). Portions of the purified cosmid clone containing a 30-kb insert were subcloned and sequenced as depicted in Fig. 4.

Chromosomal Localization of Human VAChT-Chromosomal loca- tion of human VAChT was by in situ hybridization using a biotinylated cDNA probe as described (16, 17). Specific hybridization of the human VAChT cDNA was observed at 1Oq11.2 in 52 metaphase chromosomal spreads of 92 cells examined.

RESULTS AND DISCUSSION

Rat PC-12 cells synthesize ACh and accumulate it by proton antiport within synaptic vesicles (18). Fig. 1 shows the pre- dicted protein sequence of an -2.9-kb cDNA obtained by screening of a PC-12 cell library with a probe derived from the Torpedo vesamicol-binding protein cDNA. This protein is ho- mologous (48 and 66% identity, respectively) to the previously characterized C. elegans (9) and Tbrpedo putative ACh trans- porters (10). There is also significant homology (40 and 38% sequence identity, respectively) to the rat biogenic amine trans- porters VMATl and VMAT2 (formerly referred to as CGAT and SVATMAT, respectively) (19, 20). An -2.4-kb human homolog of the rat cDNA was also cloned from a human neuroblastoma cDNA library. The amino acid conservation between rat and human VAChT is 94%. Fig. 1 shows the residues (blue) con- served among UNC-17, Torpedo vesamicol-binding protein, and their rat and human homologs, and the residues (red) con- served between these and the related vesicular monoamine transporters human VMATl: human VMAT2 (21, 221, bovine VMAT2 (23, 241, and rat VMATs 1 and 2 (19, 20).

The protein encoded by the rat cDNA whose structure is depicted in Fig. 1 was expressed using a recombinant vaccinia expression system in a non-neuronal host cell (CV-11, previ- ously demonstrated to reconstitute proton-driven accumulation of monoamines into an intracellular compartment containing a vacuolar ATPase (20). In the presence of esterase inhibition, ACh permeates intact cells from the medium (25,261. Like the quaternary amine N-methyl-4-phenylpyridinium (MPP+) (27), ACh is able to cross the plasma membrane by diffusion or by an unidentified cationic transport system and is then sequestered within an acidic intracellular compartment in fibroblastic cells transfected with a vesicular tran~porter.~

J. D. Erickson, T. I. Bonner, and L. E. Eiden, manuscript in prepa- ration.

T h e permeabilized cell uptake assay used previously to demonstrate biogenic amine transport in CV-1 vaccinia-infected cells expressing VMATZ (20) was attempted with CV-1 cells expressing rat VAChT. The low signal-to-noise ratio obtained precluded the use of this method to definitively demonstrate ACh uptake mediated by VAChT, and its in- hibition by vesamicol and ATPase inhibitors, possibry because of the high rate of ACh efflux from the vacuolar ATPase-containing intracel- lular compartment. While the intact cell assay allows characterization

Cells expressing the rat UNC-17 homolog, but not the vesic- ular monoamine transporters VMATl or VMAT2, acquired the ability to accumulate ACh in a vesamicol-sensitive manner (Fig. 2A). ACh transport was inhibited by vesamicol with an IC,, of -6 m (Fig. 2E3) but not by reserpine or tetrabenazine, inhibitors of biogenic amine transport by VMATs (Fig. 2C). Transport was blocked by inhibitors of the vacuolar ATPase (N-ethylmaleimide, tributyltin, and bafilomycin) or the proton ionophore carbonyl cyanide-p-trifluoromethoxyphenylhydra- zone, which dissipate intraorganellar proton gradients, demon- strating that vesamicol-sensitive ACh uptake occurred into an intracellular compartment that contains an electrogenic proton pump (Fig. 2C). ACh transport into VAChT-transfected CV-1 cells was not blocked by pretreatment with 10 p~ hemicho- linium-3, an inhibitor of high affinity choline uptake at the plasma membrane of cholinergic cells (data not shown). High affinity binding of [3Hlvesamicol was observed in cells express- ing the rat UNC-17 homolog, but not mock-transfected or VMAT1- or VMAT2-expressing cells (Fig. W). The Kd for [3H]vesamicol binding was -6 m, similar to the IC,, for inhi- bition of PHlACh accumulation by vesamicol. Based on this evidence, the protein encoded by this messenger RNA is iden- tified as the mammalian vesicular acetylcholine transporter, or VAChT. Based on their conservation of key portions of the transmembrane domains of rat VAChT and distinctness from the VMAT transporter family, UNC-17, the Torpedo vesamicol- binding protein, and the human homolog of rat VAChT can all reasonably be assigned this functional identity also.

In situ hybridization demonstrated high levels of expression of VAChT mRNA in all major cholinergic cell groups examined, including peripheral postganglionic parasympathetic cells, preganglionic sympathetic and parasympathetic cells, ventral spinal cord and brainstem motoneurons, scattered cell groups in the basal forebrain, the habenula, and the striatum. This distribution is virtually identical to that previously reported for ChAT mRNA and protein (28-31) and is consistent with the expression of both VAChT and ChAT exclusively in the cholin- ergic nervous system (Fig. 3). Northern blot analysis indicated that a single VAChT mRNA of -3 kb is expressed in regions of the brain containing cholinergic neurons and in PC12 cells, suggesting that, in contrast to vesicular transport in biogenic amine-containing neurons, a single protein species can account for ACh vesicular transport throughout the cholinergic nervous system.

of the general functional and pharmacologic characteristics of vesicular uptake, as reported for VMAT2 (201, kinetic parameters of ACh uptake into the vacuolar ATPase-containing intracellular compartment cannot be determined in this assay.

Vesicular Acetylcholine

h

0 60 T i e (in)

120

Dansporter

B

21931

0 -9 -8 -7

log [ m a i c d ]

'F

0 50 100 150 [3H vesamico~] nu

FIG. 2. Functional identification of rat VAChT. A, the accumulation of c3H]ACh by fibroblasts transfected with VAChT (0) was reduced by L-vesamicol (0.5 p ~ ) (0) to levels obtained in mock-transfected cells (*), or in VMAT1- or VMAT2-transfected cells (data not shown). Means of triplicate determinations, which varied less than 5%, are shown. All values are corrected for uptake at time zero (-100 pmoVmg protein). This

uptake (2 h) by L-vesamicol. Mock-transfected values are substracted from each data point. Vesamicol had no effect on background (mock- experiment was repeated with essentially identical results (mean values within 5% of mean values of experiment shown). B, inhibition of L3H1ACh

transfected) values at any concentration. IC,, for inhibition by vesamicol was 6.6 n~ (range 4.4-8.7 I", values from two separate experiments performed in quadruplicate). C, L3H]ACh accumulation during a 120-min period following pretreatment with medium alone (Control), 0.5 PM L(-)-vesamicol (L-uesamicol), 0.5 p~ d+)-vesamicol (D'uesamicol), 0.5 p~ tetrabenazine (TBZ), 0.5 p~ reserpine (Res), 200 PM N-ethylmaleimide (NEM), 50 p~ tri-n-butyltin (Tbt ) , 10 1.1~ bafilomycin (Bafilo), or 40 PM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP). Mock-

of triplicate determinations, and each experiment was repeated at least once with quantitatively and qualitatively similar results. D , analysis of transfected background values were unaffected by any of the drug treatments, and were subtracted from each data point. Values are the means

[3Hlvesamicol binding revealed an affinity constant (K,) of approximately 6 nM for VAChT (0). [3HlVesamicol binding in cells expressing VMATl or VMAT2 (m) was not different from mock-transfected cells (0). Values are uncorrected for background and represent the means of triplicate determinations from a single experiment, which was repeated once with quantitatively similar results.

The presence of common structural and regulatory motifs within the VAChT and ChAT genes might suggest a transcrip- tional mechanism for the apparently stringent co-expression of mammalian VAChT and ChAT. A human VAChT cDNA was cloned and sequenced (see Fig. 1). Chromosomal in situ hybrid- ization localized the human VAChT gene to 10q11.2, which is the position of the ChAT gene locus (32). A human genomic cosmid clone hybridizing to human VAChT cDNA was found by direct sequencing to contain the entire human cDNA sequence uninterrupted by introns (Fig. 4). In addition, it contained por- tions of the ChAT gene: including the R-type ChAT exon, which corresponds to the first exon of the farthest upstream ChAT transcript so far identified in mammals (7, 33, 341, and the first coding exon of human ChAT (8). The entire human VAChT cDNA sequence lies uninterrupted between these two exons of the ChAT gene (Fig. 4).

The 5' end of the rat VAChT cDNA maps to a position within the human VAChT gene locus 359 bases 5' of the end of the human VAChT cDNA (based on 87% identity with the first 68 bases of the rat VAChT cDNA) and 640 bases 3' of the R exon within the human cosmid clone (Fig. 4, dotted line). Since the cDNA size is nearly the same as that of the mRNA, the tran- scription start site for VAChT may be relatively close to the putative ChAT R exon (35). Whether the mammalian VAChT and ChAT start sites for transcription are identical or only in close physical proximity, the production of transcripts for both

M. A. Chireux, A. Le Van Thai, and M. J. Weber, GenBank accession numbers M96104 and M96015.

is potentially regulated by a cis-active element just upstream of the R-type exon of the VAChT/ChAT locus imparting positive transcriptional regulation by nerve growth factor (331, which promotes survival and cholinergic phenotypic expression of forebrain neurons in the mammalian central nervous system (36). Transcription of ChAT from downstream promoters con- taining nerve growth factor-responsive cis-acting elements may also occur (7, 8, 34, 37). Transcription from the R-type pro- moter, as well as downstream ChAT promoters, is also likely to be negatively regulated by cell-specific silencer elements iden- tified near the R-type promoter (38).

The structure of the ChATNAChT (chu-1 /unc-l7) gene locus in C. eleguns is similar to that reported here for the human ChATNAChT gene locus. Thus, the UNC-17 coding exons reside within the first intron of the C. eleguns ChAT gene (39). Con- servation of this gene organization was unexpected, especially since both theVAChTRJNC-17 and CHATlchu-1 genes ofhuman and nematode are significantly divergent in their exon organi- zation (Fig. 4) (38,39). Thus, human VAChT contains no introns within the coding region, while the unc-17 gene is interrupted by two introns within the coding domain. The evolutionary con- servation of the nested gene structure of the VAChT/ChAT locus therefore underscores its likely regulatory significance.

The coordination of transcription from a single gene locus of separate gene products whose cell-specific expression together determines cellular phenotype has not previously been de- scribed in the human genome. The physical contiguity of the genes encoding VAChT and ChAT may allow coordinate expres-

21932 Vesicular Acetylcholine Dansporter

A

28C

185

FIG. 3. Distribution of VAChT mRNA-positive cells in the cho- linergic nervous system. A, in situ hybridization histochemical visu- alization of representative VAChT mRNA-positive cells in the choliner- gic nervous system using a “’S-labeled VAChT cRNA antisense probe. Low power dark field (left; scale bar = 250 pm for upper panel, 500 pm for lower panels) and high power bright field (right; scale bar = 50 pm) micrographs of emulsion dipped autoradiograms. VAChT mRNA con- taining neurons in the basal nucleus of Meynert ( B ) show a scattered distribution characteristic for the cholinergic projection neurons. Scat- tered neurons in the caudate putamen (CPu) resemble striatal inter- neurons. Strongest hybridization is observed in motoneurons of the ventral horn (VH) and cranial motor nuclei (not shown), and presumed preganglionic sympathetic neurons of the thoracic intermediolateral cell column ( IML) . The intermediomedial cell column ( I M M ) of the sacral cord containing preganglionic parasympathetic neurons exhibits a less intense signal, as do presumptive post-ganglionic peripheral parasympathetic cell bodies (not shown). GP, globus pallidus; ic, inter- nal capsule; cc, central canal. B , Northern blot of polyadenylated RNA (5 pg) from rat tissues and PC12 cells hybridized to “’P-labeled rat VAChT cDNA shows a single -3-kb message in various brain regions and in PC12 cells but no signal in adrenal gland, cerebellum, or testis.

sion in cholinergic cells, and coordinate restriction of expression in non-cholinergic ones. This coordinate regulation may be im- portant in mammalian cholinergic nervous system development and functional loss of CNS cholinergic tone in disturbances of memory such as occurs in Alzheimer’s disease (40, 41).

*no\ 55t‘ sa\ s5.r‘ %$& \ Ch4T I I I I Ch4T

exm =Rexon ’ ’ ’ lncodmg

_.”” + VAChT cDNA -

0.5kb

FIG. 4. The human VAChT gene lies within the ChAT gene lo- cus. Schematic representation of the VAChT/ChAT gene locus indicat- ing that the human VAChT cDNA sequence is contained within the first putative ChAT intron. The ChAT R exon and the first ChAT coding exon are those referred to as “R” and “ M by Wu and Hersh for the human ChAT gene (35). The entire human VAChT gene locus within the BstXI sites was sequenced and shown to contain the entire human VAChT cDNA uninterrupted by introns. The dotted line indicates probable 5‘ extension of the primary human VAChT mRNA transcript compared to the human VAChT cDNA clone, based on homology to the rat VAChT cDNA as described under “Results and Discussion.”

REFERENCES

2. Kanner, B. I., and Schuldiner, S. (1987) CRC Crit. Rev. Biochem. 22, 1-38 1. Whittaker, V. P. (1988) The Cholinergic Synapse, Springer-Verlag, Berlin

3. Jahn, R., and Siidhof, T. C. (1993) J. Neurochem. 61, 12-21 4. Berrard, S., Brice, A,, Lottspeich, F., Braun, A,, Barde, Y. A,, and Mallet, J.

(1987) Proc. Nutl. Acad. Sci. U. S. A. 84,9280-9284 5. Brice, A,, Berrard, S., Raynaud, B., Ansieau, S., Coppola, T., Weber, M. J., and

Mallet, J. (1989) J. Neurosci. Res. 23,266-273 6. Ishii, K., Oda, Y., Ichikawa, T., and Deguchi, T. (1990) Mol. Bruin Res. 7,

7. Misawa, H., Ishii, K., and Deguchi, T. (1992) J. Biol. Chem. 267,20392-20399 151-159

8. Hersh, L. B., Kong, C. E , Sampson, C., Mues, G., Li, Y.-P., Fisher, A,, Hilt, D.,

9. Alfonso, A,, Grundahl, K., Duerr, J. S., Han, H.-P., and Rand, J. B. (1993)

10. Varoqui, H., Diebler, M.-F., Meunier, F.-M., Rand, J. B., Usdin, T. B., Bonner,

11. Parsons, S. M., Prior, C., and Marshall, I. G. (1993) Znt. Rev. Neurobiol. 35,

12. Seed, B. (1987) Nature 329,840-842 13. Fuerst, T. R., Niles, E. G., Studier, F. W., and Moss, B. (1986) Proc. Nutl. Acud.

Sci. U. S. A. 83,812243126 14. Schafer, M. K.-H., Day, R., Cullinan, W. E., Chretien, M., Seidah, N. G., and

Watson, S. J. (1993) J. Neurosci. 13, 1258-1279 15. Okayama, H., Kawaichi, M., Brownstein, M., Lee, F., Yokota, T., and Arai, K.

and Baetge, E. E. (1993) J. Neurochem. 61,306-314

Science 261,617-619

T. I., Eiden, L. E., and Erickson, J. D. (1994) FEBS Lett. 342,97-102

279-390

16. Lawrence, J. B., Villnave, C. A,, and Singer, R. H. (1988) Cell 52,51-61 17. Pinkel, D., Straume, T., and Gray, J. W. (1986) Proc. Nutl. Acad. Sci. U. S. A.

(1987) Methods Enzymol. 154,3-28

18. Toll, L., and Howard, B. D. (1980) J. Biol. Chem. 255, 1787-1789 19. Liu, Y., Peter, D., Roghani, A., Schuldiner, S., Prive, G. G., Eisenberg, D.,

Brecha, N., and Edwards, R. H. (1992) Cell 70,539-551 20. Erickson, J. D., Eiden, L. E., and Hoffman, B. (1992) Proc. Natl. Acud. Sci.

U. S. A. 89,10993-10997 21. Surratt, C., Persico, A,, Yang, X., Edgar, S., Bird, G., Hawkins, A,, Griffin, C.,

Li, X., Jabs, E., and Uhl , G. (1993) FEBS Lett. 318,325-330 22. Erickson, J. D., and Eiden, L. E. (1993) J. Neurochem. 61,2314-2317 23. Krejci, E., Gamier, B., Botton, D., Isambert, M . 3 , Sagne, C., Gagnon, J.,

24. Howell, M., Shirvan, A,, Stern-Bach, Y., Steiner-Mordoch, S., Strasser, J. E., Massoulie, J., and Henry, J.-P. (1993) FEBS Lett. 335, 27-32

25. Dan, Y., and Poo, M.-m. (1994) J. Neurobiol. 25, 336-341 Dean, G. E., and Schuldiner, S. (1994) FEBS Lett. 338, 16-22

26. Israel, M., Lesbats, B., Syngublakis, M., and Joliot, A. (1994) Neurochem. Znt.,

27. Liu, Y., Roghani, A,, and Edwards, R. H. (1992) Proc. Nutl. Acad. Sci. U. S. A.

28. Wainer, B. H., Levey, A. I., Mufson, E. J., and Mesulam, M. M. (1984) Neuro-

29. Oh, J. D., Woolf, N. J., Roghani, A,, Edwards, R. H., and Butcher, L. L. (1992)

83,2934-2938

in press

89,9074-9078

chem. Znt. 2, 163-182

30. Lauterborn, J. C., Isackson, P. J., Montalvo, R., and Gall, C. M. (1993) Mol. Neuroscience 47,807422

31. Mori, N., Tajima, Y., Sakaguchi, H., Vandenbergh, D. J., Nawa, H., and Sal-

32. Cohen-Haguenauer, 0.. Brice, A,, Berrard, S., Nguyen, V. C., Mallet, J., and

33. Bejanin, S., Habert, E., Berrard, S., Edwards, J.-B. D. M., Loemer, J.-P., and

34. Kengaku, M., Misawa, H., and Deguchi, T. (1993) Mol. Bruin Res. 18, 71-76 35. Wu, D., and Hersh, L. B. (1994) J. Neurochem. 62, 1653-1663 36. Lindsay, R. M., Weigand, S. J., altar, C. A,, and DiStefano, P. S. (1994) ?Fends

Bruin Res. 1 7 , 5 9 4 9

vaterra, P. M. (1993) Mol. Bruin Res. 17, 101-111

Frezal, J. (1990) Genomics 6, 374-378

Mallet, J. (1992) J. Neurochem. 58, 1580-1583

37. Ibanez, C. F., and Persson, H. (1991) Eur. J. Neurosci. 3, 1309-1315 38. Li, Y.-P., Baskin, F., Davis, R., and Hersh, L. B. (1993) J. Neurochem. 61,

Neurosci. 17, 182-190

39. Alfonso, A., Grundahl, K., McManus, J. R., Asbury, J. M., and Rand, J. B.

40. Davies, P., and Maloney, M. R. (1976) Lancet ii, 1403 41. Perry, E. K., Perry, R. H., Blessed, G., and Tomlinson, B. E. (1977)Lancet i, 189

748-751

(1994) J. Mol. Biol., in press