Proc. Nati. Acad. Sci. USAVol. 91, pp. 10933-10937, November 1994Neurobiology

Regional variation in expression of acetylcholinesterase mRNA inadult rat brain analyzed by in situ hybridization

(cholinergic neuroanatomy/transcriptional regulation of acetylcholinesterase)

P. HAMMOND, R. RAO, C. KOENIGSBERGER, AND S. BRIMuJOIN*Department of Pharmacology, Mayo Clinic, Rochester, MN 55905

Communicated by George B. Koelle, July 11, 1994

ABSTRACT To investigate the molecular basis of regionalvariation in expression of brain acetylcholinesterase (AChE;EC 3.1.1.7), steady-state levels of AChE activity and mRNAwere examined. Relative AChE activity in Triton extracts fromsix areas ofthe rat brain varied as follows: cortex < cerebellum<medulla < pons-midbrain < thalamus < striatum. In con-tralateral samples from the same brains, AChE mRNA wasassesed by Northern blotting with random-primed 32P-labeledcDNA. The regiona abundanc of the major 2.4-kb AChEtranscript differed from that of the enzyme activity: cortex <striatum < cerebellum < medulla < thalamus < pons-mid-brain. In situ hybridization with a MP-labeled antisense AChEoligonucleotide provided evidence for high levels of AChEmessage in cells of the nucleus basalis, nucleus accumbens,neostriatum, subantia nigra, motor nucleus of the facialnerve, and spinal nucleus of the trigeminal nerve. In thecaudate-putamen, large, heavily labeled neurons were notnumerous, but they were approximately as frequent as thechollnergic interneurons revealed by choline acetyltransferaseimmunocytochemistry. The relatively low number of theseAChE-expressing cells probably explains the relative dearth ofAChE mRNA-like material in the neostriatum.

Many studies have been undertaken to define the localizationof acetylcholinesterase (AChE; EC 3.1.1.7) and its molecularforms in brain (for reviews, see refs. 1-4). The results showthat AChE, like choline acetyltransferase (ChAT), is con-centrated in cholinergic centers like the neostriatum, basalforebrain, superior colliculus, and motor nuclei of cranialnerves (5-7). In certain other areas, such as substantia nigraand cerebellum, noncholinergic cells express large amountsof AChE without detectable ChAT (8). Elsewhere, AChEactivity is quite low.

Since the control ofAChE expression in nerve cells remainspoorly understood, one of our goals is to determine how thebrain acquires its characteristic pattern of AChE-rich andAChE-poor regions. In muscle, posttranscriptional mecha-nisms are thought to account for much ofAChE's spatial andtemporal variation (9). Different mechanisms may operate inbrain. It remains possible that the local AChE content in eachbrain area reflects the local rate ofAChE production, which inturn reflects the local abundance of AChE mRNA (10). Toaddress this possibility, we combined conventional AChEhistochemistry with Northern blotting and in situ hybridiza-tion histochemistry in a study of the regional and cellulardistribution of AChE in the adult rat brain.

METHODSAnimals. Ten adult male Sprague-Dawley rats (250-350 g)

and three 16-day-old rat pups (30-40 g) were handled in

accordance with the National Institutes of Health Guidelinesfor Care and Use of Laboratory Animals under a protocolapproved by the Mayo Animal Care and Use Committee. Therats were killed by overdose with ether, and the brains wererapidly removed.

Tissue Processing for Biochemistry. Brains for enzymeassay or mRNA analysis were freshly dissected as describedby Glowinski and Iversen (11). Samples were taken from sixregions: medulla, pons-midbrain, dorsal thalamus, cerebel-lum (including cortex and deep nuclei), striatum (caudate-putamen), and cerebral cortex (midline strip, frontal to oc-cipital pole). Equivalent samples from four to six rats werefrozen on dry ice, pooled to obtain 400-800 mg oftissue, andkept at -80'C until analyzed.Enzyme Assay. Tissues were thawed and homogenized in

10 volumes of 0.05 M sodium phosphate buffer (pH 7.4) with1% Triton X-100 and 0.5 M NaCl. After centrifugation at10,000 x g for 10 min, supernatant AChE activity wasmeasured spectrophotometrically with acetylthiocholine assubstrate (12). Ethopropazine (0.1 mM; Sigma) was used asa selective inhibitor of butyrylcholinesterase.

Preparation of RNA. Under a standard protocol (Invitro-gen), poly(A)+ RNA was extracted from tissues lysed inSDS-based buffer containing RNase inhibitor. Lysates wereincubated at 450C and applied to oligo(dT) cellulose columns.Differential salt elution was used to isolate mRNA, andconcentration was determined by A2Ni readings.Northern Blotting. Probes were prepared from plasmids

containing a 590-bp mouse AChE cDNA cloned into pBlue-script II KS- (ref. 13; kindly supplied by Palmer Taylor,University ofCalifornia, San Diego) or a 680-bp cDNA for ratcyclophilin, pibiS (14). Fragments ofthe expected sizes wereisolated from XL1-Blue Escherichia coli lysates by endonu-clease digestion followed by agarose gel electrophoresis.Probes were random primer labeled with [a-32P]dCTP to aspecific radioactivity of 1-3 x 108 cpm/pg of DNA.

Total RNA or poly(A)+ RNA was denatured in 50%(vol/vol) formamide (650C for 10 min) and electrophoresed in1.2% agarose gels with 6% formaldehyde. After electropho-resis, RNA was blotted onto nylon membranes, which werebaked at 800C for 2 hr, and then treated for 4 hr at 420C with50% formamide/5x standard saline citrate (SSC)/lOx Den-hardt's solution/denatured herring sperm DNA (250 gg/ml)/poly(A) (50 ,ug/ml)/0.1% SDS. Blots were hybridized in freshbuffer with random primer 32P-labeled cDNA for AChE (-5x 106 cpm/ml; 36-40 hr at 420C). Hybridized blots werewashed with 0.5 x SSC (three times at 230C, three more timesat 500C), dried, and exposed to Kodak XAR-5 film for up to10 days at -800C. AChE mRNA was standardized withreference to cyclophilin mRNA by reprobing with a 32plabeled probe for cyclophilin. For quantitation, relative banddensities were measured with a scanning densitometer.

Abbreviations: AChE, acetylcholinesterase; ChAT, choline acetyl-transferase.*To whom reprint requests should be addressed.

10933

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Mar

ch 6

, 202

1

10934 Neurobiology: Hammond et al.

In Situ Hybridization. Probes for in situ hybridization weresynthetic oligonucleotides based on the published sequencefor rat AChE mRNA (10), 3'-end-labeled with [33P]dATP(Amersham) to a specific activity of -2 x 108 cpm/pg. Threeantisense probes were evaluated, complementary to bases781-843, 1090-1152, and 1179-1228. Although the results werequalitatively similar with all probes, selective labeling was bestwith antisense 1090-1152, which was used for subsequentwork. Hybridization controls were probed with the corre-sponding radiolabeled sense DNA or were blocked by coin-cubation with a 100-fold excess of unlabeled antisense probe.As described by Dagerlind et al. (15), sections from unfixed

brains (15 /im, cut at - 12'C) were thawed onto glass slides,dried, and hybridized at 420C for 18-24 hr with probe ( 150 sII,106 cpm per section). Five posthybridization washes with 1xSSC were carried out at 550C (12 min each; a final rinse wasleft to reach room temperature over a 30-min period). Sectionswere then dipped in water, dehydrated in alcohols, and placedon Kodak XAR-5 film or coated with photographic emulsion(Kodak NTB2, diluted 1:1). Autoradiographs were exposed6-11 days (film) or 4-5 weeks (emulsion). Multiple overlap-ping grains constituted "heavy radiolabeling"; no attempt wasmade to distinguish lighter labeling from background.

Histology and Immunohistochemistry. After autoradiogra-phy, slides were Nissl stained with thionin (16). AChEhistochemistry was performed on adjacent cryostat sectionsby a modification of Koelle and Friedenwald's method (17).For ChAT immunocytochemistry, other brains were fixed insitu by perfusion with 4% buffered formalin, postfixed for 1hr in the same solution, and cryoprotected with 15% (wt/vol)sucrose. Cryostat sections were incubated with a monoclonalmurine anti-ChAT antibody (10 ,g/ml) provided by Boyd K.Hartman, Department of Psychiatry, University of Minne-sota, Minneapolis (18). Bound antibody was visualized bybiotinylated anti-mouse IgG and avidin-biotin-peroxidase(Vector Laboratories) according to Hsu et al. (19).

RESULTSNorthern blots of poly(A)+ RNA from pooled samples ofcerebellum, medulla, pons-midbrain, cerebral cortex, dorsalthalamus, and striatum developed the expected major band at-2.4 kb when probed with [32P]AChE cDNA (Fig. 1). Aweaker band at about 3.2 kb also appeared in some samples.No hybridization was seen with extracts from liver or kidney(data not shown). The 2.4-kb AChE transcript predominatedin all brain samples and was especially abundant in medulla,thalamus, and pons-midbrain. When AChE mRNA was quan-titated against cyclophilin mRNA, a wide range of relativeabundance was apparent; AChE activity also varied widely inthese samples but in different rank order (Fig. 1). Dissociationof activity and mRNA was particularly striking in pons andstriatum. Pons-midbrain had the most AChE mRNA but onlyaverage AChE activity; striatum had the most AChE activitybut below average levels of AChE mRNA.

Regional variations ofAChE mRNA were revealed in moredetail by in situ hybridization with 33P-labeled antisense DNA(bp 1090-1152). A clear pattern of selective labeling roughlyparalleled the localization of AChE activity. Film autoradi-ography (four adult brains, sagittal sections at 12 equallyspaced intervals, 0.5-5.0 mm lateral to the midline) demon-strated the abundance of transcripts selectively hybridizingwith the DNA probe in the basal forebrain, substantia nigra,and certain areas of the pons and medulla (Fig. 2). Caudatenucleus was only moderately labeled overall, consistent withthe Northern blots, but it did show many isolated "hotspots." This selective pattern was abolished when hybrid-ization was performed in the presence of a 100-fold excess ofunlabeled antisense probe. Control autoradiographs with thecorresponding "sense" oligonucleotide also showed a diffusebackground instead of specific labeling (Fig. 3).

;.O

F

Adu.) 1

-a:

Ii0 -

7 activrtv[7 rnRNA

.:.-fl I.r-Lo

C)C)C0

..'I. LI)

.-

kb

3.4-

1.8-

FIG. 1. Biochemical comparison of the regional abundance ofAChE activity and mRNA in adult rat brain. Corresponding brainregions from five rats were divided at the midline and pooledseparately. One side was used for AChE assay (bar graph showsmeans ± SEM of three independent trials). The contralateral sidewas used for Northern blotting, in which 1.7 jug of poly(A)+ RNAfrom each region was hybridized with a 590-bp 32P-labeled cDNA formurine AChE. AChE message was quantitated with reference tocyclophilin mRNA after stripping and reprobing (bar graph showsmean of three independent trials; a representative AChE Northernblot is illustrated below).

To identify specifically labeled structures more precisely,adjacent sections from the brains studied by film autoradi-ography were also prepared for AChE histochemistry anidemulsion autoradiography with cellular resolution. ChATimmunocytochemistry was carried out on comparable sec-tions from a separate, optimally fixed brain. Emulsion auto-radiography with the antisense probe showed heavily labeledcells with silver grains concentrated over the cytoplasm.Such cells were found at all of the previously identified"message-rich" sites, including the caudate putamen, wherethey corresponded to the isolated hot spots seen in filmautoradiography. Heavy cellular labeling was also noted insubstantia nigra, nucleus accumbens, olfactory tubercle, andthe principal motor nuclei of the cranial nerves, but thehighest local concentration of heavily labeled cells was foundin the nucleus basalis (Fig. 4). The enrichment in the nucleusbasalis was even more impressive in immature brains, 16days postnatal (Fig. 5). There were few if any heavily labeledcells in other parts of the brain, including the superiorcolliculus and the frontal, parietal, and occipital cortex. Inother words, the distribution of radiolabel at the cellular levelaccounted for the patterns obtained on film.

Histochemistry of adjacent sections revealed that the areaswith cells strongly labeled by antisense AChE DNA alsocontained many cells with AChE activity and usually ChATimmunoreactivity as well (data not shown). Wherever cellu-lar staining could be assessed, the prevalence of cells stainedfor AChE activity appeared commensurate with the preva-lence of cells labeled by antisense DNA. In the caudatenucleus, where the overlying AChE-rich fiber matrix ob-scured the perikarya, we compared AChE mRNA labelingwith ChAT immunostaining. Comparable sections from twobrains showed similar numbers of caudate neurons that werestrongly immunoreactive for ChAT (36 per section) and those

Proc. Natl. Acad Sci. USA 91 (1994)

..

'7-

f--% cl.-.1

1,

., 1"

M."'.."

I'i

w.

Dow

nloa

ded

by g

uest

on

Mar

ch 6

, 202

1

Proc. Natl. Acad. Sci. USA 91 (1994) 10935

A

B-

that were heavily radiolabeled by antisense AChE DNA (33per section).

DISCUSSIONAChE mRNA is readily detected in brain by in situ hybrid-ization as well as Northern blotting, even though the poly-peptide makes up less than 0.005% of total brain protein (20).There is strong evidence of specificity in that (i) the hybrid-ization was blocked by an excess of unlabeled antisenseDNA, (ii) the corresponding sense DNA failed to hybridizeappreciably under the same conditions, and (iii) similarpatterns of labeling were obtained with different, nonover-lapping oligonucleotide probes. Therefore, our autoradio-graphs should reliably portray the distribution of AChEmRNA in adult rat brain.The picture emerging from this study largely meets expec-

tations from previous work on the localization of brainAChE. Antisense AChE DNA probes detect wide variationsof AChE mRNA-like material concentrated in the principalcholinergic nuclei but also quite abundant at certain othersites. In many areas (e.g., nucleus basalis, facial nervenucleus) the patterns of heavy cellular labeling and stainingfor AChE activity or ChAT immunoreactivity are convinc-ingly similar. Such results fit the view that cholinergicneurons all produce large amounts of AChE. AbundantAChE mRNA-like material in the pars compacta of thesubstantia nigra, a "noncholinergic" site, was also expectedfrom the classical histochemical picture ofAChE activity (7).This finding is in line with the rule that cholinoceptive cellstypically express AChE at levels that compare with cholin-ergic, acetylcholine-producing cells. Finally, one would havepredicted low AChE mRNA in the superficial superior col-liculus, which is rich in AChE-containing fibers but poor inAChE-reactive perikarya. Presumably, the AChE content ofthis structure is delivered by axonal transport from cholin-ergic cell bodies in the pedunculopontine tegmental nucleus.The distribution ofAChE mRNA does have some surpris-

ing aspects. One ofthese is the apparent abundance in certain

-Fii. 2. Histochemical com-parison of the regional aibun-dance of AChE activitvGandrnRN..A in atdult raIt hrain. Adja-Cent sagittal crxostlt sections ofadult rat brain werc examined bryconventional AChF histochem-

A) and b\ in bif hybrid-ij ization with a 33P-labeled an-

tisense oligonucleoti-de probe forrat AChE mRRNA (B: film auto-radiography. 10-day exposure).Corresponding structures are in-dicated: 1. nucleus of facialnerve: 2. pons and pituitary be-

_low: 3. superior colliculus: 4.hippocampus: 5. caudate nu-cleus: 6. olfactorx tubercie andadjacent nucleus basalis: c. sub-stantia nigra.

parts of the cerebellum, where little AChE enzyme is de-tected histochemically. We suspect that much of the labelingof the granular layer may be artifactual, since it was incon-sistent (compare Figs. 2 and 3), and since this locus isnotorious for nonspecific binding of cDNA probes. On theother hand, the possibility should be considered that thecerebellum produces significant amounts of catalytically in-active AChE (21), with as yet unknown functions. Closerstudy of this issue is justified.A more convincing discrepancy between mRNA and en-

zyme levels is in the neostriatum, the brain region whereAChE is overwhelmingly most abundant. In announcing theircDNA clone for rat AChE, Legay et al. (22) noted that thestriatum was relatively poor in AChE mRNA as measured byNorthern blotting. Our results confirm this paradoxical find-ing and provide a partial explanation: the low overall level ofstriatal AChE mRNA reflects a scarcity of neurons express-ing the transcript at high levels. Comparison of antisenselabeling and ChAT immunocytochemistry does suggest thatall of the cholinergic interneurons are "AChE expressors."In fact, the distribution ofcDNA-labeled cells shows remark-able similarities to the distribution ofAChE-positive neuronsafter pulse inhibition by diisopropyl fluorophosphate as de-scribed by Butcher and coworkers (23, 24). Thus, the corre-lation is good at the cellular level.Even in the caudate nucleus, cholinergic neurons make up

only a small percentage of the cellular mass. Therefore, toreconcile the molecular data with enzymological and histo-chemical observations, one must explain how the neostria-tum maintains so much active AChE without equivalentlyhigh steady-state levels of specific mRNA. Two possibilitiesto consider are that striatal AChE (i) is chiefly located in thelocal projections of extrastriatal perikarya (import hypothe-sis) or (ii) has an unusually long half-life (turnover hypoth-esis). We favor the second of these ideas.

Available information on the import hypothesis is conflict-ing, but the weight of evidence is negative-that is, little ofthe striatal AChE seems to be derived from outside sources.

Neurobiology: Hammond et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 6

, 202

1

10936 Neurobiology: Hammond et al.

C

FIG. 3. Specificity controls forin situ hybridization: film autora-diography with 10-day exposure.(A) 33P-labeled antisense AChEprobe (bp 1090-1152). Punctate la-beling is dense in olfactory tuber-cle and nucleus basalis (1) and alsoevident in caudate nucleus (2) andhippocampus (3). (B) Same probeblocked by addition of 100-foldexcess of unlabeled antisenseAChEDNA. (C) 33P-labeled senseAChE probe (bp 1090-1152).

The AChE-rich matrix characterizing the caudate-putamenconsists mainly of fibers and terminals, as shown by the lossand recovery of fibrous AChE staining after irreversibleinhibition by organophosphates (24, 25) or immunologiclesions of terminals (26). Several studies suggest that thisAChE-rich neuropil stems from cholinergic interneurons.Histochemical staining for AChE is largely preserved afterinterruption of striatal afferents (25, 27). Although morequantitative biochemical measures have detected small re-ductions in striatal AChE after afferent lesioning (28), it islikely that the bulk of the enzyme is locally produced.Therefore, we suggest that the neostriatum accumulatesmore than average amounts of AChE, not because it pro-duces greater quantities of AChE than other regions, butbecause it retains more of the enzyme and keeps it longer.The opposite may be true of the nucleus basalis, where highlevels of AChE mRNA are associated with "export-

oriented" cholinergic projection neurons. These ideas un-derlie the turnover hypothesis.As Legay et al. (10) point out, striatal AChE is primarily of

the membrane-bound oligomeric form (G4), which is gener-ally more stable than the smaller AChE forms (G1 and G2).One study of AChE recovery after irreversible inhibition bydiisopropyl fluorophosphate did find signs of relatively slowturnover in striatum (29), but another study obtained oppositeresults with the related organophosphate soman (30). Acritical reexamination of the regional turnover of AChE inbrain is now warranted, taking into consideration the multiplemolecular forms of the enzyme, the potential existence ofpools of inactive AChE (21), and the possibility that organ-ophosphates might alter AChE synthesis.

In conclusion, we find that AChE activity and mRNA bothvary greatly from one part of the brain to another. As theabundance of the message largely parallels that of the protein

Proc. Natl. Acad. Sci. USA 91 (1994)

Dow

nloa

ded

by g

uest

on

Mar

ch 6

, 202

1

Proc. Natl. Acad. Sci. USA 91 (1994) 10937

B

am

pP N

*i.~ ~ ~ ~ ~ .%.71

Co D

-- 4- t ... b at

~hm

/ S .;

ice,: ,~~~~4-

FIG. 4. Cellular localization of AChE message. Sections werehybridized with 33P-labeled AChE probe (bp 1090-1152), dipped inemulsion, exposed for 30 days, developed, and counterstained withthionin. (A) Basal medulla, nucleus of the facial nerve. (Bar = 100Aum.) (B) Caudate nucleus. (C) Olfactory tubercle and nucleusbasalis. (D) Same as C, hybridized with sense probe instead ofantisense probe.

itself, the steady-state level of AChE mRNA in many brainregions may well reflect the enzyme's rate of biosynthesis.This relationship may not be universal, however, especiallywhen the neostriatum is taken into consideration. Hence,further studies are needed at the regional and cellular level todefine the relation between AChE mRNA, AChE protein, andAChE turnover.

We thank Drs. Palmer Taylor, William Randall, and CynthiaMcMurray for providing cDNAs for AChE and cyclophilin and Dr.Boyd Hartman for contributing anti-ChAT antibody. This study wassupported by Grant NS 29646 from the National Institute of Neu-rological Disorders and Stroke.

1. Koelle, G. B. (1963) in Handbook of Experimental Pharma-cology: Cholinesterases and Anticholinesterase Agents, ed.Koelle, G. B. (Springer, Berlin), Vol. 15, pp. 187-298.

2. Massoulid, J. & Bon, S. (1982) Annu. Rev. Neurosci. 5, 57-106.3. Brimijoin, S. (1983) Prog. Neurobiol. 21, 291-322.4. Massoulid, J., Pezzementi, L., Bon, S., Krejci, E. & Vallette,

F.-M. (1993) Prog. Neurobiol. 41, 31-91.5. Koelle, G. B. (1954) J. Comp. Neurol. 100, 211-236.6. Butcher, L. L. & Woolf, N. J. (1984) in Handbook ofChemical

Neuroanatomy, eds. Bj6rklund, A. J., Hckfelt, T. & Kuhar,M. J. (Elsevier, Amsterdam), Vol. 3.

7. Eckenstein, F. & Sofroniew, M. V. (1983) J. Neurosci. 3,2286-2291.

8. Paxinos, G. & Butcher, L. (1985) in The Rat Nervous System,ed. Paxinos, G. (Academic, New York), Vol. 1, pp. 487-521.

9. Rotundo, R. L. (1988) J. Biol. Chem. 263, 19398-19406.10. Legay, C., Bon, S., Vernier, P., Coussen, F. & MassouliU, J.

(1993) J. Neurochem. 60, 337-346.

FIG. 5. AChE message in nucleus basalis of 16-day-old rat pup.A sagittal section was hybridized with 33P-labeled AChE antisenseprobe (bp 1090-1152; see legend to Fig. 4) and photographed withbrightfield (A) and darkfield (B) optics. (Bar = 200 Pm.) bas, nucleusbasalis; th, thalamus; ic, internal capsule; gp, globus pallidus.

11. Glowinski, J. & Iversen, L. L. (1966) Biochem. Pharmacol. 15,977-987.

12. Ellman, G. L., Courtney, K. D., Andres, V. J. & Feather-stone, R. M. (1961) Biochem. Pharmacol. 7, 88-95.

13. Rachinsky, T. L., Camp, S., Li, Y., Ekstrom, T. J., Newton,M. & Taylor, P. (1990) Neuron 5, 317-327.

14. Danielson, P. E., Forss-Petter, S., Brow, M. A., Calavetta, L.,Douglass, J., Milner, R. J. & Sutcliffe, J. G. (1988) DNA 7,261-267.

15. Dagerlind, A, Friberg, K., Bean, A. J. & Hbkfelt, T. (1992)Histochemistry 98, 39-49.

16. Donovick, P. (1974) Stain Technol. 49, 49-51.17. Koelle, G. B. & Friedenwald, J. S. (1949) Proc. Soc. Exp. Biol.

Med. 70, 617-622.18. Cozzari, C., Howard, J. & Hartman, B. (1990) Soc. Neurosci.

Abstr. 16, 200.19. Hsu, S. M., Raine, L. & Fanger, H. (1981) J. Histochem.

Cytochem. 29, 577-580.20. Rakonczay, Z., Mallol, J., Schenk, H., Vincendon, G. &

Zanetta, J.-P. (1981) Biochim. Biophys. Acta 657, 243-256.21. Chatel, J.-M., Grassi, J., Frobert, Y., Massouli6, J. & Vallette,

F.-M. (1993) Proc. Natl. Acad. Sci. USA 90, 2476-2480.22. Legay, C., Bon, S. & Massoulid, J. (1993) FEBS Lett. 315,

163-166.23. Butcher, L. L. & Hodge, G. K. (1976) Brain Res. 106,223-240.24. Butcher, L. L. & Bilezikjian, L. (1975) Eur. J. Pharmacol. 34,

115-125.25. Lynch, G. S., Lucas, P. A. & Deadwyler, S. A. (1972) Brain

Res. 45, 617-621.26. Bean, A. J., Xu, Z., Chai, S. Y., Brmijoin, S. & H6kfelt, T.

(1991) Neurosci. Lett. 133, 145-149.27. McGeer, P. L., McGeer, E. G., Fibiger, H. C. & Wickson, V.

(1971) Brain Res. 35, 308-314.28. Lehmann, J. & Fibiger, H. C. (1978) J. Neurochem. 30, 615-

624.29. Kozar, M. D., Overstreet, D. H., Chippendale, T. C. & Rus-

sell, R. W. (1976) Neuropharmacology 15, 291-298.30. Yaksh, T. L., Filbert, M. G., Harris, L. W. & Yamamura,

H. I. (1975) J. Neurochem. 25, 853-860.

Neurobiology: Hammond et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 6

, 202

1