Immunoreactivity of Protein Gene Product 9.5 (PGP 9.5)in the Developing Hamster Olfactory Bulb

TAKAYUKI NAKAJIMA, CHIKAKO MURABAYASHI, KAZUSHIGE OGAWA,AND KAZUYUKI TANIGUCHI*

Department of Veterinary Anatomy, Faculty of Agriculture, Iwate University, Iwate, Japan

ABSTRACT Background: Protein gene product 9.5 (PGP 9.5) is a solubleprotein isolated from human brain and recently demonstrated to corre-spond to ubiquitin carboxyl terminal hydrolase. We examined the PGP 9.5immunoreactivity in the developing hamster olfactory bulb to determinewhether the olfactory bulb expresses PGP 9.5 from the early developmen-tal stage and whether it is a useful marker for investigating the differentia-tion of bulbar neurons and the development of laminar organization.

Methods: The developing hamster olfactory bulb was fixed in Bouin’ssolution without acetic acid or acetic alcohol (95% absolute alcohol with5% acetic acid) and immunostained with human PGP 9.5 antiserum by theavidin-biotin peroxidase complex (ABC) method.

Results: PGP 9.5 immunoreactivity was detected in the axons of olfac-tory and vomeronasal receptor cells and immature bulbar neurons fromembryonic day 13 in the tissue fixed in Bouin’s solution without acetic acidand in the immature bulbar neurons from embryonic day 11 in the tissuefixed in acetic alcohol. In some bulbar neurons, PGP 9.5 immunoreactivitywas detected not only in the cytoplasm, but also in the nuclei. The patternof immunostainings for PGP 9.5 became almost the same as in the adult atpostnatal day 15. From the immunostaining for PGP 9.5, we found that theshort axon cells in addition to mitral, tufted, and mitral/tufted cells wereidentified in the prenatal period and that the extension of the dendrites,formation of the glomeruli, and completion of the laminar organizationwere earlier in the main olfactory bulb than in the accessory olfactorybulb.

Conclusions: PGP 9.5 can be a useful marker to investigate the differentia-tion of bulbar neurons and the formation of laminar organization in thedeveloping olfactory bulb. Anat. Rec. 250:238–244, 1998. r 1998 Wiley-Liss, Inc.

Key words: accessory olfactory bulb; development; hamster; immunohis-tochemistry; main olfactory bulb; protein gene product 9.5

The olfactory bulb, a primary center of olfactorysystems, is separated into two regions: main olfactorybulb (MOB) and accessory olfactory bulb (AOB). Therodent MOB can be divided into six layers: olfactorynerve layer (ONL), glomerular layer (GL), externalplexiform layer (EPL), mitral cell layer (MCL), internalplexiform layer (IPL), and granule cell layer (GCL)(Macrides and Davis, 1983). The MOB receives theprojection from olfactory receptor neurons in the olfac-tory epithelium, constituting the main olfactory system(Macrides and Davis, 1983; Mori, 1987; Scalia andWinans, 1975). In contrast, the rodent AOB can bedivided into five layers: vomeronasal nerve layer (VNL),GL, mitral/tufted cell layer (MTL), lateral olfactorytract (LOT), and GCL (Takami et al., 1992). The AOBreceives the projection from receptor neurons in thevomeronasal organ, constituting the accessory olfactorysystem (Halpern, 1987; Macrides and Davis, 1983;

Mori, 1987; Scalia and Winans, 1975; Takami andGraziadei, 1991).

Since protein gene product 9.5 (PGP 9.5), a solubleprotein isolated from the human brain, is found predomi-nantly in the cytoplasm of neurons and neuroendocrinecells, this protein has been used as a marker for thesecells (Doran et al., 1983; Johnson et al., 1994; Kent andRowe, 1992; Nakajima et al., 1996; Shofield et al., 1995;Taniguchi et al., 1993; Thompson et al., 1983; Wilson etal., 1988). Recent studies demonstrated that PGP 9.5corresponds to ubiquitin carboxyl terminal hydrolase,which regenerates ubiquitin from ubiquitin/protein com-plexes or polyubiquitin chains formed in the proteolytic

*Correspondence to: Dr. Kazuyuki Taniguchi, Department of Veteri-nary Anatomy, Faculty of Agriculture, Iwate University, 3-18-8 Ueda,Morioka, Iwate 020, Japan.

Received 26 September 1996; Accepted 14 October 1997

THE ANATOMICAL RECORD 250:238–244 (1998)

r 1998 WILEY-LISS, INC.

pathways (Hershko and Ciechanover, 1992; Wilkinsonet al., 1989).

The olfactory marker protein (OMP) is known to be auseful marker for olfactory and vomeronasal receptorcells, although it is not expressed in neurons of theolfactory bulb (Margolis, 1985; Monti-Graziadei, 1980).However, PGP 9.5 was recently demonstrated to beexpressed in the developing and mature rat olfactoryand vomeronasal receptor cells (Johnson et al., 1994;Kent and Rowe, 1992; Shofield et al., 1995; Taniguchi etal., 1993). In addition, PGP 9.5 was demonstrated to bedistributed predominantly in the mitral, tufted, mitral/tufted and short axon cells in the rat and hamsterolfactory bulb in our previous studies (Nakajima et al.,1996; Taniguchi et al., 1993), although there has beenno report on the immunohistochemical study on PGP9.5 expression in the developing olfactory bulb.

In the present study, therefore, we examined theimmunoreactivity for PGP 9.5 in the developing ham-ster olfactory bulb in order to determine whether theolfactory bulb expresses PGP 9.5 from the early develop-mental stage and whether it is a useful marker forinvestigating the differentiation of bulbar neurons andthe development of laminar organization.

MATERIALS AND METHODS

Fetuses of golden hamster at embryonic day 11 (E11),n 5 15; E12, n 5 15; E13, n 5 15; E14, n 5 15; E15, n 515; the pups at postnatal day 0 (P0), n 5 13; P1, n 5 10;P3, n 5 10; P5, n 5 10; P7, n 5 8; P10, n 5 8; P15, n 5 8;and five adult animals were used as materials in thepresent study. The animals were kept under constantlighting conditions (12 h light:12 h dark) with waterand food ad libitum. The morning of the detection of avaginal copulatory plug was estimated as day 0 (E0) ofgestation. In the present study, two different fixatives,Bouin’s solution without acetic acid and acetic alcohol(95% absolute alcohol with 5% acetic acid), were used.In the fetuses at E11 and E12, the brain was removedquickly after decapitation and immersed in Bouin’ssolution without acetic acid for 24–48 h, or aceticalcohol for 5–6 h. In the other stages, the animals weredeeply anesthetised by intraperitoneal injection of so-dium pentobarbital (60 mg/kg body weight) (Abbott,North Chicago, IL) and perfused with physiologicalsaline, followed by Bouin’s solution without acetic acidprior to the immersed fixation for 24–48 h, or by aceticalcohol prior to the immersed fixation for 5–6 h.

Fixed tissues were routinely dehydrated and embed-ded in paraffin wax. The paraffin sections were cutsagittally at 5 µm, deparaffinized with xylene, andprocessed for immunohistochemical staining. Immuno-staining was performed by the avidin-biotin peroxidasecomplex (ABC) method using the Vectastain ABC kit(Vector, Burlingame, CA) and involved the followingsteps: (1) incubation with 10% normal goat serum at32°C for 30 min, (2) rinse in 0.02 M phosphate bufferedsaline (PBS) (pH 7.25) for 15 min, (3) incubation withrabbit antihuman PGP 9.5 serum 1:3,000 in dilution(UltraClone, Wellow, UK) at 4°C for 48 h, (4) rinse inPBS for 15 min, (5) incubation with biotinylated goatantirabbit IgG 1:600 in dilution (Vector) at 32°C for 40min, (6) rinse in PBS for 15 min, (7) incubation withABC at 32°C for 40 min, (8) rinse in PBS for 15 min, (9)incubation with 0.05 M Tris-HCl (pH 7.6) containing

0.01% 3-38 diaminobenzidine tetrahydrochloride (DAB)and 0.003% hydrogen peroxide for 20–30 min, (10) rinsein distilled water. The characterization of the antise-rum has been described elsewhere (Chen et al., 1994;Kent and Rowe 1992; Thompson et al., 1983). Controlimmunostainings were performed by the use of nonim-mune rabbit or goat serum to replace the primaryantiserum. No specific immunoreactivity was observedin these control stainings.

RESULTSFixation in Bouin’s Solution Without Acetic Acid

At E11 and E12, the olfactory bulb was observed as asmall projection at the rostral end of the cerebralhemisphere and consisted of the germinal cells sur-rounding the olfactory ventricle and the immature cellsin the superficial layer. The immunoreactivity for PGP9.5 was not detected at these stages.

At E13, the distinction between the MOB and theAOB was already observed, although the boundary wasillegible. The MOB was divided into three layers, ONL,MCL, and germinal cell layer (GMCL) (Fig. 1A). In theONL, olfactory nerve fiber bundles were stained faintlyto moderately with PGP 9.5 antiserum. In the MCL, theimmature mitral and tufted cells were arranged in fourto five rows and stained faintly. The GMCL occupiedmost of the deeper part of the MOB and consisted of alarge number of immunonegative germinal cells and asmall number of immunopositive immature cells. How-ever, the AOB consisted of only two layers, VNL andMTL (Fig. 2A). In the VNL, vomeronasal nerve fiberbundles were stained moderately. In the MTL, manyimmature mitral/tufted cells were aggregated andstained faintly. In some PGP 9.5-immunopositive neu-rons of the MOB and AOB, the immunoreactivity wasdetected not only in the cytoplasm, but also the nuclei.

At E14, the bulbar size and the immunoreactivity forPGP 9.5 increased slightly. In the MCL of the MOB, asmall number of the immature mitral or tufted cellsbegan to extend their dendrites. In the AOB, no morpho-logical changes were observed.

At E15, the bulbar size was significantly larger thanthat observed at previous stages. In the MOB, a smallnumber of immunopositive tufted cells began to mi-grate into the EPL from the MCL, although most of theimmature tufted cells remained in the MCL. Most ofthe mitral and tufted cells extended their dendriticprocesses to converge to some loci of the EPL to form thepresumptive glomeruli (Fig. 1B). In the GMCL, some ofimmunopositive immature cells observed at E13 beganto extend their dendrites at this stage (Fig. 1C). Thesecells seemed to be short axon cells according to theirmorphological features. In the AOB, the mitral/tuftedcells became larger in size and some of them began toextend their dendritic processes. These cells werestained intensely with PGP 9.5 antiserum (Fig. 2B).The PGP 9.5-immunopositive short axon cells in theAOB were also observed at this stage.

At P0, the GL was distinguishable in the boundarybetween the ONL and the EPL in the MOB, since asmall number of the well-defined glomeluri appearedand were stained moderately with PGP 9.5 antiserum(Fig. 1D). In the AOB, no morphological changes wereobserved.

239PGP 9.5 IN HAMSTER OLFACTORY BULB

Fig. 1. PGP 9.5 immunohistochemistry in the developing mainolfactory bulb (MOB) fixed in Bouin’s solution without acetic acid atE13 (A), E15 (B, C), P0 (D), P3 (E), and P7 (F). A. PGP 9.5immunoreactivity was detected in the olfactory nerve layer (ONL),mitral or tufted cells in the mitral cell layer (MCL), and immature cells(arrowheads) in the germinal cell layer (GMCL). B. Immunopositivetufted cell (arrow) began to migrate into the external plexiform layer(EPL) from the MCL. The dendrites (arrowheads) of the mitral and

tufted cells converged to some loci of the EPL to form the presumptiveglomeruli (asterisk). C. PGP 9.5-immunopositive short axon cell(arrow) in the MOB. D. The well-defined glomeruli (asterisks) wereobserved in the glomerular layer (GL) of the MOB. E. The dendrites(arrowheads) of the mitral and tufted cells increased in the thicknessand immunoreactivity. F. Six laminae of the MOB were distinctlyrecognized. GCL: granule cell layer. IPL: internal plexiform layer.Scale bars 5 25 µm.

240 T. NAKAJIMA ET AL.

At P3, the dendritic processes of the mitral and tuftedcells in the MOB increased in the thickness and immu-noreactivity for PGP 9.5 (Fig. 1E). In the GL, a smallnumber of the immunonegative periglomerular cellsappeared around the glomeruli. The GCL of the MOBwas distinguishable from the GMCL at this stage.Granule cells were immunonegative for PGP 9.5. In theAOB, a small number of the well-defined glomeruliwere first observed at this stage (Fig. 2C). Theseglomeruli were stained moderately with PGP 9.5 antise-rum.

At P7, six laminae of the MOB were distinctlyrecognized (Fig. 1F). In the GL, the glomeruli increasedin number and were surrounded by a large number ofthe immunonegative periglomerular cells. In the MCL,the mitral cells were arranged in one or two rows. Mostof the tufted cells migrated into the EPL from the MCL.The EPL and GCL were enlarged. These morphological

features as well as PGP 9.5 immunoreactivity werealmost the same as those observed in the adult MOB.

At P15, the AOB displayed distinct five laminae sincethe pattern of the distribution of the mitral/tufted cellsin the MTL became the same as that observed in theadult (Fig. 2D). The pattern of the stainings for PGP 9.5in the olfactory bulb became almost the same as thatobserved in the adult at this stage.

Fixation in Acetic Alcohol

The immature cells in the superficial layer werealready stained faintly with PGP 9.5 antiserum at E11,and the immunostainings increased at E12 (Fig. 3A,B).PGP 9.5 immunoreactivity also was detected not only inthe cytoplasm, but also the nuclei in some immunoposi-tive neurons in the tissue fixed in acetic alcohol.

At E13, PGP 9.5 immunoreactivity was detected inthe same neuronal elements showing the immunoreac-

Fig. 2. PGP 9.5 immunohistochemistry in the developing accessoryolfactory bulb (AOB) fixed in Bouin’s solution without acetic acid atE13 (A), E15 (B), P3 (C), and P15 (D). A. PGP 9.5 immunoreactivitywas detected in the vomeronasal nerve layer (VNL) and mitral/tuftedcells in the mitral/tufted cell layer (MTL). B. The mitral/tufted cells

were stained intensely with PGP 9.5 antiserum. C. The well-definedglomeruli (asterisks) were observed in the glomerular layer (GL) of theAOB. D. The AOB displayed distinct five laminae. GCL: granule celllayer. LOT: lateral olfactory tract. Scale bars 5 25 µm in A, B, C, and100 µm in D.

241PGP 9.5 IN HAMSTER OLFACTORY BULB

tivity in the tissue fixed in Bouin’s solution withoutacetic acid, although the immunoreactivity was moreintense than that in the tissue fixed in Bouin’s solutionwithout acetic acid (Fig. 3C,D).

After E15, PGP 9.5 immunoreactivity was almostequal to that in the tissue fixed in Bouin’s solutionwithout acetic acid.

DISCUSSION

We demonstrated that PGP 9.5 immunoreactivitycould be detected in the axons of olfactory and vomero-nasal receptor neurons and immature mitral, tuftedand mitral/tufted cells from E13 in the tissue fixed inBouin’s solution without acetic acid, and in the imma-ture cells in the superficial layer from E11 in the tissuefixed in acetic alcohol. This time lag in immunostain-ings for PGP 9.5 may reflect that the epitopes of PGP

9.5 in the immature cells of early developmental olfac-tory bulb are susceptible to be hidden by fixation informaldehyde (Wilson et al., 1988).

As previous studies demonstrated that PGP 9.5 corre-sponds to carboxyl terminal of ubiquitin hydrolase,which plays an important role in the intracellularproteolysis (Hershko et al., 1992; Wilkinson et al.,1989), the present findings may suggest that the bulbarneurons acquired this metabolic pathway as soon asthey began to differentiate. Since the present resultsdemonstrated that the pattern of the immunostainingsfor PGP 9.5 became almost the same as that observed inthe adult at P7 in the MOB and at P15 in the AOB, thismetabolism may not reach the same level as that in theadult until P7 in the MOB and P15 in the AOB. In somebulbar neurons, PGP 9.5 immunoreactivity was de-tected not only in the cytoplasm, but also in the nuclei.

Fig. 3. PGP 9.5 immunohistochemistry in the developing olfactorybulb fixed in acetic alcohol at E11 (A), E12 (B), and E13 (C), and fixedin Bouin’s solution without acetic acid at E13 (D). PGP 9.5 immunore-activity was already detected in the immature cells in the superficiallayer from E11 in the developing olfactory bulb fixed in acetic alcohol.

PGP 9.5 immunoreactivity was more intense in the main and acces-sory olfactory bulb fixed in acetic alcohol (C) than those fixed in Bouin’ssolution without acetic acid (D) at E13. AOB: accessory olfactory bulb.MOB: main olfactory bulb. OB: olfactory bulb. V: ventricle. Scalebars 5 100 µm.

242 T. NAKAJIMA ET AL.

This finding was consistent with previous studies onvarious nervous systems (Johnson et al., 1994; Wilsonet al., 1988) and suggest that the proteolysis in whichubiquitin may be involved also occur in the nuclei(Johnson et al., 1994).

In the GMCL, PGP 9.5-immunopositive cells werealso detected at E13. These cells seemed to be imma-ture short axon cells according to their morphology anddistribution pattern observed at E15. The short axoncells are a kind of local interneuron of the olfactory bulband are thought to make synaptic contacts with theintrinsic neurons (mitral, tufted, and mitral/tuftedcells) and interneurons (periglomerular, granule, andshort axon cells) to control their excitability (Macridesand Davis, 1983; Mori, 1987). The short axon cells weredemonstrated to be generated postnatally in the ratolfactory bulb (Bayer, 1983), although at least a part ofthem already may be generated prenatally. Alterna-tively, this time lag of the generation of short axon cellsmay be attributable to the species difference.

Previously, PGP 9.5 immunoreactivity was demon-strated in the olfactory and vomeronasal receptor neu-rons not only in the adult, but also in the earlydevelopmental stage (Johnson et al., 1994; Kent andRowe, 1992; Schofield et al., 1995; Taniguchi et al.,1993), suggesting that this protein is a useful markerfor investigation of these two receptor neurons fromearly developmental stage.

These results were consistent with the present re-sults that PGP 9.5 immunoreactivity was detected inthe olfactory and vomeronasal nerve fiber bundles inthe olfactory bulb from early developmental stage. Inthe bulbar neurons, however, there has been no reporton a substance expressed predominantly from earlydevelopmental stage when these neurons begin todifferentiate. Although OMP is known to be a usefulmarker for the olfactory and the vomeronasal receptorneurons, the expression of this protein in the olfactorybulb is restricted to the ONL, VNL, and glomeruli of theGL (Margolis, 1985; Monti-Graziadei et al., 1980).However, PGP 9.5 immunoreactivity could be detectednot only in the olfactory and vomeronasal nerve fibersin the olfactory bulb, but also in several kinds of bulbarneurons from the early developmental stage as demon-strated in the present results, suggesting that PGP 9.5can be a useful marker to investigate the differentiationof bulbar neurons and the formation of laminar organi-zation in the developing olfactory bulb. Since PGP 9.5plays an important role in maintaining cell functions asdescribed above, it may be taken for granted that thisprotein must be produced earlier than several kinds ofneuropeptides (Matsutani et al., 1988).

In the present study, the extension of the dendrites,formation of the glomeruli, and completion of thelaminar organization were observed earlier in the MOBthan in the AOB. Our previous study demonstratedthat the hamster olfactory epithelium displayed almostthe same morphological features as those observed inthe adult at P1, although the vomeronasal sensoryepithelium retained some immature properties even atP10 (Taniguchi et al., 1982). The present and previousfindings suggest that the main olfactory system maymature earlier than the accessory olfactory system. As

for the dendrites, the mitral and tufted cells in theMOB began to extend their dendrites at E14 andformed presumptive glomeruli at E15 in the presentstudy. Similarly, the formation of the glomeruli wasreported to begin at E15 in the mouse MOB (Hinds andHinds, 1976). The mitral and tufted cells in the ratMOB were demonstrated to begin the synthesis of somekinds of neuropeptides from E16 (Matsutani et al.,1988). Since the olfactory information is transmittedfrom olfactory receptor cells in the olfactory epitheliumto the dendrites of intrinsic neurons within the glo-meruli through the ONL (Macrides and Davis, 1983;Mori, 1987; Scalia and Winans, 1975), these findingssuggest that the MOB may acquire the ability of theolfactory signal transduction to some extent beforebirth in the rodents.

In summary, the present study demonstrated thatPGP 9.5 is a useful marker to investigate the differentia-tion of bulbar neurons and development of laminarorganization. Since PGP 9.5 was demonstrated to corre-spond to carboxyl terminal of ubiquitin hydrolase, itmay play an important role in maintaining cell func-tions in the olfactory bulb.

LITERATURE CITED

Bayer, S.A. 1983 3H-thymidine-radiographic studies of neurogenesisin the rat olfactory bulb. Exp. Brain Res., 50:329–340.

Chen, S.T., K.A. von Bussman, L.J. Gray, and L.S. Jen 1994 Proteingene product 9.5-immunoreactive retinal neurons in normaldeveloping rats and rats with optic nerve or tract lesion. Dev.Brain Res., 78:265–272.

Doran, J.F., P. Jackson, P.A.M. Kynoch, and R.J. Thompson 1983Isolation of PGP 9.5, a new human neurone-specific proteindetected by high-resolution two-dimensional electrophoresis. J.Neurochem., 40:1542–1547.

Halpern, M. 1987 The organization and function of the vomeronasalsystem. Annu. Rev. Neurosci., 10:325–362.

Hershko, A., and A. Ciechanover 1992 The ubiquitin system for proteindegradation. Annu. Rev. Biochem., 61:761–807.

Hinds, J.W., and P.L. Hinds 1976 Synapse formation in the olfactorybulb. 1. Quantitative studies. J. Comp. Neurol., 169:15–40.

Johnson, E.W., P.M. Eller, and B.W. Jafek 1994 Protein gene product9.5 in the developing and mature rat vomeronasal organ. Dev.Brain Res., 78:259–264.

Kent, C., and H.L. Rowe 1992 The immunolocalisation of ubiquitincarboxyl-terminal hydrolase (PGP 9.5) in developing paraneuronsin the rat. Dev. Brain Res., 68:241–246.

Macrides, F., and B.J. Davis 1983 The olfactory bulb. In: ChemicalNeuroanatomy. P.C. Emson, ed. Raven Press, New York, pp391–426.

Margolis, F.L. 1985 Olfactory marker protein: from PAGE band tocDNA clone. Trends Neurosci., Dec. 542–546.

Matsutani, S., E. Senba, and M. Tohyama 1988 Neuropeptide- andneurotransmitter-related immunoreactivities in the developingrat olfactory bulb. J. Comp. Neurol., 272:331–342.

Monti-Graziadei, G.A., R.S. Stanley, and P.P.C. Graziadei 1980 Theolfactory marker protein in the olfactory system of the mouseduring development. Neuroscience, 5:1239–1252.

Mori, K. 1987 Membrane and synaptic properties of identified neuronsin the olfactory bulb. Prog. Neurobiol., 29:275–320.

Nakajima, T., M. Okamura, K. Ogawa, and K. Taniguchi 1996Immunohistochemical and enzyme histochemical characteristicsof short axon cells in the olfactory bulb of the golden hamster. J.Vet. Med. Sci., 58:903–908.

Scalia, F., and S.H. Winans 1975 The differential projections of theolfactory bulb and accessory olfactory bulb in mammals. J. Comp.Neurol., 161:31–56.

Schofield, J.N., I.N.M. Day, R.J. Thompson, and Y.H. Edwards 1995PGP 9.5, a ubiquitin C-terminal hydrolase; pattern of mRNA and

243PGP 9.5 IN HAMSTER OLFACTORY BULB

protein expression during neural development in the mouse. Dev.Brain Res., 85:229–238.

Takami, S., and P.P.C. Graziadei 1991 Light microscopic Golgi study ofmitral/tufted cells in the accessory olfactory bulb of the adult rat.J. Comp. Neurol., 31:65–83.

Takami, S., G.D. Fernandez, and P.P.C. Graziadei 1992 The morphol-ogy of GABA-immunoreactive neurons in the accessory olfactorybulb of rats. Brain Res., 588:317–323.

Taniguchi, K., H. Saito, M. Okamura, and K. Ogawa 1993 Immunohis-tochemical demonstration of protein gene product 9.5 (PGP 9.5) in theprimary olfactory system of the rat. Neurosci. Lett., 156:24–26.

Taniguchi, K., K. Taniguchi, and K. Mochizuki 1982 Comparativedevelopmental studies on the fine structure of the vomeronasal

sensory and the olfactory epithelia in the golden hamster. Jpn. J.Vet. Sci., 44:881–890.

Thompson, R.J., J.F. Doran, P. Jackson, A.P. Dhillon, and J. Rode 1983PGP 9.5-a new marker for vertebrate neurons and neuroendo-crine cells. Brain Res., 278:224–228.

Wilkinson, K.D., K. Lee, S. Deshpande, P. Duerksen-Hughes, J.M.Boss, and J. Pohl 1989 The neuron-specific protein PGP 9.5 is aubiquitin carboxyl-terminal hydrolase. Science, 246:670–673.

Wilson, P.O.G., P.C. Barber, Q.A. Hamid, B.F. Power, A.P. Dhillon, J.Rode, I.N.M. Day, R.J. Thompson, and J.M. Polak 1988 Theimmunolocalization of protein gene product 9.5 using rabbitpolyclonal and mouse monoclonal antibodies. Br. J. Exp. Path.,69:91–104.

244 T. NAKAJIMA ET AL.