Brain Research, 587 (1992) 203-210 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006.8993/92/$05,00
203
BRES 17955
Immunohistochemical markers in rat brain: colocalization of calretinin and calbindin-D28k with tyrosine hydroxylase * J o h n H. R o g e r s Department of Physiology, University of Cambridge, Cambridge, (UK) (Accepted I0 March 1992)
Key words: Calcium-binding protein; Dopamine: Substantia nigra; Hypothalamus; Olfactory bulb
Many dopaminergic cells of the substantia nigra are known to contain the calcium.binding proteins calretinin and calbindin-D28k. Catecholaminergic cell groups throughout the rat brain were therefore examined by two-colour immunofluorescence to determine whether they too contained these calcium-binding proteins as well as tyrosine hydroxylase (TH), Some TH + cell groups are mostly positive for both calretinin and calbindin, notably in the ventral tegmental area, the interfascicular nucleus, and parts of the substantia nigra. Other TH + cell groups in the midbrain, hindbrain and hypothalamus are very diverse; different cell groups are positive for calretinin, or calbindin, or both, or neither, In the olfactory bulb, entirely separate sets of periglomerular cells are positive for TH, calretinin and calbindin. However, there is considerable heterogeneity in calcium-binding protein expression within most cell groups, even in the substantia nigra. This could be a sign that calcium-binding proteins are regulated according to aspects of neuronal activity.
INTRODUCTION Calcium-binding proteins have recently emerged as distinctive markers for many neuronal populations. Being present throughout the cytosol, often at high concentrations, they are readily detectable by antibody staining in neurones which contain them. At least 3 such proteins - parvalbumin a..~,calbindin-D28 2.3 and calretinin 13,1"/.19,20_are present in extensive, but largely non-overlapping, sets of neurones in the rat brain. While parvalbumin is mainly present in fast-firing GABAergic neurones, calbindin and calretinin are present in diverse types of cells, ranging from local inhibitory neurones to sensory projection neurones. Dopaminergic and noradrencrgic neurones form far-reaching systems projecting throughout the brain ~. The cell bodies, which can be identified by staining with antibody for tyrosine hydroxylase (TH), are not numerous; the TH + cells lie in 15 cell groups scattered through the brainstem and hypothalamus ~,.m2.A sixteenth group consists of local circuit neurones in the olfactory bulb. The TH + cell groups are numbered A1 to A16 from caudal to rostral, and dopaminergic groups AI and A2 are accompanied by adrenergic cell groups C1, C2 and C3 (there is no A3) 1~,,2. Many of the cells
are very large in keeping with their long projections. The highest concentration of TH, and of dopaminergic cells, is in the ventral midbrain. In the ventral tegmental area, retrorubral field, and parts of substantia nigra compacta, most of the TH + cells contain calbindin 7. We have also found many calretinin-containing cells in these areas 17. The cells of the ventral tier of the substantia nigra compacta, projecting to the striatal 'patches', are calbindin-negative ( C B - ) " but calretinin-positive (CR + ) 17. While mapping the distributions of calretinin and calbindin, separately and jointly (ref. 17 and Rogers and R6sibois, submitted), we noticed that both proteins are present not only in many presumed dopaminergic cells of these ventral midbrain areas, but also in areas of the central grey, hypothalamus, and olfactory bulb that are known to be rich in dopaminergic cells. The present study was undertaken to determine how general is the colocalization of calretinin and calbindin with catecholamines. MATERIALS AND METHODS lmmunohistochemical procedures were as previously described i,,i,J. In brief, adult rats were perfused with 4%
Correspondence to: J.H. Rogers, Department of Physiology, University of Cambridge, Cambridge CB2 3EG (UK). * This paper is the second of two articles, the first of which appeared in Brain Research Vol. 587/1 on pp. 147-157.
204 paraformaldehyde, and their brains were cut into 50.pro sections which were processed free-floating. Some brains were cut sagittally and others transversely, Sections were incubated with a pair of primary antibodies for 2 days at 4 °. Secondary antibodies were fluoresce(n-labelled anti-rabbit 18 (Serotec) and Texas Red-labelled anti-mouse lg (Molecular Probes Inc.), each at 1:50 dilution. The primary antibody against TH was mouse monoclonal antibody 2 / 4 0 / 1 5 from Boehringer-Mannheim. Even though the manufacturer recommended that it should be used undiluted, we used it at 1 : 100 dilution and got very distinct staining of the recognised TH + cell groups, with no staining of other cells. Antibody against caketinin was a rabbit serum is.t9 used at 1:2,000 dilution; it does not cross-react with calbindin nor any other known antigen tf.ts.t,~. Antibody against calbindin was a rabbit serum kindly given by D.E.M. Lawson 14 used at I : 1,000 dilution. It sometimes shows weak crossreaction with calretinin ,,.t~.t,), but in the present experiments no such cross-reaction was evident, as the strongly C R + cells in the thalamic reticular nucleus and parts of the substantia nigra appeared C B - . Although immunoreactivity for calretinin or calbindin is usually distinct, some cells did show very weak immunoreactivity of doubtful significance; they were only a small minority of TH + cells and were omitted from the counts. For calretinin, each cell group listed in Table i was examined in 3 or 4 rat brains, and cell counts were done on at least one section from each brain. For calbindin, each cell group for which an entry is given in Table I was examined in 2 or 3 brains. In counting the proportion of TH + cells that were immunoreactire for calretinin (CR + ). all the TH + cells in selected areas were scored as C R + or C R - at the microscope, and the number of T H - C R + cells in the same area was estimated for comparison. These counts, summed over several sections, are given in Table I. in many areas th,:re were many T H - C R + cells, showing that the presence of " : H + C R - cells was not due to a deficiency in the calretinin sta~nin~ technique. The same applied to calbindin.
RESULTS
Almost all of the defined TH + cell groups were examined for calretinin, and most of them also for calbindin. Examples of the results are shown in the figures, in which each field is shown in both red (TH) and green (calretinin or calbindin). The numerical results are summarised in Table 1. Most TH + cell groups fell into 3 categories as regards calretinin expression: almost all CR + , or almost all CR - , or mixed CR + and CR - (often with some clustering of the differently stained cells). The same was true as regards calbindin expression. Within the 'mixed' category the percentage of CR + cells sometimes varied considerably between different sections. This was at least partly attributable to spatial heterogeneity, which was evident on single sections of these cell groups. As not all levels were examined, the percentages for 'mixed' cell groups in Table I may not apply to the whole cell groups.
Hindbrain CI (t,entral medulla). Large, intensely TH + cells with impressive neurites, scattered through the ventral medulla including the lateral reticular formation. The
TABLE I
i)rcq)orthm o/ Tl! + cells htlmum~rt, acti~,¢ for calrelinht or calhimlin R¢~ion
CI C2 C3 A7 A8 A~
All) Ai0 AIfl AI0c AI0dc
All)dr AII AI2 AI3 AI4 AI4d AISv AI6
Tit + CR = Ventral medulla Nuc. sol(tar(us PrH/mlf Pons Retro-rubral field Substantia nigra: pars compacta pars lateralis pars reticulata ventral tegmentalarea Interfascicular nuc. Supramammillary nuc. Linear nuc. of raph~ Dorsal raph~/ccntral grey: mod. -large TH + + cells rood, -small TH + cells Anterior midbrain Post,/vent, border of thalamus Hypo.: Arcuate nuc. Hyl~).: dorsal area and edge ofzona incerta Hypo.: medial area Hypo,: paraventricular nuc, i lypo,: supraoptic nuc, Olfactory bulb
TH + CR + (*)
CR + TH = (**)
TIt + CB=
+ +
31 6 0 27
56 52 44 33 60
22 (28~/~) 2 (4~) 3 (6%) I) (0~) 54 (47%)
146 I 17 125 54 44 13
374 (72r/F) 65 (98¢I,) 47 (73e,;-) 223 (64e~) 139 (72%) 83 (65~;,) 88 (87%)
(-) (=) + + (=) +
49 9 48 II 124
82 (63%) 88 (91~) 21 (31)%) 142 (93%) 7 (5~)
+ + + + +
150 66 82 140 120
8 (5%) !!9 (64%) I) (0%) 3 (2%) l) (0%)
+ + + + + + + +
+ + +('. (=) + + (~ +
210 2 31 62 15
TH + CB+
CB + TH =
42 (58%) 15 (71%) 30 (100%) 0 (0%)
+
+ + +~,' + + + + (~
86 ~; I 69 73
(-) (-) (-) (-) +
(28%) (93~,) (3~) (53%) (83¢'/~)
+ + +
+ ÷t~e + + th + + + + (¢~
24 42 40
75 71 100 33 30
44 (65%) 4 0
(9%) (0%)
7 (10%) 97 (58%) 7 (7%) 71 (68%) 0 (0%)
+ + + + + (o + +
+ + + + + + (-) + +
* in brackets, percentage of TH + cells which are CR +. * * Approx, numbers of CR + TH - cells, compared to number of TH + cells: + + + , many more: + +. similar number: +, less: (-). few; -, none; (-, generally smaller than the TH + cells.
205 proportion of TH + CR + and TH + CB + cells varied from 0% to > 60% between regions. The TH - CR + cells were heterogeneous in size but included cells as large as the TH + ones. C2 (nuc. tractus solitarii, medial edge) (Fig. 1). Large, strongly TH + cells, generally CR - CB + .
C3 (nuc. prepositus hypoglossi ~medial longitudinal fasciculus). Large, strongly TH + cells, all C R - except for a few on one section; all were CB + , but not strongly. (However, a striking, strongly CB + cell group in the same region was T H - , separate from group C3 though adjacent to it.) Thus the spectacular adrenergic cells of the medulla are mostly CR - but CB + . A6 (locus coeruleus). This nucleus is devoid of calretinin 17 and of calbindin 3.
A7 (pontine reticular formation, in and around nuclei of lateral lemniscus). Giant strongly TH + cells, all CR and C B - .
A8 (retrorubral field). Moderate to large TH + cells; TH + CR + and TH + CR - cells were distinct, and their ratio differed in different regions. On one brain (data not shown) we confirmed the finding 7 that most TH + cells in the retrorubral field are CB + .
Midbrain A9 (substantia nigra) (Fig. 2). In the pars compacta, 72% of the TH + cells were CR + but only 28% were CB + ; the latter were almost all TH + CR + CB + (Rogers and R~sibois, submitted). The TH + C R and TH + CB - cells tend to occur in clusters, and the cells along the ventrolateral edge, with prominent neurites projecting into the pars reticulata, are a distinct TH + CR + CB - population (the 'ventral tier' s). In the pars lateralis, virtually all the cells were TH + CR + CB + . In the pars reticulata, there were few labelled cells at all; the few TH + cells were mostly CR + CB - , but there was also a population of scattered TH - CR + cells. AIO (ventral tegmental area) (Fig. 3). In contrast to the intense and ubiquitous TH immunoreactivity in this area, the content of calcium-binding proteins was very heterogeneous. As in substantia nigra, most of the TH + cells were CR + and many were CB + ; but the calretinin immunoreactivity was generally weak - often so weak that the proportion of CR + cells was somewhat uncertain and variable. Some parts contained more strongly TH + CR + cells. The CB + cells were more distinct but tended to be weak in TH immunoreactivity. We have shown that CR + CB - , CR + CB + , and C R - CB + cells are all present in the ventral tegmental area, with some clustering (Rogers and R6sibois, submitted). AIO (interfascicular nuc.) (Fig. 4) In this nucleus of smaller, strongly TH + cells, most cells were TH + CR + CB + . A IO (supramammillary nucleus). Only sparse, moderate to small, weakly TH + cells were see,l in this nucleus which is packed with strongly CR + cells. Most of the TH + cells, on staining for calretinin, appeared like typical members of the CR + cell population, but at least one area of TH + CR - cells was seen. AIO caudal (linear nuc. of raphd). Moderate to large, strongly TH + cells; almost all CR + .
AIO dorsocaudal (dorsal raphd/central grey) (Fig. 5). The moderate to large, strongly TH + cells with great dendrites were mostly CR + but only weakly. The moderate to small, more weakly TH + cells, close to the aqueduct, were almost all CR + . (This region contrasted with analogous TH + cells in the central grey in the A6-rostral group (data not shown), where two sections showed the TH + cells to be CR - .) Fig. 1. Cell group C2 at medial edge of nucleus tractus solitarii, in
this and Figs. 2-5, the section is coronal, with lateral to the right, In all figures, each field is shown in red for tyrosine hydroxylase (TH) and in green for calretinin (CR) or calbindin (CB). Bar = 50 ~m. Nucleus tractus solitarii is visualized by diffuse fibrous staining for calretinin and calbindin. The large TH+ cells are all CR- and mostly CB +.
AIO dorsorostral (near !11 ventricle and around fasciculus retroflexus). Moderate to small-sized, weakly TH + cells. A variable proportion was CR + , sometimes weakly.
All (posterior and posterovemral border of thalamus) (Fig. 6A). This band of sparse but spectacular giant
206
Fig. 2. Substantia nigra. Bar - 50/~m. In the dorsal tier of pars compacta (d), TH + neurons are mostly CR+ and CB+, though rome are not. In the ventral tier (v), TH + neurons are mostly C'R + and C B - though some are CR -.: they have neurites projecting into the pars retieulata (r). (The pars rcficulata is full of CB + fibres from the striatum,)
strongly TH + cells is embedded in a band of numerous CR + cells, mostly much smaller, that forms part of a calretinin-rich 'shell around the thalamus' ~7 The band includes the subparafascicular nucleus and adjacent areas. Almost all the great TH + cells were CR + , some weakly but some with moderate intensity.
Hypothalamus A I2 (arcuate nucleus). This nucleus, with its beautiful cluster of TH + cells with well-stained neurites, has very few CR + cells although its neuropil is intensely CR + and there are many CB + cells. The TH + cells were all C R - and C B - except for a few on one section.
some groups were almost entirely TH + CR + while others included TH + CR - cells, Many of the TH + cells were also CB + , and one group (stained for calretinin and for calbindin on near-adjacent sections) must have been TH + CR + CB + . A14 dorsal (parat,emricular nuc.). The cells of this nucleus (round cells without prominent neurites) were weakly TH + but all CR - and mostly CB - .
AI5 t,entral (supraoptic nuc., including retrochiasmatic part). The majority of cells in this nucleus are also
AI3 (dorsal area arid edge of zona incerta) (Fig. 6C).
weakly TH + . In accordance with that the nucleus is mostly CB + , most cells were TH + CB + . The C B - cells were all T H - except section.
Distinct groups of strongly TH + and weakly TH + cells were almost all C R - and C B - , although surrounded by many CR + and CB + cells.
Olfactory bulb AI6 (periglomendar cells) (Fig. 7). Entirely separate
AI4 (medial area, inch~ding dorsomedial and t'entromedial nuclei) (Fig. 6B). In this large field of moderatesized, strongly TH + cells with well-stained neurites,
previous findings "~ it was found that minority of CR + for a few on one
sets of periglomerular cells were positive for TH, for calretinin, and for calbindin. The TH + cells were larger than the CR + and CB + cells.
207
Fig. 3. Ventral tegmental area. Bar - 50 ~m. Many TH + cells are CR + and/or CB + although the relative intensities vary.
DISCUSSION It is clear from these results that calretinin and calbindin are not universally present in TH + cells, in spite of the striking colocalization in the substantia
Fig. 4. lnterfascicular nucleus. The arrow marks the midline, pointing dorsally. Dark 'holes' are blood vessels. Bar = 50 /~m. Most TH + cells are CR + and CB + although the relative intensities vary.
nigra and some other cell groups. In fact, the major conclusion is that many TH + cell groups show great heterogeneity in CaBP content. The cell groups mostly fall into 3 categories according to the proportion of TH + cells labelled for calretinin (and the same is true for caibindin): the proportion is < 10%, or 25-75%, or > 90%. In the first and last categories, the few minority neurones could be aberrant cells or outlying members of neighbouring populations. In the middle category, the spatial distri. bution of cells with and without the calcium-binding protein was often seen to be uneven, varying from one section to another and also within a single section. Since many of these groups were not surveyed from end to end, the percentages in the middle category in Table 1 should not be taken as absolute; they serve mainly to illustrate the heterogeneity of these cell groups, In the ventral tegmental area, there is known to be heterogeneity among the TH + cells in terms of firing rate, presence of autoreceptors, and anatomical projections t,4,5,22,2.~. It remains to be seen whether the content of calcium-binding proteins correlates with any of these features. If it does, the patchy distribution of the cells with calcium-binding proteins may serve to map the distribution of different physiological types.
208
Fig. 5. Cell group AI0-dorsocaudal in the central grey. The arrow in the ventricle marks the midline, pointing dorsally. Bar = 50 p.m. A: most of the Ttt + cells are CR +. including the very large one. B: most of the TH + cells are CB +. (The ependyrna is also CB + .)
Throughout the midbrain (substantia nigra, ventral tegmental area and retrorubral field), Gerfen et al. H showed that the TH + CB + cells were those that projected to the striatal matrix. The TH + C B - cells in the substantia nigra (in the ventral tier and in the reticulata), which project to the striatal patches (strio. somes), were selectively sensitive to 6-hydroxydopamine toxicity K and likewise disappear first in Parkinson's disease .~4. However, as these cells contain the related protein calretinin, it would he premature to
infer any relationship between calcium-buffering capacity and cell survival. The olfactory bulb provides an opposite case, where none of the TH + cells contain calretinin nor calbindin. These results are consistent with earlier studies ,),~0,~s which found that the TH + periglomerular cells (estimated as = 8 - 2 2 % of the total) were separate from the CB + and OABAergic cells, The calbindin-immunoreaetive cells were estimated as 26% of the total f' but this probably included the CR + cells
Fig, 6, Thalamus and hypothalamus (TH/calrctinin labelling only), Bar = 50 p.m, A: cell group A I I, in the hand on the ventro-posterior edge of the thalamus, Para.~agittal section. The large TH + cells are CR +, B: cell group AI4, in the medial hylx)thalamus, Parasagittal section, Most of the: T i t + culls arc CR +, hut faintly, C: cell group AI3, in the dorsal hyp~thalarnus, Coronal section..All the TH + cells are C R - .
209
Fig 7. Glomeruli of the olfactory bulb. Bar = 50 u.m, Many periglomerular cells are labelled, the CR + and CB + ones being small. All the TH + cells are C R - and C B - ,
as these authors' antibody also detected the CR + CB - granule cells. Our sections show similar numbers of CR + and CB + periglomerular cells and fewer TH + cells (Fig. 7). in the olfactory bulb, as in retina and cortex t,.tN..,o(see also accompanying paper), these studies are revealing numerous subpopulations of local inhibitory cells with little or no overlap. The heterogeneous patterns of calcium-binding protein content in other TH + cell groups may indicate that they too consist of several definite cell types. But even in the substantia nigra compacta, where most cells are so strikingly TH + CR + , a minority of cells is clearly TH + CR - . Like studies of cortex (see accompanying paper) and other regions (Rogers and Rdsibois, submitted), these results suggest that calcium-binding proteins may not be invariant components of definite cell types, but may be regulated according to aspects of neuronal activity. Acknowledl~enaents. I am grateful to Elizabeth tlousden and Eduardo Torres for technical assistance, to Dr. Eric Lawson for the calhindin antiserum, to Dr. James Fawcett for the use of equipment, and to Dr. Peter Lewis fi,' comments on the manuscript, This study was supported by a grant from the Wellcome Trust.
ABBREVIATIONS CR+ CB+ TH+ CR-
calretinin-immunoreact ive calbindin-D28k-immunoreactive tyrosine hydroxylase-immunoreactive (etc.). not immunoreactive
REFERENCES I Bj6rklund, A. and Lindvall, O,, Dopamine-containing systems in the CNS. In A, Bj6rkhmd and T. lt6kfell (Eds.), lhmdbook o.f C'hemical Nem'oanatomy, Vol,2,1, Elsevier. Amsterdam, 1984, pp. 277-379. 2 Braun. K., CaBPs in ~lvian and wnalnmaliall ccntt'~ll nervous sysrein: localiz~itiot'i,development, ~lnd possilde ftlnctions, Pro,~, lllsmchem, Cvtochem,, 21 (1990) 1~o2, 3 Cello, M,R,, ('nlhindin.D2Hk and parvalburnin in the rat nct~,ous systcnt, Nem'osch, twe,, 35 (19~}0) 375=475, 4 Chiodo, L,A,, Banr.m, M,J,, Grace, A,A,, Roth, R,II. ~md lhmo n~y, B,S,, Evidenc~ l'orthe absence of inlr~|llsc-[Icncratin[lsomatodcndritic and syntl'tcsis.rrtoduhttirtg w'lc~c terminal receptors on suhpopuhitiorts of nlcsocorticitl dopanlinc ilgUrOllS, Nem'o,wiem'e, 12 (1~)144) I=II"D, 5 Clark, D,, Enherg, G,, Peleldart, ['.,, Svcrtriort, T.I I, arid ('llrlssort, A,, An electrophysiological analysis of the actions o1' the 3-PPP enantiomers on the nigrostriatal dopamine system. NamO'n Schmiedeber~'s Arch, Pharmaco/.. 329 (1985) 344-354, 6 Garcia-Segura, L,M., Battens, D., Roth, J,, Norman. A,W. and Orci, L.. Immunohistochemical mapping of CaBP-IR in the rat central nervous system, Brain Re,s',, 296 (1984) 75-86, 7 Gcrl'cn, C,R,, Bainll'~ridgc, K,G. and Millet', J,J,, The ncostriatal mosaic: cornpartmental distribution ol' ('aBP and parvalhumin in the bas'.d ganglia of the rat and monkey, Prec. Nail, Acad. Sci, USA, ~2 (19SS) S7S0-871~4, 8 Gerl'en, C,R., Baimhridge, K,G, and Thibault, J,, The neostriatal rues,de. Ill. Biochemical and developmental dissect:ilion of patch-matrix mesostriatal systems, J. Nem'o.~ci., 7 (1987) 3¢}353944. 9 Halasz, N,, H6kfelt. T,, Ljungdahl, A., Johansson, O. and Goldstein. M., Dopaminc neurons in the olfactory bulb. Adz', 8ioc'/wm. P,s,ychop/zarmacoh, 16 (1976) 169-177. l0 l l:dasz. N,. H6kfclt, T,, Norman, A.W. :md Goldstein. M.. Tyrosine hydroxylase and 2Sk vitamin D-dependent CaBP arc localizcd in different subpopulations of periglomcrular cells of the rat olfactory bulb, Neumsci. Left,. 61 (1985) 1(13-107. I I H6kfelt. T., Johansson, O. and Goldstein, M.. Chemical anatomy of the brain. Sciem'e, 225 (1984) 1326-1334.
210 12 H6kfelt, T., Martensson, R,, Bj6rklund, A., Kleinau, S. and Goldstein, M., Distributional maps of tyrosine hydroxylase-immunoreactive neurons in the rat brain, in A. Bj6rklund and T. H6kfel! (Eds.), Handbook of Chemical Neuroanatomy. Vol.2.1, Elsevier, Amsterdam, 1984, pp, 55-115. 13 Jacobowitz, D.M. and Winsk'y, L., lmmunocytochemical localization of calrelinin in the forebrain of the rat, J, Comp. Neurol., 304 (1991) 198-218. 14 Jande, S,S., Maler, L. and Lawson, D.E.M., lmmunohistochemical mapping of vitamin D-dependent CaBP in brain, Nature, 294 (1981) 765-767. 15 Mugnaini, E., Oertel. W.H. and Wouterlood, F.F., Immunocytochemical localization of GABA neurons and dopamine neurons in the rat main and accessory olfactory bulbs. Neurosci. Lett., 47 (1984) 221-226. 16 Pasteels, B., Rogers. J., Blachier. F. and Pochet. R., Calbindin and calretinin localization in retina from different species, I,qs. Neurosci., 5 (1990) I - 16. 17 RSsibois. A. and Rogers. J.H., Calretinin in rat brain: an immunohistochemical study, N,,uroscience, 46 ( 1991 ) 101-134.
18 Rogers, J.H., Two calcium-binding proteins mark many chick sensory neurons, Neuroscience, 31 (1989) 697-709. 19 Rogers, J.H., lmmunoreactivity for calretinin and other calciumbinding proteins in cerebellum, Neuroscience, 31 (1989) 711-721. 20 Rogers. J.H., Calretinin, In C.W. tteizmann (Ed.), Not'el Calcium-Binding Proteins, Springer, Heidelberg, 1991, pp. 251-276. 21 Rogers, J.H., Khan, M. and Ellis, J.E., Calretinin and other CaBPs in the nervous system. In R. Pochet, D.E.M. Lawson and C.W. Heizmann (Eds.), Calcium-Binding Proteins in Normal and Transformed Cells (Adt'. Exp. Med. Biol. 269), Plenum, New York, 1990, pp.195-203. 22 Shepard, P.D. and German, D.C., A subpopulation of mesocortical dopaminergic neurons possess autoreceptors, Eur. J. Pharmacol., 98 (1984) 455-456. 23 White, F.J. and Wang, R.Y., AI0 dopamine neurons: role of autoreceptors in determining firing rate and sensitivity to dopamine agonists, Life Sci., 34 (1984), 1161-1170. 24 Yamada et al,, Relative sparing in Parkinson's disease of substantia nigra dopamine neurons containing calbindin-D,sK, Brain Res., 526 (1990) 303-307.