0306-4522/92 $5.00 + 0.00 Pergamon Press Ltd © 1992 IBRO

Neuroscience Vol. 49, No. 2, pp. 347-364, 1992

Printed in Great Britain

DISSOCIATED HIGH-PURITY DOPAMINERGIC N E U R O N CULTURES FROM THE SUBSTANTIA NIGRA A N D THE VENTRAL TEGMENTAL AREA OF THE POSTNATAL RAT S. MASUKO,* S. NAKAJIMAt and Y. NAKAJIMA~ Department of Biological Sciences, Purdue University, West Lafayette, IN 47907, U.S.A. Department of Pharmacology and Department of Anatomy and Cell Biology, University of Illinois, College of Medicine at Chicago, Chicago, IL 60612, U.S.A. Abstract---We have developed dissociated primary neuronal cultures obtained from the substantia nigra

and from the ventral tegmental area of postnatal rats (two to three days old). After making brain slices, the regions of the substantia nigra and the ventral tegmental area were separately dissected. The removed fragments of brain tissue were dissociated and cultured on a glial feeder layer. Double immunocytochemica1 labeling for tyrosine hydroxylase and GABA on cultures grown for two to three weeks showed the presence of 42% dopaminergic and 39% GABAergic neurons in substantia nigra cultures, whereas in ventral tegmental area cultures there were 65% dopaminergic and 21% GABAergic neurons. The dopaminergic neurons were characterized by thick and straight primary processes dividing into several branches. Varicosities were found mainly on distal parts of the processes. In contrast, GABAergic neurons possessed highly branched thick and thin primary processes with intensive arborization and numerous varicosities. Co-existence of dopamine and cholecystokinin was found in about 70% of dopaminergic neurons from the substantia nigra and in about 35% of dopaminergic neurons from the ventral tegmental area. Physiological properties of these cultured dopaminergic neurons were investigated with the whole-cell version of the patch-clamp method. After each physiological experiment, immunocytochemical labeling confirmed that the cell was dopaminergic. Properties of single action potentials, with an action potential height of 92 mV and duration of 1.6 ms, were similar to those reported for dopaminergic neurons in brain slices. The neurons showed a high resting potential, and no spontaneous firing of action potentials. Constant current depolarizations elicited trains of action potentials. In the majority of cells, the train stopped firing within a few seconds, while in some cells it lasted indefinitely. When the cell was hyperpolarized, the voltage response started to decline slowly (sag), indicating the presence of hyperpolarization-activated currents (time-dependent inward rectification). These results show that by using our culture method it is possible to obtain separate dissociated cultures of the substantia nigra and the ventral tegmental area from newborn rats. Because they are rich in functional dopaminergic neurons, these cultures will be a useful tool for studying various properties of dopaminergic neurons.

Dopaminergic (DA) neurons play important roles in brain functions. Their dysfunction leads to neurological and psychiatric disorders such as Parkinson's disease and schizophrenia. 32'72'76 Because of their functional importance, extensive studies have been

*Present address: Department of Anatomy, Saga Medical School, Nabeshima, Saga, 840-01, Japan. tPresent address: Department of Pharmacology, University of Illinois, College of Medicine at Chicago, Chicago, IL 60612, U.S.A. ~To whom correspondence should be addressed at: Department of Anatomy and Cell Biology (m/c 512), University of Illinois, College of Medicine at Chicago, 808 South Wood Street, Chicago, IL 60612, U.S.A. Abbreviations: CCK, cholecystokinin; DA, dopaminergic; DAB, 3,3'-diaminobenzidine tetrahydrochloride; EDTA, ethylenediaminetetraacetic acid; EGTA, ethyleneglycolbis(aminoethylether)tetra-acetate; GTP, guanosine 5'triphosphate; HEPES, N-[2-hydroxyethyl]piperazineN'-[2-ethanesulfonic acid]; IN, interpeduncular nucleus; PBS, phosphate-buffered saline; PIPES, piperazineN,N'-bis[2-ethanesulfonic acid]; TH, tyrosine hydroxylase; VTA, ventral tegmental area.

done on the anatomical connections and the physiological functions of D A neurons. In the mesencephalon there are two major brain nuclei containing D A neurons: the substantia nigra and the ventral tegmental area (VTA). These two groups of D A neurons differ in afferent and efferent connections. D A neurons in the substantia nigra are the source of the nigrostriatal D A system, and play a role in the control of movement. 1'32'79On the other hand, D A neurons in the V T A constitute the main source of the mesolimbic D A system, and they are involved in more complex functions such as emotion, reward, cognition or m o t i v a t i o n ] '71,73'8~ Recently, experiments on brain preparations in vivo and on brain slice preparations have contributed to our knowledge of the cell physiological aspects of D A neurons. The present paper describes the properties of dissociated cultured neurons that we developed from the substantia nigra and the VTA. Three objectives were addressed in this study. The first was to obtain cell cultures separately from the substantia nigra and the VTA. The systemic functions of these two nuclei are

347

348

S. MASUKOet al.

very different, but there have been no reports of culturing these two nuclei separately. The second was to develop functionally healthy cell cultures rich in D A neurons from the brain; this will serve as a powerful tool with which to investigate cellular and molecular mechanisms of D A cell physiology and pathology. The third was to determine electrophysiological properties of identified (by immunocytochemistry) D A neurons. Previously there have been several reports on organotypic cultures of m e s e n c e p h a l i c D A neurons. 9'35'59'66'78'80 However, to solve many problems concerning membrane channels, transmitter receptors, and signal transduction, dissociated cell cultures are often more suitable than organotypic cultures. Several investigators sa3ag'~'4~'s~'6~'7° succeeded in culturing embryonic brain D A neurons. However, because these cultures were made from the mesencephalon, not from restricted areas of the substantia nigra or VTA, the percentage population of the D A neurons seemed to be quite low, except for Ort et al., 58 who reported a 10-30% yield of D A neurons. Recently, we have devised a new method of culturing brain nuclei. A unique feature of our method is first to make brain slices of the region containing the nucleus in question, and then to take out the nucleus under the dissecting microscope. In this way, we have previously succeeded in making dissociated cultures of noradrenergic neurons from the locus coeruleus 49 and cholinergic neurons from basal forebrain nuclei. 54 We have now applied this method to culturing D A neurons from the substantia nigra and from the VTA. The present study describes the culture method and some of the properties of the cultured neurons derived separately from the substantia nigra and the VTA. Specifically, morphological characteristics of these neurons are demonstrated using phasecontrast optics and immunocytochemistry. We will describe how these neurons survive during the culture period. D a t a of physiological experiments using the patch-clamp method on D A neurons that are immunocytochemically confirmed are also presented. These cultures contain a high percentage of D A neurons (42% in the substantia nigra cultures and 65% in the V T A cultures). These D A neurons are functionally alive and produce action potentials similar to those obtained in slice or in vivo preparations. An abstract 48 describing some of these results has appeared previously. During the past few years, several abstracts and a paper on dissociated cultures of postnatal brain D A neurons have been published 4,47,61 63,68 EXPERIMENTAL PROCEDURES Cell culture

Separate cell cultures of the substantia nigra and the VTA were prepared using the method previously described49,53,54 with several modifications. Two- to three-day-old (postnatal) Long-Evans rats (Charles River Breeding Labs, Wilmington, MA) were used. Usually six to 16 newborn rats

were used for each culture series. Under ether anesthesia the brainstem, including the mesencephalon, was removed aseptically, and immersed in an oxygenated ice-cold balanced salt solution consisting of 130mM NaC1, 4.5 mM KC1, 2 mM CaC12, 33 mM glucose, and 5 mM piperazine-N,N'bis[2-ethanesulfonic acid] (PIPES) buffer (pH 7.4). The brainstem was then embedded in 2.5% agar in the balanced salt solution. Using a Vibratome (Lancer 1000), we made transverse slices (400/~m in thickness) of the embedded brainstem in the oxygenated balanced salt solution. Under a dissecting microscope the regions of substantia nigra and VTA in the brain slices were identified and separately dissected out by using hypodermic needles (Fig. IA, B). The tissue fragments of the substantia nigra and the VTA were treated for 30rain at 37°C in a papain solution, consisting of 20 units/ml papain (Pharmacia, Piscataway, NJ or Cooper Biomedical, Malverin, PA), 116 mM NaC1, 5.4 mM KC1, 26 mM NaHCOs, 1 mM NaH2PO4, 1.5 mM CaCI2, 1 mM MgSO4, 25mM glucose, 0.5mM EDTA and 1 mM cysteine. 34 The tissue fragments were then washed with culture media and were dissociated by gentle trituration using a fire-polished Pasteur pipette. The culture medium contained a modified (see below) minimum essential medium with Earle's salt (Gibco) (95%), heat-inactivated rat serum (5%) (prepared in our laboratory), L-ascorbic acid (10#g/ml), penicillin (50 units/ml), and streptomycin (50#g/ml). The minimum essential medium with Earle's salt (Gibco, Grand Island, NY, Cat. No. 330-1430) was modified by adding the following extra-ingredients: e-glutamine (0.292mg/ml), NaHCO 3 (3.7mg/ml), and D-glucose (5 mg/ml, the final concentration being 6 mg/ml). The cells were plated at a density of 2 5 x 10 4 cells/cm2 in a 1.2-cm diameter well which was made at the center of a Petri dish; the bottom of the well was made of an Aclar ltuorohalocarbon film (Allied Fibers and Plastics, Morristown, NJS°). The surface of the well had previously been coated with collagen and a feeder layer: the feeder layer consisted of nonneuronal cells, mostly astroglial cells obtained from brainstems of newborn rats. 49 The cultures were kept at 37°C in 10% COj90% air with saturated humidity. Three days after plating, we treated the cultures with 5-fluoro-2'-deoxyuridine (15/~g/ml) and uridine (35/zg/ml) for two days to suppress proliferation of non-neuronal cells. The culture medium was changed every five to seven days. Immunocytochemistry

Either indirect fluorescence immunocytochemistry or the peroxidase-antiperoxidase method of Sternberger TM was used. We used the following antibodies: (i) mouse monoclonal antibody to neurofilaments (Labsystems Inc., Chicago, IL) diluted 1:20; (ii) mouse monoclonal antibody to tyrosine hydroxylase (TH) (a gift from Dr H. Hatanaka, Protein Research Institute, Osaka University28) diluted 1:500; (iii) rabbit antiserum to cholecystokinin-8 (CCK) (Immuno Nuclear Corp., Stillwater, MN) diluted 1:1000; (iv) rabbit antiserum to GABA (Immuno Nuclear Corp., Stillwater, MN) diluted 1:3000; and (v) sheep antiserum to glutamic acid decarboxylase (a gift from Dr I. J. Kopin and Dr W. H. Oertel, Laboratory of Clinical Science, NIMH 55,56) diluted l : 1000 or 1:1200. Cultures were briefly washed with a warm (37°C) N-[2-hydroxyethyl] piperazine-N'-[2-ethanesulfonic acid] (HEPES)buffered Krebs solution (see below for the composition) and fixed for 0.5-3 h at room temperature or for 2 4 h at 4°C with one of the following fixation solutions: (i) a solution containing 2% of paraformaldehyde and 15% of saturated aqueous picric acid in 0.12M sodium phosphate buffer (pH 7.4); (ii) a solution containing 0. 1% glutaraldehyde and 4% paraformaldehyde in 0.12 M sodium phosphate buffer (pH 7.4) (used for immunocytochemistry of GABA); and (iii) 4% paraformaldehyde in 0.12M sodium phosphate

Cultures of substantia nigra and ventral tegmental area

349

Fig. 1. Transverse Vibratome slices of the mesencephalon of newborn rats. (A) A fresh, unfixed slice (400/~m thick). (B) The same slice as in A after demarcating the substantia nigra (SN) and the VTA by cutting with hypodermic needles (arrowheads). (C, E) Slices (100#m thick) fixed and treated with TH immunocytochemistry. (D) High-magnification pictures of a part of C, indicated by the symbol *, showing TH-immunoreactive cells in the substantia nigra (pars compacta) and the VTA. (F) Highmagnification picture of a part of E, indicated by the symbol*, representing the pars reticulata of the substantia nigra. RN, red nucleus; IN, interpeduncular nucleus. Scale bars in A ~ and E = 1 mm; scale bars in D and F = 50/~m.

buffer (pH 7.4) (used for immunocytochemistry of glutamic acid decarboxylase). After fixation the cultures were washed with a phosphate-buffered saline solution (PBS). In experiments involving CCK or glutamic acid decarboxylase immunocytochemistry, prior to the fixation, the cultures were treated overnight with 10-6M colchicine to accumulate the antigens in the cell bodies. (The colchicine treatment was not used in other kinds of immunocytochemistry.) When the solution containing glutaraldehyde and paraformaldehyde (for GABA immunocytochemistry) was used, the preparations were cryo-protected by treatment with a PBS containing 30% sucrose for 8-15 h at 4°C. Then the preparations were frozen on a metal block cooled in liquid nitrogen for 5 10 s, and were thawed by immersion into the same solution. This procedure permeabilized the cells and aided the penetration of the antibodies. After thawing, the preparations were treated with 1 M ethanolamine for 1 h at room temperature. The cultures were pre-incubated for 0.5-1 h at room temperature with a PBS containing 10% normal serum (taken from the same species of animal from which the secondary antibody was obtained). In addition, sometimes 0.05% Triton X-100 was included in the pre-incubation solution. The cultures were then treated generally for 1 3 h at room temperature with the primary antiserum (the

primary antiserum was dissolved in PBS, which also contained 1% normal serum again taken from the same species of animal from which the secondary antibody was obtained). Controls were treated with non-immune serum or with the same solution as above except for the absence of the primary antibodies. For fluorescence immunocytochemistry against neurofilaments or TH, cultures treated with the primary antibody were washed three times with PBS. Then they were incubated for 1-2 h at room temperature with the rhodaminelabeled secondary antibody (goat antibody against mouse IgG). For CCK or GABA immunoreactions, the cultures were incubated with fluorescein-labeled goat antibody •against rabbit IgG, and for the detection of glutamic acid decarboxylase, they were incubated with fluorescein-labeled rabbit antibody against sheep IgG. All the secondary antibodies were afffinity-purified anti-IgGs (heavy and light chain) (Kirkegaard and Perry Laboratories Inc, Gaithersburg, MD), and they were used at a dilution of 1: 80 or 1: 40. For peroxidase-antiperoxidase processing,TM after treatment with the primary and secondary antibodies, cultures were rinsed three times with PBS and incubated with a 1:100 dilution of mouse, rabbit or goat peroxidase-antiperoxidase complex (Sternberger-Meyer, Jarrettsville, MD) for 0.5 1 h at room temperature. They were then rinsed with PBS and

350

S. MASUKO et al.

treated for 15-20 min with a 0.05 M Tris buffer (pH 7.6) containing 0.02% 3-3'-diaminobenzidine tetrahydrochloride (DAB) and 0.005% HzO 2. The specimens were examined using an epifluorescence Zeiss microscope equipped with a rhodamine filter set (exciter BP 546/12 nm, beam splitter FT 580 nm and barrier LP 590 nm, Zeiss 487715) and a fluorescein filter set (exciter BP 485/20 nm, beam splitter FT 510 nm and barrier LP 520-560nm, Zeiss 787717). Some cultures were double-labeled with a combination of the mouse monoclonal antibody to TH and the rabbit antiserum to GABA, or double-labeled with the TH antibody and the antiserum to CCK. The background immunocytochemical reaction was hardly detectable with TH- and GABA-immunocytochemistry. Only a slight background label was observed with CCK- and glutamic acid decarboxylase-immunocytochemistry. In the neuron survival experiments, both TH and neurofilament immunoreactivities were simultaneously examined using the sequential immunoperoxidase technique. 75 After fixation, these cultures were first treated with the mouse monoclonal antibody to TH, and the immunoreaction was visualized with DAB precipitation by COC12,33 which resulted in black deposits. After several rinses with PBS, the same preparations were incubated with the mouse monoclonal antibody to neurofilaments and were treated with DAB, resulting in light-brown deposits. With this procedure, catecholaminergic neurons appeared black and noncatecholaminergic neurons showed a light-brown color. Electrophysiology The techniques used were similar to those previously described. 49,54 The whole-cell patch-clamp method was used. 26 During the physiological experiments, the culture was continuously superfused with a modified oxygenated Krebs solution containing 148 mM NaC1, 2.5 mM KC1, 2.4mM CaC12, 1.3mM MgC12, 5 m M HEPES NaOH buffer and 11 mM D-glucose (pH 7.4). The intrapipette solution was 120 mM potassium aspartate, 40 mM NaC1, 5 mM HEPES-KOH buffer, 0.5 mM EGTA-KOH, 0.25mM CaC12, 3 m M MgC12, 2 m M Na2ATP , 0.1mM Na3GTP (Na3GTP was omitted in early stages of the experiments) and ~ 8 mM KOH (pH 7.2). The free calcium ion concentration [Ca 2+] in this pipette solution is 160nM; this was computed with a program provided by Dr R. E. Godt. TMThis [Ca2+] is within the range of the resting intracellular [Ca2÷] for cultured mammalian brain neurons reported by Connor et al., 8 and Connor and Tseng. 7 We have included 40 mM Na in the patch pipette solution to conform with the values of the intracellular [Na] of mammalian nerve fibers measured by Krnjevi643 (~40mmol/kg water), or those reviewed by Lipton and Whittingham45 (32-45 mM Na in brain tissues). However, the real value of cytosolic [Na +] in brain neurons is unknown. The values of membrane potentials were corrected for a 9-mV liquid junction potential between the intrapipette solution and the external solution. Drugs were applied by pressure ejection from micropipettes. The electrical data were stored on an FM tape recorder. For data analysis the tape was played back at a slower speed to a chart recorder (sometimes after the data were temporarily stored in the combination of a pulse-code modulator and a video cassette recorder). The overall frequency responses of the records were measured by sinusoidal analysis, and the frequency at 3 db down was given in the figure and table legends. We investigated the effect of the frequency response on the action potential of a typical DA neuron~ A low pass RC filter with a frequency response of 3 kHz ( - 3 db) reduced the height of an action potential by ~ 2 % compared with that recorded with a 10 kHz ( - 3 db) filter, whereas a low pass filter of 1.5 kHz ( - 3 db) reduced it by -~5%. There were no detectable differences in the duration of the action potentials under these three conditions.

The bath temperature near the neuron was kept at 33 + 2°C. RESULTS Isolation o f the substantia nigra and the ventral tegmental area W h e n the b r a i n o f a n e w b o r n rat was sectioned serially at 400/~m thickness, most of the mesenc e p h a l o n was included in a single slice, as illustrated by a n unstained slice in Fig. 1A. In order to locate the D A cells, we p e r f o r m e d T H - i m m u n o c y t o c h e m i s t r y o n 100-#m-thick b r a i n slices. All catecholaminergic (DA, noradrenergic a n d adrenergic) n e u r o n s are THimmunoreactive. However, Decavel et al. 12 showed t h a t the substantia nigra a n d the V T A contain m a n y D A n e u r o n s using m o n o c l o n a l a n t i b o d y against dopamine. F u r t h e r m o r e , n o noradrenergic neurons were detected in the substantia nigra by using antib o d y to norepinephrine, iv or in the mesencephalon by using a n t i b o d y to dopamine-beta-hydroxylase. 27 Also no adrenergic n e u r o n s were f o u n d in the m e s e n c e p h a l o n by using a n t i b o d y to phenyle t h a n o l a m i n e - N - m e t h y l t r a n s f e r a s e ? 1 Thus, we used T H - i m m u n o r e a c t i v i t y as a m a r k e r for the existence of D A neurons. O n e - h u n d r e d - m i c r o m e t e r slices treated for THi m m u n o c y t o c h e m i s t r y are shown in Fig. 1C, E. Figures 1C a n d E, respectively, c o r r e s p o n d roughly to the most rostral a n d m o s t caudal level o f the b r a i n slice in Fig. 1A, the 4 0 0 - # m unstained brain slice. The distribution of T H - i m m u n o r e a c t i v e neurons in these n e w b o r n rat slices (Fig. 1) is similar to t h a t o f D A neurons described in adult rats. 1°'12'56 A dense distribution of T H - i m m u n o r e a c t i v e cell bodies was observed in the V T A a n d in the pars c o m p a c t a of the substantia nigra (Fig. 1C, D, E). Some T H - i m m u n o reactive n e u r o n s were also f o u n d in the pars reticulata of the s u b s t a n t i a nigra (Fig. 1E, F). Figure 1A, in c o m p a r i s o n with Fig. 1C a n d E, shows t h a t in a freshly sectioned b r a i n slice, the s u b s t a n t i a nigra can be identified as a dark, ovoid region in the ventrolateral part of the mesencephalon. The V T A can be identified as a s o m e w h a t translucent region b o r d e r e d dorsally by the red nucleus, laterally by the substantia nigra, a n d ventrally by the interped u n c u l a r nucleus (IN), a triangular region in the ventro-medial p a r t in Fig. 1A. The same slice as in Fig. 1A is s h o w n in Fig. 1B, after we demarcated these nuclei by cutting the borderlines with hypodermic needles (arrowheads). Once we n o t e d the location o f these l a n d m a r k s o n the slices by using TH-imm u n o h i s t o c h e m i s t r y , we were then able to locate the s u b s t a n t i a nigra a n d the V T A on unstained b r a i n slices. Culture A viability test using T r y p a n Blue exclusion was conducted after dissociation of the substantia nigra a n d the V T A from the b r a i n slices. The test indicated

Cultures of substantia nigra and ventral tegmental area

351

Fig. 2. Peroxidase-antiperoxidase immunoreaction against neurofilaments and TH. Dissociated neurons from the substantia nigra (A) and the VTA (B), grown on a feeder layer for 24 h. Cultures were treated with the sequential double-immunoperoxidase technique33. Arrows indicate TH-immunoreactive cells and arrowheads indicate neurofilament-immunoreactive cells (non-catecholaminergic neurons). Scale bars = 50 pm. that usually more than 90% of the cells were alive. The total cell yield from the substantia nigra or the VTA was 6-10 × 104 per rat (substantia nigra, 8.2_+ 0.5 × 104; VTA, 7.7 + 0.5 × 104; mean_+ S.E.M., from 11 different dissociations comprising 134 rats in total). These dissociated cells are composed of neuronal cells and non-neuronal cells such as glial cells. At this dissociation stage we did not attempt to distinguish neuronal cells immunocytochemically. The dissociated cells settled down on the feeder layer and began extending processes within 24 h (Fig. 2). During subsequent days of incubation the processes increased in length and number, and after 10 days of culture, many cells showed their characteristic shapes (see later).

Neuronal survival Three different culture series were examined. In each culture series, neurons were inoculated at the same density into five culture dishes. Twenty-four hours after the inoculation, and thereafter weekly, one culture dish per each culture series was fixed and received the sequential treatment for TH and neurofilament immunoreactivities. Immunocytochemical reaction for neurofilaments was used to identify neurons in culture. All cells that extended long processes showed positive immunoreaction to neurofilaments. On the other hand, the flat background cells did not show positive immunoreactivity to neurofilaments, indicating that they are probably glia cells of the feeder layer (Fig. 2). For each of these cultures, the numbers of DA (TH-immunoreactive) and non-DA (neurofilamentimmunoreactive, non-TH-immunoreactive) neurons in five fields were counted (1.1 x 0.9 mm for each field). The fields, which were uniformly spaced across the culture area, represented 4.4% of the total culture area of a dish. The number of total surviving neurons in these fields at 24 h after the inoculation will be called "the initial neuronal number". Subsequently,

the weekly neuronal counts (total surviving neurons and DA neurons) are expressed as a percentage of the initial neuronal number. The initial neuronal number of either substantia nigra or VTA cultures was 0.5 0.9 × 104 cells/cm2 (substantia nigra, 0 . 6 + 0 . 0 9 × 104; VTA, 0.8+ 0.06 x 104; mean + S.E.M., n = 3). The difference between this initial neuronal number and the numbers of cells plated (see above) may be due to the following reasons: (1) many glial cells existed in dissociated plated cells and (2) many dissociated cells would have died during the first day of culture. When the cultures were one week old, the number of surviving neurons sharply decreased to a level of 50% of the initial neuronal number, as shown in Fig. 3. Thereafter, there was a slower decline in the neuronal population. We did not systematically follow the survival for more than one month, but we observed that neurons were surviving in culture for over two months. One day after plating, about 30% of the initial total neuronal number were THimmunoreactive in the substantia nigra cultures, whereas about 50% were TH-immunoreactive in the VTA cultures (solid symbols in Fig. 3A, B). Seven days after plating, the number of TH-immunoreactive neurons decreased in both the substantia nigra and the VTA cultures, and subsequently their numbers continued to decline slightly. At 28 days in culture, TH-immunoreactive neurons constituted 38% of the total neuronal population in the substantia nigra cultures, whereas in the VTA cultures 57% were TH-immunoreactive (Fig. 3).

Immunocytochemical classification With the double immunofluorescence staining for TH and GABA, we found that there are three major classes of neurons: (i) T H ( + ) / G A B A ( - ) (DA cells) (Fig. 4A-I); (ii) T H ( - ) / G A B A ( + ) (GABAergic cells) (Fig. 5A-I) and (iii) T H ( - ) / G A B A ( - ) (nonDA, non-GABA cells) (Fig. 6A-C). When the double

S. MASUKOet al.

352 (~) A

1oo~ 80

Cs)

B

loo- ( ~ V T A

SN

80-

~otal

60-

(~-,

_

o

4O

. . . . . . . .

(smO

20-

. . . . . . . .

(59s)

(~

-

" "

-

C5~)

I

ol ............ 1

(~) 7

® ........

!3.,.! 14

.

.

.

,~ . . . . . . . .

.

.

21

.

.

.

•

.

Doy~

28

1

7

14

21

Days

28

Fig. 3. Neuronal survival of substantia nigra and VTA cultures. All values are expressed as percentages of the initial neuronal number, counted at 24 h of culture (average of three different cultures. See the text for the value of the initial neuronal number). Open circles represent total neurons and closed circles indicate DA (TH-immunoreactive) neurons. The values in parentheses indicate the percentage of DA neurons in reference to the total neuronal population at each culture stage. Total, total neurons; TH, TH-immunoreactive neurons.

immunofluorescence method for TH and glutamic acid decarboxylase (instead of TH and GABA) was used in colchicine-treated cultures, the result was similar to that of the double imnmnofluorescence study for TH and GABA (data not shown). The shapes of all these neurons can be classified into bipolar, tripolar and multipolar as shown in Figs 4-6.

Tyrosine hydroxylase( + ) / G A B A ( - ) (dopaminergic neurons). TH-immunoreactive neurons (DA neurons) in substantia nigra and VTA cultures grown for two to three weeks are shown in Fig. 4. THimmunoreactive neurons (DA neurons) were bipolar, tripolar (pyramidal) or multipolar in shape, the tripolar neuron being predominant in both the substantia nigra and the VTA cultures. Typical DA neurons have two to five thick and rather straight primary processes which extend from their cell bodies and give rise to relatively sparse arborization (Fig. 4). Relatively large varicosities are found mainly in the distal segment of the processes, but are rare in the proximal part. TH-immunoreactive fluorescence was clearly found in the perikarya, in the thick processes and in the varicosities along the neurites; however, the fluorescence was not prominent in the thin processes. GABA-immunoreactive varicosities (arrowheads), whose cell bodies are outside the field of view, surround the TH-immunoreactive cell bodies and neurites in Fig. 4C, I. The TH-immunoreactive cells show only a faint immunoreactivity to GABA antibody, indicating that the background immunoreaction is negligibly small compared with the positive reaction.

Tyrosine hydroxylase ( - )/GABA (+) (GABAergic neurons). GABA-immunoreactive neurons can also be categorized into three classes: bipolar, tripolar and multipolar (Fig. 5). In contrast to typical DA neurons, not only thick processes but also fine processes frequently originate directly from the cell

bodies of GABAergic neurons. These processes are highly branched and extensive arborization is particularly seen in thin processes. Thin processes and some thick processes have many varicosities. Occasionally, thin, varicose processes form fine plexiform networks as shown by arrows in Fig. 5D, F, G, I. Unlike DA neurons, GABAergic neurons tend to reveal fluorescence throughout the neuron, including the fine processes with small varicosities.

Tyrosine hydroxylase (-)/GABA ( - ) (non-dopaminergic, non-GABA neurons). About 20% of the cultured substantia nigra neurons and 15% of the VTA neurons did not show immunoreactivity to either TH or GABA (Table 1; Fig. 6A-C). The shape of this class of neurons was varied, and the size was somewhat smaller than other classes of neurons in both substantia nigra and VTA cultures (Table 1).

Tyrosine hydroxylase( + )/GABA( + ) (dopaminergic-GABA neurons). A small number of neurons ( 1 5 % of cultured substantia nigra and VTA neurons) were immunoreactive to both TH and GABA (Table 1; Fig. 6D-F). The co-existence of immunoreactivity for TH and GABA (or glutamic acid decarboxylase) was reported by Kosaka et al. 42 in some neurons of the olfactory bulb, retina, diencephalon, mesencephalic central gray, and cerebral cortex of adult rats. It was also reported by Schimchowitsch et al. 65 in tubero-infundibular and tubero-hypophyseal DA neurons. However, such co-existence has not been reported in the in viva substantia nigra and VTA of adult rats. 42'56 Occurrence and size. The frequencies and sizes of TH- and GABA-immunoreactive neurons from four different culture series are summarized in Table 1. The substantia nigra cultures yielded almost the same number of TH-immunoreactive neurons (42%) as GABA-immunoreactive cells (39%). However, the VTA cultures yielded more TH-immunoreactive cells (65%) than GABA-immunoreactive cells (21%).

Cultures of substantia nigra and ventral tegmental area

Fig. 4. Examples of cultured DA neurons. Phase-contrast micrographs (A, D, G) and immunofluorescence micrographs for TH (B, E, H) and GABA (C, F, I). (A42) TH-Immunoreactive bipolar and tripolar neurons from the substantia nigra cultured for 14 days. (D-F) TH-immunoreactive bipolar and tripolar neurons from the VTA cultured for 18 days. Arrows in E indicate a TH-immunoreactive varicose neurite from a neighboring neuron (out of picture). (G-I) A TH-immunoreactive multipolar neuron from the VTA cultured for 14 days. In C and I (arrowheads) GABA-immunoreactive varicosities surround the TH-immunoreactive cell bodies and neurites. TH, TH-immunoreactive; GABA, GABA-immunoreactive. Scale bars = 50 #m.

353

354

S, MASUKOet al,

Fig. 5. Examples of cultured GABAergic neurons. Phase-coatrast micrographs (A, D, G) and immunofluorescence micrographs for TH (B, E, H) and for GABA (C, F, I). (A~S) A GABA-immunoreactive bipolar neuron and a TH-immunoreactive neuron from the VTA culture grown for 14 days. (D-F) A GABA-immunoreactive tripolar neuron from the substantia nigra cultured for 18 days. ( G q ) A GABA-immunoreactive multipolar neuron from the substantia nigra cultured for 14 days. The neuron has two thick and two thin processes originated from the cell body. Numerous varicosities are seen along neurites of GABA-immunoreactive neurons. Arrows point to networks of GABA-immunoreactive fine varicose processes (D, F, G, I). TH, TH-immunoreactive; GABA, GABA-immunoreactive. Scale bars = 50 #m.

Cultures of substantia nigra and ventral tegmental area

355

e

Fig. 6. Phase-contrast micrographs (A, D, G) and immunofluorescence micrographs for TH (B, E, H), GABA (C, F) and CCK (I). (A-C) A neuron from the VTA, not immunoreactive to either TH or GABA, cultured for 14 days. ( D F ) A neuron from the substantia nigra, immunoreactive for both TH and GABA, cultured for 18 days. (G-I) A colchicine treated culture of the VTA showing two neurons which are immunoreactive for both TH and CCK (double arrows), one neuron which is immunoreactive for only TH (arrow) and one neuron which is not immunoreactive to either TH or CCK (arrowheads). TH, TH-immunoreactive; GABA, GABA-immunoreactive; CCK, CCK-immunoreactive. Scale bars = 50 #m.

@a

S. MASUKOet al.

356

Table 1. Frequency and size of tyrosine hydroxylase- and GABA-immunoreactive cells in substantia nigra and ventral tegmental area cultures Substantia nigra

VTA

Immunoreactivity

Frequency (%)

Diameter* (# m)

n

Frequency %

Diameter* (# m)

n

TH(+)/GABA(-) TH(-)/GABA(+) TH(-)/GABA(-) TH(+)/GABA(+)

40.3 37.0 20.8 2.0

19.0 _+0.18 17.9 _+0.23 16.3 _ 0.34 21.4 ± 1.5

99 91 51 5

64.4 20.3 14.7 0.6

18.7 ± 0.23 17.6 ± 0.38 15.0 _ 0.46 19.4

114 36 26 1

All values presented are means ± S.E.M. Measurements were conducted on 246 and 177 neuronal cells from four different cultures of the substantia nigra and the VTA, respectively. Cells were grown for two to three weeks. *Diameter was defined as the geometrical mean of the major and minor soma diameters.

Co-existence of dopamine and cholecystokinin In another series of experiments, the co-existence of immunoreactivity for C C K and T H was examined (Table 2, Fig. 6G-I). These cultures were two to three weeks old and treated with colchicine. A substantial number of CCK-immunoreactive neurons existed in both the substantia nigra and the VTA. The data of Table 2 show that almost all of CCK-positive neurons (over 93% in substantia nigra and over 83% in V T A cultures) were also immunoreactive to TH, indicating that C C K is mostly contained in D A neurons. The data also show that in substantia nigra cultures grown for 18-21 days, two-thirds of D A neurons contain CCK, whereas in V T A cultures of the same age, one-third of D A neurons contain CCK. This result indicates that different subpopulations of D A neurons exist in both the substantia nigra and the VTA.

Identification of dopaminergic neurons using phasecontrast optics It is desirable to be able to identify the neuronal type with a phase-contrast microscope. As shown in Table 1, D A neurons are slightly larger than G A B A ergic neurons both in the substantia nigra (the difference is highly significant, P < 0.01) and in the VTA (the difference is significant, P < 0.05). However, this size difference is not large enough to differentiate D A neurons from GABAergic neurons. Hence, we have to rely on the structural features described above. We tested our ability to identify D A and GABAergic neurons with phase-contrast optics. The test showed that although it was easy to identify the cells with typical structural characteristics, it was almost

impossible to do so in the case of many other neurons. Therefore, we felt it was necessary to do immunocytochemical identification after electrophysiological experiments.

Electrophysiological properties In the first stage of investigation we obtained electrophysiological data of eight cells without regard to the identity of their cell type. In the second stage, we accumulated electrophysiological data from 28 cells; in these cells we tried to establish their identity by performing immunocytochemistry on the same cells after physiological experiments. Out of 28 cells, 11 cells were immunocytochemically identified as D A neurons, three cells as non-DA, n o n - G A B A neurons (Table 3), and the rest of the cells (including all the cells which we suspected to be GABAergic) deteriorated during the fixation process following the physiological experiments. In the following descriptions of electrophysiological results, " D A neurons" always mean the cells that were immunocytochemically identified after physiological experiments.

Action potentials elicited by long-duration currents Examples of action potentials recorded from D A neurons elicited by a long-lasting step current using the constant current mode of the patch-clamp amplifier are shown in Fig. 7. Various attributes of the action potential are summarized in Table 3. In in vivo or brain slice preparations, the rising phase of action potentials recorded from substantia nigra neurons consists of two phases, 2°'24'4° probably corresponding to the initial segment and soma-dendritic spikes of spinal motoneurons. 15 In our cultured

Table 2. Co-existence of tyrosine hydroxylase and cholecystokinin immunoreactivity in cultured substantia nigra and ventral tegmental area neurons Substantia nigra Cultures (days) 14 18 21 21

VTA

TH(+) CCK(+) (%)

TH(+) CCK(-) (%)

TH(-) CCK(+) (%)

Total neurons counted

TH(+) CCK(+) (%)

TH(+) CCK(-) (%)

TH(-) CCK(+) (%)

Total neurons counted

47 70 78 71

51 28 16 26

2 2 6 3

57 87 49 94

42 35 29 31

57 64 65 65

1 1 6 4

72 72 124 164

Values presented are per cent of total neurons immunoreactive to TH and/or CCK. Four different cultures were examined.

Cultures of substantia nigra and ventral tegmental area

357

Table 3. Electrophysiological characteristics of cultured neurons from the substantia nigra and the ventral tegmental area determined by the whole-cell version of patch-clamp Non-DA Non-GABA n=3

DA n=ll Soma diameter (#m) Resting potential (mV) Input resistance (Mf~) Threshold (mV) Action potential height (mV) Overshoot (mV) Action potential duration (ms)

20 _+ 0.55 - 7 2 + 1.9 192 _+40* - 3 9 _+ 0.94 92 + 3.1 20 +_ 1.7 1.6 _+0.11

22.0 _+ 1.0 - 7 1 + 0.88 204 _+ 35** - 3 9 _+4.6 86 + 7.5 16 +_ 7.6 1.8 _+ 0.35

All values are mean _+_S.E.M. The input resistance was not measured for all cells in the sample. *n = 9, soma diameter = 20 Ftm; **n = 2, soma diameter = 22 pro. Cells were cultured for 13-17 days. Action potentials were elicited by a step current. The duration of action potentials was measured at threshold. Frequency response of the records was 3 kHz ( - 3 db). In DA cells, two cells were obtained from the VTA and nine cells from the substantia nigra. Although the small number of the cells from the VTA precludes a definite conclusion, there were no marked differences in the properties of action potentials between the DA neurons from the VTA and those from the substantia nigra. Soma diameter of the non-DA, non-GABA neurons in this sample is larger than the diameter of T H ( - - ) / G A B A ( - ) cells in Table 1, indicating that the non-DA, non-GABA cells used for physiology may not represent an unbiased sample of the T H ( - ) / G A B A ( - ) cells.

neurons, as s h o w n in Figure 7B, the differentiated record of the action potential (dV/dt) indicated t h a t the rising phase has two c o m p o n e n t s . These two c o m p o n e n t s were seen in seven cells o u t of 11 D A cells. As in the case of spinal m o t o n e u r o n s , 15 the first c o m p o n e n t m a y come f r o m near the origin of large neurite(s) a n d the second from the m a i n part of the soma a n d other dendrites.

Action potentials elicited by short pulses A c t i o n potentials evoked in a D A n e u r o n by short current pulses are displayed in Fig. 8A1, 2. In A1 the resting m e m b r a n e potential was - 7 0 m V , a n d a n action potential was evoked by a short current pulse. In A2 the m e m b r a n e potential was changed to - 6 1 m V by passing a small, steady o u t w a r d current before the action potential was evoked. Record A3 shows action potentials from the same neuron, evoked by a long-lasting step depolarization, the

resting potential at this time being - 7 1 mV. It is seen t h a t the o v e r s h o o t in A2 is higher a n d the d u r a t i o n of the action potential longer t h a n those in A1. In A3 b o t h o f these quantities are slightly larger t h a n in A2. This p h e n o m e n o n was observed in all the D A n e u r o n s tested. The average change of o v e r s h o o t by varying the m e m b r a n e potential within the range of - 7 9 to - 6 1 m V was 8.5 +_ 1.5 m V (mean _+ S.E.M., n = 5) per 10 m V of the m e m b r a n e potential change in short pulse experiments. A n action potential evoked by a short pulse was a c c o m p a n i e d by a n after-potential, its polarity a n d a p p a r e n t m a g n i t u d e being d e p e n d e n t o n the preset potential. In the example o f Fig. 8, at the resting m e m b r a n e potential of - 7 0 mV, the spike was followed by a n afterdepolarization (B1). O n the other h a n d , at - 6 1 m V a n afterhyperpolarization occurred (B2). The m a g n i t u d e of afterhyperpolarization (the preset potential m i n u s the level of the peak of

20 mV

21111V/s

10 m s 5 ms

Fig. 7. (A) An action potential elicited in a DA neuron cultured for 14 days from the substantia nigra. (B) An action potential from another DA neuron cultured for 14 days from the substantia nigra. The lower trace is a differentiated record (d V/dt) of the action potential. The cells were depolarized by a step current (202 pA in A, and 100 pA in B). Arrows indicate the zero potential levels. In all records (Figs 7 10), the whole-cell version of the patch-clamp under the constant current was used. The overall frequency response of the records was 3 kHz ( - 3 db).

358

S. MASUKO e t al.

A2

A1

20mV

2hA !

A3

10 m s

J

f

I0 ms

Fig. 8. (A1, 2) Action potentials by a short, large current pulse, from a DA neuron cultured for 13 days from the substantia nigra. The resting potential in A1 was - 7 0 mV; the overshoot was 8 mV. In A2 the potential was depolarized to - 6 1 mV by an externally applied d.c. The action potential elicited by a short pulse is now larger (overshoot, 18 mV) and of longer duration. (This is not entirely caused by an IR drop by the d.c.; this factor would account for less than 1 mV of the difference in the value of the overshoot). (A3) The same neuron; the resting potential was - 7 1 mV at this time. The action potential was elicited by a small, long-lasting step-current. The overshoot was 22 mV. Arrows indicate the zero potential levels. (B1, 2) The same records as in A1, 2, respectively, at a higher gain and a slower time-base to show afterdepolarization (B1) and afterhyperpolarization (B2). The overall frequency response in A1-3 was 6 kHz ( - 3 db), and that of B1 and 2 was 1.5 kHz ( - 3 db). afterhyperpolarization) was 3.4 _+ 0.5 mV (mean _+ S.E.M., n = 5), when a single spike was elicited from a preset potential of 62.4 _+ 0.7 mV. The duration of the hyperpolarization was 64 _ 17 ms.

Spike trains We did not find spontaneous firing of action potentials in our substantia nigra or V T A neurons. This is because the resting potential of our cultured neurons was quite high under our patch-clamp condition. When a D A neuron was depolarized by a step current, a train of repetitive spikes occurred (Fig. 9). The potential pattern during the repetitive spikes was almost the same as those reported for the substantia nigra neurons in in vivo or slice preparations, as shown in Fig. 9A1.21'44'46Each rapid downstroke of an action potential shifted into an afterhyperpolarization, which rebounded into a slow depolarizing creep until the threshold was again reached. However, as shown in Fig. 9A2, 3, 4, the maximum duration of the train elicited with varying intensities of current steps was not long in the majority of

the cells. Out of 27 neurons (11 D A neurons and 16 unidentified neurons), in which we tried to elicit a train of spikes by d.c. depolarization, 17 cells (nine D A neurons) responded with a train of spikes lasting less than 3 s. In the rest of the cells (10 out of the 27 neurons; two out of the 11 D A neurons) we could elicit trains of spikes lasting more than 50 s. One example of a D A neuron is shown in Fig. 9B. In this cell a relatively regular train was elicited for as long as the current continued (about 1 min).

Hyperpolarization-activated currents In D A neurons of in vivo or slice preparations, long hyperpolarizing currents produce a decline (sag) in voltage response with a relatively slow time-course (on the order of 10-100 ms). z°'4°'44Analysis of the sag in other types of neurons 25,51 indicated that it represents an increase in permeability toward N a and K ions. This phenomenon is called hyperpolarizationactivated currents (H-currents), 51,82 F-currents, 3 Q-currents, 25 or is sometimes called time-dependent

Cultures of substantia nigra and ventral tegmental area

359

A1

L loins

A2

A8

_2 --I

A4

/

--"[

~lll II Ill III III I m ]iJiJiJlllLItlll

llllllll

i

_

omv iI

ll[lllliLill

IiB'

'

I

l l l l i i H¢II I II111 i i l l l l i l i ¢ i

i l l i ¢ I I l l i l i l l IIH l i l l l He

Fig. 9. (A1-4) Trains of action potentials elicited by long-lasting depolarizing currents in a DA neuron from the substantia nigra cultured for 14 days. A1 shows a potential pattern during the train. A2-4 are slow time-base records. The action potentials of this cell adapted quickly. (B) Another DA neuron cultured from the VTA for 15 days. In this cell the steady current produced an indefinitely lasting train of spikes. The gap in the record was 60 s. Frequency response orAl 4 was 6 kHz ( - 3 db), and that of B was 1.5 kHz ( - 3 db). Arrows indicate the zero potential levels. Rayport et al. 62) are derived from the whole or the ventral part of the mesencephalon, and are therefore a mixture of the substantia nigra, the V T A and the neighboring regionsfl '11''3'3°'41'5s'6°Our cultures, on the other hand, were obtained separately from welldefined regions of the substantia nigra and the VTA. Thus, it is possible to investigate the difference in the characteristics between neurons from the substantia nigra and those from the VTA. The D A neuron cultures previously reported were obtained from embryo nic brains, 5,11,13,30,41,s8,60whereas cultured neurons in Cardozo, 4 L o p e z - L o z a n o et al., 47 Rayport et al., 62 Shen et al., 68 and our present study were obtained from postnatal animals. This difference could be important for some types of

inward rectification. In our cultured neurons, hyperpolarization-activated currents were observed in six out of nine D A cells tested (Fig. 10).

DISCUSSION

Culture

We have developed dissociated neuronal cultures separately from the substantia nigra and the VTA. Our cultures differ from previously reported D A neuron cultures from the mesencephalon in the following aspects. All of the previously reported D A neuron cultures (except for a preliminary report on V T A cultures by

A

B

_U--J2OmV 3 400pA i

|

100 ms Fig. 10. (A)-A small hyperpolarizing current produced a voltage response without a sag. (B) A large hyperpolarization produced a sag in the potential record, suggesting an occurrence of the H-current (hyperpolarization-activated current) in a DA neuron from the substantia nigra cultured for 14 days. Resting potential was - 73 mV. Overall frequency response was 6 kHz (-- 3 db).

360

S. MASUKOet al.

research such as experiments involving receptors, since certain neuroreceptors are known to develop postnatally. 14'16'19'64Another advantage of using postnatal animals is that the substantia nig'ra and the VTA are more developed and better defined, allowing us to dissect out the substantia nigra and the VTA precisely. Highly enriched DA neuron cultures are desirable for conducting various experiments. Several studies on enriched DA neuron cultures have been reported as follows, Di Porzio et al. 13 obtained an enrichment of DA neurons derived from embryonic mouse mesencephalon by using a cell-sorting technique combined with an immunoreaction to neural specific protein 4. However, even after the enrichment the population of DA neurons was only about 100 out of 10,000 neurons (1%). Rayport et al. 62 obtained enriched DA neuron cultures by cell fractionation and obtained a fraction containing 35 % DA neurons. Lope~Lozano et al. 47 conducted cell-sorting of a population of mesencephalon cells which were retrogradely fluorescence-labeled from the striatum. By culturing these sorted cells, they obtained cultures containing 93-95% DA neurons. Our cultures are very rich in DA neurons without any special enrichment procedures; 42% of the neurons in the substantia nigra and 65% of the neurons in the VTA cultures were dopaminergic. Characteristic f e a t u r e s o f d o p a m i n e r g i c neurons a n d GABAergic neurons

DA neurons and GABAergic neurons are two major types of neurons in our substantia nigra and VTA cultures. This is in general agreement with the results on the in rico mesencephalon by Oertel et al., 56 who showed that DA neurons and GABAergic neurons are the two principal neuronal types in the substantia nigra, and that DA neurons are seen in dense clusters in the VTA. The abundance of GABAergic neurons in our substantia nigra cultures agrees with the report by Okada et al., 57 which states that the substantia nigra contains the highest amount of GABA in the mammalian brain. In our cultures, the average size of DA neurons is only slightly larger than that of GABAergic neurons (Table 1). Thus, under phase-contrast optics it is impossible to distinguish DA neurons from other neurons merely based on their size. Nevertheless, as already explained, DA neurons and GABAergic neurons differ somewhat in the distribution and the structure of dendrites (Figs 4, 5). Our immunocytochemically identified DA neurons have similar structures to those of the DA neurons in the substantia nigra previously reported in the slice preparations after recording. 22'4°'77 In these reports, Grace and Onn 22 labeled cells by intracellular injection of Lucifer Yellow followed by TH-like immunoreaction, and Kita et al. 4° and Tepper et al., 77 who, although they did not perform immunocytochemistry, labeled cells by intracellular horseradish-peroxi-

dase-injection after physiological recording. Our cells also resemble "medium-sized" neurons in the pars compacta of the rat substantia nigra investigated by Juraska et al. 36 with the Golgi impregnation method. It was also noted that our immunocytochemically identified GABAergic neurons are similar to pars reticulata neurons (most of which could be GABAergic) investigated by Grofova et al., 23 using in vivo materials. The above-mentioned morphological resemblance between cultured neurons and neurons in vivo is particularly interesting since it suggests that brain neurons, after losing their processes by dissociation, regain their original shapes even under the cell culture condition. The co-existence of DA and CCK in in vivo materials was investigated in detail by Seroogy et al. 67 By using the double-immunocytochemical method, they reported a high incidence of co-existence of TH and CCK immunoreactivity in the pars compacta of the substantia nigra (80 90% in the rostral level, 70% in the intermediate level and 30 50% in the caudal level). They also found that there was a moderate level of co-localization of TH and CCK immunoreactivity in other areas of the ventral mesencephalon including the VTA (50 70%). These results agree with our present data indicating that the coexistence of DA and CCK was found more frequently in cultured DA neurons from the substantia nigra than in those cultured from the VTA. Non-dopaminergic, non-GABA

cells

A small proportion of our cultured neurons were non-DA, non-GABA cells. Until now, these kinds of cells have not been described in the substantia nigra or in the VTA regions. Electrophysiology

Many reports have been published on electrophysiological properties of mesencephalic DA neurons (either morphologically confirmed or unconfirmed) in in vivo preparations, in slice preparations, and under culture c o n d i t i o n s . 4'5'2°'22'4°'44'46'52'58'61'63'69'7° Our cultured DA neurons are functionally alive and healthy, and produce action potentials comparable to those reported previously. The value of the spike overshoot (20 mV) of our DA neurons is consistent with those of slice preparations ( ~ l l ~ 6 m V from figures in Grace and Onn, 22 or 12-22 mV in Mereu et a1.52). The threshold value ( - 3 9 mV) of our neurons is also within the range given by previous investigators (Grace and O n n , 22 - - 3 6 mV; Mereu et al., 52 - 4 8 m V ) . The input resistance of our sample (190 Mf~) is comparable to the values obtained in slice preparations by Grace and Onn22 (168 M~), Silva and Bunney69 (173M~), and Kita et al. 4° (200 M~). The resting potential of our cultured DA neurons ( - 7 2 m V ) is high in comparison with the data of previous reports, and these neurons did not exhibit spontaneous firing of spike potentials. One possible

Cultures of substantia nigra and ventral tegmental area explanation for this discrepancy is that our cultured neurons are not as heavily innervated as those in slice preparations or in in vivo brain. In the latter case, the cells, being thickly innervated, could be constantly bombarded by slow excitatory transmitters. However, not much emphasis should be placed on the difference in resting potential; the whole-cell patchclamp technique is not suited for measuring normal resting potentials since the cytosol is perfused with an artificial solution. It has been emphasized that one of the characteristics of DA neuron action potentials is a prominent afterhyperpolarization.21'22 In our DA cells, the amplitude of afterhyperpolarization elicited by a single spike was 3.4 mV, when membrane potential was set at - 62 mV. This small amplitude is probably due to the high membrane potential. In in vivo preparations, Grace and Bunney21 reported that the size of afterhyperpolarization elicited by a single spike was 2 mV (from Fig. 9 of their paper, but membrane potential was not given); presumably the resting potential was quite high under their recording conditions. Our value of the duration of action potentials (1.6 ms) is the same as that reported by Lacey et al. 44 (1.6 ms) in slice, but is shorter than the data from Grace and Bunney2° (2.75 ms), Grace and Onn 22 (2.7 ms), or Mereu et al. 52 (2.5-5 ms). At least part of this discrepancy can be ascribed to the fact that the overshoot and the duration of action potentials are, under certain circumstances, dependent on the membrane potentials just before the spike is elicited. Thus, as explained in relation to Fig. 8, the action potential elicited from the membrane potential around - 60 mV had a larger overshoot value and a longer duration than the one elicited from a - 7 0 - m V membrane potential. Belluzzi et al. 2 observed the same phenomenon in sympathetic neurons and ascribed it to the ability of the fast inactivating K-channel (A-current6) to prevent the spike from reaching its full height. However, the above interpretation may be only a partial explanation for the briefness of action potential. Indeed, our culture conditions may not be ideal for the development of some components of ionic

361

channel mechanisms. It may be that in our cultured cells the voltage-gated Ca channels,46 which would prolong the spike, are not as fully developed as in in vivo counterparts. The DA neurons in slice preparation are known to produce spontaneous spikes, and their firing frequency is increased by externally applied depolarizing currents. 69 In our cultured neurons (DA and unidentified neurons) steady depolarizing currents produced trains of action potentials, but in the majority of cells the duration of the train was less than 3 s. Only in about one-third of the cells (and only in two out of 11 DA neurons) were we able to invoke indefinitely-lasting trains of spikes by d.c. application. This may be another manifestation of the quantitative difference in the ionic mechanisms between our cultured DA neurons and the neurons in slice preparations. This difference might originate from the same factor as that which produces relatively short action potentials in our cultured cells. CONCLUSIONS We have developed functional cell cultures of substantia nigra and VTA, and have described their morphological, immunocytochemical and physiological properties. The cultures are rich in DA neurons (40% in the substantia nigra, and 65% in the VTA), and these DA neurons (as verified by immunocytochemical methods) produced action potentials that are almost as healthy as those reported in slice preparations. Recently, we have shown that the neurons cultured using the present methods respond well to dopaminergic agonists and GABAergic agonists. 37'38'39 Thus, our cultures provide a useful tool for studies of brain DA neurons, particularly those involving electrophysiology, neurotoxicology, neuronal development, plasticity, and transplantation. Acknowledgements--We wish to thank Dr R. S. Cohen for

critically reading the manuscript. Thanks are also due to Ms P. A. Schroeder, Ms L. Johnston and Miss A. G. Kondrat for their help. This work was supported by DA05701, AG06093, and NS24711.

REFERENCES

1. Beckstead R. M., Domesick V. B. and Nauta W. J. H. (1979) Efferent connections of the substantia nigra and ventral tegmental area in the rat. Brain Res. 175, 191~17. 2. Belluzzi O., Sacchi O. and Wanke E. (1985) A fast transient outward current in the rat sympathetic neurone studied under voltage-clamp conditions. J. Physiol., Lond. 358, 91-108. 3. Brown H. and DiFrancesco D. (1980) Voltage-clamp investigations of membrane currents underlying pace-maker activity in rabbit sino-atrial node. J. Physiol., Lond. 308, 331-351. 4. Cardozo D. L. (1990) Electrophysiologicalstudies of midbrain dopamine neurons in culture. Soc. Neurosci. Abstr. 16, 1044. 5. Chiodo L. A. and Kapatos G. (1987) Mesencephalicneurons in primary culture: immunohistochemistryand membrane physiology. In Neurophysiology o f Dopaminergic Systems--Current Status and Clinical Perspectives (eds Chiodo L. A. and Freeman A. S.), pp. 67-91. Lakeshore Publishing Co., Grosse Pointe. 6. Connor J. A. and Stevens C. F. (1971) Voltage clamp studies of a transient outward membrane current in gastropod neural somata. J. Physiol., Lond. 213, 21 30. 7. Connor J. A. and Tseng H. (1988) Measurement of intracellular Ca2+ in cerebellar Purkinje neurons in culture: resting distribution and response to glutamate. Brain Res. Bull. 21, 353-361.

362

S. MASUKO et al.

8. Connor J. A., Tseng H. and Hockberger P. E. (1987) Depolarization- and transmitter-induced changes in intracellular Ca 2+ of rat cerebellar granule cells in explant cultures. J. Neurosci. 7, 1384-1400. 9. Coyle J. T., Jacobowitz D., Klein D. and Axelrod J. (1973) Dopaminergic neurons in explants of substantia nigra in culture. J. Neurobiol. 4, 461-470. 10. Dahlstr6m A. and Fuxe K. (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta physiol, seand. 62, Suppl. 232, 1-55. 11. Dal Toso R. D., Giorgi O., Soranzo C., Kirschner G., Ferrari G., Favaron M., Benvegnfi D., Presti D., Vicini S., Toffano G., Azzone G. F. and Leon A. (1988) Development and survival of neurons in dissociated fetal mesencephalic serum-free cell cultures: I. Effects of cell density and of an adult mammalian striatal-derived neuronotrophic factor (SDNF). J. Neurosei. 8, 733 745. 12. Decavel C., Lescaudron L., Mons N. and Calas A. (1987) First visualization of dopaminergic neurons with a monoclonal antibody to dopamine: a light and electron microscopic study. J. Histochem. Cytoehem. 35, 1245-1251. 13. Di Porzio U., Rougon G., Novotny E. A. and Barker J. L. (1987) Dopaminergic neurons from embryonic mouse mesencephalon are enriched in culture through immunoreaction with monoclonal antibody to neural specific protein 4 and flow cytometry. Proe. natn. Acad. Sci. U.S.A. 84, 7334 7338. 14. Dudai Y., Ben-Barak J., Silman I. and Gazit H. (1980) Ontogenesis and modulation ofcholinergic receptors in rat brain. In Neurotransmitters and Their Receptors (eds Littauer U. Z., Dudai Y., Silman I., Teichberg V. I. and Vogel Z.), pp. 217~39. John Wiley, Chichester. 15. Eccles J. C. (1957) The Physiology o f Nerve Cells. The Johns Hopkins Press, Baltimore. 16. Evans R. A., Watson M., Yamamura H. I. and Roeske W. R. (1985) Differential ontogeny of putative MI and M 2 muscarinic receptor binding sites in the murine cerebral cortex and heart. J. Pharmac. exp. Ther. 235, 612-618. 17. Geffard M., Patel S., Dulluc J. and Rock A.-M. (1986) Specific detection of noradrenaline in the rat brain by using antibodies. Brain Res. 363. 395-400. 18. Godt R. E. and Lindley B. D. (1982) Influence of temperature upon contractile activation and isometric force production in mechanically skinned muscle fibres of the frog. J. gen. Physiol. 80, 279-297. 19. Gonzalez B. J., Leroux P., Bodenant C., Laquerri6re A., Coy D. H. and Vaudry H. (1989) Ontogeny of somatostatin receptors in the rat brain: biochemical and autoradiographic study. Neuroscience 29, 629-644. 20. Grace A. A. and Bunney B. S. (1983) Intracellular and extracellular electrophysiology of nigral dopaminergic n e u r o n a l . Identification and characterization. Neuroscience 10, 301-315. 21. Grace A. A. and Bunney B. S. (1984) The control of firing pattern in nigral dopamine neurons: single spike firing. J. Neurosci. 4, 2866~876. 22. Grace A. A. and Onn S.-P. (1989) Morphology and electrophysiological properties of immunocytochemically identified rat dopamine neurons recorded in vitro. J. Neurosci. 9, 3463 3481. 23. Grofova I., Deniau J. M. and Kitai S. T. (1982) Morphology of the substantia nigra pars reticulata projection neurons intracellularly labeled with HRP. J. comp. Neurol. 208, 352-368. 24. Guyenet P. G. and Aghajanian G. K. (1978) Antidromic identification of dopaminergic and other output neurons of the rat substantia nigra. Brain Res. 150, 69-84. 25. Halliwell J. V. and Adams P. R. (1982) Voltage-clamp analysis of muscarinic excitation in hippocampal neurons. Brain Res. 250, 71-92. 26. Hamill O. P., Marry A., Neher E., Sakmann B. and Sigworth F. J. (1981) Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pfliigers Arch. 391, 85 100. 27. Hartman B. K., Zide D. and Udenfriend S. (1972) The use of dopamine-beta-hydroxylase as a marker for the central noradrenergic nervous system in the rat. Proc. natn. Acad. Sci. U.S.A. 69, 2722-2766. 28. Hatanaka H. and Arimatsu Y. (1984) Monoclonal antibodies to tyrosine hydroxylase from rat pheochromocytoma PC12 cells with special reference to nerve growth factor-mediated increase of the immunoprecipitable enzymes. Neurosci. Res. 1, 253-263. 29. Hemmendinger L. M., Garber B. B., Hoffmann P. C. and Heller A. (1981) Target neuron-specific process formation by embryonic mesencephalic dopamine neurons in vitro. Proc. natn. Acad. Sci. U.S.A. 78, 1264 1268. 30. Heyer E. J. (1984) Electrophysiological study of mammalian neurons from ventral mesencephalon grown in primary dissociated cell culture. Brain Res. 310, 142-148. 31. H6kfelt T., Fuxe K., Goldstein M. and Johansson O. (1974) Immunohistochemical evidence for the existence of adrenaline neurons in the rat brain. Brain Res. 66, 235-251. 32. Hornykiewicz O. (1963) Die topische Lokalisation und das Verhalten von Noradrenalin und Dopamin (3-Hydroxytyramin) in der Substantia nigra des normalen und Parkinson-kranken Menschen. Wien. Klin. Wochenschr. 75, 309-312. 33. Hsu S.-M. and Soban E. (1982) Color modification of diaminobenzidine (DAB) precipitation by metallic ions and its application for double immunohistochemistry. J. Histochem. Cytochem. 30, 1079 1082. 34. Huettner J. E. and Baughman R. W. (1986) Primary culture of identified neurons from the visual cortex of postnatal rats. J. Neurosei. 6, 3044-3060. 35. Jaeger C. B. and WolfA. L. (1989) Neo-pathway formation of dissociated neural grafts demonstrated by immunocytochemistry, fluorescent microspheres, and retrograde transport. Brain Res. 487, 225-237. 36. Juraska J. M., Wilson C. J. and Groves P. M. (1977) The substantia nigra of the rat: a Golgi study. J. comp. Neurol. 172, 585-600. 37. Kim K.-M., Nakajima Y. and Nakajima S. (1990) Effects of dopaminergic agonists on cultured substantia nigra neurons. Soc. Neurosci. Abstr. 16, 81. 38. Kim K.-M., Nakajima Y. and Nakajima S. (1991) Electrophysiological and immunocytochemical characterization of cultured substantia nigra neurons in relation to dopamine receptor subtypes. Biophysical J. Abstr. 59, 20a. 39. Kim K.-M., Nakajima Y. and Nakajima S. (1991) Correlation of immunocytochemical properties with GABA responses in cultured substantia nigra neurons. Soe. Neurosci. Abstr. 17, 798. 40. Kita T., Kita H. and Kitai S. T. (1986) Electrical membrane properties of rat substantia nigra compacta neurons in an in vitro slice preparation. Brain Res. 372, 21-30. 41. K6nig N., Han V., Lieth E. and Lauder J. (1987) Effects of coculture on the morphology of identified raphe and substantia nigra neurons from the embryonic rat brain. J. Neurosci. Res. 17, 349-360.

Cultures of substantia nigra and ventral tegmental area

363

42. Kosaka T., Kosaka K., Hataguchi Y., Nagatsu I., Wu J.-Y., Ottersen O. P., Storm-Mathisen J. and Hama K. (1987) Catecholaminergic neurons containing GABA-Iike and/or glutamic acid decarboxylase-like immunoreactivities in various brain regions of the rat. Expl Brain Res. 66, 191-210. 43. Krnjevi6 K. (1955) The distribution of Na and K in cat nerves. J. Physiol., Lond. 128, 473-488. 44. Lacey M. G., Mercuri N. B. and North R. A. (1989) Two cell types in rat substantia nigra zona compacta distinguished by membrane properties and the actions of dopamine and opioids. J. Neurosci. 9, 1233 1241. 45. Lipton P. and Whittingham T. S. (1984) Energy metabolism and brain slice function. In Brain Slices (ed. Dingledine R.), pp. 113-153. Plenum Press, New York. 46. Llinhs R., Greenfield S. A. and Jahnsen H. (1984) Electrophysiology of pars compacta cells in the in vitro substantia nigra--a possible mechanism for dendritic release. Brain Res. 294, 127-132. 47. Lopez-Lozano J. J., Notter M. F. D., Gash D. M. and Leary J. F. (1989) Selective flow cytometric sorting of viable dopamine neurons. Brain Res. 486, 351 356. 48. Masuko S., Nakajima S. and Nakajima Y. (1989) Dopaminergic neuron cultures from the substantia nigra and the ventral tegmental area. Soc. Neurosci. Abstr. 15, 819. 49. Masuko S., Nakajima Y., Nakajima S. and Yamaguchi K. (1986) Noradrenergic neurons from the locus ceruleus in dissociated cell culture: culture methods, morphology, and electrophysiology. J. Neurosci. 6, 3229-3241. 50. Masurovsky E. B. and Bunge R. P. (1968) Fluoroplastic coverslips for long-term nerve tissue culture. Stain Tech. 43, 161-165. 51. Mayer M. L. and Westbrook G. L. (1983) A voltage-clamp analysis of inward (anomalous) rectification in mouse spinal sensory ganglion neurones. J. Physiol., Lond. 340, 19-45. 52. Mereu G., Costa E., Armstrong D. M. and Vicini S. (1991) Glutamate receptor subtypes mediate excitatory synaptic currents of dopamine neurons in midbrain slices. J. Neurosci. 11, 1359 1366. 53. Nakajima Y., Nakajima S., Leonard R. J. and Yamaguchi K. (1986) Acetylcholine raises excitability by inhibiting the fast transient potassium current in cultured hippocampal neurons. Proc. natn. Acad. Sci. U.S.A. 83, 3022-3026. 54. Nakajima Y., Nakajima S., Obata K., Carlson C. G. and Yamaguchi K. (1985) Dissociated cell culture of cholinergic neurons from nucleus basalis of Meynert and other basal forebrain nuclei. Proc. natn. Acad. Sci. U.S.A. 82, 6325 6329. 55. Oertel W. H., Schmechel D. E., Tappaz M. L. and Kopin I. J. (1981) Production of a specific antiserum to rat brain glutamic acid decarboxylase by injection of an antigen-antibody complex. Neuroscience 6, 26892700. 56. Oertel W. H., Tappaz M. L., Berod A. and Mugnaini E. (1982) Two-color immunohistochemistry for dopamine and GABA neurons in rat substantia nigra and zona incerta. Brain Res. Bull. 9, 463 474. 57. Okada Y., Nitsch-Hassler C., Kim J. S., Bak I. J. and Hassler R. (1971) Role of 7-aminobutyric acid (GABA) in the extrapyramidal motor system. 1. Regional distribution of GABA in rabbit, rat, guinea pig and baboon CNS. Expl Brain Res. 13, 514-518. 58. Ort C. A., Futamachi K. J. and Peacock J. H. (1988) Morphology and electrophysiology of ventral mesencephalon nerve cell cultures. Devl Brain Res. 39, 205215. 59. Ostergaard K., Schou J. P., G~ihwiler B. H. and Zimmer J. (1991) Tyrosine hydroxylase immunoreactive neurons in organotypic slice cultures of the rat striatum and neocortex. Expl Brain Res. 83, 357-365. 60. Prochiantz A., Di Porzio U., Kato A., Berger B. and Glowinski J. (1979) In vitro maturation of mesencephalic dopaminergic neurons from mouse embryos is enhanced in presence of their striatal target cells. Proc. natn. Acad. Sci. U.S.A. 76, 5387-5391. 61. Rayport S., Monaco J. and Sawasdikosol S. (1988) Identified mesoaccumbens dopamine neurons in vitro. Soc. Neurosci. Abstr. 14, 932. 62. Rayport S., Sulzer D. and Batson D. (1989) Purified cultures of ventral tegmental area (VTA) dopamine neurons. Soc. Neurosci. Abstr. 15, 819. 63. Rayport S., Sulzer D. and Shi W.-X. (1990) Heterogeneity of mesoaccumbens dopamine neurons. Soc. Neurosci. Abstr. 16, 1049. 64. Rotter A., Field P. M. and Raisman G. (1979) Muscarinic receptors in the central nervous system of the rat. III. Postnatal development of binding of [3HI propylbenzilylcholine mustard. Brain Res. Rev. 1, 185-205. 65. Schimchowitsch S., Vuillez P., Tappaz M. L., Klein M. J. and Stoeckel M. E. (1991) Systematic presence of GABA-immunoreactivity in the tubero-infundibular and tubero-hypophyseal dopaminergic axonal systems: an ultrastructural immunogold study on several mammals. Expl Brain Res. 83, 575-586. 66. Schlumpf M., Shoemaker W. J. and Bloom F. E. (1977) Explant cultures of catecholamine-containing neurons from rat brain: biochemical histofluorescence, and electron microscopic studies. Proc. natn. Acad. Sci. U.S.A. 74, 4471-4475. 67. Seroogy K. B., Dangaran K., Lim S., Haycock J. W. and Fallon J. H. (1989) Ventral mesencephalic neurons containing both cholecystokinin- and tyrosine hydroxylase-like immunoreactivities project to forebrain regions. J. cornp. Neurol. 279, 397-414. 68. Shen R., Chiodo L. A. and Kapatos G. (1990) Retrogradely labeled dopaminergic neurons in primary cultures from neonatal rats. Soc. Neurosci. Abstr. 16, 1046. 69. Silva N. L. and Bunney B. S. (1988) Intracellular studies of dopamine neurons in vitro: pacemakers modulated by dopamine. Eur. J. Pharmac. 149, 307-315. 70. Silva N. L., Mariani A. P., Harrison N. L. and Barker J. L. (1988) 5,7-dihydroxytryptamine identifies living dopaminergie neurons in mesencephalic cultures. Proc. natn. Acad. Sci. U.S.A. 85, 7346-7350. 71. Simon H., Scatton B. and Le Moal M. (1980) Dopaminergic A 10 neurones are involved in cognitive functions. Nature 286, 150-151. 72. Snyder S. H. (1973) Amphetamine psychosis: a "model" schizophrenia mediated by catecholamines. Am. J. Psychiatry 130, 61-67. 73. Stellar J. R. and Stellar E. (1985) Dopamine and ESB reward. In The Neurobiology o f Motivation and Reward (eds Stellar J. R. and Stellar E.), pp. 156-194. Springer, New York. 74. Sternberger L. A. (1979) The unlabeled antibody peroxidase-antiperoxidase (PAP) method. In Imrnunocytochemistry (ed. Sternberger L. A.), pp. 104-169. John Wiley, New York. 75. Sternberger L. A. and Joseph S. A. (1979) The unlabeled antibody method: contrasting color staining of paired pituitary hormones without antibody removal. J. Histochem. Cytochem. 27, 1424-1429. 76. Stevens J. R. (1979) Schizophrenia and dopamine regulation in the mesolimbic system. Trends Neurosci. 2, 102-105.

364

S. MASUKO et al.

77. Tepper J. M., Sawyer S. F. and Groves P. M. (1987) Electrophysiologically identified nigral dopaminergic neurons intracellularly labeled with HRP: light-microscopic analysis. J. Neurosci. 7, 2794~806. 78. Tomozawa Y. and Appel S. H. (1986) Soluble striatal extracts enhance development of mesencephalic dopaminergic neurons in vitro. Brain Res. 399, 111 124. 79. Ungerstedt U. (1971) Stereotaxic mapping of the monoamine pathways in the rat brain. Acta physiol, scand., Suppl. 367, 1 48. 80. Whetsell W. O. Jr, Mytilineou C., Shen J. and Yahr M. D. (1981) The development of the dog nigrostriatal system in organotypic culture. J. neural Transm. 52, 149-161. 81. Wise R. A. (1987) The role of reward pathways in the development of drug dependence. Pharmac. Ther. 35, 227-263. 82. Yanagihara K. and Irisawa H. (1980) Inward current activated during hyperpolarization in the rabbit sinoatrial node cell. Pfliigers Arch. 385, 11 19. (Accepted 19 January 1992)