CFJJ~ IN NEONATAL RAT HYPOTHALAMUS PRIMARY C U L T U R E - AN I M M U N O F L U O ~ N C E STUDY
. -
STEPHEN J. LOLAIT', ALAN T. LIM 3, DORIS D. DAHL 2, JOHN W. FUNDER 3'* and BAN HOCK TOH l
tMonash Department of Pathology and Immunology, Alfred Hospital, Victoria (Australia), 2Veterans Administration Hospital, 1400 Veterans of Foreign Wars, Parkway, Boston, MA (U.S.A.)and ~Medical Reseawh Centre, Prince Henry's Hospital, Victoria (Australia)
(Received July 25th, 1983; Revised version received and accepted No~ember Ist, 1983)
Hypothalami from 1 day neonatal rats were dissociated and cultured for 4-16 days. Using im- munofluorescence and antisera against neurofilament (NF) peptides, glial fibrillary acidic protein (GFAP), galactocerebroside and fibronectin we have distinguished neurons, astrocytes, oligodendrocytes and fibroblast-like cells in culture. Astrocytes initially grew as islets of 15-30 cells which dispersed as the culture aged. These cells, together with fibronectin-reactive flat cells, formed a monolayer upon which ovoid and process-bearing cells grew after 4 days in culture. Neurofilament-positive neurons constituted 5-10% of the total cell population, in maturing cultures the number of neurons decreased and fibroblasts increased. Oligodendrocytes represented less than 1% of total cell population. These studies emphasize the necessity of using the complementary techniques of morphology and immunocytochemistry for the characterization of hypothalamic neural cells in vitro.
Primary cultures of dissociated hypothalamus, previously characterized by mor- phology in light [22] and electron microscopy studies [1, 19], provide an important tool for investigating the neurosecretory activity of hypothalamic cells.
Cell-specific markers have been recently used to identify different neural cell types in cultures of sciatic nerve, optic nerve, corpus callosum and cerebellum [13], cerebral hemispheres [5] and whole brain [15, 18]. Since it is difficult to distinguish neural cell types by morphology alone, an immunohistochemical approach can serve as an ancillary aid in identifying these cells in culture.
The aim of the present study was to distinguish by morphology and im- munohistochemistry four major cells types in hypothalamus cultures, and to follow their growth as the cultures matured. The following antisera were used:
*Author for correspondence at: Medical Research Centre, Prince Henry's Hospital, St. Kilda Road, Melbourne, Victoria, Australia 3004.
(a) Anti-neurofilament (NF) [4, 51, which preferentially reacts with 70,000- and 150,000-dalton peptides of the NF 'triplet'. Not all neurons are NF-positive; e.g. neither the Purkinje cell soma nor dendrites bind NF antisera [3]. In addition, since NF-triplet expression is dependent on neuronal differentiation [17], immature, un- differentiated neurons may not express NF.
(b) Anti-glial fibrillary acidic protein (GFAP), [8, 15], widely used as a specific marker for astrocytes.
(c) Anti-galactocerebroside [14], which binds only to oligodendrocytes in central neural cultures of rodents [13].
(d) Anti-fibronectin [20], which reacts with fibroblasts, endothelial, epithelial and meningeal cells in the brain [13].
Brains aseptically removed from l-day-old Sprague-Dawley rats were cleared of meninges. Hypothalamic tissue (bounded by the optic chiasm, lateral hypothalamic sulci and posteror margin of the mammillary bodies, and to a depth of ~ 4 mm) was removed from 20-30 rats, minced and digested in M199 HEPES culture medium (Commonwealth Serum Laboratories, Melbourne, Australia) containing 1 mg/ml collagenase (Worthington, Freehold, N J) for 20 min at 37°C. The suspen- sion was centrifuged (100 g for 10 min) and the cell pellet rinsed twice with M199 HEPES containing 10070 heat-inactivated fetal calf serum (Flow Laboratories, McLean, VA), triturated with a siliconized pipet, and plated at a density of 0.5 × 10 6 cells on sterile glass coverslips (14 x 14 mm, Chance propper, Sydney, Australia) in culture wells (Sterilin, Teddington, Middlesex, U.K.). Cells were cultured in medium consisting of M199 HEPES, 5070 heat-inactivated horse serum (Gibco Laboratories, Grand Island, NY), 5o70 heat-inactivated fetal calf serum, 2 mM glutamine, !o70 non-essential amino acids (100 ×, Flow Laboratories, McLea, VA), and 40 U/ml penicillin in a humidified atmosphere of 5o70 CO2/95~0 air at 37°C. Cultures were undisturbed for 48 h after which media were changed twice weekly.
immunofluorescence staining of cultured cells was performed as previously described [l 1]. Astrocytes and neurons were fluorescently stained with rabbit an- tibodies against GFAP [8l and NF peptides [15], respectively; oligodendrocytes were identified with rabbit antisera to galactocerebroside prepared by the method of Raft et al. [ 14], and by their reactivity with a monoclonal antibody to galactocerebroside (gift of Dr. Perry Bartlett, W.E.H.I., Melbourne); fibroblast-like cells were distinguished by their reactivity with a monoclonal antibody of fibronectin [20] (Commonwealth Serum Laboratories, Melbourne, Australia). The antisera were used at l:10 to 1:80 dilutions in phosphate-buffered saline (PBS) pH 7.2. After staining, the coverslips were mounted in PBS/glycerol containing p- phenylenediamine to reduce fluorescence bleaching [7], and viewed under a Leitz Diavert epi-illumination microscope equipped with phase-contrast optics.
The hypothalamic cells attached to the glass substrate within 12 h with minimal re-aggregation. After 2 days in vitro, cells grew either in islands of 15-30 GFAP-
139
positive cells or were scattered randomly across the coverslip surface. The densely packed cells in the center of the glial islets showed a rounded morpholc~o-,y, while those at the periphery showed process formation. In contrast, cells outside the glial islands had a polyhedric, fibroblastoid morphology and were mostly fibronectin- positive. Occasional cells (< 10%0) outside the islets were GFAP-positive.
After 4 days in vitro, the glial cell islets dispersed. Highly refractile cells, growing on top of the layer of glial and fibroblast-like cells, included large (25-45 put) ovoid cells with a neuroblastic appearance, multipolar cells (20-35 pan) with thick pro- cesses (up to 50 pm in length), and small (8-12 ~m) bipolar cells. Most of the overly- ing cells, however, were bipolar and grew in clusters; of these cells, 20%0 were NF- positive, 30-40~0 GFAP-positive and the remainder negative for all 4 markers. In addition, occasional ovoid cells (--5%) and 40-50% of the large, multipolar cells were NF-positive (Fig. 1); all ovoid and multipolar cells were, however, negative for GFAP, fibronectin and galactocerebroside. A small number (2-3%)of NF-positive cells with long processes (120-250 pro) and NF-positive microneurons were random- ly distributed in the basal carpet layer (Fig. l d, e). The former cells often had pro- cesses 120-250/~m in length. At 4 days in vitro NF-positive overlying and basal cells comprised 5-10070 of the total cell population.
The basal carpet of 'non-neuronal' supporting cells in 4-day cultures comprised 20-30% glial cells, -~ 50% fibronectin-positive cells and fewer than 1% oligodt.n- drocytes. Process-bearing (fibrous) and non-process-bearing epithelioid (p~'o- toplasmic) GFAP-positive cells were present in equal numbers, with some cells of intermediate morphology (.Fig. 2); the diameter of glial cells ranged from 30 to 70 ~m. The galactocerebroside-positive oligodendrocytes were sparsely distributed throughout the cultures, and had long (60-70/~m) processes ramifying from a small (10-12 ~m) cell body (Fig. 3).
The main feature of cultures between 4 and 16 days in vitro was a profound decrease in the number of NF-positive cells, and a concomitant increase in fibronectin-positive, fibroblast-like cells. By 16 days in vitro, the number of NF- positive cells had decreased to 1-2%, and the number of fibronectin-positive cells had increased to ---70%.
Fig. I. NF-positive cells in 4-day cultures, a: a single, ovoid neuron overlying NF-negative background cells, b: a multipolar neuron with thick process (arrow). c: a cluster of bipolar neurons, d and e: microneuron (d) and process-bearing (e) neuron, with cell bodies arrowed. Scale bar = 10 gin.
~+~4~i I
140
Fig. 2+ Varying morphological forms of GFAP-positive cells seen in 4-day cultures, a: protoplasmic a~troc.vte, b and c: intermediate morphologies, d: fibrous aslrocyte. Scale bar = i0 / tm.
The growth and maintenance of primary cultures of dispersed hypothalamus derived from fetal and neonatal tissue has been extensively reported [l, 9, 19, 21, 22]. Morphological and histochemical studies assessed by light microscopy have shown the development of neurons growing upon a basal layer of glia and fibroblasts. Two putative neuron populations have been described: small, bipolar ceils and larger (up to 50 #m diameter) multipolar neurons [91. Electron microscopic studies [19] classified the overlying neurons into small, primitive neuroepithelial cells and maturing or mature bipolar and multipolar neurons.
Fig. 3. Cell surface staining cf cell body (arrow) and processes (arrowheads) of an oligodendrocyte in 4-day cultures, a: Phase-contrast. b: same cell as in a reacted with anti-galactocerebroside. Scale bar = 10 tim.
141
Our study re-emphasizes the caution that must be exercised in characterizing cells based solely on morphological criteria [5, 13, 15], since morphologically similar overlying bipolar cells may express NF or GFAP. Using antisera to NF peptides, we detected 5 morphologically distinct neuron subpopulations - bipolar, ovoid and large multipolar neurons, microneurons and small neurons possessing long pro- cesses. Since the expression of NF is dependent on neuronal differentiation [ 17], the NF-negative, ovoid cells without processes may represent undifferentiated neurons. Alternatively, since not all neurons are NF-positive [3], it is possible that NF- negative overlying cells are neurons which do not express NF peptides.
Three types of non-neuronal cells constituting the basal cell layer were identified. The description of GFAP-positive cells (fibrous and protoplasmic) is in accord with observations reported by others [14]. Galactocerebroside-positive oligodendrocy/es represented a minor cell type, as in cultures of cerebellum and cerebral cortex [13]. Fibronectin-positive fibroblast-like cells expressing fibronectin were the major cell type in culture. Techniques to select for neurons while selectively depleting non- neuronal cells (e.g. polylysine-coated culture surfaces [23], antimetabolites [18], or fibroblast lysis by anti-Thy. 1 plus complement [13]) may allow for greater neuronal enrichment in this culture system. Neuronal survival depends on the presence of growth factors [2] and in some cases appears to be due to direct cell-cell contact with metabolitically active non-neuronal cells [21]. Varying the culture conditions (e.g. serum, substrate, hormones) may compensate for any loss of non-neuronal cell sup- port, and may also reduce expression of NF in a larger number of 'putative' neurons.
To distinguish different classes of neural cells in this study, the complementary techniques of immunohistochemistry and morphological description were used. The advent of monoclonai antibodies to many neural antigens [16] i and to neurotransmitters and neuropeptides [12], will enable even mo~e precise characterization to be achieved in future studies.
1 Benda, P.. DeVitry, F., Picart, R. and Tixier-Vidal, A., Dissociated cell cultures fro~! fetal mouse hypothala~aus. Patterns of organisation and ultrastructural features, Exp. Brain Res., 23 (1975) 29-47.
2 Bottenstein, J.E., Skaper, S.D., Vavon, S.S. and Sato, G.H., Selective survival of neurons from chick embryo sensory ganglionic dissociates utilizing serum-free supplemented medium, Exp. Cell. Res., 125 (19~0) 183-190.
3 Dahl, D. and Bignami, A., Preparation of antisera to neurofilament protein from chickea brain and human sciatic nerve. J. comp. Neurol., 176 (1977) 645-658.
4 Dahl, D., Crosby, C. and Bignami, A., Filament proteins in rat optic nerves undergoing Wallerian degeneration, Exp. Neurol., 71 (198 I) 421-430.
5 Hansson, E., Sellstrom, A., Persson, L.I. and Ronback, L., Brain primary culture - a characteriza- tion, Brain Res., 188 (1980) 233-246.
6 Jirikowski, G., Reisert, l. and Pilgrim, Ch., Neuropeptides in dissociated cultures of hypothalamus and septum: quantitation of immunoreactive neurons, Neuroscience, 6 (1981) 1953-1960.
7 Johnson, G.D. and de C. Nogueira Araujo, G.M., A simple method of reducing the fading of im- munofluorescence during microscopy, J. lmmunol. Meth., 43 (1981) 349-350.
142
8 Kozak, L., Dahl, D. and Bignami, A., Glial fibrillary acidic protein in reaggregating and monolayer cultures of fetal mouse cerebral hemispheres, Brain Res., 150 (1978) 631-637.
9 Knigge, K.M., Hoffman, G., Scott, D.E. and Sladek, Jr., J.R., Identification of catecholamine and luteinizing hormone-releasing hormone 0LH-RH)-containing neurons in primary cultures of dispersed cells of the basal hypothalamus, Brain Res., 120 (1977) 393--405.
10 Liotta, A.S., Loudes, C., McKelvy, J.G. and Krieger, D.T., Biosynthesis of precursor cor- ticotrophin/endorphin-, corticotropin-, a-melanotropin-, #-Iipotropin-, and ~endorphin-like material by cultured neonatal rat hypoth~l~mic neurons, Proc. nat. Acad. Sci. U.S.A., 77 (1980) 1880--1884.
11 Lolait, S.J., Lim, A.T., Dahl, D., Khalid, B.A.K., Toh, B.H. and Funder, J.W., Neonatal rat hypo- thalamus cell culture: neuron subpopulations secrete immunoreactive4J-endorphin but not immuno- reactive-ACTH, Neuroendocrinology, 37 (1983) I I I-I 16.
12 Meo, T., Gramsch, C., lnan, R., Hellt, V., Weber, E., Herz, A. and Riethmuller, G., Monoclonal antibody to the message sequence Tyr-Gly-Gly-Phe of opioid peptides exhibits the specifidty re- quirements of mammalian opioid receptors, Proc. nat. Acad. Sci. U.S.A., 80 (1983) 4084-4088.
13 Raft, M.C., Fields, K.L., Hakomori, S.-I., Mirsky, R., Pruss, R.M. and Winter, J., Cell-type- specific markers for distinguishing and studying neurons and the major classes of glial cells in culture, Brain Res., 174 (1979) 283-308.
14 Raff, M., Mirsky, R., Fields, K., Lisak, R., Dorfman, S., Silberberg, D., Gregson, N., Leibowitz, S. and Kennedy, M., Galactocerebroside is a specific cell-surface antigenic marker for oligoden- drocytes in culture, Nature (Lond.), 272 (1978) 813-816.
15 Raju, T., Bignami, A. and Dahl, D., in vivo and in vitro differentiation of neurons and astrocytes Jn the rat embryo, Develop. Biol., 85 (1981) 344-357.
16 Schachner. M., Cell-type-specific surface antigens in the mammalian nervous system, J. Neurochem., 39 (1982) I-8.
17 Shaw, G. and Weber, K., Differential expression of neurofilament triplet proteins in brain develop- ment, Nature (Lond.), 298 (1982) 277-279.
18 Sotelo, J., Gibbs, C.J., Jr., Gajdusek, D.C., Toh, B.H. and Wurth, M., Method for preparing cultures of central neurons: Cytochemical and immunochemic.'tl studies, Proc. nat. Acad. Sci. U.S.A., 77 (1980) 653-657.
19 Tixier-Vidal, A. and De Vitry, F., Hypothalamic neurons in cell culture, Int. Rev. Cytol., 58 (1979) 291-331.
20 Vaheri0 A. and Mosler, D.F., High molecular weight, cell-surface associated glycoprotein (fibronec- tin) lost in malignant transformation, Biochim. biophys. Acta, 517 (1978) 1-25.
21 Whatley, S.A., Hall, C. and Lim, L., Hypothalamic neurons in dissociated cell culture: the mechanism of increased survival times in the presence of non-neuronal cells, J. Neurochem., 36, (! 981) 2052-2056.
22 Wilkinson, M., Gibson, C.J., Bressler, B.H. and lnman, D.R., Hypothalamic neurons in dissociated cell culture, Brain Res., 82 (1974) 129-138.
23 Yavin, E. and Yavin, Z., Attachment and culture of dissociated cells from rat embryo cerebral hemispheres on polylysine coated surface, J. Cell Biol., 62 (1974) 540-546
