Brain Research, 120 (1977) 393--405
393
© Elsevier/North-Holland Biomedical Press, Amsterdam - Printed in The Netherlands
Research Reports
IDENTIFICATION OF CATECHOLAMINE AND LUTEINIZING HORMONERELEASING HORMONE (LHRH)-CONTAINING NEURONS IN PRIMARY CULTURES OF DISPERSED CELLS OF THE BASAL HYPOTHALAMUS*
K. M. KNIGGE, G. HOFFMAN**, D. E. SCOTT*** and J. R. SLADEK, Jr.
Department of Anatomy, Universityof Rochester School of Medicine, Rochester, N. Y. 14642 (U.S.A.) (Accepted May 15th, 1976)
SUMMARY
Primary cultures of dispersed cells were prepared from 3-5 mg pieces of basal hypothalami of lO-12-day-old rats. The tissue included median eminence, arcuate nucleus and variable amounts of adjacent hypothalamus and preoptic area. The dispersion procedure consisted basically of tissue trypsinization and mechanical dissociation of cells. They were cultured in a modified L-15 medium in an air atmosphere. Neurons survived approximately 3 months. On the basis of morphological characteristics, two basic cell types could be distinguished. One was a larger (50 #m diameter) multipolar cell; microspectrofluorometric analysis revealed that a small percentage of these neurons contained a catecholamine. A second type was smaller, fusiform or ovoid and generally bipolar; a significant number of these were immunoreactive for the releasing hormone LHRH.
INTRODUCTION
The endocrine hypothalamus 16 contains several populations of cells responsible for the production of neurohormones which are delivered to the adenohypophysis and regulate the secretion of its tropic hormones. One of these hypothalamic principles, luteinizing hormone-releasing hormone (LHRH) has been visualized by immunocytochemical methods1-3,11,15,17,24,26,34. Dopamine-containing neurons have been identified by fluorescence cytochemistryS,~,s,~o,~a,xL The role of these dopa* Supported by Program Project Grant NS-11642. ** NIH Fellow 122HD00640 and PMA Foundation Awardee, Pharmacology-Morphology. *** Career Development Awardee K04GM77001.
394 minergic neurons as well as the potential influence of noradrenergic and serotonergic input upon LHRH secretion are being studied intensively. The use of dissociated cell cultures would appear to offer many opportunities for elucidating important issues of neurohormone synthesis, intrinsic cytoarchitectural relations, and hormonal feedback. In this paper, we describe the preparation of cell cultures of dissociated hypothalamic neurons which were maintained for periods up to 12 weeks and in which we identify catecholamine- and LHRH-containing neurons. MATERIAL AND METHODS Hypothalami from 10-12-day-old male and female Sprague-Dawley rats were used in this study. Brains were removed after decapitation and rinsed repeatedly with copius quantities of sterile saline containing 1 ~ penicillin-streptomycin mixture. Fine forceps and irridectomy scissors were used to remove small pieces of basal hypothalamus under a dissecting magnification of × 25. The piece of tissue removed extended from the distal end of the infundibular stalk to the anterior margin of the optic chiasm approximately 0.3 mm on each side of the midline and 0.3-0.4 mm deep. This piece of tissue weighed 3-5 mg and included the median eminence, arcuate nucleus, preoptic area, and variable amounts of additional adjacent hypothalamic tissue. Median eminence, arcuate nucleus and preoptic area were the regions of principal interest since dopaminergic and LHRH-producing neurons are localized here. Tissue (40-50 pieces per experiment) was collected in sterile ice-cold Hank's basal medium containing 0.5~ bovine serum albumin (BSA) and 1 ~ penicillin-streptomycin mixture (Hank's-BSA). Further steps in the dispersion procedure required approximately 2 h and consisted of initial treatment of the tissue pieces through the following steps: (a) 3 washes of 5 ml cold sterile Hank's-BSA, 15 min each, (b) treatment with 5 ml 0.1 ~ trypsin (Sigma, type 3) in Hank's-BSA at 37 °C for 25 rain with occasional agitation, (c) a 5 rain wash in 2 mM EDTA in calcium-magnesium-free Hank's medium at 37 °C, and (d) 3 successive washes of 5 ml of calcium-magnesium-free Hank's, 5 min each. At each step solutions were removed as completely as possible by suction. Cells were dispersed in 2 ml calcium-magnesium-free Hank's by mechanically shearing the tissue through pipettes of 500 and 100/zm tip diameters; approximately 50 excursions through the 100 #m pipette in 3 4 min produced virtually complete dissociation. This suspension was allowed to stand for 5 min after which 1.8 ml of the supernatant was removed, introduced into 5 ml calcium-magnesium-free Hank's, and centrifuged at 70 x g for 10 rain. The supernatant was removed and the pellet of ceils resuspended in final culture medium at a cell density of approximately 105 cells/ml. The total cell yield from 40-50 pieces of tissue was 1-2 x 106 cells. Cell viability was in the range of 85 ~ when examined in a hemocytometer with 0.05 ~/o nigrosin. Culture medium consisted of a stock solution of L-15 (Leibowitz) from Microbiological Associates, Inc. (Bethesda, Md.); 14.9 g were dissolved in 1080 ml triple distilled, deionized water and the following substances (in rag) added: imidazole, 60; aspartic acid, 15: fumaric acid, 25; glutamic acid, 15; proline, 15; cystine, 15; /3-alanine, 5; vitamin B-12, 2; inositol, 10; choline chloride, 10; p-aminobenzoic acid,
395 5; ascorbic acid, 2; lipoic acid, 0.5; glutathione, 0.5; DMPH4, 0.05; and biotin, 0.02. Final culture medium was prepared by adding to each 100 ml of L-15 stock 600 mg glucose, 1 ml 200 mM glutamine, 1 ml penicillin-streptomycin mixture, 150 mg BSA and 5 ml fetal calf serum. Two ml aliquots of the cell suspension were transferred to 35 mm Falcon polystyrene dishes and the cultures maintained in an incubator at 36.8 °C with 92-96 % relative humidity and air atmosphere. Cultures were left undisturbed for the first 4-5 days after which they were placed on a regular feeding routine of 3-4 day intervals. No cytotoxic agents were used to suppress dividing cells; cultures were viewed and photographed under phase optics on a Wild inverted microscope. For scanning electron microscopic studies, culture dishes were drained of nutrient media and gently washed for 1 min in 0.1 M phosphate buffer. Culture dishes were slowly infused with phosphate-buffered Karnovsky's fixative 14 at a pH of 7.3, fixed for 2 h, washed in distilled water and postosmicated in 1% osmium tetroxide for 1 h. The floors of the culture dishes with adherent cells were cut into quarters and dehydrated through ascending acetone series. Acetone was replaced by immersion in graded Freon TF solutions and transferred from absolute Freon TF into the high pressure chamber of a Bomar SPC-50 critical point drying apparatus and flooded with Freon 113, the final exchange solvent. The specimens were critically point dried, shadow casted with 8 cm of 10 rail gold in a Denten vacuum evaporator and analyzed with a JEOL JSM 35 scanning electron microscope. Fluorescence histochemistry was performed according to a modification of the glyoxylic acid technique of Watson and Barchas31. Freshly prepared 0.1 M phosphate buffer (pH 7.4) was made by combining 19.0 ml of 0.2 M NaHzPO4 with 81.0 ml of 0.2 M Na2HPO4 and diluting this with distilled water to 0.1 M. Two per cent glyoxylic acid (free acid monohydrate, Sigma, St. Louis, Mo.)-0.5 % mg MgCI~ was prepared in this buffer no more than 0.5 h prior to fluorophor formation. The final pH was adjusted with 0.1 M NaOH to 4.9-5.0 and then chilled to 0-2 °C. Dispersed cells were rinsed thoroughly with phosphate buffer and then bathed with 3 ml of glyoxylic acid solution for 3 rain at 0-2 °C. Culture dishes were drained thoroughly and air dried under a warm stream of air (45 °C) for 5 min. Finally, they were heated in an oven at 80 °C for 5 min and then stored in a vacuum over silica gel until time of examination. These preparations were infiltrated with warm immersion oil and examined in a Leitz MPV2 microspectrofluorometer equipped with Schoffel excitation and emission grating monochromoters, photomultipliers and ratio computing circuitry. Excitation and emission spectra were recorded according to the principles of Bj6rklund et al. 7. Culture dishes were scanned with either monochromatic light (370 nm) or narrow band excitation ($405, BG3) light for the purpose of identifying fluorescence within cells. Ploem illumination was used in combination with a × 63 or × 100 oil immersion objective. For immunocytochemical identification of LHRH-containing neurons, the unlabeled antibody enzyme method of Sternberger ~9 was used. Cultures were rinsed with phosphate buffer-saline (PBS) and then rapidly frozen on a cold stage at ---20 °C. They were thawed and fixed with 1 ml of Bouin's fixative. After a 2 min fixation period the cultures were rinsed with 50 % ethanol followed by 70 ~o ethanol
396
Figs. i-3. Phase micrographs of large multipolar type I neurons from a 50-day-old culture. Primary processes of these cells are fairly thick and exhibit numerous thickenings and possible sites of contact. In Figs. 1 and 2, a variety of other, unidentified processes (~) cris-cross the field. In Fig. 3, a smaller type II neuron is present..': 700. in order to remove traces of picrie acid. Once the yellow coloration was removed, the cultures were again rinsed with 50 ~o ethanol followed by two changes of PBS. Cold PBS containing 1 ~ bovine serum albumin (BSA) was poured into the dish and allowed to stand for 2-5 min. The dish was drained and 16 drops of rabbit a n t i - L H R H antiserum (1 : 1000 with PBS containing 1 ~ BSA) was added. The dishes were placed in a large petri dish on top of moistened filter paper and were incubated at 4 °C for 48 h. The cultures were rinsed with PBS, and incubated with goat anti-rabbit antisera (1:20 in PBS) for 30 min; they were rinsed again with PBS and 16 drops of rabbit peroxidase anti-peroxidase complex (PAP) were added. After 20 rain, the PAP was washed away and 16 drops o f 50 m g ~ 3,3'-diaminobenzidine HCI in 0.5 M Tris buffer, p H 7.8 containing 0.03 ~ hydrogen peroxide was added. The dishes were agi-
397
;~ :
~
~i~I it ~¸
Figs. 4-8. Phase micrographs of type II fusiform or ovoid neurons from a 65-day-old culture. Nuclei and other intracellular structures are less conspicuous than in the flattened, multipolar type I cells. Process of these neurons may extend for considerable distances. Possible sites of contact between processes of these cells and with unidentified processes are numerous (~'). x 700.
tated gently on a mechanical shaker for 10 min at room temperature, rinsed in PBS, and mounted with glycogel. For controls, identical cultures were run using antiLHRH which had been absorbed with synthetic LHRH (either 4 or 10 fig LHRH added to 1 ml diluted antisera). RESULTS
Growth and differentiation in culture Dispersed cells attached to the floor of the culture dish within 6-8 h; morpholog-
398 ical development, however, proceeded slowly and 3-4 days passed before processes of any cell types were apparent. Connective tissue cells were the first to exhibit outgrowth of processes; after 10-14 days, they produced a moderate syncytium over the floor of the culture dish. Neuroglia shared the floor of the dish with connective tissue cells; the majority were considered astrocytes on the basis of their relative size and fairly concentric distribution of short processes which branched repeatedly. Neurons at this time were identified tentatively by their highly refractile appearance; they were almost exclusively bipolar with short processes (40-100/~m) at least one of which characteristically attached to the edge of connective tissue cell processes or had grown on to their surface. By the fourth week of culture, a thin carpet of connective tissue covered approximately three-fourths of the surface of the floor of the culture dish and neurons appeared to have used this to greatly extend the field of their processes. Although not all cells could be categorically identified, two types were present in sufficiently consistent numbers to warrant designations as types I and II. Type I cells, 20-25 #m in diameter, were flat, stellate, with large nuclei and single, prominent nucleoli (Figs. 1-3); cytoplasmic granules were conspicuous. Two to 3 primary processes of 1--2/~m diameter exhibited irregular varicosities and thickenings. Type II cells (Figs. 4-8) were smaller and fusiform or ovoid in shape. These cells did not flatten in culture and maintained an identifying refractile halo in phase optics; nucleus, nucleolus and cytoplasmic structures were visible only occasionally (Figs. 6, 8). The majority of these cells were bipolar, although a third primary process was not uncommon (Figs. 6, 7). The thickness of the primary processes was more variable than those of type I cells; some possessed delicate, fairly smooth processes (Fig. 5) which extended, with branching, for several millimeters. Others were beaded with varicosities at loci along the process which gave the appearance of either points of branching or points of contact with other processes (Figs. 4-7). Both of these cell types maintained these characteristics for periods of 50-60 days in culture. After this time, processes became more irregular and gnarled in appearance; cell loss became conspicuous after 75-80 days. Scanning electron micrographs further verified basic differences in the structure of these cell types (Figs. 9, 10). In addition to the 3-4 large primary processes of type I cells seen in phase optics, scanning micrographs revealed that there may be varying numbers of additional, smaller processes which frequently form complex interconnections close to the perikaryon (Fig. 9). The basic bipolar organization of fusiform type II cells is seen to great advantage in scanning micrographs (Fig. 113).
Fluorescence analysis Cultures treated with glyoxylic acid displayed variable amounts of dull green fluorescence within the connective tissue carpet upon which neurons grew. Some large fibroblasts contained this fluorescence although many were non-fluorescent; the latter often contained dull yellow granules. A similar fluorescence was seen in many type I and II neurons. Of considerable significance with respect to the nature of this green fluorescence is the fact that it was qualitatively and quantitatively the same when glyoxylic acid was omitted from the procedure.
399
r 4,
Fig. 9. Scanning electron micrograph of a presumed type I neuron. With scanning electron microscopy this multipolar cell type is consistently observed above the matrix of the tissue culture preparation. This species of cell, significantly different in appearance and configuration than type II bipolar cells, exhibits numerous delicate processes which branch, course and ramify over the surface of the tissue culture, x 1800. Fig. 10. Scanning electron micrograph of Type II bipolar neuron. Both major branches of this cell lype eventually subdivide extensively over considerable distances. × 1600. A brilliant blue fluorescence visually characteristic of catecholamines treated with glyoxylic acid was present in relatively few neurons (Fig. 11). Fluorescence was localized to the perinuclear cytoplasm as well as in some processes and their varicosities. The size, arrangement of processes and general organization of these cells suggest that they are multipolar, type I neurons. Preliminary studies involving the use of desipramine indicate no change in the compartmentalization of either the green background fluorescence or the catecholamine-specific blue fluorescence. After treatment with reserpine, however, blue fluorescence was not seen in type I neurons. Microspectrofluorometric analysis of these two kinds of fluorescence revealed an excitation peak at 370 nm for the green fluorescence and a peak at 410 nm for the blue fluorescence (Fig. 12). Emission spectra revealed a primary emission peak at 480 nm for the blue, type I cells; the greenish fluorescence of connective tissue and
400
Fig. 11. An intense blue catecholarnine fluorescence is seen within a dispersed, organ-cultured hypothalamic neuron. Two fluorescent processes (arrow) appear to emanate from the perikaryon. One of these can be seen to extend a considerable distance across this photornontage. This process appears to branch and contains irregular fluorescent swellings (arrowhead) which are reminiscent of varicosities seen in vivo. ~
393
© Elsevier/North-Holland Biomedical Press, Amsterdam - Printed in The Netherlands
Research Reports
IDENTIFICATION OF CATECHOLAMINE AND LUTEINIZING HORMONERELEASING HORMONE (LHRH)-CONTAINING NEURONS IN PRIMARY CULTURES OF DISPERSED CELLS OF THE BASAL HYPOTHALAMUS*
K. M. KNIGGE, G. HOFFMAN**, D. E. SCOTT*** and J. R. SLADEK, Jr.
Department of Anatomy, Universityof Rochester School of Medicine, Rochester, N. Y. 14642 (U.S.A.) (Accepted May 15th, 1976)
SUMMARY
Primary cultures of dispersed cells were prepared from 3-5 mg pieces of basal hypothalami of lO-12-day-old rats. The tissue included median eminence, arcuate nucleus and variable amounts of adjacent hypothalamus and preoptic area. The dispersion procedure consisted basically of tissue trypsinization and mechanical dissociation of cells. They were cultured in a modified L-15 medium in an air atmosphere. Neurons survived approximately 3 months. On the basis of morphological characteristics, two basic cell types could be distinguished. One was a larger (50 #m diameter) multipolar cell; microspectrofluorometric analysis revealed that a small percentage of these neurons contained a catecholamine. A second type was smaller, fusiform or ovoid and generally bipolar; a significant number of these were immunoreactive for the releasing hormone LHRH.
INTRODUCTION
The endocrine hypothalamus 16 contains several populations of cells responsible for the production of neurohormones which are delivered to the adenohypophysis and regulate the secretion of its tropic hormones. One of these hypothalamic principles, luteinizing hormone-releasing hormone (LHRH) has been visualized by immunocytochemical methods1-3,11,15,17,24,26,34. Dopamine-containing neurons have been identified by fluorescence cytochemistryS,~,s,~o,~a,xL The role of these dopa* Supported by Program Project Grant NS-11642. ** NIH Fellow 122HD00640 and PMA Foundation Awardee, Pharmacology-Morphology. *** Career Development Awardee K04GM77001.
394 minergic neurons as well as the potential influence of noradrenergic and serotonergic input upon LHRH secretion are being studied intensively. The use of dissociated cell cultures would appear to offer many opportunities for elucidating important issues of neurohormone synthesis, intrinsic cytoarchitectural relations, and hormonal feedback. In this paper, we describe the preparation of cell cultures of dissociated hypothalamic neurons which were maintained for periods up to 12 weeks and in which we identify catecholamine- and LHRH-containing neurons. MATERIAL AND METHODS Hypothalami from 10-12-day-old male and female Sprague-Dawley rats were used in this study. Brains were removed after decapitation and rinsed repeatedly with copius quantities of sterile saline containing 1 ~ penicillin-streptomycin mixture. Fine forceps and irridectomy scissors were used to remove small pieces of basal hypothalamus under a dissecting magnification of × 25. The piece of tissue removed extended from the distal end of the infundibular stalk to the anterior margin of the optic chiasm approximately 0.3 mm on each side of the midline and 0.3-0.4 mm deep. This piece of tissue weighed 3-5 mg and included the median eminence, arcuate nucleus, preoptic area, and variable amounts of additional adjacent hypothalamic tissue. Median eminence, arcuate nucleus and preoptic area were the regions of principal interest since dopaminergic and LHRH-producing neurons are localized here. Tissue (40-50 pieces per experiment) was collected in sterile ice-cold Hank's basal medium containing 0.5~ bovine serum albumin (BSA) and 1 ~ penicillin-streptomycin mixture (Hank's-BSA). Further steps in the dispersion procedure required approximately 2 h and consisted of initial treatment of the tissue pieces through the following steps: (a) 3 washes of 5 ml cold sterile Hank's-BSA, 15 min each, (b) treatment with 5 ml 0.1 ~ trypsin (Sigma, type 3) in Hank's-BSA at 37 °C for 25 rain with occasional agitation, (c) a 5 rain wash in 2 mM EDTA in calcium-magnesium-free Hank's medium at 37 °C, and (d) 3 successive washes of 5 ml of calcium-magnesium-free Hank's, 5 min each. At each step solutions were removed as completely as possible by suction. Cells were dispersed in 2 ml calcium-magnesium-free Hank's by mechanically shearing the tissue through pipettes of 500 and 100/zm tip diameters; approximately 50 excursions through the 100 #m pipette in 3 4 min produced virtually complete dissociation. This suspension was allowed to stand for 5 min after which 1.8 ml of the supernatant was removed, introduced into 5 ml calcium-magnesium-free Hank's, and centrifuged at 70 x g for 10 rain. The supernatant was removed and the pellet of ceils resuspended in final culture medium at a cell density of approximately 105 cells/ml. The total cell yield from 40-50 pieces of tissue was 1-2 x 106 cells. Cell viability was in the range of 85 ~ when examined in a hemocytometer with 0.05 ~/o nigrosin. Culture medium consisted of a stock solution of L-15 (Leibowitz) from Microbiological Associates, Inc. (Bethesda, Md.); 14.9 g were dissolved in 1080 ml triple distilled, deionized water and the following substances (in rag) added: imidazole, 60; aspartic acid, 15: fumaric acid, 25; glutamic acid, 15; proline, 15; cystine, 15; /3-alanine, 5; vitamin B-12, 2; inositol, 10; choline chloride, 10; p-aminobenzoic acid,
395 5; ascorbic acid, 2; lipoic acid, 0.5; glutathione, 0.5; DMPH4, 0.05; and biotin, 0.02. Final culture medium was prepared by adding to each 100 ml of L-15 stock 600 mg glucose, 1 ml 200 mM glutamine, 1 ml penicillin-streptomycin mixture, 150 mg BSA and 5 ml fetal calf serum. Two ml aliquots of the cell suspension were transferred to 35 mm Falcon polystyrene dishes and the cultures maintained in an incubator at 36.8 °C with 92-96 % relative humidity and air atmosphere. Cultures were left undisturbed for the first 4-5 days after which they were placed on a regular feeding routine of 3-4 day intervals. No cytotoxic agents were used to suppress dividing cells; cultures were viewed and photographed under phase optics on a Wild inverted microscope. For scanning electron microscopic studies, culture dishes were drained of nutrient media and gently washed for 1 min in 0.1 M phosphate buffer. Culture dishes were slowly infused with phosphate-buffered Karnovsky's fixative 14 at a pH of 7.3, fixed for 2 h, washed in distilled water and postosmicated in 1% osmium tetroxide for 1 h. The floors of the culture dishes with adherent cells were cut into quarters and dehydrated through ascending acetone series. Acetone was replaced by immersion in graded Freon TF solutions and transferred from absolute Freon TF into the high pressure chamber of a Bomar SPC-50 critical point drying apparatus and flooded with Freon 113, the final exchange solvent. The specimens were critically point dried, shadow casted with 8 cm of 10 rail gold in a Denten vacuum evaporator and analyzed with a JEOL JSM 35 scanning electron microscope. Fluorescence histochemistry was performed according to a modification of the glyoxylic acid technique of Watson and Barchas31. Freshly prepared 0.1 M phosphate buffer (pH 7.4) was made by combining 19.0 ml of 0.2 M NaHzPO4 with 81.0 ml of 0.2 M Na2HPO4 and diluting this with distilled water to 0.1 M. Two per cent glyoxylic acid (free acid monohydrate, Sigma, St. Louis, Mo.)-0.5 % mg MgCI~ was prepared in this buffer no more than 0.5 h prior to fluorophor formation. The final pH was adjusted with 0.1 M NaOH to 4.9-5.0 and then chilled to 0-2 °C. Dispersed cells were rinsed thoroughly with phosphate buffer and then bathed with 3 ml of glyoxylic acid solution for 3 rain at 0-2 °C. Culture dishes were drained thoroughly and air dried under a warm stream of air (45 °C) for 5 min. Finally, they were heated in an oven at 80 °C for 5 min and then stored in a vacuum over silica gel until time of examination. These preparations were infiltrated with warm immersion oil and examined in a Leitz MPV2 microspectrofluorometer equipped with Schoffel excitation and emission grating monochromoters, photomultipliers and ratio computing circuitry. Excitation and emission spectra were recorded according to the principles of Bj6rklund et al. 7. Culture dishes were scanned with either monochromatic light (370 nm) or narrow band excitation ($405, BG3) light for the purpose of identifying fluorescence within cells. Ploem illumination was used in combination with a × 63 or × 100 oil immersion objective. For immunocytochemical identification of LHRH-containing neurons, the unlabeled antibody enzyme method of Sternberger ~9 was used. Cultures were rinsed with phosphate buffer-saline (PBS) and then rapidly frozen on a cold stage at ---20 °C. They were thawed and fixed with 1 ml of Bouin's fixative. After a 2 min fixation period the cultures were rinsed with 50 % ethanol followed by 70 ~o ethanol
396
Figs. i-3. Phase micrographs of large multipolar type I neurons from a 50-day-old culture. Primary processes of these cells are fairly thick and exhibit numerous thickenings and possible sites of contact. In Figs. 1 and 2, a variety of other, unidentified processes (~) cris-cross the field. In Fig. 3, a smaller type II neuron is present..': 700. in order to remove traces of picrie acid. Once the yellow coloration was removed, the cultures were again rinsed with 50 ~o ethanol followed by two changes of PBS. Cold PBS containing 1 ~ bovine serum albumin (BSA) was poured into the dish and allowed to stand for 2-5 min. The dish was drained and 16 drops of rabbit a n t i - L H R H antiserum (1 : 1000 with PBS containing 1 ~ BSA) was added. The dishes were placed in a large petri dish on top of moistened filter paper and were incubated at 4 °C for 48 h. The cultures were rinsed with PBS, and incubated with goat anti-rabbit antisera (1:20 in PBS) for 30 min; they were rinsed again with PBS and 16 drops of rabbit peroxidase anti-peroxidase complex (PAP) were added. After 20 rain, the PAP was washed away and 16 drops o f 50 m g ~ 3,3'-diaminobenzidine HCI in 0.5 M Tris buffer, p H 7.8 containing 0.03 ~ hydrogen peroxide was added. The dishes were agi-
397
;~ :
~
~i~I it ~¸
Figs. 4-8. Phase micrographs of type II fusiform or ovoid neurons from a 65-day-old culture. Nuclei and other intracellular structures are less conspicuous than in the flattened, multipolar type I cells. Process of these neurons may extend for considerable distances. Possible sites of contact between processes of these cells and with unidentified processes are numerous (~'). x 700.
tated gently on a mechanical shaker for 10 min at room temperature, rinsed in PBS, and mounted with glycogel. For controls, identical cultures were run using antiLHRH which had been absorbed with synthetic LHRH (either 4 or 10 fig LHRH added to 1 ml diluted antisera). RESULTS
Growth and differentiation in culture Dispersed cells attached to the floor of the culture dish within 6-8 h; morpholog-
398 ical development, however, proceeded slowly and 3-4 days passed before processes of any cell types were apparent. Connective tissue cells were the first to exhibit outgrowth of processes; after 10-14 days, they produced a moderate syncytium over the floor of the culture dish. Neuroglia shared the floor of the dish with connective tissue cells; the majority were considered astrocytes on the basis of their relative size and fairly concentric distribution of short processes which branched repeatedly. Neurons at this time were identified tentatively by their highly refractile appearance; they were almost exclusively bipolar with short processes (40-100/~m) at least one of which characteristically attached to the edge of connective tissue cell processes or had grown on to their surface. By the fourth week of culture, a thin carpet of connective tissue covered approximately three-fourths of the surface of the floor of the culture dish and neurons appeared to have used this to greatly extend the field of their processes. Although not all cells could be categorically identified, two types were present in sufficiently consistent numbers to warrant designations as types I and II. Type I cells, 20-25 #m in diameter, were flat, stellate, with large nuclei and single, prominent nucleoli (Figs. 1-3); cytoplasmic granules were conspicuous. Two to 3 primary processes of 1--2/~m diameter exhibited irregular varicosities and thickenings. Type II cells (Figs. 4-8) were smaller and fusiform or ovoid in shape. These cells did not flatten in culture and maintained an identifying refractile halo in phase optics; nucleus, nucleolus and cytoplasmic structures were visible only occasionally (Figs. 6, 8). The majority of these cells were bipolar, although a third primary process was not uncommon (Figs. 6, 7). The thickness of the primary processes was more variable than those of type I cells; some possessed delicate, fairly smooth processes (Fig. 5) which extended, with branching, for several millimeters. Others were beaded with varicosities at loci along the process which gave the appearance of either points of branching or points of contact with other processes (Figs. 4-7). Both of these cell types maintained these characteristics for periods of 50-60 days in culture. After this time, processes became more irregular and gnarled in appearance; cell loss became conspicuous after 75-80 days. Scanning electron micrographs further verified basic differences in the structure of these cell types (Figs. 9, 10). In addition to the 3-4 large primary processes of type I cells seen in phase optics, scanning micrographs revealed that there may be varying numbers of additional, smaller processes which frequently form complex interconnections close to the perikaryon (Fig. 9). The basic bipolar organization of fusiform type II cells is seen to great advantage in scanning micrographs (Fig. 113).
Fluorescence analysis Cultures treated with glyoxylic acid displayed variable amounts of dull green fluorescence within the connective tissue carpet upon which neurons grew. Some large fibroblasts contained this fluorescence although many were non-fluorescent; the latter often contained dull yellow granules. A similar fluorescence was seen in many type I and II neurons. Of considerable significance with respect to the nature of this green fluorescence is the fact that it was qualitatively and quantitatively the same when glyoxylic acid was omitted from the procedure.
399
r 4,
Fig. 9. Scanning electron micrograph of a presumed type I neuron. With scanning electron microscopy this multipolar cell type is consistently observed above the matrix of the tissue culture preparation. This species of cell, significantly different in appearance and configuration than type II bipolar cells, exhibits numerous delicate processes which branch, course and ramify over the surface of the tissue culture, x 1800. Fig. 10. Scanning electron micrograph of Type II bipolar neuron. Both major branches of this cell lype eventually subdivide extensively over considerable distances. × 1600. A brilliant blue fluorescence visually characteristic of catecholamines treated with glyoxylic acid was present in relatively few neurons (Fig. 11). Fluorescence was localized to the perinuclear cytoplasm as well as in some processes and their varicosities. The size, arrangement of processes and general organization of these cells suggest that they are multipolar, type I neurons. Preliminary studies involving the use of desipramine indicate no change in the compartmentalization of either the green background fluorescence or the catecholamine-specific blue fluorescence. After treatment with reserpine, however, blue fluorescence was not seen in type I neurons. Microspectrofluorometric analysis of these two kinds of fluorescence revealed an excitation peak at 370 nm for the green fluorescence and a peak at 410 nm for the blue fluorescence (Fig. 12). Emission spectra revealed a primary emission peak at 480 nm for the blue, type I cells; the greenish fluorescence of connective tissue and
400
Fig. 11. An intense blue catecholarnine fluorescence is seen within a dispersed, organ-cultured hypothalamic neuron. Two fluorescent processes (arrow) appear to emanate from the perikaryon. One of these can be seen to extend a considerable distance across this photornontage. This process appears to branch and contains irregular fluorescent swellings (arrowhead) which are reminiscent of varicosities seen in vivo. ~
