0013.7227/92/1311-0509$03.00/0 Endocrinology Copyright 0 1992 by The Endocrine
Vol. 131, No. 1 Printed in U.S.A.
Society
Immunocytochemical Colocalization Progestin Receptors and Tyrosine Steroid-Treated Monkeys* STEVEN
G. KOHAMA,
FRANCETTA
FREESH,
AND
CYNTHIA
of Hypothalamic Hydroxylase in L. BETHEA
Division of Reproductive Biology and Behavior, Oregon Regional Primate Research Center, Beaverton, Oregon 97006; and the Department of Physiology, Oregon Health Sciences University, Portland, Oregon 97201 ABSTRACT Progesterone (P)-induced PRL secretion in estradiol @-primed monkeys is not due to direct pituitary stimulation, because lactotropes do not express progestin receptors (PR). However, the hypothalamus, particularly the tuberoinfundibular dopaminergic system (TIDA), plays a major role in the regulation of PRL secretion. To determine whether hypothalamic dopamine neurons are progestin target cells, the colocalization of PR and tyrosine hydroxylase (TH), a phenotypic marker of dopaminergic neurons, was examined with double immunocytochemistry. Two methods for visualizing the antigens were applied; the first was a dual peroxidase method, and the second was a peroxidase-alkaline phosphatase method. In addition, the question of whether E induces PR in dopamine neurons was explored. Spayed female monkeys were treated with empty Silastic capsules, E-filled capsules for a period of 28 days, or E capsules supplemented with P capsules for the last 14 days of E treatment. Only the E- plus P-treated monkeys exhibited an increase in serum PRL during the P treatment period. Frontal sections at the level of the optic chiasm and arcuate nucleus were examined for the colocalization of TH and PR. After E treatment, hypothalamic PRpositive cells increased in both intensity and number. Neurons express-
ing both TH and PR were detected in the rostra1 hypothalamus, lateral to the third ventricle (All-rostral) and in a discrete subventricular population (All-subvent). The lateral population continued caudally (All-caudal). The All-subvent population exhibited little steroid regulation. Of the remaining All TH neurons, approximately 20% exhibited PR in the spayed and E-treated groups. Addition of P doubled the percentage of PR-containing TH neurons in this group. Although very few TH-positive neurons in the ventral arcuate nucleus contained PR (AlB-ventral), many double labeled neurons were observed in the dorsal arcuate region (A12-dorsal). Ventral arcuate TIDA neurons were not regulated by steroids, but E plus P increased PR expression in A12dorsal. Double labeled cells were rarely seen in the zona incerta (A13) or the emerging ventral tegmental area (AlO). In summary, P probably does not act directly on ventral arcuate TIDA neurons to stimulate PRL secretion. However, the frequency of PR-positive dopamine neurons in the All-rostral, All-caudal, and A12-dorsal groups increased with E and P treatment. Therefore, the contribution of the PR-positive periventricular dopamine neurons to progestin-stimulated PRL secretion may be important. (Endocrinology 131: 509-517,1992)
A
found no PR in TIDA neurons of guinea pigs or monkeys using double immunocytochemistry (12, 13). Because of the different combinations of techniques and species, it was difficult to conclude that TIDA neurons are target cells for P in primates. A recent study from this laboratory demonstrated a marked induction of PR in the MBH after E treatment of spayed monkeys (14). Although serum PRL increased after P treatment, PR were not down-regulated; this differs from the actions of P in the reproductive tract (15). However, the data are consistent with the hypothesis that a MBH neuronal population(s) containing E-inducible PR is acted upon by P to increase PRL. If dopamine neurons mediate the effect of P on PRL, then they should express PR, which is induced by E treatment, and P would be expected to decrease dopaminergic inhibition. To further test this hypothesis, colocalization of PR and tyrosine hydroxylase (TH), a phenotypic marker of dopaminergic neurons, was examined in the anterior and basal hypothalamus of spayed, E-treated, and Eplus P-treated female monkeys.
LTHOUGH estrogen (E) alone has little effect on serum PRL levels in monkeys, PRL secretion can be stimulated by progesterone (P) after E priming (1). Since PRL does not increase after P treatment alone (2), E is probably required for the induction of progestin receptors (PR). A direct effect of P on the pituitary is not involved, since lactotropes do not express PR (3-5). PRL levels are decreased by the administration of dopamine (6), and the surgically isolated medial basal hypothalamus (MBH) maintains inhibition of PRL secretion in primates (7, 8). Thus, the MBH tuberoinfundibular dopaminergic system (TIDA) has been implicated as a major site for inhibition of PRL secretion (9). We questioned whether TIDA neurons could mediate the effect of P on PRL in primates. PR was found in 40-90% of TIDA neurons in rats using steroid autoradiography combined with immunocytochemistry (3, 10). A minimal percentage of TIDA neurons expressed PR in a study with guinea pigs (1 l), and other studies Received November 26, 1991. Address all correspondence and requests for reprints to: Dr. Cynthia L. Bethea, Oregon Regional Primate Research Center, 505 NW 185th Avenue, Beaverton, Oregon 97006. * Publication 1844 of the Oregon Regional Primate Research Center. This work was supported by NIH Grant HD-17269 (to C.L.B.), Program Project Grant X30-HD-18185, and Grant RR-00163 for operation of the Oregon Regional Primate Research Center.
Materials Animals Tissue
cularis),
and experimental
and Methods design
was acquired from seven cynomolgus for which the steroid induction and
macaques distribution
509
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
(Macaca fasiof PR have
COLOCALIZATION
510
OF MONKEY
been previously documented (14), and from three rhesus macaques (Macaca mulattu). Briefly, after menses each animal was ovariectomized and either received an empty implant (controls) or a 3-cm implant containing crystilline 170.estradiol (E and E plus P groups). The E implant achieves serum E levels of approximately 200 pg/ml for the entire 28.day treatment period (5). After 14 days of E treatment, the E plus P group was supplemented with a 6-cm implant containing crystilline progesterone, producing serum I’ levels of approximately 8 rig/ml for the remaining 14 days of treatment (5). A total of three spayed, three E-treated, and four E- plus P-treated monkeys were examined. Only the E plus P treatment caused an increase in serum PRL (14).
Immunocytochemistry All animals were killed according to procedures recommended by the Panel on Euthanasia of the American Veterinary Association. After steroid treatment, the monkeys were deeply anesthetized, cranially perfused with 500 ml physiological saline, and the brains were removed. The hypothalamic block extended from the optic chiasm to the mamillary bodies, laterally to the hypothalamic sulci, and dorsally a few millimeters above the third ventricle. This large hypothalamic block was then bisected into anterior (rostral) and MBH (caudal) areas. Each block was bath fixed in 2-4% paraformaldehyde in 0.1 M phosphate buffer for approximately 3 h. After fixation, blocks were cryoprotected in a solution of 20% sucrose made up in 0.05 M phosphate buffer. Tissue blocks were frozen on dry ice and stored over liquid nitrogen. Sections, 10 or 15 Km thick, were cut on a cryostat and thaw mounted onto poly-L-lysinecoated slides. Slides were then stored at -70 C. Combined immunocytochemistry for PR and TH was performed using either a dual peroxidase or a sequential peroxidase and alkaline phosphatase reaction. In the dual peroxidase procedure, sections were rinsed in phosphate buffer, treated with normal goat serum (NGS), then incubated overnight at 4 C with a rat monoclonal antibody generated against human PR (JZB39, a gift from Dr. Geoffrey Greene, University of Chicago) at a concentration of 1 &ml (16). This concentration of PR antibody stained nuclei with little cytoplasmic background. After incubation, sections were rinsed in buffer, treated with NGS, rinsed, and then incubated with a 1:50 dilution of biotinylated goat antirat immunoglobulin G (Boehringer-Mannheim, Indianapolis, IN; or Chemicon, Temecula, CA). After another set of rinses, sections were incubated with an avidin-biotin complex (ABC kit, Vector Laboratories, Burlingame, CA). Diaminobenzidine (DAB) combined with 8% NiCl* in Tris buffer solution was used as the peroxidase substrate to visualize PR. This resulted in a black reaction product localized to the neuronal nucleus. Sections were rinsed and then postfixed in 4% paraformaldehyde to inactivate any residual peroxidase activity (17). After fixation, sections were rinsed extensively, treated with NGS, and then incubated overnight at 4 C in a 1:200 dilution of rabbit anti-TH antibody (Chemicon). The next day, development followed the same ABC procedure mentioned above, but used a biotinylated goat antirabbit immunoglobulin G bridged with ABC, followed by development in DAB. A cocktail of 2% normal monkey serum, 2% NGS, and biotin was added during the secondary antibody incubation to lower nonspecific background staining. TH staining with DAB resulted in a brown chromagen that was restricted to neuronal cytoplasm. To verify inactivation of peroxidase activity by fixation, control sections were incubated without substrate during PR labeling, followed by fixation, then addition of substrate. In these cases no PR labeling was seen. Also, some sections were run through the entire double labeling procedure, but without addition of the anti-TH antibody, yielding only PR-positive nuclei. The dual peroxidase method provides very precise localization of chromagen to the antigen of interest, but relies on inactivation of peroxidase from the first reaction. To corroborate the dual peroxidase results, PR was again bridged to ABC-peroxidase and visualized with DAB, but then TH was bridged to ABC-alkaline phosphatase with substrate-IV as the chromagen (Vector Laboratories). The monkey and goat serum cocktail with biotin was also added to decrease background during secondary antibody incubation. Additionally, 1 rnM levamisole was used in buffer rinses before and during substrate-IV development to inhibit endogenous alkaline phosphatase activity. The alkaline phosphatase reaction, which generates a blue deposit, produced a heavier
PR AND
Endo. 1992 Vol 131. No 1
TH
deposition of chromagen than the peroxidase-DAB reaction. Some TH neurons and their nuclei were occluded by the alkaline phosphatase reaction product in areas with a high density of TH-positive fibers and terminals. The immunocytochemical distribution of TH neurons was mapped in cynomolgus sections between the optic chiasm and mamillary bodies for comparison to earlier studies using histofluorescent and immunohistochemical techniques (18-22). To verify that the distribution of TIDA neurons in the cynomolgus MBH was similar in other monkey species, sections from female rhesus monkeys were also immunostained for TH. For double labeling experiments, coronal sections were chosen from both the anterior hypothalamic block, close to the superchiasmatic and supraoptic nucleus, and from the MBH block at the level of the arcuate (ARC) and ventromedial nucleus.
Results The distribution of hypothalamic TH neurons from the optic chiasm to the mamillary bodies is illustrated in Fig. 1. Because of brain orientation during cutting, dorsal areas of the hypothalamus were shifted forward relative to ventral structures compared to their positions in an atlas (23). In rostra1 sections, TH-positive cells were found in high density in the periventricular nucleus (Al l-rostral) and in a discrete population between the third ventricle and the optic chiasm (All-subvent; Fig. 1, level 1). Other TH neurons were found dorsolateral to the third ventricle (A13). Morphologically, the TH-positive neurons located along the lateral walls of the third ventricle were small and oval in shape, with their long
FIG. 1. Camera lucida drawing of the distribution the hypothalamus. Major collections of dopamine fied at five representative levels.
of TH neurons
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
neurons in are identi-
COLOCALIZATION
OF MONKEY
axis oriented parallel to the ventricle. Al3 neurons were larger and had more processesthan the periventricular population. Caudally, the Al3 population increased in density and becamemore lateral in their distribution (Fig. 1, level 2). Small lateral TH populations were also observed near the superchiasmatic nucleus and supraoptic nucleus. The ARC, found in levels 4 and 5, contained TIDA neurons (A12). These neurons were smaller than those found in Al3 and were distributed in two subpopulations, one in the ventral area of the ARC (A12-ventral) and the other in the dorsal area of the ARC located adjacent to the third ventricle (A12-dorsal; Fig. 2). The ventral population of the TIDA neurons was round in shape and morphologically had little ramification of processes.The dorsal TIDA population was more similar to rostra1 populations of TH-positive periventricular neurons, possessingelongated soma and processes oriented parallel to the third ventricle. Also by level 5, some of the A10 neurons were beginning to emerge (Fig. 1). The immunocytochemical localization of PR followed a distribution similar to that described previously (14). Staining was nuclear in nature, and the majority of positive cells exhibited a dense nuclear reaction product, with minimal cytoplasmic staining. The number of PR-positive cells markedly increased in the E-treated animals compared to spayed controls (Fig. 3, A and B), as previously reported (14). TH-positive neurons in the dorsal hypothalamic areas, i.e. the caudal A10 group and the rostra1 Al3 group, rarely expressedPR (Fig. 3, C and D). Steroid replacement with E or E plus P did not induce PR in these neurons (Fig. 3, E and
PR AND
TH
511
F; A10 not shown). The percentage of TH neurons that stained for PR in Al3 equalled 0%, 0%, and 5% in the spayed, E-treated, and E- plus P-treated monkeys, respectively (Table 1). Most of the neurons that label for both PR and TH occurred in the ventral half of the periventricular area. Allrostra1 neurons, which contained PR, were found along the side of the third ventricle (Fig. 4, A-C). Only 12% of these cells expressed PR in the spayed monkey. E treatment increasedcolocalization of TH and PR to 23%, and addition of P further increased the percentage of dual labeled cells to 39% (Table 1). All-rostra1 was underrepresented in the Etreated group of monkeys; however, this same population continued caudally. Similar to that in All-rostral, PR expression in the TH neurons in Al 1-caudal was markedly up-regulated by E plus P. Spayed monkeys exhibited 20% colocalization, and Etreated monkeys exhibited 23% colocalization in this area. However, the percentage of dual labeled cells increased to 44% upon the addition of P (Fig. 5A and Table 1). Since Al l-rostra1 and Al 1-caudal are a continuous population of cells, the cell counts were combined to represent an overall percent change for Al 1-periventricular (rostra1and caudal). The percentage of TH neurons expressing PR in Al l-periventricular equalled 14%, 22%, and 42% in the spayed, Etreated, and E- and P-treated groups, respectively. All-subvent was enriched with dopamine neurons that also contained PR-positive nuclei (Fig. 4D). However, the percentage of double labeled cells showed little steroid reg-
FIG. 2. Composite photomicrograph (x10) of the arcuate of a female rhesus monkey. The left panel depicts half of the ARC stained with cresyl violet to reveal the cell density in this area. The right panel depicts the ARC at the same level which has been immunocytochemically stained for TH. Two subpopulations of TH-positive cells are apparent. The ventral population (curued arrows) rarely exhibited nuclear staining for PR. The dorsal population (straight arrow) frequently exhibited nuclear staining for PR (see Fig. 5).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
FIG. 3. Composite photomicrograph of sections that were double labeled for PR and TH with either the dual peroxidase method or the peroxidase plus alkaline phosphatase method. The straight black arrows designate neurons that contain both TH and PR. The curved black arrows point to TH-positive neurons without PR. Black arrowheads point to neurons that contain PR only. White arrows denote cells that are negative for PR and TH. A, Arcuate-infundibular area from a spay monkey labeled for PR with peroxidase and labeled for TH with alkaline phosphatase. Many TH-positive cells are apparent (blue), but most of the cell nuclei are negative for PR (yellow). Magnification, 40x. B, Arcuate-infundibular area
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION
OF MONKEY
ulation. In the absence of steroids, 38% of the TH neurons contained PR, and with the addition of E or E plus P, the percentage of dual labeled cells equalled 44% and 43%, respectively. The percentage of double labeled neurons in the ventral region of the TIDA system (A12-ventral) was less than 10% (Fig. 5 C and D). Although PR was markedly induced by E treatment in the ARC and ventromedial nuclei (Fig. 3B), the induction of PR did not occur in A12-ventral neurons. However, at the same level in the hypothalamus, TH-positive neurons within the A12-dorsal group do contain PR, and the frequency of dual labeled cells in A12-dorsal increased to 62% with E plus P treatment (Fig. 58 and Table 1). Discussion Induction of PRL requires sequential E and P treatment (l), suggesting that E induction of PR is necessary for P to stimulate PRL. However, pituitary lactotropes lack PR (3-5) because dopamine is a critical PRL inhibitory factor (9), and because the primate TIDA system has been shown to project to the median eminence (22, 24), we sought to determine whether TIDA neurons, which contain estrogen receptors (25), are inducible for PR, which could function in the regulation of PRL. The periventricular hypothalamic populations of TH neurons (Al 1 -rostral, Al 1 -caudal, and A12-dorsal) exhibited steroid regulation of PR expression. Colabeled neurons were observed in the periventricular areas of spayed monkeys at a lower frequency than in E- and E- plus P-treated monkeys. The induction and maintenance of PR expression by steroids implicate these populations in progestin-stimulated PRL secretion. Further experimentation is required to determine the effects of E and E plus P on dopamine production in these populations of neurons. In contrast, the Al l-subvent group manifested a moderate level of colocalization regardless of steroid treatment. However, P alone does not stimulate PRL secretion (2). Together, these observations suggest that A-11 subvent may not contribute to progestin-stimulated PRL secretion. The ARC dopamine neurons in the macaques were found in two subpopulations, a ventral group and a dorsal periventricular group. These two populations exhibited cell-specific expression of PR. That is, the ventral population rarely contained nuclear staining for PR, whereas the dorsal population frequently stained for PR. In addition, PR in the dorsal group appears to be modestly regulated by steroid hormones. Although the percentages of PR-positive TH neurons in A12-dorsal were similar in the spayed and E-treated monkeys, there was an increase in the percentage of TH neurons with PR-positive nuclei after E plus P treatment.
PR AND
TH
513
These observations suggest that P may act directly on the dorsal TIDA neurons to stimulate PRL secretion, but that the ventral TIDA neurons play a lesser or indirect role. The distributions of both TH and PR in this study matched prior descriptions in monkeys (14, 18-20, 26, 27), and the two different double immunocytochemical techniques showed agreement with one another. However, prior reports on the percentage of colocalization are inconsistent across species and techniques. In two separate reports, female rats injected with radiolabeled progestin (R5020 or ORG.2058, respectively) and then immunostained for TH had colocalization levels of 40% (10) or 89% (3) in ARC TIDA neurons, but not in other diencephalon or mesencephalon dopaminergic neurons. In contrast, PR immunoreactivity was reported to be colocalized primarily with glutamic acid decarboxylase, and not with TH, in both monkey (13) and guinea pig ARC (12). Another report with double immunocytochemistry found minimal colocalization in guinea pig ARC (5.513%), but none in the Al 1 periventricular group or the Al3 group (11). We also found minimal expression of PR in the Al3 group and in the ventral TIDA neurons of monkeys which is similar to that in guinea pig, but differs from that in rat. However, unlike previous reports in guinea pig and monkey, we found significant PR expression in both rostra1 and MBH periventricular dopamine neurons (Alland A12dorsal). This could be due to the use of different antibodies or fixation protocols. The coincidence of PR in TH-positive neurons in the primate periventricular areas was more similar to the levels reported in rats using a combination of immunocytochemistry and autoradiographic procedures (3, 10). One group of these PR-expressing TH cells appears to reside within the anatomical borders of the arcuate, but they are morphologically distinct from the ventral arcuate TIDA neurons. Nonetheless, Goldsmith et al. (22) reported that 50-60% of TH neurons in both the periventricular areas and the ARC project to the median eminence in monkeys. Thus, both periventricular and arcuate populations in primates contain neurons that potentially function in the control of PRL secretion, but differ in their ability to respond directly to I’. This is somewhat different from the situation in rodents (24), in which the majority of hypothalamic dopaminergic neurons projecting to the median eminence originate in the ARC (73%) and not from the more rostra1 periventricular areas (19%). Although the double labeled neurons in the anterior periventricular area and the dorsal ARC were separated by several hundred microns, we suggest that they are contiguous populations. First, both sets of neurons are located adjacent to the third ventricle. At the level of the ARC, the third ventricle extends downward as the median eminence expands ventrally. Therefore, the anatomical positions of the
from an E-treated monkey labeled for PR with peroxidase and labeled for TH with alkaline phosphatase. The numerous dark brown nuclei show the marked increase in the number of PR-expressing cells. Several blue TH-positive cells are also apparent with negative nuclei. Magnification, 40x. C, A10 dopamine cells from sections that were double labeled with peroxidase plus alkaline phosphatase (left panel) or dual peroxidase (right panel). PR was rarely expressed in the TH-positive cells of this group regardless of steroid treatment. Magnification, 100x. D, Al3 dopamine cells in a spayed monkey. Sections were processed for PR and TH u&g the sequential peroxidase-alkaline phisphatase method. Magnification, 100X. E, Al3 dopamine cells in an E-treated monkey. Sections were processed for PR and TH using the sequential peroxidase-alkaline phosphatase method. Magnification, X100. F, Al3 dopamine cells in an E- plus P-treated monkey. Both staining methods are represented. PR was rarely observed in the Al3 TH-positive cells regardless of steroid treatment. Magnification, x100.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION
OF MONKEY
PR AND
TH
Endo. 1992 Voll31 *No 1
FIG. 4. Composite photomicrographs of the rostra1 periventricular area (All) from sections that were double labeled for PR and TH using both immunohistochemical methods. Magnification, all x100. A, Lateral periventricular area from spayed monkeys. B, Lateral periventricular area from E-treated monkeys. C, Lateral periventricular area from E- plus P-treated monkeys. D, Subventricular area from E- plus P-treated monkeys.
two groups appear to be located in a similar horizontal plane. Secondly, although the TIDA neurons in the dorsal ARC are located within the infundibular nucleus, the dorsal popula-
tion morphologically looks more like the rostra1periventricular neurons than the rounded ventral TIDA neurons. Finally, the extensive anatomical analysis of Goldsmith et al.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION TABLE
1, Percentage Group
Spayed 15081~cv 15093-c; 13962-rh Mean
(%)
that
contain
All-rostra1
O/60(0)
3/40
(8)
3/21
(14)
12/74
(16)
3/11
(27)
(0)
All-caudal
12
PR AND
TH
515
PR in monkeys
Al3
0
All-r&
+ caud
All-subvent
6/61
(10)
15/85
(18)
24/58 17/34 18/75
20
14
13/52 (25) 65/305 (21)
13/52 (25) 87/402 (21)
23
22
87/165 (53) 30/54 (55) 41/168 (24)
19/48 (39) 107/195 (54) 30/54 (55) 50/239 (21)
AlS-dorsal
(41) (50) (24)
21432 (%)
are given
21/46
22197
(23)
23
0138(0)
19/48 (39) 20/30
(66)
9/71
(13)
5
39
in parentheses.
Cy, Cynomolgus
44 macaque;
42 rh, rhesus
(22) stated that TH neurons in the anterior-ventral periventricular zone were continuous with neurons in the posteriorventral periventricular zone. Additional data suggestthat the dorsal ARC dopaminergic neurons are distinctly different from the ventrolateral TIDA population. In a report showing different ontogenetic origins for the dopamine neurons in the rostra1periventricular region US.ARC, the TIDA neurons in the dorsal ARC were found to be responsive to pregnancy or lactation (28). The ventrolateral TIDA population, however, was found to contain GHreleasing factor (29, 30). Thus, in addition to a significant difference in the expression of PR, these populations appear to have other morphological and functional differences. Steroid stimulation of TIDA function in rats has been shown to be pituitary dependent, with a decline due to hypophysectomy and stimulation by PRL replacement (3133). One study found that both P and E plus P increased median eminence TH activity in hypophysectomized rats, although intact rats treated with E plus P had a much greater responsedue to PRL (33). However, the samestudy reported no change in DOPA accumulation in the median eminence of hypophysectomized rats after E, I’, or E plus P treatment. Thus, although steroids may have a direct effect on some functions of TIDA neurons, PRL appears to be the primary stimulus of this system in rats (31-33). The effect of P on dopamine activity in monkeys has not been studied. Since rats and monkeys differ in several aspectsof hypothalamic function, it may be premature to extrapolate along these lines basedupon the results from rodent studies. There is an intense induction of PR in the MBH after E treatment, and the majority of PR-positive cells contained glutamic acid decarboxylase in guinea pig (12) and monkey (13). Hence, an action of P on y-aminobutyric acid (GABA)ergic interneurons to inhibit dopamine neurons should be considered. However, this speculation is weakened by the biochemistry studies in rats that found no change in dopa-
46/104
(44)
144/483
44
49/93
(52)
36/107
(46)
6/151
46
12/25
0
17/448 (3) 21/120 (17) O/76 (0)
(%)
Percentages
(0)
A12-ventral
O/33 (0)
38
W3 (0)
E- + P-treated 15083~cy 15089~cy 15090~cy 12899-rh Mean
neurons
o/103
E-treated 15088-cy 15095-cy 13866-rh Mean
of TH
OF MONKEY
(34)
(4)
2
(48) (30)
4185 (7) 5164 (7) 4196 (4)
39
6
38/50 (76) 134/262 (51) 64/80 (80) 112/280 (40)
6/135 (4) 12/176 (7) l/21 (4) 11/126 (9)
62
6
43
macaque.
mine turnover with P treatment in vim (33). Immunocytochemistry cannot quantitatively discriminate the concentration of PR in an individual cell. E treatment could increase PR expression over a basal level, and this would not be readily apparent with immunocytochemistry. Thus, the induction of PR by E may be of a greater magnitude than we can currently assessby examining the frequency of positive cells. Future investigation of PR mRNA expression with in situ hybridization will enable a more quantitative conclusion in this regard. In summary, the various populations of dopamine neurons in the macaque hypothalamus exhibit cell-specific expression of PR. PR was colocalized with TH in the periventricular regions, and the percentage of TH cells expressing PR increasedwith steroid treatment. Little or no PR was found in the dorsal hypothalamic A10 or Al3 groups. Only a small percentage (40%) of the ventral ARC TIDA neurons (A12ventral) contained PR, even after PR induction by E treatment. The absenceof PR in a significant number of ventral TIDA neurons suggeststhat P does not act directly on these cells to stimulate PRL secretion. The possibility still exists that I’ acts on GABA-ergic interneurons, which then could inhibit TIDA dopamine neurons. In addition, the periventricular dopamine neurons may play a role, since they contain inducable PR and probably project to the median eminence (22). Further work is required to unravel the roles of the PRpositive GABA interneurons and PR-positive periventricular TH cells in progestin-stimulated PRL releasein primates. References 1. Williams RF, Barber DL, Cowan BD, Lynch A, Marut EL, Hodgen GD 1981 Hyperprolactenemia in monkeys: induction by an estrogen-progesterone synergy. Steroids 38:321-331 2. Williams RF, Gianfortoni JG, Hodgen GD 1985 Hyperprolactinemia induced by an estrogen-progesterone synergy: quantitative and temporal effects of estrogen priming in monkeys. J Clin Endocrinol
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
516
COLOCALIZATION
OF MONKEY
PR AND
TH
Endo. 1992 Vul 131. No 1
FIG. 5. Composite photomicrographs of the MBH at the level of the ARC from sections that were double labeled for PR and TH using both immunohistochemical methods. Magnification, all x100. A, Midperiventricular area from an E- plus P-treated monkey (All-caudal). B, Dorsal ARC from an E- plus P-treated monkey (AlB-dorsal). C, Ventral ARC from an E-treated monkey (A12-ventral). D, Ventral ARC from an E- plus P-treated monkey (A12-ventral).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION
OF MONKEY
Metab 60:126-132 3. Fox SR, Shivers BD, Harlan RE, Pfaff DW, Progesterone receptors in the female rat are localized within dopaminergic neurons of the hypothalamic arcuate nucleus, but not within pituitary lactotrophs. 68th Annual Meeting of The Endocrine Society, Anaheim CA, 1986, p 135 (Abstract 434) 4. Sprangers SA, Brenner RM, Bethea CL 1989 Estrogen and progestin receptor immunocytochemistry in lactotropes versus gonadotropes of monkey pituitary cell cultures. Endocrinology 124:14621470 5. Sprangers SA, West NB, Brenner RM, Bethea CL 1990 Regulation and localization of estrogen and progestin receptors in the pituitary of steroid-treated monkeys, Endocrinology 126:1113-1142 6. Neil1 JD, Frawley S, Plotsky PM, Tindall GT 1981 Dopamine in hypophysial stalk blood of the rhesus monkey and its role in regulating prolactin secretion. Endocrinology 108:489-494 7. Butler WR, Krey LC, Lu K-H, Peckham WD, Knobil E 1975 Surgical disconnection of the medial basal hypothalamus and pituitary function in the rhesus monkey. IV. Prolactin secretion. Endocrinology 96:1099-1105 8. Ferin M, Antunes JL, Zimmerman E, Dyrenfurth I, Frantz AG, Robinson A, Carmel PW 1977 Endocrine function in female rhesus monkeys after hypothalamic disconnection, Endocrinology 101:1611-1620 9. Ben-Jonathan N 1985 Dopamine: a prolactin-inhibiting hormone. Endocr Rev 6:564-589 10. Sar M 1988 Distribution of progestin-concentrating cells in rat brain: colocalization of[3H]ORG.2058, a synthetic progestin, and antibodies to tyrosine hydroxylase in hypothalamus by combined autoradiography and immunocytochemistry. Endocrinology 123: 11 lo1118 11. Blaustein JD, Turcotte JC 1989 A small population of tyrosine hydroxylase-immunoreactive neurons in the guinea-pig arcuate nucleus contains progestin receptor-immunoreactivity. J Neuroendocrinol 1:333-338 12. Brown TJ, MacLusky NJ, Leranth C, Shanabrough M, Naftolin F 1990 Progestin receptor-containing cells in guinea pig hypothalamus: afferent connections, morphological characteristics, and neurotransmitter content. Mol Cell Neurosci 1:58-77 13. Leranth C, MacLusky NJ, Redmond DE, Naftolin F, Transmitter content, afferent connections of progesterone receptor (PR) containing neurons in the primate hypothalamus. 18th Annual Meeting of the Society for Neuroscience, Toronto, Ontario, Canada, 1988, p 1058 (Abstract 426.1) 14. Bethea CL, Fahrenbach WH, Sprangers SA, Freesh F 1992 Immunocytochemical localization of progestin receptors in monkey hypothalamus: effect of estrogen and progestin. Endocrinology 130:895-905 15. Brenner RM, Maslar IA 1988 The primate oviduct and endometrium. In: Knobil E, Neil1 J (eds) The Physiology of Reproduction, Raven Press, New York, pp 303-329 16. Greene GL, Harris K, Bova R, Kinders R, Moore B, Nolan C 1988 Purification of T47D human progesterone receptor and immunochemical characterization with monoclonal antibodies. Mol Endocrinol 2:714-726 17. Harfstrand A, Fuxe K, Cintra A, Agnati LF, Zini I, Wikstrom A-
PR AND
TH
517
C, Okret S, Yu Z-Y, Goldstein M, Steinbusch H, Verhofstad A, Gustafsson J-A 1986 Glucocorticoid receptor immunoreactivity in monoaminergic neurons of rat brain. Proc Nat1 Acad Sci USA 83:9779-9783 18. Schofield SP, Everitt BJ 1981 The organization of catecholaminecontaining neurons in the brain of the rhesus monkey (Macaca mulatta) J Anat 132:391-418 19. Tanaka C, Ishikawa M, Shimada S 1982 Histochemical mapping of catecholaminergic neurons and their ascending fiber pathways in the rhesus monkey brain. Brain Res Bull 9:255-270 20. Felten DL, Sladek Jr JR 1983 Monoamine distribution in primate brain monoaminergic nuclei: anatomy, pathways and local organization Brain Res Bull IO:171-284 21. Thind KK, Goldsmith PC 1989 Corticotropin-releasing factor neurons innervate dopamine neurons in the periventricular hypothalamus of juvenile macaques. Neuroendocrinology 50:351-358 22. Goldsmith PC, Thind KK, Song T, Kim EJ, Boggan JE 1990 Location of the neuroendocrine dopamine neurons in the monkey hypothalamus by retrograde tracing and immunostaining. J Neuroendocrinol 2:169-179 23. Bleier R 1984 The Hypothalamus of the Rhesus Monkey. A Cytoarchitectonic Atlas. University of Wisconsin Press, Madison 24. Kawano H, Daikoku S 1987 Functional topography of the rat hypothalamic dopamine neuron systems: retrograde tracing and immunohistochemical study. J Comp Neurol 265:242-253 25. Sar M 1984 Estradiol is concentrated in tyrosine hydroxylasecontaining neurons of the hypothalamus. Science 223:938-940 26. MacLusky NJ, Lieberburg I, Krey LC, McEwen BS 1980 Progestin receptors in the brain and pituitary of the bonnet monkey (Macca radiutu): differences between the monkey and the rat in the distnbution of progestin receptors, Endocrinology 106:185-191 27. Rees HD, Bonsall RW, Michael RI’ 1985 Localization of the synthetic progestin 3H-ORG 2058 in neurons of the primate brain: evidence for the site of action of progestins on behavior. J Comp Neurol235:336-342 28. Daikoku S, Kawano H, Okamura Y, Tokuzen M, Nagatsu I 1986 Ontogenesis of immunoreactive tyrosine hydroxylase-containing neurons in rat hypothalamus. Dev Brain Res 28:85-98 29. Okamura H, Murakami S, Chihara K, Nagatsu I, Ibata Y 1985 Coexistence of growth hormone releasing factor-like and tyrosine hydroxylase-like immunoreactivities in neurons of the rat arcuate nucleus. Neuroendo 41:177-179 30. Meister B, Hokfelt T, Vale WW, Goldstein M 1985 Growth hormone releasing factor (GRF) and dopamine exist in hypothalamic arcuate neurons. Acta Physiol Stand 124:133-136 31. Demarest KT, Moore KE 1980 Accumulation of L-DOPA in the median eminence: an index of tuberoinfundibular dopaminergic neuron activity. Endocrinology 106:463-468 32. King TS, Steger RW, Morgan WM 1985 Effect of hypophysectomy and subsequent prolactin administration on hypothalamic 5-hydroxytryptamine synthesis in ovariectomized rats. Endocrinology 116:485-491 33. Gonzalez HA, Kedzierski W, Aguila-Mansilla N, Porter JC 1989 Hormonal control of tyrosine hydroxylase in the median eminence: demonstration of a central role for the pituitary gland. Endocrinology 124:2122-2127
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
Vol. 131, No. 1 Printed in U.S.A.
Society
Immunocytochemical Colocalization Progestin Receptors and Tyrosine Steroid-Treated Monkeys* STEVEN
G. KOHAMA,
FRANCETTA
FREESH,
AND
CYNTHIA
of Hypothalamic Hydroxylase in L. BETHEA
Division of Reproductive Biology and Behavior, Oregon Regional Primate Research Center, Beaverton, Oregon 97006; and the Department of Physiology, Oregon Health Sciences University, Portland, Oregon 97201 ABSTRACT Progesterone (P)-induced PRL secretion in estradiol @-primed monkeys is not due to direct pituitary stimulation, because lactotropes do not express progestin receptors (PR). However, the hypothalamus, particularly the tuberoinfundibular dopaminergic system (TIDA), plays a major role in the regulation of PRL secretion. To determine whether hypothalamic dopamine neurons are progestin target cells, the colocalization of PR and tyrosine hydroxylase (TH), a phenotypic marker of dopaminergic neurons, was examined with double immunocytochemistry. Two methods for visualizing the antigens were applied; the first was a dual peroxidase method, and the second was a peroxidase-alkaline phosphatase method. In addition, the question of whether E induces PR in dopamine neurons was explored. Spayed female monkeys were treated with empty Silastic capsules, E-filled capsules for a period of 28 days, or E capsules supplemented with P capsules for the last 14 days of E treatment. Only the E- plus P-treated monkeys exhibited an increase in serum PRL during the P treatment period. Frontal sections at the level of the optic chiasm and arcuate nucleus were examined for the colocalization of TH and PR. After E treatment, hypothalamic PRpositive cells increased in both intensity and number. Neurons express-
ing both TH and PR were detected in the rostra1 hypothalamus, lateral to the third ventricle (All-rostral) and in a discrete subventricular population (All-subvent). The lateral population continued caudally (All-caudal). The All-subvent population exhibited little steroid regulation. Of the remaining All TH neurons, approximately 20% exhibited PR in the spayed and E-treated groups. Addition of P doubled the percentage of PR-containing TH neurons in this group. Although very few TH-positive neurons in the ventral arcuate nucleus contained PR (AlB-ventral), many double labeled neurons were observed in the dorsal arcuate region (A12-dorsal). Ventral arcuate TIDA neurons were not regulated by steroids, but E plus P increased PR expression in A12dorsal. Double labeled cells were rarely seen in the zona incerta (A13) or the emerging ventral tegmental area (AlO). In summary, P probably does not act directly on ventral arcuate TIDA neurons to stimulate PRL secretion. However, the frequency of PR-positive dopamine neurons in the All-rostral, All-caudal, and A12-dorsal groups increased with E and P treatment. Therefore, the contribution of the PR-positive periventricular dopamine neurons to progestin-stimulated PRL secretion may be important. (Endocrinology 131: 509-517,1992)
A
found no PR in TIDA neurons of guinea pigs or monkeys using double immunocytochemistry (12, 13). Because of the different combinations of techniques and species, it was difficult to conclude that TIDA neurons are target cells for P in primates. A recent study from this laboratory demonstrated a marked induction of PR in the MBH after E treatment of spayed monkeys (14). Although serum PRL increased after P treatment, PR were not down-regulated; this differs from the actions of P in the reproductive tract (15). However, the data are consistent with the hypothesis that a MBH neuronal population(s) containing E-inducible PR is acted upon by P to increase PRL. If dopamine neurons mediate the effect of P on PRL, then they should express PR, which is induced by E treatment, and P would be expected to decrease dopaminergic inhibition. To further test this hypothesis, colocalization of PR and tyrosine hydroxylase (TH), a phenotypic marker of dopaminergic neurons, was examined in the anterior and basal hypothalamus of spayed, E-treated, and Eplus P-treated female monkeys.
LTHOUGH estrogen (E) alone has little effect on serum PRL levels in monkeys, PRL secretion can be stimulated by progesterone (P) after E priming (1). Since PRL does not increase after P treatment alone (2), E is probably required for the induction of progestin receptors (PR). A direct effect of P on the pituitary is not involved, since lactotropes do not express PR (3-5). PRL levels are decreased by the administration of dopamine (6), and the surgically isolated medial basal hypothalamus (MBH) maintains inhibition of PRL secretion in primates (7, 8). Thus, the MBH tuberoinfundibular dopaminergic system (TIDA) has been implicated as a major site for inhibition of PRL secretion (9). We questioned whether TIDA neurons could mediate the effect of P on PRL in primates. PR was found in 40-90% of TIDA neurons in rats using steroid autoradiography combined with immunocytochemistry (3, 10). A minimal percentage of TIDA neurons expressed PR in a study with guinea pigs (1 l), and other studies Received November 26, 1991. Address all correspondence and requests for reprints to: Dr. Cynthia L. Bethea, Oregon Regional Primate Research Center, 505 NW 185th Avenue, Beaverton, Oregon 97006. * Publication 1844 of the Oregon Regional Primate Research Center. This work was supported by NIH Grant HD-17269 (to C.L.B.), Program Project Grant X30-HD-18185, and Grant RR-00163 for operation of the Oregon Regional Primate Research Center.
Materials Animals Tissue
cularis),
and experimental
and Methods design
was acquired from seven cynomolgus for which the steroid induction and
macaques distribution
509
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
(Macaca fasiof PR have
COLOCALIZATION
510
OF MONKEY
been previously documented (14), and from three rhesus macaques (Macaca mulattu). Briefly, after menses each animal was ovariectomized and either received an empty implant (controls) or a 3-cm implant containing crystilline 170.estradiol (E and E plus P groups). The E implant achieves serum E levels of approximately 200 pg/ml for the entire 28.day treatment period (5). After 14 days of E treatment, the E plus P group was supplemented with a 6-cm implant containing crystilline progesterone, producing serum I’ levels of approximately 8 rig/ml for the remaining 14 days of treatment (5). A total of three spayed, three E-treated, and four E- plus P-treated monkeys were examined. Only the E plus P treatment caused an increase in serum PRL (14).
Immunocytochemistry All animals were killed according to procedures recommended by the Panel on Euthanasia of the American Veterinary Association. After steroid treatment, the monkeys were deeply anesthetized, cranially perfused with 500 ml physiological saline, and the brains were removed. The hypothalamic block extended from the optic chiasm to the mamillary bodies, laterally to the hypothalamic sulci, and dorsally a few millimeters above the third ventricle. This large hypothalamic block was then bisected into anterior (rostral) and MBH (caudal) areas. Each block was bath fixed in 2-4% paraformaldehyde in 0.1 M phosphate buffer for approximately 3 h. After fixation, blocks were cryoprotected in a solution of 20% sucrose made up in 0.05 M phosphate buffer. Tissue blocks were frozen on dry ice and stored over liquid nitrogen. Sections, 10 or 15 Km thick, were cut on a cryostat and thaw mounted onto poly-L-lysinecoated slides. Slides were then stored at -70 C. Combined immunocytochemistry for PR and TH was performed using either a dual peroxidase or a sequential peroxidase and alkaline phosphatase reaction. In the dual peroxidase procedure, sections were rinsed in phosphate buffer, treated with normal goat serum (NGS), then incubated overnight at 4 C with a rat monoclonal antibody generated against human PR (JZB39, a gift from Dr. Geoffrey Greene, University of Chicago) at a concentration of 1 &ml (16). This concentration of PR antibody stained nuclei with little cytoplasmic background. After incubation, sections were rinsed in buffer, treated with NGS, rinsed, and then incubated with a 1:50 dilution of biotinylated goat antirat immunoglobulin G (Boehringer-Mannheim, Indianapolis, IN; or Chemicon, Temecula, CA). After another set of rinses, sections were incubated with an avidin-biotin complex (ABC kit, Vector Laboratories, Burlingame, CA). Diaminobenzidine (DAB) combined with 8% NiCl* in Tris buffer solution was used as the peroxidase substrate to visualize PR. This resulted in a black reaction product localized to the neuronal nucleus. Sections were rinsed and then postfixed in 4% paraformaldehyde to inactivate any residual peroxidase activity (17). After fixation, sections were rinsed extensively, treated with NGS, and then incubated overnight at 4 C in a 1:200 dilution of rabbit anti-TH antibody (Chemicon). The next day, development followed the same ABC procedure mentioned above, but used a biotinylated goat antirabbit immunoglobulin G bridged with ABC, followed by development in DAB. A cocktail of 2% normal monkey serum, 2% NGS, and biotin was added during the secondary antibody incubation to lower nonspecific background staining. TH staining with DAB resulted in a brown chromagen that was restricted to neuronal cytoplasm. To verify inactivation of peroxidase activity by fixation, control sections were incubated without substrate during PR labeling, followed by fixation, then addition of substrate. In these cases no PR labeling was seen. Also, some sections were run through the entire double labeling procedure, but without addition of the anti-TH antibody, yielding only PR-positive nuclei. The dual peroxidase method provides very precise localization of chromagen to the antigen of interest, but relies on inactivation of peroxidase from the first reaction. To corroborate the dual peroxidase results, PR was again bridged to ABC-peroxidase and visualized with DAB, but then TH was bridged to ABC-alkaline phosphatase with substrate-IV as the chromagen (Vector Laboratories). The monkey and goat serum cocktail with biotin was also added to decrease background during secondary antibody incubation. Additionally, 1 rnM levamisole was used in buffer rinses before and during substrate-IV development to inhibit endogenous alkaline phosphatase activity. The alkaline phosphatase reaction, which generates a blue deposit, produced a heavier
PR AND
Endo. 1992 Vol 131. No 1
TH
deposition of chromagen than the peroxidase-DAB reaction. Some TH neurons and their nuclei were occluded by the alkaline phosphatase reaction product in areas with a high density of TH-positive fibers and terminals. The immunocytochemical distribution of TH neurons was mapped in cynomolgus sections between the optic chiasm and mamillary bodies for comparison to earlier studies using histofluorescent and immunohistochemical techniques (18-22). To verify that the distribution of TIDA neurons in the cynomolgus MBH was similar in other monkey species, sections from female rhesus monkeys were also immunostained for TH. For double labeling experiments, coronal sections were chosen from both the anterior hypothalamic block, close to the superchiasmatic and supraoptic nucleus, and from the MBH block at the level of the arcuate (ARC) and ventromedial nucleus.
Results The distribution of hypothalamic TH neurons from the optic chiasm to the mamillary bodies is illustrated in Fig. 1. Because of brain orientation during cutting, dorsal areas of the hypothalamus were shifted forward relative to ventral structures compared to their positions in an atlas (23). In rostra1 sections, TH-positive cells were found in high density in the periventricular nucleus (Al l-rostral) and in a discrete population between the third ventricle and the optic chiasm (All-subvent; Fig. 1, level 1). Other TH neurons were found dorsolateral to the third ventricle (A13). Morphologically, the TH-positive neurons located along the lateral walls of the third ventricle were small and oval in shape, with their long
FIG. 1. Camera lucida drawing of the distribution the hypothalamus. Major collections of dopamine fied at five representative levels.
of TH neurons
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
neurons in are identi-
COLOCALIZATION
OF MONKEY
axis oriented parallel to the ventricle. Al3 neurons were larger and had more processesthan the periventricular population. Caudally, the Al3 population increased in density and becamemore lateral in their distribution (Fig. 1, level 2). Small lateral TH populations were also observed near the superchiasmatic nucleus and supraoptic nucleus. The ARC, found in levels 4 and 5, contained TIDA neurons (A12). These neurons were smaller than those found in Al3 and were distributed in two subpopulations, one in the ventral area of the ARC (A12-ventral) and the other in the dorsal area of the ARC located adjacent to the third ventricle (A12-dorsal; Fig. 2). The ventral population of the TIDA neurons was round in shape and morphologically had little ramification of processes.The dorsal TIDA population was more similar to rostra1 populations of TH-positive periventricular neurons, possessingelongated soma and processes oriented parallel to the third ventricle. Also by level 5, some of the A10 neurons were beginning to emerge (Fig. 1). The immunocytochemical localization of PR followed a distribution similar to that described previously (14). Staining was nuclear in nature, and the majority of positive cells exhibited a dense nuclear reaction product, with minimal cytoplasmic staining. The number of PR-positive cells markedly increased in the E-treated animals compared to spayed controls (Fig. 3, A and B), as previously reported (14). TH-positive neurons in the dorsal hypothalamic areas, i.e. the caudal A10 group and the rostra1 Al3 group, rarely expressedPR (Fig. 3, C and D). Steroid replacement with E or E plus P did not induce PR in these neurons (Fig. 3, E and
PR AND
TH
511
F; A10 not shown). The percentage of TH neurons that stained for PR in Al3 equalled 0%, 0%, and 5% in the spayed, E-treated, and E- plus P-treated monkeys, respectively (Table 1). Most of the neurons that label for both PR and TH occurred in the ventral half of the periventricular area. Allrostra1 neurons, which contained PR, were found along the side of the third ventricle (Fig. 4, A-C). Only 12% of these cells expressed PR in the spayed monkey. E treatment increasedcolocalization of TH and PR to 23%, and addition of P further increased the percentage of dual labeled cells to 39% (Table 1). All-rostra1 was underrepresented in the Etreated group of monkeys; however, this same population continued caudally. Similar to that in All-rostral, PR expression in the TH neurons in Al 1-caudal was markedly up-regulated by E plus P. Spayed monkeys exhibited 20% colocalization, and Etreated monkeys exhibited 23% colocalization in this area. However, the percentage of dual labeled cells increased to 44% upon the addition of P (Fig. 5A and Table 1). Since Al l-rostra1 and Al 1-caudal are a continuous population of cells, the cell counts were combined to represent an overall percent change for Al 1-periventricular (rostra1and caudal). The percentage of TH neurons expressing PR in Al l-periventricular equalled 14%, 22%, and 42% in the spayed, Etreated, and E- and P-treated groups, respectively. All-subvent was enriched with dopamine neurons that also contained PR-positive nuclei (Fig. 4D). However, the percentage of double labeled cells showed little steroid reg-
FIG. 2. Composite photomicrograph (x10) of the arcuate of a female rhesus monkey. The left panel depicts half of the ARC stained with cresyl violet to reveal the cell density in this area. The right panel depicts the ARC at the same level which has been immunocytochemically stained for TH. Two subpopulations of TH-positive cells are apparent. The ventral population (curued arrows) rarely exhibited nuclear staining for PR. The dorsal population (straight arrow) frequently exhibited nuclear staining for PR (see Fig. 5).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
FIG. 3. Composite photomicrograph of sections that were double labeled for PR and TH with either the dual peroxidase method or the peroxidase plus alkaline phosphatase method. The straight black arrows designate neurons that contain both TH and PR. The curved black arrows point to TH-positive neurons without PR. Black arrowheads point to neurons that contain PR only. White arrows denote cells that are negative for PR and TH. A, Arcuate-infundibular area from a spay monkey labeled for PR with peroxidase and labeled for TH with alkaline phosphatase. Many TH-positive cells are apparent (blue), but most of the cell nuclei are negative for PR (yellow). Magnification, 40x. B, Arcuate-infundibular area
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION
OF MONKEY
ulation. In the absence of steroids, 38% of the TH neurons contained PR, and with the addition of E or E plus P, the percentage of dual labeled cells equalled 44% and 43%, respectively. The percentage of double labeled neurons in the ventral region of the TIDA system (A12-ventral) was less than 10% (Fig. 5 C and D). Although PR was markedly induced by E treatment in the ARC and ventromedial nuclei (Fig. 3B), the induction of PR did not occur in A12-ventral neurons. However, at the same level in the hypothalamus, TH-positive neurons within the A12-dorsal group do contain PR, and the frequency of dual labeled cells in A12-dorsal increased to 62% with E plus P treatment (Fig. 58 and Table 1). Discussion Induction of PRL requires sequential E and P treatment (l), suggesting that E induction of PR is necessary for P to stimulate PRL. However, pituitary lactotropes lack PR (3-5) because dopamine is a critical PRL inhibitory factor (9), and because the primate TIDA system has been shown to project to the median eminence (22, 24), we sought to determine whether TIDA neurons, which contain estrogen receptors (25), are inducible for PR, which could function in the regulation of PRL. The periventricular hypothalamic populations of TH neurons (Al 1 -rostral, Al 1 -caudal, and A12-dorsal) exhibited steroid regulation of PR expression. Colabeled neurons were observed in the periventricular areas of spayed monkeys at a lower frequency than in E- and E- plus P-treated monkeys. The induction and maintenance of PR expression by steroids implicate these populations in progestin-stimulated PRL secretion. Further experimentation is required to determine the effects of E and E plus P on dopamine production in these populations of neurons. In contrast, the Al l-subvent group manifested a moderate level of colocalization regardless of steroid treatment. However, P alone does not stimulate PRL secretion (2). Together, these observations suggest that A-11 subvent may not contribute to progestin-stimulated PRL secretion. The ARC dopamine neurons in the macaques were found in two subpopulations, a ventral group and a dorsal periventricular group. These two populations exhibited cell-specific expression of PR. That is, the ventral population rarely contained nuclear staining for PR, whereas the dorsal population frequently stained for PR. In addition, PR in the dorsal group appears to be modestly regulated by steroid hormones. Although the percentages of PR-positive TH neurons in A12-dorsal were similar in the spayed and E-treated monkeys, there was an increase in the percentage of TH neurons with PR-positive nuclei after E plus P treatment.
PR AND
TH
513
These observations suggest that P may act directly on the dorsal TIDA neurons to stimulate PRL secretion, but that the ventral TIDA neurons play a lesser or indirect role. The distributions of both TH and PR in this study matched prior descriptions in monkeys (14, 18-20, 26, 27), and the two different double immunocytochemical techniques showed agreement with one another. However, prior reports on the percentage of colocalization are inconsistent across species and techniques. In two separate reports, female rats injected with radiolabeled progestin (R5020 or ORG.2058, respectively) and then immunostained for TH had colocalization levels of 40% (10) or 89% (3) in ARC TIDA neurons, but not in other diencephalon or mesencephalon dopaminergic neurons. In contrast, PR immunoreactivity was reported to be colocalized primarily with glutamic acid decarboxylase, and not with TH, in both monkey (13) and guinea pig ARC (12). Another report with double immunocytochemistry found minimal colocalization in guinea pig ARC (5.513%), but none in the Al 1 periventricular group or the Al3 group (11). We also found minimal expression of PR in the Al3 group and in the ventral TIDA neurons of monkeys which is similar to that in guinea pig, but differs from that in rat. However, unlike previous reports in guinea pig and monkey, we found significant PR expression in both rostra1 and MBH periventricular dopamine neurons (Alland A12dorsal). This could be due to the use of different antibodies or fixation protocols. The coincidence of PR in TH-positive neurons in the primate periventricular areas was more similar to the levels reported in rats using a combination of immunocytochemistry and autoradiographic procedures (3, 10). One group of these PR-expressing TH cells appears to reside within the anatomical borders of the arcuate, but they are morphologically distinct from the ventral arcuate TIDA neurons. Nonetheless, Goldsmith et al. (22) reported that 50-60% of TH neurons in both the periventricular areas and the ARC project to the median eminence in monkeys. Thus, both periventricular and arcuate populations in primates contain neurons that potentially function in the control of PRL secretion, but differ in their ability to respond directly to I’. This is somewhat different from the situation in rodents (24), in which the majority of hypothalamic dopaminergic neurons projecting to the median eminence originate in the ARC (73%) and not from the more rostra1 periventricular areas (19%). Although the double labeled neurons in the anterior periventricular area and the dorsal ARC were separated by several hundred microns, we suggest that they are contiguous populations. First, both sets of neurons are located adjacent to the third ventricle. At the level of the ARC, the third ventricle extends downward as the median eminence expands ventrally. Therefore, the anatomical positions of the
from an E-treated monkey labeled for PR with peroxidase and labeled for TH with alkaline phosphatase. The numerous dark brown nuclei show the marked increase in the number of PR-expressing cells. Several blue TH-positive cells are also apparent with negative nuclei. Magnification, 40x. C, A10 dopamine cells from sections that were double labeled with peroxidase plus alkaline phosphatase (left panel) or dual peroxidase (right panel). PR was rarely expressed in the TH-positive cells of this group regardless of steroid treatment. Magnification, 100x. D, Al3 dopamine cells in a spayed monkey. Sections were processed for PR and TH u&g the sequential peroxidase-alkaline phisphatase method. Magnification, 100X. E, Al3 dopamine cells in an E-treated monkey. Sections were processed for PR and TH using the sequential peroxidase-alkaline phosphatase method. Magnification, X100. F, Al3 dopamine cells in an E- plus P-treated monkey. Both staining methods are represented. PR was rarely observed in the Al3 TH-positive cells regardless of steroid treatment. Magnification, x100.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION
OF MONKEY
PR AND
TH
Endo. 1992 Voll31 *No 1
FIG. 4. Composite photomicrographs of the rostra1 periventricular area (All) from sections that were double labeled for PR and TH using both immunohistochemical methods. Magnification, all x100. A, Lateral periventricular area from spayed monkeys. B, Lateral periventricular area from E-treated monkeys. C, Lateral periventricular area from E- plus P-treated monkeys. D, Subventricular area from E- plus P-treated monkeys.
two groups appear to be located in a similar horizontal plane. Secondly, although the TIDA neurons in the dorsal ARC are located within the infundibular nucleus, the dorsal popula-
tion morphologically looks more like the rostra1periventricular neurons than the rounded ventral TIDA neurons. Finally, the extensive anatomical analysis of Goldsmith et al.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION TABLE
1, Percentage Group
Spayed 15081~cv 15093-c; 13962-rh Mean
(%)
that
contain
All-rostra1
O/60(0)
3/40
(8)
3/21
(14)
12/74
(16)
3/11
(27)
(0)
All-caudal
12
PR AND
TH
515
PR in monkeys
Al3
0
All-r&
+ caud
All-subvent
6/61
(10)
15/85
(18)
24/58 17/34 18/75
20
14
13/52 (25) 65/305 (21)
13/52 (25) 87/402 (21)
23
22
87/165 (53) 30/54 (55) 41/168 (24)
19/48 (39) 107/195 (54) 30/54 (55) 50/239 (21)
AlS-dorsal
(41) (50) (24)
21432 (%)
are given
21/46
22197
(23)
23
0138(0)
19/48 (39) 20/30
(66)
9/71
(13)
5
39
in parentheses.
Cy, Cynomolgus
44 macaque;
42 rh, rhesus
(22) stated that TH neurons in the anterior-ventral periventricular zone were continuous with neurons in the posteriorventral periventricular zone. Additional data suggestthat the dorsal ARC dopaminergic neurons are distinctly different from the ventrolateral TIDA population. In a report showing different ontogenetic origins for the dopamine neurons in the rostra1periventricular region US.ARC, the TIDA neurons in the dorsal ARC were found to be responsive to pregnancy or lactation (28). The ventrolateral TIDA population, however, was found to contain GHreleasing factor (29, 30). Thus, in addition to a significant difference in the expression of PR, these populations appear to have other morphological and functional differences. Steroid stimulation of TIDA function in rats has been shown to be pituitary dependent, with a decline due to hypophysectomy and stimulation by PRL replacement (3133). One study found that both P and E plus P increased median eminence TH activity in hypophysectomized rats, although intact rats treated with E plus P had a much greater responsedue to PRL (33). However, the samestudy reported no change in DOPA accumulation in the median eminence of hypophysectomized rats after E, I’, or E plus P treatment. Thus, although steroids may have a direct effect on some functions of TIDA neurons, PRL appears to be the primary stimulus of this system in rats (31-33). The effect of P on dopamine activity in monkeys has not been studied. Since rats and monkeys differ in several aspectsof hypothalamic function, it may be premature to extrapolate along these lines basedupon the results from rodent studies. There is an intense induction of PR in the MBH after E treatment, and the majority of PR-positive cells contained glutamic acid decarboxylase in guinea pig (12) and monkey (13). Hence, an action of P on y-aminobutyric acid (GABA)ergic interneurons to inhibit dopamine neurons should be considered. However, this speculation is weakened by the biochemistry studies in rats that found no change in dopa-
46/104
(44)
144/483
44
49/93
(52)
36/107
(46)
6/151
46
12/25
0
17/448 (3) 21/120 (17) O/76 (0)
(%)
Percentages
(0)
A12-ventral
O/33 (0)
38
W3 (0)
E- + P-treated 15083~cy 15089~cy 15090~cy 12899-rh Mean
neurons
o/103
E-treated 15088-cy 15095-cy 13866-rh Mean
of TH
OF MONKEY
(34)
(4)
2
(48) (30)
4185 (7) 5164 (7) 4196 (4)
39
6
38/50 (76) 134/262 (51) 64/80 (80) 112/280 (40)
6/135 (4) 12/176 (7) l/21 (4) 11/126 (9)
62
6
43
macaque.
mine turnover with P treatment in vim (33). Immunocytochemistry cannot quantitatively discriminate the concentration of PR in an individual cell. E treatment could increase PR expression over a basal level, and this would not be readily apparent with immunocytochemistry. Thus, the induction of PR by E may be of a greater magnitude than we can currently assessby examining the frequency of positive cells. Future investigation of PR mRNA expression with in situ hybridization will enable a more quantitative conclusion in this regard. In summary, the various populations of dopamine neurons in the macaque hypothalamus exhibit cell-specific expression of PR. PR was colocalized with TH in the periventricular regions, and the percentage of TH cells expressing PR increasedwith steroid treatment. Little or no PR was found in the dorsal hypothalamic A10 or Al3 groups. Only a small percentage (40%) of the ventral ARC TIDA neurons (A12ventral) contained PR, even after PR induction by E treatment. The absenceof PR in a significant number of ventral TIDA neurons suggeststhat P does not act directly on these cells to stimulate PRL secretion. The possibility still exists that I’ acts on GABA-ergic interneurons, which then could inhibit TIDA dopamine neurons. In addition, the periventricular dopamine neurons may play a role, since they contain inducable PR and probably project to the median eminence (22). Further work is required to unravel the roles of the PRpositive GABA interneurons and PR-positive periventricular TH cells in progestin-stimulated PRL releasein primates. References 1. Williams RF, Barber DL, Cowan BD, Lynch A, Marut EL, Hodgen GD 1981 Hyperprolactenemia in monkeys: induction by an estrogen-progesterone synergy. Steroids 38:321-331 2. Williams RF, Gianfortoni JG, Hodgen GD 1985 Hyperprolactinemia induced by an estrogen-progesterone synergy: quantitative and temporal effects of estrogen priming in monkeys. J Clin Endocrinol
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
516
COLOCALIZATION
OF MONKEY
PR AND
TH
Endo. 1992 Vul 131. No 1
FIG. 5. Composite photomicrographs of the MBH at the level of the ARC from sections that were double labeled for PR and TH using both immunohistochemical methods. Magnification, all x100. A, Midperiventricular area from an E- plus P-treated monkey (All-caudal). B, Dorsal ARC from an E- plus P-treated monkey (AlB-dorsal). C, Ventral ARC from an E-treated monkey (A12-ventral). D, Ventral ARC from an E- plus P-treated monkey (A12-ventral).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
COLOCALIZATION
OF MONKEY
Metab 60:126-132 3. Fox SR, Shivers BD, Harlan RE, Pfaff DW, Progesterone receptors in the female rat are localized within dopaminergic neurons of the hypothalamic arcuate nucleus, but not within pituitary lactotrophs. 68th Annual Meeting of The Endocrine Society, Anaheim CA, 1986, p 135 (Abstract 434) 4. Sprangers SA, Brenner RM, Bethea CL 1989 Estrogen and progestin receptor immunocytochemistry in lactotropes versus gonadotropes of monkey pituitary cell cultures. Endocrinology 124:14621470 5. Sprangers SA, West NB, Brenner RM, Bethea CL 1990 Regulation and localization of estrogen and progestin receptors in the pituitary of steroid-treated monkeys, Endocrinology 126:1113-1142 6. Neil1 JD, Frawley S, Plotsky PM, Tindall GT 1981 Dopamine in hypophysial stalk blood of the rhesus monkey and its role in regulating prolactin secretion. Endocrinology 108:489-494 7. Butler WR, Krey LC, Lu K-H, Peckham WD, Knobil E 1975 Surgical disconnection of the medial basal hypothalamus and pituitary function in the rhesus monkey. IV. Prolactin secretion. Endocrinology 96:1099-1105 8. Ferin M, Antunes JL, Zimmerman E, Dyrenfurth I, Frantz AG, Robinson A, Carmel PW 1977 Endocrine function in female rhesus monkeys after hypothalamic disconnection, Endocrinology 101:1611-1620 9. Ben-Jonathan N 1985 Dopamine: a prolactin-inhibiting hormone. Endocr Rev 6:564-589 10. Sar M 1988 Distribution of progestin-concentrating cells in rat brain: colocalization of[3H]ORG.2058, a synthetic progestin, and antibodies to tyrosine hydroxylase in hypothalamus by combined autoradiography and immunocytochemistry. Endocrinology 123: 11 lo1118 11. Blaustein JD, Turcotte JC 1989 A small population of tyrosine hydroxylase-immunoreactive neurons in the guinea-pig arcuate nucleus contains progestin receptor-immunoreactivity. J Neuroendocrinol 1:333-338 12. Brown TJ, MacLusky NJ, Leranth C, Shanabrough M, Naftolin F 1990 Progestin receptor-containing cells in guinea pig hypothalamus: afferent connections, morphological characteristics, and neurotransmitter content. Mol Cell Neurosci 1:58-77 13. Leranth C, MacLusky NJ, Redmond DE, Naftolin F, Transmitter content, afferent connections of progesterone receptor (PR) containing neurons in the primate hypothalamus. 18th Annual Meeting of the Society for Neuroscience, Toronto, Ontario, Canada, 1988, p 1058 (Abstract 426.1) 14. Bethea CL, Fahrenbach WH, Sprangers SA, Freesh F 1992 Immunocytochemical localization of progestin receptors in monkey hypothalamus: effect of estrogen and progestin. Endocrinology 130:895-905 15. Brenner RM, Maslar IA 1988 The primate oviduct and endometrium. In: Knobil E, Neil1 J (eds) The Physiology of Reproduction, Raven Press, New York, pp 303-329 16. Greene GL, Harris K, Bova R, Kinders R, Moore B, Nolan C 1988 Purification of T47D human progesterone receptor and immunochemical characterization with monoclonal antibodies. Mol Endocrinol 2:714-726 17. Harfstrand A, Fuxe K, Cintra A, Agnati LF, Zini I, Wikstrom A-
PR AND
TH
517
C, Okret S, Yu Z-Y, Goldstein M, Steinbusch H, Verhofstad A, Gustafsson J-A 1986 Glucocorticoid receptor immunoreactivity in monoaminergic neurons of rat brain. Proc Nat1 Acad Sci USA 83:9779-9783 18. Schofield SP, Everitt BJ 1981 The organization of catecholaminecontaining neurons in the brain of the rhesus monkey (Macaca mulatta) J Anat 132:391-418 19. Tanaka C, Ishikawa M, Shimada S 1982 Histochemical mapping of catecholaminergic neurons and their ascending fiber pathways in the rhesus monkey brain. Brain Res Bull 9:255-270 20. Felten DL, Sladek Jr JR 1983 Monoamine distribution in primate brain monoaminergic nuclei: anatomy, pathways and local organization Brain Res Bull IO:171-284 21. Thind KK, Goldsmith PC 1989 Corticotropin-releasing factor neurons innervate dopamine neurons in the periventricular hypothalamus of juvenile macaques. Neuroendocrinology 50:351-358 22. Goldsmith PC, Thind KK, Song T, Kim EJ, Boggan JE 1990 Location of the neuroendocrine dopamine neurons in the monkey hypothalamus by retrograde tracing and immunostaining. J Neuroendocrinol 2:169-179 23. Bleier R 1984 The Hypothalamus of the Rhesus Monkey. A Cytoarchitectonic Atlas. University of Wisconsin Press, Madison 24. Kawano H, Daikoku S 1987 Functional topography of the rat hypothalamic dopamine neuron systems: retrograde tracing and immunohistochemical study. J Comp Neurol 265:242-253 25. Sar M 1984 Estradiol is concentrated in tyrosine hydroxylasecontaining neurons of the hypothalamus. Science 223:938-940 26. MacLusky NJ, Lieberburg I, Krey LC, McEwen BS 1980 Progestin receptors in the brain and pituitary of the bonnet monkey (Macca radiutu): differences between the monkey and the rat in the distnbution of progestin receptors, Endocrinology 106:185-191 27. Rees HD, Bonsall RW, Michael RI’ 1985 Localization of the synthetic progestin 3H-ORG 2058 in neurons of the primate brain: evidence for the site of action of progestins on behavior. J Comp Neurol235:336-342 28. Daikoku S, Kawano H, Okamura Y, Tokuzen M, Nagatsu I 1986 Ontogenesis of immunoreactive tyrosine hydroxylase-containing neurons in rat hypothalamus. Dev Brain Res 28:85-98 29. Okamura H, Murakami S, Chihara K, Nagatsu I, Ibata Y 1985 Coexistence of growth hormone releasing factor-like and tyrosine hydroxylase-like immunoreactivities in neurons of the rat arcuate nucleus. Neuroendo 41:177-179 30. Meister B, Hokfelt T, Vale WW, Goldstein M 1985 Growth hormone releasing factor (GRF) and dopamine exist in hypothalamic arcuate neurons. Acta Physiol Stand 124:133-136 31. Demarest KT, Moore KE 1980 Accumulation of L-DOPA in the median eminence: an index of tuberoinfundibular dopaminergic neuron activity. Endocrinology 106:463-468 32. King TS, Steger RW, Morgan WM 1985 Effect of hypophysectomy and subsequent prolactin administration on hypothalamic 5-hydroxytryptamine synthesis in ovariectomized rats. Endocrinology 116:485-491 33. Gonzalez HA, Kedzierski W, Aguila-Mansilla N, Porter JC 1989 Hormonal control of tyrosine hydroxylase in the median eminence: demonstration of a central role for the pituitary gland. Endocrinology 124:2122-2127
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 October 2015. at 23:06 For personal use only. No other uses without permission. . All rights reserved.
