0013-7227/91/1281 -0459$02.00/0 Endocrinology Copyright© 1991 by The Endocrine Society
Vol. 128, No. 1 Printed in U.S.A.
Single Cell Levels of Hypothalamic Messenger Ribonucleic Acid Encoding Luteinizing Hormone Releasing Hormone in Intact, Castrated, and Hyperprolactinemic Male Rats* MICHAEL SELMANOFF, CHRISTINE SHUf, SANDRA L. PETERSEN, CHARLES A. BARRACLOUGH, AND R. THOMAS ZOELLER Center for Studies in Reproduction, Department of Physiology, University of Maryland School of Medicine (M.S., C.S., C.A.B.), Baltimore, Maryland 21201; and the Department of Anatomy and Neurobiology, University of Missouri School of Medicine (S.L.P., R.T.Z.), Columbia, Missouri 65212
ABSTRACT. We have examined the changes that occur in neuronal expression of LHRH mRNA in response to castration and hyperprolactinemia in male rats. Single cell levels of LHRH mRNA were determined by quantitative in situ hybridization histochemistry using an 35S-labeled synthetic 48-base oligodeoxynucleotide probe and quantitative autoradiography. Nine days postcastration, a 10.4-fold increase in mean plasma LH titers was observed which was associated with significantly increased LHRH mRNA in rostral hypothalamic neuronal cell bodies. Both increases were blocked in rats rendered hyperprolactinemic by the presence of the 7315a PRL-secreting pituitary tumor.
The location and number of neurons expressing LHRH mRNA were unchanged, indicating that these differences were attributable to amounts of mRNA expressed per neuron. Experimental differences occurred in LHRH perikarya situated throughout the rostral hypothalamus from the organum vasculosum of the lamina terminalis to the caudal extent of the medial preoptic nucleus. These results suggest that gonadal steroids and PRL are involved, either directly or indirectly, in regulating the biosynthesis of LHRH in the rostral hypothalamus. (Endocrinology 128: 459-466, 1991)
T
perfusion (10,11) studies have not detected mean LHRH increases, and this issue has become controversial. Evidence that PRL inhibits LH release is suggested by endocrine states in which an inverse relationship exists between these two pituitary hormones (e.g. postpartum lactation, pregnancy, and pseudopregnancy) (12). The antigonadotropic actions of PRL may occur at the level of the brain, pituitary, and/or gonad in male and female mammals (13). Experimentally induced hyperprolactinemic states employing PRL-secreting tumors, ectopic pituitary glands, or PRL injections inhibit circulating LH levels under several circumstances. Profound, unremitting inhibition of the postcastration LH rise occurs in the presence of transplantable 7315a, 7315b, or MtTW15 PRL-secreting pituitary tumors (14-17). This inhibitory effect in gonadectomized rats may be mediated within the hypothalamus by decreasing LHRH secretion and/or at the pituitary level by altering pituitary responsiveness to LHRH. Increasing evidence indicates that changes in the activity of peptidergic neurons is related to alterations in cellular levels of prepropeptide mRNA (18-20). In situ hybridization histochemistry allows one to identify and
HE DECAPEPTIDE LHRH stimulates the release of LH and FSH from the anterior pituitary gland (1). It has been thought that removal of the gonadal steroid negative feedback effect results in the hypothalamic mechanisms mediating LHRH release becoming chronically active. As a consequence, increased LHRH is detected in hypophyseal portal blood (2-5). In male rats, the rapidly increased mean plasma LH levels (6) are produced by increases in both the frequency and amplitude of pulsatile LH secretion (6, 7). Similarly, LHRH pulses detected in push-pull perfusates of the anterior pituitary gland in unanesthetized castrated male rats have increased amplitude and frequency (8). However, some portal blood (9) and hypothalamic push-pull Received May 14, 1990. Address all correspondence and requests for reprints to: Dr. Michael Selmanoff, Department of Physiology, University of Maryland School of Medicine, 660 West Redwood Street, Baltimore, Maryland 21201. * This work was supported in part by a stipend from the Foundation of Medical Research, France (to C.S.), and grants HD-21351 (to M.S.) and HD-02138 (to C.A.B.) from the NIH. A preliminary report of these data was made at the 18th Annual Meeting of the Society for Neuroscience, Toronto, Canada, 1988, vol 18:1069. t Present address: Roussel-UCLAF Pharmaceutical Co., Romainville, France.
459
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 August 2016. at 19:23 For personal use only. No other uses without permission. . All rights reserved.
HYPOTHALAMIC mRNA ENCODING LHRH IN RATS
460
quantify mRNA transcripts in single neurons. Hence, the effects of physiological variables on gene expression in single neurons can be evaluated. In the present study we used this approach to determine the effects of castration and elevated plasma PRL levels on regional changes in cellular levels of LHRH mRNA. We provide clear evidence that cellular levels of LHRH mRNA are increased after castration and that this increase is blocked in animals bearing a PRL-secreting tumor.
Materials and Methods Animals and tissue preparation Adult male Buffalo rats (Harlan Sprague-Dawley, Indianapolis, IN), weighing 300-350 g, were maintained on a 14-h light, 10-h dark schedule, with food and water provided ad libitum. Rats were inoculated sc with a suspension of minced 7315a tumor tissue, as previously described (21). Six weeks later, when the tumor volume was 35-50 cm3, these rats were orchidectomized (Orch) under brief ether anesthesia (n = 7). Age-matched intact (n = 9) and orchidectomized (n = 9) rats comprised the other two experimental groups. Nine days postcastration, rats were decapitated between 0900-1000 h. Brains were rapidly removed, frozen on dry ice, and stored, tightly wrapped in Parafilm, at -70 C. Trunk blood was collected, the cells were separated by centrifugation, and the plasma was stored at —20 C until RIA. Frozen 12 nm sections were prepared from the bed nucleus of the diagonal band of Broca (-A7860; DBB) through the caudal extent of the medial preoptic nucleus (~A6460; MPN). Atlas coordinates and coronal plane of section are from the atlas of Konig and Klippel (22). One of three sections (save one, discard two) was thaw-mounted on acid-washed chromalum- and gelatin-coated slides (two sections per slide) through this 1400 jim extent of the rostral hypothalamus. Thus, three sets of slides per animal were prepared and stored at —70 C until hybridization was carried out. RIA Plasma LH concentrations were determined by the heterologous ovine:ovine double antibody method using the NIDDK rat LH RP-1 standard, GDN #15 antibody, and LER-1374A ovine LH preparation for radioiodination (23). Plasma PRL was assayed by a homologous double antibody method, as previously described (24), using NIDDK rat PRL reagents provided by Dr. Albert Parlow through the National Hormone and Pituitary Program. The rat PRL standard used was rat PRL RP-3 (biopotency = 2.8 times NIDDK rat PRL RP-1). In situ hybridization In situ hybridization histochemistry was performed by a modification (25) of the method described by Mason and coworkers (26). The LHRH probe was a 48-base oligodeoxynucleotide complementary to the LHRH coding region of the rat LHRH cDNA (bases 102-149) described by Adelman and coworkers (27). This 48mer was synthesized on an Applied Bio-
Endo • 1991 Voll28«No 1
systems 380A DNA synthesizer and purified on a preparative denaturing acrylamide gel (28). The probe (0.1 MM) was 3' end labeled with [35S]dATP (1.0 MM; 1300 Ci/mmol; New England Nuclear, Boston, MA) and terminal deoxynucleotidyl transferase (100 U; Bethesda Research Laboratories, Gaithersburg, MD) to a specific activity of about 7000 Ci/mmol. Prehybridization treatment consisted of warming the sections to room temperature, after which they were fixed for 5 min in 4% formaldehyde in 0.12 M sodium PBS (pH 7.4), rinsed in 2 x SSC (1 x SSC = 0.15 M NaCl, 0.015 M sodium citrate buffer, pH 7.2), and then acetylated for 10 min in 0.25% acetic anhydride in 0.1 M triethanolamine hydrochloride-0.9% NaCl (pH 8.0). Sections were dehydrated through an ethanol series (70-100%), delipidated in chloroform, rehydrated in 95% ethanol, and air dried. Labeled probe was applied in a saturating concentration to sections in 45 n\ hybridization buffer. This buffer contained 4 x SSC, 50% (vol/vol) formamide, 10% (wt/ vol) dextran sulfate, 250 yug/ml yeast transfer RNA, 500 ^tg/ml sheared single stranded salmon sperm DNA, 1 X Denhardt's solution (0.02% Ficoll, 0.02% polyvinylpyrrolidone, and 0.02% BSA), and 50 mM fresh dithiothreitol. Hybridization proceeded under Parafilm coverslips, overnight, at 37 C in humid chambers. Sections were washed four times under stringent conditions (2 x SSC-50% formamide) at 40 C (-19 C below Tm, the mean melting temperature for this cDNA-mRNA hybrid) for 15 min each, and then twice under less stringent conditions (1 x SSC) at 22 C for 60 min each. After the last wash, slides were dipped briefly in water and then in 70% ethanol and dried. Slides were dipped in fresh Kodak NTB-3 nuclear track emulsion (Eastman Kodak, Rochester, NY), exposed for 5 days at 4 C, and developed, and the cell nuclei were counterstained with toluidine blue. Several findings indicate that the 35S-labeled oligonucleotide probe hybridized specifically to mRNA encoding the LHRH precursor: 1) labeled cells display the same distribution as reported by immunohistochemical staining of the decapeptide (29, 30); 2) this LHRH probe and a 48-mer complementary to the GAP portion of the LHRH precursor (bases 218-265) label the same cells in adjacent sections (28); 3) hybridization of adjacent sections with a vasopressin probe (31) yielded a labeled hypothalamic cell population distinct from that labeled with the LHRH probe and did not label cells in regions labeled by the LHRH probe; 4) no labeled cells were observed when sections were hybridized with probe diluent alone as a control for positive chemography, and 5), a sense oligomer, directed to the complementary DNA strand, revealed no labeling in adjacent sections. Image analysis Tissue sections were matched anatomically between animals, and 10 sections/animal were analyzed. The autoradiographic hybridization signal was quantitated in a Bioquant Image Analysis System IV (R & M Biometrics, Inc., Nashville, TN) consisting of an AT & T TARGA M8 frame grabber and AMETEK Hipad digitizing tablet attached to an IBM XT computer. Video images were obtained by a DAGE-MTI 65 video camera attached to a Leitz Laborlux D photomicroscope (Rockleigh, NJ) equipped with a X40 epiillumination darkfield
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 August 2016. at 19:23 For personal use only. No other uses without permission. . All rights reserved.
HYPOTHALAMIC mRNA ENCODING LHRH IN RATS condensor. Illumination intensity was controlled by a 3-12V DC-regulated stabilized power supply. Images were displayed on an Analog RGB monitor. In this study two separate hybridizations were performed on two sets of sections (10-14 sections/animal) from the same animals. The second hybridization was performed for three reasons: 1) to confirm the original finding, 2) to use a different image analysis approach, and 3) to increase the number of sections from each animal for the regional data analysis. The basic experimental results were the same with both hybridizations and we present data from the second, more extensive, hybridization. Image analysis of labeled cells from the first hybridization (total number of cells = 2075) was performed as follows. Clusters of grains were determined to be over cell bodies using brightfield optics at X400 magnification. Then, under darkfield optics, grain clusters associated with individual cells were moved to the center of the video screen, brought into exact microscope focus and circumscribed by the operator. The average optical density and area over which grains are distributed were then obtained. The average cellular labeling intensity was 50 times background (average background optical density reading = 2.5; average cell reading = 125.4). Analyzed in this manner, the optical brightness of each pixel in the image analysis system is assigned a number based on a gray level scale of 0-255 in the Bioquant IV system. We constructed a standard curve relating this gray level scale to brain paste standards (18, 25, 32). These standards were prepared by adding known amounts of 35S to brain homogenate as previously described (25, 33), freezing them, sectioning them at 12 fiva, processing them for autoradiography, and reading them in the image analysis system. Optical density values could then be converted to disintegrations per min of 35S. The standard curve verified that the optical density of the hybridization signal over labeled cells was on the linear portion of the curve relating optical density and radioactivity and was, thus, below saturation of the nuclear emulsion (32). Labeled cells from the second hybridization (total number of cells = 2480) were analyzed at X400 magnification using a neutral density filter as follows. The image analysis system was set for grain counting (VCMTE Video Count) and an illumination threshold was established. Silver grains were set above threshold while nuclei in these lightly counterstained sections remained below threshold. In this approach, pixels above threshold (i.e. those covered by silver grains) are counted on an open-ended scale. Due to unavoidable stacking of silver grains in autoradiographic emulsions, optical density measurements cannot strictly be considered as grain counts. Hence, we have expressed our data as the mean area of grains by dividing the number of pixels above threshold (i.e. the video count value) by the number of pixels per nm2 (4.28 pixels/^m2 at X400 magnification). A single background grain gave a reading of 4 (pixels). The average cellular labeling intensity for the second hybridization was 66 times background. Both of these image analysis approaches are comparable to cellular grain counts since grain density is linearly related to optical reflectance (34). Image analysis was performed on all cells situated ventral to the anterior commissure and medial to the medial forebrain
461
bundle. Hence, the few cells residing in the medial, lateral and triangular septal nuclei and the dorsal portion of the interstitial nucleus of the stria terminalis were not included. Occasional labeled cells also were visualized in the amygdala and in the cingulate and piriform cortices in these sections. Data analysis and statistics Approximately 100 cells/animal were analyzed and an animal mean was calculated. Group means and SEMs for intact, Orch and Orch tumor-bearing groups were then calculated based on animal numbers of 9, 9, and 7, respectively. Cochran's test of the raw data indicated that the assumption of homogeneity of variance was satisfied. Data were then analyzed using one-way analysis of variance, followed by Duncan's new multiple range test (a = 0.05) to identify significant differences between individual means (35).
Results Nine days postorchidectomy, a 10.4-fold increase in plasma LH titers was observed which was blocked by the presence of the PRL-secreting tumor (Fig. 1). Circulating PRL levels produced by these tumors are depicted on the right axis. The effects of orchidectomy and PRL on single cell levels of LHRH mRNA are illustrated in Fig. 2. Orchidectomy resulted in a significant, 19% increase in the area of grains compared with that in intact control rats [98.6 ± 4.6 (n = 9) us. 117.5 ± 3.6 (n = 9) /urn2]. This increase was blocked in tumor-bearing rats [98.6 ± 4.6 (n = 9) vs. 101.6 ± 6.3 (n = 7) jum2]. The total numbers of cells analyzed per group were 909 (intact), 871 (Orch)
• LH • PRL
M
T :r 600 i 1
2 o
»
cc
4000
500
|
_l w
-="400 -
3000
S 300 Ii
2000 g
O 200
1000 3
E
to
o a:
IOO
1 M\ Intact
HLJ Orch
- 10
BB-J Orch 7315a Tumor
FlG. 1. Plasma LH response to orchidectomy in control and tumorbearing rats. Each bar represents the mean ± SEM plasma LH and PRL levels in intact (n = 9), Orch (n = 9) and Orch, tumor-bearing (n = 7) rats. Plasma PRL levels produced by these tumors are depicted on the right axis. * Significantly different from the other two experimental groups.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 August 2016. at 19:23 For personal use only. No other uses without permission. . All rights reserved.
462
HYPOTHALAMIC mRNA ENCODING LHRH IN RATS 125 r
120
E
115
|
M0
•
Vol. 128, No. 1 Printed in U.S.A.
Single Cell Levels of Hypothalamic Messenger Ribonucleic Acid Encoding Luteinizing Hormone Releasing Hormone in Intact, Castrated, and Hyperprolactinemic Male Rats* MICHAEL SELMANOFF, CHRISTINE SHUf, SANDRA L. PETERSEN, CHARLES A. BARRACLOUGH, AND R. THOMAS ZOELLER Center for Studies in Reproduction, Department of Physiology, University of Maryland School of Medicine (M.S., C.S., C.A.B.), Baltimore, Maryland 21201; and the Department of Anatomy and Neurobiology, University of Missouri School of Medicine (S.L.P., R.T.Z.), Columbia, Missouri 65212
ABSTRACT. We have examined the changes that occur in neuronal expression of LHRH mRNA in response to castration and hyperprolactinemia in male rats. Single cell levels of LHRH mRNA were determined by quantitative in situ hybridization histochemistry using an 35S-labeled synthetic 48-base oligodeoxynucleotide probe and quantitative autoradiography. Nine days postcastration, a 10.4-fold increase in mean plasma LH titers was observed which was associated with significantly increased LHRH mRNA in rostral hypothalamic neuronal cell bodies. Both increases were blocked in rats rendered hyperprolactinemic by the presence of the 7315a PRL-secreting pituitary tumor.
The location and number of neurons expressing LHRH mRNA were unchanged, indicating that these differences were attributable to amounts of mRNA expressed per neuron. Experimental differences occurred in LHRH perikarya situated throughout the rostral hypothalamus from the organum vasculosum of the lamina terminalis to the caudal extent of the medial preoptic nucleus. These results suggest that gonadal steroids and PRL are involved, either directly or indirectly, in regulating the biosynthesis of LHRH in the rostral hypothalamus. (Endocrinology 128: 459-466, 1991)
T
perfusion (10,11) studies have not detected mean LHRH increases, and this issue has become controversial. Evidence that PRL inhibits LH release is suggested by endocrine states in which an inverse relationship exists between these two pituitary hormones (e.g. postpartum lactation, pregnancy, and pseudopregnancy) (12). The antigonadotropic actions of PRL may occur at the level of the brain, pituitary, and/or gonad in male and female mammals (13). Experimentally induced hyperprolactinemic states employing PRL-secreting tumors, ectopic pituitary glands, or PRL injections inhibit circulating LH levels under several circumstances. Profound, unremitting inhibition of the postcastration LH rise occurs in the presence of transplantable 7315a, 7315b, or MtTW15 PRL-secreting pituitary tumors (14-17). This inhibitory effect in gonadectomized rats may be mediated within the hypothalamus by decreasing LHRH secretion and/or at the pituitary level by altering pituitary responsiveness to LHRH. Increasing evidence indicates that changes in the activity of peptidergic neurons is related to alterations in cellular levels of prepropeptide mRNA (18-20). In situ hybridization histochemistry allows one to identify and
HE DECAPEPTIDE LHRH stimulates the release of LH and FSH from the anterior pituitary gland (1). It has been thought that removal of the gonadal steroid negative feedback effect results in the hypothalamic mechanisms mediating LHRH release becoming chronically active. As a consequence, increased LHRH is detected in hypophyseal portal blood (2-5). In male rats, the rapidly increased mean plasma LH levels (6) are produced by increases in both the frequency and amplitude of pulsatile LH secretion (6, 7). Similarly, LHRH pulses detected in push-pull perfusates of the anterior pituitary gland in unanesthetized castrated male rats have increased amplitude and frequency (8). However, some portal blood (9) and hypothalamic push-pull Received May 14, 1990. Address all correspondence and requests for reprints to: Dr. Michael Selmanoff, Department of Physiology, University of Maryland School of Medicine, 660 West Redwood Street, Baltimore, Maryland 21201. * This work was supported in part by a stipend from the Foundation of Medical Research, France (to C.S.), and grants HD-21351 (to M.S.) and HD-02138 (to C.A.B.) from the NIH. A preliminary report of these data was made at the 18th Annual Meeting of the Society for Neuroscience, Toronto, Canada, 1988, vol 18:1069. t Present address: Roussel-UCLAF Pharmaceutical Co., Romainville, France.
459
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 August 2016. at 19:23 For personal use only. No other uses without permission. . All rights reserved.
HYPOTHALAMIC mRNA ENCODING LHRH IN RATS
460
quantify mRNA transcripts in single neurons. Hence, the effects of physiological variables on gene expression in single neurons can be evaluated. In the present study we used this approach to determine the effects of castration and elevated plasma PRL levels on regional changes in cellular levels of LHRH mRNA. We provide clear evidence that cellular levels of LHRH mRNA are increased after castration and that this increase is blocked in animals bearing a PRL-secreting tumor.
Materials and Methods Animals and tissue preparation Adult male Buffalo rats (Harlan Sprague-Dawley, Indianapolis, IN), weighing 300-350 g, were maintained on a 14-h light, 10-h dark schedule, with food and water provided ad libitum. Rats were inoculated sc with a suspension of minced 7315a tumor tissue, as previously described (21). Six weeks later, when the tumor volume was 35-50 cm3, these rats were orchidectomized (Orch) under brief ether anesthesia (n = 7). Age-matched intact (n = 9) and orchidectomized (n = 9) rats comprised the other two experimental groups. Nine days postcastration, rats were decapitated between 0900-1000 h. Brains were rapidly removed, frozen on dry ice, and stored, tightly wrapped in Parafilm, at -70 C. Trunk blood was collected, the cells were separated by centrifugation, and the plasma was stored at —20 C until RIA. Frozen 12 nm sections were prepared from the bed nucleus of the diagonal band of Broca (-A7860; DBB) through the caudal extent of the medial preoptic nucleus (~A6460; MPN). Atlas coordinates and coronal plane of section are from the atlas of Konig and Klippel (22). One of three sections (save one, discard two) was thaw-mounted on acid-washed chromalum- and gelatin-coated slides (two sections per slide) through this 1400 jim extent of the rostral hypothalamus. Thus, three sets of slides per animal were prepared and stored at —70 C until hybridization was carried out. RIA Plasma LH concentrations were determined by the heterologous ovine:ovine double antibody method using the NIDDK rat LH RP-1 standard, GDN #15 antibody, and LER-1374A ovine LH preparation for radioiodination (23). Plasma PRL was assayed by a homologous double antibody method, as previously described (24), using NIDDK rat PRL reagents provided by Dr. Albert Parlow through the National Hormone and Pituitary Program. The rat PRL standard used was rat PRL RP-3 (biopotency = 2.8 times NIDDK rat PRL RP-1). In situ hybridization In situ hybridization histochemistry was performed by a modification (25) of the method described by Mason and coworkers (26). The LHRH probe was a 48-base oligodeoxynucleotide complementary to the LHRH coding region of the rat LHRH cDNA (bases 102-149) described by Adelman and coworkers (27). This 48mer was synthesized on an Applied Bio-
Endo • 1991 Voll28«No 1
systems 380A DNA synthesizer and purified on a preparative denaturing acrylamide gel (28). The probe (0.1 MM) was 3' end labeled with [35S]dATP (1.0 MM; 1300 Ci/mmol; New England Nuclear, Boston, MA) and terminal deoxynucleotidyl transferase (100 U; Bethesda Research Laboratories, Gaithersburg, MD) to a specific activity of about 7000 Ci/mmol. Prehybridization treatment consisted of warming the sections to room temperature, after which they were fixed for 5 min in 4% formaldehyde in 0.12 M sodium PBS (pH 7.4), rinsed in 2 x SSC (1 x SSC = 0.15 M NaCl, 0.015 M sodium citrate buffer, pH 7.2), and then acetylated for 10 min in 0.25% acetic anhydride in 0.1 M triethanolamine hydrochloride-0.9% NaCl (pH 8.0). Sections were dehydrated through an ethanol series (70-100%), delipidated in chloroform, rehydrated in 95% ethanol, and air dried. Labeled probe was applied in a saturating concentration to sections in 45 n\ hybridization buffer. This buffer contained 4 x SSC, 50% (vol/vol) formamide, 10% (wt/ vol) dextran sulfate, 250 yug/ml yeast transfer RNA, 500 ^tg/ml sheared single stranded salmon sperm DNA, 1 X Denhardt's solution (0.02% Ficoll, 0.02% polyvinylpyrrolidone, and 0.02% BSA), and 50 mM fresh dithiothreitol. Hybridization proceeded under Parafilm coverslips, overnight, at 37 C in humid chambers. Sections were washed four times under stringent conditions (2 x SSC-50% formamide) at 40 C (-19 C below Tm, the mean melting temperature for this cDNA-mRNA hybrid) for 15 min each, and then twice under less stringent conditions (1 x SSC) at 22 C for 60 min each. After the last wash, slides were dipped briefly in water and then in 70% ethanol and dried. Slides were dipped in fresh Kodak NTB-3 nuclear track emulsion (Eastman Kodak, Rochester, NY), exposed for 5 days at 4 C, and developed, and the cell nuclei were counterstained with toluidine blue. Several findings indicate that the 35S-labeled oligonucleotide probe hybridized specifically to mRNA encoding the LHRH precursor: 1) labeled cells display the same distribution as reported by immunohistochemical staining of the decapeptide (29, 30); 2) this LHRH probe and a 48-mer complementary to the GAP portion of the LHRH precursor (bases 218-265) label the same cells in adjacent sections (28); 3) hybridization of adjacent sections with a vasopressin probe (31) yielded a labeled hypothalamic cell population distinct from that labeled with the LHRH probe and did not label cells in regions labeled by the LHRH probe; 4) no labeled cells were observed when sections were hybridized with probe diluent alone as a control for positive chemography, and 5), a sense oligomer, directed to the complementary DNA strand, revealed no labeling in adjacent sections. Image analysis Tissue sections were matched anatomically between animals, and 10 sections/animal were analyzed. The autoradiographic hybridization signal was quantitated in a Bioquant Image Analysis System IV (R & M Biometrics, Inc., Nashville, TN) consisting of an AT & T TARGA M8 frame grabber and AMETEK Hipad digitizing tablet attached to an IBM XT computer. Video images were obtained by a DAGE-MTI 65 video camera attached to a Leitz Laborlux D photomicroscope (Rockleigh, NJ) equipped with a X40 epiillumination darkfield
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 August 2016. at 19:23 For personal use only. No other uses without permission. . All rights reserved.
HYPOTHALAMIC mRNA ENCODING LHRH IN RATS condensor. Illumination intensity was controlled by a 3-12V DC-regulated stabilized power supply. Images were displayed on an Analog RGB monitor. In this study two separate hybridizations were performed on two sets of sections (10-14 sections/animal) from the same animals. The second hybridization was performed for three reasons: 1) to confirm the original finding, 2) to use a different image analysis approach, and 3) to increase the number of sections from each animal for the regional data analysis. The basic experimental results were the same with both hybridizations and we present data from the second, more extensive, hybridization. Image analysis of labeled cells from the first hybridization (total number of cells = 2075) was performed as follows. Clusters of grains were determined to be over cell bodies using brightfield optics at X400 magnification. Then, under darkfield optics, grain clusters associated with individual cells were moved to the center of the video screen, brought into exact microscope focus and circumscribed by the operator. The average optical density and area over which grains are distributed were then obtained. The average cellular labeling intensity was 50 times background (average background optical density reading = 2.5; average cell reading = 125.4). Analyzed in this manner, the optical brightness of each pixel in the image analysis system is assigned a number based on a gray level scale of 0-255 in the Bioquant IV system. We constructed a standard curve relating this gray level scale to brain paste standards (18, 25, 32). These standards were prepared by adding known amounts of 35S to brain homogenate as previously described (25, 33), freezing them, sectioning them at 12 fiva, processing them for autoradiography, and reading them in the image analysis system. Optical density values could then be converted to disintegrations per min of 35S. The standard curve verified that the optical density of the hybridization signal over labeled cells was on the linear portion of the curve relating optical density and radioactivity and was, thus, below saturation of the nuclear emulsion (32). Labeled cells from the second hybridization (total number of cells = 2480) were analyzed at X400 magnification using a neutral density filter as follows. The image analysis system was set for grain counting (VCMTE Video Count) and an illumination threshold was established. Silver grains were set above threshold while nuclei in these lightly counterstained sections remained below threshold. In this approach, pixels above threshold (i.e. those covered by silver grains) are counted on an open-ended scale. Due to unavoidable stacking of silver grains in autoradiographic emulsions, optical density measurements cannot strictly be considered as grain counts. Hence, we have expressed our data as the mean area of grains by dividing the number of pixels above threshold (i.e. the video count value) by the number of pixels per nm2 (4.28 pixels/^m2 at X400 magnification). A single background grain gave a reading of 4 (pixels). The average cellular labeling intensity for the second hybridization was 66 times background. Both of these image analysis approaches are comparable to cellular grain counts since grain density is linearly related to optical reflectance (34). Image analysis was performed on all cells situated ventral to the anterior commissure and medial to the medial forebrain
461
bundle. Hence, the few cells residing in the medial, lateral and triangular septal nuclei and the dorsal portion of the interstitial nucleus of the stria terminalis were not included. Occasional labeled cells also were visualized in the amygdala and in the cingulate and piriform cortices in these sections. Data analysis and statistics Approximately 100 cells/animal were analyzed and an animal mean was calculated. Group means and SEMs for intact, Orch and Orch tumor-bearing groups were then calculated based on animal numbers of 9, 9, and 7, respectively. Cochran's test of the raw data indicated that the assumption of homogeneity of variance was satisfied. Data were then analyzed using one-way analysis of variance, followed by Duncan's new multiple range test (a = 0.05) to identify significant differences between individual means (35).
Results Nine days postorchidectomy, a 10.4-fold increase in plasma LH titers was observed which was blocked by the presence of the PRL-secreting tumor (Fig. 1). Circulating PRL levels produced by these tumors are depicted on the right axis. The effects of orchidectomy and PRL on single cell levels of LHRH mRNA are illustrated in Fig. 2. Orchidectomy resulted in a significant, 19% increase in the area of grains compared with that in intact control rats [98.6 ± 4.6 (n = 9) us. 117.5 ± 3.6 (n = 9) /urn2]. This increase was blocked in tumor-bearing rats [98.6 ± 4.6 (n = 9) vs. 101.6 ± 6.3 (n = 7) jum2]. The total numbers of cells analyzed per group were 909 (intact), 871 (Orch)
• LH • PRL
M
T :r 600 i 1
2 o
»
cc
4000
500
|
_l w
-="400 -
3000
S 300 Ii
2000 g
O 200
1000 3
E
to
o a:
IOO
1 M\ Intact
HLJ Orch
- 10
BB-J Orch 7315a Tumor
FlG. 1. Plasma LH response to orchidectomy in control and tumorbearing rats. Each bar represents the mean ± SEM plasma LH and PRL levels in intact (n = 9), Orch (n = 9) and Orch, tumor-bearing (n = 7) rats. Plasma PRL levels produced by these tumors are depicted on the right axis. * Significantly different from the other two experimental groups.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 August 2016. at 19:23 For personal use only. No other uses without permission. . All rights reserved.
462
HYPOTHALAMIC mRNA ENCODING LHRH IN RATS 125 r
120
E
115
|
M0
•
