| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Anatomy and Neuroscience (G.V.C., G.U., P.W.), University of Texas Medical Branch, Galveston, Texas 77555; and Department of Anatomy (G.V.C.), University of Arkansas School for Medical Sciences, Little Rock, Arkansas 72205
Address all correspondence and requests for reprints to: Gwen V. Childs, Ph.D., Professor and Chair, Department of Anatomy, University of Arkansas School for Medical Sciences, 4301 W Markham, Slot 510, Little Rock, Arkansas 72205. E-mail: childsgv{at}exchange.uams.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recently, Gonzalez-Parra et al. (7) showed that castration of adult or neonatal rats decreased percentages of GH cells and levels of mRNA for GH and Pit 1, which encodes a transactivating factor for the GH gene. Normal percentages of GH cells and levels of mRNA could be restored if neonatal castrates received testosterone replacement.
The mechanisms behind the sexual dimorphism appear to be complex. Gatford et al. (9) tested sex differences in circulating patterns of somatostatin and GH-releasing hormone (GHRH) in growing lambs. In experiments that controlled for sex differences in food intake, they found that rams had higher levels of the following: plasma and pituitary GH, median eminence content of GHRH and somatostatin, GH pulse amplitudes, and integrated plasma GH. However, the steady-state levels of GH mRNA or the portal blood levels of somatostatin did not differ between the sexes. Thus, they concluded that the sex differences may also be driven by GHRH pulses or changes in pituitary sensitivity to the pulses.
Recent studies in our laboratory have provided additional clues to possible cellular mechanisms behind the sexual dimorphism. In the early 1990s, we discovered that, in diestrous and proestrous female rats, 40–60% of cells with GH antigens expressed gonadotropin ß subunit mRNA. These cells contributed significantly to the overall increase in the percentages of gonadotropes seen late in diestrus and early in proestrus (10). Further tests of females during the same period of the cycle showed that about 30–38% of cells with GH antigens expressed receptors for GnRH (11), suggesting that they could function as multipotential gonadotropes. In subsequent studies, we reported that the expression of GnRH receptors by GH cells could be modulated by inhibin (12) or activin (13). We hypothesized that a subset of somatotropes might be stimulated to support the gonadotrope population (by the production of LH and FSH) and produce a cocktail of hormones that could be used by the ovary or testis. This multihormonal cell might also be a logical cellular site for steroid induced sex differences in the expression of GH.
We recognized, however that the expression of GH antigens cannot be used to prove that the cell produces GH. GH could be present in gonadotropes as a regulatory hormone bound to receptors or GH binding proteins. Therefore, our studies have sought to determine if the multihormonal cells retain their GH phenotype. In recent studies, we reported that cells with gonadotropin antigens expressed GHRH receptors during proestrus, which is an important characteristic of somatotropes (14). The present report presents parallel studies to learn if these multihormonal gonadotropes express GH mRNA. Biotinylated oligoprobes complementary to GH mRNA were produced and applied to cell populations from male rats and cycling female rats. The cytochemical data were then correlated with RNase protection assays for GH mRNA to learn if the changes in the cell populations corresponded to overall changes in GH mRNA levels in the pituitary. Our studies provide new evidence for the differential expression of GH mRNA during the estrous cycle as well as significant expression of GH mRNA by gonadotropes from both male and proestrous female rats. Furthermore, this report will confirm previous studies that reported sex differences in expression of GH mRNA. However, these differences depend on the stage of the estrous cycle being studied. Male rats express more GH mRNA only if they are compared with females in estrus, metestrus, and diestrus.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Rats were killed by decapitation within seconds of removal from their cages. The pituitaries were then removed and placed in defined medium as previously described (10). They were dissociated and the cells were plated overnight in defined media (10).
For the RNase protection assays, normally cycling female or male rats were killed as described above. Their pituitaries were collected and stored in RNAlater (Ambion, Inc.), a solution that protects cellular RNA from degradation. At the time of extraction, RNAs from each pituitary were extracted with RNAwiz (Ambion, Inc.) according to manufacturer’s instructions. Typical yields were 30–50 micrograms RNA per pituitary.
In situ hybridization
The protocol for in situ hybridization is described
in two recent techniques papers (15, 16). The probe for GH was an
oligonucleotide synthesized with biotin attached to the 5' end (DNA
International, Inc.). It was complementary to mRNA encoding amino acids
40 to 54 on the GH molecule. This is a region that carries no homology
with PRL. The biotinylated sense oligonucleotide sequence was also
produced and used as a control probe.
The pituitary cells were fixed in 2% glutaraldehyde as described previously (10). They were then stored for use in the in situ hybridization protocol for no longer than a week. Prehybridization steps were as described in techniques papers 15 and 16 and the cells were incubated with the probes overnight at 37 C. The biotinylated sense and antisense probes were then detected by antibody to biotin in a sandwich technique that ultimately used streptavidin peroxidase as the reporter molecule (15, 16). This produced a sensitive blue-black reaction in patches or linear patterns in the cell.
In parallel experiments, cells were labeled for the GH mRNA, after which the population was prepared for dual labeling for LHß, FSHß, or GH antigens. The immunocytochemical technique was identical to that described previously (15, 16, 17, 18). It produced an orange-amber reaction product that usually filled the entire cytoplasm.
Counts of at least 150 cells/slide provided the quantitative data for the in situ hybridization studies. Each set of counts represented cell populations from three different slides. Each experimental group was repeated at least three times with two rats/group. We calculated percentages of cells that contained label for GH mRNA with or without label for one of the antigens as described in previous studies (10, 11). ANOVA (1x) was run to compare averages from each set of rats (n = at least three groups) and, if the F value was significant (P <0.05), the Fisher’s least significant difference post hoc test was run to identify the values that were different.
To detect differences in density or area of label, the threshold density of the GH mRNA label was first established with a Bioquant Windows 95 computer-based densitometer (R & M Biometrics, Nashville, TN). This provided automatic measurements of the area of the label in each cell as well as the average density of the pixels on a scale of 0–255, where 0 is maximal density (no light transmitted) and 255 is the lowest density (maximal light transmitted). The light meter was set to the same reading each day, and the first 25 labeled cells/coverslip were analyzed. Periodic checks of the threshold were done to highlight the pixels over the label. This helped ensure that all label was detected. During each measurement, the area of the cell was calculated after drawing around its perimeter. Then, the computer automatically calculated the density of label inside the cell. Area of label was calculated by the "video count" reading that provided the number of pixels that covered label for GH mRNA. The area of these counts was calculated automatically. These measurements were done only on cell populations that were exposed to single labeling protocols. Averages were obtained for each experimental group (sampling 4 rats/group), and the averages for the diestrous and proestrous populations were compared by Student’s t test. A P < 0.05 was considered significant.
Assays for pituitary GH mRNA
Production of the cRNA probes. To make the GH cRNA probe, we
chose the 5' end of rat GH cDNA as a template. This 380-bp fragment was
first removed from pRGH-1 (ATCC) by PstI and
KpnI digestion, and then inserted into pBluescript II SK (+)
(Stratagene, pBSGH). The design ensured that the GH
fragment was under the control of the T7 promoter in an antisense
orientation. The resulting pBSGH plasmid was amplified in
Escherichia coli DH 5
, and purified with
QIAGEN Plasmid Midi kit (QIAGEN). An aliquot
of pBSGH was linearized with BamHI, and then purified using
QIAquick PCR Kit (QIAGEN) to remove the salt and
enzymes.
32P-labeled antisense RNA probes and RNA markers were produced using a MAXIscript T7 in vitro transcription kit (Ambion, Inc.) according to manufacturer’s instruction. The GH cRNA probe was made with the use of the linearized pBSGH as a template. In addition, pTRI-RNA-28S purchased from Ambion, Inc. was applied to make a 28S probe, which targets cellular 28S rRNA. We chose 28S as an internal control for quantitative studies since it is rarely influenced by changes in cell stages or different tissues. RNA Century Marker Templates (Ambion, Inc.) were used for the synthesis of labeled RNA that provided molecular size standards.
Initially, the conditions for the production of the labeled probe were
optimized by varying the concentrations of labeled vs.
unlabeled UTP in the reactions. Concentrations of 3.1
µM of -32P-UTP and 50
µM of unlabeled UTP were found to be optimal
for the synthesis of the GH rRNA probe. Similarly, 0.625
µM of labeled and 500
µM of unlabeled UTP were optimal for the 28S
probe. Both GH and 28S probes were gel-purified to ensure the recovery
of only full length probes.
The Ribonuclease protection assay (RPA). The RPA was performed using the RPA III kit (Ambion, Inc.). In the initial phases of the study, pilot experiments were set up to determine the optimal ratios between the amounts of sample RNA and radiolabeled probes. This was done by applying pituitary RNA ranging from 0.25 µg to 10 µg in these pilot RPAs with a constant amount of radiolabeled GH or 28S probe. Our preliminary study indicated that 1 µg of pituitary total RNA is sufficient for the detection of GH mRNA by RPA. The specificity of the GH rRNA probe was verified by size fragmentation on a polyacrylamide gel, as well as by RPA run on samples of yeast RNA or on samples that omitted the target RNA.
After the pilot studies optimized the reaction conditions and also showed the specificity of the reaction, GH mRNA was assayed following detailed procedures described in the kit instruction. Briefly, 1 µg of sample RNA plus 10 µg of yeast RNA (for expansion) were mixed with 4 x 104 cpm of GH probe and 2 x 104 cpm of 28S probe. After ethanol precipitation, they were resuspended in 10 µl of the kit hybridization buffer and incubated at 56 C for about 19 h. Unhybridized RNAs and probes were then digested with RNase A/RNase T1 at a dilution of 1:100 at 37 C for 30 min. Following RNase inactivation and precipitation, sample pellets were resuspended in 10 µl of Gel Loading Buffer II, and then run on a 5% acrylamide/8 M urea gel at 200 V for 1.5 h. Autoradiography was carried out by exposing the gel to an x-ray film with an intensifying screen at -70 C for 1–1.5 h. The radioactive signals on the gel were also quantified by an automated system, InstantImager (Packard Instrument Co., Meriden, CT). The intensity of signals for GH mRNA was depicted as cpm and normalized against signals for 28s rRNA.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). This value is significantly higher
than those from metestrous (14.9 ± 0.7%) or estrous (23 ±
3%) females. Diestrous or proestrous female rats had comparable
percentages of GH mRNA-bearing cells (29 ± 3%). When the
biotinylated sense sequence was substituted for the antisense sequence,
the percentages of labeled cells averaged 3 ± 1.5% (three
separate experiments, cells from male rats). Cells exposed to
hybridization buffer alone during the hybridization had minimal
labeling (1.5–2.4%; examples from all sexes and experimental
groups).
|
also shows the counts of GH antigen-bearing cells from these
same groups (taken from the dual-labeling experiments). Male rats had
35.7 ± 1.4% GH cells. Female rats had percentages of GH cells
that ranged from 26 ± 1.4% (metestrous rats) to 34.7%
(proestrous rats). The application of the Student’s t test
showed that metestrous female rats had significantly lower percentages
of GH cells when compared with male rats (P < 0.01) or
proestrous female rats (P <0.05).
Figure 2 is a plate that illustrates the
GH mRNA labeling data. As in our previous studies (10), the label is in
dark spots or lines in the cells and it varies in density from cell to
cell. Figure 2, a–c, compares the labeling in the male rat cells with
the use of antisense (Fig. 2a) and sense probes (Fig. 2, b and c).
Figure 2d shows that the expression in proestrous females compares
favorably with that in the male rats (Fig. 2a). Figure 2e shows that
the expression in some regions from metestrous rat populations is
virtually absent. In other fields, there are only one to two cells.
|
shows an autoradiogram that includes some of the tests of specificity
of the antisense RNA probes following the manufacturer’s instructions.
Lane B shows a strong signal on the gel when 10 µg yeast RNA was
hybridized with 4 x 104 cpm of the GH probe
and no RNase treatment was used (Lane B). In contrast, Lane A shows
that, when RNase A/RNase T1 at 1:100 was added to digest all
nonhybridized RNAs, it completely abolished the signal, indicating that
no hybrids had formed from the yeast RNA and the GH probe. The 400-bp
band seen in Lane B includes both the 380-bp GH probe fragment and
20-bp of flanking sequences from the pBluescript II SK (+) plasmid.
Lane C contains a distinctive band (380-bp) from a sample containing
total RNA (from male rat pituitary) and the GH probe. The sample was
treated with RNase, and the radioactive band indicates that hybrids had
formed. The extra bands are believed to be a nonspecific background.
This is most likely due to the excess amount of probes used to ensure a
quantitative analysis. Lane D shows the effect of adding a sample
containing male rat pituitary RNA without the GH probe. No signal was
detected when 1 µg of total RNA from a male rat pituitary was
digested with the same amount of RNase in the absence of the GH
probe.
|
illustrates a representative
autoradiogram showing samples containing total rat pituitary RNA plus
GH and 28S probes (and RNase A/RNase T1 treatment) from each
experimental group. The quantification of RNA from at least six
rats/experimental group is shown in Fig. 4b. The statistical analysis
shows that levels of GH mRNA in male rats are greater than those in
diestrous, metestrous, and estrous female rats. However, proestrous
female rats have higher levels of GH mRNA than all other groups,
including male rats. Metestrous rats contain levels of GH mRNA that are
lower than all but the estrous group.
|
indicate that similar
percentages of GH mRNA-bearing cells are found in diestrous and
proestrous groups. Densitometric analyses were conducted to learn if
the difference in levels assayed by the RPA reflected lower levels of
GH mRNA/cell. Table 1 shows that values
were higher in proestrous rats for all types of measurements, including
average areas of GH mRNA bearing cells, average areas of label for GH
mRNA/cell, and the average density of the label. When evaluating
density measurements, it is important to note that a lower number
reflects less light transmitted. Therefore, labeling density that is
more intense (darker blue or black) would let less light through and
give a reading closer to 0. Weaker labeling would let more light
through and give a higher value in the reading.
|
shows that over 70–80% of cells with
GH antigens expressed GH mRNA in all groups, except metestrous females.
Fewer than 30% of GH antigen-bearing cells expressed GH mRNA in the
metestrous rats, which is significantly lower than values from all
other groups.
|
shows that, in metestrous rats, cells
with GH antigens and GH mRNA are only 7% of the pituitary cells. In
contrast, these dual-labeled cells represent 22–29% of pituitary
cells in all other groups.
|
). In metestrous females, only 30–35% of cells with
FSHß or LHß antigens contained GH mRNA. Expression of GH mRNA by LH
cells rose significantly during diestrus to reach a peak of 57% of
cells with LHß antigens by the morning of proestrus, and it remained
high during estrus. Cells with FSH antigens exhibited a slightly
different pattern. The rise to 57% of FSH cells did not occur until
the morning of proestrus, and this peak level persisted until the
morning of estrus (which is not different from the values in proestrous
rats).
|
).
|
Finally, the data also calculated changes in percentages of GH
mRNA-bearing cells that contained each of the antigens tested. Figure 9 shows the data. In all stages of the
cycle but estrous, over 90% of GH mRNA bearing cells had translated
the mRNA and stored GH antigens. During metestrus, this value declined
significantly to 60 ± 15%. This suggests that 40% of GH
mRNA-bearing cells have stores of GH below threshold levels needed for
detection by immunolabeling. This could be the result of secretion, or
it could be that the cells have simply not translated the mRNA. When
gonadotropins were analyzed, 18–26% of GH mRNA bearing cells stored
LH or FSH during all phases of the cycle except proestrus when the
values increased to 41 ± 2%. Recall that percentages of cells
with gonadotropin antigens are lowest during estrus and metestrus
because of loss of gonadotropin stores. Thus, as the percentage of GH
mRNA bearing cells reaches its nadir, so do percentages of LH or FSH
cells. This parallel reduction maintains the low proportion of GH mRNA
bearing cells that store gonadotropins.
|
is a plate illustrating
examples of dual-labeled cells. Figure 10, a and b, compares fields
labeled for GH mRNA and GH antigens in metestrous (Fig. 10a) and
proestrous (Fig. 10b) rats. Fewer dual labeled GH cells are seen in the
metestrous field as would be expected from the values in Fig. 9. In
contrast, numerous cells with GH mRNA and/or GH antigens are evident in
the field from proestrous rats (10b).
|
shows a cell with GH mRNA and LHß antigens in the
metestrous field. It is extending a process and strands of linear
arrayed label are evident in the cytoplasm. The field in 10d
illustrates a cell with GH mRNA and LHß antigens (also shown in the
inset). Also illustrated are two cells with label for GH
mRNA only (gray label).
Figure 10, e–g, are views from male rats taken with oil immersion
objectives to show examples of the differential labeling patterns.
Figure 10e shows two labeled cells. One cell is labeled intensely
gray for GH mRNA next to a cell labeled orange
for LHß and black for GH mRNA. Faint orange label is seen
in the adjacent cell when one focuses through the gray label. Figure 10f is also from male rats and shows a cell labeled gray for GH mRNA
(with a faint orange for LH antigens) and another cell labeled dense
orange for LHß and black for GH mRNA (in a linear pattern). Figure 10 g shows a view of the circular pattern of label for GH mRNA
(vacuoles) in a cell that contains GH antigens. Figure 10h shows three
labeled cells from proestrous rats. Two of the cells are labeled
orange for FSHß only. The cell in the center contains
FSHß antigens and a dense patch of label for GH mRNA.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our earlier studies (10, 11), we recognized that the presence of GH antigens in the putative multihormonal cell did not necessarily prove that the cells were somatotropes. We reasoned that GH could be present in gonadotropes as a regulatory hormone bound to GH binding proteins (reviewed in Ref. 11). Therefore, during the past 3 years, we have continued to look at the extent to which gonadotropes might express characteristics unique to GH cells, to learn more about the significance of this coexpression of GH and gonadotropin hormones and receptors.
The first studies of GH cell phenotypes showed that GHRH binding sites could be found on GH cells and gonadotropes (14). We also reported that the highest expression of GHRH receptors by cells with gonadotropin antigens occurs in early proestrus, when the putative multihormonal cells emerge. Significant numbers of gonadotropes in male rats (20–30% of gonadotropes) also expressed GHRH. The presence of GHRH receptors added support for the hypothesis that the cells were somatotropes and not gonadotropes with endocytosed GH antigens. It also added information suggesting that these cells could respond to multiple secretagogues.
The findings from the present study provide the strongest support for
the hypothesis that the multihormonal cells represent a unique subset
that functions to support both the GH and gonadotrope population. In
all but the metestrous groups, over 91% of the mRNA-bearing cells
contain GH antigens (Fig. 9). At the same time, 41% or 32% of GH
mRNA-bearing cells contain LHß or FSHß antigens, respectively.
Adding these percentages accounts for over 100% of the GH mRNA cell
population. Therefore, it is likely that 30–40% of these GH
mRNA-bearing cells are multipotential and can store all three hormones.
These data correlate well with percentages published in 1994 (10),
which showed that 40% of cells with GH antigens contain LHß mRNA and
34% of cells with GH antigens express GnRH receptors (11).
At this point, there are four combinations of phenotypes found in these pituitary cells that point to a multihormonal function for this subset. These include expression of: GH antigens with LHß or FSHß mRNAs (10); GH antigens and GnRH receptors (11, 12, 13); GHRH receptors and LHß or FSHß antigens (14); and GH mRNA and LHß or FSHß antigens (the present report). A fifth phenotypic characteristic common to this subset is the response to inhibin and activin (12, 13), which decreases or increases expression of GnRH receptors in GH cells, respectively.
Sexual dimorphism in expression of GH mRNA
Sexual dimorphism appears during pubertal development (1, 2, 3, 4) and,
as discussed in the introduction, removal of steroid feedback produces
expression of GH mRNA or antigens more like that in the opposite sex
(6, 7). Similarly, restoration of steroid feedback allows recovery to
the sex-specific level of expression (7). The results from this study
also partially confirm studies that show sexual dimorphism in the
expression of GH mRNA by adult rat pituitaries (7, 8, 19). However, in
our experiments, the female rats were separated into groups based on
the stage of the estrous cycle. New findings in the present study show
that male rat pituitaries have more GH mRNA only when compared with
female rats in metestrus, diestrus, or estrus. Because none of the
previous studies of adult rats identified the stage of the cycle (7, 8, 19), it is likely that they assayed GH mRNA in a mixed group of females
that might contain proestrous rats in a minority of the population.
Differential expression of GH mRNA during the estrous cycle
Both the RNase protection assays and counts from the in
situ hybridization studies agree that there is an increase in GH
mRNA transcripts (or their stability) when metestrous and proestrous
groups are compared. However, the RNase protection assays showed
intermediate values in the diestrous group, whereas the counts from
in situ experiments showed maximal percentages of cell
labeled for GH mRNA. Densitometric studies were then done to learn if
the differences between the assays were due to changes in the amount of
GH mRNA/cell (as detected by area and density). The stereometric and
densitometric data showed that GH mRNA bearing cells from proestrous
rats were larger and contained more GH mRNA (detected by area and
density) than their counterparts from diestrous rats.
The current study is the first to show differential expression of GH mRNA during the estrous cycle. Past studies of the expression of GH secretory activity using reverse hemolytic plaque assays (20) showed no changes in the number of GH-secreting plaques with the bovine estrous cycle. Indeed, most studies that have counted percentages of GH cells in the cycling female rat have reported no significant differences (10, 11). The present studies reported differences in GH antigen-bearing cells only when values from metestrous rats were compared with those from males and proestrous female rats.
This study also is the first to show that differences in GH mRNA are expressed by cells bearing GH or gonadotropin antigens. The lowest expression of GH mRNA was in both GH cells and gonadotropes during metestrus. The first increase in GH mRNA-bearing cells during diestrus came from cells that stored GH antigens as well as those storing LHß antigens. The timing of the increase in expression varied with the cell type, however. This suggests a sequence of events that may underlie the differentiation of this multihormonal cell.
Proposed steps toward differentiation of the multihormonal
subset
Significant changes in GH mRNA appeared first in the GH cell
population early in diestrus. This is seen as an increase in expression
of GH mRNA from 30% of GH cells during metestrus to 70% of GH cells
by the morning of diestrus. GHRH is known to be an important regulator
of the synthesis of GH mRNA (21, 22, 23, 24, 25, 26, 27, 28, 29, 30). These data suggest that the
GHRH-GH cell axis may be triggered early in the cycle to stimulate an
increase in GH mRNA synthesis.
Comparing these results with those in the previous study (10) also showed that this increased transcriptional activity from metestrus to diestrus occurs before the appearance of significant numbers of GH cells with LH or FSH mRNAs. Hence, the data suggest that GH mRNA is the first product from this multihormonal subset. At the same time, however, there is a significant increase in the percentages of LH cells with GH mRNA. Thus, collectively this suggests that the diestrous multihormonal cells are producing three products: GH and LH antigens and GH mRNA.
In our 1994 studies (10), we showed that 30% GH cells (defined by antigen content) begin to express GnRH receptors (11) late in diestrus. These data are consistent with a multihormonal cell type that contains LH antigens. Thus, by late diestrous, we hypothesize that a fourth product (GnRH receptors) is made that allows the population to respond to GnRH pulses. On the morning of proestrus, there is an expansion in the population of cells with GH antigens and LHß and FSHß mRNA. The timing of appearance suggests that the GnRH receptors produced during the previous 12 h have allowed the expression of gonadotropins that eventually reach a peak of 12.5%–15% of pituitary cells by the afternoon of proestrus (10).
Finally, the present studies show some nonparallelism when LH and FSH cells are compared. Whereas LH cells show an increase in GH mRNA during diestrus, the same change is not seen in FSH cells until proestrus. Furthermore, during peak expression periods late in proestrus, FSH is found in a higher percentage of GH cells (60%) than LH (40%, 10). Collectively, these findings suggest the later appearance of a subset of GH-FSH bearing cells. Perhaps these cells support unique FSH functions during early estrous.
Significance of the appearance of the multihormonal cells
The significance of the coexpression of GH and gonadotropins may
be related to their ability to secrete a "cocktail" of hormones
needed by the gonads, including GH. In 1993, Blumenfeld presented
evidence that GH plays a role in ovulation induction. He suggested that
GH may be a "co-gonadotropin" because it adds to or synergizes with
gonadotropins (31). There is evidence that GH is involved in a variety
of ovarian functions (31, 32, 33, 34, 35, 36). In a recent review Adashi (37) cited
several studies showing that GH is therapeutic in humans, particularly
in hypogonadotropic conditions. For example, GH may play a permissive
role in the onset of puberty. Also, patients resistant to ovulation
induction therapy could ovulate if given GH.
In rodents, human GH (hGH) prolongs the cycles (from 5 to 10 days) and the period of fertility (37). In addition, partial suppression of endogenous GH secretion resulted in delayed puberty in female rats. Rat GH replacement therapy increased ovarian progesterone production (in response to gonadotropins) and allowed normal puberty to occur. Finally, GH replacement therapy restored follicular development in some GH deficient mice (reviewed in Ref. 37). Recent studies of the Wistar derived dwarf rat reported that rhGH therapy restored the reduced ovarian weights and promoted development of intermediate follicles (38). The phenotypic defects in the reproductive system of these GH-deficient animals (39, 40) and the fact that some defects can be corrected with GH therapy (37, 38, 41, 42) points to the importance of GH in normal development of sperm and ovarian follicles.
To summarize, this study has added more information about the GH phenotypic characteristics shared by subsets of GH and gonadotropin cells in adult pituitaries. The study shows that gonadotropes may express significant levels of GH mRNA in both male and female rats, especially females in early proestrus. This supports the working hypothesis that the multihormonal cell is derived from a subset of cells that can express normal somatotropic and gonadotropic phenotypes. It adds information about the significant expression of GH mRNA in male rat gonadotropes that suggests that GH may be important in the regulation of the male reproductive system as well. The appearance of this multihormonal cell may not be solely the result of GH binding and endocytosis by gonadotropes as was suggested after our 1994 studies (10, 11), although autocrine or paracrine regulation by GH cannot be ruled out (43, 44).
Finally, the study confirms sex differences in the expression of GH mRNA reported previously, but only if male rats are compared with females rats in metestrus, diestrus, or estrus. This study is also the first to show differential expression of GH mRNA during the estrous cycle that suggests that expression might be regulated by hormones in the reproductive axis. Because the sex differences in GH expression are clearly affected by changing the steroid feedback (gonadectomy), this presents candidate hormones for further tests of this hypothesis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received October 27, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
subunit mRNA levels in the porcine fetal and neonatal
anterior pituitary. J Neuroendocrinol 9:439–449[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  C. Crane, N. Akhter, B. W. Johnson, M. Iruthayanathan, F. Syed, A. Kudo, Y.-H. Zhou, and G. V. Childs Fasting and Glucose Effects on Pituitary Leptin Expression: Is Leptin a Local Signal for Nutrient Status? J. Histochem. Cytochem., October 1, 2007; 55(10): 1059 - 1073. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Rocha, M. S. Bonkowski, L. R. Franca, and A. Bartke Mild Calorie Restriction Does Not Affect Testosterone Levels and Testicular Gene Expression in Mutant Mice Experimental Biology and Medicine, September 1, 2007; 232(8): 1050 - 1063. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Akhter, B. W. Johnson, C. Crane, M. Iruthayanathan, Y.-H. Zhou, A. Kudo, and G. V. Childs Anterior Pituitary Leptin Expression Changes in Different Reproductive States: In Vitro Stimulation by Gonadotropin-releasing Hormone J. Histochem. Cytochem., February 1, 2007; 55(2): 151 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Pals, M. Boussemaere, E. Swinnen, H. Vankelecom, and C. Denef A Pituitary Cell Type Coexpressing Messenger Ribonucleic Acid of Proopiomelanocortin and the Glycoprotein Hormone {alpha}-Subunit in Neonatal Rat and Chicken: Rapid Decline with Age and Reappearance in Vitro under Regulatory Pressure of Corticotropin-Releasing Hormone in the Rat Endocrinology, October 1, 2006; 147(10): 4738 - 4752. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Iruthayanathan, Y.-H. Zhou, and G. V. Childs Dehydroepiandrosterone Restoration of Growth Hormone Gene Expression in Aging Female Rats, in Vivo and in Vitro: Evidence for Actions via Estrogen Receptors Endocrinology, December 1, 2005; 146(12): 5176 - 5187. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Benoit, G. L. McCoy, and C. A. Blake Localization of Fertility Factor SP22 to Specific Cell Types Within the Anterior Pituitary Gland Experimental Biology and Medicine, November 1, 2005; 230(10): 721 - 730. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Senovilla, J. Garcia-Sancho, and C. Villalobos Changes in Expression of Hypothalamic Releasing Hormone Receptors in Individual Rat Anterior Pituitary Cells during Maturation, Puberty and Senescence Endocrinology, November 1, 2005; 146(11): 4627 - 4634. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
