| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Laboratory of Molecular Dynamics (F.R.B.), Department of Cell Biology and Anatomy, Medical University of South Carolina, Charleston, South Carolina 29425; Department of Medicine, University of Colorado Health Sciences Center and Research Service (M.E.W.), Veterans Affairs Medical Center, Denver, Colorado 80220; and Department of Biochemistry (G.M.L.), Medical University of South Carolina, Charleston, South Carolina 29425
Address all correspondence and requests for reprints to: Dr. Fredric R. Boockfor, Laboratory of Molecular Dynamics, Department of Cell Biology and Anatomy, Medical University of South Carolina, 173 Ashley Avenue, Charleston, South Carolina 29425. E-mail: boockfor{at}musc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Much of this lack of information has resulted from the difficulty in studying these cells. GnRH neurons represent a very small and scattered cell population [about 1300 cells in rat brain (4)], reducing the effectiveness of many traditional methods of isolation and investigation, especially those related to the study of pulsatility. Recent development of immortalized GnRH neurons (GT1 cells), as a model system, provides a particularly useful tool for study of this pulsatile process (5). Several laboratories have shown not only that cultures of these cells release GnRH in an episodic manner (6, 7, 8), but that the interpulse interval was comparable to primary cultures of hypothalamic cells obtained from rodents (8). In our laboratory, we are identifying the cellular and molecular properties of this episodic function using GT1–7 cell cultures. With single cell techniques, we found that pulsatile GnRH release is an intrinsic characteristic of individual neurons (9). This was evidenced by observations of repeated intermittent exocytotic events from a majority of the cells in culture that continued even after disruption of cell to cell communication. Next, at the synthetic level, our use of real time measurement of GnRH promoter activity revealed that GnRH gene expression also occurred in pulses and that this activity maintained a close functional relationship with exocytotic pulses. In fact, exocytotic pulse activity was shown to be necessary for the induction of gene expression pulses (10), even though transcription and translation were found not to be necessary for pulsatile GnRH release (11). This indicates that immediate replenishment of GnRH after cellular release is important to maintain the functional capacity of GnRH neurons. When taken together, these findings suggested that this pulsatile process was quite complex involving many aspects of cell function. Moreover, it appears that one of the most basic of these is intermittent activation of GnRH synthesis.
Although little is available on the process(es) underlying pulsatile GnRH promoter activity, there has been prior investigation on GnRH transcriptional regulation. In the rat, GnRH transcription is dictated mainly by the activity of two major regulatory regions within the 5' flanking region of the GnRH gene. The first region proximal to the transcriptional initiation site is highly conserved among species (12). Deoxyribonuclease I footprint analysis of the rGnRH promoter revealed at least seven regions located between -173 and 112 (12). These regions have been shown to bind transcription factors such as Oct-1 (12, 13) and SCIP/Oct-6/Tst1 (14). Furthermore, this proximal promoter segment also contains a majority of the cis-acting elements characterized for hormonal control of GnRH transcription, such as glucocorticoid-response (15), estrogen-response (16, 17), progesterone-response (18), and 12-O-tetradecanoylphorbol 13-acetate-response elements (12, 19, 20). A second region important for transcription of the rat GnRH promoter is a 300-bp enhancer, termed the neuron specific enhancer (NSE), that is located approximately 1.8 kb upstream of the transcriptional start site of the rat GnRH gene (21). This 300-bp enhancer region confers approximately a 50-fold activation of GnRH transcription in GT1–7 cells and appears to be the site of binding of a number of proteins (21). The two most well characterized of these are regulatory sequences with affinity for Oct-1 (22, 23) and GATA factors (24). When taken together, the identification and study of these two regulatory regions provide a wide variety of potential components that may play a role in pulsatility. Our goal in the following study was to identify whether one or both of these DNA regions and their respective transacting proteins may be responsible for episodic GnRH gene expression. To accomplish this, we used real-time monitoring of luciferase reporter activity in individual GT1–7 GnRH neuronal cells to evaluate the ability of different segments of the 5' flanking region of the GnRH gene to confer episodic gene expression.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
are representative examples of the activity of each of these constructs from individual GT1–7 cells. After injection, the plasmid with the largest rGnRH promoter construct (-3026/+116) (Fig. 1A) as well as one in which approximately 1 kb from the 5' flanking region was deleted (-2012/+116) (Fig. 1B) exhibited episodes of photonic activity with frequencies [0.234 ± 0.014 (n = 35) and 0.219 ± 0.034 pulses/h (n = 22), respectively], comparable to that reported previously by our laboratory (10, 25). However, when other 5' deletions were tested (-1597/+116 and -902/+116), episodic activity virtually disappeared [0.058 ± 0.018 (n = 7) and 0.032 ± 0.009 pulses/h (n = 16), respectively; Fig. 1, C and D]. Averages of pulse frequencies found with each deletion construct are presented in Fig. 1E. Interestingly, the loss of pulsatile photonic activity was first detected using the rGnRH (-1597/+116) construct, which did not contained the 415-bp region present in the rGnRH (-2012/+116) construct. These results indicate that GnRH gene expression pulsatility is a transcriptional event originating at the promoter level, and suggest that this 415-bp region present in the -2012/+116 segment is necessary for episodic GnRH promoter activity.
|
, cells injected with a control reporter plasmid containing only the RSV180 promoter rarely exhibited intermittent activity. Of eight cells tested, only one displayed a single significant fluctuation in photonic emission, yielding a total average frequency of 0.014 ± 0.015 pulses/h. In contrast, injection of the construct containing the -2012/-1597 GnRH enhancer region upstream of the RSV180 promoter resulted in profiles of photonic activity with significant fluctuations that occurred at a frequency of 0.345 ± 0.071 pulses/h (n = 12; Fig. 2B), similar to that found with -3026/+116 GnRH promoter construct. These results indicate that the region -2012/-1597 of the rGnRH promoter encompasses regulatory sequences that are not only necessary but also sufficient to generate pulses of GnRH gene expression.
|
and found GnRH gene expression pulses to occur at a frequency of 0.498 ± 0.042 pulses/h (n = 15; Fig. 3A), a value comparable to that observed in cells not receiving IgG injections. Microinjection of anti-Oct-1 abolished pulses of GnRH gene expression (0.034 ± 0.022 pulses/h; n = 7), supporting the importance of Oct-1 in the generation of bursts of rGnRH promoter activity in GT1–7 cells (Fig. 3B). Furthermore, anti-Oct-1 also induced a gradual decrease in overall promoter activity, which reached background levels by the end of the recording period. This decline in promoter activity may be a consequence of the crucial role that Oct-1 plays in maintaining the enhanced transcriptional activity of rGnRH in GT1 cells (23).
|
, photonic pulses occurred with a frequency of 0.53 ± 0.050 pulses/h (n = 10; Fig. 4A) when the wild-type GnRH promoter was injected into cells. In contrast, when a rGnRH construct containing the mutated AT-a site in the NSE was injected, we found that intermittent GnRH gene expression was virtually abolished (0.028 ± 0.020 pulses/h; n = 8; Fig. 4B). These results confirm that Oct-1 is an important component of pulsatile GnRH promoter activity and indicate that it acts by associating with the AT-a site of the NSE region.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The critical requirement of GnRH for proper reproductive function has resulted in extensive investigation on the molecular regulation of its synthesis. One of the most interesting findings was identification of a region of the rGnRH promoter between -2012 and -1597 that was responsible for neuron-specific and enhanced promoter activity in rat GnRH cells (26). This enhanced activity appears to be conferred by a 300-bp sequence, the NSE, located between -1863 and -1571 in the rat GnRH promoter (21). In the present study, our findings that GnRH gene expression pulsatility occurs only if this region is part of the promoter raise the intriguing possibility that NSE activity is involved in the generation/timing of GnRH gene expression episodes. In fact, the presence of pulsatile reporter expression, even when a fragment containing the NSE is linked to a neutral viral promoter suggests that the mechanism for pulsatility is dependent on this promoter fragment.
Several factors have been shown to bind to this NSE region and may be part of the mechanism responsible for pulsatile expression. Among the cis-acting elements located in the NSE, there are several regions highly protected by GT1 nuclear proteins in deoxyribonuclease I footprint assays. The most well characterized of these sites bind the POU homeodomain transcription factor Oct-1 at -1787/-1783 (AT-a site) and -1702/-1695 (AT-b site) of the rat GnRH promoter (23). The first of these sites has been shown to be essential for basal transcriptional activity of the GnRH enhancer (23), whereas the second was associated only with secretagogue-mediated repression of GnRH gene expression (22). Our initial experiments demonstrated that immunoneutralization of Oct-1 blocked pulses of GnRH gene expression confirming that this factor is critical for episodic activity. Because episodic activity occurred in the absence of secretagogue treatment, we reasoned that the AT-a site was more likely to be involved in this process. This was supported by our results revealing that a rGnRH promoter construct mutated at the AT-a site, when expressed, was devoid of pulsatile activity even though reporter activity was detectable. These findings demonstrate not only that Oct-1 is important for GnRH gene expression pulsatility, but that it acts (alone or in combination with other factors) by associating with the AT-a binding site in the NSE of the GnRH promoter.
Although very little information is available on the molecular process underlying pulsatile expression of GnRH, it would appear that there are several possible ways in which Oct-1 could contribute to such intermittent activity. First, changes in the synthesis of Oct-1 may dictate the on/off transcriptional activity of the GnRH gene. In this case, pulses of GnRH gene expression would directly depend on the amount of Oct-1 synthesized at any particular point in time. Sharply varying concentrations of Oct-1 in the nucleus would modify the probability of this factor to bind to its regulatory element in the NSE, thereby resulting in a changing and potentially intermittent GnRH gene expression pattern. To date, this possibility can not be confirmed or denied because the exact temporal dynamic of Oct-1 gene transcription in neuronal cells remains unknown. A second possibility would involve an interaction of Oct-1 with other soluble transcription factors or modulators that would alter the Oct-1 DNA-binding specificity and thereby up- or down-regulate GnRH transcriptional activity. This would not necessarily require a process in which synthesis or degradation of Oct-1 is carefully modulated, but would necessitate a highly regulated group of cofactors that may alter the ability of Oct-1 to elicit its effect. In favor of this, it has been shown that selective association of Oct-1 and the virion protein 16 (VP 16) results in recruitment of Oct-1 to a new cis-regulatory site that, in the absence of VP 16, is not recognized by Oct-1 (for a review see Ref. 27). Consequently, recruitment of Oct-1 to other regulatory elements that would be determined by the identity of the cofactors associated with it may rapidly modify the availability of this factor to bind to the NSE in the GnRH promoter. A third related possibility of Oct-1 involvement in intermittent function would be that the availability of Oct-1 would also be modified by an interaction with other factors, but that these factors may be fixed sites or structures within the nucleus. Interestingly, it has been shown recently that Oct-1 interacts with nuclear scaffolding proteins such as lamin B, and that the level of this interaction is an important component of gene regulation (28). In immortalized fibroblast cell lines such as HuS-L12 and IML12–4, the expression of the collagenase gene is influenced in an age-dependent manner by Oct-1 binding to lamin B in the nuclear periphery (28). When Oct-1 is anchored to lamin B in these cells, the collagenase gene is repressed. However, certain functional changes in these cells that alter the phosphorylation state (29, 30) or binding properties of Oct-1 (28) appear to cause the Oct-1/lamin B assembly to disrupt eliminating the repression of the collagenase gene. As in these other cells, Oct-1 may also form an assembly with scaffolding proteins in the nucleus of GT1 cells. Disruption of this interaction (presumably induced by some type of a cytoplasmic signal) would provide an increase in the level of Oct-1 available for binding with the NSE. Such an increase in Oct-1 would probably be more rapid than one depending on de novo protein synthesis. As this process progresses, subsequent reestablishment of the interaction of Oct-1 with the nuclear scaffolding would again limit the availability of Oct-1 and its binding to the NSE. Such events would provide an on/off mechanism that could act to generate GnRH gene expression episodes. This possibility is rather exciting, especially when viewed in light of previous results from our laboratory demonstrating that the process of granule fusion to the cell membrane is closely associated with episodes of GnRH gene expression (10). During this process, a reorganization of the cytoskeletal structures surrounding granule fusion sites in the cell membrane could be transmitted via cytoskeletal or signaling molecules to the nuclear scaffolding proteins, thereby influencing the level of Oct-1 available to bind to the octamer element in the NSE regulatory region. Further work will be needed to determine whether one or more of these potential mechanisms may underlie pulsatile expression of GnRH.
In summary, our results demonstrate clearly that GnRH gene expression pulsatility displayed by individual GT1–7 cells is a promoter-dependent phenomenon. Moreover, our findings reveal that regulation of a specific region of the GnRH promoter (e.g. AT-a region at -1787/-1783) by the POU homeodomain transcription factor Oct-1 is important for the generation of GnRH gene expression episodes. To begin to fully understand the temporal expression of the GnRH gene, it will be necessary in future studies to determine whether Oct-1 acts alone or in conjunction with other factor(s) or nuclear components to regulate the generation of these GnRH pulses.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Promoter Constructs and Microinjection
The homologous promoter constructs [rGnRH(-3026/+116), rGnRH(-2012/+116), rGnRH(-1597/+116), and rGnRH(-903/+116)] used in this study were introduced into pA3LUC reporter plasmid upstream of the coding sequence of the firefly luciferase gene as described previously (31, 32). The heterologous constructs were made using a 180-bp segment (-130/+50) of the 3'-long terminal repeat of the RSV180 with or without the rGnRH(-2012/-1597) promoter fragment inserted into the pA3LUC vector.
Oligonucleotide-directed mutagenesis of the AT-a binding site (-1787/-1783) was achieved by using a PCR-based, site-directed mutagenesis method to mutate specifically multiple consecutive bases in DNA cloned fragments (33). Briefly, a region of 11 bp corresponding to the AT-a binding site was modified to generate unique restriction sites for two enzymes (SalI and NheI). The nucleotide sequences of the mutagenic primers were as follows: AT-a-F 5'-ACT AGG TCG ACG CTA GCT AGG GCA AGA CAG AGG TTG TTC-3', and AT-a-R 5' CAT ATG CTA GCG TCG ACA GCA GGA GGA CTG TGG GGC TA-3'. The mutated construct containing the two intended restriction sites was amplified with the ELONGASE Enzyme Mix (Invitrogen Corp.). The PCR conditions were as follows: initial incubation at 95 C for 3 min and then 36 cycles of denaturation at 94 C for 40 sec, annealing at 52 C for 30 sec and elongation at 68 C for 8.5 min. The resulting PCR products were then digested with SalI, and purified using the QIAquick Gel Extraction Kit (QIAGEN Inc., Valencia, CA). Purified DNA was ligated and transformed in Escherichia coli XL-Blue MRF’ (Stratagene, La Jolla, CA). The recombinant molecules were then screened with SalI and NheI to facilitate the identification of the mutated clones and to ensure that the 11-bp region of the AT-a site was completely changed. The integrity of the intended mutation was confirmed by nucleotide sequencing.
All plasmids were purified using the EndoFree plasmid kit (QIAGEN, Inc.) and microinjected into individual cells at a concentration of 2 µg/µl using a semiautomated micromanipulator and injector system (Eppendorf, Brinkmann Instruments, Inc., Westbury, NY). The injection conditions were maintained at constant time (0.4 sec) and pressure (50–70 hPa) to ensure reproducibility in the amount of material delivered to each individual cell (25). Cells microinjected were then cultured for 24–48 h before photon counting was performed.
Immunoneutralization of Oct-1, a transcription factor known to regulate GnRH promoter activity (12, 13, 22, 23), was achieved by first microinjecting cells with the pA3rGnRHLUC reporter plasmid and 24–48 h later again microinjecting the cells with a monoclonal antibody (1 mg/ml) raised against Oct-1 (Oncogene Research Products, Boston, MA). Anti-Oct-1 was coinjected with 1.25% rhodamine dextran (10 kDa; 1:1; Sigma-Aldrich Corp., St. Louis, MO) for identification of the cells microinjected with the antibodies. As a control, cells first injected with pA3rGnRHLUC were reinjected with a solution of 1 mg/ml purified IgG1 that was of the same isotype as that of the anti-Oct-1 antibody (Sigma-Aldrich Corp.) solution, and rhodamine dextran (1:1). Immediately after microinjection of the antibody, photon emission from individual cells was recorded.
Photonic Emission
To measure the photonic activity of individual cells microinjected with different reporter plasmids, coverslips bearing the cells were mounted in Sykes-Moore chambers (Bellco Glass Inc.) and perifused continuously at a rate of 10 µl/min with high glucose DMEM (phenol-red free) supplemented with 10% fetal bovine serum, 10 mM HEPES, 0.1 mM sodium pyruvate, 4 mM glutamine, 1% penicillin/streptomycin, and 0.1 mM luciferin (Sigma-Aldrich Corp.). Chambers were then placed on the temperature-controlled stage of an Axiovert microscope (Carl Zeiss, Jena, Germany). Microinjected cells were reidentified from their coordinates in the gridded coverslips and a brightfield image was acquired for reference purposes. Photonic emission from individual cells was captured using x20 or x40 objectives over 30 or 10 min, respectively, and collected by a Hamamatsu VIM Photon Counting Camera/Argus 20 Image Processor (Hamamatsu Photonics, Hamamatsu City, Japan). Images of photonic activity were stored and later analyzed.
Data and Statistical Analyses
Because of the large heterogeneity in GnRH gene expression among individual cells, we normalized photonic emission profiles of each cell to the first point to allow baseline comparison of all cells studied. Episodes of photonic emission were first resolved with the pulse detector software PULSAR [developed by Drs. Merriam, Kozuch, and Wachter, NICHHD (Bethesda, MD) and University of California (Berkeley, CA)]. However, because occasionally PULSAR yielded significant episodes that were questionable by visual inspection, we added another criterion for pulse identification. Increases in photonic activity had to be greater than 5% of the value at the beginning of the episode plus two times the total signal-noise variation to be considered as significant fluctuations. Signal-noise variation was calculated by the square root of each value plus the SD of the corresponding background values squared [
(signal + (SD(background)2))] as reported previously (34). The signal to background ratio was always higher than 2 at the beginning of each recording. To avoid false positives, we did not consider single point fluctuations as significant episodes of promoter activity in any of the bioluminescence profiles analyzed.
Comparisons of pulse frequencies from cells microinjected with the different reporter plasmids used in this study were assessed by one-way ANOVA followed by Newman-Keuls multiple comparison test. Comparisons of pulse frequencies displayed by cells microinjected with the heterologous constructs, with anti-Oct-1 antibody, and with the AT-a site mutated GnRH promoter constructs were achieved by one-way ANOVA followed by two-tailed, unpaired t test. All data are expressed as mean ± SEM.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Current address: Department of Molecular Embryology, Pasteur Institute, 25 rue du Dr Roux, Paris 75724, France. 
Abbreviations: NSE, Neuron-specific enhancer; rGnRH, rat GnRH; RSV, Rous sarcoma virus.
Received for publication April 9, 2002. Accepted for publication May 16, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
(ER)- and ERß-expressing GT1–7 GnRH neurons. Endocrinology 140:5045–5053This article has been cited by other articles:
 |
|
 |
|
  T. R. Pak, W. C. J. Chung, J. L. Roberts, and R. J. Handa Ligand-Independent Effects of Estrogen Receptor {beta} on Mouse Gonadotropin-Releasing Hormone Promoter Activity Endocrinology, April 1, 2006; 147(4): 1924 - 1931. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-Y. Kam, K.-H. Jeong, E. R. Norwitz, E. M. Jorgensen, and U. B. Kaiser Oct-1 and Nuclear Factor Y Bind to the SURG-1 Element to Direct Basal and Gonadotropin-Releasing Hormone (GnRH)-Stimulated Mouse GnRH Receptor Gene Transcription Mol. Endocrinol., January 1, 2005; 19(1): 148 - 162. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
