| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Physiology and Pharmacology (T.A.R., A.M., M.A.B., M.J.K., O.K.R.) and Anesthesiology and Perioperative Medicine (O.K.R.), Division of Neuroscience (M.A.B., O.K.R.), Oregon National Primate Research Center, Oregon Health & Science University, Portland, Oregon 97239
Address all correspondence and requests for reprints to: Oline K. Rønnekleiv or Martin J. Kelly, Department of Physiology and Pharmacology, Mail Code L334, Oregon Health & Science University, 3181 Southwest Sam Jackson Park Road, Portland, Oregon 97239. E-mail: ronnekle{at}ohsu.edu or kellym{at}ohsu.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(PKC), cAMP-dependent protein kinase (PKA), A-kinase anchor protein (AKAP), phospholipase C (PLC), and calmodulin. Based on these findings, we dissected the arcuate nucleus from ovariectomized guinea pigs treated with estradiol benzoate (EB) or vehicle and analyzed mRNA expression using quantitative real-time PCR. We found that EB significantly increased the expression of KCNQ5 and Kv4.1 and decreased expression of KCNQ3 and AKAP in the rostral arcuate. In the caudal arcuate, EB increased KCNQ5, Kir2.4, Kv4.1, calmodulin, PKC, PLCß4, and PI3Kp55
expression and decreased Kvß1. The effects of estrogen could be mediated by estrogen receptor-
, which we found to be highly expressed in the guinea pig arcuate nucleus and, in particular, proopiomelanocortin neurons. In addition, single-cell RT-PCR analysis revealed that about 50% of proopiomelanocortin and neuropeptide Y neurons expressed KCNQ5, about 40% expressed Kir2.4, and about 60% expressed Kv4.1. Therefore, it is evident that the diverse effects of estrogen on arcuate neurons are mediated in part by regulation of K+ channel expression, which has the potential to affect profoundly neuronal excitability and homeostatic functions, especially when coupled with the rapid effects of estrogen on K+ channel function. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To address the effects of estrogen-induced gene regulation on hypothalamic functions through a genomic/gene-regulatory approach, we produced an estrogen-regulated, brain-specific cDNA library from the guinea pig using suppression subtractive hybridization (SSH). The SSH method was used specifically to isolate rare as well as abundant neuronal genes that are differentially regulated by estradiol (11). The SSH selectively amplifies target cDNA fragments and simultaneously suppresses nontarget cDNA amplification (12, 13). The differentially expressed, estrogen-regulated cDNA fragments from the SSH were used to produce a small guinea pig gene microarray chip for the analysis of estradiol’s actions on hypothalamic gene regulation (11). Previous reports analyzing the genes isolated in the SSH found a number of estrogen-regulated transcripts in the basal hypothalamus and preoptic area of the brain (11, 14). The transcripts include genes that affect synaptic transmission and neuronal excitability such as GABAB-R2, gec-1, neurobeachin (an A-kinase anchoring protein) and vesicle-associated membrane protein (vamp2). Furthermore, numerous genes involved in signal transduction such as phosphatidylinositol 3-kinase (PI3K) subunit p55 and protein kinase inhibitor (PKIG) were also found in the cDNA library. The genes identified in these reports were only about 10% of the genes isolated during the SSH process but suggested that the SSH produced a functionally relevant set of estrogen-regulated genes.
To fully use the guinea pig cDNA library and subsequent microarray gene chip, we completed the sequencing of all 2000 clones and uncovered an additional whole set of important genes in the cDNA library. Of the approximately 710 uniquely identified genes, many are involved in transcription, translation, cell growth, and synaptic transmission. Within the cDNA library, 61 sequences aligned with channels, receptors, and other membrane proteins including subunits from three types of K+ channels associated with prominent neuronal potassium currents. Over 100 sequences aligned with signaling molecules including multiple protein kinases, monomeric G proteins, and proteins involved in calcium signaling.
One rationale for microarray studies is the development of functional models of gene transcription that can be tested using other molecular and functional techniques. Preliminary microarray analysis of basal hypothalamic genes indicated that several of the K+ channels and the signal transduction molecules known to modulate their function were regulated by 24 h estradiol treatment. These K+ channels influence various aspects of neuronal excitability including the generation of action potentials and are a key element in the hormonal control of homeostatic functions and behaviors. Further examination of estrogen-induced gene regulation using quantitative real-time PCR (qPCR) analysis of mRNA extracted from the arcuate nucleus confirmed the changes in gene expression for KCNQ, Kv4, and Kir2 channel subunits and their signaling modulators by estradiol. Estrogen-induced gene regulation was also regionally dependent in the arcuate nucleus. Using these data, we developed a model cell to illustrate the potential effects of the regulation of K+ channels and signaling molecules in arcuate neurons and how estradiol may either directly or indirectly affect channel activity.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
50 µg/kg dose, n = 6) 24 h before decapitation. The dose of EB is known to produce physiological levels of plasma estradiol necessary for induction of negative and positive feedback on LH secretion (19).
Sequencing and identifying genes in the guinea pig brain-specific cDNA
As previously described (11), we used SSH to produce an estrogen-regulated, brain-specific guinea pig cDNA library. The 2000 clones selected from the SSH experiment and printed on cDNA chips were sequenced and identified. Briefly, glycerol stocks of each clone were stored in 96-well plates at –80 C. Plasmid clones were amplified by PCR using T7 and SP6 plasmid primers in 96-well plates containing 100 µM T7, 100 µM SP6, 10x thermophilic DNA polymerase buffer (Promega, Madison, WI), 1.25 mM MgCl2, 200 mM dTNPs, 2.5 U Taq DNA polymerase, and water in 100-µl reactions. PCR amplification was done using a DNA Engine Thermo Cycler (MJ Research, Waltham, MA) for 40 cycles of 94 C for 45 sec, 55 C for 45 sec, and 72 C for 150 sec. PCR products were purified using Telechem PCR clean-up plates (Telechem International, Inc., Sunnyvale, CA).
Samples were prepared for sequencing by mixing 2 µl PCR product, 1 µl 3.2 pmol/µl T7 primer, and 9 µl water in 96-well plates. To each sample, 1 µl Big Dye Terminator version 3.1 (Applied Biosystems, Foster City, CA) and 5 µl Better Buffer (dilution buffer/sequencing enhancing solution; The Gel Company, San Francisco, CA) was added. The samples were amplified by PCR using the following protocol: 96 C for 1 min for the initial denaturation, 96 C for 15 sec, 50 C for 10 sec, and 60 C for 4 min for 30–40 cycles and held at 10–16 C. To purify or remove excess dye terminators, samples were run through a 50-gauge superfine Sephadex (45 µl) column using MilliPore HV 96-well filtration plates. The Sephadex matrix was hydrated for 3 h with 300 µl milli-Q water after which the column was centrifuged at 960 x g for about 3 min. The PCR samples were transferred to individual wells and centrifuged with a collection plate at 960 x g for about 4 min. Samples were dried, and 10 µl HiDi formamide was added to rehydrate samples. The samples were denatured at 90 C for about 3 min and placed on ice for about 10 min. Finally, samples were loaded into the Applied Biosystems 3130XL Genetic Analyzer (Applied Biosystems) instrument for sequencing. The sequencing data were collected using Data Collection Software version 3.0 and analyzed using Sequencing Analysis Software version 5.2.
Sequences were cleaned and trimmed using CodonCode Aligner (CodonCode Corp., Dedham, MA). Sequences were identified in batches by NCBI’s Basic Local Alignment Search Tool (BLAST) and sorted according to function using National Human Genome Research Institute’s Gene Ontology Consortium Gene Ontology tool (20). Preliminary microarray analysis of the regulation of basal hypothalamus (BH) genes by 24 h estradiol treatment was carried out as described by the Oregon Health & Science University Center for Biomarker Discovery (11).
Primer design and testing
Guinea pig-specific primers were designed and tested for areas of high homology between multiple species by aligning cDNA sequences from our guinea pig cDNA library (11) or known guinea pig sequences with sequences from human and rodents. All primers were designed to span introns and synthesized by Invitrogen (Carlsbad, CA) using Clone Manager 5 software (Sci Ed Software, Cary, NC). See Table 1
for a listing of all the primer sets used for both qPCR and single-cell RT-PCR.
|
) corresponding to Figs. 18–22 and Figs. 23–26, respectively, from the Bleier guinea pig atlas of the hypothalamus (21). The tissue blocks were placed in RNAlater (Ambion, Austin, TX), and the rostral and caudal parts of the arcuate nucleus were dissected from the rostral and caudal blocks, respectively, using a dissecting microscope. Dissected tissue was stored at –80 C.
|
For qPCR, 4 µl cDNA template (an equivalent of 2 ng total RNA) was amplified using PowerSyber Green master mix (Applied Biosystems) on an ABI 7500 Fast Real-time PCR instrument. Standard curves for each primer pair were prepared using serial dilutions of BH cDNA in triplicate to determine the efficiency [E = 10(–1/m)– 1; m = slope] of each primer pair. All efficiencies were similar; therefore, the relative mRNA expression data were analyzed using the 
CT method, where CT is cycle threshold (Table 2) (22, 23). The amplification protocol for all the genes was as follows: 95 C for 10 min (initial denaturing) followed by 45 cycles of amplification at 94 C for 15 sec (denaturing), 60 C for 30 sec (annealing), and 72 C for 30 sec (extension) and completed with a dissociation step for melting point analysis with 35 cycles of 95 C for 15 sec, 60 C to 95 C (in increments of 1 C) for 1 min and 95 C for 15 sec. However, primers for KCNQ5 and cAMP-dependent protein kinase-1 (PKA1) were optimized with an annealing temperature of 57 C and 62 C for 30 sec, respectively. Positive and negative controls were added to each amplification run including a water blank. See Table 2 for the slope, efficiencies, and melting points for each primer set. Quantification values were generated only from samples showing a single product at the expected melting point.
|
CT). The CT values were calculated using the pooled oil-cDNA calibrator CT [CT = (CT target gene – CT reference gene) – CT of calibrator]. The relative linear quantity of target molecules was calculated using the formula 2–CT. Therefore, all transcription data are expressed as an n-fold difference relative to the calibrator. The n-fold difference was averaged for each treatment and analyzed statistically using a two-tailed Student’s t test (P < 0.05).
Single-cell RT-PCR
Twelve ovariectomized female guinea pigs were sc injected with oil (control, n = 6) or EB (25 µg/100 µl oil 50 µg/kg dose, n = 6) 24 h before being killed. Coronal hypothalamic slices (350 µm) from adult females containing the arcuate nucleus were cut on a vibratome and prepared for cell dispersions as previously described (10). Briefly, the dissected arcuate nucleus was incubated in 5–10 ml artificial cerebral spinal fluid (aCSF) containing 1 mg/ml protease for 17 min at 37 C. The tissue was washed four times in low-Ca2+ aCSF and twice in regular aCSF. The cells were isolated by triturating the tissue with flame-polished Pasteur pipettes, dispersed on a 35-mm glass-bottomed Petri dish, and perfused continuously with aCSF at a rate of 1.5 ml/min. Cells were visualized under a Leitz inverted microscope, and approximately 15 individual neurons from each female were patched and harvested by applying negative pressure. Samples of aCSF from the Petri dish were also collected before, during, and after cells were harvested. Tissue positive and negative controls and a single-cell negative control (no reverse transcriptase) were also analyzed. The cells were randomly chosen and selected on the basis of morphological appearance (i.e. not swollen and having up to three processes). The pipette contents were expelled into a siliconized microcentrifuge tube containing 0.5 µl 10x buffer, 0.38 µl Rnasin, 0.5 µl 100 mM DTT, and 3.62 µl DEPC-water (Ambion) and stored at –80 C as previously described (10).
The harvested cell solution was denatured for 5 min at 65 C and cooled on ice for 5 min, and then single-stranded cDNA was synthesized from cellular RNA by adding 50 U murine leukemia virus reverse transcriptase (Applied Biosystems), 4 µl 5x buffer, 25 mM MgCl2, 10 mM dNTP, 100 ng random hexamer primers, 40 U/µl Rnasin, and 100 mM DTT in DEPC-water in a total volume of 20 µl as previously described (10). RT was conducted using the following protocol: 60 min at 42 C, 5 min at 95 C, and 5 min at 4 C. PCR was performed using 2.5–3 µl cDNA template from each RT reaction in a 30-µl PCR mix containing the following: 3 µl 10x buffer (Promega), 25 mM MgCl2, 0.2 mM dNTP, 0.33 µM forward and reverse primers, 2 U Taq DNA polymerase and TaqStart antibody (Clontech, Palo Alto, CA). Taq DNA polymerase and TaqStart antibody were combined and incubated at room temperature for 5 min, and the remainder of the reaction content was added to the tube. Each reaction was amplified for 50 cycles using an MJ Research PTC-100 thermocycler in 0.5-ml thin-walled PCR tubes according to protocols optimized for each primer pair. Ten microliters of PCR product were visualized with ethidium bromide on a 2.5% agarose gel. For analysis of KCNQ2, -3 and -5 subunit colocalization with POMC and NPY neurons, four untreated, intact male guinea pigs were used with 19 neurons being collected from each animal.
Cloning of the guinea pig ER, mRNA expression, and immunocytochemistry
Guinea pig-specific cDNA fragments were cloned using RT-PCR. Oligonucleotide primers were designed 100% homologous to the respective human ER (GenBank accession no. NM_000125) and ERß (GenBank accession no. AB006590) sequences (forward primers: ER, 894–913 bp; ERß, 587–606 bp; reverse primers: ER, 1266–1285; ERß, 979–998). The guinea pig ER and ERß PCR fragments were amplified from 200 ng hypothalamic RNA. The result was a single 392- and 412-bp product for ER and ERß, respectively. The guinea pig fragments for ER (GenBank accession no. DQ218311) and ERß (GenBank accession no. DQ218312) were 88 and 86% homologous to the respective human sequences. Guinea pig-specific primers to be used in the qPCR were designed based on the cloned sequences for ER and ERß. Primers were designed using the Primer Express Software (Applied Biosystems). The primer sequences were as follows: guinea pig ER (56-bp product) forward primer 5'-CTGCGCAGTGTGCAATGAC-3' and reverse primer 5'-TCACAGGACCAGACCCCATAA-3' and guinea pig ERß (58-bp product) forward primer 5'-AGAACCGGCGGAAAAGCT-3' and reverse primer 5'-CATTCCTACTCCATAGCACTTTCG-3'.
Five ovariectomized female guinea pigs were injected with oil for a period of 4 wk before sedation and decapitation. Dissections of the arcuate nucleus were performed as described above. Total RNA was extracted and cDNA produced using the protocols described above. qPCR was performed in duplicate using an equivalent of 2 ng total RNA (4 µl of a 1:20 dilution of cDNA template), 0.5 µM forward and reverse primers, and the Platinum SYBR Green qPCR SuperMix-UDG kit (Invitrogen, Carlsbad, CA) in a 20-µl reaction volume. Reaction parameters using the ABI Prism 7500 were as follows: 50 C for 2 min, 95 C for 2 min, and 45 cycles of 15 sec at 94 C and 30 sec at 60 C, followed by a dissociation step from 60 C to 95 C. The dissociation step revealed a single melting peak for all amplicons. Standard curves using diluted cDNA from guinea pig hypothalamus were prepared to determine the efficiency of the primers. The slopes of the standard curves for ER, ERß, and GAPDH were –3.39, –3.35, and –3.32, respectively. The efficiency was calculated for each primer pair using the following formula: E = [10(–1/m); m=slope]. The efficiencies were 97.3% for ER, 99.0% for ERß, and 100% for GAPDH. The similar efficiencies between the primers allowed us to make quantitative estimates between ER and ERß. The amplification data were analyzed by the ABI 7500 System version 1.3.0 software and calculated using the CT method with GAPDH as the reference gene. Positive tissue controls included guinea pig BH, preoptic area, and ovarian tissue. To compare between ER and ERß, the relative mRNA expression was determined by calibrating each sample to the average ERß CT and averaged for each receptor.
To document ER expression in guinea pig POMC neurons, we performed dual immunocytochemistry with a polyclonal antibody to ß-endorphin (24) and a monoclonal antibody to ER (H222; kindly provided by Dr. Geoffrey Greene, University of Chicago, Chicago, IL). Brains from ovariectomized female guinea pigs were harvested, and the BH was fixed with 4% paraformaldehyde for 4–5 h with gentle shaking at 4 C and cryopreserved in 20% sucrose overnight 4 C. Tissue blocks were frozen at –55 C, sectioned on a cryostat at 20 µm, and thaw-mounted on Superfrost Plus glass slides (Fisher Scientific, Pittsburgh, PA). The following day, sections were washed in PBS (0.1 M phosphate buffer, pH 7.4, and 0.15 M NaCl) for 30 min and then incubated for 2 h at room temperature in biotinylated IgG (donkey-antirabbit, 1:300). After rinsing in PBS, the sections were incubated for 1 h at room temperature in streptavidin-conjugated Cy3 (1:1000) and donkey-antirat IgG-conjugated Cy2 (1:1000), rinsed for 3–6 h in PBS, and covered with a coverslip applied using a glycerol-glycine buffer (2:1, pH 8.6) containing 5% n-propyl gallate to reduce photobleaching (25). Both the primary and secondary antisera were diluted in Tris-(hydroxymethyl)aminomethane (0.5%; Sigma Chemical Co., St. Louis, MO) in PB containing 0.7% seaweed gelatin (Sigma), 0.4% Triton X-100 (Sigma) and 3% BSA (Sigma) adjusted to pH 7.6. Slides were analyzed using confocal microscopy with a Leica TCS SP confocal system using a 40x NA 1.25PL APO objective. Individual sections, 0.488 µm apart, were imaged by sequential excitation with the 488-nm line of an argon (Ar) gas laser and the 561-nm line of a diode pumped solid state (DPSS) laser and projected into one plane. A total of 12 arcuate sections from two females were examined. The number of POMC-positive cells colocalizing with ER were counted per section and averaged per animal.
Data analysis
All values are expressed as mean ± SEM. All the data from the qPCR and single-cell RT-PCR experiments were analyzed using a two-tailed Student’s t test (P < 0.05) comparing the means of all the oil-treated and EB-treated samples. The gene sequences from the selected K+ channels and the signaling molecules were further analyzed to look for estrogen response elements (EREs) using the Dragon Estrogen Response Element Finder, version 2 (26). This program is a package for the specific discovery of EREs in DNA sequences. The consensus ERE is 5'-GGTCAnnnTGACC-3', where n can be any nucleotide. To model the ERE, the program uses the position weight matrix method in addition to the probability of pairing the half-sites by the transitional probabilities of the 3' nucleotide of the 5' half-site to the 5' nucleotide of the 3' half-site, ignoring spacer nucleotides. Details can be found at http://sdmc.lit.org.sg/ERE-V2/index.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) (20). Of the 2000 genes selected from the SSH, over 700 clones were identified by BLAST searches, whereas 480 were not identified, i.e. not aligned to a known sequence or the sequence was not clear. Many of the identified clones appeared multiple times in the SSH, thus accounting for the remaining 2000 clones. Among the genes identified were 1) 38 genes involved in synaptic transmission, chemokines, and endo- and exocytosis (cell-to-cell communication); 2) 61 genes localized in the cell membrane such as channels, transporters, and receptors; 3) 31 mitochondrial and electron transport genes; 4) 105 signaling molecules including kinases, phosphatases, G proteins, etc.; 5) 38 cytoskeletal proteins; 6) 109 genes involved in protein trafficking and catabolism; 7) 59 genes involved in metabolic processes and the synthesis of amino acids, lipids, and steroids, etc.; 8) 139 genes for transcription and translation; 9) 53 cell cycle and cell growth genes and, finally 10) 77 identified genes of unknown function. See supplemental Table 1 (published as supplemental data on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org) for a complete list of identified genes in the guinea pig cDNA library.
|
(PKC), PI3K p55, a catalytic (ß3) and regulatory (1) subunit of PKA, and PLC-like 1 were also identified. Many of these signaling molecules are vital participants in the rapid modulation of K+ channels, and several of these signaling molecules are known to be regulated by estradiol in a variety of cell types (14, 27, 28, 29, 30, 31).
Regulation of genes in the BH by 24 h EB treatment
In the BH, which includes the arcuate nucleus, the dorsomedial hypothalamus, and the ventromedial hypothalamus, microarray analysis showed that KCNQ5, Kir2.4, CaM, AKAP11, PKC, and PI3K p55 were regulated by 24 h EB treatment. There was no significant effect on the other identified genes. Our hypothesis that estradiol’s effects on hypothalamic nuclei potentially involve the regulation of gene expression of K+ channels and their signaling modulators was based on previous electrophysiological results and preliminary microarray analysis using SSH (9, 11). Therefore, for further analysis of estrogen-induced gene regulation, we turned to qPCR analysis of these genes in the arcuate nucleus, which regulates many homeostatic functions such as feeding and reproduction. KCNQ2 and KCNQ3 were not isolated during the SSH; however, due to their coexpression with KCNQ5 and association with the M-current, we added these two K+ channel subunits to our qPCR analysis. A PLC variant (PLC-like 1) was isolated during the SSH, but, due to PLCß4 involvement in the modulation of K+ channels, specifically the M-current (32), we chose to analyze the ß4 variant of PLC.
Expression of K+ channels and signaling molecules in the arcuate nucleus
Because of the different neuronal organization and functional diversity between the rostral and caudal arcuate nucleus (33, 34, 35, 36, 37), the arcuate nucleus was divided into two parts. The rostral arcuate block included the retrochiasmatic area as well as the more rostral regions of the arcuate nucleus (see Fig. 1
). The caudal block contained the main arcuate-median eminence region without any contribution of the ventromedial nucleus. All of the transcripts were expressed in both the rostral and caudal parts of the arcuate nucleus (Table 3). The average CT values for each transcript in the controls are shown and indicate the relative mRNA expression of each transcript in the arcuate sections. For example, CaM had average CT values in the 20- to 21-cycle range, a full 8–10 cycles earlier than Kv4.1 and Kir2.4 in both arcuate dissections, indicating a much higher expression of CaM in the arcuate nucleus than these two K+ channel subunits. To determine whether there are any differences in the relative level of transcript expression between the rostral and caudal arcuate microdissections (Fig. 3), we also analyzed the relative mRNA expression of each transcript normalized to the rostral mRNA expression. mRNA expression was not uniform for all transcripts between the rostral and caudal parts of the arcuate nucleus. The relative mRNA expression of Kir2.4, AKAP, PKC, and PLC was significantly lower in the caudal region (Fig. 3, A and B; P < 0.05). In contrast, the relative mRNA expression for PI3K p55 and KCNQ5 were significantly higher in the caudal region (P < 0.05). However, there were similar expression levels between the rostral and caudal regions for KCNQ3, KCNQ2, Kv4.1, Kvß1, CaM, and both PKA subunits.
|
|
). The KCNQ3 subunit is considered to be the necessary subunit for M-current function in the brain, colocalizing with both KCNQ2 and -5 (38). In the caudal arcuate nucleus, EB treatment increased the expression of KCNQ5, Kir2.4, and Kv4.1 but decreased the expression of the regulatory subunit for Kv4 channels, Kvß1 (Fig. 4B). The increase in KCNQ5 subunit in the caudal arcuate nucleus is similar to the effect of EB in the rostral arcuate nucleus. Unlike the rostral arcuate nucleus, Kvß1 was regulated in the caudal arcuate nucleus. Although there were changes in the expression of multiple channels in some portions of the arcuate nucleus, it is the functional interactions between the K+ channels that will ultimately determine the excitability of arcuate neurons.
|
). AKAPs are involved in localizing many signaling molecules such as PKA, PKC, and CaM to the membrane to facilitate the phosphorylation of target proteins in the membrane (39, 40, 41, 42). Although there was an apparent difference in PKAß3 mRNA expression between the treatment groups in the rostral arcuate nucleus, the difference was not statistically significant. In the caudal arcuate nucleus, EB treatment significantly up-regulated four of the modulators analyzed (CaM, PKC, PLCß4, and PI3K p55) but had no effect on AKAP, the 1 regulatory subunit of PKA, or the ß3 catalytic subunit of PKA. CaM and PKC increased by 2-fold, whereas PLCß4 and PI3K p55 by less than 2-fold (Fig. 5B).
|
). The M-current subunit, KCNQ5, was expressed in 57.6% of the cells in oil-treated females and in 38.7% in EB-treated females. Although there is an apparent lower expression in EB-treated females, the difference was not significant. Kv4.1 was expressed in 62.5% of oil-treated arcuate neurons and in 68.3% of EB-treated neurons whereas Kir2.4 was expressed in 35.9 and 38.4%, respectively. There was no change in the number of NPY-positive cells from EB treatment (42.4% in oil and 47.2% in EB); however, the number of cells expressing the POMC transcript did significantly increase (P < 0.01) in EB-treated females (65.5%) over oil-treated females (32.8%). The increase in POMC-expressing neurons is congruent with earlier data showing an increase in ß-endorphin immunostaining in female guinea pigs with the same treatment paradigm (36).
|
). The Kv4.1 channel was expressed in 64.1% of the POMC neurons and 60.9% of NPY neurons. The inwardly rectifying K+ channel, Kir2.4, was expressed in fewer POMC (36.6%) and NPY (39.4%) neurons than the other channels. No significant difference was found in the number of POMC or NPY neurons expressing any of the K+ channels harvested between rostral arcuate slices and those more caudal (data not shown).
|
) and did not significantly affect the number of arcuate neurons expressing KCNQ5 (Fig. 6A). Therefore, we decided to use intact males to examine the coexpression of the KCNQ subunits in POMC and NPY neurons using single-cell RT-PCR. Of the 76 neurons collected, 41 expressed POMC and 22 expressed NPY. In the POMC-positive neurons, 53% expressed KCNQ5, 93% expressed KCNQ3, and 87% expressed KCNQ2 (Fig. 7B). In the NPY-positive neurons, 59% expressed KCNQ5, 76% expressed KCNQ3, and 73% expressed KCNQ2. The coexpression of KCNQ3 and KCNQ5 occurred in 50 and 48% of the POMC- and NPY-positive neurons, respectively. The coexpression of KCNQ3 and KCNQ2, the primary functional M-current channel, occurred in 76 and 67% of the POMC and NPY neurons, respectively. Only one POMC neuron and one NPY neuron did not express at least one of the KCNQ subunits. The physiological significance of these findings awaits electrophysiological investigation.
ER expression in the arcuate nucleus
Based on the findings that estradiol regulates hundreds of genes including K+ channels and signaling molecules and that a few of these genes express EREs (see below), we decided to measure the expression of both ER and ERß transcripts using qPCR. Indeed, both ER and ERß were expressed in the arcuate nucleus of the female guinea pig (Fig. 8A). However, there was significantly (
4-fold) more ER mRNA vs. ERß mRNA expressed in the arcuate nucleus (Fig. 8B). In addition, we found that the majority (74%) of guinea pig POMC neurons colocalized ER protein as demonstrated by double immunohistochemical staining of ß-endorphin and ER in arcuate slices from female guinea pigs (Fig. 8C). ERß protein is not expressed in rat POMC neurons (43), but we do not know whether ERß is present in guinea pig POMC neurons. In addition, we have functional evidence that a mER is expressed in arcuate neurons including POMC neurons. Therefore, the estrogen regulation of K+ channel and signaling molecule expression may be due to multiple ER-driven processes.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The regionally dependent regulation of the channels and signaling molecules (more estrogen regulation in the caudal vs. the rostral arcuate nucleus) corresponds to the regional differences in ER distribution in the guinea pig arcuate nucleus (44) and to the regional differences in POMC and NPY localization (36, 45). The pattern of gene regulation in the arcuate nucleus by estradiol is relevant to the effects of estrogen on multiple homeostatic functions. One such function is feeding behavior, which is controlled by POMC and NPY neurons through the release of opposing anorectic and orexigenic peptides, respectively, that modulate the excitability of other hypothalamic neurons to control appetite and food intake (5, 6, 7). Estradiol up-regulates the expression of the POMC transcript and ß-endorphin immunoreactivity in POMC neurons (36, 46) while suppressing the expression and release of NPY in the paraventricular nucleus (47). However, estradiol also stimulates NPY expression during the positive feedback stage of the ovulatory cycle (reviewed in Ref. 48).
Estrogen regulation of K+ channel subunits and their modulators
The changes in channel expression in the arcuate nucleus potentially accounts for some of the known alterations in channel activity from sex steroid treatment. In the caudal arcuate nucleus, the mRNA expression of the regulatory subunit for the Kv4 channels, Kvß1, which is expressed in all the neurons, is suppressed by 24 h estradiol treatment. The Kv4 channels function as the rapidly inactivating, subthreshold A-current in neurons and are expressed throughout the central nervous system. These channels determine somatodendritic signal integration by adapting spiking behavior and controlling action potential duration and latency to first spiking to reduce the effects of depolarizing stimuli (49). The Kv4 channel subunits function as the A-current only when coexpressed with its regulatory subunit, Kvß1 (50), and are also modulated by PKA and PKC (49), PI3K (51), and Ca2+/calmodulin-dependent protein kinase II (49). The decrease in Kvß1 should result in a decrease in the A-type K+ current despite the concurrent increase in Kv4.1 expression. In the rostral arcuate nucleus, the increase in Kv4.1 expression without a change in Kvß1 expression suggests that neurons here would have an increase in A-type activity. The expression of Kv4.1 in over 61% of both POMC and NPY neurons suggests that any increase in Kv4.1 mRNA expression in the arcuate nucleus could affect these neuronal cell types.
The KCNQ channel subunits (KCNQ2, -3, and -5) are expressed throughout the CNS and form the voltage-sensitive, neuronal M-current, which controls neuronal excitability by constitutively hyperpolarizing the cell membrane (38). The M-current is ubiquitously expressed in most neuronal cell types (38). The coexpression of the KCNQ3 with KCNQ2 is the prominent heteromultimeric K+ channel combination that functions as the neuronal M-current. When coexpressed together in heterologous cells, they produce a robust current with similar properties to the native neuronal M-current. The other neuronally expressed KCNQ subunit, KCNQ5, also produces a robust M-current when coexpressed with the KCNQ3 subunit (38, 41). The M-current is modulated by numerous neurotransmitters such as acetylcholine, norepinephrine, serotonin, and glutamate (38, 52). Acetylcholine, for example, modulates the M-current through the PLC-mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2), which is a direct modulator of KCNQ channel activity, and the subsequent activation of PKC (38, 52) localized to the KCNQ subunits by an AKAP (40). The M-current is also modulated by calcium-dependent molecules such as calmodulin (53). Although many diverse neurotransmitters directly modulate the M-current, there are no reports of the effects of estradiol on the M-current either through transcriptional or signaling pathways. However, in this study, KCNQ3 is down-regulated in the rostral arcuate nucleus by estradiol with no change in the caudal arcuate nucleus, whereas KCNQ5 is up-regulated in both. In the caudal arcuate nucleus, the increase in KCNQ5, which is expressed in about 50% of the POMC and NPY cells, may alter the electrophysiological properties of the M-current because heteromultimers of KCNQ3/5 have activation and deactivation kinetics similar to the native M-current (38).
Another type of K+ channel is the inward rectifier K+ (Kir) channel family, which is grouped into five families. The Kir2 family is widely expressed throughout the central and peripheral nervous systems and are known to modulate neuronal excitability (54, 55). Kir2 channels are constitutively active and are referred to as a strongly rectifying channel of the brain, heart, skeletal muscle, and other peripheral tissues. The expression of the subunits (Kir2.1, -2.2, -2.3, and -2.4) are differentially distributed in the brain where they act as the prime determinant of the resting membrane potential (54). The Kir2 channels are inhibited by synaptic input through GPCR mechanisms that activate the signaling pathways of PIP2-PLC-PKC and PKA (54, 55, 56, 57). There is no evidence to suggest that estradiol regulates inwardly rectifying Kir2 channels. However, estradiol does reduce the amplitude of inwardly rectifying currents (Kir) in ventricular myocytes (58) and osteoclasts (59). Although Kir2.4 channels are expressed in only approximately 40% of POMC and NPY neurons, the increased expression would tend to hyperpolarize the neurons expressing Kir2.4 and decrease their sensitivity to synaptic stimuli.
The three types of K+ channels have overlapping roles in determining action potential generation, action potential waveform, and spike frequency. These roles are primarily the maintenance of the resting membrane potential (KCNQ and Kir2) by holding it at a hyperpolarized state below the spike threshold or by quickly repolarizing the membrane (Kv4) after an action potential, thereby controlling action potential firing. A reduction in KCNQ or Kir2 channels would diminish their control of the membrane potential, whereas an increase would strengthen that hold. A reduction in Kv4 channels would increase action potential firing by reducing the amount of repolarization controlled by A-type Kv4 channels. An increase in Kv4 channel expression or activity would increase the speed and depth of the repolarization, thereby limiting the number of action potentials. Indeed, the changes in K+ channel expression have been simulated by blocking the function of each, either alone or together, and measuring the number of action potential spikes (60, 61).
Changes in the signaling molecules that modulate K+ channel activity are another mechanism by which estradiol can affect K+ channel activity. The signaling molecules examined in this study are all negative effectors of K+ channel activity either via phosphorylation (PKC and PKA), modulation through direct association (CaM and AKAP), or hydrolysis of a positive effector (PLC) (39, 49, 54). In the caudal arcuate nucleus, the increase in modulator expression by estradiol may not directly affect the activity of the K+ channels without an external stimulus. The greater availability and presumably greater activity of the modulators prepare the neurons for multiple external signals from various neurotransmitters and neuromodulators (like estrogen) and potentially increases the modulation of neuronal excitability by the external signals.
Difference in gene expression and regulation between rostral and caudal arcuate nucleus
The arcuate nucleus is a heterogeneous population of neurons expressing many different transcripts often with opposing or complementary functions (33, 34, 35, 37). Expression of neuronal cell types, hormonal receptors, and other markers differs along the rostrocaudal and dorsoventral axis of the arcuate nucleus. The differences in cell types and receptor distribution in the arcuate nucleus allow for a diverse set of functions and the mechanisms by which they are controlled. For example, the distribution of both POMC and NPY neurons is not uniform in the arcuate nucleus. Although both cell types are expressed throughout the arcuate nucleus in the guinea pig, there is a greater density of POMC and NPY neurons in the caudal parts of the arcuate nucleus (36, 45). This regional difference in expression suggests that POMC and NPY neurons are a primary target for the gene regulation reported in this study. Furthermore, we saw significant differences in regional mRNA expression for two of the K+ channel subunits and three of the signaling molecules as well as differences in their regulation by estradiol.
The regional differences in distribution is also true for other neuronal cell types such as dopaminergic (tyrosine hydroxylase-positive, A12) neurons with more tyrosine hydroxylase-positive neurons localized in the rostromedial arcuate nucleus than in the caudal arcuate nucleus of the guinea pig (62). There are two types of dopaminergic neurons with cell bodies in the arcuate nucleus. The first type, tuberoinfundibular dopaminergic neurons, is located throughout the arcuate nucleus whereas the second, tuberohypophysial dopaminergic neurons, originate in the rostral arcuate nucleus. These two types of dopaminergic neurons terminate in different areas (median eminence vs. pituitary, respectively) and are differentially regulated by estrogen, progesterone, and prolactin during the ovulatory cycle (33, 34, 35).
Other steroids, peripheral peptides, and physical signals are known to have regional effects on arcuate function and gene expression. Testosterone increases POMC mRNA in castrated male rats only in the rostral arcuate nucleus (25% of the nucleus) (63). In neonatal rat, leptin, an adipocyte-secreted peptide that controls food intake, alters the gene expression of both POMC and NPY mRNA in the rostral arcuate nucleus but affects only NPY mRNA expression in the caudal arcuate nucleus (64). During lactation and suckling in the rat, mRNA expression of the orexigenic peptides, NPY and agouti-related peptide (AgRP), is increased in the caudal parts of the arcuate nucleus with little to no effect in the rostral part (65, 66).
Estrogen regulation of gene transcription through multiple pathways
In the guinea pig, the distribution of ER correlates with regional regulation of the signaling molecules and indicates that estrogen may have regional differences in regulation of arcuate gene expression, which has previously been inferred (44, 67, 68). In the guinea pig, ER is distributed throughout the arcuate nucleus with a great number of positive cells in the caudal/posterior part compared with the rostral/anterior part. ERß protein does not appear to be highly expressed in the guinea pig arcuate nucleus (44). However, as previously described (68), ER is abundantly expressed in guinea pig arcuate neurons (Fig. 8). Conversely, expression of ERß mRNA (in situ hybridization) or protein (immunocytochemistry) is much less than ER in the rodent arcuate nucleus (69), which has been confirmed by our qPCR measurements indicating that there is an approximately 4-fold greater expression of ER vs. ERß in the guinea pig arcuate nucleus. Past studies have reported that only a small percentage of POMC neurons express the classical ER in rodents (43, 70). Our results demonstrate a difference between other rodent models and the guinea pig in the expression of ER in POMC neurons. NPY colocalization with ER in the arcuate nucleus has been reported to be as low as 10% of the NPY neurons (70). However, we did not ascertain NPY and ER colocalization in this study due to the lack of specific NPY antibody for the guinea pig.
We used the online DRAGON ERE Finder program to search for potential EREs associated with the genetic sequences of the K+ channels or their modulators. Using the same human sequences used for the construction of the qPCR primers, we located predicted EREs for five of the 13 sequences. We located predicted EREs before and after the coding sequences of the KCNQ2, Kv4.1, PKC, PKAß3, and PI3K p55 genes. Apparently, three of the channel subunits regulated by estradiol in the arcuate nucleus (KCNQ3, KCNQ5, and Kvß1) do not have an ERE promoter region. Also, at least two of the modulators regulated by estradiol (CaM and PLCß4) in the caudal arcuate nucleus do not have predicted EREs. Conversely, PKAß3 has an ERE promoter but was not regulated by estradiol in the arcuate nucleus. Although a full gene promoter search will be needed to confirm the DRAGON ERE Finder information, the lack of EREs associated with some of these genes suggests the involvement of other signaling pathways in the estrogen-induced regulation of these genes in the guinea pig arcuate nucleus. The other signaling pathways available for estrogen are activation of the cAMP response element (CRE) via PKA phosphorylation of cAMP response element-binding protein (pCREB) or the interactions of ERs with DNA-binding proteins such as specificity protein-1 (SP-1) and activator protein 1 (AP-1) (9).
Estrogen modulation of K+ channel activity through multiple signaling mechanisms
Gene regulation of K+ channel mRNA expression may not be the primary mechanism for estrogen’s effects on neuronal excitability and neuropeptide secretion. In the central nervous system, specific hypothalamic neurons including GnRH, POMC, GABAergic, and dopaminergic neurons as well as neurons in the ventromedial hypothalamus are controlled by the rapid effects of estrogen functioning through at least two types of GPCR linked to either Gi,o (inhibitory) or Gq (modulatory) G proteins. The signaling pathways activated include the PLC-PKC-PKA pathway, calcium signaling pathways, and cAMP-mediated pathways (reviewed in Refs. 8 and 9). The K+ channels affected by the rapid estrogen responses include inward rectifiers (G protein-activated inwardly rectifying potassium channels or GIRK), small-conductance, calcium-activated K+ (SK) channels, (8, 9), and Kv4 A-current-type channels (71, 72). In the periphery, estrogen also directly modulates other KCNQ, Kv4, and Kir K+ channels through multiple signaling pathways in cardiovascular, reproductive, and digestive tissues (73, 74, 75, 76, 77). Therefore, an increase in the mRNA expression of signaling molecules would synchronize the arcuate neurons to the increase in estradiol levels wherein the rapid membrane-mediated effects of estrogen can have their greatest control over neuronal excitability.
Because estrogen is known to use multiple signaling pathways to modulate K+ channel function through the interaction with a membrane GPCR, we have developed a pharmacological tool, a selective mER agonist called STX, to examine the rapid, membrane-mediated effects of estrogen on K+ channel activity through signal transduction pathways (10, 78, 79). STX preferentially binds to the putative mER and not to the classical ER, thus delineating between rapid, membrane-mediated effects on neuronal excitability and the actions of the classical steroid receptors localized to the nucleus or to the cell plasma membrane. Using the mER-mediated signaling pathway determined in previous studies, we have formulated a presumptive model for the regulation of the K+ channels (KCNQ, Kv4, and Kir2) through the putative Gq-linked mER functionally described in POMC neurons (Fig. 9).
|
Summary
The effects of systemic 24 h estradiol treatment on the regulation of the hypothalamic-pituitary-gonadal axis correspond to the negative feedback phase of the ovulatory cycle and the suppression of pulsatile LH secretion. The suppression lasts for 40-plus hours in the guinea pig and is followed by a typical LH surge seen during the estrogen-induced positive feedback phase (19). Estrogen-induced gene regulation during the negative feedback stage would presumably prepare hypothalamic neurons for the preovulatory LH surge, ovulation, and the concurrent changes in many homeostatic functions occurring during the peak of estradiol levels. During this stage, the regulation of the channels and/or the signaling molecules are a part of the coordination of hypothalamic neuronal activity by direct modulation of K+ channels to initiate changes in the excitability of neurons, which may play an important role for the estrogenic effects on homeostatic functions and reproduction.
Ultimately, effects of estrogen on arcuate neuronal excitability and their modulation of other hypothalamic neurons will be determined using electrophysiological experiments. We recently reported on the increased mRNA expression of the T-type, calcium channel Cav 3.1 1 subunit in the arcuate nucleus using the same treatment paradigm. The increase in channel subunit expression correlated with a more robust T-type current in arcuate neurons, which may affect burst firing of these neurons (80). However, to remove inactivation of the T-type calcium channels and recruit more of these channels for burst firing, it is necessary to hyperpolarize the membrane (80, 81). Certainly, an up-regulation of K+ channels (i.e. Kir2.4, KCNQ5) during the critical estrogen negative feedback period would serve this function and potentially facilitate an increase in burst firing of arcuate (POMC) neurons. Also, the presumptive decrease in A-current activity (decreased Kvß1) in caudal arcuate neurons, where many of the POMC neurons are located, would also contribute to increase cell firing once these neurons reach threshold as has been shown by DeFazio and Moenter (71) in GnRH neurons. Conversely, the up-regulation of signaling molecules (PKC, PLC, etc.) that are negative modulators of K+ channels would set the stage for the positive feedback period in which synaptic input activates these molecules via G protein-coupled receptors to attenuate the activity of these channels (82). Therefore, the synergistic effects of estrogenic gene regulation and direct K+ channel modulation by estrogen is potentially a major mechanism through which estrogen controls multiple hypothalamic homeostatic systems during the ovulatory cycle.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online June 26, 2007
Abbreviations: aCSF, Artificial cerebral spinal fluid; AKAP, A-kinase anchor protein; BH, basal hypothalamus; CaM, calmodulin 1; CT, cycle threshold; DAG, diacylglycerol; DEPC, diethylpyrocarbonate; DTT, dithiothreitol; EB, estradiol benzoate; ER, estrogen receptor; ERE, estrogen response element; GPCR, G protein-coupled receptor; IP3, inositol 1,4,5-triphosphate; mER, membrane estrogen receptor; NPY, neuropeptide Y; PI3K, phosphatidylinositol 3-kinase; PIP
