Dynamin Is Required for GnRH Signaling to L-Type Calcium Channels and Activation of ERK.

We have shown that GnRH-mediated engagement of the cytoskeleton induces cell movement and is necessary for ERK activation. It also has previously been established that a dominant negative form of the mechano-GTPase dynamin (K44A) attenuates GnRH activation of ERK. At present, it is not clear at what level these cellular events might be linked. To explore this, we used live cell imaging in the gonadotrope-derived αT3-1 cell line to determine that dynamin-green fluorescent protein accumulated in GnRH-induced lamellipodia and plasma membrane protrusions. Coincident with translocation of dynamin-green fluorescent protein to the plasma membrane, we demonstrated that dynamin colocalizes with the actin cytoskeleton and the actin binding protein, cortactin at the leading edge of the plasma membrane. We next wanted to assess the physiological significance of these findings by inhibiting dynamin GTPase activity using dynasore. We find that dynasore suppresses activation of ERK, but not c-Jun N-terminal kinase, after exposure to GnRH agonist. Furthermore, exposure of αT3-1 cells to dynasore inhibited GnRH-induced cyto-architectural rearrangements. Recently it has been discovered that GnRH induced Ca(2+) influx via the L-type Ca(2+) channels requires an intact cytoskeleton to mediate ERK phosphorylation. Interestingly, not only does dynasore attenuate GnRH-mediated actin reorganization, it also suppresses Ca(2+) influx through L-type Ca(2+) channels visualized in living cells using total internal reflection fluorescence microscopy. Collectively, our data suggest that GnRH-induced membrane remodeling events are mediated in part by the association of dynamin and cortactin engaging the actin cytoskeleton, which then regulates Ca(2+) influx via L-type channels to facilitate ERK phosphorylation.

kinase C (PKC) isozymes (1,5). Additionally, GnRHR activation leads to increased cytoplasmic Ca 2ϩ mobilization in both ␣T3-1 cells and primary pituitary gonadotropes (2,3). IP 3 facilitates Ca 2ϩ mobilization from the endoplasmic reticulum via activation of the IP 3 receptors, whereas PKC mobilizes extracellular Ca 2ϩ influx through activation of voltage-gated L-type Ca 2ϩ channels (VGCCs) (4). GnRH-mediated Ca 2ϩ mobilization from both sources is essential for activation of multiple MAPK pathways in gonadotropes, including ERK, c-Jun N-terminal kinase (JNK), and p38-MAPK. Release of Ca 2ϩ from the endoplasmic reticulum facilitates JNK activation, whereas extracellular Ca 2ϩ influx through VGCCs facilitates ERK activation (6). The ERK signaling module has been studied extensively in gonadotropes and is linked to the transcriptional regulation of the LH␤ gene through induction of the immediate early gene Egr-1 (7)(8)(9). Female mice deficient in either in Egr-1 or ERK1/2 are infertile due to deficiencies in LH␤ production (10,11). Past work from our group has also shown that disruption of the actin cytoskeleton inhibits GnRH-induced activation of ERK (12). Presently, the biochemical link by which GnRHR activation causes both actin reorganization and ERK activation remains unclear.
The actin cytoskeleton is involved in important cellular functions underlying cell movement, shape, and intracellular and transmembrane protein trafficking (13,14). Previous work from our group and others has shown that gonadotropes rapidly (within 60 s) induce actin polymerization to form membrane ruffles, lamellipodia, and filopodia in the presence of GnRH (12,15). Recently, it has been found that GnRH induces actin remodeling events through the activation of cortactin, which facilitates its association with actin-related protein (Arp) 2/3 complex to induce actin branching and remodeling (16). Not only does cortactin localize to areas of dynamic actin-containing structures, it also directly binds to dynamin via its C-terminal SH3-domain in these areas (17)(18)(19).
Dynamin is a large GTPase and proline-rich domain (PRD)-containing protein that possesses mechanochemical properties important in membrane remodeling events and fission (20). There are 3 different dynamin isoforms that are selectively expressed in cells. Dynamin I is specifically expressed in neuronal cells (21), dynamin II is ubiquitously expressed (22), and dynamin III is restricted to the testis (23). Regardless of the isoform, dynamin appears to be associated with remodeling of the actin cytoskeleton (24). Overexpression of the dominant negative dynamin-K44A mutant, impaired in hydrolyzing GTP, perturbs many F-actin-rich cellular structures (25)(26)(27). Consistently, ␣T3-1 cells transfected with dynamin-K44A resulted in a loss of not only GnRH-induced actin remod-eling events (data not shown), but also GnRH-mediated ERK activation (28). Taken together, we suggest that GnRH-induced gonadotrope plasticity may be modulated through the interaction of dynamin and cortactin to effectively engage the actin cytoskeleton to facilitate ERK activation.
Our studies provide evidence that GnRHa activation of ␣T3-1 gonadotropes leads to redistribution of dynamin to the leading edge of gonadotropes where it colocalizes with the actin cytoskeleton and cortactin to facilitate actin assembly events. Inhibition of dynamin GTPase activity, via dynasore, results in suppression of actin reorganization and also phosphorylated ERK. Consistent with the loss of actin reorganization, dynasore also decreases Ca 2ϩ influx through VGCCs in response to GnRH, which consequently leads to a significant reduction in ERK phosphorylation. Taken together, we demonstrate that after Gn-RHR activation, dynamin associates with cortactin to effectively engage the actin cytoskeleton. Reorganization of actin is critical for extracellular Ca 2ϩ influx through L-type channels and subsequently ERK activation necessary for LH␤ synthesis.

Materials
The antiphospho-ERK1/2 (p-ERK), antirabbit horseradish peroxidase (HRP), antimouse-HRP antibodies, and protein A/G agarose beads were purchased from Santa Cruz Biotechnology, Inc. The Dyngo 4a inhibitor and the anticortactin and antidynamin I/II antibodies were purchased from Abcam. The anti-ERK1/2 (ERK1/2), antiphospho-JNK 1/2 (p-JNK), and anticortactin were purchased from Cell Signaling Technology. The anti-LH␤ antibody was purchased from the National Hormone and Peptide Program (National Institute of Diabetes and Digestive and Kidney Diseases). Antidynamin I/II antibody and Matrigel was purchased from BD Biosciences. All fluorescently labeled Alexa Fluor secondary antibodies were purchased from Molecular Probes-Life Technologies. The dynamin I and II primers were purchased from Integrated DNA Technologies. GnRH, Buserelin (GnRHa), dynasore, and the anti-␤-tubulin were purchased from Sigma. Glass bottom microwell dishes for confocal studies were obtained through MatTek. Phorbol 12-myristate 13-acetate (PMA) was purchased from Fisher Scientific The dynamin II-green fluorescent protein (GFP) was a generous gift from Dr Gary Whittaker (Cornell University, Ithaca, NY).

Cell culture
␣T3-1 or L␤T2 cells (a generous gift from Dr Pamela Mellon, University California, San Diego) were maintained in high glucose DMEM containing 2mM glutamine, 100-U penicillin/mL, 100-g streptomycin/mL, 10% fetal bovine serum by Life Technologies. All cells were grown in 5% CO 2 at 37°C in a humidified environment.

Ovine pituitary culture
White-faced Colorado ewes of similar age (2-4 y) and weight were maintained under natural conditions at the Animal Reproduction and Biotechnology Laboratory Sheep Facility, Fort Collins, CO. A follicular phase ewe was killed with an overdose of sodium pentobarbital, and the pituitary gland was removed aseptically. The pituitary was rinsed free of blood and cells dispersed enzymatically using collagenase, hyaluronidase, and deoxyribonuclease at 37°C for 90 minutes as previously described (29). Dissociated cells were then suspended in culture medium (DMEM supplemented with 10% horse serum [Gemini Bio-Products, Inc], 2.5% fetal bovine serum, 1% nonessential amino acids, 100-IU/mL penicillin, and 100-g/mL streptomycin). Cells (5 ϫ 10 5 ) in 2-mL media were plated in glass-bottom microwell dishes and cultured at 37°C in a humidified atmosphere of 5% CO 2 . All procedures involving animals were approved by the Colorado State University Animal Care and Use Committee and complied with National Institutes of Health guidelines (protocol number 13-4349A).

RNA extraction, reverse transcription, PCR, and quantitative RT-PCR
Total RNA was extracted from ␣T3-1 cells using E.Z.N.A. Total RNA kit (Omega) according to the manufacturer's instructions. To obtain single-stranded cDNA, 1.0 g of total RNA were reverse transcribed using iScript Reverse Transcription Supermix (Bio-Rad). cDNA was subject to PCR and quantitative polymerase chain reaction analysis with primers specific for the isoform of dynamin I (forward 5Ј-CTCACGTCCACCATCAGAAA-3Ј and reverse 5Ј-GGATGTGGTGGTCACAAT-3Ј) and dynamin II (forward 5Ј-CCCCTGGTATCCTCCAG-3Ј and reverse 5Ј-CTC-CCACGAAGCTCAGAAGA-3Ј). cDNA (0.5 g), ribonucleasefree H 2 O, along with iTaq SYBR green Supermix (10 L) (Bio-Rad) were added to each well on a 48-well PCR plate and then placed in real-time thermocycler. The samples were denatured at 94°C for 3 minutes followed by denaturation at 94°C for 30 seconds, annealing at 53.3°C for 40 seconds, and extension at 72°C for 1 minute. Quantitative PCR data were analyzed by the ␦/␦Ct method, in which all dynamin threshold cycle (Ct) values were adjusted to corresponding glyceraldehyde-3-phosphate dehydrogenase levels.

Live cell confocal microscopy
Live cell confocal microscopy was conducted using either a Zeiss LSM510 or LSM710 confocal laser-scanning microscope (CLSM) under a ϫ63 oil objective with appropriate fluorescent filters. ␣T3-1 cells (5 ϫ 10 5 ) were transiently transfected with dynamin-GFP using SuperFect reagent (QIAGEN) for 24 hours were grown on glass bottom microwell dishes coated with Matrigel (1:100 dilution). Cells were serum starved for 2-4 hours before treatment. Selected fluorescent cells were imaged before the administration of GnRHa. After the addition of 10nM GnRHa, the selected cell was imaged using a time series function that captured images every 30 seconds for 20 minutes.

Immunocytochemistry
␣T3-1 cells were grown in a 35-mm dish with a glass bottom coverslip coated with Matrigel (1:100) and serum starved for 2-4 hours. Cells were then treated with either vehicle (0) or GnRHa 10nM. After treatment, cells were washed twice with 0.01M PBS and then fixed by submersion in 4% paraformaldehyde (PFA) in PBS for 30 minutes at room temperature. The cells were then rinsed twice with PBS and permeabilized for 20 minutes in 0.5% Triton X-100/PBS. After permeabilization, cells were rinsed twice with PBS and blocked in 3% BSA/PBS for 1 hour at room temperature. Cells were incubated with an anticortactin antibody followed by Alexa Fluor 488-conjugated secondary antibody. Cells were washed and reblocked before the administration of antidynamin antibody. Dynamin was visualized using an Alexa Fluor 594-conjugated secondary antibody and imaged using CLSM. For visualization of dynamin and actin, cells were treated and stained for dynamin as described above. Actin was stained using Alexa Fluor 594 phalloidin diluted in 200-L 1% BSA/PBS and applied to the cells for 20 minutes at room temperature.
Ovine pituitary cells were serum starved for 2 hours and then administered vehicle or dynasore (80M) for 30 minutes. After pretreatment, either vehicle or 10nM GnRHa was administered or 10 minutes. Cells were washed with cold PBS and fixed in 4% PFA for 30 minutes, permeabilized with 0.5% Triton X-100 in PBS for 20 minutes, and blocked in PBS containing 3% BSA for 1 hour. Cells were then exposed to an anti-LH␤ antibody (1:250) in PBS containing 3% BSA for 2 hours. For visualization, cells were incubated with Alexa Fluor 594 antiguinea pig IgG secondary antibodies (1:500) for 1 hour. After immunostaining, pituitary cells were incubated with Alexa Fluor 488-conjugated phalloidin followed by 4Ј,6-diamidino-2-phenylindole (DAPI) and imaged by CLSM.

Phalloidin staining
␣T3-1 cells were plated in glass-bottom microwell dishes coated with Matrigel (1:100). Selected cells were pretreated with dynasore for 30 minutes and then treated in the presence or absence of either vehicle (0) or GnRHa (10nM) for 10 minutes. Cells were fixed in 4% PFA for 20 minutes and then permeabilized with 0.3% Triton X-100/PBS for 10 minutes. Cells were then blocked in 3% BSA/PBS for 20 minutes. For visualization of actin, 5 L of Alexa Fluor 488-phalloidin was diluted in 200-L 1% BSA/PBS and applied to the cells for 20 minutes at room temperature followed by DAPI staining. Cells were imaged by CLSM.

Immunoprecipitations
The ␣T3-1 cells (2 ϫ 10 7 ) were grown in 100-mm tissue culture dishes overnight. The next day, cells were serum starved for 2-4 hours and treated in the absence or presence of 10nM GnRHa for 15 minutes and lysed in RIPA buffer. Lysates were precleared and then incubated with anticortactin antibody (mouse) or nonimmune mouse serum for 2 hours at 4°C. Protein A/G agarose beads (30 L) were added, and the samples were rocked overnight at 4°C. Beads were then washed 3 times in PBS, resuspended in 40 L of 1ϫ SDS loading buffer, and boiled for 5 minutes. Immunoblots were performed using an antidynamin antibody. For loading controls, the membrane was stripped and probed using an anticortactin antibody.

Electrophysiology and total internal reflection fluorescence microscopy
␣T3-1 cells were plated onto Matrigel-coated glass-bottomed dishes (1:100) 24 hours before experimentation and Ca 2ϩ was imaged at previously described (30). Simultaneous electrophysiology and Ca 2ϩ imaging experiments were carried out using the conventional dialyzed whole-cell patch clamp technique as described previously (31)(32)(33)(34). Briefly, Ca 2ϩ influx through L-type Ca 2ϩ channels was visualized with a TILL Photonics through-the-lens total internal reflection fluorescence (TIRF) system built around an inverted Olympus IX-71 microscope with a 100X TIRF oil-immersion objective (numerical aperture 1.45) and an Andor iXON EMCCD camera (Andor Technology). To monitor Ca 2ϩ influx, gonadotropes were loaded with the Ca 2ϩ indicator fluo-5F (200M) and an excess of EGTA (10mM) to lower background noise while minimally interfering with fluo-5F via the patch pipette. The membrane potential was controlled with an Axopatch 200B amplifier (Molecular Devices); fluo-5F excitation was achieved with a 491-nm laser with excitation and emission light being separated with appropriate filters. Ca 2ϩ influx was recorded with 2mM external Ca 2ϩ at a frame rate of 50 Hz and holding potential of Ϫ70 mV to increase the driving force for Ca 2ϩ entry. To preclude potential contaminating Ca 2ϩ release events from the endoplasmic reticulum, the Ca 2ϩ -ATPase inhibitor thapsigargin (1M) was present during all experiments. ␣T3-1 cells were pretreated with vehicle or dynasore (80M) for 10 minutes before acute exposure to GnRH (3nM). Cells were imaged during the pretreatment and for an additional 10 minutes after GnRH treatment. All experiments were performed at room temperature (22°C-25°C).

L-type Ca 2؉ channel sparklet analysis
Background-subtracted fluo-5F fluorescence signals were converted to intracellular Ca 2ϩ concentrations ([Ca 2ϩ ] i ), as described previously (31,32,34,35). Briefly, fluo-5F fluorescence images were analyzed with custom software kindly supplied by L. Fernando Santana (University of Washington, Seattle, WA), and L-type Ca 2ϩ channel sparklet activity was determined by calculating the number of quantal levels detected and probability that the site is active (nP s ) of each site, where n is the number of quantal levels detected, and P s is the probability that the site is active. nP s values were obtained using pCLAMP version 10.0 (Molecular Devices) on imported [Ca 2ϩ ] i time-course records using an initial unitary [Ca 2ϩ ] i elevation of approximately 20nM as determined empirically. Active L-type Ca 2ϩ channel sparklet site densities (Ca 2ϩ sparklet sites per m 2 ) were calculated by dividing the number of active sites by the area of cell membrane visible in the TIRF images. Image stacks selected for analysis were obtained between 5 and 10 minutes of pharmacological manipulation.
Normally distributed data are presented as mea Ϯ SEM. Twosample comparisons of these data were performed using either a paired or unpaired (as appropriate) 2-tailed Student's t test, and comparisons between more than 2 groups were performed using a one-way ANOVA with Tukey's multiple comparison post hoc test. L-type Ca 2ϩ channel sparklet activity (ie, nP s ) datasets were bimodally distributed; thus, 2-sample comparisons of nP s data were examined with the nonparametric Wilcoxon matched pairs test (2 tailed), and comparisons between more than 2 groups were performed using the nonparametric Friedman test with Dunn's multiple comparison post hoc test. Arithmetic means of nP s datasets are indicated in the figures (solid gray horizontal lines) for nonstatistical visual purposes, and dashed gray lines mark the threshold for high-activity Ca 2ϩ sparklet sites (nP s Ն 0.2). Values of P Ͻ .05 were considered significant and asterisks used in the figures indicate a significant difference between groups.

Measurement of intracellular calcium
␣T3-1 cells were plated the day before at 3 ϫ 10 5 in 35-mm glass-bottomed dishes. Cells were loaded at room temperature with fura 2 (5M) and 0.1% Pluronic F-127 for 30 minutes in fluorescence buffer (FB) (145mM NaCl, 5mM KCl, 1mM Na 2 HPO 4 , 0.5mM MgCl 2 , 1mM CaCl 2 , 10mM HEPES, and 5mM glucose; pH 7.4). After loading, cells were washed twice in the same buffer and incubated for 30 minutes in FB to allow for ester hydrolysis. Treatment solutions (1-L Dimethyl sulfoxide (DMSO), as a vehicle control, 80M dynasore, 0.5M nicardipine, or both dynasore and nicardipine) were added directly into coverslip dishes before imaging. Cells were mounted on the stage of the inverted fluorescent microscope (Nikon TSM) attached to dual-wavelength Fluorescence Imaging System InCyt2 (Intracellular Imaging) and imaged for 5 minutes before and after the application of 3nM GnRH. Changes in [Ca 2ϩ ] i in individual cells were measured as the ratio of fluorescence at 340/380-nm excitation and 510-nm emission, using an InCyt2 imaging system. The individual responses of 14 -18 cells in one dish (14 -20 dishes per treatment) were aligned and averaged using the numerical analyses software (CalciumComp) developed by an engineer consultant (K. J. Bois, Fort Collins, CO). Only cells that responded to treatment were used for data analysis. Responses are reported as integrated [Ca 2ϩ ] i increase over baseline for 2 minutes.

Statistical analysis
tAll statistical analysis was performed using GraphPad prism software. Data are expressed as mean Ϯ SEM of at least 3 independent experiments. Results were analyzed for significance using an unpaired Student'sttestorone-wayANOVA.Posthocgroupcomparisonswere made using Tukey's honestly significant difference using the critical value P Յ .05 for declaring significance.

Dynamin II is predominately expressed in gonadotropes and colocalizes with the actin cytoskeleton after GnRH stimulation
To begin addressing the mechanistic link between dynamin, the actin cytoskeleton and MAPK activity, we first assessed the expression of different dynamin isoforms in our gonadotrope derived ␣T3-1 cell model. In this study, we used PCR and quantitative RT-PCR to examine the relative expression profiles of both dynamin I and II in ␣T3-1 cells. Dynamin III, expressed predominately in the testis and brain, was not found in our samples (data not shown). We found ␣T3-1 cells express approximately 35 times more dynamin II mRNA than dynamin I (Figure 1).
Dynamin possesses mechanochemical properties important in membrane remodeling events and fission (20). We have previously found that gonadotrope cell lines and ex vivo gonadotropes display a high degree of plasticity after GnRH treatment due to actin reorganization events (12,16,36). To investigate the role that dynamin may play in the initiation of actin remodeling events, we used live cell imaging to visualize dynamin-GFP redistribution after GnRHR activation. ␣T3-1 cells were transiently transfected with cDNA encoding a dynamin II-GFP fusion protein. Cells were imaged by CLSM before GnRHa treatment (10nM) and at 30 seconds intervals for 10 minutes after GnRHa activation. Upon GnRHa treatment, a portion of dynamin-GFP is redistributed from the cytosol to the plasma membrane in regions indicative of actin remodeling ( Figure 2A). We next wanted to confirm that dynamin reorganization was associated with the actin cytoskeleton. After 10 minutes of GnRHa treatment, endogenous dynamin I/II was colocalized with the actin cytoskeleton and a portion is redistributed to the leading edge of ␣T3-1 ( Figure 2B). These results suggest GnRHR activation leads to association of actin and dynamin in areas of cyto-architectural rearrangement.  press.endocrine.org/journal/endo

Cortactin associates with dynamin after GnRH treatment
We have previously highlighted the importance of cortactin in actin remodeling events in gonadotrope cells (16). Taken together with Figure 2, dynamin may potentially associate with cortactin to induce actin reorganization in gonadotropes. In support of this idea, cortactin has been shown to not only enhance the GTPase activity of dynamin, but also bind to dynamin via the interaction of the PRD on dynamin and the cortactin SH3 domain (37). Taken together, we wanted to determine whether GnRHa stimulation facilitates the association of dynamin and cortactin. ␣T3-1 cells were treated with either vehicle (0) or GnRHa for 15 minutes and lysed in a RIPA buffer. Lysates were immunoprecipitated using an anticortactin antibody, resolved by SDS-PAGE, and membranes were immunoblotted with an antidynamin I/II antibody. After 15 minutes of GnRHa treatment, there is an increased association between cortactin and dynamin ( Figure 3A). Using immunocytochemistry, we were able to visualize the spatial localization of cellular sites of cortactin and dynamin interaction. As expected, we find that after GnRHa treatment, cortactin and a portion of dynamin are redistributed and become colocalized at the leading edge of ␣T3-1 gonadotrope cells ( Figure 3B).

Inhibition of dynamin GTPase activity suppresses GnRHmediated activation of ERK but not JNK
Based on our data in the previous section, dynamin is likely to play a role in GnRH-mediated engagement of the cytoskeleton. Interestingly, our previous work also highlights that an intact cytoskeleton is necessary for ERK. At present, it is not clear at what level these cellular events might be linked. To begin exploring these issues, we used dynasore, a cell-permeable pharmacological inhibitor that interferes with GTPase activity of both dynamin I and II (38). ␣T3-1 cells were pretreated with either vehicle (0) or increasing doses of dynasore for 30 minutes. ERK phosphorylation was detected 10 minutes after GnRHa treatment. After dynasore treatment, GnRH-mediated ERK phosphorylation remained unchanged at 1M and 40M concentrations. However, dynasore treatment was capable of significantly decreasing ERK phosphorylation in a dose-dependent manner (80M and 160M) ( Figure 4A). The 80M dose of dynasore is consistent with other previously published work in neuronal cell lines and Chinese hamster ovary cells (39). To confirm the role of dynamin in specifically regulating ERK phosphorylation after GnRH exposure, we also investigated JNK activation. We detected JNK phosphorylation after 10 minutes of GnRHa treatment, which persisted despite exposure to increasing doses of dynasore ( Figure 4B). Thus, dynamin appears to be an upstream mediator involved in ERK activation but not that of JNK.
To confirm our results were not an artifact of either cell line or dynamin inhibitor choice, we validated our findings in the gonadotrope derived L␤T2 cell line. We also analyzed the effect of an additional inhibitor to dynamin GTPase activity, Dyngo 4a (dyngo). ␣T3-1 or L␤T2 cells were pretreated as described above with either dynasore (80M) or dyngo (30M). Cells were then treated with 10nM GnRHa for 10 minutes. Dose-response studies revealed that 30M of dyngo was an effective concentration for blocking dynamin GTPase activity in both cell lines (data not shown). We find that regardless of cell line or dynamin inhibitor, the loss of dynamin GTPase activity leads to a reduction in GnRH induced ERK phosphory- Cortactin associates with dynamin f after GnRHa treatment. A, ␣T3-1 cells were grown in 100-mm dishes and treated with either vehicle (0) or 10nM GnRHa for 15 minutes and lysed in RIPA buffer. Lysates were precleared and then immunoprecipitated (IP) with anticortactin antibody using protein A/G agarose beads overnight at 4°C. Immunoblotting (IB) was performed using an antidynamin. Anticortactin antibody was used to assess IP inputs. B, ␣T3-1 cells were grown on glass-bottom microwell dishes and incubated in the presence or absence of 10nM GnRHa for 10 minutes and then fixed in 4% PFA. Cells were incubated with an anticortactin antibody followed by Alexa Fluor 488-conjugated secondary antibody. Cells were washed and reblocked before the administration of antidynamin antibody. Dynamin was visualized using an Alexa Fluor 594-conjugated secondary antibody and imaged using CLSM. The red box highlights the colocalization of dynamin and cortactin at a higher resolution.

Inhibiting dynamin GTPase activity blunts actin reorganization in response to GnRH
If our hypothesis is correct that dynamin is involved in actin cytoskeletal reorganization after GnRH, presumably dynasore treatment would inhibit these cytoarchitectural changes. To explore this possibility, we assessed whether inhibition of dynamin would attenuate GnRHa-induced actin remodeling in both ␣T3-1 and primary ovine pituitary cells. ␣T3-1 and ovine pituicytes cells were pretreated with either vehicle or dynasore (80M) for 30 minutes and then treated with either 0 or GnRHa (10nM) for 10 minutes. Cells were fixed and stained for Alexa Fluor 488 phalloidin and imaged with CLSM. Our results confirm our previous studies that actin remodeling occurs after GnRHa stimulation; however, pretreatment with dynasore abrogates GnRH-induced actin reorganization events in both ␣T3-1 cells ( Figure 6A) and primary pituitary cells ( Figure 6B). Thus, dynamin is not only found associated with actin binding proteins but is involved in regulating actin dynamics in gonadotropes.

Dynasore decreases Ca 2؉ influx through L-type Ca 2؉ channels after GnRH treatment
It is well known that VGCC activity is necessary for ERK activation. Recent data have highlighted that (VGCC) activity is also dependent on an intact actin cytoskeleton (30). Because inhibition of dynamin blunted actin reorganization and ERK activation, we were intrigued with possibility that dynamin may have an effect on VGCC signaling. To gain a better understanding of the mechanistic actions of dynamin on VGCC activity we assessed channel activity after dynamin inhibition using dynasore. We used electrophysiology combined with TIRF microscopy to visualize subplasmalemmal, Figure 5. Inhibition of dynamin GTPase activity decreases p-ERK activity in multiple gonadotrope cell lines. A, ␣T3-1 cells were pretreated (30 min) with vehicle (0), dynasore (80M), or dyngo (30M) for 30 minutes followed by a 10-minute treatment of either 0 or GnRHa (10nM). Cells were lysed in RIPA buffer and Western blotted using an anti-p-ERK1/2 antibody. After probing with p-ERK, membranes were stripped and reprobed with an antibody that detects ␤-tubulin to assess protein lane loading. All groups were normalized to ␤-tubulin and are represented as fold change relative to 0. B, L␤T2 cells were treated as described in A. All groups compared with GnRH: *, P Ͻ .05. Data are expressed as mean Ϯ SEM of 3 independent experiments.  (0) or the dynamin inhibitor dynasore (1M, 40M, 80M, and 160M) followed by a 10-minute treatment of either 0 or GnRHa (10nM). Cells were lysed in RIPA buffer and Western blotted using an anti-p-ERK1/2 antibody. After probing with p-ERK, membranes were stripped and reprobed with an antibody that detects ERK1/2 to assess protein lane loading. All groups were normalized to ERK1/2 and are represented as fold change relative to 0. B, ␣T3-1 cells were treated as described in A. Cells were lysed in RIPA buffer and Western blotted using an anti-p-JNK antibody. Membranes were stripped and reprobed with ␤-tubulin to assess protein lane loading. All groups were normalized to ␤-tubulin and are represented as fold change relative to 0. All groups compared with GnRH: *, P Ͻ .05. doi: 10.1210/en.2015-1575 press.endocrine.org/journal/endo localized Ca 2ϩ influx through VGCCs (Ca 2ϩ sparklet activity) in ␣T3-1 gonadotropes (Figure 7, A and B). Acute exposure of ␣T3-1 cells to GnRH (3nM) increased both Ca 2ϩ sparklet activity (nPs) and sparklet site densities (number of Ca 2ϩ sparklet sites per m 2 ) (P Ͻ .05, n ϭ 19 cells; Figure 7C). However, pretreatment with dynasore (80M for 10 min) attenuated both sparklet activity and site densities after the administration of GnRH (3nM) and was not significantly different compared with control or dynasore alone (P Ͼ .05, n ϭ 14 cells) ( Figure  7C). These results suggest dynamin plays a key role in regulating Ca 2ϩ influx through VGCC by engaging the actin cytoskeleton.
Nicardipine and dynasore are equivalent in attenuating the GnRH-induced whole-cell calcium signal GnRH induces a transient increase in intracellular Ca 2ϩ by opening VGCCs and releasing intracellular Ca 2ϩ from IP 3 -mediated internal stores. If the effect of dynasore on Ca 2ϩ dynamics is confined strictly to the VGCC, then one would predict that its impact on the whole cell Ca 2ϩ signal as measured by fura 2 would be equivalent to the VGCC antagonist nicardipine. Consistent with earlier studies (30,40,41), the attenuated Ca 2ϩ signal in the presence of nicardipine reflects the loss of Ca 2ϩ entry via VGCCs and the retention of the IP 3 -mediated release of intracellular Ca 2ϩ stores. Dynasore led to an equivalent reduction in the whole cell Ca 2ϩ signal and the combination of nicardipine and dynasore was not additive, suggesting a common site of action. Thus, like nicardipine, the effects of dynasore on GnRH-induced Ca 2ϩ dynamics appear to be specific to the VGCC with little or no impact on IP 3 -mediated Ca 2ϩ release. Importantly, these data are entirely consistent with the effect of dynasore to attenuate ERK but not JNK activation in response to GnRH (see Figure 4), because the former is dependent on Ca 2ϩ entry via VGCC and the latter dependent on IP 3 mobilization of intracellular Ca 2ϩ (Figure 8) (6).

Dynamin facilitates GnRH-mediated ERK activation upstream of PKC
To gain a better mechanistic understanding of dynamin, we next examined where in the signaling cascade dynasore is working to suppress ERK phosphorylation after GnRHR activation. To investigate this, ␣T3-1 cells were pretreated with 0 or dynasore for 30 minutes and then treated with either 0, GnRHa, or PKC activator PMA for 10 minutes. As expected, ERK phosphorylation was detected 10 minutes after GnRH and PMA treatment. After dynasore treatment, PMA-mediated ERK phosphorylation remained unchanged (Figure 9). However, dynasore treatment significantly decreased GnRH-mediated ERK phosphorylation. These results suggest that dynamin is working upstream of PKC to mediate ERK activation after GnRHR activation.  press.endocrine.org/journal/endo 839

Discussion
In the current study, we have identified dynamin as a key signaling intermediate linking GnRHR activation to actin remodeling, VGCC activity and ERK activation. We found that upon GnRHR activation, dynamin associates with cortactin to engage the actin remodeling events. However,pharmacologicalinhibitionof dynamin GTPase activity attenuates GnRH-induced actin polymerization. Consistent with the loss of actin reorganization, dynasore also decreased Ca 2ϩ influx through VGCCs and consequently reduced ERK phosphorylation. Thus, dynamin mechano-GTPase activity appears to be important in coupling GnRHR signaling to the actin cytoskeleton and VGCC activity to facilitate ERK activation. Although dynamin plays a central role in fission of budding endocytic vesicles, it is clear that the PRD of this protein serves as a scaffold for actin interacting proteins. Consistent with this, we demonstrate that GnRHR activation leads to redistribution of dynamin and subsequent association with cortactin at the leading edge of ␣T3-1 cells. Previous work from our group has highlighted the necessity of cortactin in facilitating GnRH-mediated engagement of the actin cytoskeleton in gonadotropes (16). Beyond gonadotropes, other studies have also shown the importance of dynamin and cortactin association in membrane ruffles (18), podosomes (17), and actin comets (19). Interestingly, data show that after association, cortactin appears to stimulate GTPase activity of dynamin, and their association can form a dynamic actin filament remodeling complex (37). Furthermore, there is evidence linking the GTPase activity of dynamin to the formation of F-actin-rich cellular structures (17,42). In support of this, we show that inhibition of dynamin GTPase activity, using dynasore, blunts GnRH-induced actin remodeling. These data are consistent with overexpression of dominant negative dynamin-K44A that attenuates formation of dendritic spines of neurons (25), dynamic cortical ruffles (26,27), focal adhesions (43), and podosomes. Although cortactin is a likely candidate for stimulating GTPase activity of dynamin, necessary for actin reorganization, we cannot rule out a direct functional interaction of dynamin and   . Dynamin facilitates GnRH-mediated ERK activation upstream of PKC. ␣T3-1 cells were pretreated with vehicle (0) or dynasore (80M) for 30 minutes and then treated in the presence or absence of either GnRHa (10nM) or PMA (10nM) for 10 minutes. Cells were lysed in RIPA buffer and Western blotted using an anti-p-ERK antibody. After probing with an anti-p-ERK, membranes were stripped and reprobed with an antibody that detects ␤-tubulin to assess protein lane loading. All groups were normalized to ␤-tubulin and are represented as fold change relative to 0. All groups compared with GnRH: *, P Ͻ .05.

840
Edwards et al Dynamin Regulates Actin Dynamics and ERK Activity Endocrinology, February 2016, 157 (2):831-843 actin as it has been recently discovered that dynamin contains an actin binding domain (44). It is of interest that dynasore and Dyngo 4a suppressed lamellipodia and membrane ruffles in mouse fibroblasts despite the absence of all 3 dynamin isoforms, suggesting potential nonspecific actions of the dynamin inhibitors (45). However, these fibroblasts presented lamellipodia and membrane ruffles despite no ligand stimulation in both wild type and knockout (KO) cells. This is in sharp contrast to ␣T3-1 cells that only exhibit actin reorganization events after GnRH stimulation. Additionally, independent of pharmacological inhibition, dominant negative K44A dynamin has been shown to reduce ERK activation (28) and membrane remodeling events in both ␣T3-1 and primary sheep pituitary cells (data not shown) highlighting a specific role for dynamin in gonadotropes. Thus, it seems possible that, dynamin has a different mechanism of action in endocrine cells compared with embryonic fibroblasts.
Previous in vivo studies of dynamin I KO mice were met with limitations as synaptic transmission defects resulted in an average lifespan of less than 2 weeks (46). Although the KO of dynamin II in mice resulted in early embryonic lethality (47), development of dynamin II KO mouse fibroblasts was achieved in vitro; however, the defects in clathrin-mediated endocytosis was partially compensated by expression of the dynamin I isoform (47,48). The generation of dynamin I and II double KO mouse fibroblasts presented robust defects in clathrin-mediated endocytosis as expected, but these fibroblasts also failed to proliferate. This highlights the key role of dynamin in fission and actin modulation, because the double KO cells displayed defective invaginations of clathrin-coated pits that were also surrounded by F-actin and actin regulatory proteins.
At issue is the physiological importance of dynamin in gonadotrope cells. Previous work suggests that gonadotropes are near blood vessels and plasma membrane protrusions extend in the direction of pituitary vasculature during LH secretory events (36,49). Additionally, we have recently shown that after GnRH stimulation, ovine pituitary cells concentrate LH␤ into areas of high actin reorganization (16). Thus, dynamin GTPase activity may be key in linking GnRHR activation to gonadotrope plasticity to sustain high systemic levels of LH during the preovulatory period necessary for ovulation. Taken together, we suggest that interactions of dynamin and cortactin along with the mechano-GTPase activity of dynamin are necessary for GnRH-induced actin reorganization.
In addition to inducing actin polymerization, GnRH also maintains LH␤ synthesis through activation of ERK (50 -52). Using jasplakinolide (Jas), a pharmacological disruptor of the actin cytoskeleton, we have previously found that GnRH-induced actin engagement is required for ERK activation (12), whereas activation of JNK was not affected (Supplemental Figure 1). The importance of GnRH facilitating both actin assembly and ERK activation was also evident in HEK293 cells (53). Interestingly, this preferential suppression of ERK but not JNK activation is seen in a similar manner after the inhibition of dynamin GTPase activity, with similar blunted GnRHmediated actin reorganization phenotypes. Taken together, there appears to be an interdependence between actin reorganization and ERK activation that is regulated by dynamin GTPase activity.
To better understand the actions of dynamin GTPase activity at the level of MAPK activation, we used an approach that is based on a combination of voltage-clamp electrophysiology and TIRF microscopy that allowed for visualization of Ca 2ϩ influx through VGCCs. Our data demonstrate that dynasore exposure not only decreased the GnRH-mediated increase in Ca 2ϩ sparklet activity, but also the Ca 2ϩ sparklet site density. This potentially explains why dynasore attenuates ERK activation while JNK activation remained unaffected. JNK activation is known to be mediated by IP 3 calcium stores, not VGCC (6). Although it is known that activation of ERK by GnRH requires influx of extracellular Ca 2ϩ through VGCCs (6), we are the first to show a role for dynamin regulating ERK activation through VGCCs.
One question still unanswered is precisely where in the cell signaling cascade dynamin is engaging the actin cytoskeleton and associating with cortactin after GnRHR activation to facilitate Ca 2ϩ influx through VGCCs. Our data suggest that dynamin is working downstream of GnRHR activation but upstream of PKC, because direct stimulation of PKC with PMA leads to ERK activation in the presence of dynasore. Consistent with these findings, our previous work in ␣T3-1 cells has shown that actin cytoskeletal reorganization is also working downstream of the GnRHR but upstream of both PKC and VGCCs (30). These data suggest that ERK activation in gonadotropes is at least partially dependent on the ability of dynamin to associate with cortactin and engage the actin cytoskeleton to then drive PKC-dependent activation of VGCCs. Future studies are necessary to examine how dynamin GTPase activity and the actin cytoskeleton influence the trafficking of PKC in gonadotropes to drive VGCC activity.
In summary, our data support a model by where dynamin is a multifunctional protein not only involved in regulating membrane remodeling through the actin cytoskeleton but also mediating cell signaling events important in gonadotropin subunit synthesis. We propose a model where through GTP hydrolysis, dynamin associates with doi: 10.1210/en.2015-1575 press.endocrine.org/journal/endo cortactin and engages the actin cytoskeleton which is critical in driving PKC-dependent VGCC activity and subsequent ERK activation. This mechanism displays a novel role for dynamin that may underlie how gonadotropes maintain adequate sensitivity to incoming neuroendocrine stimulation to produce a physiological response and alignment towards anterior pituitary microvasculature.