Progesterone Increases the Release of Brain-Derived Neurotrophic Factor from Glia via Progesterone Receptor Membrane Component 1 (Pgrmc1)-Dependent ERK5 Signaling

Abstract

Progesterone (P4) is cytoprotective in various experimental models, but our understanding of the mechanisms involved is still incomplete. Our laboratory has implicated brain-derived neurotrophic factor (BDNF) signaling as an important mediator of P4's protective actions. We have shown that P4 increases the expression of BDNF, an effect mediated by the classical P4 receptor (PR), and that the protective effects of P4 were abolished using inhibitors of Trk receptor signaling. In an effort to extend our understanding of the interrelationship between P4 and BDNF signaling, we determined whether P4 influenced BDNF release and examined the role of the classical PR and a putative membrane PR, progesterone receptor membrane component-1 (Pgrmc1), as mediators of this response. Given recent data from our laboratory that supported the role of ERK5 in BDNF release, we also tested whether P4-induced BDNF release was mediated by ERK5. In this study, we found that P4 and the membrane-impermeable P4 (P4-BSA) both induced BDNF release from cultured C6 glial cells and primary astrocytes. Both these cells lack the classical nuclear/intracellular PR but express high levels of membrane-associated PR, including Pgrmc1. Using RNA interference-mediated knockdown of Pgrmc1 expression, we determined that P4-induced BDNF release was dependent on the expression of Pgrmc1, although pharmacological inhibition of the PR failed to alter the effects of P4. Furthermore, the BDNF release elicited by P4 was mediated by ERK5, and not ERK1/2. Collectively, our data describe that P4 elicits an increase in BDNF release from glia via a Pgrmc1-induced ERK5 signaling mechanism and identify Pgrmc1 as a potential therapeutic target for future hormone-based drug development for the treatment of such degenerative diseases as Alzheimer's disease as well as other diseases wherein neurotrophin dysregulation is noted.

A growing body of evidence suggests that progesterone (P4) has broad nonreproductive roles in the central nervous system (CNS) to regulate neural and glial functions (1). P4 is beneficial to the survival of neurons and glia after a variety of insults, such as traumatic brain injury (TBI) (2, 3), ischemia (4–6), and excitotoxicity (3, 7, 8). Clinical trials evaluating the effects of P4 on TBI have also been conducted and support the preclinical data (9–11). Despite observations in both experimental models and clinical trials of P4-mediated neuroprotection, the exact mechanisms underlying the beneficial effects of P4 have yet to be determined. It has been proposed that P4 protection involves modulation of the neurotrophin, brain-derived neurotrophic factor (BDNF) (12–16). BDNF can be synthesized and secreted by both neurons and glia to regulate a variety of brain functions, including cognition, anxiety-like behavior, and pain (17). In addition, BDNF is well known to promote the survival of several neuronal types (18–20), making P4 an intriguing therapeutic for a wide range of degenerative disorders (17). In line with these findings, we have previously shown that P4 elicited an increase in both BDNF mRNA and protein levels in the explants of the cerebral cortex and that P4 protected these cortical cultures in a manner that was dependent on neurotrophin signaling (21, 22).

The exact mechanism by which P4 regulates BDNF signaling remains unclear. Similar to other steroid hormones, P4 exerts its effects by binding to specific cellular receptors (23–25). Identified P4 receptors (PR) can be grouped into two major categories: classical nuclear/intracellular receptors (PR-A/B) and nonclassical receptors that include membrane-associated PR. The classical PR are transcription factors encoded by a single gene, which uses separate promoters and translational start sites to produce two isoforms, PR-A and -B (26). P4 influences cell function by binding to these nuclear PR and initiating the transcription of target genes. Interestingly, the effect of P4 has been reported in the brain of PR-knockout mice (27), suggesting PR other than the classical PR may mediate the effect of P4 in the CNS. Several lines of evidence recently obtained suggest that the rapid effects of P4 are mediated by cell membrane-associated PR expressed in the brain (28–31). Putative membrane PR (mPR) include seven-transmembrane domain PR (mPRα, -β, and -γ) and a single-transmembrane domain-containing receptor called progesterone receptor membrane component (Pgrmc) (32). The mPR (molecular mass approximately 40 kDa) are believed to be G protein-coupled receptors that mediate important physiological functions in male and female reproductive tracts, liver, neuroendocrine tissues, and the immune system as well as in breast and ovarian cancer (33–35). In comparison, the single-transmembrane protein Pgrmc1 (molecular mass 25–28 kDa, also known as 25-Dx in the rat) and the closely related Pgrmc2 is part of a multiprotein complex that binds to P4 and other steroids as well as pharmaceutical compounds (36). Although Pgrmc1 has been implicated in the proliferation and survival of both normal and cancer cells (37–39), very little is known regarding its potential function in the CNS. Given the aforementioned protective effects of P4 and the apparent requirement of BDNF synthesis in this process, we explored the relationship between P4 and BDNF further by investigating whether P4 influences the release of BDNF. By using complementary pharmacological and molecular approaches, we determined that P4 not only elicits the release of BDNF but also that this effect is mediated by Pgrmc-1-regulated ERK5 signaling.

Materials and Methods

Cell culture and treatment

Rat C6 glioma cells were purchased from American Type Culture Collection (Manassas, VA). Rat cortical primary astrocytes were isolated from postnatal d 4 Sprague Dawley rat pups and were supplied to us as a gift from the laboratory of Dr. James Lechleiter at University of Texas Health Science Center at San Antonio, TX. Procedures associated with generation of these cells were approved by the Institutional Animal Care and Use Committee at the University of Texas Health Science Center, San Antonio. Cells were propagated in DMEM (Invitrogen Life Technologies, Inc., Carlsbad, CA) supplemented with 10% charcoal-stripped fetal bovine serum (HyClone, Logan, UT) and maintained at 37 C in a humidified environment containing 5% CO₂. C6 cells were treated with vehicle control (dimethylsulfoxide, 0.1%), P4 (Sigma-Aldrich, St. Louis, MO), or P4-BSA (Steraloids Inc., Newport, RI) for the indicated times and concentrations. Controls for P4-BSA treatment included equimolar concentrations of BSA alone (Fisher Scientific, Fair Lawn, NJ). To avoid the potential confound of having free or unconjugated P4 in the P4-BSA preparation, the P4-BSA solution was prefiltered using a 30-kDa nominal cutoff column to remove any unconjugated steroid. The retentate (consisting of conjugated P4-BSA) was then reconstituted in the original volume.

Pharmacological inhibition of MAPK kinase (MEK), the signaling protein upstream of ERK, was achieved using the pharmacological inhibitor U0126 (1 μm; Cell Signaling Technology, Inc., Beverly, MA) or PD184352 (10 μm; gift from Pfizer Global Research and Development, Holland, MI), which were preincubated with the cells for a period of 30 min before P4 administration. Inhibition of the classical PR was achieved using the PR antagonist RU486 (1 μm; Sigma), which was applied 30 min before P4 treatment.

RNA isolation and cDNA synthesis

Total RNA was extracted from C6 cell and primary astrocyte cultures and then deoxyribonuclease treated using the RNeasy Lipid Kit (QIAGEN, Valencia, CA) according to the manufacturer's instructions. Concentrations of extracted RNA were calculated from the absorbance at 260 nm. The quality of RNA was assessed by absorption at 260 and 280 nm (A₂₆₀/A₂₈₀ ratios of 1.9–2.0 were considered acceptable). Total RNA (2 μg) was reverse transcribed into cDNA in a total volume of 20 μl using the High-Capacity DNA Archive Kit (Roche Applied Science, Indianapolis, IN) according to the manufacturer's instructions.

Primers and probes for quantitative real-time RT-PCR

PCR primers and probes for the target gene were purchased as Assay-On-Demand (Applied Biosystems Inc., Foster City, CA). The assays were supplied as a 20× mix of PCR primers (900 nm) and TaqMan probes (200 nm). The PR (Rn01448227_m1), mPRα (Paqr7, Rn02586286_s1), mPRβ (Paqr8, Rn01511662_m1), mPRγ (Rn01418618_m1), Pgrmc1 (Rn02132590_g1), BDNF (Rn02531967_s1), and ERK5 (Rn01466571_g1) assays contained FAM (6-carboxy-fluorescein phosphoramidite) dye label at the 5′ end of the probes and minor groove binder and nonfluorescent quencher at the 3′ end of the probes. The GAPDH assay (Rn9999999_s1) contained VIC-labeled probes.

Quantitative real-time RT-PCR

The reaction mixture contained water, 2× quantitative PCR Master Mix (Eurogentec, Freemont, CA), and 20× Assay-On-Demand for BDNF. A separate reaction mixture was prepared for the endogenous control, GAPDH. The reaction mixture was aliquoted in a 96-well plate, and cDNA (100 ng RNA converted to cDNA) was added to give a final volume of 30 μl. Each sample was analyzed in triplicate. Amplification and detection were performed using the ABI 7300 Sequence Detection System (Applied Biosystems) with the following profile: 2 min hold at 50 C [uracil-N-glycosylase (UNG)] and 10 min hold at 95 C, followed by 40 cycles of 15 sec at 95 C (denaturation) and 1 min at 60 C (annealing and extension). Sequence Detection Software version 1.3 (Applied Biosystems) was used for data analysis.

The comparative cycle threshold (Ct) method (2^−ΔΔCt) was used to calculate the relative changes in target gene expression (40). In the comparative Ct method, the amount of target, normalized to an endogenous control (GAPDH) and relative to a calibrator (untreated control), is given by the 2^−ΔΔCt equation. Quantity is expressed relative to a calibrator sample that is used as the basis for comparative results. Therefore, the calibrator was the baseline (vehicle-treated control) sample, and all other treatment groups were expressed as an n-fold (or percentage) difference relative to the control (40). The average and sd of 2^−ΔΔCt was calculated for the values from five independent experiments, and the relative amount of target gene expression for each sample was plotted in bar graphs using GraphPad Prism version 4 software (GraphPad, San Diego, CA).

BDNF ELISA in situ

To define the amount of endogenous BDNF released, we modified the ELISA in situ protocol developed by Promega (Madison, WI) (41). A 96-well Nunc MaxiSorp surface polystyrene flat-bottom immunoplate was precoated with an anti-BDNF monoclonal antibody [diluted 1:1000 in coating buffer containing 25 mm sodium bicarbonate and 25 mm sodium carbonate (pH 9.7)]. After rinsing off unbound antibody with Tris-buffered saline/Tween 20 buffer, TBST [20 mm Tris-HCl (pH 7.6), 150 mm NaCl, and 0.05% (vol/vol) Tween 20] and blocking the plate to minimize nonspecific binding, the culture medium was added to the plate for 2 h to equilibrate the cell growth environment. C6 cells or primary astrocytes were then plated, and after a period of time to ensure cell attachment to the plate, the appropriate treatments were applied. BDNF standards, ranging in concentration from 1.95–500 pg/ml, was added in parallel wells. At the end of hormone treatment, cells were carefully washed with TBST. The plate was then incubated with the polyclonal antihuman BDNF antibody. The amount of specifically bound polyclonal antibody was then detected through the use of the anti-IgY-horseradish peroxidase tertiary antibody (final concentration = 0.5 μg/ml), which when exposed to the chromogenic substrate (TMB reagent; Promega), changes color in proportion to the amount of BDNF present in the sample. The color intensity was quantified by measuring the absorbance at 450 nm with a Viktor3 ELISA plate reader (PerkinElmer, Waltham, MA). Only values within the linear range of the standard curve, and above the lowest standard, were considered valid. BDNF levels were normalized to protein and are reported as a percentage of vehicle control. This method allowed detection of as little as 2 pg/ml BDNF release in control cultures to approximately 250 pg/ml in P4-treated cultures.

Enriched plasma membrane (cell surface) protein isolation

The surface membrane-impermeable biotinylation of cytoplasmic membrane proteins and isolation was performed using the Cell Surface Protein Isolation Kit (Thermo Pierce, Rockford, IL), as directed by the manufacturer. Briefly, cells were washed once with ice-cold PBS, resuspended in ice-cold biotinylation mix (one 12-mg vial of M-hydroxy-sulfosuccinimide-biotin dissolved in 12 ml ice-cold PBS immediately before addition to cells) and incubated for 30 min at 4 C. The biotinylation reaction was quenched by the addition of 500 μl quenching solution. Labeled cells were washed three times in ice-cold Tris-buffered saline and pelleted by centrifugation at 500 × g for 3 min. The pellet was dispersed in 500 μl lysis buffer containing protease inhibitor (Sigma). Sonication was performed at low power on ice, and the lysate was centrifuged for 2 min at 10,000 × g. Biotinylated proteins were enriched using 500 μl NeutrAvidin agarose (Thermo Pierce) for 1 h at room temperature. The mixed lysates were centrifuged at 1000 × g for 1 min, and the supernatants were saved as the intracellular fraction. Biotinylated proteins were eluted from beads by incubation for 1 h at room temperature in 1% (wt/vol) sodium dodecyl sulfate, 50 mm dithiothreitol, and 100 mm Tris-HCl. Proteins were denatured subsequently by boiling for 5 min and then applied to Western blotting analysis.

Protein isolation from cultured cells and Western blotting

After transfection with small interfering ERK5 (siERK5) and subsequent stimulation with P4, C6 cells were harvested into lysis buffer containing protease and phosphatase inhibitors as described previously (42). After homogenization, samples were centrifuged at 99,000 × g for 15 min at 4 C, and the resulting supernatants were evaluated for total protein concentrations using the Bio-Rad DC (Bio-Rad Laboratories, Inc., Hercules, CA) protein assay kit. Sample lysates were loaded onto a sodium dodecyl sulfate/10% polyacrylamide gel, subjected to electrophoresis, and subsequently transferred onto a polyvinylidene difluoride membrane (0.22 μm pore size; Bio-Rad). The membrane was blocked for 1 h with 5% nonfat milk in 0.2% Tween-containing Tris-buffered saline solution before application of the primary antibody. The following primary antibodies were used: goat anti-Pgrmc1 (ab48058, 1:500; Abcam, Cambridge, MA), rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:2500; Cell Signaling Technology), rabbit anti-phospho-p44/42 MAPK (Thr202/Tyr204, 1:1000; Cell Signaling Technology), goat anti-ERK1 (C-16; 1:500)/goat anti-ERK2 (C-14, 1:500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), rabbit anti-phospho-ERK5 (Thr218/Tyr220, 1:1000; Cell Signaling Technology), and rabbit anti-ERK5 (1:1000; Cell Signaling Technology).

Antibody binding to the membrane was detected using a secondary antibody (either goat antirabbit or rabbit antigoat) conjugated to horseradish peroxidase (1:20,000; Pierce Chemical Co., Rockford, IL) and visualized using enzyme-linked chemiluminescence (Pierce ECL Western Blotting Substrate; Thermo Scientific, Rockford, IL) with the aid of the Alpha Innotech (San Leandro, CA) imaging system. Pgrmc1 blots that assessed the expression of this protein on the cell surface were reprobed with GAPDH to exclude the contamination of intracellular proteins in the plasma membrane fractions. Phospho-ERK1/2 and phospho-ERK5 blots were reprobed with anti- ERK1/2 or anti-ERK5 antibodies to ensure equal loading across lanes. The blots were imaged and analyzed using the FluorChem SA system (Alpha Innotech) and AlphaEase FC software. Band intensities were measured as integrated density volumes and expressed as percentage of control lane values for comparisons between treatment groups. Each experiment was repeated three times, and representative images are shown.

siRNA transfection

AllStars negative siRNA control, Pgrmc1 siRNA duplexes (5′-TTGGATAATACTCTTGGTTAA-3′), and ERK5 siRNA duplexes (5′-GGCCATCAAGAAGATACCTAA-3′) were purchased from QIAGEN. Briefly, C6 cells were transfected with the siRNA duplexes using Gene Silencer reagent (Genlantis, San Diego, CA) per manufacturer's protocol. The ratio of siRNA (micrograms) to Gene Silencer (microliters) was 1:5. RNA was isolated 24 h after transfection. Silencing of Pgrmc1 and ERK5 expression was assessed by real-time PCR. To assess the roles of Pgrmc1 and ERK5 in mediating P4-induced BDNF release, cells were plated in anti-BDNF antibody-coated Nunc MaxiSorp immunoplates, as described in BDNF ELISA in situ. Twenty-four hours after plating, cells were transfected with siRNA duplexes and returned to a CO₂ incubator for another 24 h before treatment with 1 nm P4 or P4-BSA. BDNF released into the media was measured by ELISA in situ after 18 h.

Statistical analysis

Densitometric analysis of the Western blots was conducted using Alpha Innotech Image Analysis software (Cell Biosciences, Santa Clara, CA). Densitometric data from at least three independent experiments were subjected to ANOVA, followed by Tukey's analysis for the assessment of group differences, and are presented as a bar graph depicting the average ± sem, using GraphPad Prism software.

Results

Expression profile of PR in C6 glial cells and primary astrocytes

We first determined the relative expression of classical and membrane-associated PR in both C6 glioma cells and primary cortical astrocytes. Using real-time PCR to profile the expression of PR, we found no expression of classical/nuclear PR but identified significant levels of mRNA for membrane-associated PR, especially Pgrmc1, in both C6 glial cells and primary astrocytes (presented on an ordinal scale in Table 1). The Ct values for Pgrmc1, mPRα, mPRβ, and mPRγ obtained from amplifying 20 ng template of C6 cell cDNA were 20.2 ± 0.3, 26.7 ± 1.9, 29.3 ± 0.8, and 32.9 ± 1.1, respectively. The Ct values in the primary astrocytes fell into the same range.

Table 1.

Expression profile of PR in C6 glioma cells and rat primary cortical astrocytes

Cell	Gene
	PR	mPRα	mPRβ	mPRγ	Pgrmc1	BDNF
C6 glial cells	−	+++	++	+	++++	+
Primary astrocytes	−	+++	++	+	++++	+

mRNA levels were quantified by real-time PCR using 20 ng template: −, no detectable mRNA;

⁺, Ct = 30–35;

⁺⁺, Ct = 27–30;

⁺⁺⁺, Ct = 25–27;

⁺⁺⁺⁺, Ct = 20–25. Relative abundance of transcript expression was based on the assumption of equal primer efficiency for all primers used in the amplification of the various transcripts.

P4 elicits the release of BDNF from C6 glial cells and primary astrocytes

Because astrocytes are an important source of BDNF (43–45), we sought to determine whether P4 elicits the release of BDNF from C6 cells and primary astrocytes. Detection of BDNF protein release from astrocytes has historically been technically challenging due to the low concentrations typically present in culture media (46, 47). To circumvent this problem, we developed a sensitive detection method by modifying the Promega in situ ELISA protocol, which in our hands, readily measures small quantities of endogenous BDNF released from glial cultures. Although basal BDNF levels in nonstimulated C6 cells was almost undetectable, 0.3–10 nm P4 elicited a significant increase in released BDNF (Fig. 1A). P4-BSA (a membrane-impermeable form of P4) elicited the same effect. Coupled with the lack of PR expression noted in these cells, the data support our hypothesis that a membrane-associated receptor other than the PR mediates P4's effect on BDNF release. Similarly, both P4 and P4-BSA elicited an increase in BDNF release from primary astrocytes (Fig. 1B). BSA by itself failed to elicit a statistically significant increase in BDNF release.

Fig. 1.

P4 and P4-BSA induced BDNF release in a concentration-dependent manner. C6 cells (A) or primary cortical astrocytes (B) were treated with P4, P4-BSA, or BSA at the indicated concentrations for 18 h, when BDNF release was measured by in situ ELISA. Both P4 and P4-BSA triggered a significant increase in BDNF release, whereas BSA did not. The data are representative of three independent experiments and expressed as the mean ± sem. After an ANOVA, individual comparisons were made using Tukey's post hoc analysis. Symbols denote individual group differences: *, **, and ***, P < 0.05, 0.01, and 0.001, respectively, vs. vehicle control for P4; +, ++, and +++, P < 0.05, 0.01, and 0.001, respectively, vs. vehicle control for P4-BSA; n.s., no statistically significant difference vs. vehicle control.

Expression of Pgrmc1 on cell surface in C6 glial cells and primary astrocytes

The expression profile of PR, in conjunction with the fact that cell-impermeable P4-BSA triggers BDNF release from glia, strongly implied the involvement of membrane-associated PR in mediating P4's effects. This abundance of Pgrmc1 (assuming equal efficiency of the different primers used) relative to the expression of the other membrane-associated PR prompted our analysis of Pgrmc1 in mediating the effects of P4 on BDNF release. Depending on the cell type examined, Pgrmc1 has been reported to be localized in the endoplasmic reticulum (48, 49), nuclei (50, 51), mitochondria (52), and cytoplasmic membrane (53, 54). To confirm the localization of Pgrmc1 on cell surface in glia, we evaluated the expression of Pgrmc1 protein expression on the cell surface (i.e. plasma membrane). Cell surface proteins were biotinylated in nonpermeablized C6 cells and primary astrocytes. Subsequently, biotin-labeled cell surface proteins were separated from the intracellular fraction using NeutrAvidin agarose purification. The isolated cell surface proteins were then subjected to Western blot analysis, which revealed an approximately 25 kDa Pgrmc1 band. A similar band was noted in the intracellular fraction as well (Fig. 2), supporting previous studies that Pgrmc1 is localized on both the cell surface (i.e. plasma membrane) and intracellular compartments. Reprobing of the blots with GAPDH confirmed the relative purity of the plasma membrane preparation.

Fig. 2.

Subcellular localization of Pgrmc1 in C6 glioma cells and rat primary cortical astrocytes. Cell surface proteins were labeled with biotin before cell lysis. Plasma membrane fractions were then separated from the intracellular fractions with NeutrAvidin agarose and applied to Western blotting for analysis of Pgrmc1 expression. The same blot was reprobed with GAPDH to ensure no contamination from the intracellular fraction.

Pgrmc1 mediates P4's effect on BDNF release

To determine the role of Pgrmc1 in mediating the effect of P4, we used RNA interference (RNAi)-mediated gene depletion to knock down the expression of Pgrmc1 mRNA in C6 cells. The siRNA duplex successfully decreased Pgrmc1 mRNA level to approximately 5% of control (Fig. 3A). No off-target knockdown of other mPR was noted (data not shown). Western blotting confirmed the effective decrease of Pgrmc1 protein levels as well (Fig. 3B). Importantly, this Pgrmc1 depletion completely abolished the BDNF release elicited by P4 or P4-BSA treatment (Figs. 3C and 4). These results suggest that Pgrmc1 mediates the effect of P4 on BDNF release.

Fig. 3.

Pgrmc-1 knockdown abolished P4-induced BDNF release. RNAi efficiently decreased the expression of Pgrmc1 in C6 cells. C6 cells were transfected with mock (transfection reagent only, no siRNA duplex), scrambled siRNA (siControl), or siRNA against Pgrmc-1 (siPgrmc-1) for 24 h. mRNA level of Pgrmc1 was determined by quantitative PCR (A), and protein level of Pgrmc1 was determined by Western blotting (B). Pgrmc1 mRNA level is expressed as a percentage of that seen in the scrambled siRNA control-transfected group (set at 100%). ****, P < 0.0001 vs. siControl. C, siPgrmc-1 abolished P4-induced BDNF release. C6 cells were transfected with mock, siControl, or siPgrmc1 for 24 h before cells were stimulated with P4 (1 nm, 18 h). Data are presented as the mean ± sem. ***, P < 0.001 vs. siControl; ###, P < 0.001 vs. siControl + P4.

Fig. 4.

Pgrmc-1 knockdown abolished P4-BSA-induced BDNF release. C6 cells were transfected with mock, scrambled siRNA (siControl), or siPgrmc1 for 24 h before cells were stimulated with P4-BSA (1 nm, 18 h). BDNF release was assessed by in situ ELISA. Data are presented as the mean ± sem ***, P < 0.001 vs. siControl; ###, P < 0.001 vs. siControl + P4-BSA.

P4 activates ERK5 signaling pathway via Pgrmc1

The MAPK pathway is a known downstream target of membrane-initiated P4 signaling in various types of tissues (55–57). In addition to the well-studied ERK1/2 pathway, ERK5 is a more recently identified MAPK family member that has been implicated in cardiovascular development and neuronal differentiation (58). Both ERK1/2 and ERK5 can be activated by growth factors such as epidermal growth factor and nerve growth factor and have overlapping expression patterns and functions (59–61). However, recent studies suggest that ERK5 also has unique functions distinct from those of ERK1/2 (58, 62). We recently reported that ERK5 and ERK1/2 pathways mediate opposite effects with regard to the regulation of BDNF mRNA levels in C6 cells (42). To investigate whether P4 activates either of these members of the ERK/MAPK family, we assessed the levels of phosphorylated/activated ERK1/2 or ERK5 from P4-treated C6 cells. Indeed, P4 treatment resulted in an increase in the phosphorylation of both ERK1/2 and ERK5 as early as 5 min, with a peak at 10 min (data not shown). To determine whether this effect was attributed to Pgrmc1, we assessed the effect of P4 in cells whose expression of Pgrmc1 was diminished using siRNA. The effect of P4 on ERK5 was completely blocked by knockdown of Pgrmc1 expression, whereas the effect of P4 on ERK1/2 phosphorylation was unaltered (Fig. 5). These data support the role of Pgrmc1 in mediating the effect of P4 on ERK5.

Fig. 5.

Pgrmc-1 mediates P4-induced activation of ERK5, but not ERK1/2. C6 cells were transfected with siRNA for 24 h before treatment with P4 (1 nm) for 10 min. Western blot analysis revealed that P4 elicited an increase in both ERK1/2 and ERK5 phosphorylation in scrambled siRNA (siControl)-transfected groups (p-ERK; upper right panels). In contrast, siPgrmc-1 transfection blocked P4-induced ERK5 phosphorylation, whereas no effect on ERK1/2 phosphorylation was noted. The lower right panels represent reprobing of the phospho-blot for total ERK protein to ensure equal loading across lanes. Band intensity was quantified by densitometry and normalized to total protein (left panels). Phosphorylation level is expressed as fold of the scrambled siRNA control-transfected group (set at 1). ** or ***, P < 0.01 or 0.001 vs. siControl; ##, P < 0.01 vs. siControl + P4; n.s., no statistically significant difference.

Effects of pharmacological inhibitors against ERK and classical PR on P4-induced BDNF release

The role of the classical PR in mediating P4-induced BDNF release was assessed pharmacologically, using the PR antagonist RU486 (63). RU486 failed to alter the effects of P4 or P4-BSA on BDNF release (Fig. 6, A and B). In addition, we used pharmacological inhibitors of the ERK/MAPK pathway to identify potential differences in the role of ERK1/2 and ERK5 in P4-induced BDNF release. U0126, a chemical inhibitor of both the ERK5 and the ERK1/2 signaling pathways (64), was effective at inhibiting P4-induced BDNF release (Fig. 6A). However, when we used the more selective MEK1 inhibitor PD184352, which inhibits ERK1/2 phosphorylation but not ERK5 phosphorylation/activity (42), no statistically significant attenuation of the effect of P4 was noted and suggested that the effects of P4 on BDNF release were mediated preferentially by ERK5 and not ERK1/2.

Fig. 6.

P4- and P4-BSA-induced BDNF release is inhibited by U0126 but not by RU486 or PD184352. C6 cells were treated with the PR antagonist RU486 (1 μm), the ERK1/2 and ERK5 inhibitor U0126 (1 μm), or the ERK1/2-specific inhibitor PD184352 (10 μm) for 30 min before P4 (A) or P4-BSA (B) (1 nm, 18 h) administration. BDNF release was measured by in situ ELISA. P4- and P4-BSA-induced BDNF release was blocked by U0126 but not by RU486 or PD184352. Data are presented as the mean ± sem. **, P < 0.01 vs. vehicle control; # or ##, P < 0.05 or 0.01 vs. P4 or P4-BSA group; n.s., no statistically significant difference.

RNAi-mediated ERK5 depletion abolished P4-induced BDNF release

To complement the pharmacological strategy to dissect out the role of ERK1/2 from ERK5 in mediating the effect of P4 on BDNF release, we used an siRNA-based strategy to determine whether inhibition of ERK5 expression attenuated or abolished the effect of P4 on BDNF release. This siRNA duplex decreased ERK5 mRNA levels to approximately 45% of control (42). Although P4 significantly stimulated BDNF release in both mock- and scrambled siRNA-transfected cells, this effect was completely abolished in siERK5-transfected cells (Fig. 7), confirming the requirement of the ERK5 signaling pathway in P4's effects on BDNF release.

Fig. 7.

ERK5 knockdown abolished P4-induced BDNF release. C6 cells were transfected with mock, scrambled siRNA (siControl) or siERK5 for 24 h before cells were stimulated with P4 (1 nm, 18 h). BDNF release was assessed by in situ ELISA. Data are presented as the mean ± sem. ***, P < 0.001 vs. siControl; ###, P < 0.001 vs. siControl + P4.

Discussion

P4 is now well recognized for its involvement in diverse brain functions such as regulation of mood, cognition, inflammation, and neuronal survival (1). With regards to the latter, there currently exists both experimental and clinical evidence that supports the neuroprotective effects of P4, either in conjunction with estrogens or by itself (65–69). Mechanistically, although the intracellular (classical) PR has been studied extensively as the mediator of many of P4's effects, recent research has implicated putative membrane-associated PR and their associated downstream signaling events as equally relevant pharmacological mediators of P4's actions (70–72). Particularly for the CNS, our current understanding of the expression of different PR, and their functions, in different cell types is limited. As such, we characterized the expression of known membrane-associated PR and compared them with the expression of the classical PR in C6 glioma cells and primary cortical astrocytes (Table 1) as well as for several commonly used neuronal cell lines (73). Through such analysis, we have identified several cells that are known to respond to P4 (74, 75) but fail to express the classical PR. These cells include C6 glioma cells, primary cortical astrocytes, and HT22 hippocampal neuronal cells (73).

Glia are the most abundant cell type in the CNS (76). Traditionally viewed as support cells for neurons, glia are now known to be involved in numerous functions, including neuroinflammation, synapse formation and plasticity, and neurotrophin production (43, 77–79). The facts that P4 can influence neuronal functions through glial mechanisms (80) and that astrocytes (the most abundant macroglial cell) are the primary source of neuroprogesterone synthesized de novo in the brain (81, 82) further support the potential importance of glia as cellular mediators of P4's effects on the brain.

In this study, we provide several lines of evidence that Pgrmc1, and not the classical PR, is the mediator of P4's effect on BDNF release from glia. For example, we showed that the plasma membrane-impermeable BSA-conjugated P4 induced a similar effect on BDNF release as P4 from C6 cells and primary cortical astrocytes (Fig. 1). Given the lack of classical PR expression in either of these cells, the data strongly support the involvement of a membrane-associated PR. To exclude the possibility that chemical modification of P4 was responsible for the observed effect, we included BSA by itself as a control. Although BSA appeared to have slightly increased the level of BDNF in the media, this increase was not statistically significant.

Our data also demonstrated that the classical PR antagonist RU486 failed to block P4-induced BDNF release (Fig. 6). This lack of effect is consistent with the observed absence of classical PR expression. Moreover, the lack of effect with RU486 indirectly supports the involvement of a membrane-associated PR because it is not known to inhibit either a member of the mPR family or Pgrmc1.

We also found that the effective concentration of P4 to elicit BDNF release was significantly lower than that used in most in vitro experimental paradigms that address the neuroprotective effects of P4 (typically 100 nm) (83, 84). In fact, 0.1–10 nm P4 effectively triggered BDNF release from C6 cells (Fig. 1A), whereas higher concentrations of P4 did not, suggestive of a mechanism distinct from that of the classical PR. This range of concentrations that elicited BDNF release are in agreement with the data of Peluso et al. (85), who reported that P4 bound to Pgrmc1 with an EC₅₀ of approximately 10 nm and that concentrations of P4 from 1–10 nm inhibited apoptosis via Pgrmc1 (86). Together, we believe that our data point to a mPR, instead of the classical PR, in mediating P4-induced BDNF release. To our knowledge, this is the first report that P4 elicits the release of endogenous BDNF from glia via a membrane-associated PR mechanism.

In this report, we chose to address the involvement of Pgrmc1 first because it appeared to be the most abundant membrane-associated PR among the receptors examined (Table 1). Although some reports suggest that Pgrmc1 is primarily localized in intracellular membranes (for review, see Ref. 87), there are also studies that support the localization of Pgrmc1 on the cell surface (i.e. plasma membrane). For example, immunoreactivity of Pgrmc1 protein has been localized to cell membranes of dorsal horn neurons in spinal cord (88), sperm (89), and granulosa cells (54). In our hands, when we used an antibody to Pgrmc-1, coupled with biotin labeling of surface proteins, we identified an approximately 25-kDa band, corresponding to the reported molecular mass of Pgrmc1 in both C6 cells and primary astrocytes (Fig. 2), supporting the expression of Pgrmc-1 on the cell surface.

Our data also support the conclusion that Pgrmc1 mediated the effect of both P4 and P4-BSA on BDNF release (Figs. 3 and 4). Although we recognize that Pgrmc1 might be part of a multiprotein complex that binds to other steroids besides P4 (90), our data certainly support the indispensable role of Pgrmc1 in mediating the effects of P4 on BDNF release.

Although originally cloned in 1996 (91), relatively little information is available for Pgrmc1's biological functions to date, particularly as it relates to the brain. Pgrmc1 has been implicated as an important biomarker for cancer progression and as a potential target for anticancer therapies (86, 92). Interestingly, recent reports suggested that Pgrmc1 is the putative sigma-2 receptor (93, 94), which has been validated as a biomarker for tumor proliferation. With regard to the brain, expression of Pgrmc1 has been mapped to the cerebral cortex, hypothalamus, amygdala, and cerebellum. With respect to neuroprotection, Guennoun and colleagues (95) showed the potential role of Pgrmc1 in mediating the protective effects of P4 in the injured spinal cord and in the brain after TBI. Interestingly, these same authors showed that Pgrmc1 expression is up-regulated after TBI in both cortical neurons as well as in the astrocytes near the lesion (95), suggesting that Pgrmc1 may influence P4-stimulated glial responses after injury as well. The latter lends additional support to the biological relevance of our reported model of P4 action on glia, where P4-induced BDNF release from glia may provide support to nearby neurons.

Despite the evidence for Pgrmc1 to mediate the effects of P4, knowledge regarding its downstream signaling, particularly in the CNS, is extremely limited. However, the role of the ERK/MAPK pathway has been implicated. For example, Liu et al. (56) showed that the P4-induced increase in neural progenitor cell proliferation is dependent on Pgrmc1 and can be antagonized by the MEK inhibitor U0126. Although U0126 is widely used as a MEK1/2 (and thus, ERK1/2) inhibitor, work from our laboratory and others have demonstrated that ERK5 activation is also inhibited by U0126 (42, 96). Therefore, the potential involvement of both ERK1/2 and ERK5 pathways must be considered whenever this pharmacological agent is used. Here, we report that that ERK5, not ERK1/2, is responsible for P4-induced BDNF release (Fig. 4). In addition, knocking down Pgrmc1 expression by RNAi abolished P4-induced ERK5 activation, whereas P4-induced ERK1/2 activation was not affected (Fig. 5), suggesting that a PR other than Pgrmc1 mediates P4's effect on ERK1/2. Given that ERK5 signaling can compete with the effects of ERK1/2 in neurons and glia (42, 97), our data highlight the importance of delineating different ERK pathways.

Among the variety of cellular mechanisms implicated in P4 neuroprotection, which include the suppression of neuroinflammation (98), decreasing excitotoxicity through direct inhibition of voltage-gated calcium channels (3, 98, 99), inducing neural regeneration by increasing vascular remodeling after TBI (100), and decreasing glutamate levels (101), the regulation of neurotrophic support is also important. BDNF is a key neurotrophin responsible for the survival and function of multiple cell types in the CNS and peripheral nervous system, and its dysregulation is implicated in pathogenesis of various neurological and psychiatric disorders (102). Because of its importance in multiple disease states, our laboratory has been interested in whether P4 modulates BDNF expression and whether such regulation is relevant to P4's protective effects. In our hands, P4 not only increased the expression of BDNF mRNA and protein in cerebral cortical cultures, but the cytoprotective effects of P4 were also dependent on neurotrophin signaling (22). Interestingly, we found P4 up-regulates BDNF mRNA expression via the classical PR, because this effect is blocked by RU486 and was lost in classical PR-knockout mice (21).

Although most studies have focused on quantifying changes in cellular expression of BDNF (either at the level of mRNA or protein), we are not aware of any studies that have addressed the effects of P4 on BDNF release. Here, we report that P4 elicits a robust, concentration-dependent increase in released BDNF. We propose that the regulation of BDNF secretion from glia serves to support the survival of nearby neurons against neurotoxic insult. Whether Pgrmc1 plays a similar role in neurons is unclear at this moment but is the subject of ongoing research in the laboratory. Moreover, our data implicate an alternative (receptor) mechanism in mediating the effect of P4 on ERK1/2, signaling intermediates that are reported to be relevant for P4-induced cytoprotection. Future studies will address the involvement of other PR in greater detail to gain a more complete understanding of the mechanisms underlying P4-induced protection.

Given the fact that the classical PR is important for P4-induced BDNF expression, it is conceivable that both classical PR and Pgrmc1 work together to enable sustainable protection of the brain, such that activation of the classical PR increases the expression of BDNF, thus building a cellular reserve of BDNF, whereas activation of Pgrmc1 mediates the release of BDNF so that it may be used in an autocrine or paracrine manner to support cell viability.

In summary, our data support a novel role of Pgrmc1 in mediating P4-induced BDNF release from glial cells. In addition, the data support the involvement of ERK5 signaling as a downstream effector consequent to Pgrmc1 activation. Future studies will seek to determine whether disruption of the Pgrmc1 function contributes to neurological diseases through their regulation of BDNF level. Such information may be instrumental in helping design therapeutic regimens that employ progestins for the treatment of various diseases of the brain. And given the important role of Pgrmc1 in multiple cancers, the identification of its downstream signaling targets could stimulate the development of unique anti-(brain) cancer drugs.