Previous studies on NGF treatments of PC12 cells focused on an enhanced differentiated cell morphological profile as an endpoint that required relatively long treatment periods (upregulation of DAT after two weeks of treatment with 50 ng NGF 40). We sought to achieve functional enhanced DAT levels in a shorter time frame. To develop a cell model in which robust DAT responses could be monitored for estrogen regulation after shorter differentiation times, we chose to assay DAT levels directly via an immunofluorescence assay. We treated PC12 cells with low concentrations of NGFβ for 2, 4, and 7 days. An immunoblot of DAT protein from NGF-treated cells detected an increase in DAT protein by day 4 of NGF treatment (Fig. 1A). Fig. 1B shows the appearance of DAT on the membrane of 2 day NGF-treated cells, where DAT is evident on both cell bodies and processes. Next we adapted our immunocytochemistry conditions to a quantitative immunoassay for nonpermeabilized fixed-cells in 96-well plates, similar to a previous assay we developed for other extracellular and intracellular antigens 3541. This assay (shown in Fig. 1C) assessed only membrane DAT (under these nonpermeabilizing conditions) by recognizing the second extracellular loop of the DAT protein 42. We optimized fixation conditions to prevent antibody (Ab) from entering the cells (as shown by the low value for the clathrin Ab control). Clathrin, lying just under the membrane surface, is an abundant antigen, and gives a large signal in this assay when cells are permeabilized by adding detergent during the fixation process (not shown). The NGF-treated cells show an ~8-fold increase in DAT levels vs. untreated cells after just 2 days. These data show a concentration curve with increasing Ab to determine the single concentration of Ab that would saturate the antigen for future one-point assays (1 μg/ml). Both untreated and NGF-treated levels can be measured by this assay (both are above the background values obtained using no 1° Ab or nonspecific IgG to control for nonspecific Ab binding, and no 1° or 2° Ab to control for endogenous alkaline phosphatase contributing to the signal). We chose this two-day NGF treatment as a efficient and effective way to prepare our cells for robust functional assays of DAT level changes after E₂ treatments.

We next addressed the affect of 10 nM E₂ on DA uptake. First we looked at total DA uptake in PC12 cells (not blocked by any specific inhibitors to define a particular mechanism), and found that general uptake was inhibited by E₂ (Fig 2A). We next added the specific DAT inhibitor nomifensine to define uptake specifically mediated by DAT (Fig. 2B). Again, E₂ significantly blocked DAT-specific DA uptake. DAT-specific uptake was enhanced in NGF-treated cells, in agreement with the increased levels of DAT shown in Fig. 1. NGF-enhanced DAT activity was inhibited to a similar extent as the basal levels. Finally, though a 30-min treatment with E₂ is considered a relatively short period of time for E₂ to act (and so could be attributable to the nongenomic pathway), some may still argue that E₂-induced transcription and translation could contribute to an effect in this time frame. Therefore, we next tested a shorter E₂ exposure time during the 30-min uptake assay (Fig. 2C) to more clearly link this inhibitory function of E₂ to a nongenomic mechanism of action. Again, treatment with NGF increased the DAT-specific uptake of DA, suggesting that new NGF-induced DAT is functional. When 10 nM E₂ was added for only the last 5 min of the uptake assay, it dramatically inhibited DAT activity, more effectively than during the 30-min assay; uptake was completely reversed with only a 5-min hormone exposure. The NGF-inducible portion of the transport was also completely blocked by E₂ treatment. Such effects not only show rapid and efficient E₂ inhibition, but also suggest that the hormone may reverse the direction of the transporter, causing DA taken up in the previous 25 min to be removed from the cell.

Next we wished to examine which specific subtypes of ERs may be present in PC12 cells and could be responsible for these rapid effects on DA transport. The presence of naturally expressed ERs α and β in PC12 cells and their long-term (weeks) upregulation by NGF have been previously reported 31324043, but not the presence of these proteins in the membrane. Our single protein band for ERα and the doublet bands seen for ERβ (Figs. 3a and 4a, Western blots) are similar to that seem by others 31 and confirm that ERα and ERβ are expressed in PC12 cells. The results in Fig. 3A also show that ERα can be elevated by a 2 day (or longer) NGF treatment. We were particularly interested in demonstrating membrane versions of ERs, as those are the most likely to participate in rapid nongenomic responses. Because these immunoblots contained ER protein from whole cell extracts, we next examined the subcellular location of these receptor proteins. As expected, ERα was shown to be in the nucleus of fixed, permeabilized cells (data not shown). ERα was heterogeneously present on membranes of fixed, nonpermeabilized cells (Fig. 3B), and appeared on both the cell body and processes, arranged in irregularly spaced punctate clusters. This is similar in appearance to the mERα staining that we have observed previously 444546. We next applied these nonpermeabilizing immunocytochemistry conditions to developing a quantitative plate assay that we have used previously to demonstrate mERα on other cells 3547. Fig. 3C shows saturability of the membrane antigen with increasing Ab concentrations, and comparisons of the low levels of membrane receptors to the levels of nuclear ERα measured with the same technique in cells permeabilized with detergent. The bars show the values for these proteins when cells are permeabilized, and the symbols within each bar show the same values in nonpermeabilized cells. As expected, the membrane population of these proteins is much smaller in each case than the whole-cell value in permeabilized cells. Negative controls (including no 1° Ab to detect any nonspecific binding of the 2° Ab, and no 1° or 2° Abs to detect any endogenous alkaline phosphatase contributing to the colorimetric signal) gave very low values. The clathrin signal in unpermeabilized cells was very low, and the permeabilized cell value was quite high, as expected for a protein residing just inside the plasma membrane.

The other classical ER family member, ERβ was then similarly investigated. In contrast to ERα, NGF treatment spanning 7 days did not affect the levels of this receptor protein (Fig. 4A). In Fig. 4B mERβ was assessed by immunocytochemistry on NGF-differentiated cells prepared with a nonpermeabilizing fixation technique. The appearance of mERβ was very similar to that of mERα – heterogeneous membrane staining among cells, punctate, and unevenly distributed over both cell bodies and processes. Again, measurement of this receptor was amenable to a quantitative plate assay (Fig. 4C) which showed a saturable receptor protein antigen in the membrane, with somewhat higher levels present in the whole (permeabilized) cells; low negative control values and high positive control values were similar to those shown for the ERα plate assay.

Finally, we examined RNA from these cells to determine if the rat GPR30 RNA was expressed; the lack of a specific Ab for rat GPR30 necessitated this approach. An RT-PCR amplimer of the anticipated size can be produced from PC12 cell RNA (Fig. 5) using two different sets of primers. This result matches those for the positive control cell line, MCF-7 human breast cancer cells. No signal was evident when RNA samples were omitted, showing that contaminated reagents did not produce this signal. Though this RT-PCR detection is not a quantitative method, the levels of GPR30 RNA in PC12 cells appear to be similar to that in MCF-7 cells.

In previous studies we noted that mERα and the downstream effects it mediates in pituitary (GH3/B6/F10) cells were profoundly influenced by the density at which cells were grown; with increasing cell density, the expression of mERα (in favor of intracellular ERα) declined dramatically. Cells grown at higher densities (though not unusual densities for most cell culture experiments) also became unresponsive to E₂ for nongenomic actions 46. Therefore, we examined the effect of cells grown at various densities on the 5 min, 10 nM E₂-induced inhibition of DA uptake in PC12 cells (Fig. 6A). Increasing the density from 10,000 to 15,000 cells per well of a 48-well plate increased the measurable DA uptake; E₂ was still completely effective in inhibiting this higher level of uptake. However, when cells were further crowded to 20,000 cells per well, the inhibitory effect of E₂ was lost, and instead an estrogenic stimulatory effect on DA transport was observed.

Because we have repeatedly observed non-conventional dose-response relationships for these nongenomic estrogenic responses, we always test our responses over a very wide (fM to nM) concentration range. Fig. 6B shows such an analysis for E₂'s effects on the DA uptake response. As we have observed for rapid responses in other tissues 29414849, there is more than one peak of inhibitory activity for E₂ in PC12 cells, separated by concentrations at which E₂ is less effective (10^-10 M).

To determine the rapidity and stability of this estrogenic effect on DA transport, we examined a time course of 10 nM E₂ exposure, concentrating on the more rapid (<30 min) increments expected to operate via the nongenomic mechanism. As we have seen many times in the past for nongenomic responses 4147, rapidly oscillating temporal phases of this response were evident. Because these effects changed rapidly over time (from inhibition to enhancement of uptake) at 37°C, the results had somewhat large error ranges, as expected for such a fluctuating system over short time intervals when samples have to be removed from the incubator for analysis. Therefore, we reduced the temperature of the assay (to ambient) to hopefully create a more stable experimental demonstration of this oscillation. Under these conditions (Fig. 6C), peak inhibitory effects were seen at the 9 min time point. A small transport enhancement occurred at 3 min and a more robust one at 20–30 min.

Because DAT-mediated uptake was inhibited by E₂, we next examined whether this could be due to E₂-induced DAT trafficking or reduced protein levels. We adapted our quantitative plate assay for DAT to measure cell surface vs. intracellular DAT, by permeabilizing the cells for the latter measurement with detergent during the cell fixation step. An Ab to the second extracellular loop of DAT recognizes transporter which is either on the plasma membrane or inside the cell, but not vesicle-bound (in which case the 2^nd loop would be facing the inaccessible vesicle interior). Fig. 7 shows that the amount of DAT was decreased in both of these cellular compartments. Next we looked at the whole-cell levels of DAT protein, to determine if E₂ might have an effect on rapid DAT protein stability. Such rapid turnover sometimes occurs for phosphoproteins that after activation are subsequently ubiquitinated and sent to proteosomes or lysosomes. Fig. 8 shows that E₂ treatment had no effect on DAT protein levels, compared to vehicle-treated cells, across a time course of 5 min to 1 hr.

We propagate our PC12 cells in 15% serum-containing medium, releasing them for subculture by repeatedly pipetting the cells. However, before each experiment, cells were transferred to a defined medium for 48 hrs to insure withdrawal from the effects of estrogens and other hormones and growth factors present in serum. Our defined medium is high-glucose, phenol red-free RPMI 1640 (Gibco/Invitrogen, Grand Island, NY), substituting the 15% serum with TCM serum replacement (Celox Laboratories, St. Paul, MN). During this serum starvation period NGFβ (a gift from Dr. Regino Perez-Polo) was added to (at 20 ng/ml) to designated cultures for these studies. We study "native" cell regulation using NGF stimulation to achieve levels of DAT in our cells that can be easily assayed for the inhibitory effects of E₂, instead of transfecting cells with constructs that will over-express DAT. In this way our studies differ from many previously published DAT regulation studies. Under these conditions, regulatory interactions may be subject to fewer artifacts due to over-expression, such as normal receptor level interactions with other signaling molecules.

PC12 cells plated on poly-D-lysine (10 μg/ml)-coated coverslips were cultured in 6-well plates at densities of 20,000–40,000 cells per well for 24 hours. After serum starvation ± NGF treatment, cells were fixed using 2% paraformaldehyde (Fisher Scientific, Houston, TX), and 0.2% gluteraldehyde (Electron Microscopy, Fort Washington, PA) for 30 minutes. Where applicable, cells were permeabilized during fixation by adding NP-40 (or IGEPAL CA-630 can be substituted) and sucrose 35. Unquenched aldehydes were reduced by applying a 13 mM NaBH₄, 70 mM NaHPO₄ aqueous solution for 15 minutes at room temperature. For additional reduction of autofluorescence background, Schiff's Reagent (Fisher) was applied for 15 minutes while on ice, and then rinsed off three times with sulfurous water (equal volumes of 1 N-HCl and 10% sodium metabisulfite aqueous solution). The wells were washed (for 10 minutes) with a second reducing solution (1%Na₂HPO₄/1%NaBH₄ in ddH₂O) at room temperature, with light shaking. This was followed by a 15 minute PBS wash. Cells were then blocked for 45 minutes at RT with 0.1% coldwater fish gelatin (Sigma, St. Louis MO) in PBS. The cells were incubated overnight at 4°C (with light orbital shaking) with primary Ab (diluted in fish gelatin/PBS); the 1° Ab we used for detection of ERα, at concentrations of 2–10 μg/ml was C542 (StressGen, Inc., Collegeville, PA). Anti-ERβ monoclonal Ab (clone 9.88, Sigma E1276) was used at dilutions ranging from 1:500–1:2000. Ab DAT/e2 to the DA transporter binds to the second extracellular domain 42, and was obtained from A. Levey (Emory Univ.). It was used at concentrations ranging from 0.025–10 μg/ml. Anti-clathrin Ab (ICN Biochemicals, Inc., Aurora, OH) was used in concentrations of 1–3 μg/ml as a control for cell permeabilization. We used mouse IgG1κ (Sigma) as an irrelevant Ab and isotype control. Secondary Ab (either biotinylated mouse IgG or IgM, from Vector Laboratories, Burlingame, CA) was diluted (50 μl/10 ml) in 0.1% fish gelatin/PBS and incubated with samples for 1 hour at RT. Using a kit from Vector Laboratories, ABC-AP solution (diluted in PBS) was added, followed by a series of PBS washes, and then Vector Red alkaline phosphatase substrate (Vector Laboratories) was added for 2–5 minutes. The alkaline phosphatase substrate was prepared by adding 2 drops of each reagent from the kit to 5 ml of 100 mM Tris-HCl (pH 8.2–8.5) and adding the endogenous phosphatase inhibitor levamisole (Vector Labs) to a final concentration of 0.5 mM. The solution was removed and the wells immediately rinsed with ddH₂O; a small volume was left in each well until the coverslip was ready to go through the following cycle: dehydration with two changes of 70% ethanol (30 seconds each), followed by two changes of 100% ethanol (30 seconds each). The coverslips were then cleared with three xylene washes (3 minutes each) and then mounted with Cytoseal 280 (Electron Microscopy). Images were viewed under an FITC filter with a Leitz fluorescence microscope equipped with a CoolSNAP-Pro digital camera from Media Cybernetics using Image-Pro Plus software. The cell images were viewed using a wide-spectrum FITC filter. Under these conditions the intense red staining of Vector Red is observed as orange red staining; the green staining is aldehyde-induced autofluorescence background. We used these photographs with included autofluorescence background to give a simultaneous cell outline upon which the Ab-mediated Vector Red signal is visible. Specific staining was also visible using a rhodamine filter, but without the green background (not shown).

PC12 cells were grown to 50% confluence in 100 mm Petri dishes and serum-starved ± NGF for 2–7 days. Cells were then rinsed twice with ice-cold PBS and solubilized in 0.5 ml of lysis buffer (20 mM Tris; 150 mM NaCl; 1 mM EDTA; 1 mM EGTA; 1% Triton X-100; 2.5 mM sodium pyrophosphate; 1 mM β-glycerolphosphate; 1 mM Na₃VO₄; 1 μg/ml leupeptin; 1 mM PMSF) at 4°C. After sonication (4 × s, 5 sec each), the insoluble materials were removed by centrifugation at 15000 × g for 10 min. The extract was treated with SDS sample buffer and boiled for 5 min. Aliquots were assayed for protein concentration (BioRad), and 20 μg/ml total protein was subjected to SDS-PAGE in 10% acrylamide, and then transferred to a nitrocellulose membrane. Blots were blocked (2% nonfat dry milk, 1% BSA in 10 mM Tris-buffered saline pH 7.4) for 1 hour, followed by overnight incubation with primary Ab for ERα (1 μg/ml Stressgen, SRA 1010), ERβ (1:1000, Sigma E1276), or DAT (DAT/e2, 1:500 from A. Levey) at 4°C. Blots were then rinsed and incubated with peroxidase-conjugated anti-mouse IgG (1:4000, Southern Biotech) for ERα and ERβ, and peroxidase-conjugated anti-rat (1:4000, Southern Biotech) for DAT at RT for 2 hours. Immunoreactivity was detected on X-ray film (Amersham) by enhanced chemiluminescence.

We originally developed sensitive and specific quantitative fixed cell sandwich immunoassays suitable for 96-well plates for demonstrating mERα on the cell surface of both GH3 cells and MCF-7 breast cancer cells 3547. For the present studies we further adapted the assay for use with PC12 cells, to measure DAT, ERa and ERβ. Cells were plated in serum-containing medium, fed with serum-free defined medium for 48 hrs, then fixed with 2% paraformaldehyde/0.1% glutaraldehyde. Cells were then treated with the following reagents, with washing steps between each application: NaBH₄ for free aldehyde reduction; fish gelatin for blocking; 1° Ab; biotinylated 2° Ab; avidin-conjugated alkaline phosphatase; and levamisole (to block the endogenous mammalian subtype of alkaline phosphatase). Then pNpp, a substrate for alkaline phosphatase, was added, producing a soluble yellow dephosphorylated product (pNp) measured at A_{405 nm}. Finally, the pNp reagents were washed from the wells, and the cells stained with crystal violet (CV); after extraction, this dye was read at A_{590 nm} as a measure of cell number, to which our antigen values were normalized.

This assay was adapted to measure the same antigens in the intracellular compartment by simply including membrane-permeabilizing detergents in the fixation step, so that levels of intracellular (permeabilized) and extracellular (unpermeabilized) antigens can be compared by essentially the same methods 35. We thus assayed membrane antigens for intracellular relocation (trafficking). Ab for the abundant intracellular antigen clathrin was used to determine the permeabilization status of the cells in each assay.

Cultured PC12 cells (1–4 × 10⁴/well in poly-D-lysine-coated 48-well plates) were rinsed and preincubated in buffer for 30 min under assay conditions at 37°C. Inhibitors were included in this preincubation step. Transport assays were initiated by adding buffer containing ³H-DA (Dupont NEN, 50 nM for single point assays), monoamineoxidase inhibitors pargyline or selegeline 78, ascorbic acid (for metabolic stability of E₂ and other compounds), and 50 nM DMI (to inhibit any contribution from the NET). The specific DAT inhibitors nomifensine (500 nM) or GBR-12909 (100 nM) were included in parallel samples; the differences between responses ± these inhibitors were used to define uptake due to DAT. In some cases E₂ was added concomitantly with the labeled DA; in other cases it was added after the labeled DA incubation had continued for number of minutes; thus E₂ was added only for the last min of the uptake assay. Assays were terminated by rapidly washing the wells 3 × with ice-cold buffer. Cells were then solubilized in water for 15 min at room temperature with shaking, or by a freeze-thaw cycle, and an aliquot assessed for ³H via liquid scintillation counting. In some cases, another aliquot was assayed for protein content with a Bio-Rad Bradford assay; this value determined wells where cells had become disengaged from the well bottom, for data exclusion. Uptake assays to determine the temporal changes in regulation were done at 15,000 cells per well and at room temperature, slowing the reaction and allowing for less error prone measurements.

RNA was prepared from PC12 cell lysate using the RNAqueous kit (Ambion). First-strand cDNA synthesis was performed using the SuperScript III First Strand Synthesis System for RT-PCR (Invitrogen). Briefly, 2 μg RNA, 50 μM oligo(dT), 10 mM dNTP mix, and DEPC-treated water up to 10 μl final volume was incubated at 65°C for 5 minutes and placed on ice for 1 minute. The samples were then incubated for 50 minutes at 50°C with the addition of 10 μl of cDNA synthesis mix containing RT buffer, 25 mM MgCl₂, 0.1 M DTT, RNaseOUT, and Superscript III RT (200 U/μl). The reaction was terminated by heating the samples to 85°C for 5 minutes. The PCR reaction was performed in 25 μl GoTaq Flexi DNA Polymerase buffer (Promega) containing 0.2 μM of both gene specific sense and antisense primers, plus 0.5 μl of the RT reaction. The primers (kind gift of Dr. Peter Thomas 39) were designed using GenBank sequence accession no. BC011634. Primer set 1 was: sense, 5'-GGC TTT GTG GGC AAC ATC-3'; antisense, 5'-CGG AAA GAC TGC TTG CAG G-3'. Primer set 2 was: sense, 5'-GCA GCG TCT TCT TCC TCA CC-3'; antisense, 5'-ACA GCC TGA GCT TGT CCC TG-3'. The PCR product was obtained using the GeneAmp PCR system 9700 with 35 cycles of 30 sec at 94°C, 30 sec at 55°C, and 2 min at 72°C followed by a 10-min extension at 72°C. The PCR products were electrophoresed in a 1% agarose gel containing 0.5% ethidium bromide, and bands corresponding to the anticipated products of 680 bp (primer set 1) and 585 bp (primer set 2) were identified.

A one way ANOVA (SigmaStat 3.0) was used to determine the significance of treatment effects compared to vehicle controls. Statistical significance was accepted at the p < 0.05 level.