Establishment of cell lines expressing wild type (WT) and mutated G_sα

The coding regions from wild-type and mutated G_sα were subcloned into a pcDNA3.1/myc-His expression vector (Invitrogen, Carlsbad, CA), as previously described, and the sequence confirmed (Figure S1) [15]. Cell sensor CRE-bla CHO-K1 cells (Invitrogen, Carlsbad, CA) were grown in Dulbecco's modified Eagle's medium (DMEM; Life Technologies, Carlsbad, CA) containing 10% fetal bovine serum (HyClone, Logan, UT), Penicillin (100 U/ml)-Streptomycin (100 µg/ml), L-glutamine, 25 mM HEPES, pH 7.3, 1 mM non-essential amino acids (NEAA), and 1 mM sodium pyruvate (Invitrogen, Carlsbad, CA). Cells were seeded (1.5×10⁶/100 mm plate) and grown to 80% confluence prior to transfection. Cells were transfected in serum-free condition by using 2 or 10 µg each of expression plasmids and Lipofectamine 2000 (Invitrogen, Carlsbad, CA) and incubating at 37°C for 3–5 hours. Transfection was monitored by microscopy and the presence of green fluorescence on a separate plate that was transfected with YFP-N1 (Clontech, Mountain View, CA). After 24 hours of transfection, the medium was changed and cells were grown in full culture medium containing 1000 µg/ml gentamycin (Invitrogen, Carlsbad, CA) for selection purposes. Cells were grown in selection medium for 2–3 weeks until cell colonies formed. Colonies were selected using Cloning Cylinders (Millipore, Bedford, MA) and propagated separately. Several clones were propagated for use in further screening steps. Four separate clones from individual stable clones were used for cAMP measurement using a cAMP ELISA kit (Tropix cAMP-Screen direct cyclic AMP ELISA system, Applied Biosystems, Foster City, CA).

Confirmation of equivalent expression of transfected genes

Cell extracts were prepared from each cell line using whole cell extract buffer (10% glycerol, 250 mM NaCl, 10 mM HEPES pH 7.5, 1 mM EDTA, 0.1% NP-40, 1 mM DTT and 1× protease inhibitor cocktail) (CalBiochem EMD Biosciences, La Jolla, CA). Protein concentrations were measured using BCA Protein Assay Kit (Thermo Scientific, Rockford, IL). Equal amounts of protein extracts were loaded on 4–20% polyacrylamide gel (Bio-Rad, Hercules, CA) and were then transferred onto nitrocellulose paper for immunodetection purposes using Anti-c-Myc antibody (Invitrogen, CA). Loading for each lane was normalized by using mouse anti-human β-actin antibody (Sigma, St. Louis, MO) (Figure S2).

Demonstration of differences in cAMP levels

cAMP levels were determined in stable clones by ELISA (Tropix cAMP-Screen direct cyclic AMP ELISA system, Applied Biosystems, Foster City, CA). Stable cell lines (1×10⁴ cells/96-well) were grown in full medium overnight and incubated for 30 minutes in fresh serum-free medium containing 10–50 µM of 3-isobutyl-1-methylxanthine (IBMX) (Sigma, St. Louis, MO), a known phosphodiesterase inhibitor. Cell and medium were collected and the amount of cAMP was measured following the manufacturer's instructions. One cell line from each set (WT9 cells expressing the wild-type G_sα, C6 for cells expressing the R201C G_sα and H2 for cells expressing the R201H G_sα) was selected for further studies (Figure S3).

Results from light microscopy also showed that the C6 and H2 cell lines demonstrate a more fibroblastic/stellate appearance relative to WT9 cells, consistent with morphologic changes typically induced by higher cAMP levels, as has been previously demonstrated (Figure S4) [16].

C6 cells carrying the R201C mutation exhibited very high basal cAMP levels compared to WT9 cell line, and were therefore used for the subsequent experiments. The assay was optimized for cell density and phosphodiesterase inhibitor. Cell densities of 1000–3000 cells/well did not show significant differences in basal cAMP level (Figure S5A). There were significant differences in basal cAMP levels with the addition of Ro-20-1724 (Figure S5B), though there were no major differences in the cAMP levels from 30–120 minutes (data not shown). Therefore, a 2000 cells/well density, a 30 min incubation period, and the inclusion of Ro-20-1724 were chosen as optimized conditions for the assay.

Demonstration of system inhibition and stimulation

To demonstrate that the system would be able to detect molecules that would inhibit and stimulate cAMP, adenylyl cyclase inhibitors and stimulators were tested. Cells were grown in either 96-well (5–10×10³ cells/well) overnight in full medium. The medium was then changed to core DMEM containing 50 µM of the phosphodiesterase inhibitor 4-(3-Butoxy-4-methoxybenzoyl)-2-imidazolidine (Ro-20-1724) (Sigma, St. Louis, MO) for 1–2 hours. The medium was then removed from each well and treatments were started using the same medium containing Ro-20-1724 for 30 minutes or other specified times at 37°C. cAMP levels were measured in the cells after treatment with four different adenylyl cyclase inhibitors [2′,5′-dideoxyadenosine (ddA)], (E)-2-(1H-Benzo[d]imidazol-2-ylthio)-N′-(5-bromo-2-hydroxybenzylidene) propanehydrazide (KH7), MDL-12,330A (MDL) and SQ 22,536 (SQ)], and the adenylyl cyclase activator, forskolin (compounds were purchased from Sigma, St. Louis, MO). Different concentrations of inhibitors were tested on C6 (expressing R201C) and H2 (expressing R201H) cells for 5, 15 and 30 minutes in 96-well formats. All the treatments were performed at 37°C for 5–60 minutes. A fluorescence resonance energy transfer (FRET)-based cAMP assay kit (Cisbio, Bedford, MA) was used to measure cAMP (Figure S6A–D). 1 µM ddA or SQ could inhibit the basal levels of cAMP by 60–75% after 5 minutes of incubation. The degree of inhibition was variable in the cells treated with KH7 or MDL. The less robust and more variable effect of KH7 and MDL may be due to the fact that KH7 may be more of an inhibitor of soluble rather than membrane-bound adenylyl cyclase and that MDL has recently been shown to have activity on pathways unrelated to adenylyl cyclase [17]. There was an approximately 10-fold increase in basal cAMP levels in C6 cells after forskolin treatment for 30 minutes (Figure S6E). These results were similar in the H2 cells (data not shown). These data indicated that the cAMP levels in cells expressing mutant G_sα could be inhibited and stimulated, and therefore appropriate for HTS screening.

High throughput screen (HTS) assay

Cells were plated (2,000 cells/well) in 6 µl of complete medium (DMEM with 10% fetal bovine serum and 1% penicillin/streptomycin, 0.5 mg/ml G418) 1536-well plates and grown overnight at 37°C 5% CO₂ and 95% humidity. Certain cells, both in vivo (proximal renal tubule cells) and cell lines have the intrinsic capacity to export cAMP. This transport is typically probenecid-responsive, in which probenecid inhibits cAMP transport. Export of cAMP to the media is a property of the CHO cell line used in this assay (Figure S7). Therefore, to maximize the ability to detect inhibitor-induced changes in cAMP accumulation in the conditioned medium that occurred during the overnight incubation period, a wash protocol was included (Table 1). Cells were seeded (2,000 cells/well) in 6 µl DMEM. Following the overnight incubation, the growth media was aspirated leaving 1 µl residual volume. The wells were washed with 5 µl DPBS containing 1 mM CaCl₂, 0.5 mM MgCl₂, 0.05% BSA, 0.005% Tween 20, then aspirated leaving 2 µl residual volume. To each well 23 nl/well of test compounds in DMSO were dispensed via pin transfer before 1 µl medium containing 300 µM Ro 20-1724, a known phosphodiesterase inhibitor, was added. The cells were incubated with compounds for 30 minutes at 37°C. HRTF assays were performed using the homogenous time-resolved fluorescence (HTRF) HiRange cAMP detection kit (Cisbio, Bedford, MA) by adding 1.0 µl/well cAMP-conjugated d2 and 1.0 µl/well europium cryptate conjugated anti-cAMP antibody in lysis buffer according to the manufacturer's instructions. Plates were incubated at room temperature for 30 minutes and FRET signals (665 and 615 nM) were read using an EnVision plate reader (PerkinElmer, Waltham, MA). HTRF signal was calculated as the ratio of signal from the 665 nm (acceptor) and 615 nm (donor) channels and multiplied by 10,000. % activity was calculated by normalizing each the HTRF signal from each sample well to the mean HTRF signal from the DMSO only control wells.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Assay Protocol.

https://doi.org/10.1371/journal.pone.0090766.t001

Screening of Library of Pharmacologically Active Compounds (LOPAC)

To test feasibility of proceeding to a full screening assay, the LOPAC library (1280 compounds, Sigma-Aldrich, St. Louis) was screened using a quantitative qHTS assay format as described previously (Table S1) [18]. Prior to that screen, a 1536-well DMSO test plate (solvent control), was tested for signal-to-basal (S/B) ratio and Z′ factor calculations (Figure S8). The S/B and Z-factor were calculated from 16 control wells each with CHO-WT and CHO-C6 HTRF values. For the LOPAC screen, each compound was titrated in 5 concentrations. The final concentrations of the compounds assayed ranged from 92 nM to 57.5 µM. The mean S/B was 1.6 between DMSO and forskolin treatments, and 1.6 between DMSO and ddA treatments. The mean Z-factor was 0.24 between DMSO and forskolin treatments and 0.16 between DMSO and ddA treatments. Considering the fact that the assay requires two wash steps to remove secreted cAMP in the conditioned media, and that the assay is tuned to detect both increases and decreases in cAMP levels, the S/B and Z′ factors, while low, indicate that the assay is capable of detecting both an increase and decrease in cAMP levels. The screen resulted in the identification of several inhibitors and activators, and indicated that this assay was suitable for HTS (Tables S2, S3, S4A&B, Figures S8, S9A,B,&C).

Curve response class classification from dose response HTS

Curve response classifications (CRCs) are the measure that includes potency, efficacy and reliability of the data, and estimates an IC₅₀ value directly from the primary screen [18]. To determine CRCs the plate raw data were loaded into the NCATS quantitative high-throughput screening (qHTS) database and normalized to the DMSO and forskolin control wells. The data were then used to fit 4-parameter dose-response curves, a custom grid-based algorithm, to generate curve response class (CRCs) values for each compound [19]. The resultant curves were then classified using a heuristic curve classification scheme, allowing for the distinction of high quality curves (class 1.1) from lower (2.1, 1.2, 2.2) to poor quality ones (3, 4). Briefly, a curve (and hence a compound) was classified as 1.1 if it exhibited well defined upper and lower asymptotes, with a good fit to the observed data points (R²> = 0.9) and an efficacy greater than 80%. A class 2.1 curve was similar to a 1.1 curve, but exhibited only one well-defined asymptote. A curve that exhibited poorer efficacy (between 30% and 80%) was classified as a 1.2 or a 2.2 if it had two asymptotes or one asymptote, respectively. A class 3 curve was one that was poorly fit or only exhibited activity at the highest concentration, thus representing inconclusive activity, and a class 4 was assigned to those cases where there was no dose response, and considered inactive.

Hit selection criteria

The following criteria were applied for hit selection from the primary screen: 1) hits in robust curve classes 1.1, 1.2, and 2.1 and active compounds in other curve classes (1.3/1.4/2.3/2.4/3) with maximum response (efficacy) >60% were considered active; 2) these hits were filtered for donor interference, and those compounds demonstrating donor interference were eliminated; 3) compounds were further filtered by reactive and promiscuous functional groups, as previously described [20]; 4) to group hits by structural similarity, clustering was performed using Leadscope Hosted Client (Leadscope Inc., Columbus, OH).

Molecular Libraries Small Molecule Repository library screening

For the initial screen of the 343,768 compound MLSMR library, C6 cells, following HTRF cAMP assay protocol were screened at a single dose of compounds (38 µM). The screening assay was conducted according to the protocol outlined above (High throughput screen (HTS) assay, Table 1). A total of 1,375,072 wells were screened. The signal cut-off was set at >30% change in HTRF signal from basal activity. The mean S/B was calculated to be 1.66±0.30 and Z′-factor was 0.27±0.23.

Forskolin-induced cAMP Assay in WT9 Cells

The WT9 clonal cell was selected based on its low basal level of cAMP, and robust stimulation with forskolin to induce detectable levels of cAMP using the HTRF assay. The cells were grown in DMEM, 10% FBS, 1% Pen/Strep, 0.5 mg/ml G418. The day before screening, 2000 cells/well in 3 µl in DMEM 10% FBS, 1% Pen/Strep, 0.5 mg/ml G418 were seeded in Greiner One high base solid bottom white tissue culture treated plates using a small cassette and a Multidrop (from Thermo Fisher). The plates were allowed to incubate 16–24 hr at 37°C, 5% CO2, 95% humidity covered with low evaporation stainless steel lids from Kalypsys. Prior to compound addition, 1 µl of 500 µM (final 100 µM) of the PDE inhibitor Ro-20174 solution in complete DMEM was dispensed to prevent degradation of cAMP. 23 nl of compound dose response solutions in DMSO were then dispensed using a Kalypsys pintool (diluted into 5 µL resulting in a 1∶217 dilution of compound). The control compound included the cAMP stimulator forskolin used at 4.6 µM final. Next, 1 µl of a 200 nM (final 40 nM) of forskolin solution in complete DMEM was added to induce an EC₈₀ cAMP response. The plates were incubated for 30 minutes at 37°C, 5% CO2, 95% humidity using the same stainless steel lids. Finally, the high range cAMP HTRF kit (CisBio, Bedford, MA) was used to detect the levels of cAMP. A total 1 µl/well of the HTRF reagent cAMP-d2 in lysis buffer and 1 µl/well HTRF reagent anti-cAMP antibody-K in lysis buffer were dispensed at the same time using the FRD. The plates were incubated for 30 minutes at room temperature and then FRET signal was measured with an EnVision plate reader using an HTRF protocol (Excitation at 320 nM and Emission at 665 nM for the cAMP-d2 and 615 nM for the anti-cAMP antibody-K). The ratio of 615/615 nM is calculated to normalize for any effects in the donor only channel (Figure S10).

HTS Viability Assays

For each cell line tested, a total of 500 cells per well in 5 µL of media was dispensed using a Multidrop Combi dispenser (Thermo Fisher Scientific Inc., Waltham, MA) and a small cassette into barcoded 1536 solid bottom white Greiner One tissue culture treated plates (catalog # 789173-F). The plates were then covered with stainless steel cell culture Kalypsys lids and incubated at 37°C with 5% CO₂ under 95% humidity to allow the cells to adhere. Standard DMEM -1640 supplemented with 10% FBS, 1× penicillin/streptomycin/amphotericin, 2 mM glutamine and 0.5 mg/mL G418 was used (Gibco). For the generation of the standard 11 point dose response curves the library compounds and control compound forskolin (43 µM final) was added by the pintool addition (Kalypsys) of 23 nL solubilized in DMSO. The cells were incubated for 48 hours and then 3 µL of CellTiter Glo luminescent cell viability assay reagent (Promega, Madison, WI) was added using a Bioraptor Flying Reagent Dispenser (Aurora Discovery-BD). The plates were then incubated for 15 minutes at room temperature. The signal was captured using a 10 second exposure with a ViewLux (Perkin Elmer) contacting a luminescent filter. Relative luminescence units (RLU) for each well were normalized to the median RLUs from the DMSO control wells as 100% viability.