Uveal melanoma is the most common intraocular tumor in adults and often metastasizes to the liver, leaving patients with few options. Recurrent activating mutations in the G proteins, Gαq and Gα11, are observed in approximately 93% of all uveal melanomas. Although therapeutic intervention of downstream Gαq/11 targets has been unsuccessful in treating uveal melanoma, we have found that the Gαq/11 inhibitor, FR900359 (FR), effectively inhibits oncogenic Gαq/11 signaling in uveal melanoma cells expressing either mutant Gαq or Gα11. Inhibition of oncogenic Gαq/11 by FR results in cell-cycle arrest and induction of apoptosis. Furthermore, colony formation is prevented by FR treatment of uveal melanoma cells in 3D-cell culture, providing promise for future in vivo studies. This suggests direct inhibition of activating Gαq/11 mutants may be a potential means of treating uveal melanoma.
Oncogenic Gαq/11 inhibition by FR900359 may be a potential treatment option for those with uveal melanoma.
Uveal melanoma (UM) arises from melanocytes contained in the choroid, ciliary body, and iris (together known as the uvea) of the ocular cavity, and is the most common intraocular tumor in adults (1, 2). UM accounts for approximately 5% of all lethal melanoma cases, and the current treatment options for eye localized disease are either surgery or radiation. Metastases occur in approximately 50% of patients with UM, predominantly to the liver. Patients with macrometastases have an average survival of 2 to 8 months as there are no effective therapies (1). The mutations in UM are distinct from those commonly found in cutaneous melanoma, such as the driver mutations in BRAF and NRAS (3–6). UM predominately involves activating mutations in either GNAQ or GNA11 genes that encode the highly conserved Gαq and Gα11 subunits of heterotrimeric G proteins (7, 8). Point mutations occur at the Q209 and R183 residues in the GTPase domain of Gαq/11 proteins. These residues are critical for the intrinsic GTP hydrolysis activity of the G proteins, which are rendered constitutively active in the GTP-bound state leading to aberrant downstream signaling (7–11). Gαq/11-Q209 mutants are much more prevalent in UM than Gαq/11-R183 mutants, and are unresponsive to GTP hydrolysis stimulation by regulator of G protein signaling (RGS) proteins (12). This results in a more significant increase in signaling and a more severe phenotype (8, 9).
The introduction of the Gαq- or Gα11-Q209L mutant into human or mouse melanocytes results in anchorage-independent growth and gives rise to heavily pigmented tumors in mice (8, 13, 14). Because of the redundancy of their signaling pathways, oncogenic mutations in either Gαq or Gα11 appear to cause similar cellular oncogenic properties leading to the pathogenesis of UM. For example, GNAQ and GNA11 mutations are responsible for the upregulation of the MAPK pathway in the absence of BRAF mutations in UM (15–18). Knockdown of Gαq in cell lines derived from primary or metastatic UM results in decreased MAPK signaling (7). However, inhibiting the ERK1/2 pathway with MEK inhibitors has not been clinically effective in treating UM (19) and growth factors from the liver tumor microenvironment are able to mediate resistance (20, 21). Additional Gαq/11 stimulated signaling pathways that promote tumorigenesis have been implicated in UM. Oncogenic Gαq/11 leads to aberrant Akt signaling and increased activation of small GTPases RhoA and Rac1, which promote cell growth through JNK, p38, and yes-associated protein (YAP) directed transcription of growth-promoting genes (13, 14, 22, 23). YAP is a cotranscriptional regulator involved in the cell growth regulating Hippo pathway, which when dephosphorylated translocates from the cytoplasm into the nucleus where it associates with TEAD4 to promote the transcription of growth promoting genes (24–26). Recent studies indicate that YAP dephosphorylation in UM occurs through the Gαq/11 activation of a Trio, RhoA, Rac1 pathway (14). Considering the multiple avenues of disease progression and the multiple signaling pathways that are activated as a result of oncogenic Gαq/11, it has been difficult to determine a suitable and successful treatment for UM. This raises the possibility that direct inhibition of oncogenic Gαq/11 may be advantageous in preventing the constitutive activation of these multiple pathways.
Although there are no current drugs that directly target Gαq/11, a few compounds that effectively inhibit Gαq have been identified. YM-254890, a cyclic depsipeptide purified from Chromobacterium sp. QS3666, has been shown to potently and selectively inhibit wild-type Gαq/11 and the R183 mutant, but not the Q209L mutant protein, by inhibiting GDP dissociation from GDP-bound Gαq (27, 28). FR900359 (a.k.a. UBO-QIC), a highly related analog of YM-254890, was first isolated in 1988 from the leaves of the plant Ardisia crenata, and has since been shown to be a potent and selective inhibitor of Gαq/11 (29–33). Interestingly, recent studies revealed an effect of FR900359 (FR) on cell-cycle regulation in melanoma cell lines containing wild-type or oncogenic Gαq/11 (32). These suggest that FR might serve as a useful inhibitor in UM.
In this study, we present evidence that FR effectively inhibits Gαq-Q209L, Gαq-Q209P, and Gα11-Q209L mutant proteins resulting in decreased Gαq/11 signaling, cell-cycle arrest, and eventual cell death of UM cells harboring a GNAQ/11-Q209 mutation. These data provide evidence that direct inhibition of activating Gαq/11 mutants may be a potential strategy to treat UM.
Materials and Methods
Some studies were performed using FR purchased from the Pharmazeutische Biologie at the Universität Bonn. In addition, FR was also isolated from the roots of Ardisia crenata sims using the targeted counter-current chromatography method with minor modifications (34, 35). In brief, the dry powder of the roots of Ardisia crenata sims was first extracted with 95% ethanol, then the ethanol extract was partitioned by n-butanol–water. The upper n-butanol phase was separated, concentrated, and submitted to counter-current chromatography separation using an optimum two-phase solvent system of n-hexane–ethylacetate–ethanol–water (4.5:5.5:4.5:5.5, v/v), yielding more than 95% purity of FR. A positive ESI-MS spectra showed the characteristic molecular ions at m/z 1002.27 [M+H]+ and 1024.35 [M+Na]+ implying its molecular formula (C49H75N7O15).1H NMR and 13C NMR data were in close agreement with previous data (29). Similar results were obtained with both sources of FR.
Cell lines and culture
OMM1.3 cells were obtained from Dr. Bruce Ksander's lab and were confirmed to harbor the Q209P mutation in GNAQ by Sanger sequencing. 92.1 cells were obtained from Dr. Martine Jager and contained a GNAQ-Q209L mutation that was confirmed by Sanger sequencing, whereas OCM3 cells were obtained from Dr. Bruce Ksander and contained a BRAF-V600E mutation and wild-type GNAQ/11 as confirmed by Sanger sequencing. UM002B cells were derived from a brain metastasis and are PTEN negative and contain a heterozygous Q209L mutation in GNA11 and no change in GNAQ as determined by Sanger sequencing (Supplementary Fig. S1). 92.1, OMM1.3, UM002B, and OCM3 cells were cultured in RPMI1640 containing 10% FBS, 2 mmol/L l-glutamine, 0.2 units/mL penicillin, and 100 μg/mL streptomycin. Cells were cultured at 37°C and 5% CO2 in a humidified incubator. Cells were not tested for mycoplasma.
Wild-type rat Gαq was purified from High Five insect cells as previously described (36). Briefly, Gαq and glutathione-S-transferase (GST)-tagged Ric-8A were coexpressed, GST-Ric-8A:Gαq complexes were captured from the clarified lysates with glutathione Sepharose 4B, and Gαq was eluted. The Gαq was evaluated for activity using a Ric-8A-stimulated [35S]-GTPγS binding assay (37). Mutant GαqΔ34-Q209L was purified from High-Five insect cells. GαqΔ34-Q209L was engineered for high-level expression using a chimeric form of Gαq/i that contained an N-terminal hexahistidine tag followed by residues 1 to 28 of rat Gαi1, a tobacco etch virus protease site, and residues 35 to 359 of mouse Gαq-Q209L. Purification of GαqΔ34-Q209L followed protocols published previously (38).
GTPγS binding of purified Gαq and Gαq-Q209L was measured using a filter-binding method. Spontaneous GTPγS binding to the Gα proteins was promoted in the presence of (NH4)2SO4, as previously described (39). Briefly, purified Gαq (250 nmol/L) was pre-incubated with or without FR for 2 hours at 20°C in assay buffer A (50 mmol/L HEPES, pH 7.5, 1 mmol/L EDTA, 1 mmol/L dithiothreitol, 0.9 mmol/L MgSO4, 0.05% Lubrol). Reactions were started by the addition of 10 μmol/L [35S]GTPγS and 300 mmol/L (NH4)2SO4. Reactions were stopped by the addition of ice-cold wash buffer (20 mmol/L Tris-HCl, pH 7.5, 100 mmol/L NaCl, 2 mmol/L MgSO4), and the mixture was filtered through nitrocellulose membranes. The membranes were washed four times with ice-cold wash buffer, air-dried, and [35S]GTPγS binding was quantitated by liquid scintillation counting.
92.1, OMM1.3, UM002B, and OCM3 cells were plated on poly-l-lysine coated 96-well white bottom plates at 5 × 103 cells per well and allowed to sit overnight at 37°C and 5% CO2 in a humidified incubator. Cells were then treated with DMSO or 100 nmol/L, 300 nmol/L or 1 μmol/L FR for 1 to 4 days and harvested by removing the media and lysing the cells with 100 μL of cell-grade water for 1 hour at 37°C. One hundred microliters of 200-fold diluted Quant-iT PicoGreen (Invitrogen) reagent in TE buffer (10 mmol/L Tris-HCl, 1 mmol/L EDTA, pH 7.5) was added to the mixture and incubated for 1 hour at room temperature in the dark. DNA-bound PicoGreen fluorescence was detected using a Tecan Infinite F500 microplate reader.
Cell lysates were separated by SDS-PAGE and proteins were transferred to PVDF membrane. Immunoreactivity was detected using horseradish peroxidase-conjugated secondary antibodies (Vector) and chemiluminescence substrate (Thermo Scientific) on autoradiography film. HER3/ErbB3 (#4754), p-STAT3 (Tyr705, #4113), p-S6 (Ser235/236, #2211), p-S6 (Ser240/244, #2215), p-Rb (Ser807/811, #9308), and PLK1 (#4513) antibodies were all from Cell Signaling Technology. pCDK1 (Thr14/Tyr15, #44686G) antibody was from Invitrogen.
Cell lysates were separated by SDS-PAGE and proteins were transferred to PVDF membranes and incubated with ERK2 (Santa Cruz Biotechnology) and pERK1/2 (Cell Signaling Technology) antibodies followed by IRDye 800 CW and IRDye 680RD fluorescent secondary antibodies (LI-COR). Detection was with a LI-COR Odyssey and fluorescence was quantified using ImageStudioLite.
Cells were grown on coverslips in RPMI1640 containing 10% FBS. After 48 hours, cells were washed in serum-free RPMI1640, maintained overnight in serum-free RPMI1640, and then treated for 6 hours with vehicle (DMSO) or 0.1 to 100 nmol/L FR900359. Cells were fixed in 3.7% formaldehyde, and processed for immunofluorescence using an anti-YAP antibody (Santa Cruz Biotechnology; mouse monoclonal Ab 63.7, 1:100) followed by an anti-mouse antibody conjugated to Alexa Fluor 594 (Invitrogen, A11032; 1:250). Coverslips were mounted on glass slides using ProLong Diamond Antifade (Invitrogen; P36970). Subcellular localization of YAP in individual cells was scored as: C > N, indicating the majority of YAP staining in the cytoplasm; C = N, indicating that YAP staining is equally distributed in cytoplasm and nucleus; C < N, indicating that YAP staining is predominantly detected in the nucleus. Data are presented as the percentage of cells exhibiting the above criteria. One hundred cells were counted in each of three separate experiments. Images were recorded with an Olympus BX-61 microscope and 60× PlanApo objective with an ORCA-ER (Hamamatsu) cooled charge-coupled device camera controlled by Slidebook version 4.0 (Intelligent Imaging Innovation).
Reverse-phase protein array analysis
92.1 and OMM1.3 cells were plated in six-well dishes at 1.5 × 105 cells per well. Cells at ∼80% confluency were treated with DMSO or 1 μmol/L FR for 24 hours. Cells were lysed in reverse-phase protein array (RPPA) lysis buffer (1% Triton X-100, 50 mmol/L HEPES, pH 7.4, 150 mmol/L NaCl, 1.5 mmol/L MgCl2, 1 mmol/L EGTA, 100 mmol/L NaF, 10 mmol/L Na pyrophosphate, 1 mmol/L Na3VO4, 10% glycerol, freshly added protease and phosphatase inhibitors from Roche Applied Science; Catalog No. 05056489001 and 04906837001), and proteins were prepared and analyzed by RPPA, as previously described (40). RPPA data were used to determine proteins that are significantly different in both 92.1 and OMM1.3 cell lines treated with FR or DMSO. Comparisons were performed for each cell line by the two-sample t test method with 10,000 permutations and assumed unequal variance. Proteins with a Storey FDR < 0.05 and a log-2 ratio > 1 were considered significant. Supervised hierarchical clustering was performed based on median-centered log2-transformed expression values. Statistical calculations were performed in Matlab (v2017a) using the mattest and mafdr functions. RPPA was performed at MD Anderson where antibodies are extensively validated before being included in the panel. Antibodies must give one major band by western blotting that is within a dynamic range as determined by drug treatment or siRNA. Full details are provided at https://www.mdanderson.org/content/dam/mdanderson/documents/core-facilities/Functional%20Proteomics%20RPPA%20Core%20Facility/RPPA_Ab%20Validation%20Example.pdf. The RPPA results were validated by Western blotting for key targets.
EdU and propidium iodide incorporation assays
Cells were treated with DMSO or 1 μmol/L FR for 6 to 24 hours, and then incubated with 10 μmol/L EdU. Cells were prepared according to the Click-iT Plus EdU Alexa Fluor 647 Flow Cytometry Assay Kit (Invitrogen) protocol. Cells were incubated in 500 μL of Click-iT saponin-based permeabilization and wash reagent containing 40 μg/mL propidium iodide (PI) and 100 μg/mL RNaseA (Thermo Scientific). EdU and PI fluorescence was measured and analyzed using a FACS Calibur Flow Cytometer.
Caspase 3/7 activation assays
Cells were treated with DMSO or FR for 24 hours. Cells were prepared following the CellEvent Caspase-3/7 Green Flow Cytometry Assay Kit (Thermo Scientific) procedure. Cells were incubated with 500 nmol/L CellEvent Caspase-3/7 Green Detection Reagent for 30 minutes at 37°C, protected from light. Caspase-3/7 Green Detection Reagent fluorescence was measured and analyzed using a FACS Calibur Flow Cytometer.
3D cultures in Matrigel
The morphogenesis analysis was done using 3D cultures in Cultrex. Briefly, 5 × 103 OMM1.3 cells treated or not with FR were seeded on top of a Cultrex matrix (Trevigen) as described (41). Cells were cultured under these conditions for 7 to 12 days and the morphology of the colonies was monitored daily by phase contrast microscopy and by confocal microscopy at the end of the experiment to obtain optical sections to determine the effect of the treatment on cell survival and/or differentiation. The 3D cultures were processed as described (41) and mounted in Prolong gold with DAPI (Molecular Probes) to visualize the nuclei and activated caspase 3 and Melan-A to determine induction of apoptosis and melanocyte differentiation, respectively.
FR900359 inhibits GTPγS binding to Gαq and Gαq-Q209L proteins in vitro
FR has recently been reported to be capable of inhibiting wild-type Gαq, Gα11, and Gα14 and is thought to act as a guanosine nucleotide dissociation inhibitor (GDI) by inhibiting the GDP/GTP exchange of the Gα protein (30, 32, 33). Because the Q209L mutation is unlikely to alter the structure of Gαq/11, it is possible that FR might be able to bind to and inhibit Gαq/11-Q209L. Indeed, FR effectively blocked spontaneous [35S]GTPγS binding to both purified Gαq and Gαq-Q209L in a dose-dependent manner with an IC50 ∼75 nmol/L (Fig. 1). This provides evidence that FR can inhibit nucleotide binding to Gαq-Q209L, suggesting that FR can directly bind to mutant Gαq.
UM cell growth is inhibited upon treatment with FR900359
To determine if FR can inhibit the Gαq mutant protein in cells, we next treated UM cells containing GNAQ-Q209L/P or GNA11-Q209L mutations with varying concentrations of FR and tracked their growth over 4 days (Fig. 2A and B). The growth of primary 92.1 and metastatic OMM1.3 cells, containing GNAQ-Q209L and GNAQ-Q209P mutations, respectively, was effectively inhibited by treatment of FR at concentrations as low as 100 nmol/L. We also observed inhibited growth of metastatic UM002B cells, containing a GNA11-Q209L mutation, suggesting FR is also able to inhibit mutant Gα11 (Fig. 2C). In contrast, FR had no effect on the growth of OCM3 melanoma cells, containing a BRAF-V600E mutation (Fig. 2D). Thus, FR effectively inhibits the growth of UM cells containing activating mutations of Gαq or Gα11.
Oncogenic Gαq inhibition in UM cells abolishes MAPK signaling and prevents YAP localization to the nucleus
Increased MAPK and YAP signaling is observed in cells expressing oncogenic Gαq (13, 17, 42). To test if Gαq inhibition affects MAPK signaling in UM cells, we immunoblotted for pERK1/2 after treating 92.1, OMM1.3, and UM002B cells with FR (Fig. 3). ERK1/2 activation in 92.1 and OMM1.3 cells was completely abolished while it was decreased by ∼65% in UM002B cells after 24 hours treatment with 1 μmol/L FR (Fig. 3A and B). In contrast, ERK1/2 phosphorylation was unchanged by FR treatment of OCM3 cells (Fig. 3A and B). ERK1/2 activation slowly decreased in 92.1 cells over the course of 5 hours with a half-life of ≤30 minutes, the earliest time point tested, whereas it decreased more rapidly in OMM1.3 cells and was almost completely abolished after 30-minute treatment (Fig. 3C). We also detected YAP localization by fluorescence microscopy after treating OMM1.3 cells with FR for 6 hours. FR significantly decreased the amount of YAP found in the nucleus, whereas cytoplasmic YAP levels were increased (Fig. 4A). FR concentrations as low as 0.1 nmol/L were sufficient to observe a significant change in YAP localization (Fig. 4B). Significant displacement of YAP from the nucleus to the cytoplasm was also observed in 92.1 and UM002B cells, but was not observed in OCM3 cells when treated with FR for 6 hours (Fig. 4C). These data suggest direct inhibition of activated Gαq/11 in UM cells results in decreased activation of MAPK and YAP pathways.
Oncogenic Gαq inhibition results in decreased levels of proteins involved in cell cycle and cell proliferation
Because Gαq activation is responsible for initiating many different signaling cascades, we wanted to take a more global approach to determine pathways that are affected in UM cells when oncogenic Gαq is inhibited by FR. This was done using a high-throughput antibody-based RPPA analysis that utilized validated antibodies. Cell lysates from DMSO or FR treated UM cells were probed for 218 different proteins and phosphoproteins and analyzed for changes in protein levels (Fig. 5). The 20 proteins that were most significantly up- or downregulated upon Gαq inhibition by FR in 92.1 and OMM1.3 cells are shown in Fig. 5A. Many of the same proteins were significantly altered in both cell lines tested, whereas a few were specific to one or the other. In both cell lines, proteins that promote cell-cycle progression, such as CDK1, cyclin-B, PLK1, and p-Rb, were significantly decreased by FR treatment, whereas the observed decreased phosphorylation of the S6 ribosomal subunit and decreased p-90RSK are indicative of cell growth arrest (43). Interestingly, we also observed a number of proteins that increased in expression following FR treatment. These included Pdcd4, a programmed cell death protein that suppresses tumor progression (44), HER3 and pSTAT3 (Fig. 5A). The changes in expression or phosphorylation for many of these proteins was validated by immunoblotting of control and FR-treated UM cell lysates (Fig. 5B). In contrast, OCM3 cells showed no changes in expression or phosphorylation of those proteins when treated with FR for 24 hours (Fig. 5B). These data are in line with the observed effects of FR on UM cell growth (Fig. 2).
FR treatment of UM cells induces cell growth arrest and apoptosis
We next investigated the effect of FR treatment on cell cycle, as measured by PI staining and EdU incorporation. FR reduced S phase entry in 92.1 cells (Fig. 6A). After 24 hours treatment with FR, ∼96% of cells were in the G1 phase whereas ∼0.3% were in S phase (Fig. 6B). This compared with ∼61% of untreated cells in G1 phase and 30% in S phase. Twenty-four hour treatment with FR also resulted in a decrease in cells in G2–M phase from ∼8.5% in untreated cells to ∼3.7% in FR-treated cells. Similar results were observed in FR-treated OMM1.3 and UM002B cells, whereas OCM3 cells were not affected by FR (Fig. 6C). We also assessed whether FR induced apoptosis of UM cells using an activated caspase-3/7 detection reagent and FACS analysis. We found that by 24 hours there were significantly more 92.1 cells undergoing apoptosis when treated with 1 μmol/L FR compared with untreated cells (Fig. 7A and B). The same effect was observed in FR-treated OMM1.3 and UM002B cells, but not in OCM3 cells (Fig. 7C). By comparison, treatment with 1 μmol/L staurosporine, which induces apoptosis through caspase-3 activation (45), for 3 to 5 hours induced significant apoptosis in all cell lines. Therefore, FR induces G1 cell growth arrest and leads to apoptosis of UM cells containing activating mutations of Gαq/11.
FR treatment prevents UM colony formation and forces cellular differentiation in 3D culture
To obtain a better idea of how UM cells might be affected by FR-treatment in vivo, we utilized 3D-Matrigel-cell culture, which allows cells to grow in contact with extracellular matrix and provides positional information. OMM1.3 cells were seeded in the Matrigel culture and allowed to grow and form 3D cellular colonies of at least 20 cells in size, analogous to how they would initiate colonization in the liver. Untreated cells formed colonies within 12 days, whereas cells that were treated with 30 nmol/L FR immediately after being seeded were completely incapable of forming colonies during that same timespan (Fig. 8A). FR treatment also led to the disassembly of cells that had already formed OMM1.3 colonies, which continued to be unable to re-form colonies after 35 days of no additional treatment with FR (Fig. 8B). Lower doses of FR appear to induce OMM1.3 cell differentiation as evidenced by an increase in Melan-A expression with 1 and 10 nmol/L FR-treatment. However, at higher doses we observed little Melan-A expression and an increase in the proportion of cells undergoing apoptosis (Fig. 8C). These data convey that FR prevents colony formation and, depending on the degree of inhibition, disrupts colonies through the induction of cell differentiation or apoptosis to ultimately elicit a long-lasting effect.
UM is the most common intraocular tumor in adults and has a high rate of metastasis, leaving patients with a short survival time (1). UM is predominantly driven by activating mutations in Gαq/11 proteins, which promotes increased signaling in various cell proliferative pathways, such as the ERK1/2–MAPK and Hippo/YAP pathways (7, 8, 14, 22). To date, targeting proteins downstream of Gαq/11 has yielded unsuccessful results in clinical trials (19, 46). Therefore, it may be advantageous to inhibit oncogenic Gαq/11 directly. Here, we provide evidence that shows that a depsipeptide, that has previously been shown to inhibit wild-type Gαq/11, is also capable of inhibiting oncogenic Gαq and Gα11 in UM cells, promoting G1 cell-cycle arrest and either cell death or differentiation.
Previous studies have shown that the related compounds YM-254890 and FR900359 act as GDIs to inhibit GDP dissociation from Gαq and thereby inhibit GTP binding and activation of Gαq (27, 28, 31). FR has been shown to bind specifically to Gαq, Gα11, and Gα14 and inhibit activity in a pseudo-irreversible manner (32). These studies also demonstrated that FR can inhibit basal signaling mediated by Gαq-R183C and Gαq-Q209L expressed in HEK293 cells as well as signaling in some melanoma cell lines expressing mutant Gαq (32). Moreover, a very recent study has shown that FR is capable of inhibiting oncogenic Gαq/11-Q209L in UM cells (47). FR appears to function by trapping mutant Gαq in the inactive GDP-bound form by inhibiting nucleotide exchange and promoting Gαq association with Gβγ (47). In this study, we show that FR effectively inhibits signaling in UM cells expressing Gαq-Q209L, Gαq-Q209P, and Gα11-Q209L. We show in vitro that FR inhibits GTPγS-binding to Gαq-Q209L with an IC50 ∼75 nmol/L (Fig. 1). The IC50 determined in these studies is likely driven by the high concentration of Gαq that was used in these assays and thus does not reflect the 1 to 10 nmol/L IC50s previously reported for FR binding to Gαq (32, 47, 48). In addition, the high IC50 may also reflect the fact that we used Gαq in these studies whereas YM-254890 interacts with both the Gαq and Gβγ subunits in the heterotrimer (presumably FR does as well; ref. 28).
Because mutationally activated Gα subunits are thought to be constitutively associated with GTP, it is unclear how FR might inhibit the activity of activated Gαq/11. It might bind to GTP-loaded Gαq/11-Q209L and promote GTP dissociation, or there may be a slow exchange of GDP/GTP on Gαq/11-Q209L such that FR can bind in the GDP state. This latter possibility seems likely given the recent report on FR inhibition of Gαq-Q209L in UM cells (47), although this inhibition occurs within minutes in some cells (Fig. 3C). It is also possible that FR binds to newly synthesized GDP-bound Gαq/11 with the degradation of GTP-bound Gαq/11 being required to observe an effect. However, this seems unlikely because we see rapid inhibition of ERK activity (Fig. 3C) whereas the half-life of Gαq/11 is reported to be about 5 hours (49). Although further characterization of the mechanism of FR inhibition of mutant Gαq/11 is warranted, the effects of FR treatment on UM cells are striking.
Gαq/11 inhibition in UM cells containing oncogenic GNAQ/11-Q209 mutations leads to cell growth arrest within 24 hours (Figs. 2 and 6). FR-mediated G1 cell growth arrest is also detected in melanoma cells with (Hcmel12) or without (B16) a GNAQ-Q209L mutation (32) as well as in primary UM cell lines containing a GNAQ-Q209L mutation (47). In addition, we show that FR treatment is capable of producing similar results in metastatic UM cells containing GNAQ-Q209P or GNA11-Q209L mutations. Given the prevalence of Q209L and Q209P mutations in UM, it is worth noting that Gαq-Q209P has been shown to have altered binding to various effectors and Gβγ compared with Gαq-Q209L, although it still effectively activates downstream signaling (50). FR treatment likely does not affect the growth of mutant BRAF cell lines such as OCM3 due to the lack of dependency on GNAQ/11 for cell growth (Figs. 2 and 6). Thus, OCM3 cells have been a valuable control in our studies, providing support for the specificity and noncytotoxic nature of FR. These results are similar to recent studies in which OCM-1A cells were also used as a control for FR treatment (47).
Inhibition of Gαq/11 in UM cells harboring GNAQ-Q209L/P or GNA11-Q209L, but not in melanoma cells containing wild-type GNAQ/11, abolishes MAPK signaling (Fig. 3) and inhibits YAP translocation from the cytoplasm to the nucleus (Fig. 4), thereby hindering YAP-mediated transcription of cell proliferative genes. In general, these results are similar to studies of mutant GNAQ-Q209L melanoma and UM cells treated with GNAQ siRNA or FR (13, 14, 32, 47, 51). However, we see observable differences in YAP localization among the different cell lines. In OCM3 cells, YAP is largely equally distributed between the nucleus and cytoplasm and is unchanged by FR treatment. In both the OMM1.3 and 92.1 cells, YAP is primarily nuclear before treatment and moves to the cytoplasm with FR treatment. The effects of FR on YAP localization in UM002B cells are not as dramatic because YAP is not highly localized in the nucleus in UM002B cells, although FR does significantly decrease the YAP levels in the nucleus. The inhibition of the MAPK and YAP pathways likely contributes to the changes in cell cycle and cell growth associated proteins we detected by RPPA.
Upon FR treatment, we observed a significant decrease in the expression of proteins involved in regulating cell cycle, such as CDK1, cyclin B, PLK1, pRb, and Wee1 (Fig. 5), as further evidence of a disruption in cell cycle. We also observed a significant decrease in the phosphorylation of the S6 ribosomal subunit and of its activating kinase, p90RSK, which is characteristic of cell growth arrest (Fig. 5; ref. 43).
Utilizing a Matrigel 3D culture, UM cells were allowed to grow in a 3D manner as they would in vivo. FR treatment prevents the colonization of single UM cells, and disrupts colonies that were allowed to form before treatment (Fig. 8). Of note, UM cells remain unable to form colonies even after 35 days of no treatment/washout (Fig. 8B). These suggest that Gαq/11 inhibition causes long-term effects in UM cell behavior consistent with a reprogramming into a more differentiated phenotype. Further, if used as a treatment option, FR might not need to be continuously administered to patients. Upon treatment with lower doses of FR, UM cells show an increased expression of melan-A, a protein expressed on the surface of melanocytes (52), but lose that expression and begin dying at higher doses of the molecule (Fig. 8C). Therefore, it appears that lower doses of FR have a correcting effect on aberrant Gαq/11 signaling that causes the cells to differentiate into normal melanocytes. The differentiation of B16 melanoma cells by FR has previously been reported (32), whereas recent studies suggest FR promotes differentiation of primary 92.1 and Mel202 UM cells containing Gαq-Q209L by promoting polycomb-mediated gene repression (47). UM cells treated with higher doses of FR may succumb to cell death (Figs. 7 and 8C), although there appears to be some compensatory mechanisms in play. For example, a significant increase in HER3 expression was observed upon FR treatment (Fig. 5). This is thought to be a compensatory mechanism for cell survival as a result of decreased ERK signaling, which has previously been observed in melanoma cells harboring a BRAFV600E mutation when treated with a RAF inhibitor, effectively suppressing ERK activation (53, 54). In the presence of its ligand (neuregulin 1), HER3 heterodimerizes with HER2 to recruit PI3 kinase and activate Akt, promoting cell survival (54, 55). An increase in activated STAT3 was also detected upon FR treatment (Fig. 5). STAT3 also activates Akt and may be enhanced by HER3 signaling (56).
In summary, the data presented here show that FR inhibits oncogenic Gαq/11-Q209 protein in vitro and in UM cells, promoting cell-cycle arrest and eventual cell death, while also preventing the colonization of UM cells in 3D culture. Our work highlights the potential use of Gαq/11 inhibitors such as FR in the treatment of UM.
Disclosure of Potential Conflicts of Interest
A.E. Aplin reports receiving a commercial research grant from Pfizer. J. Aguirre-Ghiso has ownership interest (including stock, patents, etc.) in HiberCell; and is a consultant/advisory board member of HiberCell and Eli Lilly. No potential conflicts of interest were disclosed by the other authors.
Conception and design: D. Lapadula, E. Farias, P.B. Wedegaertner, J. Aguirre-Ghiso, J.L. Benovic
Development of methodology: D. Lapadula, D. McGrath, T. Sato
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D. Lapadula, E. Farias, C.E. Randolph, D. McGrath, T.H. Charpentier, L. Zhang, T. Sato, J. Aguirre-Ghiso
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D. Lapadula, E. Farias, C.E. Randolph, T.J. Purwin, T.H. Charpentier, L. Zhang, S. Wu, P.B. Wedegaertner, A.E. Aplin, J. Aguirre-Ghiso, J.L. Benovic
Writing, review, and/or revision of the manuscript: D. Lapadula, E. Farias, C.E. Randolph, T.H. Charpentier, L. Zhang, S. Wu, M. Terai, T. Sato, N. Zhou, A.E. Aplin, J. Aguirre-Ghiso, J.L. Benovic
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Wu, M. Terai, G.G. Tall, N. Zhou
Study supervision: J. Aguirre-Ghiso, J.L. Benovic
The authors would like to thank Dr. John Sondek for providing purified GαqΔ34-Q209L and Christian Heine for technical support. This research was supported by the Dr. Ralph and Marian Falk Medical Research Trust Bank of America, N.A., Trustee (to J. Benovic, A. Aplin, P. Wedegaertner, T. Sato, and J. Aguirre-Ghiso) and NIH awards P01 HL114471 (to J. Benovic), R01 GM56444 (to P. Wedegaertner), R03 CA202316 (to P. Wedegaertner), F31 CA225064 (to D. Lapadula), and F31 CA224803 (to C. Randolph). Dr. A. Aplin is also supported by a Cure Ocular Melanoma/Melanoma Research Foundation Established Investigator award. S. Wu and N. Zhou were supported by the Foundation of Research Center of Siyuan Natural Pharmacy and Biotoxicology and the National Natural Science Foundation of China. RPPA studies were supported by the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation (to A. Aplin). Research reported in this publication utilized the Flow Cytometry and MetaOmics Shared Resources at the Sidney Kimmel Cancer Center at Jefferson Health and was supported by the NCI of the NIH under Award Number P30CA056036. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.