Targeting nucleotide exchange to inhibit constitutively active G … · tagged [FLAG and StrepII (FS)] wild-type or constitutively active G a q. The data indicate that FR was able

SC I ENCE S I GNAL ING | R E S EARCH ART I C L E

DRUG DEVELOPMENT

1Department of Biochemistry and Molecular Biophysics, Washington UniversitySchool of Medicine, St. Louis, MO 63110, USA. 2Department of Cell Biology andPhysiology, Washington University School of Medicine, St. Louis, MO 63110, USA.3Department of Ophthalmology, Washington University School of Medicine, St. Louis,MO 63110, USA.*Corresponding author. Email: [email protected] (M.D.O.); [email protected](K.J.B.)

Onken et al., Sci. Signal. 11, eaao6852 (2018) 4 September 2018

Copyright © 2018

The Authors, some

rights reserved;

exclusive licensee

American Association

for the Advancement

of Science. No claim

to original U.S.

Government Works

Dow

nloaded

Targeting nucleotide exchange to inhibit constitutivelyactive G protein a subunits in cancer cellsMichael D. Onken1*, Carol M. Makepeace2, Kevin M. Kaltenbronn2, Stanley M. Kanai2,Tyson D. Todd2, Shiqi Wang1, Thomas J. Broekelmann2, Prabakar Kumar Rao3,John A. Cooper1,2, Kendall J. Blumer2*

Constitutively active G protein a subunits cause cancer, cholera, Sturge-Weber syndrome, and other disorders. Ther-apeutic intervention by targeted inhibition of constitutively active Ga subunits in these disorders has yet to beachieved. We found that constitutively active Gaq in uveal melanoma (UM) cells was inhibited by the cyclic depsipep-tide FR900359 (FR). FR allosterically inhibited guanosine diphosphate–for–guanosine triphosphate (GDP/GTP) ex-change to trap constitutively active Gaq in inactive, GDP-bound Gabg heterotrimers. Allosteric inhibition of otherGa subunits was achieved by the introduction of an FR-binding site. In UM cells driven by constitutively active Gaq,FR inhibited secondmessenger signaling, arrested cell proliferation, reinstatedmelanocytic differentiation, and stimu-lated apoptosis. In contrast, FR had no effect on BRAF-driven UM cells. FR promoted UM cell differentiation by reacti-vating polycomb repressive complex 2 (PRC2)–mediated gene silencing, a heretofore unrecognized effector systemofconstitutively active Gaq in UM. Constitutively active Gaq and PRC2 therefore provide therapeutic targets for UM. Thedevelopment of FR analogs specific for other Ga subunit subtypes may provide novel therapeutic approaches fordiseases driven by constitutively active Ga subunits ormultiple G protein–coupled receptors (GPCRs) where targetinga single receptor is ineffective.

fro

on June 12, 2021

http://stke.sciencemag.org/

m
INTRODUCTION
Heterotrimeric G proteins transduce signals from hundreds of cell sur-face G protein–coupled receptors (GPCRs) to intracellular signalingnetworks that regulate diverse biological processes. By undergoingGPCR-stimulated guanosine diphosphate–for–guanosine triphosphate(GDP/GTP) exchange followed by GTP hydrolysis, G protein a sub-units cycle between inactive GDP-bound and active GTP-bound statesto determine the duration, magnitude, and specificity of biological re-sponses (1). In cholera, certain cancers (2), Sturge-Weber syndrome (3),and other disorders, this cycle is disrupted by mutant or covalentlymodified Ga subunits that, by failing to hydrolyze GTP, are consti-tutively active.

Constitutively active mutant forms of Gaq or its close relative Ga11are the oncogenic drivers in nearly 90% of uveal melanoma (UM) pa-tients (4–6). UM is the most common cancer of the eye, and the eye isthe second most common site of melanoma. Regardless of primary tu-mor treatment, nearly half of UM patients develop metastatic disease(7) with amean survival of less than 1 year (8). Therapies to treat primarytumors and treat or prevent metastatic disease are needed. Inhibitors ofindividual signaling pathways downstreamofGaq/11 are being studied inUM clinical trials, but all have failed thus far (9). Thus, therapeuticapproaches that directly target constitutively active Gaq/11 may be re-quired to inhibit all necessary downstreamoncogenic signalingnetworks.

Constitutively active Ga subunits have yet to be targeted pharmaco-logically in disease due to challenges analogous to those of inhibitingoncogenic Ras (10–12). GTP hydrolysis defects would be extremely dif-ficult to correct pharmacologically, and the high affinity of Ga subunits

for GTP or GDP precludes the generation of effective competitive in-hibitors of guanine nucleotide binding.

However, other evidence led us to consider that constitutively activeGaq can be targeted in UM by pharmacologically inhibiting GDP/GTPexchange. Although nucleotide exchange by soluble Gaq is very slowin vitro (13), it is enhanced markedly by lipid membranes (14, 15) andRic-8a (16, 17), a nonreceptor guanine nucleotide exchange factor(GEF) and folding chaperone. Nucleotide exchange, therefore, may oc-cur at appreciable rates in cells; however, constitutively active Gaq stillwould exist predominantly in the active GTP-bound state because av-erageGTP/GDPratios in human cells are~8:1 (18), andGTPdissociates~10-fold slower than GDP (13). Nevertheless, we reasoned that thisequilibriummight be driven toward the GDP-bound state by inhibitingGDP dissociation, which would cause constitutively active Gaq to as-semble into inactive GDP-bound Gabg heterotrimers and thereby at-tenuate all downstream oncogenic signaling networks.

Here, we found that constitutively active Gaq can be targeted phar-macologically in UM cells by FR900359 (FR), a naturally occurring,cyclic depsipeptide that has been shown previously to inhibit wild-typeGaq by interfering allosterically with GDP dissociation (19–21). Weshow that FR can trap constitutively active Gaq in the GDP-bound stateand inhibit downstream signaling in UM cells. We also show thatconstitutively active Gaq drives oncogenesis by a previously unknownmechanism that antagonizes epigenetic silencing. Our results suggestthat targeting nucleotide exchange is a novel, general strategy forinhibiting Ga subunits in cancer and other diseases.

RESULTSFR traps mutant, constitutively active Gaq in theGDP-bound stateTo determine whether constitutively active Gaq undergoes appreciableGDP/GTP exchange in cells, we investigated whether Gaq(Q209L), acommon oncogenic guanosine triphosphatase–defective mutant inUM, could be trapped in the GDP-bound state by FR. The GDP- and

1 of 11



on June 12, 2021http://stke.sciencem

ag.org/D

ownloaded from

GTP-bound states of Gaq(Q209L) were assessed by detecting interactionwithGbg subunits, whichbindpreferentially toGDP-loadedGa subunits(1), or with regulator of G protein signaling 2 (RGS2), which binds GTP-loaded but not GDP-loaded Gaq (22). To detect protein-protein interac-tions, we used split luciferase complementation assays (23) using clickbeetle green (CBG) luciferase to study activation state–dependent inter-action between Ga subunits and cognate binding partners in living cells


(24) or copurificationwith tandemaffinity-tagged [FLAG and StrepII (FS)] wild-typeor constitutively active Gaq.

The data indicate that FR was able totrap constitutively active Gaq in the GDP-bound state. Split luciferase complemen-tation between Gb1/Gg2 and wild-type orconstitutively active Gaq was increased byFR [half-maximal effective concentration(EC50), 3 and 9 nM, respectively; Fig. 1, Aand B]. FR also increased the associationbetween Gbg and affinity-tagged wild-typeor constitutively active Gaq (19- and 9-fold,respectively), as indicated by copurification(Fig. 1C). Conversely, FR drove consti-tutively active Gaq out of the active GTP-bound state, as indicated by inhibition ofsplit luciferase complementation betweenconstitutively active Gaq and RGS2 [half-maximal inhibitory concentration (IC50),0.4 nM; Fig. 1D]. FR was selective for Gaq,as revealed by its lack of effect on the inter-action between wild-type or constitutivelyactive Ga13(Q226L) and Gb1g2 (Fig. 1A)or between constitutively active Ga13 andthe RGS domain of leukemia-associatedRhoGEF(LARG) (Fig. 1D).Thus,FRdrivesconstitutively active Gaq from its activeGTP-bound state into inactive GDP-boundGabg complexes.

As a further test of the ability of FR toinhibit constitutively active Gaq, we mea-sured downstream signaling. FR inhi-bited the induction of a transcriptionalreporter driven by constitutively activeGaq (IC50, 1 nM; Fig. 1E) but had noeffect on the expression of the reporterwhen driven by constitutively active Ga13(Fig. 1E). Crystallographic and muta-genesis studies of wild-type Gaq identi-fied amino acid residues (Arg60, Val184,and Ile190) that are important for inhibi-tion by YM-254890 (19), an inhibitor near-ly identical to FR. We found that singleamino acid substitutions at any of thesesites (R60K, V184S, and I190N) in con-stitutively active Gaq were sufficient toblunt the inhibitory potency of FR (IC50,30 to 70 nM; Fig. 1E), demonstrating thatFR targets constitutively active and wild-type Gaq by using the same binding site.

FR inhibits Gai1 bearing an engineered FR-binding siteFR inhibits receptor-evoked signaling by wild-type Gaq and its closerelativesGa11 andGa14 but not by otherGa subunits (20).We thereforedetermined whether an FR-insensitive Ga subunit could be convertedinto anFR-sensitive formby introducing an FR-binding site. Thismightbe possible because (i) all Ga subunits release GDP by a common allo-steric mechanism (25), (ii) structural elements of the allosteric relay

Fig. 1. FR traps mutant constitutively active Gaq in the inactive GDP-bound state. (A) Split luciferase com-plementation assays in human embryonic kidney (HEK) 293 cells, measured as reconstitution of CBG luciferase activitynormalized to cotransfected, constitutively expressed Renilla luciferase to assess the effect of FR on the activity state ofGa subunits as determined by interaction of Gb1g2with the indicatedwild-type (WT) ormutant constitutively active formsof Gaq (q) or Ga13 (13) (Q209L and Q226L, respectively). Data are means ± SEM of three experiments. (B) Split luciferasecomplementation assays in HEK293 cells, as described in (A), to assess the potency of FR as a driver of interaction betweenGb1g2 andwild-type ormutant constitutively active Gaq. Data aremeans ± SEM from three experiments. CBGC, C-terminalportion of CBG luciferase; CBGN, N-terminal portion of CBG luciferase. (C) Effect of FR treatment of HEK293 cells on co-purification of endogenous Gbgwith overexpressed, affinity-taggedwild-type or constitutively active Gaq(Q209L). Affinity-tagged Gaq and Gbg were detected by immunoblotting (IB) with FLAG and Gb antibodies, respectively. Data shown arerepresentative of three independent experiments. (D) Split luciferase complementation assays in HEK293 cells, as de-scribed in (A), to assess the potency of FR as an inhibitor of interaction between RGS2 and mutant constitutively activeGaq(Q209L) or between the RGS domain of LARG [LARG(RGS)] and mutant constitutively active Ga13(Q226L). Data aremeans ± SEM from three experiments. (E) Potency of FR as inhibitor of signal transduction by constitutively active GaqorGa13 detectedwith an SRE(L) promoter-driven transcriptional reporter inHEK293 cells. Data aremeans ± SEM from threeexperiments. Constitutively active Gaq(Q209L), constitutively active Gaq bearing the indicated FR-binding site mutations,and constitutively Ga13(Q226L) were studied. Concentration-response curves were fit by nonlinear regression to deriveEC50 or IC50 values, whichwere compared by t test. *P < 0.05, **P< 0.01 by t test; significancewas confirmed using q < 0.05or q < 0.01 by the false discovery rate (FDR) method of Benjamini and Hochberg. n.s., not significant.

2 of 11



D

include the FR-binding site (19, 25), and (iii) FR-insensitive Ga sub-units contain a similar but diverged form of the FR-binding site.

To test this hypothesis, we engineered an FR-binding site into Gai1by changing eight diverged amino acids to match their counterparts inthe FR-binding site of Gaq, producing “Gi/q” a subunits (illustrated inFig. 2A).We chose Gai1 because it is insensitive to FR (20) and becauseits function is studied readily in biochemical and cell-based assays. As acontrol, we also made a Gai/q(R54K) mutant, corresponding to a Gaqmutant that is less sensitive to FR (19). We found that FR inhibited therate of nucleotide exchange byGai/q in vitro (IC50, 4.6 mM), as indicatedby the binding kinetics of a fluorescent, nonhydrolyzable GTP analog(BODIPY-GTPgS; Fig. 2B and fig. S1A), which is rate-limited by GDPrelease. As expected, FR was >10-fold less potent toward Gai/q(R54K)(IC50, 76 mM;Fig. 2B and fig. S1A). Similarly,we found that FR inhibitedGai/q signaling in cells. FR attenuated the ability of Gai/q activated bycannabinoid type 1 receptors to inhibit forskolin-induced cyclic aden-osine 3′,5′-monophosphate (cAMP) accumulation (IC50, 25 nM; Fig.2C and fig. S2, C and D). FR was ~30-fold less potent toward Gai/q(R54K) (IC50, ~800 nM; Fig. 2C and fig. S2, C and D). Thus, FR targets


Gai/q andGaq by using the same binding site. This finding suggests thatFR-like molecules could be created to target the analogous, but distinct,binding sites of other Ga subtypes, providing a general approach to dis-cover novel chemical probes of Ga function and potential therapeuticsfor various diseases.

FR targets constitutively active Ga subunits by inhibitingnucleotide exchangeIn principle, FR could target constitutively active Ga subunits byinhibiting nucleotide exchange or restoring GTP hydrolysis. We testedboth hypotheses by using Gai/q; Gaq was unsuitable due to its unusualnucleotide binding properties that confoundmeasuringGTPhydrolysisin vitro. We found that constitutively active Gai/q(Q204L) [equivalentto Gaq(Q209L)] exhibited a severe defect in the catalytic rate of GTPhydrolysis that was not corrected by FR (Fig. 2D), whereas wild-typeGai/q hydrolyzed GTP effectively (Fig. 2D). In contrast, FR effectivelyinhibited nucleotide exchange by constitutively active Gai/q (IC50,2.7 mM; Fig. 2E). Thus, FR targets constitutively active Ga subunitsby inhibiting nucleotide exchange rather than restoringGTPhydrolysis.


ag.org/ow

nloaded from

FR inhibits signaling byconstitutively active Gaq in UM cellsTo determine whether FR inhibits signaltransduction by constitutively active Gaqin UM cells, we analyzed twoUM cell lines(Mel202 and 92.1) driven by constitutivelyactive Gaq(Q209L). A third UM cell line(OCM-1A) driven by constitutively activeBRAF(V600E) served as a negative control.Signaling by Gaq-stimulated phospho-lipase Cb was quantified on the basis ofthe abundance of inositol monophosphate(IP1), ametabolically stable product of ino-sitol 1,4,5-trisphosphate produced bycleavage of phosphatidylinositol 4,5-bisphosphate. In the absence of FR, IP1 was>50-fold more abundant in Gaq(Q209L)-driven Mel202 and 92.1 cells relative toBRAF(V600E)-driven OCM-1A cells(Fig. 3A). FR reduced IP1 abundance inMel202 and 92.1 cells >50-fold (Fig. 3B)but had only modest effect (~2-fold) onOCM-1A cells (Fig. 3B). Thus, FRmarkedlyinhibitedsecondmessengerproductiondriv-en by constitutively active Gaq in UM cells.

FR inhibits UM tumor cellproliferation and survivalDuring the preceding experiments, we al-so observed that FR treatment decreasedthe viability ofGaq(Q209L)-drivenUMtu-mor cells.Quantification confirmed that FRinhibited the proliferation of Gaq(Q209L)-driven Mel202 and 92.1 UM cells (EC50,6 and 2 nM, respectively; Fig. 4, A and B),with no effect on proliferation of BRAF(V600E)-drivenOCM-1Acells evenwhenapplied at a high concentration (10 mM).Flow cytometry revealed that FR induced

Fig. 2. Inhibition of Gai1 bearing an engineered FR-binding site. (A) Amino acid residues in Gai1 identical to ordiverged from the FR-binding site in Gaq are indicated in blue and red, respectively. An FR-binding site was engineered inGai1 by introducing the eight indicated amino acid substitutions so as tomatch the corresponding residues of Gaq, there-by producing a chimeric Ga subunit termed Gi/q. (B) Effect of FR on guanine nucleotide exchange by Gai/q in vitro.Nucleotide exchange was assayed by measuring increase in BODIPY-GTPgS fluorescence upon binding to the indicatedpurified His-tagged Ga subunits in the absence or presence of the indicated concentrations of FR. Gi/q(R54K) correspondsto a Gaq mutant that is less sensitive to inhibition by FR. Nucleotide exchange rates (kobs) of Gai/q and Gai/q(R54K) in theabsence of FR were 0.24 and 0.12 min−1, respectively. Data are means ± SEM from three independent experiments.(C) Effect of FR on agonist-evoked signaling mediated by Gi/q. Inhibition of forskolin-induced cAMP Förster by Gi-coupledcannabinoid receptors was measured in Neuro2A cells transfected with a cAMP fluorescence resonance energy transfer(FRET) reporter and pertussis toxin (PTX)–resistant and EE epitope–tagged forms of the indicated Ga subunits and treatedwith PTX to inactivate endogenous Gi. Inhibition of forskolin-induced cAMP formation by a cannabinoid receptor agonist(WIN 55,212-2; WIN) was measured by FRET. Attenuation of this inhibitory effect by FR was quantified relative to vehiclecontrols. Data aremeans ± SEM from three experiments. IC50 values of FR toward cells expressing Gi/q and Gi/q(R54K) were4.6 and 76 mM, respectively. (D) Failure of FR to correct the GTP hydrolysis defect of Gi/q(Q204L) in vitro. Hydrolysis of g

32P-GTP by the indicated Ga subunits present at 10-fold molar excess over GTP was measured over time. FR was added at theindicated time point (arrow). Data are means ± SEM from three independent experiments. (E) Effect of FR on guaninenucleotide exchange by constitutively active Gi/q(Q204L) in vitro, as determined by methods described in (B). The rateof nucleotide exchange (kobs) by Gi/q(Q204L) in the absence of FR was 0.14 min

−1. Concentration-response curves werefit by nonlinear regression to derive IC50 values, which were compared by t test. *P < 0.05, **P < 0.01 by t test; significancewas confirmed using q < 0.05 or q < 0.01 by the FDR method of Benjamini and Hochberg.

3 of 11




ag.org/D

ownloaded from

cell cycle inhibition (decreased fraction of S or G2-M phase cells) andapoptosis (increased fraction of sub-G1 phase cells) in Mel202 and 92.1cells but had no effect on OCM-1A cells (Fig. 4, C and D). FR thereforeinhibited the proliferation and survival ofUMcells inwhich constitutivelyactive Gaq is the oncogenic driver.


FU(AFsinddcfotemapUdDfl(9cincMSss

FR promotes melanocytic redifferentiation of UM cell linesWe observed that FR treatment caused Gaq(Q209L)-driven Mel202and 92.1 cells to undergo morphological changes indicative of rediffer-entiation. FR-treated Mel202 and 92.1 cells lost spindle morphology,became flatter, and produced multiple projections when compared with

ig. 4. FR-sensitive growth and viability ofM cells driven by constitutively active Gaq.) Changes in viability of UM cells treated withR. Cell viability was quantified using a water-oluble tetrazolium salt assay. The fold changecell viability over time is shown for Gaq(Q209L)-riven 92.1 and Mel202 cells and for BRAF(V600E)-riven OCM-1A cells in response to increasingoncentrations of FR. Data are means ± SEM fromur experiments performed in triplicate. (B) Po-ncy of FR as an inhibitor of UM cell viabilityeasured in 92.1, Mel202, and OCM-1A cell liness in (A). Data are means ± SEM from four ex-eriments performed in triplicate. The indicatedM cell lines were treated for 3 days with the in-icated concentrations of FR and analyzed forNA content. (C) Representative histograms fromow cytometry of the Gaq(Q209L)-driven cell lines2.1 and Mel202) and BRAF(V600E)-driven OCM-1Aells. PI, propidium iodide. (D) Potency of FR asducer of apoptosis (sub-G1 cells) and inhibitor ofell proliferation (S and G2-M phase cells) in 92.1,el202, and OCM-1A cell lines; data are means ±EM from four experiments. *P < 0.01 by t test;ignificance was corrected for multiple compari-ons using the Holm-Sidak method.

Fig. 3. FR inhibits signaling by constitutivelyactive Gaq in UM cells. (A) Inhibition of signalingbyconstitutivelyactiveGaq inUMcell lineswasquan-tified bymeasuring intracellular IP1, ametabolicallystable product of inositol 1,4,5-trisphosphate pro-duced by Gaq-stimulated phospholipase Cb. BasalIP1 values in UM cell lines driven by constitutivelyactive Gaq(Q209L) (92.1 and Mel202 cells) andBRAF(V600E) (OCM-1A cells). Data are means ±SEM from three experiments performed in triplicate.(B) Effect of FR on IP1 abundance in 92.1, Mel202,and OCM-1A cells. Data are means ± SEM of fourexperiments performed in triplicate. *P < 0.01 byt test; significance was confirmed using q < 0.01by the FDR method of Benjamini and Hochberg.

4 of 11



vehicle-treated cells or FR-treated OCM-1A cells (Fig. 5A). Further-more, FR increased melanocytic pigmentation in Mel202 and 92.1cells as compared to vehicle-treated cells or FR-treated OCM-1Acells (Fig. 5B). Similarly, FR increased the expression of two pigmen-tation enzymes [tyrosinase (TYR) and dopachrome tautomerase(DCT)] and premelanosome protein (PMEL) in Mel202 and 92.1cells but not in OCM-1A cells (Fig. 5C). FR increased the proportionof DCT-positive 92.1 and Mel202 cells (3.5- and 1.6-fold, respective-ly), TYR-positive Mel202 cells (1.8-fold), and PMEL-positive 92.1 andMel202 cells (6.8- and 3.9-fold, respectively). Thus, FR antagonizedthe dedifferentiation process driven by constitutively active Gaq inUM cells.

FR alters expression of genes regulated by constitutivelyactive Gaq in UM cellsTo explore how FR regulates phenotypes of UM cells driven byconstitutively active Gaq, we analyzed global gene expression by RNA se-


quencing (RNA-seq). Although FR caused an apoptotic response in UMcells, it did not markedly increase the expression of proapoptoticgenes or decrease the expression of survival genes. Instead, FR mod-estly decreased the expression of two proapoptotic BCL2 familymembers (BBC3/PUMA by 3.5-fold and PMAIP1/NOXA by 2.5-fold;fig. S2). Other BCL2 family members showed insignificant changesin expression, and broader examination of apoptosis-related genesshowed only small effects (fig. S2). These results are consistent withevidence that intrinsic and extrinsic apoptotic pathways function in-dependently of gene transcription (26, 27).

Cell cycle genes were also relatively unaffected by FR, as indicated bygene set enrichment analysis (GSEA) and direct comparison of the RNA-seqdata (fig. S2).The cyclin-dependentkinase inhibitorp21CIP1 (CDKN1A,down threefold) showed reduced expression, and several cell cycle genesshowed upward trends lacking statistical significance (fig. S2). Targets ofE2F, a transcription factor positively regulated by cyclin-dependent kinases,were positively enriched (data file S1). These results suggest that the anti-


ag.org/D

ownloaded from

proliferative effects of FR are not mediatedprimarily by short-term changes in cell ex-pression of cycle genes.

FR restores UM cell differentiationby promoting polycomb-mediatedgene repressionIn contrast to the results obtained for ap-optosis and cell cycle genes, large changesin expressionwere observed for differenti-ation and developmental genes in FR-treated Gaq(Q209L)-driven 92.1 cells(data files S1 and S2). Results of a multi-dimensional gene expression analysiscomparing relative patterns of expressionof all genes across all samples showed thata large number of gene expression changeswere associated specifically with FR treat-ment (Fig. 6A). Most strikingly, a distinctgene cluster showed marked reduction inexpression [4- to ~160-fold; Fig. 6B(circled) and data file S3]. Among genesin this cluster, 38% are associated with celldifferentiation and development, as re-vealed by gene ontology (GO) analysis(Fig. 6C and data file S4). A total of 42%of genes in this cluster are known targetsof the polycomb repressive complex 2(PRC2) during differentiation from em-bryonic stem cells, as indicated by GSEA(Fig. 6D and data file S5). These resultswere confirmed by quantitative real-timepolymerase chain reaction (PCR) analysisof FR-treated 92.1 cells (fig. S3). In con-trast, expression of PRC2-regulated genesin BRAF(V600E)-driven OCM-1A cellswas unaffected by FR (fig. S3), demon-strating specificity of FR for developmen-tal and differentiation genes targeted byconstitutively active Gaq in UM cells.

TheRNA-seq analyses suggested anovelmechanism for Gaq-induced oncogenesis

Fig. 5. FR induces redifferentiation of UM cells driven by constitutively active Gaq. (A) Morphological changeselicited by FR in UM cell lines driven by constitutively active Gaq (92.1 and Mel202) or BRAF-V600E (OCM-1A) treatedfor 3 days with FR and imaged by phase-contrast microscopy. Representative fields from one of three experiments areshown. (B) Melanocytic differentiation of FR-treated UM cells indicated by pigmentation. 92.1, Mel202, and OCM-1AUM cell lines were treated for 3 days with FR. Cells were pelleted and examined macroscopically; representativeimages fromone of three experiments performed in triplicate. (C) Induction ofmelanocytic markers by FR as indicatedby immunofluorescence staining of tyrosinase (TYR), dopachrome tautomerase (DCT), and premelanosome protein(PMEL) of Gaq-mutant cell lines (92.1 and Mel202) but not BRAF-driven UM cells (OCM-1A). Images are representativefields from one of three experiments. Scale bars, 50 mm.

5 of 11




ag.org/D

ownloaded from

in UM in which constitutively active Gaq antagonizes PRC2-mediatedgene repression, thereby reactivating genes associated with stemnessand driving dedifferentiation of UM cells into a more stem-like pheno-type. FR treatment inhibits constitutively activeGaq, relieves blockade ofPRC2-mediated repression, resilences these genes, and returns UM cellsto a melanocytic state. Consistent with this hypothesis, we found thatinhibiting the catalytic subunits of PRC2 complexes (EZH1/2) withGSK503 maintained 92.1 cells in an undifferentiated state and blockedthe ability of FR to redifferentiate these cells, as indicated bymorphology(Fig. 6E) and pigmentation (Fig. 6F). The effectiveness of GSK503 atinhibiting histoneH3 Lys27 (H3K27)methylationwas confirmed by im-munoblotting histones isolated from control and FR-treated 92.1 cells(Fig. 6G).


DISCUSSIONGDP/GTP exchange can be exploited as an Achilles heel ofconstitutively active Ga subunitsThemost important discovery provided by our studies is that GDP/GTPexchange is an underappreciated vulnerability of constitutively active Gasubunits and one that can be exploited pharmacologically in UM andother diseases. Although Ga subunits undergo GDP/GTP exchangeslowly in vitro, we discovered that nucleotide exchange occurs in cellsat rates sufficient for constitutively activeGaq to be trapped in the inactiveGDP-bound state by treatment with FR, an allosteric inhibitor of GDPrelease. When trapped by FR, GDP-bound constitutively active Gaq as-sembles into Gabg heterotrimers, further suppressing GDP release andstabilizing the inactive state. Because FR-bound Gq heterotrimers are

Fig. 6. FR represses expression of differentia-tion genes by restoring function of the PRC2.(A) Gaq-mutant 92.1 UM cells were treated withFR or vehicle, and RNA was collected 1 and 3 days(1d and 3d, respectively) after treatment for RNA-seq analysis. Results of a multidimensional geneexpression analysis that compares the relativepatterns of expression of all genes across allsamples and groups genes with similar patterns.The graph shows samples positioned by their rel-ative gene expression values within each pattern.Dimension 1 (x axis; the most represented pat-tern) shows separation based on vehicle treat-ment (red balls) versus FR treatment (blue balls),whereas dimension 2 (y axis; the secondmost rep-resented pattern) shows separation based on timein culture (indicated by 1d or 3d on balls). (B) MAplot (M, log ratio; A, mean average) comparinggene expression between FR- and vehicle-treated92.1 samples identifies a group of significantly re-duced genes (circled; fold change, >2; FDR, q< 0.01)associated with FR treatment. (C) GO analysis ofthe FR-repressed gene set [circled in (B) with arrow].(D) FR-repressed genes [circled in (B) with arrow]identified as targets of the polycomb repressivecomplex 2 (PRC2) by GSEA. EGF, epidermal growthfactor; BMP2, bone morphogenetic protein; hESC,human embryonic stem cells. (E) Effect of theEZH1/2 inhibitor GSK503 on morphological dif-ferentiation elicited by FR. Representative fields areshown from one of three experiments of 92.1 UMcells treated for 7 days with GSK503 and for 3 dayswith FR and then imaged by phase-contrast mi-croscopy. Scale bar, 100 µm. (F) Effect of GSK503on pigmentation of FR-treated cells, visualized bymacroscopic inspection. 92.1 cells were treated for7 days with GSK503 and for 3 dayswith FR and pel-leted; representative images from one of three ex-periments. (G) PRC2 inhibitionbyGSK503. Immunoblotsof 92.1 cells treated for 7 days with GSK503 showreduced histone H3K27 trimethylation. Plot showsrelative fraction of trimethyl-histone H3K27 com-pared to dimethyl sulfoxide (DMSO) control andnormalized to total histone H3 from densitometrydata from three independent experiments. *P <0.01 by t test; significance was confirmed using q <0.01by theFDRmethodof Benjamini andHochberg.

6 of 11




ag.org/D

ownloaded from

refractory to activation byGPCRs (20), signaling networks downstreamof constitutively active Gaq are attenuated.

Although our study involved constitutively active Gaq in UM, weanticipate that constitutively active forms of Ga subunit subtypes thatdrive other types of cancermay also be vulnerable to allosteric inhibitorsof GDP release. Constitutively active Ga11 in UM (12) and Ga14 in vas-cular tumors (28) should be susceptible because wild-type forms ofthese Ga subunits are sensitive to FR (20). Although other subtypesof Ga subunits are insensitive to FR, all Ga subunits have a divergedbut related form of the allosteric regulatory site in Gaq that binds FR.This site includes conserved features of linker 1, which stabilizes theGDP-bound state by interacting with helix 1, helix A, and helix F as partof the universal mechanism that regulates GDP release. As predicted bythis hypothesis, we found that engineering an FR-binding site into anFR-insensitive Ga subunit was sufficient to confer FR sensitivity. Thus,we speculate that a collection of FR-like inhibitors, each of which selec-tively targets the diverged allosteric regulatory site of certain Ga sub-units, may provide a novel approach toward therapeutic developmentin cancers associated with othermutant constitutively active Ga subunits(2), cholera, andSturge-Weber syndrome. In addition, this approachmaybe efficacious for diseases that are driven by multiple GPCRs in whichblocking a single receptor is ineffective.

UM cells are addicted to constitutively active GaqAnother important discovery emerging from our study is that con-stitutively active Gaq has unexpectedly diverse functional roles as an on-cogenic driver in UM. Instead of affecting a single-cell biological process,FR inhibits proliferation, triggers apoptosis, and drives melanocytic redif-ferentiation ofUMcells.Our findings help to explainwhyprevious studiesusing mitogen-activated protein kinase kinase, Akt, and protein kinase Cinhibitors to target individual signaling pathways downstream of Gaq/11failed in clinical trials of UM (9). By inhibiting constitutively active Gaqand consequently attenuating all downstream signaling networks, FR orrelated inhibitors may impair the growth and survival of cancer cells inUMprimary tumors, and theymay also decrease the probability or extentof metastasis because melanocytic differentiation correlates inversely withmetastatic potential in UM (29, 30). For primary tumors, FR may slowconversion from the indolent state to aggressive class 2 tumors (31), orit may impair the growth or spread of metastatic lesions, including thosethat are clinically undetectable. Tumor-targeted or focal delivery of FRmaybe required because systemic administrationmight cause unacceptable sideeffects by inhibitingGaq/11 in normal tissues.Nevertheless, themarkedFR-sensitivity of UM tumor cells driven by constitutively active Gaq suggeststhat clinical investigation of this or related inhibitors could be considered.

Constitutively active Gaq antagonizes gene silencingby the PRC2We found a marked and unanticipated consequence of inhibitingconstitutively active Gaq in UM—the repression of gene sets that con-trol differentiation and development.Many of these repressed genes areinvolved in embryonic stem cell lineage specification and differentiationand are targets of epigenetic silencing by the PRC2, which acts throughH3K27 trimethylation (32–34). These repressed genes includeADRA2A(a2A-adrenergic receptor) andHAND2 (heart andneural crest derivativesexpressed-2). TheADRA2A genepromoterwas identified in independentscreens for PRC2 subunit binding and forH3K27 trimethylation (32–34),and ADRA2A has been linked to cancer progression and severity (35).The HAND2 gene promoter is also targeted by PRC2 binding andH3K27 trimethylation (32–34), especially in migrating cranial neural


crest cells, where HAND2 expression distinguishes neural crest celllineages during facial development (36).

Together, our findings reveal a novel mechanism in which signalingby constitutively active Gaq in UM cells antagonizes PRC2-mediatedgene silencing, thereby maintaining UM cells in a less differentiated statesimilar to premelanocytic cranial neural crest cells (36). This finding,coupled with previous studies of BAP1 (BRCA1-associated protein-1)(30,37), indicates that a temporal hierarchyof epigenetic regulationdrivestumorigenesis and progression inUM. Early in tumorigenesis, mutationsthat constitutively activate Gaq are acquired, which inhibits PRC2-mediated repression. Subsequent loss of BAP1, a histone H2A(Lys119)deubiquitinase that antagonizes repressionbyPRC1, then leads tometastasis.

MATERIALS AND METHODSFR900359FR was purified from Ardisia crenata according to published methods(20). The structure of purified FR relative to a commercially availableequivalent (UBO-QIC; University of Bonn, Germany) was establishedby nuclear magnetic resonance.

Biochemical assaysSplit luciferase assays were performed as described previously (24). TheN-terminal portion ofCBG luciferase (CBGN) was inserted into theaB-aC loopwithin the helical domain of wild-type and constitutively active(c.626A>T; Q209L) mutant forms of GNAQ (Gaq) and GNA13 (Ga13)(Q226L). Insertion of foreign proteins at this site preserves Ga subunitfunction (38). The C-terminal region of CBG luciferase (CBGC) wasfused to the N termini of GNB1 (Gb1), which was cotransfected withuntagged GNG2 (Gg2); RGS2; and the RGS domain of ARHGEF12(LARG), which interacts with Ga13 only in the active GTP-boundstate (39). HEK293 cells transiently transfected with various combi-nations of fusion constructs generating tagged proteins were treatedfor 18 hours with vehicle (DMSO) or FR, and luciferase assays wereperformed to measure reconstituted luciferase activity. Assays oftranscriptional reporters driven by Ga subunits were performed asdescribed previously (40). Ga-driven firefly luciferase activity wasnormalized to cotransfected Renilla luciferase expressed from aconstitutive promoter.

Experiments used tomeasure agonist-evoked inhibition of forskolin-induced cAMP formation in Neuro2A cells were performed in a 96-wellplate format, as described in (24) with slight modifications. Cells weretransfected with a cAMP FRET reporter and PTX-resistant forms ofGai1, Gai/q, or Gai/q(R54K). After cells were treated for 16 hours withPTX (100 ng/ml) to inactivate endogenously expressed Gi, FRET wasused tomeasure inhibition of forskolin (FSK)–induced cAMP formationby aCB1 cannabinoid receptor agonist (WIN55,212-2;WIN).A SynergyH4hybrid plate reader (BioTek) was used tomeasure FRET every 57 s byexciting the FRET donor (420/20-nm bandpass filter) and simulta-neously detecting donor and acceptor emissions (480/20- and 540/20-nm bandpass filters, respectively). FRET measurements of changesin intracellular cAMP were expressed ratiometrically as

DRR0

¼ FRET ratio� baseline FRET ratiobaseline FRET ratio

where FRET ratio is (480-nm emission/540-nm emission) at a giventime point, and the baseline FRET ratio is the average of (480-nmemission/540-nm emission) before FSK stimulation. Independent

7 of 11




ag.org/D

ownloaded from

experiments were performed in triplicate. Effects of FR on agonist-evoked adenylyl cyclase inhibition were quantified as the percentage ofDR/Ro relative to vehicle control. FR concentration-response curves weregenerated by a three-parameter fit.

Accumulation of IP1 inUMcells wasmeasured using the IP-OneKit(catalog number 62IPAPEB, Cisbio Inc.), according to the supplier’s in-structions. A total of 10,000 Mel 202 cells, 20,000 92.1 cells, and 20,000OCM-1Acellswere seeded intowhite-bottom tissue culture grade 384-wellplates. After an overnight incubation, cells were treated with FR or DMSOand returned to the incubator. The next day, stimulation buffer wasadded for 1 hour after which IP1-d2 and Ab-Cryp were added, andthe cells were incubated at room temperature for 60 min. Plates wereread in a Synergy H4 hybrid plate reader. Standard curves were gener-ated using reagents supplied with the kit.

Copurification of tandem affinity-tagged Gaq and endogenous Gbgwas assayed as follows. A gBlock gene fragment (IntegratedDNATech-nologies) encoding a glycine/serine linker (GGGSSGGG) followed by aFLAG-StrepII-StrepII (FS) tag and another glycine/serine linker(GGGSSGGG) was inserted between sequences encoding amino acidresidues 124 and 125 of mouse Gaq (wild-type and Q209L mutant),cloned into pcDNA3.1+, and verified by DNA sequencing. HEK293cells were plated in 10-cm dishes and transfected the next day withthe indicated plasmids. After 24 hours, transfection medium was re-moved and replaced with fresh media containing one FR (1 mM) or ve-hicle (DMSO). Twenty-four hours after treatment, media wereremoved, and cells were washed with Dulbecco’s phosphate-bufferedsaline (PBS) and processed for tandem affinity purification (TAP).TAP was performed as described previously (24) with the followingmodifications: All subsequent steps were performed on ice or at 4°C.HEK293 cells were lysed in [50 mM tris (pH 8.0), 5 mM EDTA, 100 mMNaCl, 0.5% (v/v) IGEPALCA-630, 1mMMgCl2, and complete proteaseinhibitor mix (catalog number 11697498001, Roche), with or without1 mMFR] by sonication on ice for 2min (30 s on, 30 s off, 60%A), rotatedend-over-end for 30 min, and cleared by ultracentrifugation at 100,000gfor 15 min. Cleared lysates were incubated with Strep-Tactin resin (cat-alog number 2-1206-010, lot number 1206-0350, IBA) overnight withend-over-end rotation and washed three times in batch with 10 columnvolumes of wash buffer [50 mM tris (pH 8.0), 5 mM EDTA, 100 mMNaCl, 0.5% (v/v) IGEPAL CA-630, 1 mMMgCl2, complete protease in-hibitor mix, and ±1 mMFR]. Protein complexes were eluted two consec-utive times by incubating with a fivefold column volume of elution buffer[desthiobiotin buffer E (catalog number 2-1000-025, lot number 1000-5170, IBA), 0.1% (v/v) IGEPAL CA-630, 1 mM magnesium chloride,complete protease inhibitormixture, and ±1 mMFR] for 30min in batch.Strep-Tactin elution fractions were combined and incubated with anti-FLAG M2-Agarose (catalog number A2220, lot number SLBW1929,Sigma-Aldrich) in batch for 2 hours andwashed three times in batchwitha 10-fold column volume of wash buffer. Protein complexes were elutedfrom FLAG agarose by incubating with a fourfold column volume ofFLAG elution buffer [3xFLAG peptide (200 mg/ml); catalog numberF4799, lot number SLBM1190V in wash buffer, Sigma-Aldrich) for30 min in batch. Lysates and FLAG eluates were resolved on 12%SDS–polyacrylamide gel electrophoresis (PAGE) gels and transferredto Immobilon(PSQ) polyvinylidene difluoride (PVDF)membranes (cat-alog number ISEQ00010,Millipore).Membranes were blockedwith 5%(w/v) milk in TBST [25mM tris (pH 7.2), 150mMNaCl , 2.7 mMKCl,and 0.1% (v/v) Tween 20] and incubated with primary antibodiesovernight at 4°C. Lysate primary antibody mixture was ANTI-FLAGM2 (1:50,000; catalog number F1804, lot number SLBN5629V,


Sigma-Aldrich) and Gb H-1 (1:500; catalog number sc-166123, lotnumber G2414, Santa Cruz Biotechnology). TAP eluates were probedwith a primary antibodymixture consisting of anti-FLAGM2 (1:50,000;catalog number F1804, lot number SLBN5629V, Sigma-Aldrich) andanti-Gb1 H-1 (1:500; catalog number sc-166123, lot number G2414,Santa Cruz Biotechnology). Membranes were washed with TBST atleast three times and incubated with goat anti-mouse immunoglobulinG coupled to IRDye 800CW (catalog number 926-32210, lot numberC70712-15, LI-COR Biosciences). After incubation, membranes werewashed at least three times with TBST, and signals were detected usingan Odyssey model 9120 imaging system (LI-COR Biosciences).

Nucleotide exchange by purified His-tagged Ga subunits was as-sayed as follows. Plasmid pet14B-6xHIS-Gai1 was a gift fromM. Linder(Cornell University, New York). A pet14B-6xHIS-Gai/q plasmid wasgenerated by cloning a custom synthesized gBlock gene fragment(Integrated DNA Technologies) containing mutations encoding eightamino acid substitutions (V50I, K54R, Y69F, V72L, K180P, V185I,T187Y, and H188P) in 6xHIS-Gai1. Site-directed mutagenesis of thisplasmid was used to generate pet14B-6xHIS-Gai/q(R54K). Recombi-nant His-tagged Ga subunits were expressed and purified fromEscherichia coli according to publishedmethods (41). To detect nucle-otide exchange, fluorescence of the GTP analog BODIPY-GTPgS (cat-alog number G22183, Thermo Fisher Scientific) was recorded with aSynergy H4 hybrid plate reader in the absence or presence of recombi-nantGa subunits (42). The indicatedGa subunits (1mM)were incubatedin a black-wall clear-bottom 96-well plate (catalog number 3603, Costar)with vehicle (DMSO) or FR at the indicated concentrations for 20minat 25°C in reaction buffer [50 mM tris-HCl (pH 8.0), 1 mM EDTA,and 10 mMMgCl2]. BODIPY-GTPgS then was added to the reactionmixture at a final concentration of 25 nM to initiate nucleotide ex-change. The intensity of fluorescence emission was recorded at 30°Cevery 10 s for 30 min with excitation and emission wavelengths of 485and 528 nm, respectively, and the monochromator was set to a band-width of 9 nm. Specific fluorescence was normalized to backgroundfluorescence as follows

DFF0

¼

ðfluorescence from BODIPY‐GTPgS�bound GaÞ � ðfluorescence from BODIPY‐GTPgS aloneÞfluorescence from BODIPY‐GTPgS alone

GraphPad Prism was used to fit fluorescence (DF/F0) curves to apseudo first-order rate equation and obtain kobs values relative to ve-hicle controls. FR concentration-response curves were generated by athree-parameter fit of normalized kobs values.

Assays of GTP hydrolysis by His-tagged Gai/q subunits were per-formed as follows. Reactions contained wild-type Gai/q (1.5 mM) orconstitutively active Gai/q(Q204L) (3 mM) in reaction buffer [50 mMtris-HCl, 5 mM MgCl2, 1 mM EDTA, and 0.5 mM dithiothreitol(pH 8.0)]. Reactions performed in triplicate were started by adding re-action buffer containing g32P-GTP (final concentration, 100 nM; spe-cific activity, 100Ci/mmol). Ga subunits were present at >10-foldmolarexcess over GTP to assess the catalytic rather than steady-state rate ofGTP hydrolysis. As indicated, vehicle (DMSO) or FR (final concentra-tion, 25 mM) was added. Aliquots were removed at intervals, quenchedby addition of 1 N formic acid, and spotted on polyethyleneimine-cellulose thin-layer plates (catalog number Z122882, Sigma-Aldrich).Plates were developed in 0.5 M LiCl2 and 0.5 M formic acid, dried,wrapped in plastic, and exposed 1 hour to a phosphor storage screen

8 of 11




ag.org/D

ownloaded from

(GEHealthcare Life Sciences) thatwas visualized on aTyphoonFLA9500scanner (GEHealthcare Life Sciences) at a resolution of 100 mmand pho-tomultiplier voltage of 750V. Scanswere analyzed using Image StudioLiteto quantify the amount of labeled GTP and orthophosphate in each lane.GTP hydrolysis was calculated by dividing the magnitude of the ortho-phosphate signal by the sum of the orthophosphate and GTP signals.Datawere fit usingGraphPadPrism to a pseudo first-order rate equation.

UM cell culture assaysCells were cultured at 37°C in 5% CO2. Human UM cell lines 92.1,Mel202, and OCM-1A were derived by M. Jager (Laboratory of Oph-thalmology, Leiden University), B. Ksander (Schepens Eye Institute,Massachusetts Eye and Ear Infirmary), and J. Kan-Mitchell (BiologicalSciences, University of Texas at El Paso). UM cell lines were grown inRPMI 1640 medium (Life Technologies), supplemented with 10% fetalbovine serum and antibiotics. Cell viability wasmeasured using awater-soluble tetrazolium salt,WTS-8 (Bimake), following themanufacturer’sprotocol. Flow cytometry for analysis of cell proliferation and apoptosiswas performed at the SitemanCancerCenter FlowCytometryCore on aFACScan analyzer (BD Biosciences) using a standard propidium iodidestaining protocol, as described previously (43).

Immunofluorescence staining of UM cell lines was carried out byadding an equal volume of 2× fixative (PBS with 4% paraformaldehydeand 0.4% glutaraldehyde) to UM cells in RPMI 1640 growth medium.After 15min at 37°C, cells were permeabilizedwith 0.1%Triton X-100 inPBS for 5 min, washed with PBS, and blocked with 2% fish gelatin(Sigma-Aldrich) in PBS. Primary and secondary antibodies were dilutedin 2% fish gelatin inPBS. Primary antibodies includedmousemonoclonalanti-premelanosomal protein (One World Lab), rabbit polyclonal anti-TYR (One World Lab), rabbit polyclonal anti-DCT (One World Lab),rabbit polyclonal anti-S100 (DakoCytomation), and mouse monoclonalanti-BrdU (Life Technologies). Secondary antibodies wereAlexa Fluorconjugates (Life Technologies), and the mounting agent was ProLongGold (Life Technologies). Cell morphology was assessed by phase-contrast imaging with an inverted microscope (Olympus IX72) usinga 10× objective. Images were analyzed to compare the percentage ofcell populations that were positive for a givenmarker in representativefields from three independent samples. Significant difference relativeto vehicle control was defined as P < 0.01 by Fisher’s exact test for allcomparisons.

Immunoblotting of UM cell lines was performed by lysing cells in aradioimmunoprecipitation assay buffer [150 mM sodium chloride, 1%Triton X-100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate,and 50mM tris (pH 8.0)] with 1× complete protease inhibitormix (cat-alog number 11697498001, Roche). Cleared lysates were analyzed byimmunoblotting, as described above for HEK293 cells.

Histones were isolated fromUM cell lines by using the ActiveMotifHistone Purification Mini Kit (catalog number 40026, Active Motif).Lysates were resolved on 15% SDS-PAGE gels and transferred toan Immobilon(P) PVDF membrane (catalog number IPVH00010,Millipore). Membranes were blocked with 5% (w/v) milk in TBST[25mM tris (pH 7.2), 150mMNaCl, 2.7 mMKCl, and 0.1% (v/v) Tween20] and incubatedwith primary antibodies.Membranes werewashedwithTBSTat least three times and incubatedwith IRDye680–coupledgoat anti-rabbit and IRDye800or goat anti-mouse antibodies (LI-CORBiosciences).After incubation, membranes were washed at least three times with TBST,and signals were detected using Odyssey model 9120 imaging system(LI-COR Biosciences). Other primary antibodies used for immuno-blottingwere anti-EE (catalognumberMMS-115P, lotnumberE12BF00285,


Covance), anti-actin C4 (catalog number MAB1501, Millipore), anti-histone H3 (clone A3S; catalog number 05-928, MIllipore), and anti-histone H3-trimethyl-K27 (catalog number 6002, Abcam).

Statistical analysesAll statistical analyses were performed with GraphPad Prism. For splitluciferase complementation assays, IP1 assays, and viability assays, cellswere plated in triplicate wells and treated in parallel (technical repli-cates), and each experiment was performed three times on differentdays (biological replicates). Mean values and SEMs were calculatedfrom at least nine replicate values for each condition, and t tests wereperformed comparing each condition to controls to determine statisticalsignificance of FR treatment. For guanine nucleotide exchange andGTPhydrolysis assays, experiments were performed in triplicate, and each ex-periment was performed at least three times on different days with dif-ferent protein preparations.Mean values and SEMswere calculated fromat least nine replicate values for each condition, and t tests were per-formed comparing each condition to controls to determine statistical sig-nificance of FR treatment. For FRET reporter assays, FRET fluorescencesignals were quantified relative to vehicle controls from three independentexperiments, and mean values and SEM were calculated from the com-bined data of three experiments. IC50 values and confidence intervals werecalculated by the least-square nonlinear curve-fitting method for normal-ized response to FR. For histone methylation assays, densitometry wasperformed on immunoblots for trimethyl-histone H3K27 and com-pared to DMSO control and normalized to total histone H3 from threeindependent experiments. The statistical significance of response to FRtreatment was determined with t tests.

Gene expression analysis92.1UMcells were treatedwith 100 nMFRor vehicle (DMSO) inRPMIgrowthmedium and collected after 1 and 3 days of treatment. RNAwasisolated using the RNeasy Mini Kit (QIAGEN) following the manufac-turer’s protocol and including the optional DNase I treatment step.RNA quality was assessed on a Bioanalyzer 2100 (Agilent Technolo-gies). mRNA was extracted from total RNA using a Dynal mRNADirect kit, fragmented, and reverse-transcribed to double-strandedcomplementary DNAwith randomprimers before addition of adaptersfor library preparation. Library preparation andHiSeq 2500 sequencingwere performed by the Washington University Genome TechnologyAccess Center (gtac.wustl.edu). FastQ files were aligned to the tran-scriptome and the whole genome with STAR (Spliced TranscriptsAlignment to a Reference). Biologic replicates were simultaneously ana-lyzed by edgeR and Sailfish analyses of gene-level/exon-level features.Unexpressed genes and exons were removed from the analyses. Un-supervised principal component analysis and volcano plots were gener-ated in Bioconductor using edgeR. Significance Analysis ofMicroarraysversion 4.0 was used to generate a ranked gene list, and a threshold of q <10% and a fold change of >2.0 were then used to select the most highlystatistically significant genes that showed reduced expression in FR-treatedversus vehicle control cells. This list was used as signature gene sets for GOanalysis andGSEA (44). GOgroupswere assembled bymerging the lists ofgenes from related GO terms that were significantly enriched (P < 0.01using the Kolmogorov-Smirnov statistic) in the signature gene set. Signif-icant gene sets (P < 0.01 using the Kolmogorov-Smirnov statistic) fromGSEA analyses were combined such that genes associated with multiplerelated signatureswere only counted once and each genewas assigned onlyto a single combined group based on the signature with the highest enrich-ment score for that gene. Gene expression changes were validated in all

9 of 11

gtac.wustl.eduhttp://stke.sciencemag.org/


threeUMcell lines byquantitativePCRusing fast SYBRGreenMasterMix(ThermoFisherScientific), following themanufacturers’protocol.GAPDH(glyceraldehyde-3-phosphate dehydrogenase) was used as an endogenouscontrol. Primer sets used for the assay are listed in data file S6.

SUPPLEMENTARY MATERIALSwww.sciencesignaling.org/cgi/content/full/11/546/eaao6852/DC1Fig. S1. FR inhibition of Gai1 bearing an engineered FR-binding site.Fig. S2. Heatmaps of cell cycle and apoptosis gene expression in response to FR.Fig. S3. Validation of selected FR target genes.Data file S1. Gene sets with positive correlation to FR treatment in GSEA.Data file S2. Gene sets with negative correlation to FR treatment in GSEA.Data file S3. Genes within the FR-responsive, reduced-expression cluster.Data file S4. GO analysis results for the FR-responsive gene cluster.Data file S5. GSEA results for the FR-responsive gene cluster.Data file S6. Primers used for quantitative PCR gene expression analysis.


ag.org/D

ownloaded from

REFERENCES AND NOTES1. A. G. Gilman, G proteins: Transducers of receptor-generated signals. Annu. Rev. Biochem.

56, 615–649 (1987).2. M. O’Hayre, M. S. Degese, J. S. Gutkind, Novel insights into G protein and G protein-

coupled receptor signaling in cancer. Curr. Opin. Cell Biol. 27, 126–135 (2014).3. M. D. Shirley, H. Tang, C. J. Gallione, J. D. Baugher, L. P. Frelin, B. Cohen, P. E. North,

D. A. Marchuk, A. M. Comi, J. Pevsner, Sturge–Weber syndrome and port-wine stainscaused by somatic mutation in GNAQ. N. Engl. J. Med. 368, 1971–1979 (2013).

4. M. D. Onken, L. A. Worley, M. D. Long, S. Duan, M. L. Council, A. M. Bowcock, J. W. Harbour,Oncogenic mutations in GNAQ occur early in uveal melanoma. Invest. Ophthalmol. Vis. Sci.49, 5230–5234 (2008).

5. C. D. Van Raamsdonk, V. Bezrookove, G. Green, J. Bauer, L. Gaugler, J. M. O’Brien,E. M. Simpson, G. S. Barsh, B. C. Bastian, Frequent somatic mutations of GNAQ in uvealmelanoma and blue naevi. Nature 457, 599–602 (2009).

6. C. D. Van Raamsdonk, K. G. Griewank, M. B. Crosby, M. C. Garrido, S. Vemula, T. Wiesner,A. C. Obenauf, W. Wackernagel, G. Green, N. Bouvier, M. M. Sozen, G. Baimukanova,R. Roy, A. Heguy, I. Dolgalev, R. Khanin, K. Busam, M. R. Speicher, J. Brien, B. C. Bastian,Mutations in GNA11 in uveal melanoma. N. Engl. J. Med. 363, 2191–2199 (2010).

7. Y. Yonekawa, I. K. Kim, Epidemiology and management of uveal melanoma.Hematol. Oncol. Clin. North Am. 26, 1169–1184 (2012).

8. C. M. Balch, J. E. Gershenwald, S. J. Soong, J. F. Thompson, M. B. Atkins, D. R. Byrd,A. C. Buzaid, A. J. Cochran, D. G. Coit, S. Ding, A. M. Eggermont, K. T. Flaherty, P. A. Gimotty,J. M. Kirkwood, K. M. McMasters, M. C. Mihm Jr., D. L. Morton, M. I. Ross, A. J. Sober,V. K. Sondak, Final version of 2009 AJCC melanoma staging and classification. J. Clin. Oncol.27, 6199–6206 (2009).

9. R. D. Carvajal, G. K. Schwartz, T. Tezel, B. Marr, J. H. Francis, P. D. Nathan, Metastaticdisease from uveal melanoma: Treatment options and future prospects. Br. J. Ophthalmol.101, 38–44 (2017).

10. A. V. Smrcka, Molecular targeting of Ga and Gbg subunits: A potential approach forcancer therapeutics. Trends Pharmacol. Sci. 34, 290–298 (2013).

11. J. M. L. Ostrem, K. M. Shokat, Direct small-molecule inhibitors of KRAS: From structuralinsights to mechanism-based design. Nat. Rev. Drug Discov. 15, 771–785 (2016).

12. V. Chua, D. Lapadula, C. Randolph, J. L. Benovic, P. Wedegaertner, A. E. Aplin,Dysregulated GPCR signaling and therapeutic options in uveal melanoma.Mol. Cancer Res. 15, 501–506 (2017).

13. P. Chidiac, V. S. Markin, E. M. Ross, Kinetic control of guanine nucleotide binding tosoluble Gaq. Biochem. Pharmacol. 58, 39–48 (1999).

14. J. L. Blank, A. H. Ross, J. H. Exton, Purification and characterization of two G-proteins thatactivate the b1 isozyme of phosphoinositide-specific phospholipase C. Identification asmembers of the Gq class. J. Biol. Chem. 266, 18206–18216 (1991).

15. G. Berstein, J. L. Blank, A. V. Smrcka, T. Higashijima, P. C. Sternweis, J. H. Exton, E. M. Ross,Reconstitution of agonist-stimulated phosphatidylinositol 4,5-bisphosphate hydrolysisusing purified m1 muscarinic receptor, Gq/11, and phospholipase C-b1. J. Biol. Chem. 267,8081–8088 (1992).

16. G. G. Tall, A. M. Krumins, A. G. Gilman, Mammalian Ric-8A (synembryn) is a heterotrimericGa protein guanine nucleotide exchange factor. J. Biol. Chem. 278, 8356–8362 (2003).

17. P. Chan, C. J. Thomas, S. R. Sprang, G. G. Tall, Molecular chaperoning function of Ric-8 is tofold nascent heterotrimeric G protein a subunits. Proc. Natl. Acad. Sci. U.S.A. 110,3794–3799 (2013).

18. T. W. Traut, Physiological concentrations of purines and pyrimidines. Mol. Cell. Biochem.140, 1–22 (1994).


19. A. Nishimura, K. Kitano, J. Takasaki, M. Taniguchi, N. Mizuno, K. Tago, T. Hakoshima, H. Itoh,Structural basis for the specific inhibition of heterotrimeric Gq protein by a smallmolecule. Proc. Natl. Acad. Sci. U.S.A. 107, 13666–13671 (2010).

20. R. Schrage, A.-L. Schmitz, E. Gaffal, S. Annala, S. Kehraus, D. Wenzel, K. M. Büllesbach,T. Bald, A. Inoue, Y. Shinjo, S. Galandrin, N. Shridhar, M. Hesse, M. Grundmann, N. Merten,T. H. Charpentier, M. Martz, A. J. Butcher, T. Slodczyk, S. Armando, M. Effern, Y. Namkung,L. Jenkins, V. Horn, A. Stößel, H. Dargatz, D. Tietze, D. Imhof, C. Galés, C. Drewke,C. E. Müller, M. Hölzel, G. Milligan, A. B. Tobin, J. Gomeza, H. G. Dohlman, J. Sondek,T. K. Harden, M. Bouvier, S. A. Laporte, J. Aoki, B. K. Fleischmann, K. Mohr, G. M. König,T. Tüting, E. Kostenis, The experimental power of FR900359 to study Gq-regulatedbiological processes. Nat. Commun. 6, 10156 (2015).

21. J. Takasaki, T. Saito, M. Taniguchi, T. Kawasaki, Y. Moritani, K. Hayashi, M. Kobori, A novelGaq/11-selective inhibitor. J. Biol. Chem. 279, 47438–47445 (2004).

22. S. P. Heximer, N. Watson, M. E. Linder, K. J. Blumer, J. R. Hepler, RGS2/G0S8 is a selectiveinhibitor of Gqa function. Proc. Natl. Acad. Sci. U.S.A. 94, 14389–14393 (1997).

23. V. Villalobos, S. Naik, M. Bruinsma, R. S. Dothager, M.-H. Pan, M. Samrakandi, B. Moss,A. Elhammali, D. Piwnica-Worms, Dual-color click beetle luciferase heteroproteinfragment complementation assays. Chem. Biol. 17, 1018–1029 (2010).

24. S. L. Scherer, M. D. Cain, S. M. Kanai, K. M. Kaltenbronn, K. J. Blumer, Regulation of neuritemorphogenesis by interaction between R7 regulator of G protein signaling complexesand G protein subunit Ga13. J. Biol. Chem. 292, 9906–9918 (2017).

25. T. Flock, C. N. J. Ravarani, D. Sun, A. J. Venkatakrishnan, M. Kayikci, C. G. Tate,D. B. Veprintsev, M. M. Babu, Universal allosteric mechanism for Ga activation by GPCRs.Nature 524, 173–179 (2015).

26. T. G. Cotter, Apoptosis and cancer: The genesis of a research field. Nat. Rev. Cancer 9,501–507 (2009).

27. K. W. Yip, J. C. Reed, Bcl-2 family proteins and cancer. Oncogene 27, 6398–6406 (2008).28. Y. H. Lim, A. Bacchiocchi, J. Qiu, R. Straub, A. Bruckner, L. Bercovitch, D. Narayan;

Yale Center for Mendelian Genomics, J. McNiff, C. Ko, L. Robinson-Bostom, R. Antaya,R. Halaban, K. A. Choate, K. A. Choate, GNA14 somatic mutation causes congenital andsporadic vascular tumors by MAPK activation. Am. J. Hum. Genet. 99, 443–450 (2016).

29. M. D. Onken, J. P. Ehlers, L. A. Worley, J. Makita, Y. Yokota, J. W. Harbour, Functionalgene expression analysis uncovers phenotypic switch in aggressive uveal melanomas.Cancer Res. 66, 4602–4609 (2006).

30. K. A. Matatall, O. A. Agapova, M. D. Onken, L. A. Worley, A. M. Bowcock, J. W. Harbour,BAP1 deficiency causes loss of melanocytic cell identity in uveal melanoma. BMC Cancer13, 371 (2013).

31. M. D. Onken, L. A. Worley, J. P. Ehlers, J. W. Harbour, Gene expression profiling in uvealmelanoma reveals two molecular classes and predicts metastatic death. Cancer Res. 64,7205–7209 (2004).

32. T. S. Mikkelsen, J. Hanna, X. Zhang, M. Ku, M. Wernig, P. Schorderet, B. E. Bernstein,R. Jaenisch, E. S. Lander, A. Meissner, Dissecting direct reprogramming throughintegrative genomic analysis. Nature 454, 49–55 (2008).

33. A. Meissner, T. S. Mikkelsen, H. Gu, M. Wernig, J. Hanna, A. Sivachenko, X. Zhang,B. E. Bernstein, C. Nusbaum, D. B. Jaffe, A. Gnirke, R. Jaenisch, E. S. Lander, Genome-scaleDNA methylation maps of pluripotent and differentiated cells. Nature 454, 766–770 (2008).

34. I. Ben-Porath, M. W. Thomson, V. J. Carey, R. Ge, G. W. Bell, A. Regev, R. A. Weinberg, Anembryonic stem cell–like gene expression signature in poorly differentiated aggressivehuman tumors. Nat. Genet. 40, 499–507 (2008).

35. B. Kaabi, G. Belaaloui, W. Benbrahim, K. Hamizi, M. Sadelaoud, W. Toumi, H. Bounecer,ADRA2A germline gene polymorphism is associated to the severity, but not to the risk, ofbreast cancer. Pathol. Oncol. Res. 22, 357–365 (2016).

36. M. Minoux, S. Holwerda, A. Vitobello, T. Kitazawa, H. Kohler, M. B. Stadler, F. M. Rijli, Genebivalency at Polycomb domains regulates cranial neural crest positional identity.Science 355, eaal2913 (2017).

37. S. Landreville, O. A. Agapova, K. A. Matatall, Z. T. Kneass, M. D. Onken, R. S. Lee,A. M. Bowcock, J. W. Harbour, Histone deacetylase inhibitors induce growth arrest anddifferentiation in uveal melanoma. Clin. Cancer Res. 18, 408–416 (2012).

38. T. E. Hughes, H. Zhang, D. E. Logothetis, C. H. Berlot, Visualization of a functionalGaq-green fluorescent protein fusion in living cells. Association with the plasmamembrane is disrupted by mutational activation and by elimination of palmitoylationsites, but not be activation mediated by receptors or AlF4

–. J. Biol. Chem. 276, 4227–4235(2001).

39. M.A. Booden,D. P. Siderovski, C. J.Der, Leukemia-associatedRhoguanine nucleotide exchangefactor promotes Gaq-coupled activation of RhoA. Mol. Cell. Biol. 22, 4053–4061 (2002).

40. C. R. Evelyn, S. M. Wade, Q. Wang, M. Wu, J. A. Iñiguez-Lluhí, S. D. Merajver, R. R. Neubig,CCG-1423: A small-molecule inhibitor of RhoA transcriptional signaling. Mol. CancerTher. 6, 2249–2260 (2007).

41. W. K. Greentree, M. E. Linder, Purification of recombinant G protein a subunits fromEscherichia coli. Methods Mol. Biol. 237, 3–20 (2004).

42. D. P. McEwen, K. R. Gee, H. C. Kang, R. R. Neubig, Fluorescent BODIPY-GTP analogs: Real-timemeasurement of nucleotide binding to G proteins. Anal. Biochem. 291, 109–117 (2001).

10 of 11

http://www.sciencesignaling.org/cgi/content/full/11/546/eaao6852/DC1http://stke.sciencemag.org/


43. R. B. Delston, K. A. Matatall, Y. Sun, M. D. Onken, J. W. Harbour, p38 phosphorylates Rb onSer567 by a novel, cell cycle-independent mechanism that triggers Rb–Hdm2 interactionand apoptosis. Oncogene 30, 588–599 (2011).

44. A. Subramanian, P. Tamayo, V. K. Mootha, S. Mukherjee, B. L. Ebert, M. A. Gillette,A. Paulovich, S. L. Pomeroy, T. R. Golub, E. S. Lander, J. P. Mesirov, Gene set enrichmentanalysis: A knowledge-based approach for interpreting genome-wide expression profiles.Proc. Natl. Acad. Sci. U.S.A. 102, 15545–15550 (2005).

Acknowledgments: We are grateful to J. Y. Niederkorn (University of Texas SouthwesternMedical Center) for providing UM cell lines and R. P. Mecham (Washington Universityin St. Louis) for supporting this project. Funding: This study was supported by a multi-investigator grant from the Siteman Cancer Center and Pedal the Cause (to K.J.B. and M.D.O.)and grants from the NIH (GM044592 and GM124093 to K.J.B. and GM118171 to J.A.C.).Author contributions: M.D.O. and K.J.B. planned experiments. M.D.O., C.M.M., S.W., K.M.K.,S.M.K., T.D.T., and K.J.B. performed experiments. T.J.B. and K.J.B. purified FR. M.D.O., T.J.B.,


P.K.R., J.A.C., and K.J.B. analyzed data and provided comments. M.D.O., C.M.M., K.M.K., S.M.K.,T.D.T., P.K.R., J.A.C., and K.J.B. wrote the paper. Competing interests: The authors declarethat they have no competing interests. Data and materials availability: The RNA-seqdata used in this manuscript are deposited at National Center for Biotechnology Information’sGene Expression Omnibus (GEO) (GSE103761). All other data needed to evaluate theconclusions in the paper are present in the paper or the Supplementary Materials.

Submitted 15 August 2017Accepted 1 August 2018Published 4 September 201810.1126/scisignal.aao6852

Citation: M. D. Onken, C. M. Makepeace, K. M. Kaltenbronn, S. M. Kanai, T. D. Todd, S. Wang,T. J. Broekelmann, P. K. Rao, J. A. Cooper, K. J. Blumer, Targeting nucleotide exchange to inhibitconstitutively active G protein a subunits in cancer cells. Sci. Signal. 11, eaao6852 (2018).

11 of 11


ag.org/D

ownloaded from


cells subunits in cancerαTargeting nucleotide exchange to inhibit constitutively active G protein

Broekelmann, Prabakar Kumar Rao, John A. Cooper and Kendall J. BlumerMichael D. Onken, Carol M. Makepeace, Kevin M. Kaltenbronn, Stanley M. Kanai, Tyson D. Todd, Shiqi Wang, Thomas J.

DOI: 10.1126/scisignal.aao6852 (546), eaao6852.11Sci. Signal.

treatment for patients.and cell death. Targeting this compound, if safe, or synthetic derivatives to the uveal tumor tissue may be an effective

signaling in UM cells induced redifferentiationqα, thereby trapping it in inactive heterotrimers. The loss of Gqαactive G-mutant UM cells in culture. FR900359 allosterically inhibited the guanine nucleotide exchange activity of constitutivelyα

. found that the plant-derived compound FR900359 blocked the growth of Get al(UM), an aggressive eye cancer. Onken subunits cause various diseases, including some forms of uveal melanomaαActivating mutations in G protein

αTargeting mutant G

ARTICLE TOOLS http://stke.sciencemag.org/content/11/546/eaao6852

MATERIALSSUPPLEMENTARY http://stke.sciencemag.org/content/suppl/2018/08/30/11.546.eaao6852.DC1

CONTENTRELATED

http://stke.sciencemag.org/content/sigtrans/13/617/eaax8620.fullhttp://stke.sciencemag.org/content/sigtrans/12/573/eaau5948.fullhttp://stke.sciencemag.org/content/sigtrans/12/569/eaav2449.fullhttp://science.sciencemag.org/content/sci/360/6395/eaao4927.fullhttp://stke.sciencemag.org/content/sigtrans/11/532/eaap8113.fullhttp://stke.sciencemag.org/content/sigtrans/11/534/eaan3677.full

REFERENCES

http://stke.sciencemag.org/content/11/546/eaao6852#BIBLThis article cites 44 articles, 20 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.Science SignalingYork Avenue NW, Washington, DC 20005. The title (ISSN 1937-9145) is published by the American Association for the Advancement of Science, 1200 NewScience Signaling

Science. No claim to original U.S. Government WorksCopyright © 2018 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of


ag.org/D

ownloaded from

http://stke.sciencemag.org/content/11/546/eaao6852http://stke.sciencemag.org/content/suppl/2018/08/30/11.546.eaao6852.DC1http://stke.sciencemag.org/content/sigtrans/11/534/eaan3677.fullhttp://stke.sciencemag.org/content/sigtrans/11/532/eaap8113.fullhttp://science.sciencemag.org/content/sci/360/6395/eaao4927.fullhttp://stke.sciencemag.org/content/sigtrans/12/569/eaav2449.fullhttp://stke.sciencemag.org/content/sigtrans/12/573/eaau5948.fullhttp://stke.sciencemag.org/content/sigtrans/13/617/eaax8620.fullhttp://stke.sciencemag.org/content/11/546/eaao6852#BIBLhttp://www.sciencemag.org/help/reprints-and-permissionshttp://www.sciencemag.org/about/terms-servicehttp://stke.sciencemag.org/

Documents

Targeting nucleotide exchange to inhibit constitutively active G … · tagged [FLAG and StrepII (FS)] wild-type or constitutively active G a q. The data indicate that FR was able