Racgap1 Is a Novel Downstream Effector of E2F7-Dependent

Published OnlineFirst May 27, 2015; DOI: 10.1158/1535-7163.MCT-15-0076

Cancer Biology and Signal Transduction Molecular Cancer Therapeutics RacGAP1 Is a Novel Downstream Effector of E2F7-Dependent Resistance to Doxorubicin and Is Prognostic for Overall Survival in Squamous Cell Carcinoma Mehlika Hazar-Rethinam1, Lilia Merida de Long1, Orla M. Gannon1, Samuel Boros2, Ana Cristina Vargas2, Marcin Dzienis3, Pamela Mukhopadhyay1, Natalia Saenz-Ponce1, Daniel D.E. Dantzic1, Fiona Simpson1, and Nicholas A. Saunders1

Abstract

We have previously shown that E2F7 contributes to drug E2F7-dependent upregulation of RacGAP1 in doxorubicin- resistance in head and neck squamous cell carcinoma (HNSCC) insensitive SCC25 cells. Extending this, we found that selective cells. Considering that dysregulation of responses to chemother- upregulation of RacGAP1 induced doxorubicin resistance in apy-induced cytotoxicity is one of the major reasons for treatment previously sensitive KJDSV40. Similarly, stable knockdown of failure in HNSCC, identifying the downstream effectors that RacGAP1 in insensitive SCC25 cells induced sensitivity to regulate E2F7-dependent sensitivity to chemotherapeutic agents doxorubicin in vitro and in vivo. RacGAP1 expression was may have direct clinical impact. We used transcriptomic profiling validated in a TMA, and we showed that HNSCCs that over- to identify candidate pathways that contribute to E2F7-dependent express RacGAP1 are associated with a poorer patient overall resistance to doxorubicin. We then manipulated the expression of survival. Furthermore, E2F7-induced doxorubicin resistance the candidate pathway using overexpression and knockdown in in was mediated via RacGAP1-dependent activation of AKT. Final- vitro and in vivo models of SCC to demonstrate causality. In ly, we show that SCC cells deficient in RacGAP1 grow slower addition, we examined the expression of E2F7 and RacGAP1 in and are sensitized to the cytotoxic actions of doxorubicin in a custom tissue microarray (TMA) generated from HNSCC patient vivo.Thesefindings identify RacGAP1 overexpression as a novel samples. Transcriptomic profiling identified RacGAP1 as a poten- prognostic marker of survival and a potential target to sensitize tial mediator of E2F7-dependent drug resistance. We validated SCC to doxorubicin. Mol Cancer Ther; 14(8); 1939–50. 2015 AACR.

Introduction terized by the development of drug resistance. Thus, there is a need to identify new therapeutic strategies that can bypass the emer- Cutaneous squamous cell carcinomas (CSCC) and head and gence of a drug-resistant phenotype. neck SCC (HNSCC) are among the most common malignancies The E2F transcription factor complex comprises a family of afﬂicting man (1, 2). Current treatment options for advanced SCC activating (E2F1, 2, 3a) or repressive/inhibitory (E2F3b, 4, 5, 6, 7, include adjuvant chemotherapy with platinum-based drugs, such or 8) E2Fs that regulate key cellular functions, such as transcrip- as taxanes, 5-Fluorouracil, or therapeutic antibodies against EGFR tion, differentiation, and apoptosis. In the context of keratino- (3, 4). However, the response is generally transient and charac- cytes (KC), the E2F transcription factor family has been shown to control (i) proliferation, (ii) differentiation, (iii) stress responses, and (iv) apoptosis (5–10). Consistent with their roles in KCs, 1 Epithelial Pathobiology Group, University of Queensland Diamantina dysregulation of E2F is a common occurrence in SCC (11, 12), and Institute, Princess Alexandra Hospital,Translational Research Institute, Woolloongabba, Queensland, Australia. 2Department of Pathology, overexpression of E2Fs, such as E2F1 and E2F7, occurs in the Princess Alexandra Hospital, Woolloongabba, Queensland, Australia. majority of CSCCs and HNSCCs (12–15). E2F1 and E2F7 are 3 Department of Medical Oncology, Princess Alexandra Hospital,Wool- known to have opposing actions in the regulation of proliferation loongabba, Queensland, Australia. (5), differentiation (14), and apoptosis (9, 14). For example, Note: Supplementary data for this article are available at Molecular Cancer recent reports have shown that treatment of wild-type cells with Therapeutics Online (http://mct.aacrjournals.org/). DNA-damaging agents, such as doxorubicin or etoposide, induces Current address for P. Mukhopadhyay: The QIMR Berghofer Medical Research E2F7 protein levels and subsequent inhibition of the E2F1-medi- Institute, Brisbane, Queensland, Australia. ated DNA damage response (9, 10). In the context of KCs, E2F7 Corresponding Author: Nicholas A. Saunders, The University of Queensland was shown to causally modify responses to conventional che- Diamantina Institute, Translational Research Institute, 37 Kent St, Woolloon- motherapeutics (15) and UV responses in vitro (14). Thus, sen- gabba, Queensland 4102, Australia. Phone: 61-7-34437098; Fax: 61-7-34436966; sitivity to common cytotoxic agents and stimuli appear to be E-mail: [email protected] regulated by the relative ratio of E2F1 to E2F7 in the tissue. Given doi: 10.1158/1535-7163.MCT-15-0076 that both E2F1 and E2F7 are known to be overexpressed in SCC 2015 American Association for Cancer Research. (14, 15), it is reasonable to speculate that this may also

www.aacrjournals.org 1939

Hazar-Rethinam et al.

contribute to drug resistance in SCC. In this regard, we recently AGTCATGG, RacGAP1 Reverse: GCTCAAACAGATTCCGCACA; showed that the sphingosine kinase 1 (Sphk1)geneisadirect Sphk1 Forward: AAGACCTCCTGACCAACTGC, Sphk1 Reverse: target of E2F7 in SCC (15). E2F7-dependent overexpression of GGCTGAGCACAGAGAAGAGG. Sphk1 in SCC induces increased production of the antiapop- totic phospholipid, sphingosine-1-phosphate (S1P), which in Gene expression analysis turn invokes anthracycline resistance via activation of the PI3K/ Each sample was analyzed in duplicate. Complementary RNA AKT pathway (15). Thus, E2F dysregulation in SCC induced the was generated from samples using the Illumina TotalPrep RNA activation of a Sphk1/S1P-dependent drug-resistant phenotype Amplification Kit and hybridized with Illumina HumanHT-12 v4 (15). Identification of this novel pathway was noteworthy Expression BeadChips (Illumina) as the per manufacturer's pro- because anthracyclines such as doxorubicin are not in clinical tocol. Expression data from the microarrays were analyzed as use for the treatment of SCC, and thus the ability to sensitize previously described (19). Only genes with a fold change of 1 (in SCCs to an existing class of chemotherapeutics would be of either direction) or greater and a B value of greater than 3 clinical value. However, the activation of the Sphk1 pathway (exceeding the 95% probability of differential expression) were was clearly only part of the explanation for the anthracycline considered to be differentially expressed and further analyzed. resistance observed in SCC. Thus, other pathways that control Differentially expressed probe sets were analyzed as pair-wise drug resistance in SCC were likely to exist. contrasts. Microarray data have been uploaded to Gene Expres- In the present study, we used transcriptomic profiling to iden- sion Omnibus under the reference: GSE58074. tify a novel druggable E2F7/RacGAP1/AKT pathway that selec- tively induces anthracycline resistance in SCC. Colony-forming assay Known number of SCC cells was plated and allowed to grow for 15 days. Plates were fixed and stained with Coomassie Blue and Materials and Methods counted as previously described (20). Colony-forming efficiency Chemicals and viability assays was expressed as the total number of colonies/total number of The following drugs were purchased: AZA1 (Millipore), doxo- cells plated 100. rubicin (Sigma Aldrich), NSC23766 (Abcam), S1P (Cayman Chemicals), Y-27632 (Sigma Aldrich), and ZVAD-fmk (Alexis DNA synthesis Biochemicals). BGT26 was provided by Novartis, and stocks of DNA synthesis was measured using a colorimetric ELISA 5- BGT226 were prepared as described (16). ZVAD-fmk was added bromo-2-deoxyuridine (BrdUrd) incorporation assay (Roche 30 minutes before other treatments. Cell viability was performed Diagnostics) in accordance with the manufacturer's instructions. by trypan blue exclusion or using Cell Titer 96 Aqueous One Solution Cell Proliferation Assay (Promega). Chromatin immunoprecipitation Chromatin immunoprecipitation (ChIP) was performed using Tissue culture, adenovirus infection, and transfection the SimpleChIP Enzymatic Chromatin IP Kit (Magnetic Beads; Murine epidermal keratinocytes (MEK) and human epider- Cell Signaling Technology) in accordance with the manufacturer's mal keratinocytes (HEK) were isolated and cultured as instructions. ChIP enrichment was determined by conducting described (17, 18). E2F7 KO KCs were generated by ready- qRT-PCR as described above. The primers used were as follows: to-use adenovirus harboring Cre recombinase infection of 50-GAAGTGAGTAGTGGGGGTGC-30 (RacGAP1 Forward); 50- murine epidermal keratinocytes as per the manufacturer's TCCATCTTTCACACGAACACTCT-30 (RacGAP1 Reverse). recommendations (MOI of 50; Vector Biolabs). Detroit562 and SCC25 cells were obtained from the American Type Culture Immunoblot Collection, and cultures were not passaged for longer than 6 Immunoblotting was carried out according to previously pub- months. SCC15 was a kind gift from Dr. Elizabeth Musgrove lished procedures (21) using the following primary antibodies: (Garvan Institute, New South Wales, Australia) and was verified Anti-RacGAP1 (EPR9018; 1:2,000; Abcam), Anti-Sphk1 (1:1,000; by short tandem repeat (STR) genotyping (12). KJDSV40 cells Sigma Aldrich), PARP (1:1,000; Cell Signaling Technology), were maintained as described previously (12) and were verified phospho-Akt (Ser473; D9E; XP; 1:2,000; Cell Signaling Technol- by STR genotyping. STR-verified cells were not passaged for ogy), Akt (1:2,000; Cell Signaling Technology), phospho-p44/42 longer than 6 months after verification. All cell lines used were MAPK (Erk1/2; Thr202/Tyr204; E10; 1:2,000; Cell Signaling regularly tested and validated to be Mycoplasma free. Control Technology), ERK 1 (C-16; 1:2,000; Santa Cruz), and b-actin and overexpression plasmids used for manipulating E2F7 and (1:10,000; Sigma Aldrich). the siRNA used for targeting E2F7 have been described previously (9, 14). SureSilencing shRNA plasmids directed against Immunohistochemistry RacGAP1 or Sphk1 were purchased from SuperArray Bioscience Immunohistochemistry was carried out according to previous- Corp (SA Biosciences). A RacGAP1 expression (TrueORF Gold ly published procedures (21, 22) using the following primary Clones) and control plasmids were purchased from OriGene antibodies: Anti-PCNA (1:3,000; Sigma Aldrich), Anti-RacGAP1 Technologies (Australian Biosearch). (EPR9018; 1:100; Abcam), cleaved caspase-3 (Asp 175; 1:50; Cell Signaling Technology), and phospho-Akt (Ser473; D9E; XP; 1:50; RNA isolation and quantitative RT-PCR Cell Signaling Technology). Secondary antibody was Starr Trek Total RNA was isolated, cDNA was prepared, and qRT-PCR was Universal HRP Detection System (Biocare Medical) followed by performed as described (15). Primer sequences were E2F7 For- colorimetric immunohistochemical staining with Cardassian ward: GTCAGCCCTCACTAAACCTAAG, E2F7 Reverse: TGCGTT- DAB Chromogen as per the manufacturer's instructions (Biocare GGATGCTCTTGG; RacGAP1 Forward: GACGTTGAATAGGATG- Medical).

1940 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

RacGAP1 Is Overexpressed in Drug-Resistant SCC

Tissue microarrays silenced with siRNA (Fig. 2A). We then used these two lists to Generation and composition of the patient tissue microarrays identify those transcripts (referred to as List A; Supplementary (TMA) have been previously described (15). Immunohistochem- Table S1) that displayed E2F7-dependent expression between the istry was conducted using a Dako EnVision þ System-HRP (DAB) SCC25 cell lines (Fig. 2A). We also generated an additional list of kit in accordance with the manufacturer's instructions (DAKO). transcripts for SCC25 cells or SCC25 cells in which E2F7 is Sections were incubated with Anti-E2F7 (1:250; Abcam) and Anti- silenced by siRNAs that have been treated with 1 mmol/L of RacGAP1 (EPR9018; 1:100; Abcam) antibodies. Staining inten- doxorubicin. The transcripts that were found to be E2F7-depen- sity was evaluated by two pathologists in a blinded fashion using a dant in the context of doxorubicin-treated SCC25 cells were then modified quickscore method as described (23). referred to as List B (Fig. 2A; Supplementary Table S1). By combining Lists A and B, we identified RacGAP1 as the most Determination of RhoA and Rac1 activity differentially overexpressed genes with a B value of 17 (Supple- RhoA and Rac1 activities were measured with RhoA/Rac1/ mentary Table S1). Cdc42 activation assay combo biochem kit (Cytoskeleton) in RacGAP1 (also known as MgcRacGAP and CYK4) is an evolu- accordance with the manufacturer's instructions. tionarily conserved GTPase activating protein (GAP) that displays activity toward the Rho family of GTPases. The Rho family of Animal studies GTPases is a subfamily of the Ras superfamily and consists of All animal experiments were approved by the Institutional small signaling G proteins: Rho (A, B, and C isoforms), Rac (1,2,3 Animal Ethics Committee. In vivo tumor studies used 6-week-old isoforms and RhoG), and Cdc42 (Cdc42, Tc10, TCL, Chp/Wrch-2, female nonobese diabetic/severe combined immunodeficient and Wrch-1; ref. 24), which function as molecular switches mice. Mice were injected subcutaneously on the flank with 2 between a GTP-loaded "ON" and a GDP-loaded "OFF" state 106 cells. Groups of 6 mice received treatments (intraperitoneal (24). Thus, RacGAP1 has the potential to regulate a diverse array injections twice/week) when tumors were around 4 mm in of cellular functions through its central role as a regulator of the diameter. Animal weight and tumor growth were monitored activation state of the Rho family of GTPases. In particular, for a period of up to 3 weeks, and animals were sacrificed when RacGAP1 is known to play important roles in the completion of tumors reached 10 mm in diameter. cytokinesis (25), cell transformation, motility, migration, and metastasis (26–29). RacGAP1 is also involved in IL6-induced Statistical analysis macrophage differentiation (30) and nuclear transport of Statistical significance was calculated by a Student t test with a STAT3/5 transcription factors (31). However, a role for RacGAP1 95% confidence level using GraphPad Prism v5 (GraphPad in SCC or doxorubicin sensitivity has not been shown previously. software). Quantitative RT-PCR and Western blotting were used to confirm that RacGAP1 was more highly expressed in SCC25 (doxorubicin insensitive) cells than in KJDSV40 (doxorubicin sensitive) cells Results (Fig. 2B and C). Similarly, we showed that knockdown of E2F7 by RacGAP1 is a novel downstream effector of E2F7 siRNA in SCC25 cells caused a reduction in RacGAP1 mRNA (Fig. We generated E2f7 knockdown (KD) murine KCs via adeno- 2D) and protein level (Fig. 2E). Conversely, overexpression of E2F7 virus-mediated Cre deletion of floxed sequences in primary in KJDSV40 cells resulted in elevated levels of RacGAP1 transcript KCs isolated from E2f7-floxed mice (15). KD KCs were treated (Fig. 2F) as well as RacGAP1 protein (Fig. 2G). These data suggest for 48 hours with increasing concentrations of doxorubicin that RacGAP1 is a downstream target of E2F7 in SCC cells. (0–1 mmol/L), etoposide (0–100 mmol/L), and cisplatin (0–20 Supporting this, ChIP analysis of E2F7 binding showed that mmol/L). Figures 1A–C show that E2F7 deficiency sensitizes KCs E2F7 could bind the RacGAP1 promoter, suggesting that RacGAP1 to doxorubicin, modestly to cisplatin, but not at all to etoposide. is a direct transcriptional target of E2F7 (Fig. 2H). This is the first These data suggest that E2F7-mediated doxorubicin resistance is report to show that RacGAP1 is a downstream effector of E2F7. not attributable to topoisomerase inhibition because etoposide sensitivity was not modified by E2F7. Moreover, pan-caspase RacGAP1expression is elevated in SCCs inhibition significantly protected E2F7-deficient cells from doxo- We examined RacGAP1 expression levels by immunohis- rubicin-induced cytotoxicity (Fig. 1D), indicating that apoptotic tochemistry using a TMA consisting of 35 paired normal, primary pathways are being activated. tumor, and matched metastasis from HNSCC patients treated at We undertook a screen of doxorubicin sensitivity in HEKs and 4 the Princess Alexandra Hospital (PAH). The TMAs were stained for SCC cell lines. The KJDSV40 cell line exhibited the highest sen- E2F7 and RacGAP1 protein expression and scored blinded by two sitivity to doxorubicin (IC50 of 0.082 mmol/L; Fig. 1E), whereas pathologists. All matched adjacent "normal" epithelia demon- SCC25 cells displayed the least sensitivity (IC50 of 0.55 mmol/L; strated either negative or weak staining for RacGAP1, which was Fig. 1E) and HEKs displayed intermediate sensitivity (IC50 of predominantly nuclear in location (Fig. 3A). Conversely, mod- 0.29 mmol/L; Fig. 1E). We have previously shown that the insen- erate to high levels of RacGAP1 expression were consistently sitive SCC25 cell line expresses high levels of E2F7, whereas the recorded for the primary tumor (Fig. 3B) and its matched lymph sensitive KJDSV40 cell line expresses low levels of E2F7 (15). Based node metastasis (Fig. 3C). The tumor epithelial cells showed on these data, we selected the sensitive KJDSV40 cell line and the nuclear and cytosolic expression for RacGAP1. RacGAP1 was insensitive SCC25 cell line for transcriptomic profiling. significantly overexpressed in 73% of primary and metastatic Specifically, we generated a list of genes that were poorly human SCCs compared with matched adjacent normal tissue expressed in KJDSV40 cells and highly expressed in SCC25 (P < 0.0001; Fig. 3D). In addition, our analyses showed that E2F7 (Fig. 2A). We also generated a second list of genes that were expression is significantly upregulated in HNSCC compared with differentially regulated in SCC25 cells in which E2F7 had been matched adjacent normal tissue (P < 0.0001; Fig. 3D). The

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1941

Hazar-Rethinam et al.

Floxed MEK+ZVAD-fmk+doxoru ADE2F7 Ad-Cre-GFP+ZVAD-fmK+ 1.0 120 E2F7 Ad-Cre-GFP+doxorubicin Floxed MEK Floxed MEK+doxorubicin 100 E2F7 Ad-Cre-GFP 80 Figure 1. Cytotoxic responses to doxorubicin 0.5 60 in E2F7-deﬁcient murine KCs. E2F7- 40 ﬂoxed KCs were incubated with

Cell viability (% control) 20 Absorbance 490 nm (squares) or without (circles) Ad-Cre- 0.0 0 GFP for 48 hours and then incubated 0 0 with varying doses of doxorubicin (A), 100200 300 500 0.1 0.2 0.3 0.5 1.0 1,000 Doxorubicin (μmol/L) etoposide (B), or cisplatin (C). Doxorubicin (nmol/L) Viability (absorbance 490 nm) was assessed 48 hours after treatment B E and is expressed as the mean SEM obtained from triplicate HEK 1.5 150 determinations of three independent Detroit562 ns Floxed MEK experiments. D, Ad-Cre-GFP– KJDSV40 ns E2F7 Ad-Cre-GFP uninfected E2F7-ﬂoxed or Ad-Cre- 1.0 100 SCC15 GFP–infected E2F7-deﬁcient SCC25 proliferative KCs were treated with 1 mmol/L doxorubicin in the presence or

0.5 Viability (%) 50 absence of ZVAD-fmk and viability

Absorbance 490 nm determined 48 hours later. Viability 0.0 0 is plotted as a percentage of 0 –14 –12 –10 –8 –6 –4 doxorubicin-treated uninfected E2F7- 5,00010,000 log [Doxurubicin] (mol/L) 10,00020,00030,00050,000100,000 ﬂoxed MEKs and represents the mean Etoposide (nmol/L) SEM obtained from triplicate determinations of three independent experiments. E, HEK, Detroit562, C KJDSV40, SCC15, and SCC25 cells 150 1.5 HEK were treated with doxorubicin for Floxed MEK KJDSV40 48 hours and viability plotted as E2F7 Ad-Cre-GFP 100 SCC25 percentage of untreated cells. Inset, 1.0 estimated IC50 values for doxorubicin in HEK, KJDSV40, and SCC25 cells

Viability (%) 50 determined by nonlinear regression 0.5 analysis in Prism. ns, not signiﬁcant; Absorbance 490 nm , P < 0.05; , P < 0.01; , P < 0.001; 0 P < 0.0 –14 –12 –10 –8 –6 –4 , 0.0001. 0 5,000 10,000 15,000 20,000 log [Doxurubicin] (mol/L) Cisplatin (nmol/L) HEK KJDSV40 SCC25 IC (μmol/L) 50 0.29 0.082 0.55

Kaplan–Meier analysis revealed an inverse relationship between of RacGAP1 in sensitive KJDSV40 cells resulted in increases in RacGAP1 expression levels and progression-free survival (PFS) of RacGAP1 protein level (Fig. 4F), and reduced sensitivity to doxo- HNSCC patients studied over a period of 42 months whose rubicin compared with vector control (Fig. 4G). Combined, these samples were arrayed on the TMA (Fig. 3E). These data show, data indicate that RacGAP1 can promote proliferation and inhibit for the first time, that RacGAP1 is overexpressed in HNSCC and is doxorubicin-induced cell death in SCCs. associated with a poorer PFS. RacGAP1 differentially regulates the GTP-loaded state of RhoA RacGAP1 expression/activity determines sensitivity to and Rac1 in SCC cells doxorubicin We examined whether the overexpression of RacGAP1 in the We examined the effect of shRNA-mediated knockdown or SCC cell lines was reflected in alterations of the GTP loading RacGAP1 overexpression on sensitivity to doxorubicin. RacGAP1 (activation status) of the model targets RhoA and Rac1. Specif- gene silencing was achieved using 4 different constructs of which ically, RhoA GTP loading was constitutively higher in KJDSV40 shRNA.3 displayed the greatest knockdown in RacGAP1 protein cells, which express very low levels of E2F7 and RacGAP1, com- level (Fig. 4A). For subsequent experiments, the shRNA complex pared with SCC25 cells which express high levels of E2F7 and shRNA.3 was employed. Consistent with previous reports (32), RacGAP1 (Fig. 5A). In contrast, GTP loading of Rac1 was higher in RacGAP1 shRNA-transfected SCC25 cells displayed significant SCC25 cells when compared with KJDSV40 cells (Fig. 5B), and the reductions in proliferation (Fig. 4B), colony-forming efficiency GTP loading of Rac1 was significantly reduced in SCC25 cells in vitro (Fig. 4C), and induced a modest increase in cleaved PARP1 following RacGAP1 knockdown (Fig. 5B). Finally, knockdown of (Fig. 4D) compared with control vector–transfected cells. Finally, RacGAP1 in SCC25 cells resulted in an increase in the GTP loading silencing RacGAP1 significantly enhanced the sensitivity of of RhoA (Fig. 5A). These data suggest that E2F/RacGAP1-depen- SCC25 cells to doxorubicin (Fig. 4E). Conversely, overexpression dent resistance to doxorubicin may be mediated by activated Rac1

1942 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

RacGAP1 Is Overexpressed in Drug-Resistant SCC

KJDSV40 vs. SCC25 SCC25 vs. SCC25 vs. SCC25 vs. SCC25.E2F7siRNA SCC25.E2F7siRNA List B SCC25.E2F7siRNA+dxr List A 4 Microarray DOWN 3 Microarray UP Microarray UP Microarray UP

671 215 215 199

Figure 2. E2F7-modulated transcripts in modulating RacGAP1 is a downstream effector of E2F7-modulated transcripts sensitivity to doxorubicin E2F7. A, summary of the strategy used to identify E2F7-dependent transcripts that associate with doxorubicin resistance in SCC cells. B and C, RacGAP1 mRNA and protein Downstream effectors of E2F7-dependent levels were determined in KJDSV40 cytotoxic sensitivity and SCC25 cells by qRT-PCR and immunoblotting, respectively. D and E, the expression of RacGAP1 BCDE transcripts and protein was determined by qRT-PCR and 2.0 2.0 immunoblotting, respectively, of SCC25Control E2F7siRNA siRNA extracts derived from SCC25 and HEK SCC25 KJDSV40 1.5 SCC25 cells in which E2F7 was RacGAP1 1.5 RacGAP1 silenced with siRNA for 48 hours. 1.0 F and G, RNA was extracted from β-Actin 1.0 β-Actin KJDSV40 and KJDSV40 cells in which E2F7 was overexpressed from an 0.5 0.5 expression plasmid. The expression of RacGAP1 transcripts and protein was mRNA expression Relative mRNA expression Relative 0.0 determined 48 hours after 0.0 transfection by qRT-PCR and SCC25 immunoblotting, respectively. H, to KJDSV40 SCC25 quantitate E2F7 binding to the RacGAP1 promoter, ChIPs were performed using an E2F7 antibody or SCC25_E2F7.siRNA nonimmune IgG as control in HEK, FGH KJDSV40, and SCC25 cells. Each ChIP and qRT-PCR was repeated 2 or 3 RacGAP1 promoter 0.8 times, respectively. Data, mean SEM 0.04 of duplicate determinants normalized pcDNAE2F7 o/exp for expression of the housekeeping 0.6 RacGAP1 HEK n ¼ b 0.03 gene TBP for B, D, and F; 3. -Actin KJDSV40 is provided as a loading control for C, β 0.4 -Actin SCC25 E, and G. Western blot ﬁgure is 0.02 representative of three independent P < 0.2 experiments. , 0.01; 0.01 P <

, 0.0001. mRNA expression Relative Signal relative to input Signal relative 0.0 0.00

IgG ChIP Antibodies E2F7 KJDSV40

KJDSV40_E2F7 o/exp or inactivated RhoA. To determine which, if any, of these possi- preferred substrate for RacGAP1 in SCC cells. This is reﬂected by bilities may apply, we coincubated SCC25 cells with doxorubicin the high level of RhoA-GTP loading compared with Rac1-GTP þ/– an established selective Rho/ROCK1 inhibitor (10 mmol/L Y- loading as well as the increase in GTP loading observed following 27632; ref. 33), an established Rac1/Cdc42 inhibitor (10 mmol/L RacGAP1 knockdown in the SCC25 cells. Secondly, although AZA1; ref. 34), or a Rac1-selective inhibitor (25 mmol/L Rac1-GTP loading behavior is not indicative of it being a preferred NSC23766; ref. 35) for 48 hours and established viability (Fig. substrate of RacGAP1, it is clear that alterations in RacGAP1 5C). These experiments showed that doxorubicin resistance was activity modify Rac1-GTP loading. Thirdly, use of a Rac1-selective unaltered following incubation with a Rho/ROCK1 inhibitor and inhibitor phenocopies the inhibition of doxorubicin resistance was moderately reduced following incubation with the Rac1/ observed with RacGAP1 knockdown (Fig. 4E). Finally, the pref- cdc42 inhibitor (Fig. 5C). In contrast, treatment of SCC25 cells erence for RhoA by RacGAP1, in SCC cells, is consistent with a with the Rac1-selective inhibitor signiﬁcantly increased sensitivity previous report showing that the conventional preference for Rac1 of SCC25 cells to doxorubicin (Fig. 5C). These results indicate a can be switched to RhoA following phosphorylation of the Serine number of important points. Firstly, RhoA appears to be a 387 site of RacGAP1 by Aurora B kinase (36, 37).

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1943

Hazar-Rethinam et al.

ABC

D E2F7 RacGAP1 Figure 3. ns Expression of RacGAP1 in HNSCC 250 ns and adjacent normal tissue and 200 its association with PFS. Representative images of adjacent 150 normal tissue (A) and HNSCC specimens stained for RacGAP1 (B 100 Normal and C). D, quantitation of E2F7 and RacGAP1 staining intensities in Primary tumor

IHC quickscore 50 matched samples of primary tumor, Metastatic tumor normal squamous epithelium, and 0 lymph node metastases (n ¼ 35). Tissue sections were scored using a fi Normal Normal modi ed quickscore method to determine the percentage of cells Primary tumor Primary tumor stained (0%–100%) and the intensity Metastatic tumor Metastatic tumor of staining (1þ to 3þ). E, Kaplan– Meier analysis of PFS stratified by E 100 High RacGAP1 expression in the HNSCC Low patient cohort. ns, not significant, , P < 0.0001.

50 Percent survival Percent

0 010203040 50 Months RacGAP1 Range Median Mean Primary tumor 0–200 171 158

RacGAP1 modulates doxorubicin sensitivity via downstream AKT signaling and induce apoptosis in SCC cell lines (16). We activation of the PI3K/AKT pathway compared the sensitivity of SCC25 cells with BGT226 in SCC25 There is an existing literature showing that the PI3K/AKT cells or SCC25 cells in which RacGAP1 was knocked down. Figure pathway is an important component of RacGAP1 signaling 5G indicates that knockdown of RacGAP1 is able to reduce SCC25 (38). However, whether PI3K/AKT signaling lies upstream or cell viability to 70% that of control cells. Similarly, inhibition of downstream of the Rho family of GTPases remains less clear and PI3K activity using a dose of BGT226 known to induce maximal appears to be context-specific (38). Dysregulation of the PI3K/ inhibition (16) reduced SCC25 cell viability to approximately AKT pathway is a common event in HNSCC, which can be 50% (Fig. 5G). Finally, exposure of RacGAP1-deficient SCC25 attributed to multiple factors, such as mutations, amplifications, cells to BGT226 resulted in a further decrease in viability to below and signal-induced activation of the pathway (39). For example, 20% (Fig. 5G). These data indicate that RacGAP1 participates in we recently showed that E2F7 overexpression or knockdown AKT-dependent and AKT-independent events. caused an increase and decrease in p-AKT levels in SCC cells We recently reported that E2F7 is able to directly activate the respectively (15). Since RacGAP1 is a downstream effector of E2F7 Sphk1/S1P axis in SCC cells, which induces doxorubicin resis- in SCC cells, we examined whether RacGAP1 could modify the tance (15). It is also interesting to note that both the E2F7/ PI3K/AKT signaling pathway in SCC cells. In the first instance, we RacGAP1 pathway identified in this study and the E2F7/Sphk1/ noted that knockdown of RacGAP1 in SCC25 cells had no impact S1P pathway (15) induced doxorubicin resistance and converged on the activation status of the ERK pathway (Fig. 5D). In contrast, on the AKT pathway. Therefore, we examined whether the Sphk1 RacGAP1 knockdown in SCC25 cells significantly reduced the and RacGAP1 pathways may interact with one another. Figure 5H level of p-AKT (Fig. 5E), whereas RacGAP1 overexpression in shows that knockdown of Sphk1 can induce loss of RacGAP1 KJDSV40 cells increased p-AKT levels (Fig. 5F). We had previously mRNA, whereas knockdown of RacGAP1 induces loss of Sphk1 shown that the PI3K/mTOR inhibitor, BGT226, was able to ablate mRNA expression. Although the mechanism controlling this

1944 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

RacGAP1 Is Overexpressed in Drug-Resistant SCC

ABSCC25 C SCC25 Figure 4. 150 150 The sensitivity to doxorubicin is mediated via an E2F7/RacGAP1 axis in SCC. A, SCC25 cells were transfected 100 100 with 4 different constructs coding for RacGAP1shRNA.1RacGAP1shRNA.2RacGAP1shRNA.3RacGAP1shRNA.4Control shRNA shRNAs directed against RacGAP1. RacGAP1 After 48 hours, RacGAP1 protein expression was determined by 50 CFE (%) 50 β immunoblotting. b-Actin is provided -Actin as a loading control. SCC25 cells were transfected with the BrdUrd incorporation (%) 0 0 RacGAP1shRNA.3 or a scrambled shRNA construct. After 48 hours, BrdUrd incorporation (B) and CFE (C) were determined. Data are expressed as a percentage of that observed for Vector control Vector control control shRNA. D, cleavage of PARP RacGAP1 shRNA RacGAP1 shRNA was determined by immunoblotting extracts of RacGAP1shRNA and DE control shRNA–transfected SCC25 SCC25 cells. E, RacGAP1shRNA and control Control shRNA – shRNA transfected SCC25 cells were 120 treated with doxorubicin (1 mmol/L) RacGAP1 shRNA for 48 hours and viability estimated by Control shRNARacGAP1 shRNA 100 trypan blue exclusion. Viability is PARP plotted as percentage control 80 (untreated). F, KJDSV40 cells were Cleaved PARP transfected with RacGAP1 expression 60 plasmid or a noncoding empty vector. 40 After 48 hours, RacGAP1 protein expression was determined by 20 immunoblotting. b-Actin is a loading control. G, RacGAP1 expression Cell count (% control shRNA) 0 plasmid or empty vector–transfected Doxorubicin (1 μmol/L) –+ KJDSV40 cells were treated with doxorubicin (1 mmol/L) for 48 hours and viability estimated by trypan blue exclusion. Viability was FGKJDSV40 expressed as percentage control (untreated). Western blot figures are representative of three independent Empty vector experiments. All quantitative data 120 RacGAP1 o/exp presented as mean SEM. For Empty vectorRacGAP1 o/exp viability, values represent triplicate 100 determinations of three independent RacGAP1 experiments; for CFE, expression 80 values represent duplicate β determinations from at least three -Actin 60 independent experiments; for BrdUrd, 40 values represent triplicate determinations from 4 independent 20 experiments. , P < 0.01; , P < 0.001; , P < 0.0001. Cell count (% empty vector) 0 Doxorubicin (1 μmol/L) – + feedback is unknown, it is clear that targeted inhibition of either When tumors were around 4 mm in diameter, mice were ran- the RacGAP1 pathway or the Sphk1 pathway is likely to impact domized into four groups and treated with vehicle, dimethyl one another. To illustrate this point, knockdown of RacGAP1 or sulfoxide (DMSO), or 0.5 mg/kg doxorubicin by i.p. injections Sphk1 in SCC25 cells results in reduced p-AKT levels (Fig. 5I) and twice per week. RacGAP1 knockdown was confirmed by Western increased sensitivity to doxorubicin (Fig. 5J), which can be blotting immediately before the inoculation of SCC25 cells reversed by the addition of exogenous S1P (Fig. 5I and J). (Fig. 6A). Treatment with doxorubicin was well tolerated by the NOD/SCID mice, and the body weights remained stable through- RacGAP1 suppression enhances sensitivity of SCC25 to out the study (Fig. 6A). RacGAP1-deficient cells showed reduced doxorubicin in vivo tumor growth rate (Fig. 6B). Treatment of mice bearing vector SCC25 cells were generated to stably express either vector control SCC25 tumors with/without 0.5 mg/kg doxorubicin control or RacGAP1 shRNA and inoculated into NOD/SCID mice. had minimal effect on tumor growth rates (Fig. 6B). However,

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1945

Downloaded from mct.aacrjournals.org on September 25, 2021. © 2015 American Association for Cancer Research. 1946 o acrTe;1()Ags 2015 August 14(8) Ther; Cancer Mol aa-ehnme al. et Hazar-Rethinam leain nRcA1atvt oiyRc-T odn,adRcA1lie RacGAP1 and loading, Rac1-GTP modify activity RacGAP1 in Alterations 5. Figure iaeIbnigdmi PKPD )asy eepromdo h ex the on performed were assays B) (PAK-PBD; domain I-binding kinase iecdb hN,a niae.Teaon fatvtdo oa hAadRc a eetdby 10 detected with was combination Rac1 in and doxorubicin RhoA con To of total antibodies. doses or Rac1 or increasing activated RhoA with with of lysate amount cell The whole the indicated. and as shRNA, by silenced pk RA(ih)lvl eedtrie yqR by RacGAP1shR Ra determined and G, were (left) overexpressed. levels Sphk1shRNA was (right) e with viability RacGAP1 mRNA and transfected which Sphk1 hours were in 48 cells for cells nmol/L) p-AK SCC25 KJDSV40 (300 (E) H, BGT226 for and with shown treated levels were are protein cells levels ERK SCC25 protein total S AKT and treated. p-ERK total only hours and doxorubicin 48 percentage after as D, plotted construct. and assessed then was C2 nwihRcA1o pk a enslne eetetdwt 1 with treated were silenced been dox percentage had as Sphk1 plotted and or assessed RacGAP1 which in 1 SCC25 with treated were cells transfection, rsne smean as presented ulct eemnnsnraie o xrsino h oskeiggn B o H; for TBP gene housekeeping the of expression for normalized determinants duplicate Downloaded from G26(0 mlL – BGT226 (300nmol/L)

E bandfo rpiaedtriain ftreidpneteprmnsfrC ,adJ aaaetemean the are Data J. and G, C, for experiments independent three of determinations triplicate from obtained SEM Cell viability (% control)

Cell count 100 120 20 40 60 80 mct.aacrjournals.org S1P (1 0 CD GH (% control shRNA) Published OnlineFirstMay27,2015;DOI:10.1158/1535-7163.MCT-15-0076 100 120 . . . 0.6 0.4 0.2 0.0 Whole celllysate 20 40 60 80 AB IJ 0 μmol/L) p-AKT AKT Doxorubicin (μmol/L) m o/ 1 o 4hus -K n oa K rti eeswr hndtrie yimnbotn.J C2 and SCC25 J, immunoblotting. by determined then were levels protein AKT total and p-AKT hours. 24 for S1P mol/L rbcnol rae.Wsenblot Western treated. orubicin-only RBD

–+–+–+ SCC25 SCC25

Sphk1KD SCC25.RacGAP1KD Actin Total RhoA Active RhoA ﬁ RacGAP1 shRNA Control shRNA

m P < 0.0001 NSC23766+doxorubicin AZA1+doxorubicin Y-27632+doxorubicin Doxorubicin meulipt h ebaewsrpoe iha ci nioy ,SCclswr treated were cells SCC C, antibody. actin an with reprobed was membrane the input, equal rm o/ Z1o 10 or AZA1 mol/L

Cell viability (% untreated) Whole celllysate 100 120 20 40 60 80 t r a c t sf r o mK J D S V 4 0 ,S C C 2 5 ,a n dS C C 2 0 ptemo K.Roei-idn oan(B;A idn n 2 activat p21 and binding A) (RBD; domain Rhotekin-binding AKT. of upstream s A(ih)a ela oto hN.Atr4 or,RcA1(et and (left) RacGAP1 hours, 48 After shRNA. control as well as (right) NA ﬁ C5clswr rnfce ihteRcA1hN rasrmldshRNA scrambled a or RacGAP1shRNA the with transfected were cells CC25

m SCC25 ue r ersnaieo he needn xeiet.Alqatttv data quantitative All experiments. independent three of representative are gures o/ oouii n 1 and doxorubicin mol/L

n oa K rti eeswr eemndb mubotn.F p-AKT F, immunblotting. by determined were levels protein AKT total and T RacGAP1 KD tmtdb rpnbu xlso n lte spretg oto shRN control percentage as plotted and exclusion blue trypan by stimated m PBD o/ S276o 25 or NSC23766 mol/L Control shRNA Relative mRNA expression 1.0 1.5 2.0 0.0 0.5 Sphk1 KD Control shRNA n

RacGAP1 expression KJDSV40 Sphk1 shRNA ¼ RacGAP1 shRNA 3. Without S1P

With S1P SCC25 TotalERK p-ERK

, SCC25.RacGAP1KD P A1o pk hNs ot-ih or after hours Forty-eight shRNAs. Sphk1 or GAP1 muoltigteRDadPKPDsamples PAK-PBD and RBD the immunoblotting m < Actin Total Rac1 Active Rac1 o/ 1 o 4hus iblt a then was Viability hours. 24 for S1P mol/L 0.01; E GPsRAadcnrlshRNA control and cGAP1shRNA

m Control shRNA o/ S276fr4 or.Viability hours. 48 for NSC23766 mol/L Control shRNA Relative mRNA expression RacGAP1 shRNA , TotalAKT p-AKT P 0 1 2 3 4 el nwihRcA1hdbeen had RacGAP1 which in cells 5

< RacGAP1 shRNA Sphk1 expression 0.001; oeua acrTherapeutics Cancer Molecular F

Empty vector ,

P RacGAP1 o/exp < AKT p-AKT 0.0001. – transfected E of SEM ed A. Published OnlineFirst May 27, 2015; DOI: 10.1158/1535-7163.MCT-15-0076

RacGAP1 Is Overexpressed in Drug-Resistant SCC

A B Vector control + vehicle (DMSO) Vector control + 0.5 mg/kg doxorubicin Racgap1shRNA + vehicle (DMSO) Vector controlRacGAP1 shRNA 300 Racgap1shRNA + 0.5 mg/kg doxorubicin RacGAP1 )

3 250 β-Actin 30 200 SCC25 150

100 20 P < 0.05 P < 0.05

Tumor volume (mm volume Tumor 50

0 Weight (g) Weight 10 1 4 7 10 13 Days of treatment Figure 6. Racgap1shRNA + vehicle (DMSO) RacGAP1 suppression enhanced sensitivity 0 100 Racgap1shRNA + 0.5 mg/kg

051015 ) of SCC25 to the cytotoxic actions of 3 doxorubicin Days of treatment 80 doxorubicin in vivo. A, on the day of subcutaneous injections, RacGAP1 Vector control + vehicle (DMSO) 60 deficiency was confirmed by Vector control + 0.5 mg/kg doxorubicin 40 C immunoblotting using protein extracts RacGAP1shRNA + vehicle (DMSO) from SCC25 cells in which RacGAP1 had RacGAP1shRNA + 0.5 mg/kg doxorubicin 20 Tumor volume (mm volume Tumor stably silenced with shRNA. b-Actin is 0 provided as a loading control. All animals 1 4 7 10 13 were inoculated subcutaneously with Days of treatment 2 106 SCC25 cells expressing vector alone (scrambled shRNA) or Sphk1 shRNA and tumors allowed to establish until they reached the indicated sizes. Established D Vector control, Vector control, RacGAP1shRNA RacGAP1shRNA, IgG tumors were treated with vehicle or 0.5 vehicle doxorubicin vehicle doxorubicin mg/kg doxorubicin twice per week. A, animal weight was determined twice per week. B, tumor volumes were monitored twice weekly. The inset includes vehicle or RacGAP1 0.5 mg/kg doxorubicin-treated SCC tumors harboring RacGAP1shRNA. C, after 13 days of treatment, animals were sacrificed and tumors excised. PCNA Representative results from distinct groups are shown. þ, tumors formed from control shRNA–transfected SCC25 cells; , tumors formed from RacGAP1shRNA–transfected SCC25 cells. D, immunostaining for RacGAP1, PCNA, cleaved caspase-3, and p-AKT or normal Rabbit IgG as a negative Cleaved caspase-3 Cleaved control. Representative images of at least three independent tumors are shown for each group (Bar, 100 mm). Data presented p-AKT as mean SEM of individual measurements from 6 mice per group. E, schematic model showing how disruption of the novel E2F/RacGAP1/ Rac1/AKT pathway can lead to doxorubicin E resistance in SCC. Sphk1

E2F Sphk1 Sph S1P

Doxorubicin

E2F RacGAP1 AKT

Cell death RhoA Rac1

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1947

Hazar-Rethinam et al.

RacGAP1-deficient SCC25 tumors treated with doxorubicin tional complex (41). Similarly, E2F7 has been shown to bind the started to regress by day 4 after treatment, which continued for Sphk1 promoter in SCC cells and is associated with increased a further 7 days at which time all mice were sacrificed due to the Sphk1 transcription (15). The precise mechanism by which E2F7 tumor burden in control mice. The subcutaneous tumors were regulates the transcription of RacGAP1 and Sphk1 is currently excised, photographed, and examined histologically (Fig. 6C). under examination. Regardless of the mechanism, our functional Immunohistochemical examination of the excised tumors data show that E2F7 regulates doxorubicin-induced cytotoxicity showed that knockdown of RacGAP1 was maintained throughout via transcriptional induction of RacGAP1. the study (Fig. 6D). Tumors from vehicle and doxorubicin-treated To our knowledge, this is the first report showing overexpres- control mice stained strongly for PCNA, indicating a higher sion of RacGAP1 in HNSCC tumor samples. This is also the first proportion of proliferating cells in control tumors compared with report to show that overexpression of tumor-associated RacGAP1 RacGAP1-deficient tumors (treated or untreated; Fig. 6D). In is directly controlled by E2F7, which itself is known to be over- contrast, doxorubicin induced higher apoptosis indices in tumors expressed in SCC and to induce drug resistance (15). Overexpres- derived from SCC25/RacGAP1shRNA than in the SCC25/vector sion of RacGAP1 has been reported in high-grade meningiomas, control as estimated by immunostaining for cleaved caspase-3 non–small cell lung cancer, gastric cancer, hepatocellular carci- (Fig. 6D). Ser473 p-AKT levels in RacGAP1-deficient cells were noma, breast cancer (24, 38, 42–44) as well as in the more also decreased (Fig. 6D). Collectively, these results suggest that aggressive tumor phenotypes of epithelial ovarian cancer, high- RacGAP1 contributes to the growth of HNSCC in vivo and that grade breast cancer, and invasive cervical cancer (45–47). How- targeted inhibition of RacGAP1-overexpressing tumors may sen- ever, it is unknown whether the overexpression of RacGAP1 in sitize them to the cytotoxic actions of doxorubicin. these tumors is linked to overexpression of E2F7. Although the overexpression of RacGAP1 is not considered to be simply a – Discussion "passenger" in other cancer types (26 29), its contribution to HNSCC is unknown. Our data show that loss of RacGAP1 This is the first study to identify an E2F7/RacGAP1/AKT axis expression is able to reduce SCC growth in a xenotransplant through which SCC cells acquire resistance to doxorubicin model via inhibition of proliferation and increased basal apo- (Fig. 6E). Specifically, we show that (i) RacGAP1 is a novel ptosis. In addition, we show that high levels of expression are downstream effector of E2F7, (ii) RacGAP1 is overexpressed in associated with poor PFS of HNSCC patients. Thus, our clinical patient SCC and is associated with poor PFS, (iii) RacGAP1 and preclinical data would suggest that E2F7-dependent over- overexpression is associated with inactivation of RhoA and acti- expression of RacGAP1 is likely to be a driver of tumor growth and vation of Rac1, (iv) pharmacologic inhibition of Rac1, but not drug resistance in HNSCC. RhoA, reinstates doxorubicin sensitivity similar to that observed The functional consequences of RacGAP1 overexpression are following RacGAP1 knockdown in SCC cells, and (v) E2F7- not reflective of a generalized loss of GTP loading of the Rho/Rac dependent doxorubicin resistance is mediated via induction of family of GTPases. Our data indicated that RacGAP1 favored the RacGAP1/Rac1 and Sphk1, which in turn activates AKT-depen- conversion of RhoGTP to RhoGDP in HNSCC cells. However, it dent and AKT-independent pathways in vitro and in vivo. was clear that the GTP-loading status of Rac1 was also responsive The E2F transcription factor family is involved in a diverse array to changes in RACGAP1 expression. For example, RacGAP1 of cellular functions that are controlled by the relative ratio of appeared to negatively regulate the GTP-loading status of Rac1 atypical E2F (e.g., E2F7) to activating E2F (e.g., E2F1). For such that knockdown of RacGAP1 resulted in a reduction in the example, the apoptotic actions of E2F1, in SCC cells, can be GTP-loading state of Rac1. Although this seemed counterintui- antagonized by E2F7 overexpression (14). Similarly, E2F7 inhi- tive, it has been shown that phosphorylation of RacGAP1, by bits doxorubicin-induced cytotoxicity by inducing the expression Aurora B kinase, on Serine 387 shifts its GAP activity from Rac to of Sphk1 resulting in increased levels of S1P which enhance the Rho, resulting in increased GTP loading (activation) of Rac1 and Ser473 p-AKT–dependent prosurvival response (15). These data reduced GTP loading of Rho (inactivation; ref. 36). Significantly, are of particular relevance because we know that the majority of we show that incubation of doxorubicin-resistant SCC cells with a human SCCs express high levels of both E2F1 and E2F7. In the selective Rac1 inhibitor was able to reinstate doxorubicin sensi- present study, we found that high levels of RacGAP1 in advanced tivity, whereas incubation with a RhoA/ROCK1 selective inhibitor SCC patients were associated with a poor PFS. Moreover, we did not. These data suggest that the resistance to doxorubicin demonstrated that forced overexpression of E2F7 was able to correlates with the GTP-loaded "active" state of Rac1. This con- induce RacGAP1 overexpression and doxorubicin resistance, clusion is supported by a previous study showing that the anti- whereas knockdown of E2F7 reduced RacGAP1 expression and apoptotic protein, Bcl-2, interacts with Rac1 to protect tumor cells induced sensitivity to doxorubicin in vitro and in vivo. These data from the cytotoxic actions of etoposide and daunorubicin (48). indicate that RacGAP1 is a direct downstream transcriptional Moreover, our conclusion is also consistent with a previous report target of E2F7. The RacGAP1 promoter contains E2F-binding sites, suggesting that Rac1 was a potential therapeutic target in chemor- and E2F activation has been reported to be required for the adioresistant HNSCC (49). Thus, the E2F7-dependent drug-resis- initiation of transcription at the RacGAP1 promoter in human tant phenotype we observed in SCC cells may result from an lymphocyte cell line (40). Consistent with this, we showed that Aurora-B kinase–mediated switch of RacGAP1 activity from Rac1 elevated E2F7 levels in SCC are associated with increased binding to RhoA leading to Rac1-dependent activation of AKT-mediated of E2F7 to the RacGAP1 promoter and increased expression of doxorubicin resistance. RacGAP1 in SCCs. E2F7 is traditionally considered to be a There is an existing literature on the role of PI3K/AKT in transcriptional repressor; however, it has also been shown that controlling RacGAP1 activity and the GTP loading of the E2F7 can function as a direct transcriptional activator of the Rho family of GTPases. In particular, it has been shown that VEGFA promoter via the formation of an E2F7–HIF1a transcrip- AKT directly binds to and activates RacGAP1 activity via

1948 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

RacGAP1 Is Overexpressed in Drug-Resistant SCC

phosphorylation of T249 (50). In the present study, we show that Acquisition of data (provided animals, acquired and managed patients, E2F7 induces RacGAP1 expression, which is associated with (i) an provided facilities, etc.): M. Hazar-Rethinam, S. Boros, M. Dzienis, D.D.E. increase in activated Rac1, (ii) a decrease in RhoA activity, (iii) an Dantzic Analysis and interpretation of data (e.g., statistical analysis, biostatistics, increase in Ser473 p-AKT, and (iv) resistance to doxorubicin. computational analysis): M.Hazar-Rethinam,P.Mukhopadhyay,N.A. These findings are consistent with the observation that PI3K/AKT Saunders signaling is among the most significantly altered canonical path- Writing, review, and/or revision of the manuscript: M. Hazar-Rethinam, O.M. ways following RacGAP1 silencing in HCC (38). Our observa- Gannon, M. Dzienis, F. Simpson, N.A. Saunders tions also suggest that the overexpression of RacGAP1 in SCC may Administrative, technical, or material support (i.e., reporting or organizing contribute to the activation of the AKT pathway that is seen in data, constructing databases): M. Hazar-Rethinam, L.M. de Long, A.C. Vargas, N. Saenz-Ponce, N.A. Saunders more than 40% of all HNSCCs. However, how RacGAP1 con- Study supervision: F. Simpson, N.A. Saunders tributes to AKT activation remains unclear. We certainly know that fi overexpressing or knocking down E2F7 or RacGAP1 modi es Acknowledgments Ser473 phosphorylation of AKT. However, we also know that The authors acknowledge the generous gift of E2f7Flox/Flox or E2f8Flox/Flox E2F7 directly induces Sphk1 expression and S1P levels in SCC, mice from professor Gustavo Leone, The Ohio State University (Columbus, leading to increased Ser473 phosphorylation of AKT (15). Finally, OH). They also acknowledge the generous donations of tissue samples from the we now show that Sphk1 and RacGAP1 indirectly modify one patients without which this project could not have been possible. another's expression. Thus, it is difficult to determine whether RacGAP1-dependent effects on AKT phosphorylation status are Grant Support modified by RacGAP1 or indirectly via changes in Sphk1/S1P. N.A. Saunders is supported by grants from the Australian NHMRC Regardless of the mechanism, our results demonstrate the exis- (#APP1049182) and the Cancer Council Queensland (#APP1025479). He is also supported by a Senior Research Fellowship awarded by the Cancer Council tence of a novel, complex, and interdependent network between Queensland. O.M. Gannon is supported by a grant from the Wesley St Andrews E2F7, RacGAP1, and Sphk1/S1P, and the importance of such a Research Institute (#WRI2011-17). M. Hazar-Rethinam is supported by an network in chemosensitivity. Australian Postgraduate Award. The costs of publication of this article were defrayed in part by the Disclosure of Potential Conflicts of Interest payment of page charges. This article must therefore be hereby marked advertisement No potential conflicts of interest were disclosed. in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Authors' Contributions Conception and design: M. Hazar-Rethinam, N.A. Saunders Development of methodology: M. Hazar-Rethinam, O.M. Gannon, A.C. Var- Received January 27, 2015; revised May 13, 2015; accepted May 15, 2015; gas, N. Saenz-Ponce, F. Simpson, N.A. Saunders published OnlineFirst May 27, 2015.

References 1. Shen B, Dong P, Li D, Gao S. Expression and function of ABCG2 in head 11. Wong CF, Barnes LM, Dahler AL, Smith L, Serewko-Auret MM, Popa C, et al. and neck squamous cell carcinoma and cell lines. Exp Ther Med 2011; E2F modulates keratinocyte squamous differentiation: implications for 2:1151–7. E2F inhibition in squamous cell carcinoma. J Biol Chem 2003;278: 2. Li R, Zang Y, Li C, Patel NS, Grandis JR, Johnson DE. ABT-737 synergizes 28516–22. with chemotherapy to kill head and neck squamous cell carcinoma cells via 12. Dicker AJ, Popa C, Dahler AL, Serewko MM, Hilditch-Maguire PA, Frazer a Noxa-mediated pathway. Mol Pharmacol 2009;75:1231–9. IH, et al. E2F-1 induces proliferation-specific genes and suppresses squa- 3. Haddad R, Colevas AD, Tishler R, Busse P, Goguen L, Sullivan C, et al. mous differentiation-specific genes in human epidermal keratinocytes. Docetaxel, cisplatin, and 5-fluorouracil-based induction chemotherapy Oncogene 2000;19:2887–94. in patients with locally advanced squamous cell carcinoma of the head 13. Kwong RA, Nguyen TV, Bova RJ, Kench JG, Cole IE, Musgrove EA, et al. and neck: the Dana Farber Cancer Institute experience. Cancer 2003; Overexpression of E2F-1 is associated with increased disease-free survival 97:412–8. in squamous cell carcinoma of the anterior tongue. Clin Cancer Res 4. Sharafinski ME, Ferris RL, Ferrone S, Grandis JR. Epidermal growth factor 2003;9:3705–11. receptor targeted therapy of squamous cell carcinoma of the head and neck. 14. Endo-Munoz L, Dahler A, Teakle N, Rickwood D, Hazar-Rethinam M, Head Neck 2010;32:1412–21. Abdul-Jabbar I, et al. E2F7 can regulate proliferation, differentiation, and 5. Hazar-Rethinam M, Endo-Munoz L, Gannon O, Saunders N. The role of the apoptotic responses in human keratinocytes: implications for cutaneous E2F transcription factor family in UV-induced apoptosis. Int J Mol Sci squamous cell carcinoma formation. Cancer Res 2009;69:1800–8. 2011;12:8947–60. 15. Hazar-Rethinam M, Merida de Long L, Gannon O, Topkas E, Boros S, 6. Johnson DG, Degregori J. Putting the oncogenic and tumor suppressive Vargas AC, et al. A novel E2F/Sphingosine kinase 1 axis regulates anthra- activities of E2F into context. Curr Mol Med 2006;6:731–8. cycline response in squamous cell carcinoma. Clin Cancer Res 2015;21: 7. Wikonkal NM, Remenyik E, Knezevic D, Zhang W, Liu M, Zhao H, et al. 417–27. Inactivating E2f1 reverts apoptosis resistance and cancer sensitivity in 16. Erlich RB, Kherrouche Z, Rickwood D, Endo-Munoz L, Cameron S, Dahler Trp53-deficient mice. Nat Cell Biol 2003;5:655–60. A, et al. Preclinical evaluation of dual PI3K-mTOR inhibitors and histone 8. Berton TR, Mitchell DL, Guo R, Johnson DG. Regulation of epidermal deacetylase inhibitors in head and neck squamous cell carcinoma. Br J apoptosis and DNA repair by E2F1 in response to ultraviolet B radiation. Cancer 2012;106:107–15. Oncogene 2005;24:2449–60. 17. Zhao KN, Gu W, Fang NX, Saunders NA, Frazer IH. Gene codon 9. Panagiotis Zalmas L, Zhao X, Graham AL, Fisher R, Reilly C, Coutts AS, et al. composition determines differentiation-dependent expression of a viral DNA-damage response control of E2F7 and E2F8. EMBO Rep 2008;9: capsid gene in keratinocytes in vitro and in vivo. Mol Cell Biol 2005; 252–9. 25:8643–55. 10. Li J, Ran C, Li E, Gordon F, Comstock G, Siddiqui H, et al. Synergistic 18. Jones SJ, Dicker AJ, Dahler AL, Saunders NA. E2F as a regulator of function of E2F7 and E2F8 is essential for cell survival and embryonic keratinocyte proliferation: implications for skin tumor development. development. Dev Cell 2008;14:62–75. J Investig Dermatol 1997;109:187–93.

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1949

Hazar-Rethinam et al.

19. Endo-Munoz L, Cumming A, Rickwood D, Wilson D, Cueva C, Ng C, et al. 36. Doki N, Kawashima T, Nomura Y, Tsuchiya A, Oneyama C, Akagi T, et al. Loss of osteoclasts contributes to development of osteosarcoma pulmo- Constitutive phosphorylation of a Rac GAP MgcRacGAP is implicated in v- nary metastases. Cancer Res 2010;70:7063–72. Src-induced transformation of NIH3T3 cells. Cancer Sci 2009;100: 20. Dicker AJ, Serewko MM, Dahler AL, Khanna KK, Kaur P, Li A, et al. 1675–79. Functional characterization of cultured cells derived from an intraepider- 37. Minoshima Y, Kawashima T, Hirose K, Tonozuka Y, Kawajiri A, Bao YC, mal carcinoma of the skin (IEC-1). Exp Cell Res 2000;258:352–60. et al. Phosphorylation by aurora B converts MgcRacGAP to a RhoGAP 21. Erlich RB, Rickwood D, Coman WB, Saunders NA, Guminski A. Valproic during cytokinesis. Dev Cell 2003;4:549–60. acid as a therapeutic agent for head and neck squamous cell carcinomas. 38. Wang SM, Ooi LL, Hui KM. Upregulation of Rac GTPase-activating protein Cancer Chemother Pharmacol 2009;63:381–9. 1 is significantly associated with the early recurrence of human hepato- 22. Cameron SR, Dahler AL, Endo-Munoz LB, Jabbar I, Thomas GP, Leo PJ, cellular carcinoma. Clin Cancer Res 2011;17:6040–51. et al. Tumor-initiating activity and tumor morphology of HNSCC is 39. Iglesias-Bartolome R, Martin D, Gutkind JS. Exploiting the head and neck modulated by interactions between clonal variants within the tumor. Lab cancer oncogenome: widespread PI3K-mTOR pathway alterations and Invest 2010;90:1594–603. novel molecular targets. Cancer Discov 2013;3:722–5. 23. Detre S, Saclani Jotti G, Dowsett M. A "quickscore" method for immuno- 40. Seguin L, Liot C, Mzali R, Harada R, Siret A, Nepveu A, et al. CUX1 and histochemical semiquantitation: validation for oestrogen receptor in E2F1 regulate coordinated expression of the mitotic complex genes breast carcinomas. J Clin Pathol 1995;48:876–8. Ect2, MgcRacGAP, and MKLP1 in S phase. Mol Cell Biol 2009;29: 24. Pliarchopoulou K, Kalogeras KT, Kronenwett R, Wirtz RM, Eleftheraki AG, 570–81. Batistatou A, et al. Prognostic significance of RACGAP1 mRNA expression 41. Weijts BGMW, Bakker WJ, Cornelissen PWA, Liang K-H, Schaftenaar FH, in high-risk early breast cancer: a study in primary tumors of breast cancer Westendorp B, et al. E2F7 and E2F8 promote angiogenesis through patients participating in a randomized Hellenic Cooperative Oncology transcriptional activation of VEGFA in cooperation with HIF1. EMBO J Group trial. Cancer Chemother Pharmacol 2013;71:245–55. 2012;31:3871–84. 25. Zhao WM, Fang G. MgcRacGAP controls the assembly of the contractile 42. Ke HL, Ke RH, Li ST, Li B, Lu HT, Wang XQ. Expression of RACGAP1 in high ring and the initiation of cytokinesis. Proc Natl Acad Sci U S A 2005;102: grade meningiomas: a potential role in cancer progression. J Neurooncol 13158–63. 2013;113:327–32. 26. Sahai E. Mechanisms of cancer cell invasion. Curr Opin Genet Dev 43. Liang Y, Liu M, Wang P, Ding X, Cao Y. Analysis of 20 genes at chromosome 2005;15:87–96. band 12q13: RACGAP1 and MCRS1 overexpression in nonsmall-cell lung 27. Sanz-Moreno V, Gadea G, Ahn J, Paterson H, Marra P, Pinner S, et al. Rac cancer. Genes Chromosomes Cancer 2013;52:305–15. activation and inactivation control plasticity of tumor cell movement. Cell 44. Saigusa S, Tanaka K, Mohri Y, Ohi M, Shimura T, Kitajima T, et al. Clinical 2008;135:510–23. significance of RacGAP1 expression at the invasive front of gastric cancer. 28. Vega FM, Ridley AJ. Rho GTPases in cancer cell biology. FEBS Lett 2008; Gastric Cancer 2015;18:84–92. 582:2093–101. 45. Lu KH, Patterson AP, Wang L, Marquez RT, Atkinson EN, Baggerly KA, et al. 29. Yamazaki D, Kurisu S, Takenawa T. Involvement of Rac and Rho signaling Selection of potential markers for epithelial ovarian cancer with gene in cancer cell motility in 3D substrates. Oncogene 2009;28:1570–83. expression arrays and recursive descent partition analysis. Clin Cancer Res 30. Kawashima T, Hirose K, Satoh T, Kaneko A, Ikeda Y, Kaziro Y, et al. 2004;10:3291–300. MgcRacGAP is involved in the control of growth and differentiation of 46. Rosty C, Sheffer M, Tsafrir D, Stransky N, Tsafrir I, Peter M, et al. Identi- hematopoietic cells. Blood 2000;96:2116–24. fication of a proliferation gene cluster associated with HPV E6/E7 expres- 31. Tonozuka Y, Minoshima Y, Bao YC, Moon Y, Tsubono Y, Hatori T, et al. A sion level and viral DNA load in invasive cervical carcinoma. Oncogene GTPase-activating protein binds STAT3 and is required for IL-6-induced 2005;24:7094–104. STAT3 activation and for differentiation of a leukemic cell line. Blood 47. Ma XJ, Salunga R, Tuggle JT, Gaudet J, Enright E, McQuary P, et al. Gene 2004;104:3550–7. expression profiles of human breast cancer progression. Proc Natl Acad Sci 32. Nishimura K, Oki T, Kitaura J, Kuninaka S, Saya H, Sakaue-Sawano A, et al. U S A 2003;100:5974–9. APC(CDH1) targets MgcRacGAP for destruction in the late M phase. PLoS 48. Velaithan R, Kang J, Hirpara JL, Loh T, Goh BC, Le Bras M, et al. The small ONE 2013;8:e63001. GTPase Rac1 is a novel binding partner of Bcl-2 and stabilizes its anti- 33. Olson MF. Applications for ROCK kinase inhibition. Curr Opin Cell Biol apoptotic activity. Blood 2011;117:6214–26. 2008;20:242–8. 49. Skvortsov S, Dudas J, Eichberger P, Witsch-Baumgartner M, Loeffler-Ragg J, 34. Zins K, Lucas T, Reichl P, Abraham D, Aharinejad S. A Rac1/Cdc42 GTPase- Pritz C, et al. Rac1 as a potential therapeutic target for chemo-radioresistant specific small molecule inhibitor suppresses growth of primary human head and neck squamous cell carcinomas (HNSCC). Br J Cancer 2014; prostate cancer xenografts and prolongs survival in mice. PLoS ONE 110:2677–87. 2013;8:e74924. 50. Jacquemet G, Green DM, Bridgewater RE, von Kriegsheim A, Humphries 35. Akbar H, Cancelas J, Williams DA, Zheng J, Zheng Y. Rational design and MJ, Norman JC, et al. RCP-driven a5b1 recycling suppresses Rac and applications of a Rac GTPase-specific small molecule inhibitor. Methods promotes RhoA activity via the RacGAP1–IQGAP1 complex. J Cell Biol Enzymol 2006;406:554–65. 2013;202:917–35.

1950 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

RacGAP1 Is a Novel Downstream Effector of E2F7-Dependent Resistance to Doxorubicin and Is Prognostic for Overall Survival in Squamous Cell Carcinoma

Mehlika Hazar-Rethinam, Lilia Merida de Long, Orla M. Gannon, et al.

Mol Cancer Ther 2015;14:1939-1950. Published OnlineFirst May 27, 2015.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-15-0076

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2015/05/28/1535-7163.MCT-15-0076.DC1

Cited articles This article cites 50 articles, 20 of which you can access for free at: http://mct.aacrjournals.org/content/14/8/1939.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/14/8/1939.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/14/8/1939. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.