Nuclear ARRB 1 induces pseudohypoxia and cellular metabolism reprogramming in prostate cancer

• DOI: 10.15252/embj.201386874
• Publication Date: 2014-05-17T05:02:31Z
• Authors: Vincent Zecchini; Basetti Madhu; Roslin Russell; Nelma Pértega‐Gomes; Anne Warren; Edoardo Gaude; Joana Borlido; Rory Stark; Heather Ireland‐Z...; Roheet Rao; Helen Scott; Joan Boren; Charlie Massie; Mohammad Asim; Kevin Brindle; John Griffiths; Christian Frezza; David E Neal; Ian G Mills
• Keywords: HIF1A

Article16 May 2014Open Access Source Data Nuclear ARRB1 induces pseudohypoxia and cellular metabolism reprogramming in prostate cancer Vincent Zecchini Corresponding Author Department of CRUK, CRUK Cambridge Institute, University Cambridge, UK Search for more papers by this author Basetti Madhu Roslin Russell Nelma Pértega-Gomes Life Health Sciences Research School Sciences, Minho, Braga, Portugal Anne Warren Pathology, Edoardo Gaude Medical Council Cancer Cell Unit, Hutchison/MRC Centre, Joana Borlido Rory Stark Heather Ireland-Zecchini Roheet Rao Helen Scott Joan Boren Charlie Massie Mohammad Asim Kevin Brindle John Griffiths Christian Frezza David E Neal Ian G Mills Prostate Group, Centre Molecular Medicine Norway (NCMM), Nordic EMBL Partnership, Oslo Hospital, Oslo, Prevention Urology, Institute Information 1, Madhu1, Russell1, Pértega-Gomes2, Warren3, Gaude4, Borlido1,7, Stark1, Ireland-Zecchini1, Rao1, Scott1, Boren1, Massie1, Asim1, Brindle1, Griffiths1, Frezza4, Neal1,‡ Mills5,6,‡ 1Department 2Life 3Department 4Medical 5Prostate 6Department 7Present address: Biochemistry Biophysics Department, Cardiovascular California San Francisco, CA, USA ‡These authors contributed equally to the study *Corresponding author. Tel: [phone]; Fax: 763241; E-mail: [email protected] The EMBO Journal (2014)33:1365-1382https://doi.org/10.15252/embj.201386874 PDFDownload PDF article text main figures. Peer ReviewDownload a summary editorial decision process including letters, reviewer comments responses feedback. ToolsAdd favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract Tumour cells sustain their high proliferation rate through metabolic reprogramming, whereby shifts from oxidative phosphorylation aerobic glycolysis, even under normal oxygen levels. Hypoxia-inducible factor 1A (HIF1A) is major regulator process, but its activation normoxic conditions, termed pseudohypoxia, not well documented. Here, using an integrative approach combining first genome-wide mapping chromatin binding endocytic adaptor, ARRB1, both vitro vivo with gene expression profiling, we demonstrate that nuclear contributes shift via regulation HIF1A transcriptional activity conditions succinate dehydrogenase A (SDHA) fumarate hydratase (FH) expression. ARRB1-induced may facilitate adaptation growth harsh are frequently encountered within solid tumours. Our example adaptor protein regulating pathways. It implicates as potential tumour promoter highlights importance alterations cancer. Synopsis Global occupancy- data, together 'pseudohypoxic' HIF1alpha stability establish predominantly function Unbiased, occupancy map Integrated data highlight ARRB1-regulated networks Discovery novel cell cycle control Evidence pseudohypoxic stabilisation HIFA Introduction Beta-arrestin1 (ARRB1) ubiquitously expressed wide range molecular functions (Lefkowitz Shenoy, 2005). Recently, it has been shown contribute number diseases, (Dasgupta et al, 2006, 2011; Rosano 2009; Liu Lundgren 2011). About decade ago, landmark brought light role transcription (Kang Since then, other studies have confirmed described contribution growth, invasion metastasis lung breast carcinoma lines Shenoy 2004). One such showed co-localise physically interact hypoxia-inducible nucleus potentiate HIF1-dependent transcription, thereby mediating metastatic (Shenoy Under hydroxylated at specific proline residues prolyl hydroxylases (PHDs), tagging ubiquitination subsequent degradation proteasome pathway (Maxwell 1999). Hypoxia inhibits hydroxylation resulting upon which can translocate into heterodimerise HIF1B form functional (TF) binds regions activate target genes (Semenza, 2007b). Hypoxic switch plethora results increased glycolysis reduced mitochondrial 2007c, 2010). This Warburg effect, allows meet increase biomass production required rapid proliferation. Based on report interacts regulates given important critical played progression (Park 2012), carried out detailed determine genomics metabolomics. Herein, cistrome transcriptome adaptor. We identify interaction between recruited HIF1A-dependent manner where co-regulator activity. go show acts modulator facilitates glucose uptake that, modulation TCA metabolites, named pseudohypoxia. cells. Results upregulated maps chromosome locus 11q13, often amplified human cancers (Schwab, 1998; Kenny 1999; Buchanan 2006) (Supplementary Fig S1A). region shows regional gain or amplification 15% (El Gedaily 2001), recent 11q13 revealed multiple independent loci associated risk (Zheng Chung In addition, found be top 1% overexpressed compared tissue clinical (Wallace 2008) S1B C). Despite this, there no involvement Using two microarrays (TMAs) cancer, examined levels immunohistochemically non-neoplastic (Fig Supplementary S1D E). Although cytoplasmic staining was present tissue, stronger overall intensity TMAs 1B S1F). Importantly, additional presence significantly higher 1C), strong seen high-grade areas tumours S1E). correlated Gleason score, stage biochemical recurrence, suggesting association aggressive disease 1D Consistent previous reports (Buchanan 2006; also secondary bone metastases 1A). Figure 1. Representative pattern malignant tissues. Non-neoplastic weak moderate luminal basal Staining stromal (s). Moderate intense noted 4 (G4) tumour. Intense scattered Quantification (A) (Porto TMA, see information details). NN=non-neoplastic, TT=tumour tissue. P < 0.001 total positive cases TT versus NN. (solely + nuclear) solely Assessment expressions (total only samples clinicopathological data. comparisons were statistical significance Pearson's chi-square (χ2) test, 0.05 being threshold significance. Distribution low (< 7) (≥ grade Download figure PowerPoint correlate neoplastic phenotype panel lines, faster growing, highly tumourigenic C4-2s C4-2bs, lesser extend PC3s DU145s, display LNCaPs VCaPs 2A S2A). used C4-2 generate stably expressing AcGFP-tagged wild-type (wtARRB1) (nucARRB1) stable knock-down (KD) S2B–E). subcellular fractionation constructs localise expected intracellular compartments S2F G). NucARRB1 wtARRB1 proliferated than cells, whereas KD decreased 2B–D), implying dependency adapter As localises cytoplasm S2E F), suggests fraction largely responsible observed effect 2. aggressiveness line Cytoplasmic lines. GFP control, nucARRB1 Proliferation control. Anchorage-independent (left) migration/invasion (right) nucARRB1. Migration/invasion LNCaP information: (B C) N = 3, (D E) 6, values mean ± s.e.m., *P 0.05, **P 0.01, ***P 0.001. available online figure. [embj201386874-SourceData-Fig2A.pdf] Previous chemotaxis, regulate spread (Ge enhanced transformed indicated anchorage-independent migratory invasive potential, had opposite 2D S2H). when LNCaPs, lower endogenous C4-2s, resulted relative 2E). Thus, positively Genomic landscape order dissect mechanism behind generated whole-genome ChIP-seq analysis genes. previously p300 complex 2005), p300. selected better suited ChIP. characterised ARRB antibody ChIP 2005) Illumina platform library preparation sequencing. MACS algorithm (Zhang peaks matched inputs. 11,129 41,727 sites p300, respectively, identified S3A). distribution CEAS (Ji enriched gene-proximal 17.3% located promoters (0–3,000 bp start (TSS)) 38.9% intronic 3A). majority (86.1%) promoter-associated situated 1,000 TSS Such recruitment cis-regulatory elements supports our system. 3. A. CEAS-generated genomic H3K4me1 H3K4me3 frequency considered. pie chart ARRB1-binding proximal regions. B. Venn diagram showing overlap (minimum 1 bp) p300-binding C. sites, either alone shared C4-2s. D, E. Distance peak centres nearest (D) RNAPolII site (E). F. G. Genome Browser view ChIP-Seq enrichment profiles H3K4me3, RNAPII parental C4-2, To refine focus functionally active genome, sets mono (H3K4me1)- tri (H3K4me3)-methylated lysine residue histone H3, markers actively regulated enhancer regions, respectively (Barski 2007; Heintzman 2007, Wang 2008, Hon Bernstein 2013). 129,397 29,172 3A Out initially identified, 7,681 39,349 marks. These considered analysis. Comparison cistromes partial they act same regulatory 3B). However, non-overlapping indicates modulate independently Of loci, 82% overlapped 18% H3K4me1, bias 3C). Shared ARRB1/p300 preferentially (51.7% 15.8% H3K4me1), proportion bound enhancers (30.3% 2.2% respectively) greater affinity consistent physical complexes 2005; Dasgupta tightly centred RNA PolII (RNAPII), typically transcribed genes, further emphasising ARRB1's 3D these S3A–F), indicating constitutive overexpression does alter landscape. Having validated findings examining (66.5%) H3K4me3. these, 47% 3F). several representative (IGB) illustrates consistency 3G). performed profiling bead arrays. WtARRB1 displayed clearly different clustering patterns 4A S4A), large differentially (DEGs) common nucARRB1, pool many changes 4B Table S1A B). Real-time PCR validation yielded experimental false discovery approximately 1.6% S4B). 4. Characterisation Gene heatmap GFP, Overlap DEG GSEA-enrichment hypoxia-responsive DU145 incubated 2, 4, 8 12 h hypoxic conditions. Ingenuity Pathway Analysis (IPA) 854 direct targets. small cartoon categories. IPA analyses p300/ARRB1- alone-regulated subgroups shown. Functional DAVID ontology (GO) involved S4C S1C). note, HIF1 known targets angiogenesis (VEGFA VEGFB), (ALDOA, ALDOC, ENO3, PGM1, HK2), transport (GLUT12), (MXI1, BNIP3, BNIP3L), consumption (LONP1), lipid synthesis (PPARG) (STC2). confirm unbiased manner, set obtained reporting hypoxia-induced response DU145, (Starmans 2012). Set Enrichment robust correlation early hours incubation hypoxia, confirming signature 4C). qPCR Q394L mutation prevents translocation keeps cytoplasmic, (Scott 2002; 2003) S4D find integrated derived core S4F S1D). GO amongst most significant 4D, S4H I S1E F). exclusively closely related (nucleosome organisation, nucleosome assembly disassembly, protein/DNA assembly), vast linked 4D). annotation processes subsets S1G), likely vivo. hypothesised might dysregulated similar way Five case S4J). conserved dependent and, importantly, fashion occupies modulates tested whether response. real-time quantitative mRNA extracted cultured (1% O2 h) assess combined hypoxia additive effects prevented full induction 5A S5A). 5. (top) scramble shRNA (bottom) O2) 0, h. Heatmap VEGFA, LDHA, MXI1 BNIP3L measured qRT-PCR. Expression transiently transfected (scr) siRNAs (siRNA1 siRNA2) grown 48 post-transfection. TF motif over-representation DREME. ARRB1-associated motifs JASPAR_CORE