Role of VHL, HIF1A and SDH on the expression of miR-210: Implications for tumoral pseudo-hypoxic fate

Oncotarget1 www.impactjournals.com/oncotarget Role of VHL, HIF1A and SDH on the expression of miR-210: Implications for tumoral pseudo-hypoxic fate Anna Merlo1,*, Cristóbal Bernardo-Castiñeira1,*, Inés Sáenz-de-Santa-María1, Ana S Pitiot2, Milagros Balbín2, Aurora Astudillo3, Nuria Valdés4, Bartolomé Scola5, Raquel Del Toro6,7, Simón Méndez-Ferrer6,8, José I Piruat7, Carlos Suarez1, María-Dolores Chiara1 1Otorhinolaryngology Service, Hospital Universitario Central de Asturias, Instituto Universitario de Oncología del Principado de Asturias, Universidad de Oviedo, CIBERONC, Oviedo, Spain 2Service of Molecular Oncology, Hospital Universitario Central de Asturias, Instituto Universitario de Oncología del Principado de Asturias, Universidad de Oviedo, Oviedo, Spain 3Service of Pathology, Hospital Universitario Central de Asturias, Instituto Universitario de Oncología del Principado de Asturias, Universidad de Oviedo, Oviedo, Spain 4Service of Endocrinology and Nutrition, Hospital Universitario Central de Asturias, Instituto Universitario de Oncología del Principado de Asturias, Universidad de Oviedo, Oviedo, Spain 5Otorhinolaryngology Service, Hospital Gregorio Marañón, Madrid, Spain 6Stem Cell Niche Pathophysiology Group, Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain 7Department of Cardiovascular Physiopahology, Instituto de Biomedicina de Sevilla (IBiS), Hospital Universitario Virgen del Rocío, CSIC, Universidad de Sevilla, Sevilla, Spain 8Stem Cell Institute and Department of Haematology, University of Cambridge and National Health Service Blood and Transplant, Cambridge Biomedical Campus, UK *These authors have contributed equally to this work Correspondence to: María-Dolores Chiara, email: [email protected] Keywords: succinate dehydrogenase, hypoxia inducible factor, von hippel lindau, paragangliomas, miR-210 Received: July 02, 2016 Accepted: December 13, 2016 Published: December 27, 2016 ABSTRACT The hypoxia-inducible factor 1a (HIF-1a) and its microRNA target, miR-210, are candidate tumor-drivers of metabolic reprogramming in cancer. Neuroendocrine neoplasms such as paragangliomas (PGLs) are particularly appealing for understanding the cancer metabolic adjustments because of their associations with deregulations of metabolic enzymes, such as succinate dehydrogenase (SDH), and the von Hippel Lindau (VHL) gene involved in HIF-1α stabilization. However, the role of miR-210 in the pathogenesis of SDH-related tumors remains an unmet challenge. Herein is described an in vivo genetic analysis of the role of VHL, HIF1A and SDH on miR-210 by using knockout murine models, siRNA gene silencing, and analyses of human tumors. HIF-1a knockout abolished hypoxia-induced miR-210 expression in vivo but did not alter its constitutive expression in paraganglia. Normoxic miR-210 levels substantially increased by complete, but not partial, VHL silencing in paraganglia of knockout VHL-mice and by over-expression of p76del-mutated pVHL. Similarly, VHL- mutated PGLs, not those with decreased VHL-gene/mRNA dosage, over-expressed miR-210 and accumulate HIF-1a in most tumor cells. Ablation of SDH activity in SDHD-null cell lines or reduction of the SDHD or SDHB protein levels elicited by siRNA-induced gene silencing did not induce miR-210 whereas the presence of SDH mutations in PGLs and tumor-derived cell lines was associated with mild increase of miR-210 and the presence of a heterogeneous, HIF-1a-positive and HIF-1a-negative, tumor cell population. Thus, activation of HIF-1a is likely an early event in VHL- defective PGLs directly linked to VHL mutations, but it is a late event favored but www.impactjournals.com/oncotarget/ Oncotarget, Advance Publications 2016 Oncotarget2 www.impactjournals.com/oncotarget not directly triggered by SDHx mutations. This combined analysis provides insights into the mechanisms of HIF-1a/miR-210 regulation in normal and tumor tissues potentially useful for understanding the pathogenesis of cancer and other diseases sharing similar underpinnings. INTRODUCTION A defining hallmark of cancer is metabolic reprograming. Most of cancer cells use glycolysis-based metabolism for proliferation and reprogramming of the tricarboxylic acid cycle and mitochondrial oxidative phosphorylation to adapt to challenging conditions [1]. Aside this general cancer feature, there is a wealth of evidences proving that altered metabolic enzymes and their metabolites are also oncogenic and not simply a product of tumor proliferation [2]. This is particularly relevant in pheochromocytomas and paragangliomas (PCC/PGLs) which frequently develop in patients with germline mutations of genes encoding the succinate dehydrogenase mitochondrial complex II (SDH). These associations provide a useful tumor model to unravel the molecular connections between mitochondrial activity and cancer. Importantly, besides PCC/PGLs, SDH inactivation is increasingly observed in other neuroendocrine malignancies as well as in gastrointestinal neoplasms and renal cell carcinomas [3], and, thus, it is expected that it will acquire a more general clinical relevance in cancer pathogenesis. PGLs are rare neuroendocrine tumors of the autonomous nervous system. Although the majority of PGLs occur sporadically, about 30-45% of these tumors are hereditary due to germline mutations in the succinate dehydrogenase SDHB, SDHC, SDHD, SDHA, and SDHAF2 genes (hereafter SDHx genes), among other predisposing genes [4]. Mutations in any of the SDHx genes have been associated to succinate accumulation. In vitro assays revealed that excess of intracellular succinate may lead to increased transcriptional activity of hypoxia-inducible factors (HIF) [5, 6] and up-regulation of pro-tumorigenic HIF target genes, such as carbonic anhydrase 9 (CA9) and hypoxia-inducible factor prolyl hydroxylase 3 (EGLN3), irrespective of oxygen levels (a phenomenon defined as ‘pseudohypoxia’). PGLs may also associate with germline mutations in the von Hippel-Lindau (VHL) tumor suppressor [7]. Inherited mutations in the VHL gene cause the VHL disease, an autosomal-dominant neoplastic disease that is associated with various tumour types, including clear cell renal cell carcinomas, haemangioblastomas, pancreatic neuroendocrine tumours and PCC/PGLs [8]. VHL type 1 disease is characterized by development of clear cell renal cell carcinomas and hemangioblastomas, but not PCCs, and is associated with gross deletions in VHL. Type 2 disease is predominantly associated with VHL missense mutations and the development of PCCs (type 2C) or PCCs and hemangioblastomas (type 2A) or PCCs, hemangioblastomas and RCCs (type 2B). Somatic VHL mutations (i.e. mutations in tumor but not germline DNA) have been also described in about 9% of PGLs [9–11]. Intriguingly, the two types of VHL somatic mutations (type 1 and type 2) have been identified in parasympathetic PGLs [11] and are thus thought to be involved in the development of these tumors via a molecular mechanism not completely understood. VHL functions as the substrate recognition component of an E3-ubiquitin ligase, which targets HIF for proteasomal degradation under normoxic conditions [12]. Thus, HIF represent the molecule where SDH and VHL dysfunctions converge in PGLs. Nonetheless, the role of HIF in SDH-related tumorigenesis remains controversial. Some reported transcription profiles have shown that SDHx-mutated PGLs (SDHx-PGLs) showed enrichment in genes associated with the HIF- pathway and accumulate HIF-1α [13, 14] whereas others did not find such an association thus arguing against the pseudohypoxia hypothesis in SDHx-PGLs [11, 15, 16]. In contrast to SDHx-PGLs, the canonical HIF signaling pathway has been more clearly detected in most tumors that carry VHL mutations [11]. Studies in VHL-related PGLs (VHL-PGLs) revealed that the HIF-pathway involves, among other cues, the HIF-1α-microRNA target, miR-210 [11, 16]. Importantly, VHL-PGLs, in addition to accumulate HIF-1α, express low levels of SDHB likely as a result of miR-210 upregulation which also represses expression of other genes essential for mitochondrial metabolism [17]. This suggests that miR-210 is a potential oncomiR in VHL-deficient tumors which impact on mitochondria metabolism [16]. Thus, understanding of the in vivo mechanisms of miR-210 expression should increase our knowledge on cancer pathogenesis and allow the identification of cancer-specific vulnerabilities that could be exploited therapeutically. However, the role of miR-210 in the pathogenesis of SDH-related tumors remains an unmet challenge. These premises encouraged us to carry out an in-depth analysis of the role of HIF-1α, SDH, and VHL on miR-210 expression by using knockout murine models and analyses of human tumor material. RESULTS In vivo roles of VHL and HIF-1α on the expression of miR-210 We first addressed whether the expression of miR- 210 in the paraganglionic system is regulated by VHL and/or HIF-1α activity in vivo. To this end, we analyzed miR-210 levels in tissues isolated from mice with either Oncotarget3 www.impactjournals.com/oncotarget VHL or HIF1A conditionally deleted in the neural crest derived cells. We present here data obtained from adrenal medulla. Figure 1A shows a significant 13-fold increase of miR-210 levels in the adrenal gland of VHL-null mice in comparison with wild type mice. However, miR-210 levels did not change in heterozygous tissues obtained from VHL +/- thus indicating that a single functional copy of VHL produce enough functional VHL. In contrast to VHL, absence of HIF-1α did not alter the constitutive expression of miR-210. As shown in Figure 1B, miR-210 levels in the adrenal gland of normoxic knockout HIF1A -/- or HIF1A +/- mice were similar to those of wild type HIF1A +/+. Similarly, mRNA levels of other HIF-1α- targets, EGLN3 and CA9, did not decrease in the absence or partial loss of HIF-1α. To further analyze the role of HIF-1α on miR-210 expression, wild type HIF1A +/+ and knockout HIF1A -/- mice were exposed to hypoxia (10% O2) for 30 days prior to organ isolation to allow for HIF- 1α accumulation in tissues. Under these conditions, the functional inactivation of HIF-1α significantly reduced hypoxic expression of miR-210 as well as EGLN3 in the adrenal medulla (Figure 1C). Given that most tumors carrying VHL inactivating mutations are not necessarily accompanied by loss of heterozygosity, we also analyzed the impact of a VHL cancer-associated mutation, F76del VHL, on miR-210 expression by using SCC40 cells which, endogenously, express wild type VHL [18, 19]. F76del mutant was selected because represents the VHL gene alteration (c.227_229delTCT) associated with a wide spectrum of cancers including PGLs and no-neuroendocrine neoplasms related to VHL disease. We first used a squamous cell carcinoma cell line, SCC40, which has been well-characterized at the genetic and functional level with regard to HIF signaling [18, 19]. SCC40 cells were transfected with a VHL-expressing construct encoding for F76del-VHL. As controls, we used a VHL-expressing construct containing the wild type gene (wtVHL) or a LifeAct construct expressing an actin binding peptide (mock). Figure 1D shows that HIF-1α protein was not detected in mock-transfected cells incubated under normoxic conditions. By contrast, HIF-1α accumulated under normoxic conditions in cells transfected with F76del-VHL with further increase upon hypoxic treatment (1% O2, 24 h). As expected, normoxic accumulation of HIF-1α was not observed by ectopic expression of wild type-VHL. Consistent with our previously reported findings in PGL tissues carrying mutant F76del-pVHL [11], a ~ 3- and ~ 90-fold increase of miR-210 and CA9 RNA levels, respectively, was detected upon ectopic expression of F76del-VHL under normoxic conditions as compared with wild type VHL or mock-transfected cells (Figure 1E and 1F). Because SCC40 cells endogenously express wild type VHL, these data show that a mutant non-functional F76del-VHL may abolish ubiquitination of HIF-1α by wild type VHL thus exerting a dominant negative effect that lead to the activation of the HIF-1/ miR-210 pathway. The above results were replicated in RCC4 renal carcinoma cells expressing wild type VHL (RCC4[+]). In the absence of a human cell model suitable for in vitro analysis of PGLs, cancer cells derived from clear cell renal cell carcinomas are valuable tools for studies of parasympathetic PGLs because of the genetic similarities between both types of tumors [11]. As shown in Figure 1G, a significant increase of miR-210 and CA9 was detected upon expression of the F76del-VHL mutant in RCC4[+] cells. This result was not replicated in 786-O[+] cells that express wild type VHL and the HIF-2α rather than HIF-1α subunit thus suggesting that HIF-1α is the predominant HIF-α subunit involved in the pseudohypoxic up-regulation of miR-210. Roles of SDHD and SDHB on the expression of miR-210 To determine whether SDH loss of function exert any effect on miR-210 expression, we used previously reported established cell lines (MEF and BMK) obtained from a tamoxifen-inducible SdhD knockout mutant mouse [20]. Homozygous SdhD -/-, but not heterozygous SdhD +/-, MEF and BMK cell lines are known to partially recapitulate the HIF pathway in comparison to wild type homozygous SdhD +/+ cell lines. By contrast, tissues (such as adrenal gland) derived from conditional homozygous SdhD -/- mice do not over-express HIF-1α nor its targets [20] and therefore they were not expected to over-express miR-210. Analysis of miR-210 in MEF and BMK cells revealed that ablation of the two (homozygous SdhD -/-) or one (heterozygous SdhD +/-) SdhD alleles failed to induce miR-210 over-expression in BMK-cells and even caused a reduction of miR-210 expression levels in MEF cells (Figure 2A). These data contrasted with the reported link between SDHD deficiency and HIF-1α accumulation what prompted us to use another cell-system to clarify the role of decreased mitochondrial complex II activity on miR-210 expression. We first used thenoyltrifluoroacetate (TTFA) which binds to the ubiquinone docking sites and abrogates the electron flow at the SDH complex. Figure 2B shows that inhibition of the succinate:ubiquinone oxidoreductase activity by 0.5 mM TTFA did not induce miR-210 over-expression in VHL-deficient (786-O[−]) or VHL-proficient (786-O[+]) cells. On the contrary, TTFA significantly reduced miR-210 levels in VHL-deficient cells. 786-O cells are known to mainly express HIF- 2α subunit (although minimal levels of HIF-1α protein have been detected under hypoxic conditions [11]). Thus, we also tested the effect of TTFA on SCC40 cells which express HIF-1α subunit. As shown in Figure 2C, TTFA treatment did not induce miR-210 expression but slightly, although significantly, reduced it in SCC40 cells. Decreased mRNA levels were also detected for EGLN3, Oncotarget4 www.impactjournals.com/oncotarget Figure 1: In vitro and in vivo analysis of the role of VHL and HIF-1α on miR-210 expression. A, B, C. miR-210, CA9 and EGLN3 RNA levels were quantified by RT-qPCR in adrenal gland obtained from the indicated mice genotypes. For analysis of HIF1A knockouts, HIF1A +/+ and HIF1A -/- mice were exposed to hypoxia (10% O2) for 30 days prior to isolation of adrenal glands (panel C, hypoxia) or left under atmospheric O2 conditions (panel B, normoxia). D. Western blot showing HIF-1α protein levels upon ectopic expression of F76del VHL or wild type VHL in SCC40 cells incubated under normoxic (Nx: 21% O2) or hypoxic (Hx: 1% O2) conditions for 24 hours. Cells transiently transfected with LifeAct plasmid were used as controls (mock). β-actin, total protein-loading control. Fold change of HIF-1α protein levels upon normalization to total protein expression is shown below the Western blot image. E, F, G. miR-210 and CA9 RNA relative levels in SCC40 (E, F) and RCC4[+] or 786-O[+] cells (G) treated as indicated in (D). CA9 mRNA levels were undetectable in 786-O[+] cells even when incubated under hypoxic conditions. * P < 0.05 ** indicates P < 0.01. Oncotarget5 www.impactjournals.com/oncotarget another HIF-target gene, in SCC40 cells treated with TTFA and in 786-O[-] cells (Figure 2D). The role of the SDH complex on miR-210 expression was further explored in cultured cell lines by using siRNAs targeting the SDHD or SDHB genes. For this analysis, we used cells expressing wild type VHL such as 786-O[+] cells, which mainly express HIF-2α subunit, and RCC4[+] and SCC40 cells, which mainly express HIF-1α subunit. Endogenous SDHD and SDHB mRNA levels were found to be 40% to 90% lower in cells transfected with either of the two siRNAs (siSDHB or siSDHD) as compared to control siRNA (siCtrl)- transfected cells (data not shown). However, increased miR-210 levels were not found in any of the transfected cells (Figure 3), thus suggesting that full SDH function is not required for HIF-α-independent miR-210 expression. The same experimental strategy was then applied but using cells that were incubated under hypoxic conditions (1% O2 for 48 h) after siRNA transfection to allow for HIF-α-dependent miR-210 over-expression (Figure 3). As expected, increased levels of miR-210 were detected in hypoxic versus normoxic conditions. However, siSDHD or siSDHB did not further increase the hypoxic effect on miR-210 levels. Analysis of CA9 and EGLN3 mRNA Figure 2: Analysis of the role of SDHD on miR-210 expression. A. Mouse embryonic fibroblasts (MEFs) and baby mouse kidney (BMK) cells obtained from SDHD-ESR -/- mice and their homozygous SDHD flox/+ (+/+) and heterozygous SDHD flox/− (+/−) littermates were cultured in medium supplemented with 4-hydroxy-tamoxifen for 24 hours. B, C, D. VHL-proficient (786-O[+]) and VHL-deficient (786-O[-]) (B, D) or SCC40 cells (C, D) were incubated with 0.5 mM TTFA (TTFA(+)) or solvent (TTFA(-)) for 16 hours prior to miR-210 (B, C) or EGLN3 (G) quantification by RT-qPCR. Values are expressed as mean ratios ± SD **, P <0.01, unpaired t test. Oncotarget6 www.impactjournals.com/oncotarget levels also revealed that other HIF-α-target genes were not induced by decreased SDHD or SDHB expression either in normoxia or in hypoxia. Only a subtle inductive effect was found for CA9 in RCC4 [+] cells (CA9 gene could not be analyzed in 786-O[+] cells because its mRNA was not detected neither in normoxia nor in hypoxia). In agreement with those data, immunocytochemical analysis of HIF-1α revealed nuclear accumulation of this protein in siSDHB- or siSDHD-transfected cells incubated under hypoxic conditions but not under normoxic conditions (Figure 4). The above data shows that HIF-1α accumulation is not triggered, at least in a significant and detectable level, by absence of SDHD activity or decreased SDHB or SDHD protein levels. We next tested whether HIF- 1α accumulation could be associated with the presence of SDHx mutations instead of absence/decreased protein levels. All of SDHx genes are tumor suppressors, showing loss of heterozygosity, the loss of the normal allele in the tumors, in conjunction with the germline mutation. To analyze the impact of SDHx cancer-associated homozygous mutations on the expression of HIF-1α and miR-210, we used two PGL-derived cell lines, one composed of tumor cells which had been obtained from a tympanic PGL with homozygous SDHC exon 2 deletion (SDHC-PGL) [16] and other composed of tumor-associated fibroblasts which were derived from a SDHD-mutated (c.88_314+247del, p.His30Asn fs*11) jugular PGL (SDHD-TAFs). Thus, SDHC-PGL cells lack wild type SDHC protein whereas SDHD-TAFs express mutated and wild type SDHD gene (data not shown). Analysis of HIF-1α protein expression revealed that the SDHC-mutated cell line contains mixed populations of Figure 3: Absence of miR-210 up-regulation by siRNA-mediated silencing of SDHB or SDHD. miR-210, CA9 and EGLN3 RNA levels were analyzed in the indicated cells following treatment with SDHD (siSDHD), SDHB (siSDHB) or control (siCtrl) siRNAs. 24 hours after transfection, cell were incubated under normoxic (Nx) or hypoxic (Hx, 1% O2 for 25 hours) conditions. Values are expressed as mean ratios ± SD. Oncotarget7 www.impactjournals.com/oncotarget Figure 4: Absence of HIF-1α protein accumulation in nuclei of cells treated with SDHB or SDHD-siRNAs. Immunocytochemical analysis of HIF-1α in SCC40 cells treated with Ctrl- (A, B), SDHB- (C, D), or SDHD- (E, F) -siRNAs and exposed to normoxia (A, C, E) or hypoxia (1% O2 for 24 h) (B, D, F). (G, H) Immunocytochemical analysis of HIF-1α in tumor associated fibroblasts derived from human PGL carrying SDHD mutation (SDHD-TAFs) and incubated under normoxic G. or hypoxic (1% O2 for 24 h) H. conditions; note the higher intensity of HIF-1α immunostaining (pink) in hypoxic SDHD-TAF in comparison with siSDHB- or siSDHD- treated SCC40 cells. Quantification of the SDHB or SDHD mRNA levels shows significant reduction of the mRNA levels after siRNA treatments. Oncotarget8 www.impactjournals.com/oncotarget tumor cells, i.e. 72% were negative for HIF-1α protein expression but about 28% of cells accumulate nuclear HIF-1α even under normoxic conditions (Figure 5A). No HIF-1α accumulation was observed, however, in SDHD- TAFs in normoxia but they did accumulate higher HIF- 1α levels, under hypoxic conditions, than the siSDHB or siSDHD-treated SCC40 cells thus suggesting that they are primed for a more intense HIF-1α-related cellular response (Figure 4). Figure 5B shows that both types of cells, SDHC-PGLs and SDHD-TAFs, have slightly but significant higher levels of miR-210 expression in normoxia than human paraganglia (normal carotid body) or non-tumoral fibroblasts, respectively. Analysis of EGLN3 mRNA showed ~ 3-fold and ~ 18-fold higher levels in SDHD-TAF and SDHC-PGLs than in normal fibroblasts or carotid body, respectively (Figure 5C). Thus, in contrast to the data obtained in the SDHD knock-out mouse model or in the cell-based assays, cells carrying SDHC or SDHD mutations express higher levels of HIF- 1α protein and its targets, miR-210, CA9 and EGLN3 than their normal counterparts. Nevertheless, the effect of mutations on HIF activation, if anything, seems to be rather subtle as compared with the stronger effect exerted by hypoxia (Figure 5B). Collectively, the data indicates that the constitutive expression of miR-210 requires a functional wild type VHL protein but it is not influenced by partial VHL losses. HIF- 1α, in turn, is involved in the hypoxic-induction of miR- 210 but do not participate in its constitutive expression. Deleterious VHL gene mutations seems to have dominant negative effects on wild type VHL protein and as such activate the HIF-1α pathway. Finally, loss of SDHD or decreased SDHB or SDHD expression levels does not induce the HIF-1α/miR-210 pathway. Instead, mutations of SDHx genes may exert subtle inductive effects on the HIF-1α/miR-210 pathway. Pseudo-hypoxic fate of VHL- and SDHx-related PGLs The above data cannot be directly extrapolated to the biology of tumor tissues inasmuch as they all represent cell-based models that do not necessarily recapitulate cell pathophysiology of PGLs. Thus, we next sought to determine whether the genetic-phenotypic correlations identified in the murine and cellular models were also observed in tumor tissues. To this end, we used a cohort of PGLs of parasympathetic origin with known VHL Figure 5: Mild heterogeneous activation of HIF-1α/miR-210 signaling axis in PGL-derived cells carrying SDHx mutations. A. Immunocytochemical analysis of HIF-1α in cells derived from human PGL carrying SDHC exon 2 deletion (SDHC-PGL). Note the heterogeneity of the SDHC-PGL cell line which is composed of HIF-1α-positive and -negative cells. B and C. miR-210 (B), EGLN3 and CA9 (C) mRNA levels were analyzed by RT-qPCR in SDHC-PGL or SDHD-TAFs and compared with those of human normal carotid body (CB) or normal fibroblasts (NF), respectively. Two NF cell lines (NF1 and NF2) were used in this analysis. NF1/2 refers to the miR-210 data (mean ± SD value) obtained in NF1 and NF2 cells. CA9 mRNA levels were undetectable in NF and TAF cells. Values are expressed as mean ratios ± SD. Oncotarget9 www.impactjournals.com/oncotarget and SDH gene mutations. Germline mutations of SDHD or SDHB, and somatic heterozygous mutations of VHL were found in 11, 7, and 2 patients, respectively, while 25 patients did not carry SDHx and VHL somatic/germline mutations. VHL expression data was unknown for most of those tumors. Thus, we first performed a comprehensive analysis of the VHL gene by examining VHL DNA copy number and mRNA levels. Because, unexpectedly, we had previously identified somatic VHL gene mutations in parasympathetic PGLs [11], analysis of VHL gene dosage was also performed even though chromosomal losses on the 3p have not been previously described as frequent events in PCCs/PGLs [21]. VHL DNA copy number was analyzed by MLPA assay. We identified 6 samples exhibiting decreased copy number of exon 1 of VHL (hereafter VHLdel- PGLs) (Table 1, Figure 6A) in tumor genomic DNA. This was significantly observed in PGLs lacking SDHx mutations (SDHxwt) (6/27, ~22%) but not in SDHx mutant (SDHxm)-PGLs (0/18) (P=0.02) and was not found in peripheral blood lymphocytes from any of these patients. These genetic defects were confirmed by an optimized quantitative PCR method in all tumors (data not shown), and by CGH microarray analysis in one tumor (Figure 6B). We next measured transcriptional levels of the VHL gene by RT-qPCR (n=36) to compare tumor samples with and without VHLdel. As shown in Figure 6C, low VHL transcript levels were encountered in VHLdel-PGLs as compared with PGLs lacking VHL or SDHx gene alterations (SDHwt/VHLwt) or tumors with SDHx mutations. Thus, 9 out of the 32 (28%) non-SDHx mutant PGLs included in this study carry VHL gene alterations (3 with VHL mutations -hereafter VHLm- and 6 with VHLdel). This places this tumor suppressor among the most frequently altered genes described thus far in parasympathetic PGLs. Table 1 shows the association between the clinicopathological characteristics of the patients and the presence of VHLdel. No association was found with age, location, functionality, multiplicity, or malignancy of tumors. We next analyzed the relative HIF-1α and miR- 210 levels in tumors with specific genetic backgrounds: SDHxwt/VHLwt, VHLm, VHLdel, and SDHxm (SDHB or SDHD)m. miR-210 quantifications revealed that, as previously described, VHLm-PGLs expressed higher levels of miR-210 than SDHx m - or SDHx wt /VHL wt -PGLs (Figure 7A). However, VHLdel-PGLs expressed low levels of miR- 210 which were similar to those detected in SDHxwt/VHLwt- PGLs. In addition, SDHx m -PGLs expressed slightly higher, statistically significant, levels of miR-210 compared to SDHxwt/VHLwt-PGLs (p=0.045). Then, we analyzed HIF-1α protein expression in PGL tumor tissues (n=46) by immunohistochemistry. Given that many of the samples showed nonhomogeneous distribution of positive nuclei and distinct intensity of immunostaining, both, the percentage of positively stained cells and the relative intensity of immunostaining, were considered for HIF-1α quantification. Some of the tumors had been chemoembolized prior to surgery, but we did not find correlation between the percentage of positively immunostained cells and presurgical chemoembolization thus ruling out a “iatrogenic” effect of embolization not related to tumor biology. Overall, HIF-1α immunoscores ranged from 0 to 240 with a mean value of 52 and median value of 17. Comparison of the immunoscores in tumors revealed a gradient in the levels of HIF- 1α protein spanning from very low (SDHx wt /VHL wt - and VHLdel-PGLs) to low (SDHxm-PGLs) and high (VHLm- PGLs) expression levels (Figure 7B). The difference of immunostaining in VHLm-PGLs and SDHxm-PGLs lies in that the first has a high percentage of positive nuclei homogeneously distributed in the whole tumor section whereas the last takes on a patched appearance apparently not related to the distance to blood vessels (see representative immunohistochemical images in Figure 7C). In line with the miR-210 expression data, differences in HIF-1α immunostainings were statistically significant between VHLdel-PGLs and VHLm- PGLs (p<0.0001) or SDHx-PGLs (p=0.03), but not between VHL del -PGLs and SDHx wt /VHL wt -PGLs. Differences were also statistically significant between SDHxm-PGLs and SDHxwt/VHLwt- (p=0.044). Next, HIF-1α protein expression and SDHx genetic analysis was extended to a total of 112 parasympathetic PGLs (41 SDHxwt/VHLwt-PGLs and 39 SDHx m -PGLs all of them lacking VHL gene mutations) for validation. In addition to genetic analysis, the presence of SDHx mutations was analyzed by SDHB immunohistochemistry in 87 of the tumors given that negative immunostaining has been identified as a biomarker of SDHx mutations [22]. Indeed, our data showed that SDHB immunohistochemical negativity was significantly associated with the presence of mutations in any of the SDHx genes (p<0.0001). Comparison of the HIF-1α protein levels in tumors harboring or lacking SDHx mutations confirmed previously reported data revealing that SDHx-PGLs have a significantly higher percentage of HIF-1α positive cells than tumors lacking SDHx mutations (Figure 7D, p=0.008). No differences in HIF-1α protein levels were found, however, between SDHD- and SDHB-mutated PGLs/PCCs (p=0.999). Importantly, none of the SDHx-mutated tumors showed the very high percentage of positive nuclei that had been observed in VHL-PGL. DISCUSSION In this report we have used cellular- and knockout murine-models and neuroendocrine PGLs to determine whether the HIF-1α and VHL proteins, implicated in the miR-210-mediated hypoxia/pseudohypoxia response Oncotarget16 www.impactjournals.com/oncotarget Red Temática de Investigación Cooperativa en Cáncer [RD12/0036/0015] Instituto de Salud Carlos III (ISCIII), Spanish Ministry of Economy and Competitiveness & European Regional Development Fund (ERDF); the European Regional Development's funds (FEDER) (CIBERONC) and Obra Social CajAstur-Instituto Universitario de Oncología del Principado de Asturias. REFERENCES 1. Schulze A, Harris AL. How cancer metabolism is tuned for proliferation and vulnerable to disruption. Nature. 2012; 491: 364–73. 2. Morin A, Letouzé E, Gimenez-Roqueplo A-P, Favier J. Oncometabolites-driven tumorigenesis: From genetics to targeted therapy. Int J cancer. 2014; 135: 2237–48. 3. Evenepoel L, Papathomas TG, Krol N, Korpershoek E, de Krijger RR, Persu A, Dinjens WNM. Toward an improved definition of the genetic and tumor spectrum associated with SDH germ-line mutations. Genet Med. 2015; 17: 610–20. 4. Gimenez-Roqueplo A-P, Tischler AS. Pheochromocytoma and Paraganglioma: progress on all fronts. Endocr Pathol. 2012; 23: 1–3. 5. Selak MA, Armour SM, MacKenzie ED, Boulahbel H, Watson DG, Mansfield KD, Pan Y, Simon MC, Thompson CB, Gottlieb E. Succinate links TCA cycle dysfunction to oncogenesis by inhibiting HIF-alpha prolyl hydroxylase. Cancer Cell. 2005; 7: 77–85. 6. Tannahill GM, Curtis AM, Adamik J, Palsson-McDermott EM, McGettrick AF, Goel G, Frezza C, Bernard NJ, Kelly B, Foley NH, Zheng L, Gardet A, Tong Z, et al. Succinate is an inflammatory signal that induces IL-1β through HIF-1α. Nature. 2013; 496: 238–42. 7. King KS, Pacak K. Familial pheochromocytomas and paragangliomas. Mol Cell Endocrinol. 2014; 386: 92–100. 8. Maher ER, Neumann HP, Richard S. von Hippel-Lindau disease: a clinical and scientific review. Eur J Hum Genet. 2011; 19: 617–23. 9. Gaal J, van Nederveen FH, Erlic Z, Korpershoek E, Oldenburg R, Boedeker CC, Kontny U, Neumann HP, Dinjens WNM, de Krijger RR. Parasympathetic paragangliomas are part of the Von Hippel-Lindau syndrome. J Clin Endocrinol Metab. 2009; 94: 4367–71. 10. Burnichon N, Vescovo L, Amar L, Libé R, de Reynies A, Venisse A, Jouanno E, Laurendeau I, Parfait B, Bertherat J, Plouin P-F, Jeunemaitre X, Favier J, et al. Integrative genomic analysis reveals somatic mutations in pheochromocytoma and paraganglioma. Hum Mol Genet. 2011; 20: 3974–85. 11. Merlo A, de Quirós SB, de Santa-María IS, Pitiot AS, Balbín M, Astudillo A, Scola B, Arístegui M, Quer M, Suarez C, Chiara M-D. Identification of somatic VHL gene mutations in sporadic head and neck paragangliomas in association with activation of the HIF-1α/miR-210 signaling pathway. J Clin Endocrinol Metab. 2013; 98: E1661–6. 12. Maxwell PH, Wiesener MS, Chang GW, Clifford SC, Vaux EC, Cockman ME, Wykoff CC, Pugh CW, Maher ER, Ratcliffe PJ. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature. 1999; 399: 271–5. 13. Pollard PJ, Brière JJ, Alam NA, Barwell J, Barclay E, Wortham NC, Hunt T, Mitchell M, Olpin S, Moat SJ, Hargreaves IP, Heales SJ, Chung YL, et al. Accumulation of Krebs cycle intermediates and over-expression of HIF1alpha in tumours which result from germline FH and SDH mutations. Hum Mol Genet. 2005; 14: 2231–9. 14. Pollard PJ, El-Bahrawy M, Poulsom R, Elia G, Killick P, Kelly G, Hunt T, Jeffery R, Seedhar P, Barwell J, Latif F, Gleeson MJ, Hodgson S V, et al. Expression of HIF- 1alpha, HIF-2alpha (EPAS1), and their target genes in paraganglioma and pheochromocytoma with VHL and SDH mutations. J Clin Endocrinol Metab. 2006; 91: 4593–8. 15. Blank A, Schmitt AM, Korpershoek E, van Nederveen F, Rudolph T, Weber N, Strebel RT, de Krijger R, Komminoth P, Perren A. SDHB loss predicts malignancy in pheochromocytomas/sympathethic paragangliomas, but not through hypoxia signalling. Endocr Relat Cancer. 2010; 17: 919–28. 16. Merlo A, de Quiros SB, Secades P, Zambrano I, Balbín M, Astudillo A, Scola B, Arístegui M, Suarez C, Chiara M-D. Identification of a signaling axis HIF-1α/microRNA-210/ ISCU independent of SDH mutation that defines a subgroup of head and neck paragangliomas. J Clin Endocrinol Metab. 2012; 97: E2194–200. 17. Ivan M, Huang X. miR-210: fine-tuning the hypoxic response. Adv Exp Med Biol. 2014; 772: 205–27. 18. Secades P, de Santa-María IS, Merlo A, Suarez C CM. In vitro study of normoxic epidermal growth factor receptorinduced hypoxia-inducible factor-1-alpha, vascular endothelial growth factor, and BNIP3 expression in head and neck squamous cell carcinoma cell lines: Implications for anti-epidermal growth fact. Head Neck. 2015; 37: 1150–62. 19. Secades P, Rodrigo JP, Hermsen M, Alvarez C, Suarez C, Chiara M-D. Increase in gene dosage is a mechanism of HIF-1alpha constitutive expression in head and neck squamous cell carcinomas. Genes Chromosomes Cancer. 2009; 48: 441–54. 20. Millán-Uclés A, Díaz-Castro B, García-Flores P, Báez A, Pérez-Simón JA, López-Barneo J, Piruat JI. A conditional mouse mutant in the tumor suppressor SdhD gene unveils a link between p21(WAF1/Cip1) induction and mitochondrial dysfunction. PLoS One. 2014; 9: e85528. 21. Castro-Vega LJ, Letouzé E, Burnichon N, Buffet A, Disderot P-H, Khalifa E, Loriot C, Elarouci N, Morin A, Menara M, Lepoutre-Lussey C, Badoual C, Sibony M, et al. Multi-omics analysis defines core genomic alterations in pheochromocytomas and paragangliomas. Nat Commun. 2015; 6: 6044. Oncotarget17 www.impactjournals.com/oncotarget 22. van Nederveen FH, Gaal J, Favier J, Korpershoek E, Oldenburg RA, de Bruyn EM, Sleddens HF, Derkx P, Rivière J, Dannenberg H, Petri BJ, Komminoth P, Pacak K, Hop WC, Pollard PJ, Mannelli M, Bayley JP, Perren A, Niemann S, Verhofstad AA, de Bruïne AP, Maher de KRD. An immunohistochemical procedure to detect patients with paraganglioma and phaeochromocytoma with germline SDHB, SDHC, or SDHD gene mutations: a retrospective and prospective analysis. Lancet Oncol. 2009; 10: 764–71. 23. McCormick RI, Blick C, Ragoussis J, Schoedel J, Mole DR, Young AC, Selby PJ, Banks RE, Harris AL. miR-210 is a target of hypoxia-inducible factors 1 and 2 in renal cancer, regulates ISCU and correlates with good prognosis. Br J Cancer. 2013; 108: 1133–42. 24. Liu T, Zhao L, Chen W, Li Z, Hou H, Ding L, Li X. Inactivation of von Hippel-Lindau increases ovarian cancer cell aggressiveness through the HIF1α/miR-210/VMP1 signaling pathway. Int J Mol Med. 2014; 33: 1236–42. 25. López-Jiménez E, Gómez-López G, Leandro-García LJ, Muñoz I, Schiavi F, Montero-Conde C, de Cubas AA, Ramires R, Landa I, Leskelä S, Maliszewska A, Inglada-Pérez L, de la Vega L, et al. Research resource: Transcriptional profiling reveals different pseudohypoxic signatures in SDHB and VHL-related pheochromocytomas. Mol Endocrinol. 2010; 24: 2382–91. 26. Tsang VHM, Dwight T, Benn DE, Meyer-Rochow GY, Gill AJ, Sywak M, Sidhu S, Veivers D, Sue CM, Robinson BG, Clifton-Bligh RJ, Parker NR. Overexpression of miR- 210 is associated with SDH-related pheochromocytomas, paragangliomas, and gastrointestinal stromal tumours. Endocr Relat Cancer. 2014; 21: 415–26. 27. de Cubas AA, Leandro-García LJ, Schiavi F, Mancikova V, Comino-Méndez I, Inglada-Pérez L, Perez-Martinez M, Ibarz N, Ximénez-Embún P, López-Jiménez E, Maliszewska A, Letón R, Gómez Graña A, et al. Integrative analysis of miRNA and mRNA expression profiles in pheochromocytoma and paraganglioma identifies genotypespecific markers and potentially regulated pathways. Endocr Relat Cancer. 2013; 20: 477–93. 28. Sun F, Huo X, Zhai Y, Wang A, Xu J, Su D, Bartlam M, Rao Z. Crystal structure of mitochondrial respiratory membrane protein complex II. Cell. 2005; 121: 1043–57. 29. Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC, Schumacker PT. Mitochondrial reactive oxygen species trigger hypoxia-induced transcription. Proc Natl Acad Sci U S A. 1998; 95: 11715–20. 30. Martínez-Reyes I, Diebold LP, Kong H, Schieber M, Huang H, Hensley CT, Mehta MM, Wang T, Santos JH, Woychik R, Dufour E, Spelbrink JN, Weinberg SE, et al. TCA Cycle and Mitochondrial Membrane Potential Are Necessary for Diverse Biological Functions. Mol Cell. 2016; 61: 199–209. 31. Bender BU, Gutsche M, Gläsker S, Müller B, Kirste G, Eng C, Neumann HP. Differential genetic alterations in von Hippel-Lindau syndrome-associated and sporadic pheochromocytomas. J Clin Endocrinol Metab. 2000; 85: 4568–74. 32. Friedrich CA. Genotype-phenotype correlation in von Hippel-Lindau syndrome. Hum Mol Genet. 2001; 10: 763–7. 33. Brauch H, Weirich G, Brieger J, Glavac D, Rödl H, Eichinger M, Feurer M, Weidt E, Puranakanitstha C, Neuhaus C, Pomer S, Brenner W, Schirmacher P, et al. VHL alterations in human clear cell renal cell carcinoma: association with advanced tumor stage and a novel hot spot mutation. Cancer Res. 2000; 60: 1942–8. 34. Gao Y, Zhang Y, Zhang D, Dai X, Estelle M, Zhao Y. Auxin binding protein 1 (ABP1) is not required for either auxin signaling or Arabidopsis development. Proc Natl Acad Sci U S A. 2015 [cited 2016 Oct 28]; 112: 2275–80. 35. De Souza AT, Dai X, Spencer AG, Reppen T, Menzie A, Roesch PL, He Y, Caguyong MJ, Bloomer S, Herweijer H, Wolff JA, Hagstrom JE, Lewis DL, et al. Transcriptional and phenotypic comparisons of Ppara knockout and siRNA knockdown mice. Nucleic Acids Res. 2006; 34: 4486–94. 36. Law SHW, Sargent TD. The serine-threonine protein kinase PAK4 is dispensable in zebrafish: identification of a morpholino-generated pseudophenotype. PLoS One. 2014 [cited 2016 Oct 28]; 9: e100268. 37. Kok FO, Shin M, Ni C-W, Gupta A, Grosse AS, van Impel A, Kirchmaier BC, Peterson-Maduro J, Kourkoulis G, Male I, DeSantis DF, Sheppard-Tindell S, Ebarasi L, et al. Reverse genetic screening reveals poor correlation between morpholino-induced and mutant phenotypes in zebrafish. Dev Cell. 2015; 32: 97–108. 38. Rossi A, Kontarakis Z, Gerri C, Nolte H, Hölper S, Krüger M, Stainier DYR. Genetic compensation induced by deleterious mutations but not gene knockdowns. Nature. 2015; 524: 230–3. 39. Kanno H, Saljooque F, Yamamoto I, Hattori S, Yao M, Shuin T, U HS. Role of the von Hippel-Lindau tumor suppressor protein during neuronal differentiation. Cancer Res. 2000; 60: 2820–4. 40. Gossage L, Eisen T, Maher ER. VHL, the story of a tumour suppressor gene. Nat Rev Cancer. 2015; 15: 55–64. 41. Leisz S, Schulz K, Erb S, Oefner P, Dettmer K, Mougiakakos D, Wang E, Marincola FM, Stehle F, Seliger B. Distinct von Hippel-Lindau gene and hypoxia-regulated alterations in gene and protein expression patterns of renal cell carcinoma and their effects on metabolism. Oncotarget. 2015; 6: 11395–406. doi: 10.18632/oncotarget.3456. 42. Limaverde-Sousa G, Barreto E de A, Ferreira CG, Casalida-Rocha JC. Simulation of the mutation F76del on the von Hippel-Lindau tumor suppressor protein: mechanism of the disease and implications for drug development. Proteins. 2013; 81: 349–63. 43. Clifford SC, Cockman ME, Smallwood AC, Mole DR, Woodward ER, Maxwell PH, Ratcliffe PJ, Maher ER. Contrasting effects on HIF-1alpha regulation by Oncotarget18 www.impactjournals.com/oncotarget disease-causing pVHL mutations correlate with patterns of tumourigenesis in von Hippel-Lindau disease. Hum Mol Genet. 2001; 10: 1029–38. 44. Rathmell WK, Hickey MM, Bezman NA, Chmielecki CA, Carraway NC, Simon MC. In vitro and in vivo models analyzing von Hippel-Lindau disease-specific mutations. Cancer Res. 2004; 64: 8595–603. 45. Costa MHS, Ortiga-Carvalho TM, Violante AD, Vaisman M. Pheochromocytomas and Paragangliomas: Clinical and Genetic Approaches. Front Endocrinol (Lausanne). 2015; 6: 126. 46. Hoebeeck J, van der Luijt R, Poppe B, De Smet E, Yigit N, Claes K, Zewald R, de Jong G-J, De Paepe A, Speleman F, Vandesompele J. Rapid detection of VHL exon deletions using real-time quantitative PCR. Lab Invest. 2005; 85: 24–33. 47. Cramer T, Yamanishi Y, Clausen BE, Förster I, Pawlinski R, Mackman N, Haase VH, Jaenisch R, Corr M, Nizet V, Firestein GS, Gerber HP, Ferrara N, et al. HIF-1alpha is essential for myeloid cell-mediated inflammation. Cell. 2003; 112: 645–57. 48. Haase VH, Glickman JN, Socolovsky M, Jaenisch R. Vascular tumors in livers with targeted inactivation of the von Hippel-Lindau tumor suppressor. Proc Natl Acad Sci U S A. 2001; 98: 1583–8. 49. Balordi F, Fishell G. Mosaic removal of hedgehog signaling in the adult SVZ reveals that the residual wild-type stem cells have a limited capacity for self-renewal. J Neurosci. 2007; 27: 14248–59. 50. Gong L, Zhang W, Zhou J, Lu J, Xiong H, Shi X, Chen J. Prognostic value of HIFs expression in head and neck cancer: a systematic review. PLoS One. 2013; 8: e75094.