Purpose: Mutations in Krebs cycle genes are frequently found in patients with pheochromocytomas/paragangliomas. Disruption of SDH, FH or MDH2 enzymatic activities lead to accumulation of specific metabolites, which give rise to epigenetic changes in the genome that cause a characteristic hypermethylated phenotype. Tumors showing this phenotype, but no alterations in the known predisposing genes, could harbor mutations in other Krebs cycle genes.

Experimental Design: We used downregulation and methylation of RBP1, as a marker of a hypermethylation phenotype, to select eleven pheochromocytomas and paragangliomas for targeted exome sequencing of a panel of Krebs cycle-related genes. Methylation profiling, metabolite assessment and additional analyses were also performed in selected cases.

Results: One of the 11 tumors was found to carry a known cancer-predisposing somatic mutation in IDH1. A variant in GOT2, c.357A>T, found in a patient with multiple tumors, was associated with higher tumor mRNA and protein expression levels, increased GOT2 enzymatic activity in lymphoblastic cells, and altered metabolite ratios both in tumors and in GOT2 knockdown HeLa cells transfected with the variant. Array methylation-based analysis uncovered a somatic epigenetic mutation in SDHC in a patient with multiple pheochromocytomas and a gastrointestinal stromal tumor. Finally, a truncating germline IDH3B mutation was found in a patient with a single paraganglioma showing an altered α-ketoglutarate/isocitrate ratio.

Conclusions: This study further attests to the relevance of the Krebs cycle in the development of PCC and PGL, and points to a potential role of other metabolic enzymes involved in metabolite exchange between mitochondria and cytosol. Clin Cancer Res; 23(20); 6315–24. ©2017 AACR.

Translational Relevance

Mutations in pheochromocytoma (PCC)/paraganglioma (PGL) predisposition genes encoding enzymes involved in the Krebs cycle lead to a particular CpG island hypermethylation phenotype in these tumors. We used selected tumors with this phenotype for targeted exome sequencing of a panel of Krebs cycle-related genes to find new PCC/PGL susceptibility genes. Apart from known cancer-predisposing somatic alterations in IDH1 and SDHC, we found a germline variant in GOT2. This variant increased GOT2 activity and expression, and caused alterations in Krebs cycle metabolite ratios, both in tumors and in GOT2 KD cells transfected with the variant, suggesting that GOT2 is involved in PCC/PGL predisposition. In addition, we found a truncating germline IDH3B mutation in a patient with a PGL showing an altered α-ketoglutarate/isocitrate ratio. This study further attests to the relevance of the Krebs cycle in the development of PCC/PGL, some of which might be effectively treated with DNA-demethylating agents.

Despite their low prevalence, pheochromocytomas (PCC) and paragangliomas (PGL) represent a paradigm for hereditary cancer due to the highest degree of heritability of these tumors among all human neoplasms (1). Up to 40% of individuals with PCC and/or PGL have a hereditary background due to germline mutations affecting one of 13 susceptibility genes (2–13). The mutations result in two broad groups of tumors characterized by either activation of hypoxia or kinase receptor signaling pathways (1). Furthermore, somatic mutations in several of these genes or in two other PCC/PGL predisposition genes, HIF2A and HRAS (14, 15), are found in an additional approximately 30% of tumors (16). Finally, the presence of features of heritability amongst some of the patients without germline mutations in the known susceptibility genes, strongly suggests the implication of additional genes in this multigenetic disease. Of note, more than half of the PCC/PGL predisposition genes encode enzymes involved in the Krebs cycle (SDHA, SDHB, SDHC, SDHD, SDHAF2, FH, and MDH2), which stresses the crucial role of this pathway in PCC/PGL development and suggests that alterations in other genes from this pathway could account for more patients. This is even more likely for cases showing a noradrenergic or dopaminergic phenotype, which is associated with the presence of mutations in the aforementioned Krebs cycle-related genes.

In recent years, the application of high-throughput “omics” technologies to PCC/PGL research has led to improved understanding of the biology of these tumors, and paved the way for discovery of new predisposing genes (7, 12, 17). It has been known for years that pseudo-hypoxic PCCs/PGLs, such as SDH- and VHL-mutated tumors, share a similar transcriptional profile (18) that can be further segregated into different gene expression signatures (19). More recently, high-throughput methylation studies performed in PCCs/PGLs have revealed a particular CpG island methylation phenotype (CIMP) associated with the presence of Krebs cycle gene (SDH, FH, or MDH2) mutations (11, 12). This CIMP profile is caused by impaired histone demethylation and 5-mC hydroxylation (5-hmC) due to the enzymatic inhibition of multiple α-KG–dependent dioxygenases by the accumulation of succinate and fumarate (20). In addition, mutations in other Krebs cycle-related genes (IDH1 and IDH2) also lead to a similar CIMP in gliomas (21), caused in this case by the accumulation of D-2-hydroxyglutarate (22). IDH1 has been also found to be somatically mutated in two PGLs (23, 24). Though the majority of genes that define the altered methylation profile in PCC/PGL and gliomas are different, a substantial percentage of them (∼25%) are silenced in both neoplasias (12). That is the case for RBP1, which is consistently methylated and downregulated in both types of tumors carrying Krebs cycle mutations (11, 25). Tumors carrying mutations in other PCC/PGL susceptibility genes, such as RET, NF1, MAX or TMEM127, all belonging to the so-called expression cluster 2, do not exhibit methylation of RBP1 (11).

In the present study, we use the low expression of RBP1 as a marker of hypermethylation to select PCCs and PGLs as candidates for targeted exome sequencing of a panel of Krebs cycle-related genes to find new PCC/PGL predisposition genes.

Samples

Forty-nine tumors from patients showing noradrenergic or dopaminergic catecholamine phenotypes or no evidence of catecholamine production (when available), and testing negative for mutations in the major susceptibility genes (including all known PCC/PGL Krebs cycle-related genes: SDH, FH and MDH2), were included in the study. Immunostaining of SDHB was performed when formalin-fixed paraffin-embedded (FFPE) tissues were available, to rule out hidden mutations affecting the SDH genes, because negative SDHB immunohistochemistry is associated with the presence of SDH gene mutations in PCC/PGL (26, 27). The Instituto de Salud Carlos III (ISCIII) ethics committee approved the study, and all the patients provided written informed consent.

Quantitative real-time PCR

Total RNA was obtained from FFPE or from frozen material using the RNeasy FFPE (Qiagen) or TriReagent (MRC) Kit, respectively, according to the manufacturers' instructions. cDNA was prepared from 1 μg of total RNA using oligo (dT) primers and SuperScript III RT (Invitrogen). RBP1, SDHC and GOT2 mRNA levels were determined by quantitative PCR on a 7500 fast real-time PCR system (Applied Biosystems) using the Universal ProbeLibrary set (https://lifescience.roche.com/en_us.html), as described by the manufacturer. Relative mRNA levels were estimated by the 2-Ct method (28) and normalized using β-actin (ACTB) as housekeeping gene. The results are shown as mean + SD (n ≥ 3). mRNA obtained from tumors carrying mutations in other known pheochromocytoma susceptibility genes were used as controls. Tumors with RBP1 expression below 15 relative units (RU; n = 14) were considered candidates for subsequent analyses.

Pyrosequencing

Specific primers were used for PCR amplification and sequencing using the PyroMark assay (design version 2.0.01.15) to interrogate the methylation status of CpG sites from RBP1, as previously described (11).

Targeted next-generation sequencing

DNA extracted from 11 selected tumor samples was sequenced for a set of genes involved in the Krebs cycle by TruSight sequencing technology (Illumina), which comprises oligo probes targeting the genes of interest. The makeup of the panel was planned to cover the complete coding region of 37 human genes found to be directly or indirectly involved in the Krebs cycle based on the KEGG (http://www.genome.jp/kegg/pathway.html) and Genecard (http://www.genecards.org/) databases (Supplementary Table S1). Designstudio software (Illumina) was used to design the 733 amplicons included in the panel (cumulative Target: 65525 bp and coverage of 99%). This tool avoids designing for amplification primers that include known polymorphisms. Once the library was prepared following the manufacturer's instructions, next-generation sequencing was performed using MiSeq desktop sequencer (Illumina) and sequence alignment was carried out using MiSeq Reporter and Illumina VariantStudio softwares (Illumina). Variant calling was performed using GATK and Somatic Variant Caller, and identified variants were filtered considering mapping quality, variant score, depth, strand bias and annotation quality. To establish a cutoff value to consider a nucleotide substitution reported in controls as a candidate pathogenic variant, we used publically available data for SDHB germline mutations because they have the lowest penetrance amongst mutations in the known PCC/PGL susceptibility genes (29). The highest frequency for a known pathogenic SDHB mutation, found in the Genome Aggregation Consortium (gnomAD) database (http://gnomad.broadinstitute.org/), was 1.805 × 10−5, which was therefore applied as the cutoff value. Targeted regions without appropriate coverage and quality or with low mappability were re-amplified by Sanger sequencing. The PredictSNP consensus classifier (30) was used to predict the effect of the substitutions that passed all filtering steps. Tumor DNA samples from a series of 63 unselected patients with PCC or PGL, and negative for germline mutations in the known susceptibility genes, were tested by targeted next-generation sequencing for the presence of mutations in TCA-related genes.

Methylation assay and data processing

DNA was extracted and purified from the 11 selected cases, and from five SDH gene-mutated controls (two SDHB, one SDHA, one SDHD, and one SDHC-mutated tumors), using the DNeasy Blood & Tissue Kit (Qiagen) according to manufacturer's recommendations. Bisulfite conversion of DNA was performed using the EZ DNA Methylation Kit (Zymo Research) and genome-wide DNA methylation was assayed using the Infinium HumanMethylation450 BeadChip (Illumina) at the Centro Nacional de Genotipado (CEGEN-ISCIII), as previously described (31). This BeadChip interrogates more than 485,000 methylation sites per sample. Beta values for each interrogated CpG were assigned using the Genome Studio Methylation module. Methylomes from the 19 tumors were profiled together with 13 additional controls (four non-SDH gene-mutated and nine SDH gene-mutated tumors) obtained from the GEO database (accession number GSE43298). We used the clustering average for linkage and City-block as the distance measure, and assumed a standard deviation of 0.22.

Immunohistochemistry

Immunohistochemical SDHB (HPA002868, Sigma) analysis of 3-μm formalin-fixed paraffin-embedded (FFPE) tissues was used for all cases, where tissue was available, to further rule out the presence of SDH gene mutations, as previously reported (26). Specific immunohistochemical staining of GOT2 (NBP1-80521, Novus) was performed using FFPE sections from the two available GOT2-mutated tumors, following standard procedures; five tumors carrying mutations in PCC susceptibility genes were used as controls. Finally, immunohistochemical staining of 5-hmc (Active Motif; 39770) was performed using 3-μm FFPE sections from the IDH3B-mutated tumors, following standard procedures.

Enzymatic activity analyses

Cytosolic and mitochondrial fractions were obtained after processing of lymphoblastoid cells obtained from the GOT2-mutation carrier and from three controls, as previously described (32). Protein concentration was determined using Bradford method with BSA as standard. The cytosolic and mitochondrial fractions were used to determine the enzymatic activity of GOT1 and GOT2, quantified as a decrease in NADH fluorescence, after the addition of aspartate, using a BMG plate reader, as previously described (33). Control and mutant GOT activity in cell fractions were measured in duplicate at different dilutions in 3 independent paired experiments. Lactate dehydrogenase (LDH, measured as the increase in NADH caused by lactate addition) and citrate synthase (CS, measured as the increase in absorption of 5,5′-dithiobis-2-nitrobenzoic acid after the addition of OAA) were assayed using standard procedures (33, 34). GOT2 activity was corrected by subtracting the estimated cytosolic contamination (% of LDH present) from the activity measured in the mitochondrial fraction.

Mutagenesis

GOT2 c.357A>T and c.223T>G variants and the OGDHL c.750G>T substitution were generated by mutagenesis using the QuikChange II Site-Directed Mutagenesis Kit (Stratagene), according to the manufacturer's instructions. Wild-type and mutagenized inserts were verified by sequencing both strands.

Cell cultures

Lymphocytes from the carrier of the c.357A>T GOT2 mutation were immortalized using Epstein-Barr virus (EBV), as previously described (35). HeLa cell line (kindly provided by Flow Cytometry Core Unit, CNIO, Madrid) was authenticated by short tandem repeat profiling (GenePrint 10 System, Promega) and periodically confirmed to be Mycoplasma free by qPCR. HeLa cells, used within 10 passages after authentication, were cultured in Dulbecco's modified Eagle medium Gluta MAX (DMEM; Invitrogen), supplemented with 10% (v/v) FBS (PAA laboratories), 1% (v/v) penicillin/streptomycin, and 0.6% (v/v) Fungizone (Gibco), and maintained at 37°C in a humidified incubator with 5% CO2. MISSION shRNA lentiviral transduction particles were used to specifically knockdown GOT2 (TRCN0000034827; Sigma) and OGDHL (TRCN0000426732; Sigma). Stable gene knockdown (KD) was established by cellular resistance to puromycin (1 μg/mL). Scrambled non-target shRNA control vectors served as negative controls. HeLa (GOT2 or OGDHL) KD cells were seeded at 550,000 cells per well on 6-well plates for 24 hours. Each well was transiently transfected with 2 μg of pCMV6-AC plasmid containing the full cDNA sequence of GOT2 (SC127269, Origene), or OGDHL (RC205225, Origene), or with 2 μg of pCMV6-AC plasmids carrying the corresponding variant (GOT2 c.357A>T and c.223T>G, or OGDHL c.750G>T) using Lipofectamine 2000 (Invitrogen), according to the manufacturer's recommendations. pCMV6-AC empty vector (EV) was used as a control. In the proliferation assay, GOT2 KD HeLa cells were transfected with EV, GOT2-WT or GOT2-c.357A>T cDNA, and were seeded 24 hours post-transfection at 30,000 cells per well on 12-well plates. Then, cells were trypsinized and counted after 48, 72, and 96 hours post-transfection.

Western blot

To demonstrate the absence of full-length GOT2 or OGDHL proteins in KD HeLa cells, we performed Western blot analysis using polyclonal rabbit anti-GOT2 (NBP1-80521; Novus) and anti-OGDHL (NBP1-91486; Novus) antibodies. Proteins were separated by 10% SDS-PAGE and transferred to a polyvinylidene fluoride membrane. The membrane was blocked and then incubated with a 1:2,000 or a 1:1,000 dilution of GOT2 and OGDHL, respectively, following the manufacturer's instructions. Equal protein loading was assessed using a 1:12,000 dilution of monoclonal anti–β-actin mouse antibody (A5441, Sigma-Aldrich).

Liquid chromatographic tandem-mass spectrometric determination of metabolites

Fresh frozen or FFPE tumor tissue (5–10 mg) from available selected cases and from 15 controls, was immersed in 500 μL LC/MS grade methanol containing isotope-labeled internal standards and processes as previously described (11). For cell cultures, 500,000 cells were cultured, washed 4 times with PBS and extracted in methanol containing isotope-labeled internal standards. Analysis of metabolites was carried out using an AB Sciex 5500 QTRAP mass spectrometer coupled to an Acquity ultra–high-performance liquid chromatographic system (Waters), as previously described (11).

Clinical and molecular features of cases showing downregulation and methylation of RBP1

Fourteen out of the 49 tumors analyzed by qPCR showed low (lower than 15 RU) or almost no RBP1 expression levels compared with the remaining 35 assessed tumors and 10 control tumors carrying mutations in PCC/PGL susceptibility genes (Supplementary Fig. S1A). Pyrosequencing analysis confirmed gene methylation as the cause of the RBP1 downregulation in 11 out of the 14 tumors, and therefore further analysis were focused on these 11 cases. Five of the 11 patients had multiple PCCs/PGLs, one of them suffering also from a gastrointestinal stromal tumor (GIST), and one case showed distant metastases (Table 1). Nine cases showed a positive immunostaining for SDHB that ruled out the presence of SDH gene causal mutations in these patients. Only one H&N PGL developed by the patient who also suffered from GIST tested negative (Supplementary Fig. S1B), suggesting that a hidden SDH gene alteration could account for this patient.

Table 1.

Clinical and genetic findings of studied cases

IDTumor_ 1Tumor_ 2Tumor_ 3Tumor_ 4Tumor_ 5Tumor_ 6Tumor_ 7Tumor_ 8Tumor_ 9Tumor_ 10Tumor_ 11
Gene(s) altered IDH1 GOT2 GOT2 OGDHL PCK2 — — — — — — — — 
DNA variant c.394C>T c.223T>G c.357A>T c.750G>T c.1742C>T — — — — — — — — 
Protein change p.Arg132 Cys p.Tyr75 Asp p.Glu119Asp p.Arg250 Ser p.Ala581 Val — — — — — — — — 
Frequency in gnomAD Not found 3/246192 2/277174 Not found 2/246224 — — — — — — — — 
Frequency in Spaniardsa 0/816 0/816 0/816 0/816 0/816 — — — — — — — — 
PredictSNP Del Del Neutral Neutral Del — — — — — — — — 
Methylation cluster CIMP CIMP non-CIMP CIMP CIMP CIMP non-CIMP CIMP CIMP CIMP CIMP 
SDHB IHC Positive Positive Positive Negative Positive Positive Positive Positive NA Positive Positive 
Metabolites S/F S/F — — — — — NA — — 
Location TAP TAP PCC H&N H&N H&N H&N H&N H&N H&N TAP 
Biochemical Phenotype NM NM NM, D NS NM NS NS NM 
Number of tumors M (n = 9) M (n = 3), GIST M (n = 2) M (n = 5) M (n = 3) 
Clinical behavior Mg 
IDTumor_ 1Tumor_ 2Tumor_ 3Tumor_ 4Tumor_ 5Tumor_ 6Tumor_ 7Tumor_ 8Tumor_ 9Tumor_ 10Tumor_ 11
Gene(s) altered IDH1 GOT2 GOT2 OGDHL PCK2 — — — — — — — — 
DNA variant c.394C>T c.223T>G c.357A>T c.750G>T c.1742C>T — — — — — — — — 
Protein change p.Arg132 Cys p.Tyr75 Asp p.Glu119Asp p.Arg250 Ser p.Ala581 Val — — — — — — — — 
Frequency in gnomAD Not found 3/246192 2/277174 Not found 2/246224 — — — — — — — — 
Frequency in Spaniardsa 0/816 0/816 0/816 0/816 0/816 — — — — — — — — 
PredictSNP Del Del Neutral Neutral Del — — — — — — — — 
Methylation cluster CIMP CIMP non-CIMP CIMP CIMP CIMP non-CIMP CIMP CIMP CIMP CIMP 
SDHB IHC Positive Positive Positive Negative Positive Positive Positive Positive NA Positive Positive 
Metabolites S/F S/F — — — — — NA — — 
Location TAP TAP PCC H&N H&N H&N H&N H&N H&N H&N TAP 
Biochemical Phenotype NM NM NM, D NS NM NS NS NM 
Number of tumors M (n = 9) M (n = 3), GIST M (n = 2) M (n = 5) M (n = 3) 
Clinical behavior Mg 

Abbreviations: B, Benign; D, dopamine; Del, deleterious; GIST, gastrointestinal stromal tumor; H&N, head and neck PGL; H, high hydroxyglutarate levels; M, multiple; Mg, malignant; NA, not available; NM, normetanephrine; NS, predominant secretion not specified; PCC, pheochromocytoma; S/F, high succinate/fumarate ratio; S, single; TAP, thoracic-abdominal PGL; U, unknown.

aCIBERER Spanish exome database (http://csvs.babelomics.org/).

Next-generation sequencing findings

Five missense variants, affecting four genes and detected in three samples, passed the filtering process after excluding variants found at a frequency >2.471 × 10−5 of controls in gnomAD (Table 1). Sanger sequencing was used to validate both the presence of the variants in the corresponding tumor (Fig. 1A), and the presence/absence of the variant in the germline of the patients. All variants were found in the germline of the corresponding patient, with the exception of the c.394C>T IDH1 somatic mutation. Three variants were predicted to be deleterious and two neutral by PredictSNP (data not shown).

Figure 1.

A, Sanger sequencing validation (in tumor DNA) of targeted exome findings. B, Sanger sequencing validation (in tumor DNA) of the IDH3B mutation found in the extended series. C, Hierarchical clustering of the 11 pheochromocytomas/paragangliomas showing RBP1 mRNA downregulation, as well as SDH gene-mutated and non-SDH gene-mutated controls. Tumors hybridized and analyzed in the present study are denoted in black letters, whereas tumors downloaded from GEO are denoted in gray. Two clusters were observed and named CIMP (including all SDH gene-mutated controls), and non-CIMP (including all non-SDH gene-mutated controls). Tumor_3: tumor carrying the OGDHL and PCK2 variants; tumor_2: tumor carrying the c.357A>T GOT2 variant and tumor_1: tumor carrying the IDH1 and the c.223T>G GOT2 variants.

Figure 1.

A, Sanger sequencing validation (in tumor DNA) of targeted exome findings. B, Sanger sequencing validation (in tumor DNA) of the IDH3B mutation found in the extended series. C, Hierarchical clustering of the 11 pheochromocytomas/paragangliomas showing RBP1 mRNA downregulation, as well as SDH gene-mutated and non-SDH gene-mutated controls. Tumors hybridized and analyzed in the present study are denoted in black letters, whereas tumors downloaded from GEO are denoted in gray. Two clusters were observed and named CIMP (including all SDH gene-mutated controls), and non-CIMP (including all non-SDH gene-mutated controls). Tumor_3: tumor carrying the OGDHL and PCK2 variants; tumor_2: tumor carrying the c.357A>T GOT2 variant and tumor_1: tumor carrying the IDH1 and the c.223T>G GOT2 variants.

Close modal

Seven variants (six missense and one frameshift), affecting four genes and found in seven samples, were detected and validated in the extended series of 63 tumors. When we applied the same filtering process to that used for the selected series of tumors, only a novel IDH3B truncating mutation remained (Fig. 1B). The variant was found in the germline of a patient with a single non-secreting jugular PGL at age 51 years.

Methylation profiling classified most tumors within the CIMP cluster

Methylomes from 16 tumors (11 selected cases and five SDH gene-mutated controls) were profiled together with methylomes from 13 additional genotyped PCCs/PGLs obtained from GEO. Ten RBP1-downregulated and methylated tumors showed a “CIMP-like” profile similar to that observed for PCC/PGLs carrying SDH gene mutations, whereas only one case clustered with hypomethylated tumors carrying mutations in NF1, RET or MAX (Fig. 1C). When we focused on the methylation status of the Krebs cycle genes included in the panel, a tumor (Tumor_4 in Table 1) was found to show significantly higher methylation levels for 11 CpG targets encompassing the SDHC promoter region in 1q23.3 (Fig. 2A). To test for constitutional SDHC methylation in the patient, we analyzed DNA obtained from blood and found no silencing of the SDHC locus (Supplementary Fig. S1C). Absence of SDHC mRNA expression was demonstrated by qPCR for the tumor-exhibiting methylation of the SDHC promoter, compared to controls (Fig. 2B). Of note, and as mentioned above, the SDHC-methylated tumor was the only one from our series that showed negative SDHB immunohistochemistry. Finally, the methylome from the IDH3B-mutated tumor also showed a “CIMP-like” profile (data not shown), further supported by a negative 5-hmC immunohistochemistry (Supplementary Fig. S1D).

Figure 2.

A, DNA methylation (M_values) for 17 CpG island probes located within the SDHC locus in the 11 analyzed tumors, compared with IVD (in vitro methylated DNA). The results of the immunostaining for SDHB (SDHB IHC) are also represented. NA: not available. B, mRNA expression of SDHC in the SDHC-methylated tumor (tumor_4) compared with controls (n = 5). Error bars represent standard deviations.

Figure 2.

A, DNA methylation (M_values) for 17 CpG island probes located within the SDHC locus in the 11 analyzed tumors, compared with IVD (in vitro methylated DNA). The results of the immunostaining for SDHB (SDHB IHC) are also represented. NA: not available. B, mRNA expression of SDHC in the SDHC-methylated tumor (tumor_4) compared with controls (n = 5). Error bars represent standard deviations.

Close modal

Higher GOT2 expression and activity in tissues carrying the c.357A>T mutation

Because we found two different variants in GOT2 in two different cases, and both tumors were grouped within the “CIMP-like” cluster, we assessed the expression of the gene in the tumors. Two available tumors from the patient carrying the GOT2 c.357A>T variant (Fig. 3A) showed a significantly higher GOT2 mRNA expression (Fig. 3B), as well as a granular cytoplasmic differential immunostaining, compared with controls (Fig. 3C). Tumor_1 that carried the other GOT2 variant, c.223T>G, showed normal GOT2 expression. EBV-immortalized lymphoblastoid cells carrying the c.357A>T GOT2 mutation exhibited significantly higher GOT2 enzymatic activity compared with three GOT2–wild-type lymphoblastoid cell lines (Fig. 3D).

Figure 3.

A,GOT2 c.357A>T variant in the germline and in two tumors from the patient revealed by Sanger sequencing. The asterisk marks the tumor subjected to exome sequencing. B, GOT2 mRNA expression in two tumors (tumor_2A and tumor_2B) carrying the GOT2 c.357A>T variant, compared with controls (C1–C5). Expression level was normalized to β-actin (ACTB) and presented as mean and standard deviation (n ≥ 3). Error bars represent standard deviations. C, Immunohistochemical staining of GOT2 in two tumors carrying the GOT2 c.357A>T variant and in one control with WT-GOT2. Cytoplasmic aggregates were observed only in GOT2-mutated tumors. D, GOT2 activity measured in lymphoblastic GOT2-mutated (c.357A>T) and non-mutated cells (C1–C3).

Figure 3.

A,GOT2 c.357A>T variant in the germline and in two tumors from the patient revealed by Sanger sequencing. The asterisk marks the tumor subjected to exome sequencing. B, GOT2 mRNA expression in two tumors (tumor_2A and tumor_2B) carrying the GOT2 c.357A>T variant, compared with controls (C1–C5). Expression level was normalized to β-actin (ACTB) and presented as mean and standard deviation (n ≥ 3). Error bars represent standard deviations. C, Immunohistochemical staining of GOT2 in two tumors carrying the GOT2 c.357A>T variant and in one control with WT-GOT2. Cytoplasmic aggregates were observed only in GOT2-mutated tumors. D, GOT2 activity measured in lymphoblastic GOT2-mutated (c.357A>T) and non-mutated cells (C1–C3).

Close modal

Tumoral metabolite ratios

Liquid chromatographic tandem-mass spectrometric was applied to all available tissues to assess whether the tumors carrying the identified variants showed significant alterations in the metabolites of the Krebs cycle. Amongst the 10 cases tested (Tumor_9 could not be analyzed) only the tumors carrying the GOT2 c.357A>T variant (Tumor_2) and the OGDHL/PCK2 variations (Tumor_3), showed a high succinate/fumarate ratio (Fig. 4A) similar to that observed for SDH gene-mutated tumors. This abnormal succinate/fumarate ratio was not associated with SDH alterations in either of the two tumors, because they resulted positive for SDHB immunohistochemistry (Table 1). In addition, the tumor carrying the IDH1 c.394C>T mutation, showed an elevated hydroxyglutarate/isocitrate ratio compared to controls (n = 8; Fig. 4B), which is consistent with the presence of pathological IDH1/2 mutations. None of the other tumors (those not carrying these variants) had altered metabolite ratios. The IDH3B-mutated tumor showed an elevated α-ketoglutarate/isocitrate ratio compared to controls (Fig. 4C).

Figure 4.

A, Succinate/fumarate ratios assessed by liquid chromatographic tandem-mass spectrometry (LC/MS) in 13 tumors assayed in the present study compared with SDH gene-mutated (n = 9) and non-SDH gene-mutated controls (n = 6). Tumor_1: tumor carrying the IDH1 and the c.223T>G GOT2 variants; tumor_2: tumor carrying the c.357A>T GOT2 variant; tumor_3: tumor carrying the OGDHL and PCK2 variants; tumor_4: tumor showing DNA methylation of SDHC. The black lines represent medians. B, Hydroxyglutarate/isocitrate ratios assessed by LC/MS in nine tumors from the present study compared with SDH gene-mutated (n = 9) and non-SDH gene-mutated controls (n = 6). The IDH1-mutated tumor is indicated. The black lines represent medians. C, α-Ketoglutarate/isocitrate ratios assessed by LC/MS in the IDH3B-mutated tumor compared with 10 tumors from the present study, nine SDH gene-mutated cases and six non-SDH gene-mutated controls. The black lines represent medians. D, Aspartate/glutamate ratios assessed by LC-MS in GOT2 KD Hela cells transfected with empty vector (EV), GOT2 wild-type (WT) cDNA, and GOT2- c.357A>T cDNA. The ratios were reported as mean and standard deviation (n = 3). Error bars represent standard deviations. A t test was applied to test for differences between GOT WT and GOT2- c.357A>T-transfected cells. E, α-Ketoglutarate/citrate ratios assessed by LC-MS in GOT2 KD Hela cells transfected with EV, GOT2 WT cDNA, and GOT2- c.357A>T cDNA. The ratios were reported as mean and standard deviation (n = 3). Error bars represent standard deviations. A t test was applied to test for differences between GOT WT and GOT2- c.357A>T-transfected cells.

Figure 4.

A, Succinate/fumarate ratios assessed by liquid chromatographic tandem-mass spectrometry (LC/MS) in 13 tumors assayed in the present study compared with SDH gene-mutated (n = 9) and non-SDH gene-mutated controls (n = 6). Tumor_1: tumor carrying the IDH1 and the c.223T>G GOT2 variants; tumor_2: tumor carrying the c.357A>T GOT2 variant; tumor_3: tumor carrying the OGDHL and PCK2 variants; tumor_4: tumor showing DNA methylation of SDHC. The black lines represent medians. B, Hydroxyglutarate/isocitrate ratios assessed by LC/MS in nine tumors from the present study compared with SDH gene-mutated (n = 9) and non-SDH gene-mutated controls (n = 6). The IDH1-mutated tumor is indicated. The black lines represent medians. C, α-Ketoglutarate/isocitrate ratios assessed by LC/MS in the IDH3B-mutated tumor compared with 10 tumors from the present study, nine SDH gene-mutated cases and six non-SDH gene-mutated controls. The black lines represent medians. D, Aspartate/glutamate ratios assessed by LC-MS in GOT2 KD Hela cells transfected with empty vector (EV), GOT2 wild-type (WT) cDNA, and GOT2- c.357A>T cDNA. The ratios were reported as mean and standard deviation (n = 3). Error bars represent standard deviations. A t test was applied to test for differences between GOT WT and GOT2- c.357A>T-transfected cells. E, α-Ketoglutarate/citrate ratios assessed by LC-MS in GOT2 KD Hela cells transfected with EV, GOT2 WT cDNA, and GOT2- c.357A>T cDNA. The ratios were reported as mean and standard deviation (n = 3). Error bars represent standard deviations. A t test was applied to test for differences between GOT WT and GOT2- c.357A>T-transfected cells.

Close modal

c.357A>T GOT2 introduction in GOT2 KD cells increased metabolite ratios

To demonstrate the relevance of the c.357A>T variant in GOT2 enzymatic activity, we first silenced GOT2 (GOT2 KD) by shRNA in HeLa cells (Supplementary Fig. S2A). Subsequent transient introduction of GOT2 cDNA carrying the variant c.357A>T in GOT2 KD cells triggered a slight increase in the succinate/fumarate ratio and a significant increase of α-ketoglutarate/citrate and aspartate/glutamate ratios, compared with GOT2 KD cells transfected with wild-type GOT2 cDNA (Fig. 4D and E). Moreover, the growth rate of GOT2 KD cells transfected with c.357A>T-GOT2 was higher (although not significantly so) than that observed for GOT2 KD cells transfected with WT-GOT2, and significantly higher than that observed for KD cells (Supplementary Fig. S2B). Furthermore, GOT2 KD cells transfected with the c.223T>G variant showed aspartate/glutamate ratios similar to the control (Supplementary Fig. S3A).

Depletion of OGDHL had no effect on metabolite ratios

No alteration in the succinate/fumarate ratio was observed after silencing OGDHL in HeLa cells by shRNA, or after transient introduction of the c.750G>T variant in KD cells (Supplementary Fig. S3B). Silencing PCK2 in HeLa cells was not possible either by shRNA or CRISPR-Cas9 technologies (data not shown).

PCC/PGL are paradigmatic for illustrating the importance of human genetics in cancer, not only because of the high degree of heritability of the tumors and involvement of a large number of genes (more than 13), but also because these tumors are the main manifestation of genetic alterations in SDHD. SDHD is a gene that went down in history as the first metabolic gene involved in the Krebs cycle and the respiratory chain whose mutations were associated with the development of cancer. Since then, mutations affecting six additional genes involved in the Krebs cycle energy pathway have been associated with the development of PCC/PGL and with a particular hypermethylator phenotype. Herein, we describe how the selection of samples based on the expression of a methylation marker, identifies PCCs/PGLs harboring candidate variants affecting other Krebs cycle-related genes. These results highlight both the relevance of this pathway in PCC/PGL development and the need for marker-based selection of samples for the discovery of cancer susceptibility genes to avoid genetic heterogeneity.

To date, only two studies have focused on the involvement of IDH1/2 in PCC/PGL development, yielding 1/365 and 0/104 tumors carrying a pathogenic mutation affecting IDH1 (23). Another IDH1 mutation has also been identified in one PGL among 173 samples from The Cancer Genome Atlas (http://www.cbioportal.org/index.do; ref. 24), which further confirms the low frequency of alterations affecting this gene in PCC/PGL. The two mentioned IDH1 mutations, as well as the one reported herein, affect the same amino acid (p.Arg132Cys) and were found in older patients (>61 years) with PGLs, which stresses the relevance of this cancer-prone alteration to extra-adrenal tumors. The accumulation of hydroxyglutarate observed in the tumor, and the absence of any other somatic or germline alteration in the known PCC/PGL susceptibility genes, confirmed the driver role of the IDH1 mutation in this case and explains its CIMP profile.

Glutamic-oxaloacetic transaminase 2 (GOT2) is a mitochondrial enzyme that plays a role in amino acid metabolism and the urea and Krebs cycles. GOT2 converts oxaloacetate into aspartate by transamination, with the consequent conversion of glutamate to α-ketoglutarate. In SDH-deficient cells, glutamine and glutamate are used to provide metabolic intermediates to the truncated Krebs cycle, after conversion to α-ketoglutarate by GOT2 or GLUD1 enzymes (36, 37). It has been recently described that lysine acetylation of GOT2 enhances the protein association between GOT2 and MDH2, stimulating the malate shuttle activity and thus promoting pancreatic cell proliferation and tumor growth in vivo (38). Mutant KRAS promotes the reprogramming of glutamine metabolism in pancreatic cancer through GOT1/GOT2-mediated transamination pathway (39).

Both the expression of GOT2 and its enzymatic activity found in c.357A>T-mutated tissues suggest an activating role of the mutation. Moreover, the high succinate/fumarate ratio observed in the GOT2-mutated tumor, as well as the slight increment of α-ketoglutarate compared to controls, could be explained by an excess of this latter metabolite due to the higher enzymatic activity of GOT2. Introduction of the mutated p.Glu119Asp variant, but not p.Tyr75Asp, in GOT2 KD HeLa cells recapitulates the accumulation of metabolites observed in tumors carrying the variant, and increases cell proliferation. Interestingly, GOT2 knockdown led to an accumulation of glutamate as well as to significantly lower levels of Krebs cycle metabolites fumarate and malate in melanoma cells (40). Thus, it seems plausible that a higher activity of GOT2 could lead to an increment of anaplerotic incorporation of α-ketoglutarate to the Krebs cycle, and hence to the oncogenic accumulation of succinate. The absence of GOT2 variants in unselected cases suggests their prevalence is low in PCC/PGL patients, which has also reported for other TCA-related genes (11, 41).

The presence of aberrant hypermethylation of SDHC has been reported to be a novel mechanism of tumor development in Carney triad (PGL, GIST, and pulmonary chondroma) patients (42). Moreover, epimutations in SDHC have also been reported in hypermethylated SDH-deficient GISTs (43), mainly associated with Carney triad. Interestingly, in both studies the aberrant methylation of SDHC occurred exclusively in females. Recently, epigenetic mutation of the SDHC promoter has been found in another female with two PGLs (44). As far as we know, the case reported in the present study is the first example of a SDHC epimutation affecting a patient with Carney-Stratakis syndrome (co-occurrence of PGL and GIST). Regarding the mechanism involved in silencing SDHC, a potential role of sex chromosome or hormone biology has been proposed (43). The patient described herein is a female, and this explanation could also account for this case. The complete absence of SDHC expression in the tumor suggested that, in addition to the observed hemimethylation, a second hit affecting the gene was also present. High-density SNP genotyping performed in the tumor showing SDHC methylation revealed no alterations affecting chromosome 1, where SDHC is located (data not shown). Furthermore, the tumor had a normal DNA copy-number profile, which is consistent with previous findings describing that DNA copy-number changes are infrequent in parasympathetic PGLs (45). The absence of high succinate/fumarate ratio in this tumor could be due to the lower reliability of measurements of these metabolites in H&N PGLs (46), presumably due to a higher content of stromal cells diluting the signal from tumor cells. The absence of aberrant hypermethylation in known susceptibility genes in the other tumors analyzed (Supplementary Fig. S4) suggests that this mechanism is extremely rare in PCC/PGL.

The altered succinate/fumarate ratio observed in a tumor carrying two rare variants in OGDHL and PCK2 suggested that either gene could be involved in the disease. Interestingly, the promoter of OGDHL is differentially methylated in different tissue types, and it is thought that inactivation of OGDHL can contribute to cervical tumorigenesis (47). On the other hand, the mitochondrial phosphoenolpyruvate carboxykinase (PCK2) is an enzyme involved in gluconeogenesis, converting oxaloacetate into phosphoenolpyruvate, which also has a cataplerotic function, maintaining metabolic flux through the Krebs cycle by removing excess oxaloacetate. It has been reported that PCK2 activation mediates an adaptive response to glucose depletion in lung cancer (48). Although we have ruled out a link between OGDHL depletion and an altered succinate/fumarate ratio, we were not able to demonstrate the relevance of PCK2 in PCC/PGL development.

Homozygous loss-of-function mutations in IDH3B have been found in two families with retinitis pigmentosa (49), and somatic mutations in IDH3B have been recently found in acute myelogenous leukemia (50). IDH3B encodes the beta subunit of NAD-specific isocitrate dehydrogenase 3 (IDH3), which is involved in the oxidation of isocitrate to α-ketoglutarate in the Krebs cycle. It has been demonstrated that IDH3 activity in lysates from cells carrying heterozygous truncating IDH3B mutations was only 24% of that observed in normal controls (49). In addition, the altered α-ketoglutarate/isocitrate ratio detected in the tumor carrying the truncating mutation, and the associated CIMP-like profile further suggest a causative role for this variant in PGL development.

In summary, based on our selection of cases for exome sequencing of Krebs cycle genes and whole-genome DNA methylation assessment, we have identified rare pathological alterations in known PCC/PGL susceptibility genes (SDHC and IDH1), and two new candidate genes possibly involved in the hereditary predisposition (GOT2 and IDH3B). This study further strengthens the evidence for the relevance of the Krebs cycle in the development of PCC and PGL, and points to other enzymes involved in metabolite exchange between mitochondria and cytosol.

No potential conflicts of interest were disclosed.

Conception and design: L. Remacha, M. Robledo, A. Cascón

Development of methodology: L. Remacha, I. Comino-Méndez, M. Currás-Freixes, R. Letón, R. Torres-Perez, S. Jiménez, L. Maestre, J. Satrústegui, G. Eisenhofer, A. Cascón

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Richter, L. Contreras, A. Galarreta, E. Honrado, S. Moran, M. Esteller, G. Eisenhofer, M. Robledo

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): I. Comino-Méndez, S. Richter, L. Contreras, G. Pita, A. Galarreta, R. Torres-Perez, M. Esteller, J. Satrústegui, G. Eisenhofer, A. Cascón

Writing, review, and/or revision of the manuscript: L. Remacha, S. Richter, G. Eisenhofer, M. Robledo, A. Cascón

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): I. Comino-Méndez, M. Currás-Freixes, M. Robledo

Study supervision: A. Cascón

This work was supported by the Fondo de Investigaciones Sanitarias project PI15/00783, FEDER 2014-2020 (to A. Cascón) and the Deutsche Forschungsgemeinschaft (grant RI 2684/1-1; to S. Richter). CEGEN-PRB2-ISCIII is supported by grant PT13/0001, ISCIII-SGEFI/FEDER.

1.
Dahia
PL
. 
Pheochromocytoma and paraganglioma pathogenesis: learning from genetic heterogeneity
.
Nat Rev Cancer
2014
;
14
:
108
19
.
2.
Astuti
D
,
Latif
F
,
Dallol
A
,
Dahia
PL
,
Douglas
F
,
George
E
, et al
Gene mutations in the succinate dehydrogenase subunit SDHB cause susceptibility to familial pheochromocytoma and to familial paraganglioma
.
Am J Hum Genet
2001
;
69
:
49
54
.
3.
Burnichon
N
,
Briere
JJ
,
Libe
R
,
Vescovo
L
,
Riviere
J
,
Tissier
F
, et al
SDHA is a tumor suppressor gene causing paraganglioma
.
Hum Mol Genet
2010
;
19
:
3011
20
.
4.
Baysal
BE
,
Ferrell
RE
,
Willett-Brozick
JE
,
Lawrence
EC
,
Myssiorek
D
,
Bosch
A
, et al
Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma
.
Science
2000
;
287
:
848
51
.
5.
Latif
F
,
Tory
K
,
Gnarra
J
,
Yao
M
,
Duh
FM
,
Orcutt
ML
, et al
Identification of the von Hippel-Lindau disease tumor suppressor gene
.
Science
1993
;
260
:
1317
20
.
6.
Mulligan
LM
,
Kwok
JB
,
Healey
CS
,
Elsdon
MJ
,
Eng
C
,
Gardner
E
, et al
Germ-line mutations of the RET proto-oncogene in multiple endocrine neoplasia type 2A
.
Nature
1993
;
363
:
458
60
.
7.
Qin
Y
,
Yao
L
,
King
EE
,
Buddavarapu
K
,
Lenci
RE
,
Chocron
ES
, et al
Germline mutations in TMEM127 confer susceptibility to pheochromocytoma
.
Nat Genet
2010
;
42
:
229
33
.
8.
Niemann
S
,
Muller
U
. 
Mutations in SDHC cause autosomal dominant paraganglioma, type 3
.
Nat Genet
2000
;
26
:
268
70
.
9.
Wallace
MR
,
Marchuk
DA
,
Andersen
LB
,
Letcher
R
,
Odeh
HM
,
Saulino
AM
, et al
Type 1 neurofibromatosis gene: identification of a large transcript disrupted in three NF1 patients
.
Science
1990
;
249
:
181
6
.
10.
Hao
HX
,
Khalimonchuk
O
,
Schraders
M
,
Dephoure
N
,
Bayley
JP
,
Kunst
H
, et al
SDH5, a gene required for flavination of succinate dehydrogenase, is mutated in paraganglioma
.
Science
2009
;
325
:
1139
42
.
11.
Cascon
A
,
Comino-Mendez
I
,
Curras-Freixes
M
,
de Cubas
AA
,
Contreras
L
,
Richter
S
, et al
Whole-exome sequencing identifies MDH2 as a new familial paraganglioma gene
.
J Natl Cancer Inst
2015
;
107
:
pii
:
djv053
.
12.
Letouze
E
,
Martinelli
C
,
Loriot
C
,
Burnichon
N
,
Abermil
N
,
Ottolenghi
C
, et al
SDH mutations establish a hypermethylator phenotype in paraganglioma
.
Cancer Cell
2013
;
23
:
739
52
.
13.
Yang
C
,
Zhuang
Z
,
Fliedner
SM
,
Shankavaram
U
,
Sun
MG
,
Bullova
P
, et al
Germ-line PHD1 and PHD2 mutations detected in patients with pheochromocytoma/paraganglioma-polycythemia
.
J Mol Med
2015
;
93
:
93
104
.
14.
Crona
J
,
Delgado Verdugo
A
,
Maharjan
R
,
Stalberg
P
,
Granberg
D
,
Hellman
P
, et al
Somatic mutations in H-RAS in sporadic pheochromocytoma and paraganglioma identified by exome sequencing
.
J Clin Endocrinol Metab
2013
;
98
:
E1266
71
.
15.
Zhuang
Z
,
Yang
C
,
Lorenzo
F
,
Merino
M
,
Fojo
T
,
Kebebew
E
, et al
Somatic HIF2A gain-of-function mutations in paraganglioma with polycythemia
.
N Engl J Med
2012
;
367
:
922
30
.
16.
Curras-Freixes
M
,
Inglada-Perez
L
,
Mancikova
V
,
Montero-Conde
C
,
Leton
R
,
Comino-Mendez
I
, et al
Recommendations for somatic and germline genetic testing of single pheochromocytoma and paraganglioma based on findings from a series of 329 patients
.
J Med Genet
2015
;
52
:
647
56
.
17.
Comino-Mendez
I
,
Gracia-Aznarez
FJ
,
Schiavi
F
,
Landa
I
,
Leandro-Garcia
LJ
,
Leton
R
, et al
Exome sequencing identifies MAX mutations as a cause of hereditary pheochromocytoma
.
Nat Genet
2011
;
43
:
663
7
.
18.
Dahia
PL
,
Ross
KN
,
Wright
ME
,
Hayashida
CY
,
Santagata
S
,
Barontini
M
, et al
A HIF1alpha regulatory loop links hypoxia and mitochondrial signals in pheochromocytomas
.
PLoS Genet
2005
;
1
:
72
80
.
19.
Lopez-Jimenez
E
,
Gomez-Lopez
G
,
Leandro-Garcia
LJ
,
Munoz
I
,
Schiavi
F
,
Montero-Conde
C
, et al
Research resource: Transcriptional profiling reveals different pseudohypoxic signatures in SDHB and VHL-related pheochromocytomas
.
Mol Endocrinol
2010
;
24
:
2382
91
.
20.
Xiao
M
,
Yang
H
,
Xu
W
,
Ma
S
,
Lin
H
,
Zhu
H
, et al
Inhibition of alpha-KG-dependent histone and DNA demethylases by fumarate and succinate that are accumulated in mutations of FH and SDH tumor suppressors
.
Genes Dev
2012
;
26
:
1326
38
.
21.
Turcan
S
,
Rohle
D
,
Goenka
A
,
Walsh
LA
,
Fang
F
,
Yilmaz
E
, et al
IDH1 mutation is sufficient to establish the glioma hypermethylator phenotype
.
Nature
2012
;
483
:
479
83
.
22.
Dang
L
,
White
DW
,
Gross
S
,
Bennett
BD
,
Bittinger
MA
,
Driggers
EM
, et al
Cancer-associated IDH1 mutations produce 2-hydroxyglutarate
.
Nature
2009
;
462
:
739
44
.
23.
Gaal
J
,
Burnichon
N
,
Korpershoek
E
,
Roncelin
I
,
Bertherat
J
,
Plouin
PF
, et al
Isocitrate dehydrogenase mutations are rare in pheochromocytomas and paragangliomas
.
J Clin Endocrinol Metab
2010
;
95
:
1274
8
.
24.
Fishbein
L
,
Leshchiner
I
,
Walter
V
,
Danilova
L
,
Robertson
AG
,
Johnson
AR
, et al
Comprehensive molecular characterization of pheochromocytoma and paraganglioma
.
Cancer Cell
2017
;
31
:
181
93
.
25.
Chou
AP
,
Chowdhury
R
,
Li
S
,
Chen
W
,
Kim
AJ
,
Piccioni
DE
, et al
Identification of retinol binding protein 1 promoter hypermethylation in isocitrate dehydrogenase 1 and 2 mutant gliomas
.
J Natl Cancer Inst
2012
;
104
:
1458
69
.
26.
van Nederveen
FH
,
Gaal
J
,
Favier
J
,
Korpershoek
E
,
Oldenburg
RA
,
de Bruyn
EM
, et al
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
.
27.
Gill
AJ
,
Benn
DE
,
Chou
A
,
Clarkson
A
,
Muljono
A
,
Meyer-Rochow
GY
, et al
Immunohistochemistry for SDHB triages genetic testing of SDHB, SDHC, and SDHD in paraganglioma-pheochromocytoma syndromes
.
Hum Pathol
2010
;
41
:
805
14
.
28.
Livak
KJ
,
Schmittgen
TD
. 
Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method
.
Methods
2001
;
25
:
402
8
.
29.
Schiavi
F
,
Milne
RL
,
Anda
E
,
Blay
P
,
Castellano
M
,
Opocher
G
, et al
Are we overestimating the penetrance of mutations in SDHB?
Hum Mutat
2010
;
31
:
761
2
.
30.
Bendl
J
,
Stourac
J
,
Salanda
O
,
Pavelka
A
,
Wieben
ED
,
Zendulka
J
, et al
PredictSNP: robust and accurate consensus classifier for prediction of disease-related mutations
.
PLoS Comput Biol
2014
;
10
:
e1003440
.
31.
Bibikova
M
,
Le
J
,
Barnes
B
,
Saedinia-Melnyk
S
,
Zhou
L
,
Shen
R
, et al
Genome-wide DNA methylation profiling using Infinium(R) assay
.
Epigenomics
2009
;
1
:
177
200
.
32.
Trounce
IA
,
Kim
YL
,
Jun
AS
,
Wallace
DC
. 
Assessment of mitochondrial oxidative phosphorylation in patient muscle biopsies, lymphoblasts, and transmitochondrial cell lines
.
Methods Enzymol
1996
;
264
:
484
509
.
33.
Bergmeyer
HU
,
Bergmeyer
J
,
Grassl
M
.
Methods of enzymatic analysis / V. 3, Enzymes 1 oxidoreductases, transferases. 3rd ed. Weinheim: Verlag Chemie
; 
1983
.
34.
Bergmeyer
HU
,
Bergmeyer
J
,
Grabl
M
. 
Methods of enzymatic analysis
.
IV, Enzymes 2,. Esterases, glycosidases, lyases, ligases. 3rd ed. Weinheim [etc.]: Verlagchemie
; 
1984
.
35.
Ausubel
FM
. 
Current protocols in molecular biology
.
New York
:
Greene Pub. Associates; Wiley-Interscience
; 
1988
.
p. v. (loose-leaf)
.
36.
Saxena
N
,
Maio
N
,
Crooks
DR
,
Ricketts
CJ
,
Yang
Y
,
Wei
MH
, et al
SDHB-deficient cancers: the role of mutations that impair iron sulfur cluster delivery
.
J Natl Cancer Inst
2016
;
108
:
djv287
.
37.
Lussey-Lepoutre
C
,
Hollinshead
KE
,
Ludwig
C
,
Menara
M
,
Morin
A
,
Castro-Vega
LJ
, et al
Loss of succinate dehydrogenase activity results in dependency on pyruvate carboxylation for cellular anabolism
.
Nat Commun
2015
;
6
:
8784
.
38.
Yang
H
,
Zhou
L
,
Shi
Q
,
Zhao
Y
,
Lin
H
,
Zhang
M
, et al
SIRT3-dependent GOT2 acetylation status affects the malate-aspartate NADH shuttle activity and pancreatic tumor growth
.
EMBO J
2015
;
34
:
1110
25
.
39.
Son
J
,
Lyssiotis
CA
,
Ying
H
,
Wang
X
,
Hua
S
,
Ligorio
M
, et al
Glutamine supports pancreatic cancer growth through a KRAS-regulated metabolic pathway
.
Nature
2013
;
496
:
101
5
.
40.
Ratnikov
B
,
Aza-Blanc
P
,
Ronai
ZA
,
Smith
JW
,
Osterman
AL
,
Scott
DA
. 
Glutamate and asparagine cataplerosis underlie glutamine addiction in melanoma
.
Oncotarget
2015
;
6
:
7379
89
.
41.
Bayley
JP
,
Kunst
HP
,
Cascon
A
,
Sampietro
ML
,
Gaal
J
,
Korpershoek
E
, et al
SDHAF2 mutations in familial and sporadic paraganglioma and phaeochromocytoma
.
Lancet Oncol
2010
;
11
:
366
72
.
42.
Haller
F
,
Moskalev
EA
,
Faucz
FR
,
Barthelmess
S
,
Wiemann
S
,
Bieg
M
, et al
Aberrant DNA hypermethylation of SDHC: a novel mechanism of tumor development in Carney triad
.
Endocr Relat Cancer
2014
;
21
:
567
77
.
43.
Killian
JK
,
Miettinen
M
,
Walker
RL
,
Wang
Y
,
Zhu
YJ
,
Waterfall
JJ
, et al
Recurrent epimutation of SDHC in gastrointestinal stromal tumors
.
Sci Transl Med
2014
;
6
:
268ra177
.
44.
Richter
S
,
Klink
B
,
Nacke
B
,
deCubas
AA
,
Mangelis
A
,
Rapizzi
E
, et al
Epigenetic mutation of the succinate dehydrogenase C promoter in a patient with two paragangliomas
.
J Clin Endocrinol Metab
2016
:
101
:
359
63
.
45.
Dannenberg
H
,
de Krijger
RR
,
Zhao
J
,
Speel
EJ
,
Saremaslani
P
,
Dinjens
WN
, et al
Differential loss of chromosome 11q in familial and sporadic parasympathetic paragangliomas detected by comparative genomic hybridization
.
Am J Pathol
2001
;
158
:
1937
42
.
46.
Richter
S
,
Peitzsch
M
,
Rapizzi
E
,
Lenders
JW
,
Qin
N
,
de Cubas
AA
, et al
Krebs cycle metabolite profiling for identification and stratification of pheochromocytomas/paragangliomas due to succinate dehydrogenase deficiency
.
J Clin Endocrinol Metab
2014
;
99
:
3903
11
.
47.
Sen
T
,
Sen
N
,
Noordhuis
MG
,
Ravi
R
,
Wu
TC
,
Ha
PK
, et al
OGDHL is a modifier of AKT-dependent signaling and NF-kappaB function
.
PLoS One
2012
;
7
:
e48770
.
48.
Leithner
K
,
Hrzenjak
A
,
Trotzmuller
M
,
Moustafa
T
,
Kofeler
HC
,
Wohlkoenig
C
, et al
PCK2 activation mediates an adaptive response to glucose depletion in lung cancer
.
Oncogene
2015
;
34
:
1044
50
.
49.
Hartong
DT
,
Dange
M
,
McGee
TL
,
Berson
EL
,
Dryja
TP
,
Colman
RF
. 
Insights from retinitis pigmentosa into the roles of isocitrate dehydrogenases in the Krebs cycle
.
Nat Genet
2008
;
40
:
1230
4
.
50.
Garg
M
,
Nagata
Y
,
Kanojia
D
,
Mayakonda
A
,
Yoshida
K
,
Haridas Keloth
S
, et al
Profiling of somatic mutations in acute myeloid leukemia with FLT3-ITD at diagnosis and relapse
.
Blood
2015
;
126
:
2491
501
.

Supplementary data