Introduction

Cowden syndrome (CS, OMIM# 158350) is an underdiagnosed, difficult-to-recognize (affecting 1 in 200,000 individuals), autosomal dominant cancer syndrome with high penetrance. It is primarily associated with an increased risk of breast, follicular thyroid, and endometrial cancers; however, papillary thyroid cancer and renal cell carcinoma have also been reported.1,2,3 Clinical diagnosis of CS is made when an individual meets diagnostic criteria, a combination of pathognomonic major and/or minor criteria, established by the International Cowden Consortium.1 Many individuals meet partial diagnostic criteria for CS (defined as full criteria minus one) and are referred to as “CS-like.” Germline mutations or deletions in phosphatase and tensin homologue deleted on chromosome 10 (PTEN, OMIM +601728), a ubiquitously expressed tumor suppressor, have been identified in ~25% of individuals with CS, and somatic PTEN mutations have been variably observed in a large number of sporadic malignancies.2,4

Germline heterozygous mutations in succinate dehydrogenase (SDH) B subunit (SDHB) and SDH D subunit (SDHD) have been identified in ~10% of PTEN mutation-negative individuals with CS.5 SDHB and SDHD encode the B and D subunits of SDH, a Krebs-cycle enzyme that catalyzes oxidation of succinate to fumarate and also participates in the electron transport chain (complex II, succinate-ubiquinone oxidoreductase).5,6,7

Similar to CS, female breast cancer, papillary thyroid cancer, and renal cell carcinoma have been variably associated with individuals with germline heterozygous mutations in SDHB and SDHD.5,6,7 As with PTEN, the proteins associated with SDHx genes also function as tumor suppressors and mutations in these genes result in mitochondrial dysfunction and tumorigenesis via upregulation of angiogenic and hypoxic pathways.7,8 Mutations in SDHA, SDHB, SDHC, SDHD, and SDH complex assembly factor 2 (SDHAF2) underlie most cases of familial paraganglioma (PGL), giving rise to PGL syndromes type 4 (PGL-4, SDHB, OMIM# 115310), type 3 (PGL-3, SDHC, OMIM# 605373), type 1 (PGL-1, SDHD, OMIM# 168000), and type 2 (PGL-2, SDHAF2, OMIM# 601650), respectively.9,10,11 Although PGLs and pheochromocytomas both arise from paraganglial cells, PGLs are confined to the head and neck and pheochromocytomas are localized to adrenal glands and extra-adrenal abdominal and thoracic locations.6 This is how we have defined these terms; however, we recognize that these definitions may vary depending on group or country. In addition to SDHA, SDHB, SDHC, SDHD, and SDHAF2, four additional genes are associated with the development of hereditary PGL and/or pheochromocytoma, including RET (multiple endocrine neoplasia type 2, OMIM# 164761), VHL (associated with von Hippel-Lindau disease, OMIM# 193300), TMEM127 (encoding transmembrane spanning protein 127, associated with Golgi, endosomes, and lysosomes, and hypothesized to play a role in protein trafficking and reported to negatively regulate the target of rapamycin (TOR) signaling pathway, OMIM# 171300), and, rarely, NF1, encoding neurofibromatosis type 1 (OMIM# 162200).7,12

In addition to their recent association with PGL in its heterozygous state, germline homozygous or compound heterozygous SDHA mutations have more commonly been associated with Leigh syndrome (OMIM# 256000), a rare neurometabolic disorder. Succinate accumulation has been observed in SDHA mutant fibroblasts and in SDHB mutant tumor tissues,13,14 and elevated urinary succinate has been associated with, but is not specific to, mitochondrial disorders, hypoxia, and seizures. Therefore, our study was designed to ascertain whether or not common biochemical alterations could be identified in individuals with CS, CS-like, and SDHx-associated pheochromocytomas and PGLs. We hypothesized that elevated succinate could be measured in urine and plasma from patients with SDHx mutations and also in individuals with PTEN mutations meeting full or partial CS diagnostic criteria. A PubMed literature search failed to identify other studies that evaluated urine and plasma succinate in patients with tumors, suggesting that ours is the first such report.

Materials and Methods

Participants

Between October 2007 and February 2011, patients identified in the Center for Personalized Genetic Healthcare of the Genomic Medicine Institute at the Cleveland Clinic were recruited for study. Inclusion criteria included patients who met full operational diagnostic criteria for CS, partial criteria (full criteria minus one) for CS (termed CS-like), or a personal or family history of pheochromocytoma or PGL. Written informed consent was obtained from all participants. The study received ethical approval by the Cleveland Clinic Institutional Review Board for Human Subjects’ Protection.

Procedures

Genomic DNA was isolated and PTEN, SDHB, SDHC, and SDHD sequencing were performed based on patient phenotype using Light Scanner technology (Idaho Technology Inc., Salt Lake City, Utah). Multiplex ligation-dependent probe amplification was performed to identify PTEN, SDHB, SDHC, and SDHD gene duplications or deletions in select mutation-negative individuals.

Random urine and/or blood samples were obtained during routine patient visits and de-identified. Urine and plasma were aliquoted and frozen within 1 h of collection and were stored at −80 °C until organic acid analyses were performed. Organic acid concentrations were determined using gas chromatography-mass spectrometry in the Biochemical Genetics Laboratory of ARUP, Salt Lake City, UT. This lab was blinded to the mutation status and clinical diagnosis associated with the plasma and urine samples. Measured organic acids in urine included, but were not limited to lactic acid, pyruvic acid, succinic acid, fumaric acid, 2-ketoglutaric acid, methylmalonic acid, 3-hydroxybutyric acid, acetoacetic acid, 2-keto-3-methylvaleric acid, 2-ketoisocaproic acid, 2-ketoisovaleric acid, ethylmalonic acid, adipic acid, suberic acid, sebacic acid, 4-hydroxyphenylacetic acid, 4-hydroxyphenyllactic acid, 4-hydroxy-phenylpyruvic acid, and succinylacetone. Measured organic acids in plasma included, but were not limited to, lactic acid, pyruvic acid, succinic acid, 3-hydroxybutyric acid, acetoacetic acid, 2-keto-3-methylvaleric acid, 2-ketoisocaproic acid, 2-ketoisovaleric acid, and citric acid. Reference ranges for urine and plasma organic acids were established in an age-matched population by the Biochemical Genetics Laboratory at ARUP. Urinary organic acids were reported as mmol of acid/mole of creatinine and plasma values as µmol/l.

Statistical analysis

Comparison of frequencies was performed with the Fisher’s two-tailed exact test, with P < 0.05 considered statistically significant.

Results

A total of 66 patients (55 plasma samples; 65 urine samples) were enrolled in the study. All individuals presented to our cancer genetics clinic for an initial visit or follow-up care. Study participants whose phenotypes and family history were consistent with CS or CSL were screened for germline PTEN, SDHB, SDHC, and SDHD mutations. The majority of PTEN mutation–negative CS and CSL individuals were also assessed for PTEN duplications and deletions. We identified 15 PTEN mutation–positive individuals who met CS diagnostic criteria (PTEN mutation–positive), 15 PTEN mutation–negative individuals who met CS diagnostic criteria (PTEN mutation–negative, CS), 4 CSL individuals (PTEN mutation–negative, CSL), and 3 individuals with PTEN variants of unknown significance (PTEN VUS) ( Table 1 ).

Table 1 Summary of elevated urine and plasma succinate data
Full size table

Individuals who presented with PGL, or a family history of a known SDH mutation, were screened for germline SDHB, SDHC, SDHD, and PTEN mutations. Deletions and duplications in these genes were assessed in select mutation–negative individuals. In total, we enrolled 1 SDHB mutation–positive individual; 5 SDHD mutation–positive individuals; 10 individuals with PGL with no identifiable mutations, duplications, or deletions in SDHB, SDHC, or SDHD (SDH mutation–negative, PGL), 1 individual with a known VHL mutation, and 1 individual with a known TMEM127 single-nucleotide polymorphism ( Table 1 ).

Organic acid analyses revealed elevated plasma succinate in 13/21 (62%) individuals with germline mutations in any examined gene as compared with 5/32 (16%) mutation-negative controls (P < 0.001). The majority of PTEN mutation–positive individuals (10/15; 67%) had elevated plasma succinate; this finding was not observed in PTEN mutation–negative CS individuals (3/15; 20%) or the PTEN mutation–negative, CSL group (1/4; 25%) or in individuals with PTEN variants of unknown significance (1/3; 33%) ( Table 1 ).

Elevated plasma succinate was recorded in individuals with SDHB (1/1; 100%) and SDHD mutations (2/5; 40%), and in one individual harboring a TMEM127 single-nucleotide polymorphism (1/1; 100%). Elevated plasma succinate was not found in SDH mutation–negative individuals with PGL (0/10; 0%) or in one individual with a mutation in VHL (0/1; 0%) ( Table 1 ). Elevated urine succinate was observed in some (6/19, 32%), but not all, individuals with elevated plasma succinate (Supplementary Table S1 online). No other organic acids in plasma or urine were consistently elevated or decreased for any patient group (data not shown).

Discussion

This report demonstrates that elevated plasma succinate is a common finding in individuals with known pathogenic mutations in PTEN, SDHB, SDHD, and TMEM127. Previous studies have reported elevated succinate levels in tumor-derived tissue from patients with SDHB mutations and in SDHA mutant fibroblasts,13,14 but to the best of our knowledge, this is the first report demonstrating an elevation of succinate in plasma from patients with germline mutations in SDHB, SDHD, TMEM127, and PTEN. Although elevated plasma succinate levels might be expected for individuals with SDHx mutations, it was unexpected for PTEN mutation–positive individuals and implies that PTEN mutations somehow reduce the catalytic activity of the SDH protein complex.

Consistent with previous studies assessing succinate levels in tumor-derived tissue,14 more than half of our SDHD mutation–positive individuals (3/5, 60%) did not demonstrate elevated plasma succinate. Similarly, 33% of PTEN mutation–positive individuals did not exhibit elevated plasma succinate. Although the reason for this finding is unclear, our data suggest that elevated plasma succinate does not correlate with a specific mutation(s) or phenotype(s) (Supplementary Table S1 online). The conversion of succinate to fumarate, along with several other Krebs-cycle enzymatic reactions, is reversible. It is possible that some mutations and/or variants impair and/or enhance the reversibility of these reactions, thereby reducing succinate to normal levels. This is one possible explanation for the SDHD and PTEN mutation–positive individuals who do not exhibit elevated plasma succinate.

Elevated urine succinate was observed in some, but not all, individuals with elevated plasma succinate. The reason for this inconsistency is not entirely clear; however, it likely stems from the fact that the succinate elevations we observed in plasma samples were not severe but were mild to moderate in nature.

One of three patients harboring PTEN polymorphisms exhibited elevated succinate in both plasma and urine. This patient, unlike the other two, met full CS diagnostic criteria. It is conceivable that this intronic variant, c.210-7del5, although currently classified as a polymorphism, may actually be a pathogenic mutation leading to splicing defects.

One plausible explanation for the link between PTEN mutations and elevated plasma succinate is impaired PTEN-induced kinase 1 (PINK1), a mitochondrial localized serine-threonine kinase, transcriptionally activated by PTEN.15 Studies of PINK1 knockout mice showed impaired mitochondrial respiration in striatum, specifically; a significant decrease was seen in the state III activities for mitochondrial complex I and complex II.15 Because the oxidation of succinate to fumarate is coupled with the reduction of ubiquinone to ubiquinol, we expect that a reduction in complex II activities would simultaneously be associated with a reduction in the activity of SDH in the Krebs cycle. Therefore, mutations that affect the stability or activity of PTEN likely affect PINK1 transcription and downstream function of mitochondrial complex II.

Currently, individuals presenting to genetics clinics who meet CS diagnostic criteria or who present with familial PGL are offered the option of gene sequencing to establish the underlying cause of disease and to provide disease management. Estimated cost for PTEN sequencing, deletion, and duplication analysis is ~US$2,000 per sample.16 Likewise, cost for SDHB, SDHC, and SDHD mutation analyses are ~$1,000, $1,300, and $700, respectively. Clinical multiplex ligation-dependent probe amplification analysis for SDHB, SDHC, and SDHD is ~$550. In contrast, plasma organic acid analysis is a relatively low-cost assay. The cost of plasma organic acid analysis is ~$230.00 per sample.16 It is likely that the cost under contract to a medical institute would be even less. Therefore, based upon our finding that a large proportion of individuals with pathogenic PTEN, SDHB, and SDHD mutations exhibit elevated plasma succinate, we suggest that plasma organic acid analysis may be a useful and cost-effective preliminary screening tool for identifying individuals for whom more costly gene sequencing is warranted.

In conclusion, we have demonstrated that elevated plasma succinate is a common biochemical disturbance in the majority of PTEN, SDHB, and SDHD mutation–positive individuals and provides a plausible biochemical link for the shared phenotypic findings across these groups. Furthermore, screening for elevated succinate provides a rapid, inexpensive analytical tool for identifying individuals who are likely to harbor a germline mutation in PTEN, SDHB, or SDHD.

Disclosure

The authors declare no conflict of interest.