Spectrum and Prevalence of FP/TMEM127 Gene Mutations in Pheochromocytomas and Paragangliomas

ORIGINAL CONTRIBUTION

Li Yao, PhD Context Pheochromocytomas and paragangliomas are genetically heterogeneous neu- Francesca Schiavi, MD, PhD ral crest–derived neoplasms. We recently identified germline mutations of the novel Alberto Cascon, PhD transmembrane-encoding gene FP/TMEM127 in familial and sporadic pheochromocytomas consistent with a tumor suppressor effect. Yuejuan Qin, MD, PhD Objectives To examine the prevalence and spectrum of FP/TMEM127 mutations Lucia Inglada-Pe´rez, PhD in pheochromocytomas and paragangliomas and to test the effect of mutations in vitro. Elizabeth E. King, MD Design, Setting, and Participants We sequenced the FP/TMEM127 gene in 990 Rodrigo A. Toledo, PhD individuals with pheochromocytomas and/or paragangliomas, including 898 previously unreported cases without mutations in other susceptibility genes from 8 inde- Tonino Ercolino, PhD pendent worldwide referral centers between January 2009 and June 2010. A multi- Elena Rapizzi, PhD plex polymerase chain reaction–based method was developed to screen for large gene Christopher J. Ricketts, PhD deletions in 545 of these samples. Confocal microscopy of 5 transfected mutant proteins was used to determine their subcellular localization. Luigi Mori, PhD Main Outcome Measures The frequency and type of FP/TMEM127 mutation or Mara Giacchè, PhD deletion was assessed and correlated with clinical variables; the subcellular localization Antonella Mendola, PhD of 5 overexpressed mutants was compared with wild-type FP/TMEM127 protein. Elisa Taschin, PhD Results We identified 19 potentially pathogenic FP/TMEM127 germline mutations in 20 independent families, but no large deletions were detected. All mutation carriers Francesca Boaretto, PhD had adrenal tumors, including 7 bilateral (P=2.7ϫ10−4) and/or with familial disease Paola Loli, MD (5 of 20 samples; P=.005). The median age at disease onset in the FP/TMEM127 mutation group was similar to that of patients without a mutation (41.5 vs 45 years, re- Maurizio Iacobone, MD spectively; P=.54). The most common presentation was that of a single benign ad- Gian-Paolo Rossi, MD renal tumor in patients older than 40 years. Malignancy was seen in 1 mutation carrier Bernadette Biondi, MD (5%). Expression of 5 novel FP/TMEM127 mutations in cell lines revealed diffuse localization of the mutant proteins in contrast with the discrete multiorganelle distribu- Jose´ Viana Lima-Junior, MD tion of wild-type TMEM127. Claudio E. Kater, MD Conclusions Germline mutations of FP/TMEM127 were associated with pheochro- Marie Bex, MD mocytoma but not paraganglioma and occured in an age group frequently excluded from genetic screening algorithms. Disease-associated mutations disrupt intracellular Miikka Vikkula, PhD distribution of the FP/TMEM127 protein. Ashley B. Grossman, MD JAMA. 2010;304(23):2611-2619 www.jama.com Stephen B. Gruber, MD Marta Barontini, MD HEOCHROMOCYTOMAS AND geneous, with at least one-third of paragangliomas are chromaf- cases resulting from germline but not Alexandre Persu, MD fin cell tumors of neural crest somatic mutations in 1 of several Maurizio Castellano, MD origin that arise from the independent genes: RET, VHL, NF1, adrenalP medulla or extra-adrenal sym- and succinate dehydrogenase (SDH) Sergio P. A. Toledo, MD 2-5 Eamonn R. Maher, MD, FMedSci pathetic paraganglia, respectively, and subunit B, C, and D genes. More are frequently catecholamine secret- recently, other candidate susceptibil- Massimo Mannelli, MD ing.1 These tumors are usually benign Giuseppe Opocher, MD and can occur as a single entity or as Author Affiliations are listed at the end of this article. Mercedes Robledo, PhD part of various hereditary tumor syn- Corresponding Author: Patricia L. M. Dahia, MD, PhD, dromes. Genetically, pheochromocy- Department of Cellular and Structural Biology, Uni- Patricia L. M. Dahia, MD, PhD versity of Texas Health Science Center at San Anto- tomas and paragangliomas are hetero- nio, San Antonio, TX 78229 ([email protected]).

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

ity genes have also been reported, mined the subcellular localization of 5 tations in other known pheochromocy- including KIF1Bβ,6,7 EgIN1/PHD2,8 novel mutations by in vitro confocal toma susceptibility genes were ex- SDHAF2,9,10 and SDHA,11 although microscopy. cluded.14 The main features of each series these findings remain restricted to 1 of the entire cohort are shown in TABLE 1. or 2 occurrences of the reported METHODS mutations. Despite this broad spec- Patients DNA Isolation trum of susceptibility genes, the A total of 990 samples from patients with Of the new series of 898 samples, DNA molecular basis for the majority of pheochromocytomas and paraganglio- was isolated from blood in 774 cases pheochromocytomas and paragangliomas without mutations in RET, VHL, and from tumor tissue in 124 cases, fol- mas, including most of the sporadic SDHB, SDHC, and SDHD and who had lowing standard procedures.14 and rare familial cases, remains no clinical features of neurofibromato- unknown. These observations support sis type 1 were studied. In addition, 319 Control Population the existence of additional pheochro- of these samples were negative for A control group composed of samples of mocytoma susceptibility genes, which SDHAF2 and 52 for KIF1Bβ gene muta- Europeans, South Americans (heteroge- may account for some of the geneti- tions. A cohort of 898 samples were re- neous population mainly of Latino ori- cally undefined cases.12,13 cruited from a multi-institutional col- gin but also including non-Hispanic We recently identified TMEM12714 laborative effort derived from the whites, Asians, African Americans, and (NM_017849.3) as the pathogenic tar- International Familial Pheochromocy- indigenous Native Americans), North get of the familial pheochromocytoma toma Consortium, which encompassed Americans (including both Hispanic and (FP) locus, which we previously mapped referral centers based in the United States non-Hispanic whites and African Ameri- to chromosome 2q11.13 Germline (San Antonio, Texas, and Ann Arbor, cans), and Asians totaling 1064 alleles FP/TMEM127 mutations were found Michigan), Italy (Padova, Florence, and was used as reference samples, as re- both in familial and sporadic-appear- Brescia), Spain (Madrid), England (Bir- ported.14 In addition to these, 718 new ing pheochromocytomas, with loss of mingham and London), Brazil (Sa˜o alleles were screened for exon 3 of the the wild-type allele in tumor DNA con- Paulo), Belgium (Brussels), and Argen- FP/TMEM127 gene only. This group sistent with its role as a tumor suppres- tina (Buenos Aires). Written informed comprised 318 alleles of white ancestry sor gene. FP/TMEM127 encodes a highly consent was obtained from patients in ac- and 404 alleles of South American ori- conserved 3-spanner transmembrane cordance with institutional review gin similar in racial/ethnic composition protein that localizes to multiple intra- board–approved protocols from each to those described above. cellular organelles and is linked to regu- center. Eighteen samples were ob- lation of the mTORC1 signaling com- tained from the Cancer Therapy and Re- Polymerase Chain Reaction plex, a critical control node for protein search Center Pathology Tumor Bank at and Direct Sequencing synthesis and cell survival.14 All but 1 the University of Texas Health Science To define the prevalence of FP/TMEM127 of the 7 initially discovered mutations Center at San Antonio. Diagnosis of mutationinalarge,multi-institutionalco- were truncating and resulted in mark- pheochromocytoma and/or paragan- hortofpheochromocytomasandparagan- edly reduced expression of the glioma, including tumors of both sym- gliomas,all4exonsofFP/TMEM127were FP/TMEM127 gene in the evaluable tu- pathetic (thoracic or abdominal) and amplified by polymerase chain reaction mors, suggesting that the mutations re- parasympathetic (head and neck) ori- (PCR) and directly sequenced as previ- sult in loss of FP/TMEM127 function in gin, was established following conven- ously reported.14 Primer sequences the tumor tissue. tional procedures (including clinical, bio- are listed in eTable 1 (available at http: In recent years, specific genotype- chemical, and imaging tests) and the //www.jama.com). phenotype associations in pheochromo- diagnosis was confirmed histologically The accession numbers for nucleo- cytomas due to mutations in the vari- in every case. Tumors were considered tides (NM_017849.3) and respective ous susceptibility genes have spurred the familial when more than 1 affected in- protein variations (NP_060319.1) de- development of guidelines for genetic dividual was identified in the family. The tected in this study were deposited in screening of these patients, with a goal study was performed between January the National Center for Biotechnology of improving the outcomes of affected 2009 and June 2010. Information (NCBI) single-nucleotide and at-risk individuals.3,5,15,16 To define We previously reported FP/TMEM127 polymorphism (SNP) database (http: the prevalence and genotype-pheno- sequence variations in a cohort of 103 //www.ncbi.nlm.nih.gov/SNP/tranSNP type correlations in FP/TMEM127- patients with pheochromocytoma and/or /VarBatchSub.cgi). mutated cases, we sequenced the paraganglioma.14 In the present study, 92 FP/TMEM127 gene in 990 pheochromo- of these cases were used for genotype- Pathogenic Mutation Call cytoma and paraganglioma patients and phenotype associations with the 898 new Variations that included nonsynony- examined 545 of these samples for germ- cases described above. Eleven cases from mous sequence substitutions, inser- line deletions of the gene. We also deter- the original cohort that had germline mu- tions, deletions, duplications, or splice

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

Table 1. Summary of Clinical Features of the Entire Pheochromocytoma and Paraganglioma Cohort by Original Study Sites and Previously Reported Samples United Padua, Florence, Brescia, United Qin et al,14 States Spain Italy Italy Italy Belgium Brazil Kingdom 2010 Total (n = 176) (n = 168) (n = 200) (n = 106) (n = 52) (n = 67) (n = 52) (n = 77) (n = 92) (N = 990) Age at diagnosis, mean 46.3 47.0 47.1 44.1 52.1 48 39.3 25.5 41.1 43.3 (range), y (7-78) (11-80) (5-79) (9-80) (20-83) (12-84) (7-74) (5-65) (4-76) (4-84) Sex, No. Female 107 94 104 58 27 46 32 51 60 579 Male 63 65 96 48 25 21 20 26 32 396 Unknown 6 9 0 0 0 0 0 0 0 15 Tumor location, No. Total adrenal 144 133 190 80 42 37 43 52 76 798 Bilateral adrenal 7 11 15 6 1 1 1 15 8 65 Total extra-adrenal 32 35 10 26 10 30 9 21 16 189 Extra-adrenal head and neck 3 17 0 0 8 25 1 4 2 60 Adrenal and extra-adrenala 24110000 0 8 Unknown 0 0 0 0 0 0 0 4 0 4 Malignancy, No. 12 4 13 3 3 4 5 4 4 52 Familial history, No. 5 4 6 0 1 1 2 0 12 31 a Adrenal and extra-adrenal samples are included in the total adrenal samples.

site or nonsense changes and that were of short fluorescent fragments method has features of a tumor suppressor gene. absent in the control group were con- was designed. The assay comprised 6 In classic tumor suppressor genes, LOH sidered potentially pathogenic in the sets of primers spanning the 5Ј UTR, consists of loss of the wild-type allele present analysis. Synonymous substi- exons 1 through 4, and 3Ј UTR of the in tumor DNA, or the “second hit” (the tutions, variants affecting intronic re- FP/TMEM127 gene (eTable 1). Each first hit being the germline mutation in gions not immediately adjacent to the forward primer was labeled with a hereditary tumors), which results in loss intron-exon border, or changes de- 6-FAM-5Ј fluorescent tag, and frag- of functional copies of the putative tu- tected in the control group or in the ments were amplified in a multiplex mor suppressor gene in the tumor tis- SNP database (dbSNP, NCBI) were not PCR as previously described.20 In ad- sue. To identify the second hit of FP/ considered of pathogenic relevance for dition, amplicons spanning distinct TMEM127 inactivation in samples the purposes of this study. We also used chromosomal regions (chromosomes 1, carrying a mutation, LOH analysis was 3 prediction software modules to evalu- 11, or 17) were included in the multi- performed in 4 cases in which tumor ate the pathogenic potential of plex reaction as internal controls. Two DNA was available. Loss of heterozy- the identified variants. Two of them, to 3 normal DNA samples were run gosity was identified when the wild- PolyPhen-2 (http://genetics.bwh within each batch to serve as refer- type–mutant allele ratio was less than .harvard.edu/pph2/)17 and SIFT (http: ences for normal, 2-copy gene pat- 0.5 based on sequence data from germ- //sift.jcvi.org/),18 are algorithms that pre- tern. A total of 545 germline samples line and tumor DNA, as described.14 dict the possible effect of an amino acid with high-quality, high-molecular- Loss of heterozygosity data were avail- substitution on the structure and func- weight DNA were successfully pro- able from 6 previously reported tion of a human protein based on physi- cessed through this assay. Results were samples.14 cal/structural considerations (eg, amino analyzed using Peak Scanner soft- acid properties) and sequence homol- ware, version 1.0 (Applied Biosys- RNA and Reverse ogy. The third program, NetStart 1.0, tems, Foster City, California) as previ- Transcription–PCR Analysis predicts initiation of translation sites in ously reported.20 Reliability of the assay To demonstrate the effect of the eukaryotes based on a training data was validated by the detection of mono- c.409ϩ1GϾT mutation on splicing of set of mRNA derived from genomic allelic FP/TMEM127 loss in tumor DNA the TMEM127 transcript, RNA was ob- sequences with known start sites from previously reported mutant tained from peripheral lymphocytes of (http://www.cbs.dtu.dk/services samples, which were known to carry the patient carrying this mutation. /NetStart/).19 loss of the wild-type TMEM127 al- Complementary DNA was prepared by lele14 and, thus, represented models of reverse transcription using standard Quantitative Multiplex PCR single-copy TMEM127 loss (eFigure 1). procedures14 and used for a PCR span- of Short Fluorescent Fragments ning nucleotides 180 to 677 of the To assess larger deletions that might dis- Loss of Heterozygosity Analysis FP/TMEM127 open reading frame rupt the FP/TMEM127 gene and that Loss of heterozygosity (LOH) analysis (primer sequences listed in eTable 1), would not be detected by direct se- is a classic method to test whether a followed by direct sequencing of the quencing, a quantitative multiplex PCR gene involved in tumor susceptibility products.

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

Constructs, Transfections, sequence cloned into HindIII/EcoRI sites transfected into HEK293T, HeLa, or and Confocal Microscopy of the pEGFP-C2 vector (Clontech, MPC9/30 (mouse pheochromocytoma To begin to determine the effects of Mountain View, California), as previ- cell line21) cells. After 24 hours, cells were FP/TMEM127 mutations in vitro, we gen- ously described.14 Additionally, the fixed and imaged by confocal micros- erated clones carrying 5 novel substitu- c.3GϾT (p.Met1?) mutation was gen- copy as previously reported.14 Green fluo- tion mutations identified in the present erated by PCR, using as the start codon rescent protein (GFP) fluorescence of series. Four of the mutations (c.280CϾT, an in-frame methionine at position 85 transfected cells was examined for sub- p.Arg94Trp; c.208GϾA, p.Asp70Asn; and cloned as above. All constructs were cellular localization as punctate (wild- c.419GϾA, p.Cys140Tyr; and verified by sequencing. The remaining type) or diffuse patterns. c.418TϾC, p.Cys140Arg) were obtained constructs that led to truncating prod- using site-directed mutagenesis (Phu- ucts had previously shown to be unstable Statistical Analysis sion, New England Biolabs, Ipswich, Mas- for protein detection.14 Each construct, Statistical analyses were carried out sachusetts) from wild-type FP/TMEM127 as well as the wild-type version, was using SPSS software, version 17.0 (SPSS

Figure 1. Pathogenic FP/TMEM127 Gene Mutations Identified in Pheochromocytomas

TMEM127 (2q11.2)

p q Chromosome 2

Exon 1 2 34

Mis Missense mutation

Non Nonsense mutation

Ins Insertion

Del Deletion

SS Splice site mutation

Dup Duplication

Init Modified translation c.217G>C c.-18C>T c.3G>T c.76C>T c.116_119delTGTC c.149_150insA c.158G>C c.208G>A c.245-1G>T c.265_268delACAG c.268G>A c.280C>T c.409+1G>T c.410-2A>C c.418T>C c.419G>A c.447G>A c.475C>T c.627_640dup GCTGCTCTCAGAGA initiation site SS SS SS Ins Init Init Del Del Mis Mis Mis Mis Mis Mis Mis Non Non Non Dup Mutations expected to affect the integrity of TM domains 5 UTR Exon 2 Exon 3 Exon 4 3 UTR

TMEM127 protein

TM1 TM2 TM3 NH3 COOH

191 113 128 150 171 193 238

Amino acid number

Variants considered to be pathogenic by in silico and/or in vitro predictions are shown by their location along the coding gene and corresponding protein structure. The type of mutation is shown in the key. Mutations newly presented are indicated in boldface. An identical splice site mutation at C.410−2AϾC was detected in 2 unrelated families. TM indicates transmembrane domain.

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

Inc, Chicago, Illinois). Because riers were assessed for sex, bilaterality, fied Bonferroni-corrected nominal FP/TMEM127 mutations were de- familial history, and malignancy using threshold of P=.05/N* was used to cor- tected exclusively in patients with pheo- a ␹2 test or the Fisher exact test when rect for multiple hypothesis testing chromocytoma, only this group was appropriate. The Mann-Whitney test where N* is the number of indepen- used for statistical studies. Differ- was applied for testing age differ- dent comparisons. We further evalu- ences between pheochromocytoma mu- ences. Nominal 2-sided PϽ.05 was con- ated age at disease onset of the current tation carriers and non–mutation car- sidered statistically significant. A modi- series in comparison with 2 previ-

Table 2. Genetic and Clinical Features of the 20 Probands With FP/TMEM127 Mutations Considered Likely to be Pathogenica % of Total Samples Wth Mutations/ Predicted Total Adrenal Truncation Samples/ % of Tumor or Aberrantly Total No. of Samples Sex/ Tumor Malignancy Family Other Sized Predicted Adrenal Mutation Wth Mutationb Age, yc Location Status History Disease LOH Proteind Pathogenicitye Tumorsf Samples Mutationsg United States 176/144 3 1.70/2.08 c.280CϾT, p.Arg94Trp F/43 A B N ND N Y c.208GϾA, p.Asp70Asn F/50 A B N Y N Y c.3GϾT, p.Met1? F/61 A B N Acrocyanosis Y Y Y Padua, Italy 200/190 5 2.50/2.63 c.217GϾC, p.Gly73Arg M/44 A B N ND N Y c.76CϾT, p.Gln26X F/37 A B N ND Y ND c.447GϾA, p.Trp149X F/40 A(Bi) B N ND Y ND c.158GϾC, p.Trp53Ser M/21 A B N Bone marrow YN Y failure c.419GϾA, p. Cys140Tyr F/59 A(Bi) M N Diffuse ND N Y metastases Spain 168/133 1 0.60/0.75 c.409 ϩ 1GϾT, F/38 A B Y Breast cancer ND Y ND r.245_409del (p.Asp82_Thr136del) Florence, Italy 106/80 1 0.94/1.25 c.627_640dupGCTGCT F/26 A(Bi) B N Papillary ND Y ND CTCAGAGA, thyroid p.Met214Serfs98X carcinoma, medullary thyroid hyperplasia United Kingdom 77/52 1 1.30/1.92 c.-18CϾT, p.? M/44 A B N ND Y Y Brazil 52/43 1 1.92/2.32 c.116_119delTGTC, F/34 A B N Macrovascular ND Y ND p.Ile41ArgfsX39 disease Belgium 67/37 1 1.49/2.70 c.418TϾC, p.Cys140Arg F/47 A B N ND N Y Qin et al,14 2010 92/76 7 7.61/9.21 c.410-2AϾC, r.410_417del F/34 A(Bi) B Y Y Y ND (p.Leu138CysfsX12) c.149_150insA, F/25 A B U Y Y ND p.Pro51ThrfsX57 c.475CϾT, p.Gln159X F/72 A(Bi) B Y Y Y ND c.265_268delACAG, M/46 A(Bi) B N Y Y ND p.Thr89fsX35 c.410-2AϾC, r.410_417del F/37 A B Y Y Y ND (p.Leu138CysfsX12) c.245-1GϾT, r.245_264del F/66 A(Bi) B Y Y Y ND (p.Phe83SerfsX18) c.268GϾA, p.Val90Met F/32 A B N ND N Y Abbreviations: A, adrenal; B, benign; Bi, bilateral; LOH, loss of heterozygosity; M, malignant; ND, not done; U, unknown (adopted); Uni, unilateral. a Only data from probands are shown in the table. b Nucleotide and protein nomenclature according to Human Genome Variation Society. c Summary mean age, 42.8 years. d Predicted changes that affect protein size or frame are considered truncating mutations; others are missense variants. e Pathogenicity potential of missense variants was examined by PolyPhen2, SIFT, and NetStart 1.0 software. f Only study groups in which a mutation was found are shown; the total number of tested samples is 990. g Total frequency of FP/TMEM127 mutations does not include the 92 samples previously reported because they are biased by the presence of familial pheochromocytoma–linked samples.14 Summary mean frequency of mutations is 1.89%.

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

ously published pheochromocytoma/ Table 3. Relative Frequency of FP/TMEM127 Mutation in Patients With Single Adrenal Tumor Presentation Without Family History (Truly Sporadic Samples) vs Other paraganglioma cohorts (the Spanish Pheochromocytoma Susceptibility Genes, by Age Group series4 and the Italian series5). A pro- No. of No. of No. of portions-difference test for 2 indepen- Mutations/Total Mutations/Cases Mutations/Cases dent samples was applied to deter- Mutated Gene Cases (%)a Aged Ͼ40 y (%) Aged Ͼ45 y (%) mine the degree of heterogeneity Current series TMEM127 11/547 (2.01) 6/335 (1.8) 3/269 (1.1) between these series. This test showed Spanish series4 that the cohorts were not heteroge- VHL 2/95 (2.1) 0 0 neous and, thus, could be combined for RET 000the comparative analysis. SDHB 2/95 (2.1) 1/66 (1.52) 1/51 (1.96) SDHC 000RESULTS SDHD 000A total of 44 distinct FP/TMEM127 vari- Italian series5 VHL 6/233 (2.57) 0 0 ants were detected in 990 samples from RET 2/233 (0.86) 0 0 pheochromocytoma or paragan- SDHB 3/233 (1.29) 0 0 glioma patients (FIGURE 1, TABLE 2, and SDHC 000eTable 2). Of these, 19 mutations found SDHD 3/233 (1.29) 1/157 (0.6) 1/141 (0.71) in 20 patients were considered of po- Combined series4,5 tential pathogenic significance (2.02%) VHL 8/328 (2.43) 0 0 (Table 2). Thirteen of these variants RET 2/328 (0.60) 0 0 were novel changes, while the remain- SDHB 5/328 (1.52) 1/223 (0.44) 1/192 (0.52) der had been previously reported.14 Ap- SDHC 000proximately half of the mutations (10/ SDHD 3/328 (0.91) 1/223 (0.44) 1/192 (0.52) 19) comprised small deletions, a Numbers of mutations and total cases included only unilateral adrenal tumors without family history for the 3 cohorts duplications, or nonsense or splice site examined. substitutions that led to truncation or extension (1 mutation) of the pre- Figure 2. Age-Related Distribution of Patients and Affected Relatives Carrying FP/TMEM127 Mutations dicted FP/TMEM127 product. The other 9 variants were nonsynonymous Age at Onset, y FP/TMEM127 Mutation 0 20 40 60 80 missense mutations affecting con- c.158G>C Individual probands served codons of the FP/TMEM127 se- c.149_150insA (no family samples) c.627_640dup quence (Table 2). None of these 19 se- c.268G>A quence variants was detected in an c.116_119delTGTC Probands and other c.410-2A>C affected family members ethnically matched control group com- c.410-2A>C Family 1 prising 1064 alleles (or, in the case of > c.410-2A C' Family 2 c.76C>T p.Gln26X exon 3, 718 additional alleles). The con- Family 3 c.410-2A>C' stitutive (germline) nature of the mu- c.409+1G>T Family 4 c.447G>A tation could be determined in all but 1 c.280C>T case in which only somatic tissue was c.217G>C c.410-2A>C available for the analysis (c.208GϾA, c.-18C>T p.Asp70Asn) (eFigure 1A). Loss of het- c.410-2A>C c.265_268delACAG erozygosity of the wild-type allele was c.245-1G>T c.418T>C detected in 4 new cases in which tumor c.410-2A>C DNA was available (Table 2 and eFigure c.475C>T c.208G>A 1A). None of the 545 germline samples c.410-2A>C analyzed for larger FP/TMEM127 dele- c.410-2A>C c.419G>A tions showed evidence of partial or c.3G>T complete deletion of the gene. c.245-1G>T c.475C>T The mutations spanned all 3 coding 0 20 40 60 80 exons of FP/TMEM127 (Figure 1), and Age at Onset, y 13 of them affected 1 or more transmembrane domains of the predicted Dotted line represents mean age at onset (42.8 years). Mutations from families with a single affected individual are protein (Figure 1, Table 2, and eFig- indicated by open circles. Families with more than 1 affected individual (c.245-1GϾT, c.475CϾT, c.410-2AϾC, and ure 2), highlighting the relevance of c.410-2AϾCЈ) are color-coded. The latter 2 are unrelated families with an identical FP/TMEM127 mutation. these domains for FP/TMEM127 func-

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

tion. All of the missense mutations were fected individuals were available for in the pediatric group (younger than 20 considered to be potentially patho- analysis, 7 of 11 individuals carrying the years) had an FP/TMEM127 mutation. genic by at least 1 of the 3 prediction c.410-2AϾC mutation had clinical dis- A clear family history of pheochromo- programs used to preliminarily evalu- ease by age 55 years, while 1 mutation cytoma was present in only a quarter ate the likelihood of pathogenicity of carrier remained free of disease at age of patients carrying a mutation (5/20; the novel missense variants (Table 2 and 60 years.14 P=.005). eFigure 2) and/or by in vitro assays, as All mutated tumors arose from the Most tumors with an FP/TMEM17 described herein. adrenal medulla. No mutations were mutation were benign. However, Next, we examined the association of detected among the 189 extra-adrenal 1 patient carrying a missense muta- FP/TMEM127 mutations with various tumors, including 60 head and neck tion (c.419GϾA, p.Cys140Tyr) had clinical variables, including age at on- paragangliomas. Overall, one-third multiple vertebral metastases. In set, tumor location, and familial history of patients with a mutation had bilat- another patient (c.116_119delTGTC, of pheochromocytoma. The mean age at eral tumors (7/20; P =2.7ϫ 10−4). p.Ile41ArgfsX39), features of a more development of FP/TMEM127-mutated FP/TMEM127 mutations were detected aggressive histological profile (capsu- tumors was 42.8 years (95% confi- in 11% (7/65) of all bilateral pheochro- lar and vascular tumor invasion) were dence interval [CI], 36.44-49.25 years) mocytomas without other genetic cause. reported; however, stringent criteria and the median was 41.5 years. This age None of the 9 bilateral cases diagnosed for malignancy were not met. at onset is similar to the mean age of non- mutated cases in this series, 43.2 years Figure 3. Confocal Microscopy of HeLa Cells Transfected With c.280CϾT, p.Arg94Trp; (95% CI, 41.88-44.52 years) (median, 45 c.208GϾA, p.Asp70Asn; c.419GϾA, p.Cys140Tyr; c.418TϾC, p.Cys140Arg; and c.3GϾT, years) and to the reported average diag- p.Met1? Mutant or Wild-Type FP/TMEM127 Constructs nostic age for sporadic pheochromocytomas (47.01 years; 95% CI, 45.42-48.6 Wild-type FP/TMEM127 protein Met1? years).4,5 This is in contrast with patients with hereditary pheochromocytomas due to mutations in other susceptibility genes, in whom the disease has an earlier manifestation (eTable 3). The single exception are individuals with RET

mutations (eTable 3), who can usually 5 µm be distinguished by their unique clinical presentation.2,4,5 Asp70Asn Arg94Trp Furthermore, FP/TMEM127 mutation frequency in patients presenting with unilateral pheochromocytomas without a family history or other syn- dromic features (11/547 cases [2%])— ie, truly sporadic-appearing cases—is similar to that reported for mutations of the VHL and SDHB genes and higher than that described for other pheochromocytoma susceptibility genes Cys140Tyr Cys140Arg (TABLE 3).4,5 In the group diagnosed after age 45 years, FP/TMEM127 showed the highest frequency of mutations (1.1%) compared with all other susceptibility genes (Table 3). When all 20 probands and other affected members from 4 different families were combined (n=29), the age at diagnosis of pheochromocytoma was older than 40 years in two-thirds of these individu- Each construct expresses green fluorescent protein fused to the FP/TMEM127 sequence at the N terminus. Cells were fixed and stained with 4Ј,6-diamidino-2-phenylindole (nuclei) 24 hours after transfection. Green als and older than 50 years in almost fluorescent protein fluorescence was measured at 488 nm, as previously reported.14 Scale bar is indicated. The one-third of the cases (FIGURE 2). In a 5 mutant constructs have a diffuse cytoplasmic location in contrast with the wild-type protein, which is dis- tributed in the plasma membrane and in discrete organelles within the cytoplasm. single family in which multiple af-

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

All tumors with FP/TMEM127 of the mutant cases, more than one- moter, which has not yet been charac- mutations were catecholamine third of the mutation carriers pre- terized, can also occur in some cases. secreting, with no preferential pro- sented with a sporadic-appearing, be- Finally, the aberrant subcellular dis- duction of either norepinephrine or nign pheochromocytoma after age 40 tribution of newly identified epinephrine. Moreover, no recurrent years; (4) malignancy was rarely found FP/TMEM127 mutations strongly sug- clinical manifestations other than among FP/TMEM127 mutation carri- gests that intracellular distribution pheochromocytoma were detected in ers; and (5) no recurrent pattern of within the endomembrane system is rel- the FP/TMEM127-mutated cases, manifestations other than pheochro- evant for FP/TMEM127 function. Fu- although other neoplasias were mocytoma was detected among the af- ture studies should determine quanti- reported in 3 of these patients fected individuals or families, al- tative intracellular effects of individual (Table 2). though the presence of other neoplastic variants. We previously reported an Enforced expression of GFP fusion manifestations in a few patients, in- effect of FP/TMEM127 on mTOR constructs of 5 novel mutations in HeLa cluding medullary carcinoma of the thy- (mammalian target of rapamycin) sig- cells revealed that these variants had a roid, supports the idea that clinical dis- nals.14 The mislocalization of mutated subcellular distribution distinct from ease associated with FP/TMEM127 proteins raises the possibility that that of wild-type FP/TMEM127, which mutations may mimic other pheochro- FP/TMEM127 might affect mTOR ac- shows a typical plasma membrane and mocytoma-associated syndromes (eg, cessibility to its regulators and offers a punctate cytoplasmic localization pat- multiple endocrine neoplasia type 2A). useful platform to test functions of tern corresponding to endomembrane A clear familial history of pheochro- FP/TMEM127 in protein trafficking and organelles (endosome, lysosome, and mocytoma was observed in only a quar- mTOR signaling in vitro. Golgi body) (FIGURE 3 and eFigure 3). ter of cases, suggesting low pen- Author Affiliations: Divisions of Hematology and Medi- The mutant proteins were localized dif- etrance of FP/TMEM127 mutant alleles. cal Oncology (Drs Yao, Qin, King, and Dahia) and En- fusely within the cytoplasm (Figure 3). However, a comprehensive assess- docrinology (Dr King), Department of Medicine, Can- Similar results were obtained with 2 in- ment of age-related penetrance of cer Therapy and Research Center (Drs Yao, Qin, and Dahia), and Department of Cellular and Structural Bi- dependent cell lines, HEK293T and FP/TMEM127-associated disease was ology (Dr Dahia), University of Texas Health Science MPC9/30 (eFigure 3), indicating that not available in the current families and Center at San Antonio, San Antonio; Familial Cancer Clinic, Veneto Institute of Oncology (Drs Schiavi, Tas- these effects were not cell-specific. awaits further analyses. The prelimi- chin, Boaretto, and Opocher), and Departments of Sur- nary genotype-phenotype associa- gical and Gastroenterological Science (Dr Iacobone), COMMENT Internal Medicine 4 (Dr Rossi), and Medical and Sur- tions reported in the present study can gical Science (Dr Opocher), University of Padua, Padua, In this study, we assessed the preva- be used to further refine genetic test- Italy; Hereditary Endocrine Cancer Group, Spanish Na- tional Cancer Research Centre, and Instituto de Salud lence of mutations of FP/TMEM127,a ing priorities in patients with pheo- Carlos III Center for Biomedical Research on Rare Dis- recently identified pheochromocy- chromocytomas and suggest that eases, Madrid, Spain (Drs Cascon, Inglada-Pe´ rez, and toma susceptibility gene, in a large, FP/TMEM127 mutation screening may Robledo); Endocrine Genetics Unit (LIM 25), Divi- sion of Endocrinology, Department of Medicine, Uni- multi-institutional, ethnically diverse co- be recommended for patients present- versity of Sa˜ o Paulo School of Medicine (Drs R. A. To- hort of 990 individuals with pheochro- ing at an older age with adrenal tu- ledo and S. P. A. Toledo) and Division of Endocrinology, Department of Medicine, Federal University of Sa˜o mocytomas and paragangliomas mors, especially but not exclusively Paulo (Drs Lima-Junior and Kater), Sa˜ o Paulo, Brazil; (Table 1). Overall, 19 mutations de- those with bilateral disease. Department of Clinical Pathophysiology, University of Florence (Drs Ercolino, Rapizzi, and Mannelli), and Is- tected in 20 independent families were Loss of the wild-type FP/TMEM127 tituto Toscano Tumori (Dr Mannelli), Florence, Italy; considered potentially pathogenic (2%) allele was detected in all informative Medical and Molecular Genetics, School of Clinical and based on a combination of sequence con- cases, suggesting a classic mechanism Experimental Medicine, College of Medical and Den- tal Sciences and Centre for Rare Diseases and Person- servation data, prediction algorithms, of tumor suppressor gene inactivation alised Medicine, University of Birmingham, Birming- LOH, clinical segregation of the muta- similar to other pheochromocytoma ham, England (Drs Ricketts and Maher); Endocrine and Metabolic Unit, Clinica Medica, University of Bres- tion in families, and in vitro studies. susceptibility genes. However, in con- cia, Brescia, Italy (Drs Mori, Giacchè, and Castel- The main phenotypic associations trast with a minority of patients with lano); Laboratory of Human Molecular Genetics, de Duve Institute (Drs Mendola and Vikkula), and Divi- that could be drawn from our analysis SDHB, SDHC, SDHD,orVHL gene de- sion of Cardiology, Cliniques Universitaires Saint-Luc suggest the following pattern in asso- letions,4,5,22 our investigation of alter- (Dr Persu), Universite´ Catholique de Louvain, Brus- sels, Belgium; Department of Endocrinology, Osped- ciation with FP/TMEM127: (1) muta- native mechanisms of primary gene dis- ale Niguarda Ca’Granda, Milan, Italy (Dr Loli); De- tions were detected only in patients with ruption via germline deletions of the partment of Clinical and Molecular Endocrinology and tumors of adrenal localization (pheo- entire gene or specific exons did not re- Oncology, University of Naples Federico II, Naples, Italy (Dr Biondi); Department of Endocrinology, Univer- chromocytomas) but not with paragan- veal any abnormalities in more than 500 sity Hospital Leuven, Leuven, Belgium (Dr Bex); Cen- gliomas; (2) age at onset of tumors in samples tested. These results suggest tre for Endocrinology, St Bartholomew’s Hospital, Barts and the London School of Medicine, London, United patients with mutations was similar to that small mutations are the main Kingdom (Dr Grossman); Division of Molecular Medi- that of patients with sporadic disease; mechanism of FP/TMEM127 disrup- cine and Genetics, University of Michigan, Ann Ar- bor (Dr Gruber); and Centro de Investigaciones En- (3) while combined bilateral and fa- tion, although it cannot be excluded docrinolo´ gicas, Hospital de Nin˜ os Dr Ricardo Gutie´ rrez, milial cases accounted for almost half that methylation of the gene pro- Buenos Aires, Argentina (Dr Barontini).

Downloaded From: https://jamanetwork.com/ on 09/26/2021 MUTATIONS IN PHEOCHROMOCYTOMAS AND PARAGANGLIOMAS

Author Contributions: Dr Dahia had full access to all Imaging Core is also supported by National Institute Bruder, MD, Sunil Sudarshan, MD, I-Tien Yeh, MD of the data in the study and takes responsibility for on Aging (NIA) grant P30 AG013319 (Nathan Shock (UTSHCSA), Neil Aronin, MD (University of Massa- the integrity of the data and the accuracy of the data Center) and NIA grant P01AG19316. The Nucleic Ac- chusetts), Jessica Everett, MS, CGC, and Shu Chen analysis. Drs Yao, Schiavi, Cascon, Opocher, and ids Core Facility receives the support of the Depart- Huang, PhD (University of Michigan), Gabriela Sanso, Robledo contributed equally to this work. ment of Microbiology, UTHSCSA. Funds from the Fun- PhD (Centro de Investigaziones Endocrinologicas, Bue- Study concept and design: Opocher, Robledo, daça˜ o Faculdade de Medicina and the Division of nos Aires, Argentina), Berenice B. Mendonça, MD, Dahia. Endocrinology provided support for the screening of Maria Adelaide Albergaria Pereira, MD, Maria Caˆn- Acquisition of data: Yao, Schiavi, Cascon, Qin, the Brazilian cohort. Dr R. A. Toledo is supported by dida Barisson Villares Fragoso, MD, PhD, and Anto- Inglada-Pe´ rez, King, R. A. Toledo, Ercolino, Rapizzi, aSa˜ o Paulo State Research Foundation (FAPESP grant nio Marcondes Lera´ rio, MD, PhD (Hospital das Clıńi- Ricketts, Mori, Giacchè, Mendola, Taschin, Boaretto, 2009/15386-6) postdoctoral fellowship. Drs Ricketts cas, University of Sa˜ o Paulo), Carmen Fajardo Loli, Iacobone, Rossi, Biondi, Lima-Junior, Kater, and Maher are supported by Cancer Research UK. Dr Montan˜ ana, MD (Hospital Universitario de La Ri- Bex, Grossman, Gruber, Barontini, Persu, Castellano, Vikkula is supported by the Interuniversity Attraction bera, Alzira, Spain), Antonio Pico´ , MD (Hospital Gen- S. P. A. Toledo, Maher, Mannelli, Opocher, Robledo, Poles initiated by the Belgian Federal Science Policy, eral Universitario de Alicante, Alicante, Spain), Marta Dahia. network 6/05; Concerted Research Actions Conven- Korbonits, MD, PhD (St Bartholomew’s Hospital, Barts Analysis and interpretation of data: Yao, Schiavi, tion No. 07/12-005 of the Belgian French Commu- and the London School of Medicine), Annamaria Co- Cascon, Qin, Inglada-Pe´ rez, King, R. A. Toledo, Mori, nity Ministry; la Communaute´ Française de Wallonie- lao, MD (University of Naples Federico II), Gianpaolo Taschin, Loli, Vikkula, Grossman, Gruber, Persu, Bruxelles et la Lotterie Nationale. Drs Vikkula and Persu Bernini, MD (University of Pisa), Maria Rosaria Am- Castellano, S. P. A. Toledo, Maher, Opocher, Dahia. are supported by the FRS-FNRS (Fonds de la Recher- brosio, MD, and Chiara Zatelli, MD (University of Fer- Drafting of the manuscript: Cascon, Inglada-Pe´ rez, che Scientifique), Belgium. Dr Gruber is supported by rara, Ferrara, Italy), Emanuela Lucci Cordisco, MD (Uni- Robledo, Dahia. NCI grant 5 P30 CA465920. Dr Castellano is sup- versity of Rome), Gennaro Favia, MD, and Antonio Critical revision of the manuscript for important in- ported by funds from the Italian University and Re- Toniato, MD (University of Padova), Serena De- tellectual content: Yao, Schiavi, Cascon, Qin, search Ministry (grant 2006060473) and by the Fonda- mattè, MD (Ospedale S. Chiara, Trento, Italy), Mas- Inglada-Pe´ rez, King, R. A. Toledo, Ercolino, Rapizzi, zione della Comunità Bresciana. Dr S. P. A. Toledo is simo Terzolo, MD, Franco Veglio, MD, and Paolo Mu- Ricketts, Mori, Giacchè, Mendola, Taschin, Boaretto, supported by a Conselho Nacional de Desenvolvi- latero, MD (University of Turin), Ermanno Rossi, MD Loli, Iacobone, Rossi, Biondi, Lima-Junior, Kater, Bex, mento Cientı´fico e Tecnolo´ gico grant. Dr Robledo is (Ospedale Reggio Emilia, Italy), Giorgio Bertola, MD Vikkula, Grossman, Gruber, Barontini, Persu, supported by Fondo de Investigaciones Sanitarias (Ospedale Garbagnate Milanese, Milan, Italy), Maria Castellano, S. P. A. Toledo, Maher, Mannelli, Opocher, grant PI 08/080883 and Fundacio´ n Mutua Madrilenã Vittoria Davì, MD (Policlinico GB Rossi, Verona, Italy), Robledo, Dahia. (reference AP2775/2008). Dr Dahia is supported Giuseppe Picca, MD (Azienda Ospedaliero- Statistical analysis: Inglada-Pe´ rez, Dahia. by the Voelcker Fund, Alex’s Lemonade Stand Foun- Universitaria, Foggia, Italy), Adele Nardecchia, MD Obtained funding: Vikkula, Castellano, S. P. A. Toledo, dation, Concern Foundation, and Clinical Transla- (Azienda Ospedaliera Policlinico, Bari, Italy), Gilberta Maher, Robledo, Dahia. tional and Science Awards (National Institutes of Giacchetti, MD, and Giorgio Arnaldi, MD (University Administrative, technical, or material support: Yao, Health). Politenica delle Marche, Ancona, Italy), Julian Donckier, Schiavi, King, Ercolino, Rapizzi, Ricketts, Mori, Giacchè, Role of the Sponsors: The sponsors had no role in the MD, Pierre Garin, MD, Marc Hamoir, MD, Domi- Mendola, Taschin, Boaretto, Loli, Iacobone, Grossman, design and conduct of the study; in the collection, nique Maiter, MD, and Michel Mourad, MD (Univer- Gruber, Persu, Castellano, Opocher, Dahia. analysis, and interpretation of the data; or in the prepa- site´ Catholique de Louvain, Brussels, Belgium), Flo- Study supervision: Schiavi, Vikkula, Persu, Maher, ration, review, or approval of the manuscript. rence Bosly, MD (Clinique Saint-Joseph, Arlon, Mannelli, Opocher, Robledo, Dahia. Online-Only Material: eFigures 1 through 3 and Belgium), Annick Claessens, MD (Clinique Saint- Financial Disclosures: None reported. eTables 1 through 3 are available online at http://www Joseph, Arlon, Belgium), Brigitte Decallonne, MD (Uni- Funding/Support: The Cancer Therapy and Research .jama.com. versity Hospital Leuven), Pierre Kleynen, MD (CHU- Center Pathology Core and the Optical Imaging Core Additional Contributions: We thank the members of Brugmann, Brussels, Belgium), Michel Ponchon, MD are supported by University of Texas Health Science the Familial Pheochromocytoma Consortium for their (Clinique Saint-Etienne, Brussels, Belgium), An Van Au- Center at San Antonio (UTHSCSA) and National Can- support and earlier contributions. We are grateful to denhove, MD (Clinique Sainte-Elisabeth, Brussels, Bel- cer Institute (NCI) grant P30 CA54174 (Cancer Therapy the following colleagues who contributed to the re- gium), and Inge Van Pottelbergh, MD (O. L. Vrouw- and Research Center at UTHSCSA). The Optical cruitment of patients and/or clinical information: Jan ziekenhuis, Aalst, Belgium).

REFERENCES 1. Manger WM. An overview of pheochromocy- ral and nonneural tumors. Hum Genet. 2008;124 15. Casco´ n A, Lo´ pez-Jime´ nez E, Landa I, et al. Ra- toma: history, current concepts, vagaries, and diag- (3):279-285. tionalization of genetic testing in patients with appar- nostic challenges. Ann N Y Acad Sci. 2006;1073: 8. Ladroue C, Carcenac R, Leporrier M, et al. PHD2 ently sporadic pheochromocytoma/paraganglioma. 1-20. mutation and congenital erythrocytosis with Horm Metab Res. 2009;41(9):672-675. 2. Neumann HP, Bausch B, McWhinney SR, et al; paraganglioma. N Engl J Med. 2008;359(25):2685- 16. Neumann HP, Eng C. The approach to the pa- Freiburg-Warsaw-Columbus Pheochromocytoma 2692. tient with paraganglioma. J Clin Endocrinol Metab. Study Group. Germ-line mutations in nonsyndromic 9. Yao G, Xu L, Wu X, Xu L, Yang J, Chen H. Pre- 2009;94(8):2677-2683. pheochromocytoma. N Engl J Med. 2002;346(19): ventive effects of salvianolic acid B on transforming 17. Adzhubei IA, Schmidt S, Peshkin L, et al. A method 1459-1466. growth factor-␤1-induced epithelial-to-mesenchy- and server for predicting damaging missense mutations. 3. Gimenez-Roqueplo AP, Lehnert H, Mannelli M, et al; mal transition of human kidney cells. Biol Pharm Bull. Nat Methods. 2010;7(4):248-249. European Network for the Study of Adrenal Tumours 2009;32(5):882-886. 18. Ng PC, Henikoff S. Predicting deleterious amino (ENS@T) Pheochromocytoma Working Group. Phaeo- 10. Bayley JP, Kunst HP, Cascon A, et al. SDHAF2 mu- acid substitutions. Genome Res. 2001;11(5):863- chromocytoma, new genes and screening strategies. tations in familial and sporadic paraganglioma and 874. Clin Endocrinol (Oxf ). 2006;65(6):699-705. phaeochromocytoma. Lancet Oncol. 2010;11(4): 19. Pedersen AG, Nielsen H. Neural network predic- 4. Cascon A, Pita G, Burnichon N, et al. Genetics of 366-372. tion of translation initiation sites in eukaryotes: per- pheochromocytoma and paraganglioma in Spanish 11. Burnichon N, Brière JJ, Libe´ R,etal.SDHA is a spectives for EST and genome analysis. Proc Int Conf patients. J Clin Endocrinol Metab. 2009;94(5): tumor suppressor gene causing paraganglioma. Hum Intell Syst Mol Biol. 1997;5:226-233. 1701-1705. Mol Genet. 2010;19(15):3011-3020. 20. Casco´ n A, Landa I, Lo´ pez-Jime´ nez E, et al. Mo- 5. Mannelli M, Castellano M, Schiavi F, et al; Italian 12. Opocher G, Schiavi F, Conton P, Scaroni C, lecular characterisation of a common SDHB deletion Pheochromocytoma/Paraganglioma Network. Clini- Mantero F. Clinical and genetic aspects of in paraganglioma patients. J Med Genet. 2008; cally guided genetic screening in a large cohort of Ital- phaeochromocytoma. Horm Res. 2003;59(suppl 1): 45(4):233-238. ian patients with pheochromocytomas and/or func- 56-61. 21. Powers JF, Evinger MJ, Tsokas P, et al. Pheochro- tional or nonfunctional paragangliomas. J Clin 13. Dahia PL, Hao K, Rogus J, et al; Familial Pheo- mocytoma cell lines from heterozygous neurofibro- Endocrinol Metab. 2009;94(5):1541-1547. chromocytoma Consortium. Novel pheochromocy- matosis knockout mice. Cell Tissue Res. 2000; 6. Schlisio S, Kenchappa RS, Vredeveld LC, et al. The toma susceptibility loci identified by integrative 302(3):309-320. kinesin KIF1B␤ acts downstream from EgIN3 to in- genomics. Cancer Res. 2005;65(21):9651-9658. 22. Burnichon N, Rohmer V, Amar L, et al; PGL.NET duce apoptosis and is a potential 1p36 tumor 14. Qin Y, Yao L, King EE, et al. Germline muta- Network. The succinate dehydrogenase genetic test- suppressor. Genes Dev. 2008;22(7):884-893. tions in TMEM127 confer susceptibility to ing in a large prospective series of patients with 7. Yeh IT, Lenci RE, Qin Y, et al. A germline muta- pheochromocytoma. Nat Genet. 2010;42(3):229- paragangliomas. J Clin Endocrinol Metab. 2009; tion of the KIF1B␤ gene on 1p36 in a family with neu- 233. 94(8):2817-2827.

Downloaded From: https://jamanetwork.com/ on 09/26/2021