Prolyl hydroxylase domain (PHD) proteins play a major role in regulating the hypoxia-inducible factor (HIF) that induces expression of genes involved in angiogenesis, erythropoiesis, and cell metabolism, proliferation, and survival. Germ-line mutations in the prolyl hydroxylase domain 2 gene (PHD2) have been reported in patients with familial erythrocytosis but not in association with tumors. We describe a patient with erythrocytosis and recurrent paraganglioma who carries a newly discovered PHD2 mutation. This mutation affects PHD2 function and stabilizes HIF-
proteins. In addition, we demonstrate loss of heterozygosity of PHD2 in the tumor, suggesting that PHD2 could be a tumor-suppressor gene.
HIF is an alpha/beta heterodimer consisting of a tightly regulated alpha subunit (1, 2, or 3) and a constitutively expressed beta subunit. Under normoxic conditions, HIF- subunits are hydroxylated by dioxygenases that use oxygen as a cosubstrate.2,3 The PHD proteins (also called EGLN proteins) are dioxygenases that hydroxylate key proline residues of HIF-, allowing for the attachment of HIF- to the von Hippel–Lindau (VHL) tumor-suppressor protein, part of a complex that leads to the ubiquitination and degradation of HIF- in the proteasome. In mammalian cells, there is a family of PHD proteins (PHD1, PHD2, and PHD3), among which PHD2 (also called EGLN1) seems to be the critical oxygen sensor that tags HIF-1 for destruction.2,4 In addition, factor-inhibiting HIF hydroxylates the asparagine residue in the C-terminal domain of HIF, which blocks the attachment of HIF to an essential transcriptional coactivator.5
Germ-line mutations in genes involved in the HIF pathway have been described in association with syndromes that predispose patients to either tumors or familial polycythemia. The most frequent alterations affect the VHL tumor-suppressor gene; heterozygous germ-line mutations of VHL cause VHL disease.6 This autosomal dominant condition predisposes patients to multiple vascularized tumors, including hemangioblastomas of the central nervous system and retina, clear-cell renal-cell carcinomas, pheochromocytomas, endocrine pancreatic tumors, and endolymphatic sac tumors. In contrast, a homozygous 598C
T (R200W) germ-line mutation of VHL causes Chuvash congenital polycythemia, an autosomal recessive disease reported first in residents of the Chuvash Autonomous Republic of the Russian Federation.7 Homozygous carriers of the R200W mutation have erythrocytosis, with increased erythropoietin production. Remarkably, carriers of the R200W mutation are not unduly susceptible to cancer. Heterozygous germ-line mutations in the PHD2 gene and the gene for HIF-2 (HIF2A) have been found in patients with familial erythrocytosis due to an elevated erythropoietin level, but tumors have not been reported.8,9,10,11
Germ-line mutations in genes encoding two mitochondrial enzymes operating in the Krebs cycle, fumarate hydratase (FH) and succinate dehydrogenase (SDHB, SDHC, and SDHD), have been implicated in the development of the hereditary leiomyomatosis and renal-cell cancer syndrome and the hereditary paraganglioma–pheochromocytoma syndrome.12 In tumors deficient in FH and SDH, accumulation of the substrates (fumarate and succinate) inhibits PHD function and in this way causes overexpression of HIF.
We investigated the main genes of the HIF pathway in a patient with congenital erythrocytosis and recurrent abdominal paraganglioma.
Case Report
The patient was referred to us in 1988, when he was 30 years old and had a hematocrit value within the upper end of the normal range (Table 1). Clinical and laboratory workups confirmed the diagnosis of erythrocytosis, but both routine and specialized tests failed to identify a specific cause. The erythropoietin-dependent in vitro growth of erythroid colonies ruled out the diagnosis of polycythemia vera. Phlebotomy was initiated, to maintain the hematocrit value below 45%. In addition, mild hypertension was treated with atenolol. Because the mean corpuscular volume and ferritin level remained normal on follow-up, despite the phlebotomy regimen, the patient was evaluated for, and was found to be homozygous for, the C282Y hot-spot mutation in the hemochromatosis gene HFE, which has no causal relationship to polycythemia.13 All members of the proband's family were tested, and his brother was also found to be homozygous for the C282Y mutation (and is now undergoing treatment for hemochromatosis). Neither erythrocytosis nor tumor was found in any family member other than the proband (Figure 1A).
|
|
Subsequently, the blood pressure returned to normal, atenolol was discontinued, and the hematocrit value became stabilized within the normal range without further phlebotomy. In 2003, however, urinary excretion of metanephrines was in the upper end of the normal range, and the hematocrit value rose, raising suspicion of recurrence of the tumor, which was substantiated by 131I-metaiodobenzylguanidine scintigraphy scanning in 2004.
During a second operation, the tumor, which was histologically similar to the previous one, was completely resected. The serum erythropoietin level (normal range, 5 to 25 mU per milliliter) was 48 mU per milliliter before the resection and 32 mU per milliliter 8 days after the resection. A relation between the erythrocytosis and erythropoietin production by the tumor was ruled out, however, since erythropoietin messenger RNA was not found on gene-expression assays (Taqman assay, with human hepatocellular carcinoma cells [Hep3B cells] and an erythropoietin-producing tumor associated with secondary polycythemia as controls) (data not shown). From 2004 through the time of publication of this article, the patient's clinical status has remained unchanged, with a slight elevation of hematocrit and blood pressure, both requiring treatment with phlebotomy and antihypertensive drugs. Repeated CT and 131I-metaiodobenzylguanidine scanning have not detected tumor recurrence, although urinary excretion of metanephrines has remained at twice the upper limit of the normal range during this period. The possibilities of VHL disease or the hereditary paraganglioma–pheochromocytoma syndrome were ruled out through routine genetic testing: no germ-line mutation was identified in the VHL, SDHB, SDHC, or SDHD genes.
Methods
DNA Sequencing and Loss-of-Heterozygosity Analysis
After receiving approval from our local ethics committee and obtaining written informed consent from the proband and all family members, blood samples were collected for genetic analysis. Genomic DNA was extracted from peripheral blood and paraganglioma and purified on a spin column (Qiagen). Screening for mutations in the PHD2 gene was performed by means of direct sequencing (involving a BigDye Terminator v3.1 Kit and an ABI 3730 Genetic Analyzer, both from Applied Biosystems). Sequence data were aligned using 4Peaks software (http://mekentosj.com/4peaks). Sequences of all primers used in this study are given in the Supplementary Appendix, available with the full text of this article at www.nejm.org. DNA samples from normal persons were screened for the c.1121AG mutation (with 280 alleles tested). In silico prediction of the deleterious effect of the mutant was performed through the PolyPhen Web site (http://genetics.bwh.harvard.edu/pph) and the Panther Web site (www.pantherdb.org).
For loss-of-heterozygosity analysis, six microsatellite markers flanking the PHD2 gene (located at 1q42.1) were analyzed: D1S2877, D1S2833, D1S251, D1S2709, D1S2785, and D1S2836 (ABI PRISM Linkage Mapping Set, Applied Biosystems). Amplification was performed with the use of the Multiplex PCR Master Mix (Qiagen), according to the manufacturer's instructions. Amplified fragments were run on the genetic analyzer (ABI 3730). Data were analyzed with the use of GeneMapper software (Applied Biosystems).
Functional Analysis
The H374R PHD2 mutation was inserted into a pcDNA3-HA-PHD2 vector by means of site-directed mutagenesis (Promega). Luciferase assays were performed on human embryonic kidney 293T cells cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal-calf serum seeded in 12-well plates (2x105 cells per well) 24 hours before transfection. The transfection reagent Lipofectamine 2000 (Invitrogen) was used for transient transfections. The pcDNA3-HA-HIF-2 or HIF-1 vector (100 ng), pGL3 erythropoietin-promoter luciferase vector (200 ng), and pRenilla vector (20 ng) were cotransfected with a PHD2 (wild type, or H374R mutant) expression vector (5 to 125 ng). Luciferase activity was measured 24 hours after transfection with the use of a Dual-Luciferase Reporter Assay System (Promega). For immunoblotting, 15-µl aliquots of the lysates used in luciferase assays were run, and anti-HA (F7) antibody (Tebu) was used.
Results
Sequencing and Loss-of-Heterozygosity Analysis
We sequenced germ-line DNA obtained from the patient for genes that could be implicated in congenital polycythemia associated with a high serum erythropoietin level: PHD1, PHD2, PHD3, and HIF2A. We identified a unique c.1121AG (p.H374R) heterozygous variation in the PHD2 gene (Figure 2A). The absence of this variant in a control population (in which 280 alleles were tested) ruled out a rare polymorphism. None of the three first-degree relatives of the patient were carriers of the PHD2 mutation. DNA of the patient's tumor was sequenced for the PHD2 gene, and the mutant allele was detected (Figure 2A). Interestingly, the chromatogram at position 1121 revealed predominantly the mutant allele, indicating a loss-of-heterozygosity of PHD2; the residual signal for the wild-type allele could be accounted for through contamination by normal tissue. Loss of heterozygosity was confirmed through microsatellite analysis of the PHD2 region, which showed monozygosity for all the markers tested (Figure 2B).
|
Consequences of the amino acid change were analyzed by computer programs, and the PolyPhen and Panther Web sites predicted that the changes observed would have a negative effect on protein function. H374 is a highly conserved residue, present in various life forms (from worms to flies to humans) and even prokaryotic pathogens (Figure 3A). Prolyl hydroxylases are dioxygenases dependent on Fe2+ and 2-oxoglutarate, and the H374 amino acid is one of the three critical residues that coordinate binding of Fe2+ ion (Figure 3B).14
|
subunits (1 or 2) were coexpressed with a luciferase gene driven by hypoxia-responsive elements from the erythropoietin promoter gene. The addition of wild-type PHD2 caused the dose-related suppression of HIF-–mediated induction of the reporter gene, whereas the response to the H374R PHD2 mutant was impaired. On immunoblotting used to monitor expression of the PHD2 protein, for equivalent amounts of transfected plasmid, the wild-type protein accumulated to a level that was higher than that of the mutant protein, suggesting that the mutant protein is unstable (Figure 3C). To take this effect into account, the amounts of transfected plasmid were adjusted to yield equivalent amounts of protein. With this adjustment, the impaired activity of H374R mutant was confirmed — the ratio of the activity of luciferase and that of renilla for wild-type PHD2 was, on average, 2.5 times that for mutant PHD2 (Figure 3D, which shows the results obtained for HIF-2; the same results were observed for HIF-1). This effect was not restricted to erythropoietin-responsive elements, since it was also obtained with the luciferase reporter gene under the control of hypoxia-responsive elements from the vascular endothelial growth factor promoter (data not shown). Discussion
The PHD2–VHL–HIF-2 pathway plays a crucial role in erythropoiesis; a partial interruption of the pathway can cause erythrocytosis, whereas drastic alterations of the pathway are associated with the development of tumors.11,15,16,17 Germ-line mutations in the PHD2 gene have been reported in patients with familial erythrocytosis, but not in association with tumors.8,9,10,11 We describe here a patient with a newly discovered PHD2 mutation who first presented with isolated erythrocytosis and subsequently was found to have a recurrent paraganglioma. Tumor analysis showing loss of heterozygosity involving the wild-type PHD2 suggests that PHD2 can act as a tumor-suppressor gene. Tumor-suppressor activity of PHD2 has recently been incriminated in sporadic endometrial cancer cells and cell lines.18,19 The oncogenic processes induced by the loss of PHD2 are unknown but could involve HIF-independent pathways, since the H374 amino acid of the enzyme is part of a highly conserved catalytic core that is present even in prokaryotic pathogens that do not possess HIF. Since the H374R mutant described here perturbs the down-regulation of HIF by PHD2, and considering that paragangliomas are highly vascularized tumors, alteration of the PHD2–HIF pathway could contribute to the growth of a paraganglioma. Up to 25% of paragangliomas are hereditary and are mainly associated with germ-line mutations in the SDHB, SDHD, and VHL genes or the proto-oncogene RET.20,21 Our work suggests that PHD2 is a candidate gene involved in the development of paraganglioma and that mutation status should be explored in families with unidentified genetic alterations in the tumor.
From a clinical point of view, it is important to point out that carriers of the heterozygous germ-line PHD2 mutation who have been described in the literature are younger than our patient was at the time of diagnosis of paraganglioma (43 years of age). One exception is a 54-year-old patient with hypertension; this case could have been due to catecholamine secretion from a misdiagnosed paraganglioma.10 We recommend stringent follow-up of carriers of the germ-line PHD2 mutation, since their risk of a paraganglioma may be abnormally elevated. In addition, preclinical models with prolyl hydroxylase inhibitors to correct anemia are currently being developed. The potential oncogenic effect of this inhibition should be evaluated before the approach is considered for therapeutic use.3,22
Supported by grants from the French National Cancer Institute (Programme National d'Excellence Spécialisée Rein) and the French League against Cancer (Comités du Cher et de l'Indre).
No potential conflict of interest relevant to this article was reported.
We thank the patient and his family for their participation and Prof. Dominique Maïza, Prof. Nicole Casadevall, Prof. Patrick Bruneval, Dr. Brigitte Bressac de Paillerets, Dr. Anne-Paule Gimenez-Roqueplo, Dr. Ahyan Ulusakarya, Johny Bombled, Caroline Chordi, and Prof. Gilbert Lenoir for their assistance.
Source Information
From Génétique Oncologique, Ecole Pratique des Hautes Etudes (EPHE) and Centre National de la Recherche Scientifique (CNRS) (FRE 2939), Institut de Cancérologie Gustave Roussy, Villejuif (C. Ladroue, R.C., S.G., J.F., S.R., B.G.); Faculté de Médecine Paris-Sud, Université Paris-Sud (C. Ladroue, R.C., S.G., S.R., B.G.), and Centre Pilote Tumeurs Rares, Institut National du Cancer and Assistance Publique–Hôpitaux de Paris, Service d'Urologie, Centre Hospitalier Universitaire de Bicêtre (S.R.) — both in Le Kremlin Bicêtre; Service d'Hématologie Clinique (M.L.), Service de Chirurgie Thoracique et Cardio-Vasculaire (C. Le Hello), and Laboratoire d'Anatomie et Cytologie Pathologiques (F.G.-S.), Centre Hospitalier Universitaire, Caen; Institute of Developmental Biology and Cancer Research, CNRS (UMR 6543) University of Nice, Nice (J.P.); and Service de Néphrologie, Hôpital Necker, Paris (S.R.) — all in France.
Ms. Ladroue and Mr. Carcenac contributed equally to this article.
Address reprint requests to Dr. Richard at Génétique Oncologique EPHE, Faculté de Médecine Paris-Sud, 63 Rue Gabriel Péri, 94276 Le Kremlin-Bicêtre, France, or at stephane.richard{at}u-psud.fr.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
PHD2 Mutation and Congenital Erythrocytosis with Paraganglioma
Eltzschig H. K., Eckle T., Grenz A., Gardie B., Feunteun J., Richard S.
Extract |
Full Text |
PDF
N Engl J Med 2009;
360:1361-1362, Mar 26, 2009.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
