| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Nephrology (B.B., D.B., T.H., C.P., H.P.H.N.) and Laboratory Medicine (M.M.H.), University Medical Center Freiburg, D-79106 Freiburg, Germany; Institute for Human Genetics and Anthropology (W.B., J.K.), University of Freiburg, Freiburg, Germany; Center of Human Genetics (W.B., J.K.), Freiburg, Germany; Department of Maxillofacial Surgery (V.F.M.), University Hospital Eppendorf, Hamburg, Germany; Hereditary Endocrine Cancer Group (M.R., A.C.), Centro Nacional de Investigaciones Oncologicas, Madrid, Spain; Istituto Oncologico Veneto Instituto di Ricovero e Cura a Carattere Scientifico (F.S., G.O.), Padova, Italy; Department of Hypertension (M.P., A.J.), Institute of Cardiology, Warsaw, Poland; Departments of Clinical Sciences (C.L.) and Dermatology (S.C.), University of Rome La Sapienza, Rome, Italy; Department of Endocrinology (G.A.), Azienda Ospedaliero-Universitaria Ospedali Riuniti di Ancona, Ancona, Italy; Medical Clinic I (R.D.K.-N.), Universitätsklinikum Schleswig-Holstein, Campus Lübeck, Lübeck, Germany; Department of Endocrinology (N.R.), Ludwig-Maximilians-University of Munich, Munich, Germany; Department of Pathology (A.F.), University of Padova, Padova, Italy; Department of Digestive and Endocrine Surgery (L.B.), University Hospital Nancy, University of Nancy, Nancy, France; Institute of Nuclear Medicine (M.A.W.), Division of Endocrinology, University of Basel, Basel, Switzerland; Department of Clinical Pathophysiology, Endocrine Unit (M.Ma.), University of Florence, Florence, Italy; Blood Pressure Unit (G.M.), Department of Cardiovascular Sciences, St. George’s University, London, United Kingdom; Endocrine Surgery Unit (F.F.P.), Hammersmith Hospital, London, United Kingdom; Centro de Investigaciones Endocrinologicas-Consejo Nacional de Investigaciones Científicas y Técnicas (M.B.), Buenos Aires, Argentina; Department of Surgery (M.K.W.), Kliniken Essen-Mitte, Essen, Germany; Department of Pediatrics (B.K.), University of Essen, Essen, Germany; Department of Endocrinology (G.B.), Medizinische Hochschule, Hannover, Germany; Department of Pediatrics (R.P.), University of Leipzig, Leipzig, Germany; Department of Internal Medicine I-Endocrine and Diabetes (A.-C.K.), University of Wuerzburg, Wuerzburg, Germany; Department of Surgery (F.L.), Hospital of the German Red Cross, Berlin, Germany; Department of Nephrology and Hypertension (M.Mo.), University of Berne, Berne, Switzerland; Department of Visceral Surgery (O.G.), University of Halle, Halle, Germany; Department of Nuclear Medicine and Endocrine Oncology (B.J.), M. Sklodowska-Curie Memorial Cancer Centre and Institute of Oncology, Gliwice, Poland; Genomic Medicine Institute (S.R.M., C.E.), Cleveland Clinic Foundation, Cleveland, Ohio 44195; and Department of Genetics (S.R.M., C.E.), Case Western Reserve University School of Medicine, Cleveland, Ohio 44106
Address all correspondence and requests for reprints to: Hartmut P. H. Neumann, Medizinische Universitätsklinik, Hugstetter Straße 55, D-79106 Freiburg, Germany. E-mail: neumann{at}med1.ukl.uni-freiburg.de.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Materials and Methods: An international registry for NF1-pheochromocytomas was established. Mutation scanning was performed using denaturing HPLC for intragenic variation and quantitative PCR for large deletions. Loss-of-heterozygosity analysis using markers in and around NF1 was performed.
Results: There were 37 eligible subjects (ages 14–70 yr). Of 21 patients with corresponding tumor available, 67% showed somatic loss of the nonmutated allele at the NF1 locus vs. 0 of 12 sporadic tumors (P = 0.0002). Overall, 86% of the 37 patients had exonic or splice site mutations, 14% large deletions or duplications; 79% of the mutations are novel. The cysteine-serine rich domain (CSR) was affected in 35% but the RAS GTPase activating protein domain (RGD) in only 13%. There did not appear to be an association between any clinical features, particularly pheochromocytoma presentation and severity, and NF1 mutation genotype.
Conclusions: The germline NF1 mutational spectra comprise intragenic mutations and deletions in individuals with pheochromocytoma and NF1. NF1 mutations tended to cluster in the CSR over the RAS-GAP domain, suggesting that CSR plays a more prominent role in individuals with NF1-pheochromocytoma than in NF1 individuals without this tumor. Loss-of-heterozygosity of NF1 markers in NF1-related pheochromocytoma was significantly more frequent than in sporadic pheochromocytoma, providing further molecular evidence that pheochromocytoma is a true component of NF1.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
NF1 is caused by germline mutations of the NF1 gene, located on chromosome sub-band 17q11.2 (8, 9). It contains 57 coding exons spanning approximately 300 kb of genomic DNA encoding a 12,255-bp transcript with an open reading frame of 8,454 nucleotides and 2,818 amino acids, respectively (10). Three known alternative splice variants exist (11, 12).
Mutation detection in the NF1 gene remains a considerable challenge because of its large size and the presence of 36 pseudogenes (13, 14, 15, 16, 17, 18, 19). Because of the technical challenges and the low prevalence of pheochromocytoma in NF1, only a first-limited series of molecular genetically characterized patients with NF1 and pheochromocytoma has been presented so far by members of this study group (1).
According to Knudson’s two-hit theory, pheochromocytoma development requires biallelic inactivation (20), which was shown to be true in murine models (21) and humans (22) by loss-of-heterozygosity (LOH) of markers in the NF1 17q11.2 region and for silencing of neurofibromin expression (23).
We designed this current study to describe and document the full germline mutational spectra of NF1 in individuals with pheochromocytoma in the context of NF1. As a secondary end point, we sought to demonstrate a high frequency of somatic loss of the remaining wild-type NF1 allele in NF1-related pheochromocytomas because molecular evidence adjunctive to the epidemiological association that pheochromocytoma is a true component of NF1.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Study subjects have been recruited among those in the Pheochromocytoma Registry founded separately in Freiburg and Warsaw, and combined in 1995 ("European-American Phaeochromocytoma Registry") (4, 5). In addition, because of the rarity of the co-occurrence, we extended the special NF1-associated pheochromocytoma group by asking for further contributions to the registry from collaborators in Europe and America should they have eligible subjects with pheochromocytoma and NF1. We also contacted the neurofibromatosis outpatient clinic and the Department of Maxillofacial Surgery at the University of Hamburg, all departments of dermatology in hospitals in Germany, many private practice dermatologists, and the German neurofibromatosis self-support group (only one accrued in this manner). The project was approved by the ethics committees of the respective institutions. All subjects provided informed consent in accordance with the accepted standards for each respective country.
Pheochromocytoma was defined as a tumor of the adrenal or of extra-adrenal sites of the abdomen and thorax. All neoplasias were histologically confirmed. Patients eligible for this study included those with pheochromocytoma and NF1 registered until June 30, 2006. For all subjects, demographic data as well as tumor number, location, and biology were registered. Presence of distant metastases were the criterion for malignant pheochromocytoma. Patients with NF1 were diagnosed according to the National Institutes of Health consensus criteria (6, 7). Clinical presentation of NF1 was classified as severe according to occurrence of malignant tumors, intellectual handicap, or epilepsy, or as mild if only associated with neurofibromas, axillary freckling, café-au-lait spots, and Lisch nodules, whereas additional features were regarded as intermediate NF1. EDTA-anticoagulated blood samples were obtained from each patient after obtaining informed consent. In addition, we collected fresh-frozen or paraffin-embedded tumor tissue of NF1-associated pheochromocytomas.
Molecular genetic analyses
DNA was extracted from peripheral blood lymphocytes by standard procedures. Mutation analysis of the NF1 gene was performed on genomic DNA from all 37 subjects with NF1-associated pheochromocytoma. Mutation analysis of the 57 exons and flanking intronic regions, as well as the untranslated 5' and 3' regions of the NF1 gene, required redesign of PCR primer pairs to exclude amplification of any of the 36 pseudogenes (13, 14, 15, 16, 17, 18, 19) (Table 1
).
|
|
Samples without germline intragenic mutations were candidates for large deletion and rearrangement analysis. Fine, exon-by-exon deletion detection was performed using a quantitative PCR approach with SYBR Green I detection (SYBR Green PCR Master Mix; QIAGEN, Hilden, Germany). The ABI Prism 7900 Sequence Detection System (PE Applied Biosystems, Foster City, CA) was used. The Sequence Detection System software (version 2.2.1; PE Applied Biosystems) was used to analyze the obtained data. All amplicons were 100–350 bp in length. Quantitative PCR was performed as described with the exception of an annealing temperature of 57 C (24, 25). Two subtelomeric amplicons were used as internal references. As a deletion negative control, genomic DNA from healthy, unrelated ethnically matched individuals was amplified. Melting curve analysis and gel electrophoresis were performed after the amplification to exclude the presence of nonspecific PCR products. Absolute quantification of the patient’s DNA was performed by interpolation of the threshold cycle number against the corresponding standard curve and normalized against a normal diploid reference genome. Ratio values of 1.0 indicated a diploid situation, whereas a ratio of 0.5 or 1.5 indicated a partial haploidy or partial triploidy, respectively.
The detected DNA missense alterations were investigated for pathogenicity by analyzing whether they were present in 100 healthy, unrelated, and ethnically matched subjects.
In silico analysis
An additional in silico analysis was performed for missense mutations/sequence variants that have not been previously described to evaluate the significance of the amino acid exchange. The ClustalW program (http://www.ebi.ac.uk/clustalw/) was used to generate a multiple sequence alignment among the sequences of human, mouse, chicken, and the Drosophila NF1 proteins.
Genotyping for LOH
LOH analysis was performed to test the hypothesis of inactivation of the remaining NF1 wild-type allele in tumor DNA. Genomic DNA from fresh-frozen or paraffin-embedded tumor tissue was extracted using standard procedures. LOH was determined by typing genomic DNA from 21-germline point mutation positive paired tumor and lymphocyte samples with four microsatellite markers (D17S1849, D17S1166, D17S798, and D17S1873) spanning the NF1 region. The markers were located centromeric, intragenic, and telomeric of the NF1 gene. DHPLC was used to investigate the LOH status.
Statistical analyses
Fisher’s 2-tailed exact test was used with 
= 0.05 regarded as statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
). Male to female ratio was 18:19. Mean age at pheochromocytoma diagnosis was 41 yr, with an age range of 14–70 yr. Of the 37 patients, 62% had solitary adrenal tumors, 27% had bilateral adrenal disease, and 5% had extra-adrenal abdominal pheochromocytomas. Malignant pheochromocytomas were identified in 6% (n = 2) of the patients. One patient had an extra-paraganglial malignant pheochromocytoma, a medullary thyroid cancer, and hyperparathyroidism, and he was RET mutation negative. Interestingly, an unrelated patient with the identical NF1 mutation had neither medullary thyroid cancer nor other features of MEN2 (Table 3). There were also three women with breast cancer (10%). These tumors are indicated in Table 3 as "malignancies related to NF1."
|
|
). None of the mutations were found in 100 healthy, ethnically matched, control subjects. A total of 31 (86%) of the germline mutations were intra-exonic and splice site mutations. The spectrum of these point mutations comprised 17 truncating mutations (nonsense and frameshift type), nine affecting splicing, including NF1 c. 1466 A/G, creating a new splice site as described (26), and five missense mutations. The missense mutations all affected amino acids, which are conserved among human, mice, chicken, and Drosophila. Two of the five missense mutations (L90P, L578P) change the invariant amino acid leucine into proline, known as a helix-breaker in globular proteins (27), an exchange that may influence the secondary structure of NF1 protein. One mutation (L303R) affected a hydrophobic amino acid, which is present in mammals and Drosophila, changing it into a positively charged one. Cys 93 is mutated into a positively charged arginine, which lacks sulfhydryl groups (C93R). The fifth missense mutation is the exchange of a hydrophilic against a hydrophobic residue (S2580A).
Because of the phylogenetic conservation and the absence of these changes in healthy controls, these sequence variations can be classified as pathogenic missense mutations. Interestingly, one patient showed a homozygous 2-bp insertion (NF1 c. 2849 ins TT) in exon 16 of the NF1 gene. This result was repeated twice and also confirmed by using a second DNA probe from the patient’s blood. Her mother and grandmother were reported to have NF1, but blood DNA from them was not available. The patient’s father is unknown.
Of note, five (14%) of the germline mutations were large deletions or duplications. One patient was found to carry a deletion spanning exon 7 of the NF1 gene to exon 3 of the downstream RAB11.4 gene. Her phenotype did not appear different from the other deletion cases. Two patients had a single exon deletion of exon 12a and of exon 31, respectively, of the NF1 gene. In one patient, a deletion affecting exon 23b to exon 33 of the NF1 gene was detected, and one single exon duplication involving exon 15 was found in one patient.
The detected NF1 mutations, germline point mutations, as well as large deletions and duplications, were randomly distributed throughout the whole coding region of the NF1 gene, without any statistically significant correlations with mutational clustering or specific hot spot regions (Fig. 1). However, it is interesting to observe that although approximately one fourth of all reported NF1 cases had mutations within the cysteine-rich region [cysteine-serine rich domain (CSR)] and another one fourth within the RAS-GTPase activating protein domain (RGD) (28), 35% of our NF1-associated pheochromocytoma cases had mutations in the CSR and only 13% in the RGD (Fig. 1). The relative distribution of germline mutations within the RGD and CSR in our NF1-associated pheochromocytoma cases was reversed compared with that found historically among all NF1 cases (28) (P = 0.023).
|
). Yet, case 8 was operationally classified to have mild disease and case 9, severe.
LOH studies were performed for 21 blood/tumor pairs belonging to 19 patients with identified germline mutations (this study) and two additional NF1 patients in whom the germline mutation was not detected. LOH and, thus, inactivation of both NF1 alleles were found in 14 (67%) of the 21 paired pheochromocytoma/lymphocyte samples involving at least one of the four microsatellite markers spanning the NF1 region (Fig. 2). In contrast, no LOH was found in any of these four markers in 12 blood/tumor pairs belonging to sporadic pheochromocytoma (P = 0.0002).
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Our current study was able to overcome these challenges. First, and most importantly, we have, at our disposal, a large registry such that both clinical data and genetic data could be coupled. Second, we were able to design primers specific for the gene, avoiding amplification of the pseudogenes. Third, we were able to use the high throughput yet sensitive mutation scanning technology DHPLC. Finally, we extended our studies so that large deletions and rearrangements were systematically searched for, including fine, exon by exon deletion analysis using the highly sensitive quantitative real-time PCR approach.
Because of the low frequency of pheochromocytoma reported in NF1, there remains the question of whether pheochromocytoma is a true component of NF1 or a coincidental finding. One molecular clue that might help suggest that pheochromocytoma occurring in NF1 is a true component is to observe the loss of the remaining wild-type allele in NF1-related pheochromocytomas (20, 21, 22, 23). We believe that the 67% 17q LOH frequency in NF1-related pheochromocytoma compared with 0% in sporadic tumors (P = 0.0002) does provide molecular evidence that, indeed, pheochromocytoma is a true component of NF1.
In total, we detected germline NF1 mutations in 37 subjects with pheochromocytoma displaying the NF1 phenotype. To date, almost 80% of the 36 different germline mutations have not been previously described compared with 21%, which are summarized in the international database of NF1 mutations (http://www.nfmutation.org/). Of all germline mutations, 86% were intra-exonic and splice site mutations. In contrast with the previously reported 4–7% (34, 35) detected by fluorescence in situ hybridization or multiplex ligation-mediated probe amplification analysis, 14% were large deletions ranging from single exon deletion to nearly entire gene deletions (51 of 57 exons). Instead of the severe phenotype with dysmorphic features and intellectual impairment described for patients with large deletions, e.g. of 1.5 Mb of the NF1 gene and 11 adjacent genes, patients in our study with truncating mutations, including stop codons, frameshifts, and deletions ranging from single exon deletions to a loss of a major part of the NF1 gene, show a phenotype that does not differ from the phenotype caused by NF1 point mutations (36, 37). It is possible that the severity of NF1 phenotype might be modulated by differential deletion of the genes surrounding NF1 (38).
Despite almost 15 yr since the cloning of NF1, the full spectrum of its protein function is likely still not completely understood. One of the most important functional domains in neurofibromin, encoded by NF1, is the GAP related domain, which is encoded by exons 20–27a (Fig. 1) and has been found to interact with the RAS-GTPase p21ras (therefore, RAS-GAP domain or RGD) (39). Activation of p21ras by neurofibromin’s RGD stimulates GTP hydrolysis that can mediate the control of cellular proliferation (40, 41) and apoptosis (42). Consistently, missense mutations of the RGD with moderate reduction of GAP activity have already been reported (43) and may be sufficient to cause NF1.
Functional studies have demonstrated the role of NF1 in the pathogenesis of pheochromocytoma, and of the importance of the RGD in this regard (28, 36). To date, at least a quarter of NF1 mutations, mainly in series ascertained by NF1, are in the RGD. However, only 13% of the germline mutations in our series of pheochromocytoma patients who had NF1were within the RGD, in contrast to the 35% in the CSR. The relative distribution of mutations favoring CSR over RGD in our cases compared with historical reports is significant (P = 0.023). Consistent with our genetic observations here, there is recent emerging evidence that the CSR could be an equally important functional domain in neurofibromin. Using transfection experiments in human, rat, and avian central nervous system cells and cell lines, it has been shown that in response to epidermal growth factor, neurofibromin is phosphorylated on serine residues by protein kinase (PK) C (PKC) and that this phosphorylation was prominent in the CSR domain (44). They also showed that CSR can allosterically regulate RGD dependent on PKC-mediated neurofibromin phosphorylation of the serines within the CSR. The CSR also has three cysteine pairs suggestive of a region for ATP binding and has three cAMP-dependent PK recognition sites phosphorylated by PKA (45). Interestingly, the regulatory subunit 1 of PKA is encoded by a gene (PRKAR1A), which when mutated in the germline, causes a subset of Carney complex, another neural crest-associated phakomatosis (46). These and our genetic data suggest that the CSR could be a prominent mutational target, at least in NF1-associated pheochromocytoma, and perhaps more universally in all heritable and sporadic pheochromocytoma as well.
In summary, we have documented the germline NF1 mutational spectra in 37 subjects with pheochromocytoma and NF1. In contrast to the previously described mutational spectra of NF1 cases, we have shown that NF1-associated pheochromocytoma patients have germline NF1 mutations that favor the CSR over the RGD. These genetic data might help direct functional evaluation of neurofibromin in its role in both heritable and sporadic pheochromocytoma and paraganglial tumors.
|
Acknowledgments |
|---|
The members of the Neurofibromatosis Type 1 Study-Group are: R. Stadler, Minden; H. Ising, Sindelfingen; J. Kunze, Duisburg; D. Hördemann-Ebler, Gaggenau; T. Bierwirth, Bielefeld; H. Engler, Freiburg; B. Merdausl, Kempen; H.-G. Bongartz, Fritzlar; U. Wlotzke, Ulm; R. Leitz, Stuttgart; A. Rübben, Aachen; G. Hübner, Halle; T. Schwarz, Kiel; E. Wassmer, Augsburg; J. Nekwasil, Nordhausen; J. Reifenberger, Düsseldorf; H. Kirchesch, Pulheim; M. Schlaeger, Oldenburg; R. Barth, Kirchzarten; and C. G. Schirren, Darmstadt.
We thank B. Wehrle (technician), M. Buchta (technician), M. Sullivan, Ph.D., G. Franke, Ph.D., G. Schluh (research assistant), and Z. Nabulsi and S. Schonhardt (study nurses) for their excellent assistance. We also thank the German Neurofibromatosis Self Support Group for its support.
|
Footnotes |
|---|
Disclosure Summary: B.B., W.B., V.F.M., M.H.H., D.B., M.R., A.C., T.H., F.S., C.P., M.P., C.L., S.C., G.A., R.D.K.-N., N.R., A.F., L.B., M.A.W., M.Ma., G.M., F.F.P., M.B., M.K.W., B.K., G.B., R.P., A.-C.K., F.L., M.Mo., O.G., B.J., S.R.M., G.O., A.J., J.K., C.E., and H.P.H.N. have nothing to declare. Nobody was previously employed by a company or has equity interests. Nobody consulted for a company. Nobody was previously employed or received lecture fees from a company. Nobody is an inventor on a patent.
First Published Online April 10, 2007
1 For a list of members of the European-American Phaeochromocytoma Registry and Study Group, see Acknowledgments. 
Abbreviations: CSR, Cysteine-serine rich domain; DHPLC, denaturing HPLC analysis; LOH, loss-of-heterozygosity; MEN2, multiple endocrine neoplasia type 2; NF1, neurofibromatosis type 1; PGL, paraganglioma syndrome; PK, protein kinase; RGD, RAS-GTPase activating protein domain.
Received December 21, 2006.
Accepted April 3, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
regulatory subunit in patients with the Carney complex. Nat Genet 26:89–92[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  A. Karagiannis, D. P Mikhailidis, V. G Athyros, and F. Harsoulis Pheochromocytoma: an update on genetics and management Endocr. Relat. Cancer, December 1, 2007; 14(4): 935 - 956. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
