|
|
| |||||||||||||||||||||||||||||||
Background The mutation rate of the neurofibromatosis type 1 (NF1) gene is one of the highest in the human genome, with about 50 percent of cases being due to new mutations. We describe a family in which neurofibromatosis type 1 occurred in two siblings with clinically normal parents, and we demonstrate germ-line mosaicism in the father.
Methods We studied lymphocyte DNA from each member of the family and the father's spermatozoa for several polymorphic intragenic markers of the NF1 gene. Southern blots of DNA digested with several enzymes were hybridized with complementary DNA and individual NF1 exon probes to search for alterations in the gene.
Results The affected siblings, with a clinically severe form of neurofibromatosis type 1, showed no inheritance of paternal alleles for a marker in intron 38 of the NF1 gene, whereas they received alleles from both parents for other NF1 markers. Analysis with probes from this region of the NF1 gene showed a 12-kb deletion of the NF1 gene, involving exons 32 to 39, in the affected offspring. Ten percent of the father's spermatozoa carried the same NF1 deletion, but the abnormality was not detected in DNA from his lymphocytes.
Conclusions The presence of the NF1 mutation in 10 percent of the clinically normal father's spermatozoa supports the hypothesis that most germ-line mutations occur in precursors of gametes. In cases of spontaneous mutation, analyzing the specific NF1 mutation in the father's sperm might help in the detection of mosaicism and thus facilitate genetic counseling about further pregnancies.
Neurofibromatosis type 1 has many interesting genetic features. Almost everyone who inherits the NF1 gene has clinical features of the disease by the age of five years. Despite this high penetrance of the NF1 gene, the clinical expression of the disease is variable, even in members of the same family10. The mutation rate of the NF1 gene (1 x 10-4) is one of the highest in the human genome; about 50 percent of cases of neurofibromatosis type 1 are due to new mutations11,12. Factors that might contribute to the high incidence of new cases of neurofibromatosis type 1 are the large size of the gene (about 350 kb), an unequal exchange of genetic material between homologous chromosomes during meiosis, and sequences in the gene that are highly susceptible to mutation. Interestingly, the majority of sporadic mutations in neurofibromatosis type 1 arise in paternally inherited alleles13.
In this report we describe a family in which neurofibromatosis type 1 occurred in two siblings with clinically unaffected parents. Three explanations for this unusual family history were possible: nonpenetrance of the disease in one of the parents, germ-line mosaicism in one of the parents, or two independent mutation events. The presence of mutations in DNA from the sperm (or possibly ova) of the clinically unaffected parent should provide evidence of germ-line mosaicism. Molecular analysis of the NF1 gene in this family demonstrated germ-line mosaicism in the father of the two siblings with neurofibromatosis type 1. The finding that approximately 10 percent of the father's sperm had the NF1 mutation supports the hypothesis that most germ-line mutations occur in the precursor cells of gametes. Male germ-line mosaicism could also explain the predominance of paternally derived alleles in sporadic cases of neurofibromatosis type 1.
Methods
Patients
As part of our work in genetic counseling and the diagnosis of neurofibromatosis type 1, we have analyzed 130 families with the disease since 1990 -- 128 from Spain and 2 from Italy. The cases of neurofibromatosis type 1 were identified by standard criteria1,2. DNA from peripheral blood was obtained from every member of each family14. In the family discussed here (Figure 1), DNA was also obtained from the father's spermatozoa.
|
Several microsatellite markers (short tandem-repeat sequences), located in intronic regions of the NF1 gene (IVS27AAAT2.1 and IVS27AC33.1 in intron 27 and IVS38GT53.0 in intron 38), were analyzed by independent polymerase-chain-reaction amplification, as previously described15,16,17. Paternity was assessed by analyzing five microsatellites (D1S117, D6S89, D11S35, APOC2, and D21S168) located on different chromosomes18.
Southern Blot Analysis
For pulsed-field gel electrophoresis, 0.5 million fresh human lymphocytes were embedded in agarose plugs and lysed; the DNA was then digested with BamHI, ClaI, and XhoI (New England Biolabs)19. The DNA fragments were separated by pulsed-field electrophoresis in 1.2 percent agarose gels with TRIS-borate-EDTA electrophoresis buffer (45 mM TRIS borate and 1mM EDTA) at 4 °C. Separation of DNA fragments ranging from 5 to 60 kb was achieved after 4 hours of electrophoresis at 450 V with pulse times of 0.3 second for 1 hour, 0.5 second for 1 hour, and 0.65 second for 2 hours with an LKB Pulsaphor (Pharmacia).
For non-pulsed-field Southern blotting, 5 µg of DNA was digested with EcoRI, HindIII, PstI, and MspI, and the fragments were separated on a 0.8 percent agarose gel. After blotting, nylon filters were hybridized with the NF1 complementary DNA (cDNA) probes AE25 and P5,5 as well as with individual NF1 exon probes. Hybridization and autoradiography were performed as previously described20. Densitometric analysis of the autoradiographs was performed on a Preference HR densitometer (Sebia).
Results
Clinical Data
Both affected siblings had a clinically severe form of neurofibromatosis type 1. The affected daughter had cafe au lait spots and axillary freckling shortly after birth. From the age of 12 she had several cutaneous and two plexiform neurofibromas. A retroperitoneal neurofibrosarcoma found at the age of 26 caused her death six months later. Her 14-year-old brother had cafe au lait spots and axillary freckling at birth. He had Lisch nodules in the iris, but no cutaneous neurofibromas. He also had a learning disability, problems at school, and a speech impairment like his sister's. A magnetic resonance scan detected a left paraventricular cerebellar astrocytoma. Both siblings had mild dorsal scoliosis. The parents and an asymptomatic sister were examined thoroughly and found to be clinically normal. Additional studies of the father, including a cranial magnetic resonance scan, skeletal radiography, and an ophthalmologic examination, failed to detect any sign of neurofibromatosis type 1.
Microsatellite Studies
Analysis of NF1 microsatellite markers in lymphocyte DNA from family members showed that the two affected siblings did not inherit paternal alleles for microsatellite marker IVS38GT53.0, whereas they inherited alleles from both parents for other intragenic polymorphic markers for neurofibromatosis type 1 (Figure 1). The absence of the paternal allele for the IVS38GT53.0 marker in the siblings with neurofibromatosis type 1 was indicative of a deletion in the NF1 gene. Since this intragenic deletion of NF1 was not present in the father's lymphocytes, we can postulate that it arose spontaneously in the father's germ-line cells. Paternity was confirmed with five microsatellite markers located on different chromosomes (D1S117, D6S89, D11S35, APOC2, and D21S168),18 with a probability greater than 99.9 percent.
Identification of a 12-kb Deletion in the NF1 Gene
To confirm the putative NF1 deletion in the affected siblings with neurofibromatosis type 1, we used probes of individual exons of the NF1 gene to hybridize Southern blots of lymphocyte DNA digested with several restriction enzymes (EcoRI, MspI, HindIII, and PstI). Each member of the family was studied in this way; Figure 2A shows the results for the father and the affected boy. When the digested DNA was probed with exon 31 of the NF1 gene (see the Methods section), bands were detected in both siblings with neurofibromatosis type 1 that were not found in normal persons and the parents. These extra bands were present in fragments produced by each of the four restriction enzymes. Probes including exons 32 to 39 detected normal fragments under the same conditions, whereas a probe from exon 40 produced the same abnormal pattern as exon 31 (data not shown). These additional bands were indicative of a deletion in the NF1 gene and confirmed the data obtained with the IVS38GT53.0 marker (Figure 1).
|
We examined DNA obtained from the spermatozoa of the father and compared it with DNA from his lymphocytes and DNA from his two offspring with neurofibromatosis type 1. The same abnormal fragments that were found in the digested DNA from the lymphocytes of the offspring were detected in the father's sperm DNA, but not in the DNA from his lymphocytes (Figure 3). We estimated by densitometric scanning of the gel autoradiograph that approximately 10 percent of the father's spermatozoa carried the NF1 deletion, whereas the mutation was undetectable in his lymphocytes, even after long autoradiographic exposures. Therefore, approximately 10 percent of the father's spermatozoa contained the 12-kb NF1 deletion.
|
We have studied a family in which a brother and sister with neurofibromatosis type 1 had clinically normal parents. Molecular analysis detected a 12-kb deletion in the NF1 gene in the lymphocytes of the offspring with neurofibromatosis type 1, but only in the sperm of the clinically unaffected father. These results indicate germ-line mosaicism of the NF1 gene in the father.
Although the NF1 gene was isolated in 1990, the number of mutations that have been identified so far is very low (about 10 percent). Therefore, identifying the specific NF1 mutation in each patient and family for diagnostic purposes is possible in only a few cases. Intragenic microsatellite markers, which span the NF1 gene,15,17 allow indirect genetic analysis in cases in which the NF1 mutation is unknown16. The large number of alleles detected by microsatellite markers permits a distinction between the normal gene and the disease gene by analysis of the segregation of alleles in a family. These markers can also identify deletions through the absence of inheritance of alleles in the transmission from a parent to an offspring with neurofibromatosis type 116.
In the course of characterizing families with neurofibromatosis type 1 with NF1 intragenic markers, we found that two siblings with the disease did not inherit the paternal allele for a microsatellite located in intron 38 of the NF1 gene, although they received the expected allele from the mother. Since other markers of the NF1 gene and markers of other chromosomes segregated correctly, demonstrating paternity with a high degree of certainty, the finding in this family suggests an intragenic deletion involving intron 38. The presence of an intron-38 allele marker from the mother and the absence of an allele from the father demonstrate that the spontaneous mutation was paternal in origin. Further analysis with probes from this region of the NF1 gene showed that the phenotype for severe neurofibromatosis type 1 in the two siblings was associated with a 12-kb deletion that was inherited from the clinically normal father. Analysis of DNA from the father's sperm showed that about 10 percent of his germ cells harbored the same 12-kb deletion that his affected offspring had. Since the father had the mutation in his germ-line cells but not in his lymphocytes, we can conclude that the father was mosaic for the 12-kb deletion. The fact that only 10 percent of the spermatozoa contained the mutation yet two of his three offspring were affected is due solely to chance.
Mosaicism (the coexistence in a person of normal and mutated cell populations) has been implicated in families in which the parents are phenotypically normal, but more than one of their offspring is affected with a dominant or X-linked disorder21. Mutations during early embryonic development, before the determination of the germ line, will cause gonosomic mosaicism (affecting the majority of somatic tissues and also the germ-line cells). Mutations that occur later can affect either the germ cells alone (germ-line mosaicism), or the somatic cells alone (somatic mosaicism). Germ-line mosaicism can explain situations in which clinically normal parents have more than one child with a genetic disorder. If the mutation also affects cell types other than gametes (somatic and gonosomic mosaicism), a mild form of the disease can also occur in one of the parents. These types of mosaicism have been reported,21,22,23,24,25,26,27,28,29,30,31 but the molecular demonstration of somatic or germ-line mosaicism has been possible in only a few instances, either by linkage analysis or, exceptionally, by detection of the mutation26,27,28,29,30,31.
The demonstration that about 10 percent of the sperm from the clinically normal father of two affected offspring harbored the NF1 mutation has important implications, both for our understanding of mosaicism and for its role in the transmission of some genetic disorders. The 10 percent of the father's sperm that carried a mutated NF1 gene must have resulted from a mitotic mutation in a germ-line stem cell, a precursor of the male gametes; this finding supports the hypothesis that most germ-line mutations are mitotic in origin21. Since male germ cells have greater mitotic activity than female germ cells and the risk of mutation increases with the number of cell cycles, the sex bias observed in neurofibromatosis type 1, in which about 90 percent of sporadic mutations arise in paternally derived alleles,13 may be related to male mosaicism. Until very recently, molecular methods were not available to confirm heterozygosity for NF1 mutations in the sperm of fathers of patients with neurofibromatosis type 1 caused by sporadic mutations. Analysis of these cases should provide further information on the possible role of mosaicism in the high mutation rate observed in neurofibromatosis type 1 (3.1 x 10-5 to 1.04 x 10-4)12.
The identification of mosaicism for a genetic disorder is important for counseling in regard to further pregnancies and for assessing the clinical status of the disease in the person with the mosaicism. Detecting a spontaneous mutation in gametes and determining the proportion of cells that carry the defect provide the basis for estimating the risk of the disease in siblings of the proband. For neurofibromatosis type 1, in which about 50 percent of cases are due to new mutations, genetic counseling should take into account the possibility that the mutation is of paternal origin and (if applicable and possible) should assess the proportion of mutated cells in spermatozoa. Penetrance of the NF1 mutation may vary between ancestors and descendants of the affected person10. Penetrance is almost 100 percent in descendants, but the parent of a person with neurofibromatosis type 1 may express the mutant phenotype only partially or not at all10,12,32. The absence of expression in one of the parents is probably due to mosaicism. Cases of segmental neurofibromatosis type 1, in which only one part of the body is affected,32 and cases of limited manifestation (e.g., only Lisch nodules)10 could also be explained by mosaicism.
The high incidence of sporadic cases of neurofibromatosis type 1, the male bias in the transmission of new mutations, and the possibility of germ-line mosaicism make genetic counseling about this disease particularly difficult. Since about 90 percent of sporadic NF1 mutations arise in paternally inherited alleles, analyzing the specific NF1 mutation in the father's sperm may help in detecting mosaicism and may facilitate genetic counseling. Finally, information on the wide spectrum of NF1 mutations or the discovery of a special mutation mechanism in the NF1 gene may provide the basis for an understanding of the genetic puzzle of neurofibromatosis type 1 and should supply us with better tools for diagnosis and prevention.
Supported by grants from the Institut Catala de la Salut and the Fondo de Investigaciones Sanitarias de la Seguridad Social (92-0532). Dr. Lazaro is a fellow of the Comissio Interdepartamental de Recerca i Innovacio Tecnologica of the Generalitat of Catalonia.
We are indebted to F.S. Collins for probes AE25 and P5; to Michael Lynch, Melanie Pritchard, and Helena Kruyer for advice and help with the manuscript; and to the members of the study family for their collaboration.
Source Information
From the Institut de Recerca Oncologica, Molecular Genetics Department, Hospital Duran i Reynals, L'Hospitalet de Llobregat (C.L., A.R., A.G., V.V., X.E.); the Hospital de la Creu Roja (A.R.); and the Genetics Service of the Hospital Clinic (X.E.) -- all in Barcelona, Spain.
Address reprint requests to Dr. Estivill at the Institut de Recerca Oncologica, Hospital Duran i Reynals, L'Hospitalet de Llobregat, E08907 Barcelona, Catalonia, Spain.
References
| |||||||||||||||||||||||||||||||
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 © 2010 Massachusetts Medical Society. All rights reserved. |
Previous
