American Journal of Medical Genetics 46.182-184 (1993)

Mutations in von Recklinghausen Neurofibromatosis: An Hypothesis Jog1 Zlotogora Department of Human Genetics, Hadassah Medical Center, Hebrew University, Jerusalem, Israel

Von Recklinghausen neurofibromatosis or neurofibromatosis type I (NF1) is a relatively frequent (1/3,000 livebirths) autosomal domi- nant condition. Some unusual aspects are noted in this disorder: new mutations are fre- quent and almost all are of paternal origin without parental age effect. The recurrence of NF1 among children of healthy parents is rare as opposed to other dominant disorders. I propose that in NFl(1) new mutations occur often in somatic cells or in late germinal cells, however, they occur very rarely in early ger- minal cells leading to germinal mosaicism and (2) the individual with somatic mosaicism presents symptoms of the disease. Therefore, an NF1 patient with an apparent new muta- tion is often a somatic mosaic for the mutation and if the mosaic is also present in germinal cells some of his children will be affected. This hypothesis may explain the unusual aspects of mutation in NFl. o 1993 WilepLiss, Inc.

KEY WORDS germinal mosaic, mosaicism, mutation, neurofibromatosis type I, somatic mosaic, Von Recklinghausen neurofibro- matosis

INTRODUCTION Von Recklinghausen neurofibromatosis or neurofibro-

matosis type I (NF1) is an autosomal dominant genetic disorder which is found in all populations with a rela- tively high incidence [Riccardi and Eichner, 19861. The disorder presents with some interesting and unusual characteristics: (1) the mutation rate is relatively high [Crowe et al., 1956; Huson et al., 19891; (2) almost all of the new mutations are of paternal origin [Jadayel et al., 1990; Stephens et al., 19921; (3) no parental age effect is

Received for publication July 21, 1992; revision received Octo- ber 28, 1992.

Address reprint requests to Joel Zlotogora, M.D., Department of Human Genetics, Hadassah Medical Center, POB 12,000, Jerusa- lem 91120 Israel.

0 1993 Wiley-Liss, Inc.

seen in the new mutations [Huson et al., 19891; (4) ger- minal mosaicism is rare [Huson et al., 1989; Riccardi and Eichner, 19861.

The Mutations Rate Is Very High in NFI The mutation rate in NF1 was estimated to be

1 x [Crowe et al., 1956; Huson et al., 19891, which is much higher than the rate found in most other disor- ders: between 1 x and 1 x [Vogel and Mo- tulsky, 19861. The molecular basis for the high mutation rate is not known. The size of the gene (300 Kb) cannot fully account for the high mutation rate and it was suggested that the size of the gene as a target for new mutations may only explain an increase by l o x [Marchuk et al., 19911. In Duchenne muscular dystro- phy (DMD) an X-linked disorder, a similarly high muta- tion rate is found. However, the DMD gene is large and more than 50% of the mutations are caused by deletions [Darras et al., 19881. In NF1, up to now only few dele- tions have been found and most cases are probably due to point mutations [Marchuk et al., 19911.

Most of the Mutations Are of Paternal Origin In 2 different studies [Jadayel et al., 1990; Stephens et

al., 19921 both using DNA polymorphisms which al- lowed a determination of the haplotypes around the NF1 gene in families of patients with NF1; in 22 out of 24 cases the new mutation occurred on the paternal chro- mosome. This is significantly different from ratio ex- pected if the new NF1 mutations occurred at random on the maternal or paternal chromosome 17.

No Parental Age Effect on the Frequency of NF1 Mutations

In various autosomal dominant diseases, for instance achondroplasia, Marfan, or Aped syndrome, it was dem- onstrated that the frequency of new mutations increases with paternal age [Vogel and Motulsky, 19861. It is thought in these cases that most mutations arise by copy error at the time of DNA synthesis. It was suggested that the mutations occur mostly in the paternal germ cells and are influenced by age, since they are related to the number of cell division. In NF1 no effect of the paternal age on the frequency of new NF1 mutations was observed [Huson et al., 19891. However, in some 34% of the cases of NF1 secondary to a new mutation an-

Mutations in Neurofibromatosis 183

alysed by Riccardi et al. [19841 had a father of 35 years or older. In this study it was demonstrated that older fathers have an increased risk for a new mutation in the NF1 gene.

Germinal Mosaicism Is Rare in NF’1 Only a few families were reported in which both par-

ents who have no signs of NF1 had more than one af- fected child [Huson et al., 1989; Riccardi and Lewis, 19881. This is unexpected since germinal mosaicism has been found to be relatively frequent in many other auto- soma1 dominant disorders such as pseudoachondro- plasia [Hall et al., 19871 or osteogenesis imperfecta [Cohn et al., 19901. For instance, after the birth to nor- mal parents of a child with lethal osteogenesis imper- fecta type 11, the recurrence risk is estimated to be 6-7% [Cohn et al., 19901. NF1 is more frequent than other dominant disorders and often caused by new mutations; if a similar rate of germinal mosaicism was present in NFl, many families in which both parents are healthy and more than one child is affected would have been reported. The rarity of such families cannot be ac- counted for by difficulties in diagnosis since clinical criteria for diagnosis of NF1 are well known and the penetrance is almost 100% in adults [Riccardi and Lewis, 19881. Therefore it seems that the rarity of “ger- minal mosaicism” represents a true difference between NF1 and other dominant diseases.

Hypothesis: Many New Mutations in NF1 Are Somatic Mutations Which

Are Clinically Expressed Because of the knowledge of the biochemical and mo-

lecular basis of disorders caused by mutations in col- lagen genes in some of these disorders the origin of new mutations may be studied. For instance in osteogenesis imperfecta type 11, a lethal form of the disease, collagen I is abnormal and mutations have been reported in the genes COLlAl and COLlA2 [Beighton et al., 19921. Biochemical investigations may determine the presence of the abnormal type I collagen and molecular analysis may detect the mutation. The analysis may be done in various tissues of an affected patient and his parents; a particularly powerful tool is PCR amplification which allows the detection of mosaicism even at very low levels. Cohn et al. [19903 demonstrated the existence of paren- tal germinal mosaicism in 11 families with children affected with type I1 osteogenesis imperfeda, out of ap- proximatively 140 families studied. Among the families in which osteogenesis imperfecta type I1 was due to germinal mosaicism in one of the parents, it was demon- strated that a t least in 4, a somatic mosaicism was present with subtle or no clinical symptoms in the indi- vidual [Beighton et al., 19921. In addition, a male af- fected with a relatively mild form of osteogenesis imper- fecta had a child with the lethal form of the disease [Wallis et al., 19921. The molecular analysis demon- strated that both the father and the child had the same mutation; the father had a mild disease since he was mosaic for the mutation.

I proposed that in NFl(1) the lack of apparent ‘‘germi-

occur early in germinal cells giving rise to a number of mutated daughter cells. (2) Many of the new mutations are somatic and in some cases occur early in develop- ment and therefore are present in many tissues includ- ing the germinal cells. (3) Somatic mutations are clini- cally expressed. If the somatic mutations occurred early they will present clinically as general neurofibromato- sis: NF1, arid perhaps if they occurred later they will appear as segmental neurofibromatosis: NF5. There- fore, many patients with somatic mosaicism appear as affected with NF1; they will present as cases of “new mutations” and some will be able to transmit the disease because of the germinal mosaicism. The occurrence of most of the new mutations in somatic cells explains why there is no influence of paternal age. The preferential occurrence of the mutations on the paternal chromo- some may be because of a difference between the 2 genes such as methylation as has been suggested [Jadayel et al., 1990; Stephens et al., 19921. As in other dominant diseases, new mutations may also occur in paternal ger- minal cells and their incidence is influenced by the paternal age, explaining the increased risk for fathers older than 35 years [Riccardi et al., 19841. The hypoth- esis may also explain, at least in part, the apparently high mutation rate in NF1 since in fact there are 2 different types of mutations which are seen as new, one somatic, the other germinal, while in other disorders, only germinal mutations are taken in account to calcu- late the mutation rate.

With the completion of the sequence of the NF1 gene [Marchuk et. al., 19911 and analysis of mutations in patients with NF1 it will be possible to test the proposed hypothesis and determine whether many of the cases with “new mutations” in NF1 have in fact a somatic mosaicism.

Clinical analysis of mutations in other disorders [Hall, 19883 suggests that this may be a general phe- nomenon: mutations in early germ cells may be very rare, while somatic mutations are relatively frequent. Therefore, germinal mosaicism in dominant disorders is mostly secondary to somatic mosaicism. The difference between NFY and other autosomal dominant disorders is that most Df the individuals with somatic mosaicism in other disorders are healthy.

REFERENCES Beighton P, De Paepe A, Hall JG, Hollister DW, Pope FM, Pyeritz RE,

Steinmann H, Tsipouras P (1992): Molecular nosology of heritable disorders of connective tissue. Am J Med Genet 42431-448.

Cohn DH, Starman BJ, Blumberg B, Byers PH (1990): Recurrence of lethal osteogenesis imperfecta due to parental mosaicism for a dominant mutation in human type I collagen gene (COLlAl). Am J Hum Genet 46591-601.

Crowe FW, Schull WJ, Nee1 JV (1956): “A Clinical, Pathological and Genetic Study of Multiple Neurofibromatosis.” Springfield, I L Charles Thomas.

Darras BT, Blattner P, Harper JF, Spin AJ, Alter S, Francke U (1988): Intragenic deletions in 21 Duchenne muscular dystrophy (DMDY Becker muscular dystrophy (BMD) families studied with the dys- trophin cDNA: Location of breakpoints on HindIII and BglII exon containing fragment maps, meiotic and metotic origin of the muta- tions. Am J Hum Genet 43:620-629.

Hall JG (1988): Review and hypotheses: somatic mosaicism: Observa- tions related to clinical genetics. Am J Hum Genet 43:355-363. nalmosaicism” is due to the absence of mutations-which

184 Zlotogora

Hall JG, Dorst JP, Rotta J, McKusick VA (1987): Gonadal mosaicism in pseudoachondroplasia. Am J Med Genet 28143-151.

Huson SM, Compston DAS, Clark P, Harper PS (1989): A genetic study of von Recklinghauusen neurofibromatosis in Southeast Wales. I Prevalence, fitness, mutation rate, and effect of parental transmis- sion on severity. J Med Genet 26:704-711.

Jadayel D, Fain P, Upadhyaya M, Ponder MA, Huson SM, Carey J, Fryer A, Mathew CGP, Barker DF, Ponder BAJ (1990): Paternal origin of new mutations in von Recklinghausen neurofibromatosis. Nature (London) 343:558-559.

Marchuk DA, Saulino AM, ‘Igvakkol R, Swaroop M, Wallace MR, Andersen LB, Mitchell AL, Gutmann GH, Boguski M, Collins FS (1991): cDNA cloning of the type 1 neurofibromatosis gene: Com- plete sequence of the NF1 gene product. Genomics 11:931-940.

Riccardi VN, Dobson CE, Chakraborty R, Bontke C (1984): The pa- thophysiology of neurofibromatosis: IX. Paternal age effect on the origin of new mutations. Am J Med Genet 18:169-176.

Riccardi VM, Eichner JE (1986): “Neurofibromatosis.” Baltimore: Johns Hopkins University Press.

Riccardi VM, Lewis RA (1988): Penetrance of von Recklinghausen neurofibromatosis: A distinction between predecessors and descen- dant. Am d Hum Genet 42284-289.

Stephens K, Kayes L, Riccardi VM, Rising M, Sybert VP, Pagon RA (1992): Preferential mutation of the neurofibromatosis type 1 gene in paternally derived chromosomes. Hum Genet 88:279-282.

Vogel F, Motulsky AG (1986): “Human Genetics. Problems and Ap- proaches,” 2nd ed. Berlin: Springer-Verlag.

Wallis GA, Starman BJ, Zinn AB, Byers PH (1990): Variable expres- sion of osteogenesis imperfecta in a nuclear family is explained by somatic mosaicism for a lethal point mutation in the aKl) gene (COLlAl) type of collagen in a parent. Am J Hum Genet 461034-1040.