| |||||||||||||||||||||||||||||||
Tuberous sclerosis is an autosomal dominant disorder characterized by the development of unusual tumor-like growths (hamartomas) in multiple organs.3,4 Arguably the most important hamartomas are cerebral cortical tubers, which are regions of abnormal cortical architecture with distinctive large neuronal cells. Cortical tubers cause some of the most important clinical manifestations of tuberous sclerosis: epilepsy, mental retardation, and abnormal behavior including autism.3,5,6 Other hamartomatous lesions in tuberous sclerosis include subependymal nodules, facial angiofibromas, subungual fibromas, forehead plaques, shagreen patches, cardiac rhabdomyomas, and renal angiomyolipomas and cysts. The incidence of tuberous sclerosis is 1 in 6000 births, and about two thirds of cases are sporadic, occurring in the absence of a family history of the disorder.7,8 Mutations in one of two genes, TSC1 and TSC2, cause tuberous sclerosis.9,10
We describe a patient with severe tuberous sclerosis in whom a mutated TSC1 allele was present in only one third of leukocytes and in different proportions in other tissues. This case report illustrates the importance of considering mosaicism in many clinical settings and the limitations of molecular diagnostic methods, including complete gene sequencing, for the detection of mosaicism.
Case Report
An infant girl weighed 3000 g when born at term. Her mother was 20 years old, and her father was 24 years old; both were healthy. The baby's development was normal until the age of 18 months, when myoclonic seizures occurred, which were treated with corticotropin. Although the seizures stopped, the child subsequently learned few additional words and withdrew from interaction with her parents and others. At the age of 12 years she began to have absence seizures, and she was seen at a tertiary center to assess whether a genetic disease was present. Evaluation revealed a normal body habitus but limited activity and diminished social interaction. The child spoke simple sentences and could not read or count, but she could obey simple commands. On formal testing, the IQ score was less than 40. Several angiofibromas were present in the malar regions. Magnetic resonance imaging and computed tomography of the brain showed several calcified subependymal nodules, two large cortical tubers — one of which was calcified — and many smaller cortical tubers (Figure 1A, Figure 1B, Figure 1C, and Figure 1D). The results of renal ultrasonography, echocardiography, and retinal examination were all normal.
|
Leukocyte DNA from the patient was included in a set of DNA samples from 161 patients who met formal diagnostic criteria for tuberous sclerosis.4 These samples were screened for mutations in the TSC1 gene by heteroduplex analysis of all coding exons after amplification with the polymerase chain reaction (PCR).9,11 All patients or their parents gave informed consent, and the study was approved by the human research committee of Brigham and Women's Hospital.
Heteroduplex analysis is a method for the detection of variations in DNA sequences that depends on the different patterns of migration during electrophoresis of DNA duplexes that are mismatched at one or more internal nucleotides (Figure 2). We used conformation-sensitive gel electrophoresis11 followed by SYBR Green I or silver staining to identify heteroduplexes. Sequencing was performed with Taq polymerase in cycle sequencing with 32P-labeled oligonucleotide primers. The PCR products were subcloned into pUC18. Multiple pairs of oligonucleotide primers were used to detect the wild-type allele and the mutant allele with a deletion of the adenine (A) and cytosine (C) residues at position 2122 (2122delAC) in hybridization.
|
1–2 x h, where m is the degree of mosaicism and h is the heteroduplex fraction. This equation reflects the fact that more mutant alleles form heteroduplexes as the fraction of mutant alleles decreases. Results
Amplification of exon 15 of TSC1 of the patient's leukocyte DNA demonstrated a heteroduplex shift (Figure 2) suggestive of a mutation, but the intensity of the heteroduplex bands was less than that in DNA samples from other patients, including a previously described patient with tuberous sclerosis with an identical heteroduplex shift9 (Figure 3A). Sequence analysis of the patient's amplified exon 15 initially demonstrated no changes, but closer inspection revealed weak additional bands, suggestive of the presence of mosaicism, with a reduced contribution of the mutant allele (Figure 3B). The mutation appeared to be 2122delAC, identical to that found in the other patient with the same but more intensely staining heteroduplex shift.
|
To explore the extent of mosaicism in different tissues of this patient, we amplified exon 15 in DNA prepared from urine, hair roots, and buccal mucosa (Figure 3C). DNA from urine and hair roots formed heteroduplexes, indicating that they contained the mutant allele. However, a sample of buccal-mucosa DNA had no heteroduplex product, suggesting that the mutant allele was absent in that sample. As expected, corresponding samples from the patient's father and mother showed no heteroduplex formation.
To confirm and quantify these findings, we used PCR to amplify the region with the mutation with eight different pairs of flanking primers, and we analyzed both the original set of DNA samples and a second set. We then assessed the extent of heteroduplex formation in a quantitative manner (a representative gel is shown in Figure 3D). The DNA sample from the patient with the same mutation as the index patient but without mosaicism had a mean (±SD) heteroduplex fraction of 0.51±0.04 (eight samples), as compared with a predicted value of 0.50. The heteroduplex fraction in leukocyte DNA from the index patient was 0.29±0.04 (eight samples), which corresponds to a frequency of the mutated allele of 0.35. DNA samples from other sites had heteroduplex fractions ranging from 0.11 to 0.33, corresponding to a frequency of the mutated allele ranging from 0.12 to 0.42.
Discussion
Our patient had severe tuberous sclerosis characterized by seizures, mental retardation, and facial angiofibromas (the Vogt triad3), as well as by abnormalities on magnetic resonance imaging of the brain, and an inactivating mutation in the TSC1 gene. However, the mutation was present in only about one third of leukocytes. Its frequency in uroepithelial DNA and hair-root DNA varied, and it was undetectable in one sample of buccal-mucosa DNA. These observations have implications with respect to both the pathogenesis of tuberous sclerosis and genetic diagnosis and counseling for this and other genetic disorders in which mosaicism occurs.
Tuberous sclerosis belongs to the family of tumor-suppressor–gene syndromes. Affected patients typically have one inactivating mutation in either the TSC1 or TSC2 gene in all germ-line and somatic cells. Many renal angiomyolipomas in these patients have a loss of heterozygosity (allelic loss) in either the TSC1 or the TSC2 genomic region, consistent with the occurrence of a "second hit."12,13,14 A lower frequency of allelic loss has also been demonstrated in other hamartomas in these patients, but this difference may reflect the mechanism by which alleles are inactivated.15 In our patient, all four types of tissue that were examined showed mosaicism, suggesting that mosaicism is present in all organs and tissues. Since this patient had cortical tubers and facial angiofibromas typical of tuberous sclerosis, our findings suggest that a hamartomatous lesion can occur if even a small proportion of cells in an organ have a TSC1 mutation. This observation is consistent with the two-hit hypothesis for the development of hamartoma in patients with tuberous sclerosis.
The clinical presentation of tuberous sclerosis is highly variable, but the disorder is usually considered fully penetrant.3,4,16 Segmental expression of the phenotype is known to occur, with manifestations limited to the brain, kidney, or one side of the face, but the TSC1 and TSC2 mutations in such cases have not been characterized.17,18,19 There have also been several reports of patients with mosaicism for TSC2 mutations,20,21,22 but these patients have had milder clinical findings than the patient described here or have had tuberous sclerosis and polycystic kidney disease.21
Our case report points out a serious potential problem in the genetic diagnosis of tuberous sclerosis and other disorders caused by spontaneous mutations. Since two thirds of the cases of tuberous sclerosis are sporadic, a substantial fraction could involve mosaicism. Our results demonstrate that it is possible for a TSC1 mutation to be present at low frequency in leukocyte DNA, despite the presence of severe disease. Our method of screening for mutations was a relatively sensitive method for the detection of mutations in the presence of mosaicism. Because of the low frequency of the mutant allele, our patient's mutation probably would not have been detected by direct sequencing of the exons of TSC1 — a method that is often considered the gold standard for the detection of mutations.23 Moreover, the extent of somatic mosaicism in different tissues may vary to such a degree that no mutant alleles are present in leukocytes, whereas they are present in moderate amounts in other organ systems affected by tuberous sclerosis, particularly the brain. The failure to detect mosaicism could help explain why relatively few TSC1 and TSC2 mutations have been identified to date in patients with tuberous sclerosis,9,11,24,25,26 despite moderately intensive screening efforts.
Supported by grants from the National Institutes of Health and the National Tuberous Sclerosis Association.
We are indebted to Fiona Hall for technical assistance.
Source Information
From the Division of Experimental Medicine and Medical Oncology, Brigham and Women's Hospital, Boston (J.K., D.J.K.); the Institute of Human Genetics, Polish Academy of Science, Poznan, Poland (J.K., D.N., R.S.); and the Department of Developmental Neurology, University of Medical Sciences, Poznan, Poland (J.W.-S.).
Address reprint requests to Dr. Kwiatkowski at the Division of Experimental Medicine, 221 Longwood Ave., Boston, MA 02115, or at kwiatkowski{at}calvin.harvard.edu.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
