The Journal of Dermatology Vol. 19: 914-919, 1992
WS VI-9
Tuberous Sclerosis Complex: Genetic Aspects Hope Northrup Abstract Much has been learned about tuberous sclerosis complex (TSC) since it was described at the end of the nineteenth century. TSC was recognized to be a genetic disease with autosomal dominant inheritance in the early twentieth century. The prevalence in the general population is at least 1 in 10,000 with two-thirds of cases occurring sporadically and one-third of cases being familial. The disease exhibits variable expression which may cause mildly affected individuals to be undiagnosed. Because the aberrant or missing proteins which result in TSC have eluded investigators, a positional cloning approach has been pursued to find the mutated genes. Genetic. linkages have been reported to chromosomes 9, 11, and 12. There is definite evidence for a TSC-causing locus on chromosome 9 which is thought to account for between one-third and one-half of all familial cases. Investigators have narrowed the location on chromosome 9 to approximately two megabases of physical distance. There is some evidence for a locus on chromosome 11 which probably accounts only for a small percentage of familial cases. The locus proposed on chromosome 12 was reported by a single group and has not been confirmed by other research groups. Evidence for genetic heterogeneity is abundant. There is definitely a TSC-causing locus on chromosome 9q (TSC-l) and there is at least one additional locus, maybe more than one. As the molecular basis of TSC unfolds, new insight will be gained about the protean nature of the disorder and the genetic heterogeneity.
Key words: tuberous sclerosis complex (TSC); autosomal dominant inheritance; variable expression; genetic heterogeneity
Historical Aspects Tuberous sclerosis complex (TSC) was first described by Friedrich Daniel von Recklinghausen in 1862 in a presentation to the Obstetrical Society of Berlin (1). The autopsy findings he reported were on a newborn and included cardiac tumors (which were described in some detail) and "scleroses" of the brain. The first detailed manuscript concerning TSC was authored by Desire-Magloire Bourneville in 1880 (2). Bourneville, after whom the term "Bourneville's Disease" refers, described the pathological findings on brains from three individuals with neurological problems. One of the individuals was a female with seizures, mental retardation, and a facial rash. In describing her brain on autopsy, Bourneville coined the term "tuberous sclerosis of the cerebral convolutions" from which the term tuberous sclerosis was derived. The first descriptions of the facial lesions were by Balzer and Assistant Professor, Department of Pediatrics, Division of Medical Genetics, University of Texas Medical SchoolHouston, 6431 Fannin, Houston, Texas 77030, U.S.A
Menetrier in 1885 (3) and by Pringle in 1890 (4). The facial lesions, erroneously called "adenoma sebaceum" initially, were noted to be hereditary prior to the recognition that TSC was of a hereditary nature (5). The first to note that TSC was an inherited condition was Kirpicznik who reported a family in which three generations were affected with TSC and who also described the condition in identical and fraternal twins (6). Berg, in 1913, also recognized that TSC was hereditary (7). Gunther and Penrose first brought attention to the dominant inheritance pattern and were also the first to suggest a high mutation rate as causal in TSC (8). In 1951, Dickerson published an article describing three families with multiple members affected with TSC and reviewed the literature concerning all that was known about familial cases (9). Gomez, in 1967, authored the first large study that was not from an institution for the mentally retarded or the epileptic (71 patients were included) (10). Gomez was one of the first to note that there were many affected individuals who were not mentally retarded and that TSC was common in the general population (10, 11).
Tuberous Sclerosis Complex (TSC)
Incidence and Prevalence The earliest studies determining the incidence of TSC were conducted on the populations of institutions for the mentally retarded or epileptic (8, 12-14). These early studies found an incidence of TSC of 0.3-0.7% among the residents of these institutions. By using the frequency of mental retardation in the general population, Gunther and Penrose used the data from such a study to determine an incidence in the general population of 1 in 30,000 (8). Other subsequent studies on such populations found similar incidence and extrapolation to the general population gave results ranging from an incidence as low as 1 in 150,000 to as high as 1 in 20,000 (15-18). Surveys of the general population have also been performed to determine the prevalence of TSC (19-23). The earlier studies (19-21) found TSC to be a rare disease with highest prevalence of 1 in 100,000. The three most recent studies (11, 22, 23) were all done in the 1980s and found TSC to be much more common (prevalence ranging from 1 in 9,407 to I in 34,200). The recent studies were aided by modern imaging techniques. It has also been recognized that TSC can be a very benign disorder that is easily missed in casual examination of a mildly affected individual. Thus, the more recent estimates are presumed to be more accurate. Sampson et al. surveyed completely for TSC in a defined population (the west of Scotland; population 2,763,000) and reported on all genetic aspects of TSC in that population (23). The overall minimum prevalence of TSC was 1 in 27,000 but the minimum prevalence for children under ten years was 1 in 12,000. The ratio of isolated to familial cases was 60:40. In reviewing the previous surveys, Sampson et al. found that the reported proportion of isolated cases ranged from 50-86% of cases. Best current estimates are that isolated TSC accounts for 60-70% of cases (24). Variable Expression Expression, when referring to the clinical findings produced by a mutant gene, is the term used to describe the different clinical findings that may result from having the mutant gene. The phenotype produced may be different based on a number of factors (modifier genes; environmental factors; stochastic factors). Variable expression is known to be a characteristic of many autosomal dominant diseases. For example, neurofibromatosis type 1 (NF-l) is known to show variable expression in
915
different individuals affected with the disease. Some individuals with NF-l have hundreds of neurofibromas while others have no neurofibromas but have other clinical findings which are known to be diagnostic of the disease. TSC has been recognized for many years to show variable expression (10, 25, 26). The variability is seen not only between isolated cases or between individuals from different families but also between affected individuals who are members of the same family. The unpredictable nature of TSC causes clinical difficulties for genetic counseling. For example, if a mildly affected individual whose brother has severe findings bears an affected child, will the child be like the parent or the uncle? There is currently no way to predict the phenotype of an individual who receives a mutant TSC gene. There are numerous examples of individuals who have few or no external stigmata of TSC but have diagnostic findings on evaluation of internal organs (27-33). In most of these cases, when 'parents of affected children were evaluated extensively with the contemporary technology, one of them was diagnosed with TSC. Counseling for transmission risks for these parents changed dramatically as a result of careful evaluation. At least seven sets of parents are reported who have more than one child affected with TSC in which neither parent had any signs of TSC after careful examination and some evaluation ofinternal organs (23, 27, 33-36). There are two other reports, one involving a father and one a mother, in which two affected children were born to the same individual and the individual had no signs of TSC despite extensive careful evaluation (37, 38). A parent with no signs of TSC but several affected children can be explained at least four ways: 1) two random mutations occurring within the same family, 2) lack of detectable expression of the TSC gene in ODe of the parents, 3) gonadal mosaicism, 4) non-paternity: Two random mutations leading to TSC oCcUning>within the same immediate family are highly ttnlilQelybased on purported statistical data on ptevalente, (probability 2.25 x 10-8) . Lack of deteetablee¥S$wnofthe TSCgene in one of the pareDtS\Ma ,.ore Ilkelyexplanation for these cases. An indiWd.....·withaffected offspring by different spousesaff~d with NF-l who has only minimal findip.gshasbeenreported (39). Gonadal/somatic mosaiciSllinu been documented as the cause of recurnmril.,f,someautosomal dominant as well as X-littketf;,rece&sive diseases (40-42) and may also be the explanation, for these "unaffected" parents with
916
Northrup
more than one affected child. As an example, Cohn and colleagues (42) showed that an unaffected father who had two children with lethal osteogenesis imperfecta by different mothers. was mosaic for the mutation causing the disease in lymphocytes and sperm but not in fibroblast cells. The explanation for these parents with multiple children affected with TSC will remain unknown until the mutations which result in TSC are discovered. Eight sets of identical twins affected with TSC are reported in the literature (33, 43, 44). In six of the cases, both twins are affected, one having more severe symptoms than the other (43). Two other twin sets (one case of female twins, the other male twins) are reported with one twin affected with TSC and the other twin having no detectable signs of the disorder despite careful thorough evaluation (33, 44). There are two other reports in the literature in which affected individuals are related through seemingly unaffected individuals (30, 45). These four individuals (the two unaffected twins and the other two individuals mentioned above) represent examples where we are unable to detect the disease but the individuals probably have the mutated gene. The twin cases could, alternatively, be the result of the TSC mutation occurring after the separation of the twins implying somatic mosaicism. All of these cases (individuals with subtle findings, "unaffected" parents with several affected offspring, "identical" twins in whom one has TSC and one does not) emphasize the need to determine the molecular basis of TSC and caution clinicians that TSC is sometimes difficult to detect with current techniques. The diagnosis of TSC remains a possibility in an at-risk individual even after thorough negative evaluation, including testing of internal organs. These cases also raise the issue that TSC may be a much more common disorder than thought because mild cases are not detected. The Molecular Basis of TSC
Approximately one-third of cases of TSC are familial and autosomal dominant inheritance in the familial cases is well-cocumented (24). As many as five generations in a family have been reported affected with TSC (10). For many years, investigators have tried to determine the biochemical defect that causes an individual to have TSC. Unfortunately, these studies have not led to finding the missing or malfunctioning protein resultingly in TSC. Recently, positional cloning (determining the sequence of a disease-causing gene by its location in the genome and then determining the missing or aberrant
function) has characterized several disease-causing genes for which the gene products and their functions were unknown (46-50). In the absence of a chromosomal clue (a translocation or deletion in an affected individual) leading to the location of a disease gene, linkage analysis is the only technique available at the present time to attempt to locate a disease gene for which the function is not known. A linkage study tests polymorphisms from known locations in the genome in families with multiple members affected by a specific disease. When a marker tracks with the disease, the location of the disease gene can then be determined.
linkage to Chromosome 9 In 1987, a British group led by Fryer (51) reported linkage between the ABO blood group and a TSCcausing locus. Fryer et al. studied 19 families with multiple members affected with TSC by testing 26 polymorphic protein markers. Eight of the 19 families were informative for ABO blood grouping. A LOD (logarithm of the odds) score of 3.85 at zero recombination for linkage between the ABO blood group locus and a TSC-causing locus was found (odds close to 10,000 to 1 that the ABO blood group locus was very near a TSC-causing gene). The ABO blood group locus was known to map to chromosome 9q34. The linkage was confirmed by Connor et al. (52) with a- DNA polymorphism from the Abelson oncogene which also maps to 9q34. Northrup et al., however, reported a family in which the TSC-causing gene did not appear to be located at 9q34 (53). Several other investigators could not confirm the linkage to 9q34 (54, 55). These conflicting findings, both of which were proven to be correct, resulted in worldwide collaboration that is still continuing today in the search for the molecular basis of TSC. linkage to Chromosome 11 In 1990, another genetic linkage was reported for TSC (55). Smith and colleagues studied 15 multigenerational families affected with TSC. They found a LOD score of 3.26 with an anonymous DNA probe, MCTI28.I (DllSI44), from chromosome llq22-q23. A probe from the tyrosinase gene, TYR, which maps to llq14-q22 also gave a suggestive LOD score of 2.88. The positive result with tyrosinase was especially intriguing because of the hypomelanotic macules seen in 97-98% of individuals affected with TSC. There was a chromosomal clue which led to the search on chromosome 11 (56). An infant who died soon after birth was found
Tuberous Sclerosis Complex (TSC) on autopsy to have TSC (cardiac rhabdomyoma of the right ventricle and multifocal tubers in the cerebral cortex of the brain with subependymal giant cell tumors) in addition to other congenital anomalies (cleft palate and dysplastic right ear). The infant's mother had a balanced translocation between chromosomes 11. and 22 and the infant received an unbalanced product resulting in triplication of a segment of chromosome 11. The karyotype of the infant determined following a third trimester amniocentesis for marked polyhydramnios was 47.xx,+der(22)t(II ;22)(q23.3;qII.2)mat. Other investigators began to test TSC families for linkage of a TSC-causing locus in this region of chromosome llq. Genetic Heterogeneity The possibility of genetic heterogeneity (more than one locus in the genome resulting in the disease) was addressed because ofthe two findings described above: families affected with TSC in which the TSC-causing locus did not appear to map to chromosome 9q and the linkage in some families to a locus on chromosome llq. Sampson et al. found positive evidence for genetic heterogeneity when testing eight families using nine polymorphic markers from 9q and three polymorphic markers from llq (57). In the eight families, there was definite evidence for a locus on chromosome 9 and for genetic heterogeneity but no significant evidence for a TSC-causing locus on chromosome 11. A Dutch group, studying a small set of families (nine), used a novel analysis approach (an imaginary chromosome) to test for the possibility of loci on chromosomes 9 and 11 (58). They found evidence for TSC-causing loci on chromosomes 9 and 11 which again supported genetic heterogeneity in TSC. A large International Collaborative Study was undertaken to test the question of heterogeneity because individual investigators were limited by the number of families available to them for study (59). The International Collaborative Study included 111 multigenerational families affected with TSC. Eight loci from chromosome 9 and eight loci from chromosome 11 were included in the analysis. The results showed that heterogeneity clearly exists for TSC and that one locus resides between D9S10 and ORM on chromosome 9q34. Approximately 40% of the families studied showed evidence for a TSCcausing locus on chromosome 9. There was evidence (although less firm) for a TSC-causing locus on chromosome 11; however, only a limited num-
917
ber of the families had been tested with more than one marker on chromosome 11. Two subsequent studies, one on 22 families and the other on 14 families, also supported genetic heterogeneity for TSC and a TSC-causing locus on chromosome 9q (60,61). In these studies, 33% and 46%, respectively, of families exhibited evidence for linkage to chromosome 9q. At the International TSC Workshop held in Cambridge, England on March 22, 1992, it was determined that the TSC-causing locus (TSC-l) on chromosome 9q is between the anonymous DNA markers D9S64 and D9SIO/D9S66. According to physical data presented, the actual distance in which the TSC-I gene is confined is less than two megabases. Hopefully, the TSC-l gene will be found in the near future.
Clinical Ymdings The study by Northrup et al. (61) compared clinical findings of affecteds in the families with evidence for linkage to chromosome 9q (posterior probability >80% by multipoint analysis of markers from 9q) to findings in the families that did not appear to be linked to chromosome 9q (posterior probability
WS VI-9
Tuberous Sclerosis Complex: Genetic Aspects Hope Northrup Abstract Much has been learned about tuberous sclerosis complex (TSC) since it was described at the end of the nineteenth century. TSC was recognized to be a genetic disease with autosomal dominant inheritance in the early twentieth century. The prevalence in the general population is at least 1 in 10,000 with two-thirds of cases occurring sporadically and one-third of cases being familial. The disease exhibits variable expression which may cause mildly affected individuals to be undiagnosed. Because the aberrant or missing proteins which result in TSC have eluded investigators, a positional cloning approach has been pursued to find the mutated genes. Genetic. linkages have been reported to chromosomes 9, 11, and 12. There is definite evidence for a TSC-causing locus on chromosome 9 which is thought to account for between one-third and one-half of all familial cases. Investigators have narrowed the location on chromosome 9 to approximately two megabases of physical distance. There is some evidence for a locus on chromosome 11 which probably accounts only for a small percentage of familial cases. The locus proposed on chromosome 12 was reported by a single group and has not been confirmed by other research groups. Evidence for genetic heterogeneity is abundant. There is definitely a TSC-causing locus on chromosome 9q (TSC-l) and there is at least one additional locus, maybe more than one. As the molecular basis of TSC unfolds, new insight will be gained about the protean nature of the disorder and the genetic heterogeneity.
Key words: tuberous sclerosis complex (TSC); autosomal dominant inheritance; variable expression; genetic heterogeneity
Historical Aspects Tuberous sclerosis complex (TSC) was first described by Friedrich Daniel von Recklinghausen in 1862 in a presentation to the Obstetrical Society of Berlin (1). The autopsy findings he reported were on a newborn and included cardiac tumors (which were described in some detail) and "scleroses" of the brain. The first detailed manuscript concerning TSC was authored by Desire-Magloire Bourneville in 1880 (2). Bourneville, after whom the term "Bourneville's Disease" refers, described the pathological findings on brains from three individuals with neurological problems. One of the individuals was a female with seizures, mental retardation, and a facial rash. In describing her brain on autopsy, Bourneville coined the term "tuberous sclerosis of the cerebral convolutions" from which the term tuberous sclerosis was derived. The first descriptions of the facial lesions were by Balzer and Assistant Professor, Department of Pediatrics, Division of Medical Genetics, University of Texas Medical SchoolHouston, 6431 Fannin, Houston, Texas 77030, U.S.A
Menetrier in 1885 (3) and by Pringle in 1890 (4). The facial lesions, erroneously called "adenoma sebaceum" initially, were noted to be hereditary prior to the recognition that TSC was of a hereditary nature (5). The first to note that TSC was an inherited condition was Kirpicznik who reported a family in which three generations were affected with TSC and who also described the condition in identical and fraternal twins (6). Berg, in 1913, also recognized that TSC was hereditary (7). Gunther and Penrose first brought attention to the dominant inheritance pattern and were also the first to suggest a high mutation rate as causal in TSC (8). In 1951, Dickerson published an article describing three families with multiple members affected with TSC and reviewed the literature concerning all that was known about familial cases (9). Gomez, in 1967, authored the first large study that was not from an institution for the mentally retarded or the epileptic (71 patients were included) (10). Gomez was one of the first to note that there were many affected individuals who were not mentally retarded and that TSC was common in the general population (10, 11).
Tuberous Sclerosis Complex (TSC)
Incidence and Prevalence The earliest studies determining the incidence of TSC were conducted on the populations of institutions for the mentally retarded or epileptic (8, 12-14). These early studies found an incidence of TSC of 0.3-0.7% among the residents of these institutions. By using the frequency of mental retardation in the general population, Gunther and Penrose used the data from such a study to determine an incidence in the general population of 1 in 30,000 (8). Other subsequent studies on such populations found similar incidence and extrapolation to the general population gave results ranging from an incidence as low as 1 in 150,000 to as high as 1 in 20,000 (15-18). Surveys of the general population have also been performed to determine the prevalence of TSC (19-23). The earlier studies (19-21) found TSC to be a rare disease with highest prevalence of 1 in 100,000. The three most recent studies (11, 22, 23) were all done in the 1980s and found TSC to be much more common (prevalence ranging from 1 in 9,407 to I in 34,200). The recent studies were aided by modern imaging techniques. It has also been recognized that TSC can be a very benign disorder that is easily missed in casual examination of a mildly affected individual. Thus, the more recent estimates are presumed to be more accurate. Sampson et al. surveyed completely for TSC in a defined population (the west of Scotland; population 2,763,000) and reported on all genetic aspects of TSC in that population (23). The overall minimum prevalence of TSC was 1 in 27,000 but the minimum prevalence for children under ten years was 1 in 12,000. The ratio of isolated to familial cases was 60:40. In reviewing the previous surveys, Sampson et al. found that the reported proportion of isolated cases ranged from 50-86% of cases. Best current estimates are that isolated TSC accounts for 60-70% of cases (24). Variable Expression Expression, when referring to the clinical findings produced by a mutant gene, is the term used to describe the different clinical findings that may result from having the mutant gene. The phenotype produced may be different based on a number of factors (modifier genes; environmental factors; stochastic factors). Variable expression is known to be a characteristic of many autosomal dominant diseases. For example, neurofibromatosis type 1 (NF-l) is known to show variable expression in
915
different individuals affected with the disease. Some individuals with NF-l have hundreds of neurofibromas while others have no neurofibromas but have other clinical findings which are known to be diagnostic of the disease. TSC has been recognized for many years to show variable expression (10, 25, 26). The variability is seen not only between isolated cases or between individuals from different families but also between affected individuals who are members of the same family. The unpredictable nature of TSC causes clinical difficulties for genetic counseling. For example, if a mildly affected individual whose brother has severe findings bears an affected child, will the child be like the parent or the uncle? There is currently no way to predict the phenotype of an individual who receives a mutant TSC gene. There are numerous examples of individuals who have few or no external stigmata of TSC but have diagnostic findings on evaluation of internal organs (27-33). In most of these cases, when 'parents of affected children were evaluated extensively with the contemporary technology, one of them was diagnosed with TSC. Counseling for transmission risks for these parents changed dramatically as a result of careful evaluation. At least seven sets of parents are reported who have more than one child affected with TSC in which neither parent had any signs of TSC after careful examination and some evaluation ofinternal organs (23, 27, 33-36). There are two other reports, one involving a father and one a mother, in which two affected children were born to the same individual and the individual had no signs of TSC despite extensive careful evaluation (37, 38). A parent with no signs of TSC but several affected children can be explained at least four ways: 1) two random mutations occurring within the same family, 2) lack of detectable expression of the TSC gene in ODe of the parents, 3) gonadal mosaicism, 4) non-paternity: Two random mutations leading to TSC oCcUning>within the same immediate family are highly ttnlilQelybased on purported statistical data on ptevalente, (probability 2.25 x 10-8) . Lack of deteetablee¥S$wnofthe TSCgene in one of the pareDtS\Ma ,.ore Ilkelyexplanation for these cases. An indiWd.....·withaffected offspring by different spousesaff~d with NF-l who has only minimal findip.gshasbeenreported (39). Gonadal/somatic mosaiciSllinu been documented as the cause of recurnmril.,f,someautosomal dominant as well as X-littketf;,rece&sive diseases (40-42) and may also be the explanation, for these "unaffected" parents with
916
Northrup
more than one affected child. As an example, Cohn and colleagues (42) showed that an unaffected father who had two children with lethal osteogenesis imperfecta by different mothers. was mosaic for the mutation causing the disease in lymphocytes and sperm but not in fibroblast cells. The explanation for these parents with multiple children affected with TSC will remain unknown until the mutations which result in TSC are discovered. Eight sets of identical twins affected with TSC are reported in the literature (33, 43, 44). In six of the cases, both twins are affected, one having more severe symptoms than the other (43). Two other twin sets (one case of female twins, the other male twins) are reported with one twin affected with TSC and the other twin having no detectable signs of the disorder despite careful thorough evaluation (33, 44). There are two other reports in the literature in which affected individuals are related through seemingly unaffected individuals (30, 45). These four individuals (the two unaffected twins and the other two individuals mentioned above) represent examples where we are unable to detect the disease but the individuals probably have the mutated gene. The twin cases could, alternatively, be the result of the TSC mutation occurring after the separation of the twins implying somatic mosaicism. All of these cases (individuals with subtle findings, "unaffected" parents with several affected offspring, "identical" twins in whom one has TSC and one does not) emphasize the need to determine the molecular basis of TSC and caution clinicians that TSC is sometimes difficult to detect with current techniques. The diagnosis of TSC remains a possibility in an at-risk individual even after thorough negative evaluation, including testing of internal organs. These cases also raise the issue that TSC may be a much more common disorder than thought because mild cases are not detected. The Molecular Basis of TSC
Approximately one-third of cases of TSC are familial and autosomal dominant inheritance in the familial cases is well-cocumented (24). As many as five generations in a family have been reported affected with TSC (10). For many years, investigators have tried to determine the biochemical defect that causes an individual to have TSC. Unfortunately, these studies have not led to finding the missing or malfunctioning protein resultingly in TSC. Recently, positional cloning (determining the sequence of a disease-causing gene by its location in the genome and then determining the missing or aberrant
function) has characterized several disease-causing genes for which the gene products and their functions were unknown (46-50). In the absence of a chromosomal clue (a translocation or deletion in an affected individual) leading to the location of a disease gene, linkage analysis is the only technique available at the present time to attempt to locate a disease gene for which the function is not known. A linkage study tests polymorphisms from known locations in the genome in families with multiple members affected by a specific disease. When a marker tracks with the disease, the location of the disease gene can then be determined.
linkage to Chromosome 9 In 1987, a British group led by Fryer (51) reported linkage between the ABO blood group and a TSCcausing locus. Fryer et al. studied 19 families with multiple members affected with TSC by testing 26 polymorphic protein markers. Eight of the 19 families were informative for ABO blood grouping. A LOD (logarithm of the odds) score of 3.85 at zero recombination for linkage between the ABO blood group locus and a TSC-causing locus was found (odds close to 10,000 to 1 that the ABO blood group locus was very near a TSC-causing gene). The ABO blood group locus was known to map to chromosome 9q34. The linkage was confirmed by Connor et al. (52) with a- DNA polymorphism from the Abelson oncogene which also maps to 9q34. Northrup et al., however, reported a family in which the TSC-causing gene did not appear to be located at 9q34 (53). Several other investigators could not confirm the linkage to 9q34 (54, 55). These conflicting findings, both of which were proven to be correct, resulted in worldwide collaboration that is still continuing today in the search for the molecular basis of TSC. linkage to Chromosome 11 In 1990, another genetic linkage was reported for TSC (55). Smith and colleagues studied 15 multigenerational families affected with TSC. They found a LOD score of 3.26 with an anonymous DNA probe, MCTI28.I (DllSI44), from chromosome llq22-q23. A probe from the tyrosinase gene, TYR, which maps to llq14-q22 also gave a suggestive LOD score of 2.88. The positive result with tyrosinase was especially intriguing because of the hypomelanotic macules seen in 97-98% of individuals affected with TSC. There was a chromosomal clue which led to the search on chromosome 11 (56). An infant who died soon after birth was found
Tuberous Sclerosis Complex (TSC) on autopsy to have TSC (cardiac rhabdomyoma of the right ventricle and multifocal tubers in the cerebral cortex of the brain with subependymal giant cell tumors) in addition to other congenital anomalies (cleft palate and dysplastic right ear). The infant's mother had a balanced translocation between chromosomes 11. and 22 and the infant received an unbalanced product resulting in triplication of a segment of chromosome 11. The karyotype of the infant determined following a third trimester amniocentesis for marked polyhydramnios was 47.xx,+der(22)t(II ;22)(q23.3;qII.2)mat. Other investigators began to test TSC families for linkage of a TSC-causing locus in this region of chromosome llq. Genetic Heterogeneity The possibility of genetic heterogeneity (more than one locus in the genome resulting in the disease) was addressed because ofthe two findings described above: families affected with TSC in which the TSC-causing locus did not appear to map to chromosome 9q and the linkage in some families to a locus on chromosome llq. Sampson et al. found positive evidence for genetic heterogeneity when testing eight families using nine polymorphic markers from 9q and three polymorphic markers from llq (57). In the eight families, there was definite evidence for a locus on chromosome 9 and for genetic heterogeneity but no significant evidence for a TSC-causing locus on chromosome 11. A Dutch group, studying a small set of families (nine), used a novel analysis approach (an imaginary chromosome) to test for the possibility of loci on chromosomes 9 and 11 (58). They found evidence for TSC-causing loci on chromosomes 9 and 11 which again supported genetic heterogeneity in TSC. A large International Collaborative Study was undertaken to test the question of heterogeneity because individual investigators were limited by the number of families available to them for study (59). The International Collaborative Study included 111 multigenerational families affected with TSC. Eight loci from chromosome 9 and eight loci from chromosome 11 were included in the analysis. The results showed that heterogeneity clearly exists for TSC and that one locus resides between D9S10 and ORM on chromosome 9q34. Approximately 40% of the families studied showed evidence for a TSCcausing locus on chromosome 9. There was evidence (although less firm) for a TSC-causing locus on chromosome 11; however, only a limited num-
917
ber of the families had been tested with more than one marker on chromosome 11. Two subsequent studies, one on 22 families and the other on 14 families, also supported genetic heterogeneity for TSC and a TSC-causing locus on chromosome 9q (60,61). In these studies, 33% and 46%, respectively, of families exhibited evidence for linkage to chromosome 9q. At the International TSC Workshop held in Cambridge, England on March 22, 1992, it was determined that the TSC-causing locus (TSC-l) on chromosome 9q is between the anonymous DNA markers D9S64 and D9SIO/D9S66. According to physical data presented, the actual distance in which the TSC-I gene is confined is less than two megabases. Hopefully, the TSC-l gene will be found in the near future.
Clinical Ymdings The study by Northrup et al. (61) compared clinical findings of affecteds in the families with evidence for linkage to chromosome 9q (posterior probability >80% by multipoint analysis of markers from 9q) to findings in the families that did not appear to be linked to chromosome 9q (posterior probability
