CCM2 gene polymorphisms in Italian sporadic patients with cerebral cavernous malformation: a case-control study

Cerebral cavernous malformations (CCMs) are vascular lesions of the CNS characterized by abnormally enlarged capillary cavities that can occur sporadically or as a familial autosomal dominant condition with incomplete penetrance and variable clinical expression attributable to mutations in three different genes: CCM1 (Krit1), CCM2 (MGC4607) and CCM3 (PDCD10). Among our group of CCM Italian patients, we selected a cohort of sporadic cases negative for mutations in CCM genes. In this cohort, five variants in CCM2 gene were detected, which proved to be the known polymorphisms in intronic regions (IVS2-36A>G and IVS8 +119 C>T) and in coding sequence (c.157 G>A in exon 2, c.358 G>A in exon 4 and c.915 G>A in exon 8). Therefore, we undertook a case-control study to investigate the possible association of these polymorphisms with sporadic CCMs. The five polymorphisms were identified in 91 CCM sporadic patients and in 100 healthy controls by direct sequencing methods using lymphocyte DNA. Polymorphisms IVS2-36A>G and c.915 G>A showed statistically significant differences in frequencies between patients and controls [(χ2, 6.583; P<0.037); (χ2, 14.205; P<0.001)]. The prevalence of the wild-type genotype was significantly lower in the CCM group than in the control sample. Patients with the A/G and G/G genotypes (IVS2-36A>G) had a significant increase for CCM risk (OR, 3.08; 95% CI, 1.5-5.9 and OR, 4.3; 95% CI, 1.4-22.6) and the same was observed for the polymorphism c.915 G> A (genotype G/A OR, 6.1; 95% CI, 3.0-12.6 and genotype A/A OR, 2.79). In addition, the polymorphisms c.358 G>A in exon 4 (χ2, 15.977; P<0.04) and c.915 G>A in exon 8 (χ2, 18.109; P<0.02) were significantly associated with different types of symptoms. Haplotype analysis, performed only on polymorphisms c.358 G>A (p.Val120Ile), c.915 G>A (p.Thr305 Thr) and IVS2-36A>G, shows that haplotype GAG (+--) significantly increased among CCM sporadic patients compared to the control group. Significant differences between patients and controls were observed only for IVS2-36A>G and c.915 G>A polymorphisms indicating their possible association with sporadic CCMs and an increased risk of CCM. On the other hand, polymorphisms c.358 G>A and c.915 G>A were associated with a more benign course of the disease. These data were confirmed by the haplotype GAG (+--) frequencies.     

Highly variable penetrance in subjects affected with cavernous cerebral angiomas (CCM) carrying novel CCM1 and CCM2 mutations.

Cavernous vascular malformations may affect brain and out-of-brain tissues. In most cases, cerebral cavernous malformations (CCMs) involve the brain alone, and are rarely associated with skin hemangiomas, spinal cord, retinal, hepatic or vertebral lesions. CCMs can cause seizures, intracranial and spinal haemorrhages, focal neurological deficits, and migraine-like headaches. After collecting CCM families of Italian origin and investigating the genetic basis of the disorder we disclosed two novel molecular variations in the KRIT1 and MGC4607 genes. We found a novel CCM1 gene mutation (Q66X) in a family with apparently asymptomatic old-aged mutation carriers and patients who either had skin angiomas alone or the full association of cerebral, spinal, and skin lesions. In this family we report the highest variability in mutation penetrance so far described, including the presence of CCM in one subject since birth (surgery at 19 months of age), a condition to our knowledge so far unreported. In a CCM2 affected family, we also report a novel causative mutation, (54_55delAC) in exon 2 of the MGC4607 gene, that produces a truncated protein containing only 22 amino acids. These data describe novel CCM mutations associated with a particularly high variability of the penetrance causing, in some cases, reduced expression of clinical symptoms and sporadic cases with apparent negative family history. Hence they emphasize the importance of DNA-based diagnostics and genetic counseling to identify unaffected mutation carriers subjects, even at advanced age.     

Deep intronic KRIT1 mutation in a family with clinically silent multiple cerebral cavernous malformations

Loss‐of‐function mutations in CCM1/KRIT1, CCM2/MGC4607 and CCM3/PDCD10 genes are identified in the vast majority of familial cases with multiple cerebral cavernous malformations (CCMs). However, genomic DNA sequencing combined to large rearrangement screening fails to detect a mutation in 5% of those cases. We report a family in which CCM lesions were discovered fortuitously because of the investigation of a developmental delay in a boy. Three members of the family on three generations had typical multiple CCM lesions and no clinical signs related to CCM. No mutation was detected using genomic DNA sequencing and quantitative multiplex PCR of short fluorescent fragments (QMPSF). cDNA sequencing showed a 99‐nucleotide insertion between exons 5 and 6 of CCM1, resulting from a mutation located deep into intron 5 (c.262+132_262+133del) that activates a cryptic splice site. This pseudoexon leads to a premature stop codon. These data highly suggest that deep intronic mutations explain part of the incomplete mutation detection rate in CCM patients and underline the importance of analyzing the cDNA to provide comprehensive CCM diagnostic tests. This kind of mutation may be responsible for apparent sporadic presentations due to a reduced penetrance.     

C329X in KRIT1 is a founder mutation among CCM patients in Sardinia

Cerebral cavernous malformations (CCMs) are CNS vascular anomalies associated with seizures, headaches and hemorrhagic strokes and represent 10–20% of cerebral lesions. CCM is present in 0.1–0.5 of the population. This disorder most often occurs sporadically but may also be familial. Familial cases are inherited as a dominant trait with incomplete penetrance and are estimated to account for KRIT1 10–40% of the patients. The identification of the genes involved in such disorders allows to characterize carriers of the mutations without clear symptoms. The first gene involved in CCM1 is KRIT1. In addition to two other genes have been described: MGC4607 (CCM2) and PDCD10 (CCM3). We selected 13 patients belonging to seven Sardinian families on the basis of clinical symptoms and Magnetic Resonance results. In MGC4607 gene an undescribed exon five deletion likely producing a truncated protein was identified in one family. In two patients with clear phenotype and in three asymptomatic relatives a 4 bp deletion in exon 9 of KRIT1 gene, leading to a premature stop codon, was detected. A unique nonsense mutation (C329X) has been found in seven patients and two asymptomatic subjects belonging to four unrelated families. Haplotype analysis revealed a common origin of this mutation. These data suggest a “founder effect” in Sardinia for the C329X mutation, similar to other mutations described in different populations.     

PINK1, Parkin, and DJ-1 mutations in Italian patients with early-onset parkinsonism

Recessively inherited early-onset parkinsonism (EOP) has been associated with mutations in the Parkin, DJ-1, and PINK1 genes. We studied the prevalence of mutations in all three genes in 65 Italian patients (mean age of onset: 43.2+/-5.4 years, 62 sporadic, three familial), selected by age at onset equal or younger than 51 years. Clinical features were compatible with idiopathic Parkinson's disease in all cases. To detect small sequence alterations in Parkin, DJ-1, and PINK1, we performed a conventional mutational analysis (SSCP/dHPLC/sequencing) of all coding exons of these genes. To test for the presence of exon rearrangements in PINK1, we established a new quantitative duplex PCR assay. Gene dosage alterations in Parkin and DJ-1 were excluded using previously reported protocols. Five patients (8%; one woman/four men; mean age at onset: 38.2+/-9.7 (range 25-49) years) carried mutations in one of the genes studied: three cases had novel PINK1 mutations, one of which occurred twice (homozygous c.1602_1603insCAA; heterozygous c.1602_1603insCAA; heterozygous c.836G>A), and two patients had known Parkin mutations (heterozygous c.734A>T and c.924C>T; heterozygous c.924C>T). Family history was negative for all mutation carriers, but one with a history of tremor. Additionally, we detected one novel polymorphism (c.344A>T) and four novel PINK1 changes of unknown pathogenic significance (-21G/A; IVS1+97A/G; IVS3+38_40delTTT; c.852C>T), but no exon rearrangements. No mutations were found in the DJ-1 gene. The number of mutation carriers in both the Parkin and the PINK1 gene in our cohort is low but comparable, suggesting that PINK1 has to be considered in EOP.     

CCM1 and CCM2 protein interactions in cell signaling: implications for cerebral cavernous malformations pathogenesis

Cerebral cavernous malformations (CCMs) are sporadically acquired or inherited vascular lesions of the central nervous system consisting of clusters of dilated thin-walled blood vessels that predispose individuals to seizures and stroke. Familial CCM is caused by mutations in KRIT1 (CCM1) or in malcavernin (CCM2), the murine ortholog of which was concurrently characterized as osmosensing scaffold for MEKK3 (OSM). The roles of the CCM proteins in the pathogenesis of the disorder remain largely unknown. Here, we use co-immunoprecipitation, fluorescence resonance energy transfer and subcellular localization strategies to show that the CCM1 gene product, KRIT1, interacts with the CCM2 gene product, malcavernin/OSM. Analogous to the established interactions of CCM1 and beta1 integrin with ICAP1, the CCM1/CCM2 association is dependent upon the phosphotyrosine binding (PTB) domain of CCM2. A familial CCM2 missense mutation abrogates the CCM1/CCM2 interaction, suggesting that loss of this interaction may be critical in CCM pathogenesis. CCM2 and ICAP1 bound to CCM1 via their respective PTB domains differentially influence the subcellular localization of CCM1. Furthermore, we expand upon the established involvement of CCM2 in the p38 mitogen-activated protein kinase signaling module by demonstrating that CCM1 associates with CCM2 and MEKK3 in a ternary complex. These data indicate that the genetic heterogeneity observed in familial CCM may reflect mutation of different molecular members of a coordinated signaling complex.     

颅内海绵状血管瘤 CCM1基因第12外显子及其5′端内含子新突变位点

目的　探讨 CCM1基因突变在中国人颅内海绵状血管瘤 ( intracranial cavernous angiomas,ICCA)发病中所起的作用。方法　收集我院神经外科 2 0 0 2年 6月～ 2 0 0 3年 2月收治并经手术病理证实的2 1例 ICCA患者及 15名正常健康对照者 ,从外周静脉血中提取 DNA,PCR法扩增 CCM1基因第 12外显子及其两侧部分内含子序列 ,应用 DNA直接测序技术对扩增产物进行检测。结果　 5例患者中检测出 3处 CCM1基因突变 ,均为首次发现。其中 ,5例患者中均存在 1172 C→ T的错义突变 ,使编码 KRIT1蛋白391位的氨基酸由丝氨酸变成苯丙氨酸。另有 1例患者存在 116 0 A→ C的错义突变 ,使编码 KRIT1蛋白387位氨基酸的谷氨酰胺变成脯氨酸。另一个突变发生在第 12外显子 5′端内含子区域 ,5例患者中有 4例第 4个碱基 C被 T取代。对照组检测结果无异常。结论　中国 ICCA患者存在 CCM1基因第 12外显子的突变 ,并与 ICCA的发病有关     

汉族家族性中枢神经系统海绵状血管畸形致病基因的一个新突变位点

目的 探查汉族家族性中枢神经系统海绵状血管畸形(CCM)的突变基因。方法 选择经神经科确诊的2个家系(A及B)和8例散发性中枢神经系统海绵状血管畸形病例。所有受检对象临床表现有癫痫、突发头痛;2例伴有皮肤病变。候选基因为已确认的Krit-1,对该基因的16个编码外显子进行PCR扩增,扩增产物进行直接测序。结果 受检的A家系在第14外显子的第1289(从起始密码子计算,或2308即从mRNA的第一个碱基计算)位核苷酸一个拷贝有C→G的取代,出现编码改变(S430X),形成终止密码子(UGA),使翻译过程提前终止,导致的基因异常属无义点突变。同法筛查,A家族中有1名成员突变未获得遗传;散发病例发现一个正常多态性变化,未见到突变。结论 发现第1例汉族人CCM基因的点突变,也是国际上未见报道的新位点。这一点突变将会导致一种截短蛋白,是CCM发生的基本原因。研究结果可用于症状前基因诊断。     

Gap junction beta 1 (<Emphasis Type="Italic">GJB1</Emphasis>) gene mutations in Italian patients with X-linked Charcot-Marie-Tooth disease

X-linked Charcot-Marie-Tooth disease (CMT1X) is a peripheral neuropathy transmitted in a dominant manner and caused by mutations in the Connexin 32 (Cx32) gene (GJB1, gap junction beta 1). Here we report the mutation analysis of the GJB1 gene in 76 subjects with possible CMT1 and absence of 17p11.2 duplication, and in 38 CMT2 patients without mutations in CMT2-associated-genes, selected from a cohort of 684 patients with peripheral sensory-motor neuropathy. The analysis was performed by direct sequencing of the coding sequence and exon/intron boundaries of the GJB1 gene. The mutation screening identified 22 mutations in GJB1, eight of which have not been previously published: six point mutations (c.50C > G, c.107T > A, c.545C > T, c.545C > G, c.548G > C, c.791G > T) and two deletions (c.84delC, c.573_581delCGTCTTCAT). The GJB1 mutation frequency (19.3%) and the clinical heterogeneity of our patients suggest searching for GJB1 mutations in all CMT cases without the 17p11.2 duplication, regardless of the gender of the proband, as well as in CMT2 patients with possible X-linked inheritance.     

Low frequency of PDCD10 mutations in a panel of CCM3 probands: potential for a fourth CCM locus

Cerebral cavernous malformations (CCMs) are vascular abnormalities of the brain that can result in a variety of neurological disabilities, including stroke and seizures. Linkage analyses using autosomal dominant families manifesting CCMs have identified three different causative loci on chromosomes 7q21.2 (CCM1), 7p13 (CCM2), and 3q25.2-q27 (CCM3). Mutations in the gene Krit1 are responsible for CCM1, mutations in the gene MGC4607 are responsible for CCM2, and mutations in the gene PDCD10 were recently reported to be responsible for CCM3. We report here that sequence analysis of PDCD10 in a panel of 29 probands lacking Krit1 and MGC4607 mutations revealed only three mutations. The frequency of identified mutations in the PDCD10 gene was surprisingly low, especially given that this panel was heavily biased towards non-CCM1, non-CCM2 probands. These data are in stark contrast with the linkage data, which suggests that 40% of inherited cases would be due to mutations in this gene. Interestingly, when examining the haplotypes of previously published CCM3 families, we found a distinct recombination event in one of the largest CCM3 families that excludes the PDCD10 gene. Although there are many potential explanations for this observation, when combined with the apparent under-representation of causative CCM mutations in PDCD10, this recombination event in a CCM3-linked family suggests that there may be an additional CCM gene in the same chromosomal region.     

Genetic analysis in patients with left ventricular noncompaction and evidence for genetic heterogeneity

Left ventricular noncompaction (LVNC) is a cardiomyopathy characterized by numerous excessively trabeculations and deep intertrabecular recesses. This study was performed to investigate Japanese LVNC patients for disease-causing mutations in a series of selected candidate genes. DNA was isolated from the peripheral blood of 79 cases including 20 familial cases and 59 sporadic cases. DNA samples were screened for mutations in the genes encoding G4.5 (TAZ), alpha-dystrobrevin (DTNA), alpha1-syntrophin (SNTA1), FK506 Binding protein 1A (FKBP1A or FKPB12: FKBP1A), and LIM Domain Binding protein 3 (Cypher/ZASP: LDB3), using single-strand conformational polymorphism analysis and DNA sequencing. DNA variants were identified in 6 of the 79 cases, including four familial cases and two sporadic cases. A splice acceptor mutation of intron 8 in TAZ (IVS8-1G>C) was identified in one family with isolated LVNC, resulting in deletion of exon 9 from mRNA. In a sporadic case of isolated LVNC and Barth syndrome (BTHS), a 158insC in exon 2 of TAZ resulting in a frame-shift mutation was identified. A 1876G>A substitution changing an aspartic acid to asparagine (D626N) was identified in LDB3 in four members of two families with LVNC. A 163G>A polymorphism was identified in LDB3, which changed a valine to isoleucine (V55I) in one patient with isolated LVNC. In addition, in a family with nonisolated LVNC, a 362C>T mutation was identified in DTNA. LVNC, like other forms of inherited cardiomyopathy, is a genetically heterogeneous disease, associated with variable clinical symptoms and can be inherited as an autosomal or X-linked recessive disorder.     

PLA2G6 mutations in PARK14-linked young-onset parkinsonism and sporadic Parkinson's disease

Mutations of PLA2G6 gene have been lately proposed to be the causative gene for PARK14 in patients with autosomal recessive young-onset parkinsonism (YOPD). The role of PLA2G6 mutations as a risk factor for Parkinson's disease is not clear. To study the PLA2G6 mutations in PARK14-linked patients and its association with the onset of sporadic Parkinson's disease (sPD), sequencing and gene dosage analyses were carried out in 25 patients (onset age ≦30 years) then the identified variants were assessed in 956 sporadic PD (sPD) patients and 802 age-matched healthy controls. Four genetic variants were identified; one patient had homozygous c.991G > T (p.Asp331Tyr) mutation, two had compound heterozygous c.991G > T/c.1077G > A (p.Met358IlefsX) mutation, one had single c.1976A > G (p.Asn659Ser) mutation, and one patient had an exon 1 hetero-deletion. The c.1077G > A mutation resulted in a 4-bp deletion in leukocyte mRNA by activating a cryptic splice site in exon 7. Only p.Asp331Tyr was identified in four sPD patients and four controls. The onset age for PLA2G6 mutation carriers was younger than that for sPD (29.86 ± 8.59 vs. 56.84 ± 11.33 years, P = 0.0002). The analysis of previously reported PARK14 patients revealed that those who carried a truncated mutation tended to have a complicated phenotype and atrophies of cortex and cerebellum. In conclusion, PLA2G6 mutation was the second common genetic cause after PRKN mutation in our YOPD patients and might be a risk factor for early-onset PD in Han Chinese. Additionally, mutation data should be interpreted carefully because even a synonymous mutation could cause abnormal mRNA splicing.     

The G6055A (G2019S) mutation in LRRK2 is frequent in both early and late onset Parkinson's disease and originates from a common ancestor

Background: Mutations in the gene Leucine-Rich Repeat Kinase 2 (LRRK2) were recently identified as the cause of PARK8 linked autosomal dominant Parkinson''s disease. Objective: To study recurrent LRRK2 mutations in a large sample of patients from Italy, including early (<50 years) and late onset familial and sporadic Parkinson''s disease. Results: Among 629 probands, 13 (2.1%) were heterozygous carriers of the G2019S mutation. The mutation frequency was higher among familial (5.1%, 9/177) than among sporadic probands (0.9%, 4/452) (p<0.002), and highest among probands with one affected parent (8.7%, 6/69) (p<0.001). There was no difference in the frequency of the G2019S mutation in probands with early v late onset disease. Among 600 probands, one heterozygous R1441C but no R1441G or Y1699C mutations were detected. None of the four mutations was found in Italian controls. Haplotype analysis in families from five countries suggested that the G2019S mutation originated from a single ancient founder. The G2019S mutation was associated with the classical Parkinson''s disease phenotype and a broad range of onset age (34 to 73 years). Conclusions: G2019S is the most common genetic determinant of Parkinson''s disease identified so far. It is especially frequent among cases with familial Parkinson''s disease of both early and late onset, but less common among sporadic cases. These findings have important implications for diagnosis and genetic counselling in Parkinson''s disease.     

Analysis of both TSC1 and TSC2 for germline mutations in 126 unrelated patients with tuberous sclerosis.

Tuberous sclerosis complex (TSC) is an autosomal dominant disorder characterized by the development of multiple hamartomas involving many organs. About two-thirds of the cases are sporadic and appear to represent new mutations. With the cloning of two causative genes, TSC1 and TSC2 it is now possible to analyze both genes in TSC patients and identify germline mutations. Here we report the mutational analysis of the entire coding region of both TSC1 and TSC2 genes in 126 unrelated TSC patients, including 40 familial and 86 sporadic cases, by single-stranded conformational polymorphism (SSCP) analysis followed by direct sequencing. Mutations were identified in a total of 74 (59%) cases, including 16 TSC1 mutations (5 sporadic and 11 familial cases) and 58 TSC2 mutations (42 sporadic and 16 familial cases). Overall, significantly more TSC2 mutations were found in our population, with a relatively equal distribution of mutations between TSC1 and TSC2 among the familial cases, but a marked underrepresentation of TSC1 mutations among the sporadic cases (P = 0.0035, Fisher's exact test). All TSC1 mutations were predicted to be protein truncating. However, in TSC2 13 missense mutations were found, five clustering in the GAP-related domain and three others occurring in exon 16. Upon comparison of clinical manifestations, including the incidence of intellectual disability, we could not find any observable differences between TSC1 and TSC2 patients. Our data help define the distribution and spectrum of mutations associated with the TSC loci and will be useful for both understanding the function of these genes as well as genetic counseling in patients with the disease.     

Identification of eight novel glucokinase mutations in Italian children with maturity-onset diabetes of the young

Maturity-onset diabetes of the young (MODY) is a clinically heterogeneous group of disorders characterized by early onset non-insulin-dependent diabetes mellitus, autosomal dominant inheritance, and primary defect in the function of the beta cells of the pancreas. Mutations in the glucokinase (GCK) gene account for 8%-56% of MODY, with the highest prevalences being found in the southern Europe. While screening for GCK mutations in 28 MODY families of Italian origin, we identified 17 different mutations (corresponding to 61% prevalence), including eight previously undescribed ones. The novel sequence variants included five missense mutations (p.Lys161Asn c.483G>C in exon 4, p.Phe171Leu c.511T>C in exon 5 and p.Thr228Ala c.682A>G, p.Thr228Arg c.683C>G, p.Gly258Cys c.772G>T in exon 7), one nonsense mutation (p.Ser383Ter c.1148C>A in exon 9), the splice site variant c.1253+1G>T in intron 9, and the deletion of 12 nucleotides in exon 10 (p.Ser433_Ile436del c.1298_1309del12). Our study indicates that mutations in the GCK/MODY2 gene are a very common cause of MODY in the Italian population and broadens our knowledge of the naturally occurring GCK mutation repertoire.     

Specific polymorphisms in the RET proto-oncogene are over-represented in patients with Hirschsprung disease and may represent loci modifying phenotypic expression

Hirschsprung disease (HSCR) is a common genetic disorder presenting with functional intestinal obstruction secondary to enteric aganglionosis. HSCR can be familial or sporadic. Although five putative susceptibility genes have been identified, only germline mutations in the RET proto-oncogene account for a significant minority (up to 50%) of familial HSCR; 3% of sporadic HSCR in a population based series carry RET mutations. From 1998 to February 1999, we prospectively ascertained 64 cases of sporadic HSCR from the western Andalusia region. To determine if polymorphic sequence variants within RET could act as low penetrance predisposing alleles, we examined allelic frequencies at seven polymorphic loci in this population based series. Whether allele frequencies differed from those in the control population were determined by either chi-squared analysis or Fisher's exact test. For two sequence variants, A45A (c 135G-->A) (exon 2) and L769L (c 2307T-->G) (exon 13), the rarer polymorphic allele was over-represented among HSCR cases versus controls (p<0.0006). In contrast, two other polymorphisms, G691S (c 2071C-->A) (exon 11) and S904S (c 2712C-->G) (exon 15), were under-represented in the HSCR patients compared to controls (p=0.02). Polymorphisms in the RET proto-oncogene appear to predispose to HSCR in a complex, low penetrance fashion and may also modify phenotypic expression.     

Mutations in the TSC2 gene: analysis of the complete coding sequence using the protein truncation test (PTT)

Mutations in the TSC2 gene on chromosome 16p13.3 are responsible for approximately 50% of familial tuberous sclerosis (TSC). The gene has 41 small exons spanning 45 kb of genomic DNA and encoding a 5.5 kb mRNA. Large germline deletions of TSC2 occur in <5% of cases, and a number of small intragenic mutations have been described. We analysed mRNA from 18 unrelated cases of TSC for TSC2 mutations using the protein truncation test (PTT). Three cases were predicted to be TSC2 mutations on the basis of linkage analysis or because a hamartoma from the patient showed loss of heterozygosity for 16p13.3 markers. Three overlapping PCR products, covering the complete coding sequence of mRNA, were generated from lymphoblastoid cell lines, translated into 35S-methionine labelled protein, and analysed by SDS-PAGE. PCR products showing PTT shifts were directly sequenced, and mutations confirmed by restriction enzyme digestion where possible. Six PTT shifts were identified. Five of these were caused by mutations predicted to produce a truncated protein: (i) a sporadic case showed a 32 bp deletion in exon 11, and a mutant mRNA without exon 11 was produced; the normal exon 10 was also spliced out; (ii) a sporadic case had a 1 bp deletion in exon 12 (1634delT); (iii) a TSC2-linked mother and daughter pair had a G-->T transversion in exon 23 (G2715T) introducing a cryptic splice site causing a 29 bp truncation of mRNA from exon 23; (iv) a sporadic case showed a 2 bp deletion in exon 36; (v) a sporadic case showed a 1 bp insertion disrupting the donor splice site of exon 37 (5007+2insA), resulting in the use of an upstream exonic cryptic splice site to cause a 29 bp truncation of mRNA from exon 37. In one case, the PTT shift was explained by in-frame splicing out of exon 10, in the presence of a normal exon 10 genomic sequence. Alternative splicing of exon 10 of the TSC2 gene may be a normal variant. Three 3rd base substitution polymorphisms were also detected during direct sequencing of PCR products. Confirmed mutations were identified in 28% of the families studied and on the assumption that half of the sporadic cases should have TSC2 mutations, a crude estimate of the detection rate would be 60%. This compares favourably with other screening methods used for TSC2, notably SSCP, and since PTT involves much less work it may be the method of choice.