Human Mendelian pain disorders: a key to discovery and validation of novel analgesics.

We have utilized a novel application of human genetics, illuminating the important role that rare genetic disorders can play in the development of novel drugs that may be of relevance for the treatment of both rare and common diseases. By studying a very rare Mendelian disorder of absent pain perception, congenital indifference to pain, we have defined Nav1.7 (endocded by SCN9A) as a critical and novel target for analgesic development. Strong human validation has emerged with SCN9A gain-of-function mutations causing inherited erythromelalgia (IEM) and paroxysmal extreme pain disorder, both Mendelian disorder of spontaneous or easily evoked pain. Furthermore, variations in the Nav1.7 channel also modulate pain perception in healthy subjects as well as in painful conditions such as osteoarthritis and Parkinson disease. On the basis of this, we have developed a novel compound (XEN402) that exhibits potent, voltage-dependent block of Nav1.7. In a small pilot study, we showed that XEN402 blocks Nav1.7 mediated pain associated with IEM thereby demonstrating the use of rare genetic disorders with mutant target channels as a novel approach to rapid proof-of-concept. Our approach underscores the critical role that human genetics can play by illuminating novel and critical pathways pertinent for drug discovery.

Loss-of-function mutations in the Nav1.7 gene underlie congenital indifference to pain in multiple human populations.

Congenital indifference to pain (CIP) is a rare condition in which patients have severely impaired pain perception, but are otherwise essentially normal. We identified and collected DNA from individuals from nine families of seven different nationalities in which the affected individuals meet the diagnostic criteria for CIP. Using homozygosity mapping and haplotype sharing methods, we narrowed the CIP locus to chromosome 2q24-q31, a region known to contain a cluster of voltage-gated sodium channel genes. From these prioritized candidate sodium channels, we identified 10 mutations in the SCN9A gene encoding the sodium channel protein Nav1.7. The mutations completely co-segregated with the disease phenotype, and nine of these SCN9A mutations resulted in truncation and loss-of-function of the Nav1.7 channel. These genetic data further support the evidence that Nav1.7 plays an essential role in mediating pain in humans, and that SCN9A mutations identified in multiple different populations underlie CIP.

A genome-wide scan points to a susceptibility locus for bipolar disorder on chromosome 12.

Our previous results pointed to a putative gene for susceptibility to bipolar affective disorder located on the chromosomal region 12q23-q24 that segregated in the Saguenay-Lac-St-Jean population of Quebec. We report here results from a second genome-wide scan based on the analysis of 380 polymorphic microsatellite markers. For the purpose of this analysis, an additional 18 families were recruited from the Saguenay-Lac-St-Jean region and pooled to our previous sample to improve its statistical power, giving a total of 394 sampled individuals. This work confirms the presence of a susceptibility locus for affective disorder on chromosome 12q24 with parametric LOD score value of 3.35 at D12S378 when pedigrees were broken into nuclear families and analysed under a recessive segregation model. This result was supported by neighbouring markers and by a LOD score value of 5.05 at D12S378 under model-free analysis. Other regions of lower interest were indicated on chromosomes 2, 5, 7, 9, 10, 17 and 20.

A schizophrenia-susceptibility locus at 6q25, in one of the world's largest reported pedigrees.

We have completed a genome scan of a 12-generation, 3,400-member pedigree with schizophrenia. Samples from 210 individuals were collected from the pedigree. We performed an "affecteds-only" genome-scan analysis using 43 members of the pedigree. The affected individuals included 29 patients with schizophrenia, 10 with schizoaffective disorders, and 4 with psychosis not otherwise specified. Two sets of white-European allele frequencies were used-one from a Swedish control population (46 unrelated individuals) and one from the pedigree (210 individuals). All analyses pointed to the same region: D6S264, located at 6q25.2, showed a maximum LOD score of 3.45 when allele frequencies in the Swedish control population were used, compared with a maximum LOD score of 2.59 when the pedigree's allele frequencies were used. We analyzed additional markers in the 6q25 region and found a maximum LOD score of 6.6 with marker D6S253, as well as a 6-cM haplotype (markers D6S253-D6S264) that segregated, after 12 generations, with the majority of the affected individuals. Multipoint analysis was performed with the markers in the 6q25 region, and a maximum LOD score of 7.7 was obtained. To evaluate the significance of the genome scan, we simulated the complete analysis under the assumption of no linkage. The results showed that a LOD score >2.2 should be considered as suggestive of linkage, whereas a LOD score >3.7 should be considered as significant. These results suggest that a common ancestral region was inherited by the affected individuals in this large pedigree.

Genomewide genetic linkage analysis confirms the presence of susceptibility loci for schizophrenia, on chromosomes 1q32.2, 5q33.2, and 8p21-22 and provides support for linkage to schizophrenia, on chromosomes 11q23.3-24 and 20q12.1-11.23.

We have performed genetic linkage analysis in 13 large multiply affected families, to test the hypothesis that there is extensive heterogeneity of linkage for genetic subtypes of schizophrenia. Our strategy consisted of selecting 13 kindreds containing multiple affected cases in three or more generations, an absence of bipolar affective disorder, and a single progenitor source of schizophrenia with unilineal transmission into the branch of the kindred sampled. DNA samples from these families were genotyped with 365 microsatellite markers spaced at approximately 10-cM intervals across the whole genome. We observed LOD scores >3.0 at five distinct loci, either in the sample as a whole or within single families, strongly suggesting etiological heterogeneity. Heterogeneity LOD scores >3.0 in the sample as a whole were found at 1q33.2 (LOD score 3.2; P=.0003), 5q33.2 (LOD score 3.6; P=.0001), 8p22.1-22 (LOD score 3.6; P=.0001), and 11q21 (LOD score 3.1; P=.0004). LOD scores >3.0 within single pedigrees were found at 4q13-31 (LOD score 3.2; P=.0003) and at 11q23.3-24 (LOD score 3.2; P=.0003). A LOD score of 2.9 was also found at 20q12.1-11.23 within in a single family. The fact that other studies have also detected LOD scores >3.0 at 1q33.2, 5q33.2, 8p21-22 and 11q21 suggests that these regions do indeed harbor schizophrenia-susceptibility loci. We believe that the weight of evidence for linkage to the chromosome 1q22, 5q33.2, and 8p21-22 loci is now sufficient to justify intensive investigation of these regions by methods based on linkage disequilibrium. Such studies will soon allow the identification of mutations having a direct effect on susceptibility to schizophrenia.

Failure to establish linkage on the X chromosome in 301 families with schizophrenia or schizoaffective disorder.

The hypothesis that a gene for susceptibility to psychosis (specifically in the X-Y homologous class) is located on the sex chromosomes has been proposed. Such a gene would account for the excess of sex chromosome anomalous males and females in populations of patients with psychosis, a tendency towards concordance by sex within families, and sex differences associated with psychosis and its underlying brain pathology. In earlier studies we observed small positive LOD scores in Xp11, and in a more recent and larger cohort of 178 sibling pairs, a peak multipoint nonparametric LOD score of 1. 55 at the locus DXS8032 in Xq21. The present study with a new set of markers extended the cohort to 301 ill sibling pairs and their parents. Despite the increase in sample size, the LOD score did not increase. A peak NPL of 1.55 was observed at the locus DXS1068 in proximal Xp, a region remote from the previous report. Separating families into those who were more likely to have X chromosome inheritance (maternal with no male to male transmission) did not yield stronger findings. In spite of the evidence that psychosis is related to a sex-dependent dimension of cerebral asymmetry, it is concluded that no consistent linkage of schizophrenia to the X chromosome can be demonstrated. In the context of the general failure of replication of linkage in psychosis, the possibility that the genetic predisposition to psychosis is contributed to by epigenetic modification rather than variations in the nucleotide sequence has to be considered.

Lack of evidence for linkage to chromosomes 13 and 8 for schizophrenia and schizoaffective disorder.

A previous report [Blouin et al., 1998: Nat Genet 20:70-73] suggesting linkage to chromosomes 13q32 and 8p21 in families with schizophrenia led us to investigate these regions in a large set of 301 multiplex families with schizophrenia. Multipoint analyses failed to reveal evidence for linkage to any portion of chromosome 13, while only a weakly positive score was present on 8p using the identical marker reported in the earlier report. Failure to confirm the Blouin et al claims in a substantially larger cohort adds emphasis to the inconsistency of the findings concerning linkage in schizophrenia. Am. J. Med. Genet. (Neuropsychiatr. Genet.) 96:235-239, 2000.

New DNA markers with increased informativeness show diminished support for a chromosome 5q11-13 schizophrenia susceptibility locus and exclude linkage in two new cohorts of British and Icelandic families.

Genetic linkage of schizophrenia to markers at 5q11.2-13.3 had been reported previously in five Icelandic and two British families, but attempts at replication in independent samples have been unsuccessful. We report here an update on the diagnoses and results of linkage analyses using newer highly polymorphic microsatellite markers at or near the loci D5S76 and D5S39 in the original sample of pedigrees and in two new family samples from Iceland and from Britain. The new results show a reduction in evidence for linkage in the original sample and evidence against linkage in the two new family samples. Although it is possible that a rare locus is present, perhaps in the region 5p14.1-13.1 rather than 5q11.2-13.3, it appears most likely that the original positive lod scores represent an exaggeration of the 'true' lod scores due to random effects and that the small lod scores we now obtain could have arisen by chance.

A genome-wide search for schizophrenia susceptibility genes.

We completed a systematic genome-wide search for evidence of loci linked to schizophrenia using a collection of 70 pedigrees containing multiple affected individuals according to three phenotype classifications: schizophrenia only (48 pedigrees; 70 sib-pairs); schizophrenia plus schizoaffective disorder (70 pedigrees; 101 sib-pairs); and a broad category consisting of schizophrenia, schizoaffective disorder, paranoid or schizotypal personality disorder, psychosis not otherwise specified (NOS), delusional disorder, and brief reactive psychosis (70 pedigrees; 111 sib-pairs). All 70 families contained at least one individual affected with chronic schizophrenia according to DSM-III-R criteria. Three hundred and thirty-eight markers spanning the genome were typed in all pedigrees for an average resolution of 10.5 cM (range, 0-31 cM) and an average heterozygosity of 74.3% per marker. The data were analyzed using multipoint nonparametric allele-sharing and traditional two-point lod score analyses using dominant and recessive, affecteds-only models. Twelve chromosomes (1, 2, 4, 5, 8, 10, 11, 12, 13, 14, 16, and 22) had at least one region with a nominal P value <0.05, and two of these chromosomes had a nominal P value <0.01 (chromosomes 13 and 16), using allele-sharing tests in GENEHUNTER. Five chromosomes (1, 2, 4, 11, and 13) had at least one marker with a lod score >2.0, allowing for heterogeneity. These regions will be saturated with additional markers and investigated in a new, larger set of families to test for replication.

No evidence of expansion of CAG or GAA repeats in schizophrenia families and monozygotic twins.

Many diseases caused by trinucleotide expansion exhibit increased severity and decreased age of onset (genetic anticipation) in successive generations. Apparent evidence of genetic anticipation in schizophrenia has led to a search for trinucleotide repeat expansions. We have used several techniques, including Southern blot hybridization, repeat expansion detection (RED) and locus-specific PCR to search for expanded CAG/CTG repeats in 12 families from the United Kingdom and 11 from Iceland that are multiplex for schizophrenia and demonstrate anticipation. The unstable DNA theory could also explain discordance of phenotype for schizophrenia in pairs of monozygotic twins, where the affected twin has a greater number of repeats than the unaffected twin. We used these techniques to look for evidence of different CAG/CTG repeat size in 27 pairs of monozygotic twins who are either concordant or discordant for schizophrenia. We have found no evidence of an increase in CAG/CTG repeat size for affected members in the families, or for the affected twins in the MZ twin sample. Southern hybridization and RED analysis were also performed for the twin and family samples to look for evidence of expansion of GAA/TTC repeats. However, no evidence of expansion was found in either sample. Whilst these results suggest that these repeats are not involved in the etiology of schizophrenia, the techniques used for detecting repeat expansions have limits to their sensitivity. The involvement of other trinucleotide repeats or other expandable repeat sequences cannot be ruled out.

Amyloid-beta-protein isoforms in brain of subjects with PS1-linked, beta APP-linked and sporadic Alzheimer disease.

To determine whether similar abnormalities of various soluble full-length and N-terminal truncated Abeta peptides occur in postmortem cerebral cortex of affected PS1 mutation carriers, we examined the amounts of two amyloid species ending at residue 40 or at residues 42(43) using sandwich ELISA systems. Our results indicate that PS1 mutations effect a dramatic accumulation in brain of the highly insoluble potentially neurotoxic long-tailed isoforms of the Abeta peptide such as Abeta1-42(43) and Abetax-42(43). This enhancing effect of PS1 mutation on Abetax-42(43) deposition was highly similar to that of a betaAPP mutation (Val717Ile) but the effects on Abetax-40 production were significantly different between these two causal genes. In contrast to previous studies of soluble Abeta in plasma and in supernatants from cultured fibroblasts of subjects with PS1 mutations, our studies also show that there is an increase in insoluble Abetax-40 peptides in brain of subjects with PS1 mutations.

Estimation of the genetic contribution of presenilin-1 and -2 mutations in a population-based study of presenile Alzheimer disease.

Two closely related genes, the presenilins ( PS ), located at chromosomes 14q24.3 and 1q42.1, have been identified for autosomal dominant Alzheimer disease (AD) with onset age below 65 years (presenile AD). We performed a systematic mutation analysis of all coding and 5'-non-coding exons of PS -1 and PS -2 in a population-based epidemiological series of 101 unrelated familial and sporadic presenile AD cases. The familial cases included 10 patients of autosomal dominant AD families sampled for linkage analysis studies. In all patients mutations in the amyloid precursor protein gene ( APP ) had previously been excluded. Four different PS -1 missense mutations were identified in six familial cases, two of which where autosomal dominant cases. Three mutations resulted in onset ages above 55 years, with one segregating in an autosomal dominant family with mean onset age 64 years (range 50-78 years). One PS -2 mutation was identified in a sporadic case with onset age 62 years. Our mutation data provided estimates for PS -1 and PS -2 mutation frequencies in presenile AD of 6 and 1% respectively. When family history was accounted for mutation frequencies for PS -1 were 9% in familial cases and 18% in autosomal dominant cases. Further, polymorphisms were detected in the promoter and the 5'-non-coding region of PS -1 and in intronic and exonic sequences of PS -2 that will be useful in genetic association studies.

Linkage of bipolar affective disorder to chromosome 18 markers in a new pedigree series.

Several groups have reported evidence suggesting linkage of bipolar affective disorder (BPAD) to chromosome 18. We have reported data from 28 pedigrees that showed linkage to marker loci on 18p and to loci 40 cM distant on 18q. Most of the linkage evidence derived from families with affected phenotypes in only the paternal lineage and from marker alleles transmitted on the paternal chromosome. We now report results from a series of 30 new pedigrees (259 individuals) genotyped for 13 polymorphic markers spanning chromosome 18. Subjects were interviewed by a psychiatrist and were diagnosed by highly reliable methods. Genotypes were generated with automated technology and were scored blind to phenotype. Affected sib pairs showed excess allele sharing at the 18q markers D18S541 and D18S38. A parent-of-origin effect was observed, but it was not consistently paternal. No robust evidence of linkage was detected for markers elsewhere on chromosome 18. Multipoint nonparametric linkage analysis in the new sample combined with the original sample of families supports linkage on chromosome 18q, but the susceptibility gene is not well localized.

Linkage analysis of manic depression (bipolar affective disorder) in Icelandic and British kindreds using markers on the short arm of chromosome 18.

Attempts were made to follow up results of a previous linkage study which suggested that a locus-modifying susceptibility to bipolar and related unipolar affective disorder might be present in the pericentromeric region of the short arm of chromosome 18. Twenty-three multiply affected pedigrees collected from Iceland and the UK were genotyped using three highly polymorphic microsatellite markers at D18S37, D18S40 and D18S44 which span the region implicated. Lod score analyses under the assumption of heterogeneity and non-parametric linkage analyses were performed. The total lod scores obtained were strongly negative, and analysis allowing for heterogeneity did not suggest that any subgroup of the families was linked. Model-free linkage analysis using extended relative pair analysis and MFLINK also failed to detect any evidence for linkage. Our study provides no support for the presence of a locus-modifying genetic susceptibility to bipolar affective disorder in the pericentromeric region of chromosome 18q11. Further analyses in independent samples should help to reveal whether our negative results are due to locus heterogeneity or whether the original results were false-positive.

Analysis of the 5' sequence, genomic structure, and alternative splicing of the presenilin-1 gene (PSEN1) associated with early onset Alzheimer disease.

Mutations in the human presenilin genes (PSEN1 and PSEN2) are associated with early onset familial Alzheimer disease. The presenilin genes encode integral membrane proteins with similar structures, which suggests that they may have closely related, but as yet unknown functions. Analysis of the 5' upstream sequence and the structure of the PSEN1 gene reveals that the 5' sequence contains multiple putative transcription regulatory elements including clusters of STAT elements involved in transcriptional activation in response to signal transduction. The first four exons contain untranslated sequences, with Exons 1 and 2 representing alternate initial transcription sites. The function of these alternate initial exons is unclear. Exon 4 bears the first ATG sequence. The last 12 bp of Exon 4 is used as an alternative splice donor site. Exon 9 is alternately spliced in leukocytes, but not in most other tissues. Splicing of Exon 9 is predicted to cause significant structural changes to the protein. The majority of transcripts expressed in most tissues are polyadenylated 1127 bp from the TAG stop codon in Exon 13. A small proportion of transcripts contain the same 5'UTR and ORF but are polyadenylated 4435 bp from the stop codon. The longer polyadenylated transcripts contain three additional palindromes and at least one additional stem-loop structure with stabilities greater than -16 kcal/mol.

Presenilin proteins undergo heterogeneous endoproteolysis between Thr291 and Ala299 and occur as stable N- and C-terminal fragments in normal and Alzheimer brain tissue.

Humans inheriting missense mutations in the presenilin (PS)1 and -2 genes undergo progressive cerebral deposition of the amyloid beta-protein at an early age and develop a clinically and pathologically severe form of familial Alzheimer's disease (FAD). Because PS1 mutations cause the most aggressive known form of AD, it is important to elucidate the structure and function of this multitransmembrane protein in the brain. Using a panel of region-specific PS antibodies, we characterized the presenilin polypeptides in mammalian tissues, including brains of normal, AD, and PS1-linked FAD subjects, and in transfected and nontransfected cell lines. Very little full-length PS1 or -2 was detected in brain and untransfected cells; instead the protein occurred as a heterogeneous array of stable N- and C-terminal proteolytic fragments that differed subtly among cell types and mammalian tissues. Sequencing of the major C-terminal fragment from PS1-transfected human 293 cells showed that the principal endoproteolytic cleavage occurs at and near Met298 in the proximal portion of the large hydrophilic loop. Full-length PS1 in these cells is quickly turned over (T1/2 approximately 60 min), in part to the two major fragments. The sizes and amounts of the PS fragments were not significantly altered in four FAD brains with the Cys410Tyr PS1 missense mutation. Our results indicate that presenilins are rapidly processed to N- and C-terminal fragments in both neural and nonneural cells and that interference with this processing is not an obligatory feature of FAD-causing mutations.

Mutant presenilins of Alzheimer's disease increase production of 42-residue amyloid beta-protein in both transfected cells and transgenic mice.

The mechanism by which mutations in the presenilin (PS) genes cause the most aggressive form of early-onset Alzheimer's disease (AD) is unknown, but fibroblasts from mutation carriers secrete increased levels of the amyloidogenic A beta 42 peptide, the main component of AD plaques. We established transfected cell and transgenic mouse models that coexpress human PS and amyloid beta-protein precursor (APP) genes and analyzed quantitatively the effects of PS expression on APP processing. In both models, expression of wild-type PS genes did not alter APP levels, alpha- and beta-secretase activity and A beta production. In the transfected cells, PS1 and PS2 mutations caused a highly significant increase in A beta 42 secretion in all mutant clones. Likewise, mutant but not wildtype PS1 transgenic mice showed significant overproduction of A beta 42 in the brain, and this effect was detectable as early as 2-4 months of age. Different PS mutations had differential effects on A beta generation. The extent of A beta 42 increase did not correlate with presenilin expression levels. Our data demonstrate that the presenilin mutations cause a dominant gain of function and may induce AD by enhancing A beta 42 production, thus promoting cerebral beta-amyloidosis.

The clinical phenotype of two missense mutations in the presenilin I gene in Japanese patients.

We report the clinical and neuropathologic phenotypes associated with two different missense mutations in the presenilin 1 (PS-1) gene in Japanese patients with early-onset familial Alzheimer's disease (FAD). In the AM/JPN1 pedigree a missense mutation (C-->T) was found at nucleotide 1102, which is predicted to cause an alanine-to-valine missense substitution at codon 260. In this family, the disease had a mean age of onset of 40.3 years and an indolent course (range, 8-19 years). Neuropathologic studies in 3 members of this pedigree showed widespread senile plaques, neurofibrillary tangles, and neuronal loss, as well as abundant perivascular subpial amyloid deposits in the Virchow-Robin spaces and the presence of Pick-like intraneuronal inclusions in the dentate gyrus. In the second pedigree, transmitting a C-->T nucleotide substitution at position 1027, leading to the missense mutation of alanine to valine at codon 285, the disease had a later onset (mean, 51 years) but a more rapid course. Comparison of the disease phenotypes associated with other missense mutations in exon 9 of PS-1 reveals no clinical or pathological phenotype, which uniquely distinguishes Alzheimer's disease associated with PS-1 mutations from other forms of early-onset FAD, implying that direct mutation screening is required to identify these cases.

Intergenerational instability of the CAG repeat of the gene for Machado-Joseph disease (MJD1) is affected by the genotype of the normal chromosome: implications for the molecular mechanisms of the instability of the CAG repeat.

Machado-Joseph disease (MJD) is an autosomal dominant neurodegenerative disorder caused by unstable expansion of a CAG repeat in the MJD1 gene at 14q32.1. To identify elements affecting the intergenerational instability of the CAG repeat, we investigated whether the CGG/GGG polymorphism at the 3' end of the CAG repeat affects intergenerational instability of the CAG repeat. The [expanded (CAG)n-CGG]/[normal (CAG)n-GGG] haplotypes were found to result in significantly greater instability of the CAG repeat compared to the [expanded (CAG)n-CGG]/[normal (CAG)n-CGG] or [expanded (CAG)nGGG]/[normal (CAG)n-GGG] haplotypes. Multiple stepwise logistic regression analysis revealed that the relative risk for a large intergenerational change in the number of CAG repeat units (< -2 or > 2) is 7.7-fold (95% CI: 2.5-23.9) higher in the case of paternal transmission than in that of maternal transmission and 7.4-fold (95% CI: 2.4-23.3) higher in the case of transmission from a parent with the [expanded (CAG)n-CGG]/[normal (CAG)n-GGG] haplotypes than in that of transmission from a parent with the [expanded (CAG)n-CGG]/[normal (CAG)n-CGG] or [expanded (CAG)n-GGG]/[normal (CAG)n-GGG] haplotypes. The combination of paternal transmission and the [expanded (CAG)n-CGG]/[normal (CAG)n-GGG] haplotypes resulted in a 75.2-fold (95% CI: 9.0-625.0) increase in the relative risk compared with that of maternal transmission and the [expanded (CAG)n-CGG]/[normal (CAG)n-CGG] or [expanded (CAG)n-GGG]/[normal (CAG)n-GGG] haplotypes. The results suggest that an inter-allelic interaction is involved in the intergenerational instability of the expanded CAG repeat.