Three apolipoprotein B (apoB) genetic polymorphisms, the XbaI, MspI and EcoRI restriction fragment length polymorphisms (RFLPs), were analysed for 221 individuals in Senegal by polymerase chain reaction. Allelic frequency determination revealed that this population has 0.79 XbaI− (X−), 0.96 MspI+ (M+), and 0.89 EcoRI+ (E+). Major genotypes were X−/X− (0.62), M+/M+ (0.92) and E+/E+ (0.80). The XbaI allele frequency is different (p<10−9) from that in Caucasians (0.47) and from that in Mongoloids (0.98). Significant differences between Senegalese and Caucasians, and between Senegalese and Mongoloids were also observed for the EcoRI and MspI alleles of the apoB genes.

RALGDS is a 115 kDa protein which was identified by its ability to enhance guanine nucleotide exchange for the ras family member ral. It also binds to activated ras and rap1, and appears to function as part of a signalling complex in downstream events following rap1 activation. Here we report the identification of full-length cDNA clones for human RALGDS, isolated from a brain cDNA library. The predicted protein has strong sequence homology to rat and murine isoforms of RALGDS in the N- and C-terminal regions, but an internal region (aa 250–380) shows relatively high divergence with only 42% identical amino acid residues. The human RALGDS gene is contained within a 30 kb region of 9q34, approximately 200 kb proximal to the ABO gene, within the current critical region for the tuberous sclerosis gene TSC1. Partial genomic structure was determined; it consists of at least 11 exons. Based upon analysis of Southern blots from 110 TSC patients, genomic DNA SSCP analysis, and RT-PCR analysis which demonstrated RNA expression of both alleles in patients from 9q34-linked TSC families using intragenic polymorphisms, we conclude that RALGDS is not likely to be TSC1.

For many complex diseases, study has suggested that the disease genes influence not only the occurrence of the disease, but also the age of onset. Current methods in linkage analysis are mainly concentrated on affected relative pairs or affected family members, and age of onset information is either ignored or is taken into account by specifying age-dependent penetrances for liability classes. In fact, affected relatives with different ages of onset may be the result of different genetic aetiologies and unaffected relatives are censored at the study time. Therefore, incorporation of age of onset and including contrasts between affected and unaffected pedigree members are important components of effective analysis for the detection of linkage with genetic markers. We use multiple markers to infer the inheritance vector in order to extract information about the inheritance pattern of the disease allele in a pedigree. For a given inheritance vector, we define two neighbour sets for each individual based on allele identical by descent (IBD). We then use the within-set and between-sets conditional hazard ratios to characterize the dependence of age of onset among relatives. A pseudolikelihood ratio test is proposed for testing linkage. Simulated and real data sets are used to illustrate these new statistical methods.

Four sequence variants in the 11β-hydroxylase (CYP11B1) gene are reported. One of the sequence changes occurs in exon 1 and is in linkage disequilibrium with a second variant in intron 3. The other two changes occur at adjacent nucleotides in intron 1. The finding of easily demonstrable, intragenic variants will be beneficial to the study of the role of the CYP11B1 and the adjacent aldosterone synthase (CYP11B2) gene in hypertensive disease.

We have prepared a transferrin (TF) cDNA from total RNA of peripheral blood cells of an individual with a TF B variant (TF Bv) phenotype to analyze the molecular basis for the TF Bv. The TF B variant allele (TF Bv) was found to contain an A to G transition at nucleotide 1879 in the coding region that may result in substitution of glutamic acid for lysine at codon 627 located in exon 16. In addition, a transient expression study by transfection of each plasmid containing TF C1, TF C2 or TF Bv cDNA into monkey kidney (COS7) cells indicated that the products from specific TF alleles were able to be identified respective phenotypes by isoelectrofocusing electrophoresis and immunoblotting analyses. Based on this nucleotide substitution, we also established a PCR–RFLP method to detect the TF Bv allele from genomic DNA. Thus, peripheral blood cells were useful as an alternative source for preparation of cDNA from human liver for the molecular analysis of TF polymorphism.

LCN1 gene encodes the tear lipocalin; the lipocalins are a large and growing family of proteins characterized by their ability to bind small hydrophobic molecules. We report here the location of a dinucleotide repeat microsatellite marker (D9S1826) close to LCN1 gene. Using the CEPH reference families, the position of LCN1 is located within the 9q34 genetic map between D9S23 and D9S158.

A powerful test for population association of a disease with alleles at a bi-allelic marker locus is the transmission/disequilibrium test (TDT). A generalization of the test to multi-allelic marker locus is proposed which utilizes the maximal association of individual alleles with the disease, given by the maximum TDT statistic, TDT(max). To overcome the multiple testing problem encountered when using the maximal association to test the null hypothesis of no disease-marker association, a randomization procedure is developed. An investigation of the power of the test suggests that the randomization procedure performs almost as well as a recently proposed likelihood based test of linkage disequilibrium. The advantage of the new test is that it can be applied sequentially, based on a one-sided version of the TDT statistic, for investigating patterns of association of several individual alleles with the disease.

Linkage tests to localize oligogenes have been extended during the past year. Using simulated data and multiplex selection we find that several tests on affected sib pairs have comparable power and type I error. Three variants of SIBPAL2 are favoured when substantial numbers of normal sibs are included, but performance relative to the BETA benchmark degrades rapidly as normal sibs are depleted by selective sampling or typing. Neglect of this fact may explain recent failure of retrospective collaboration to confirm asthma candidates in the 5q cytokine region that are supported by other studies. A fully quantitative trait favours variance components under complete ascertainment and two options in SIBPAL2 under multiplex selection, with substantial gain in power from covariance analysis if the covariate is independent of the candidate locus. A dichotomy and liability threshold give virtually identical results in the SOLAR variance components program. Comparison with single-marker parametric analysis suggests that extension to multiple markers would be competitive with nonparametric methods in power, and superior in depth of genetic analysis. The simulated examples illustrate common problems encountered with linkage scans for oligogenes. They show that nonparametric methods provide no panacea for analytical problems posed by different phenotypes and methods of ascertainment.

Most studies of the pathogenesis of coronary heart disease occur between gene variants and biochemical or physiological variables known to be atherogenic. In many situations, however, the gene products are not necessarily known. We studied 17 families (n = 122) with mutations in the low density lipoprotein (LDL) receptor gene as a model in which to test formally for linkage directly between an atherogenic genotype and ischemic heart disease (IHD) or aorto-coronary calcified atherosclerosis. In each family one of three different mutations was found: the Trp66–Gly mutation, the Trp23–Stop mutation, or a ten kilobase deletion removing exons 3–6 of the LDL receptor gene. Genomic DNA was used to determine these mutations by either enzymatic cleavage assays or Southern blotting. Aorto-coronary calcification was significantly associated with age and plasma cholesterol. Sex, hypertension, BMI and smoking were not associated with aorto-coronary calcification. Nonparametric analysis indicated significant linkage of the LDL receptor gene locus to aortic (p < 0.00005) and to aorto-coronary calcified atherosclerosis (p < 0.00001). Assuming a dominant mode of inheritance, significant linkage was detected for aortic (LOD = 3.89) and aorto-coronary calcified atherosclerosis (LOD = 4.10). We suggest that the atherogenicity of variations in other genes could be assessed by a similar approach.

Two popular and robust approaches to analysing affected sib pair (ASP) data for linkage are the traditional excess sharing methods and the transmission/disequilibrium test (TDT). Here we derive an overall test of linkage for multi-allelic ASP marker data which comprises two component tests: one for excess sharing and one for transmission disequilibrium. This method has several advantages. Firstly the overall test of linkage is often more powerful than either of the two component tests. Secondly the method makes it possible to determine the contribution of linkage disequilibrium (LD), in addition to linkage, to an overall positive linkage result. This is useful because the presence of LD in addition to linkage may suggest that the marker locus is in very close proximity to a disease susceptibility gene. Thirdly the method provides estimates of the risk associated with transmission of the different marker alleles.

Penrose's sib-pair papers (1935–1953) are discussed in relation to recent applications. His essential contribution that parental typing was an inefficient addition to sib-pair data for linkage detection remains. Parents now have even less to offer with contemporary markers in the detection of linkage, although not the enumeration of haplotypes. Attention is drawn to two little known papers by Penrose on multifactorial disease, and it is suggested that this term should replace polygenic in relation to family analyses. The convention established by Penrose, both as author and editor, of raw data being published alongside its analysis, which later fell into disuse on grounds of sheer bulk, can now be remedied by the Internet, and has been for the largest set of sib-pair data on diabetes (Davies et al. Nature371 (1994), 130–136). This is essential if the contributions to the identity and location of alleles possibly relevant to common disorders, none of which are likely to be more than suggestive in any single study, can be combined to assess consistency and joint evidence. Attention is drawn to a third little known paper by Penrose of high relevance to the strategy of familial investigations in multifactorial disorders. All five of the Penrose papers discussed here can be viewed at http://www.gene.ucl.ac.uk/anhumgen/

Parametric linkage analysis of simultaneous mapping of the two disease loci of a qualitative trait governed by a two-locus model has been shown to provide greater power in detecting linkage than standard lod-score analysis that maps a single disease locus. Despite its great potential for power gains, two-locus parametric analysis has not been used routinely in disease gene mapping, due to the computational intensity of currently available methods and programs. In this paper, we propose a Markov chain Monte Carlo (MCMC) method for performing lod-score analysis of qualitative traits governed by two-locus models. This method obtains lod-score estimates that can be arbitrarily close to their corresponding exact values. The algorithm implementing this MCMC method is linear in the number of markers. This feature enables us to perform two-locus analysis mapping each trait to a set of markers, instead of just to a single marker. We analyzed an alcohol dependence dataset composed of 105 pedigrees with various sizes and various degrees of missingness in the observed marker and disease data. The estimates from our MCMC procedure match up well with the lod scores from exact analysis, but it took much less time for the MCMC procedure to obtain the results. We also performed a simulation study to investigate power gains with additional markers. Our results indicate that an additional marker on each map can provide a great deal more information for linkage measured in terms of the magnitude of lod scores.

When dominant mutations of different genes may lead to the same disease, it is often difficult to detect in a particular patient which gene is involved. A strategy is to make genealogical extensions to find affected relatives that should have inherited the same mutation. In particular, for diseases with late age of onset or short survival time, only poor information may be obtained from close relatives of probands and it can be particularly efficient to make genealogical extensions to detect pairs of distantly related affected individuals. Such a pair of affecteds may provide information concerning the region of the genome where the mutated gene should map. Two situations may be encountered depending on whether or not prior information on the location of mutated genes involved in the disease are available. If we already know, from previous linkage studies, that a gene located in a given region R of the genome may be involved in the disease, the problem is then to confirm that it is indeed a mutation of this gene that is involved in the affected pair. Once the implication of a gene in region R has been confirmed the affected pair of relatives may give information to restrict the length of this region R. In this paper we discuss these two points by deriving analytically first the lod score expected and second the expected reduction of the length of the region where the mutation is suspected to map as a function of the number of meioses between the two affected individuals and of the polymorphism of the markers available in the region.

We present a new method for the fine scale mapping of disease loci based on samples of simplex families, each containing an affected child. The method is based on a generalisation of a single locus allele transmission model to multiple marker loci. The model is developed under the assumption of a single ancestral mutation and allows for the calculation of posterior probabilities that each allele at a particular marker was present on the founder chromosome. We illustrate the method using simulated family data for cystic fibrosis and Huntingtons disease, for which the locations of mutations in the disease genes are now known. For both diseases, our new method provides good estimates of the location of the mutations.

A crucial step beyond the identification of genetic linkage of a disease to a chromosomal region is the production of a physical map that will allow the identification of candidate genes. Although the process of physical map building has been facilitated by the flow of data released by the Human Genome Project, gathering all the information together requires significant effort. In a previous study, we reported linkage between Bipolar Affective Disorder and the chromosomal location 4p15.3–p16.1. In this review we use this example to describe how to collect publicly available sequence, DNA fingerprint, and genetic marker data and integrate these with empirical data to build a large scale high resolution physical map of a region. Methods used to identify new genetic markers and candidate genes within a circumscribed region are also presented.

Phosphoglucomutase 1 (PGM1) deficiency is a stable characteristic of the erythroleukaemic cell line, K562, whereas the activity of the isozymes of the other two PGM loci (PGM2 and PGM3) is slightly elevated. In this study the molecular basis of PGM1 deficiency was investigated by a combined approach utilising protein electrophoresis, immunodetection, cytogenetic techniques, and DNA and RNA analysis. Isoelectric focusing and activity staining confirmed that K562 has no detectable PGM1 activity. Immunoblot analysis of extracts, separated by isoelectric focusing, starch gel and SDS gel electrophoresis, using monospecific anti-PGM1 antibodies showed that K562 contained no detectable immunoreactive material. Karyotype analysis revealed the presence of two intact chromosomes 1 and a derivative chromosome 1, der(1)t(1; 11), each of which carried a copy of the PGM1 gene as demonstrated by fluorescence in situ hybridization using a PGM1 cosmid as probe. Southern blot analysis using a PGM1 cDNA clone as probe suggested that the PGM1 genes had not been subject to any gross structural rearrangements. We were also able to determine that K562 is type PGM1 2+1+ by restriction endonuclease analysis of genomic DNA. Very low levels of PGM1 mRNA which appeared to be full length transcripts were detected in K562 using a reverse transcriptase PCR technique. We conclude that the most likely cause of PGM1 enzyme deficiency in K562 is abnormal regulation of transcription.

We addressed the question: Is there evidence that allelic variation in a single unmeasured gene that has a large effect on maximal activity of erythrocyte sodium-lithium countertransport (Na-Li CNT) also has pleiotropic effects on variation in plasma triglyceride levels? Complex segregation analysis models that included plasma triglyceride levels as a covariate were considered as explanations for interindividual variation in Na-Li CNT. A sample of 711 healthy adults from 254 pedigrees enrolled in the Rochester Family Heart Study was selected for this study. The majority of the pedigrees supported the hypothesis that variations in a single unmeasured non-transmitted environmental factor have large effects on the Na-Li CNT distribution. Only gender-specific first-order covariate parameters were necessary in the complex segregation models suggesting that the form of the relationship between Na-Li CNT and plasma triglyceride level was not influenced by variation in the inferred environmental factor with large effects. Stratification of the sample by this inferred environmental factor resulted in three classes of individuals with significant differences in the distributions of coronary heart disease risk factor traits, as well as interindividual variation in both Na-Li CNT and plasma triglyceride levels. These results, along with other observations from the Rochester Family Heart Study sample, emphasize the complex and multifactorial nature of the causes of interindividual variation in Na-Li CNT. Our study further suggests that new research strategies are needed for studying the relationships between genetic and environmental variation and variation in quantitative traits such as Na-Li CNT that have been identified as risk factors for hypertension.

Duchenne and Becker muscular dystrophies are among the most severe and frequent inherited disorders. Being still incurable, medical treatment is concentrated on the carrier diagnosis of the members of the affected families. Here we report the results of the studies of 151 members of 41 Hungarian families, obtained with multiplex PCR amplification of 18 exons as well as the muscle specific promoter region, and haplotype analysis of two polymorphic (CA)n repeat microsatellite loci in introns 45 and 49 of the dystrophin gene. The analysis of 15 deletion-type families revealed a frequency of new mutations not differing significantly from that in the other regions of Europe. We also compared the allele distributions of the two microsatellites in randomly selected normal individuals and affected family members. The allele distribution of STRP45 shows interesting differences between the two populations.

In its usual form, the multifactorial model of disease transmission assumes that the liabilities to disease have a multivariate normal distribution. This paper studies how sensitive to this assumption are the quantitative results from the model. Accordingly, bounds are established for the probability of a child having a disease, given that both parents have it and taking the heritability of the disease to be known. Unfortunately, these bounds turn out to be wide. For example, a probability that is 0.38 under the trivariate normal model may be as low as 0.12 or as high as 0.78 under other trivariate models, even if attention is restricted to those of variables-in-common form. The broader statistical issue of the meaning of trivariate dependence, as distinct from bivariate dependence, is also discussed.