The current state of multiple sclerosis genetic research

INTRODUCTION

Multiple sclerosis (MS) is the most common genetic disease of the nervous system with onset usually in young adulthood. Four genome-wide searches in different Caucasian populations for MS susceptibility loci have been performed, but none reported any linkage at a level that would be regarded as significant according to current criteria. Significant linkage of MS to allelic variants of the major histocompatibility (MHC) locus on chromosome 6p21 has been established although its overall contribution to MS susceptibility has proven difficult to quantify. The objective of this review is not only to provide the reader with an update of MS genetics research, but also to provide a basic knowledge of the techniques being employed to map MS susceptibility genes. The different methodologies are discussed, and specific studies are reviewed in context.

METHODS

This review is based on findings from original articles, however, the results of recent candidate gene studies are intended to update previous review articles.

RESULTS

There remains no concrete non-MHC locus for MS, although there are enough findings of sufficient interest to warrant further investigation and optimism. Stratification of genome scan data based on MHC class II suggests that it interacts differentially with non-MHC loci and that it contributes moderately to disease susceptibility. Candidate gene studies have continued to return negative and ambiguous results, and follow-up fine mapping of suggestive linkages from the UK genome scan has proven unsuccessful in identifying significant linkages. Genetic analysis of crosses between mouse strains that are differentially susceptible to experimental allergic encephalomyelitis (EAE) has yielded linkages corresponding to putative MS susceptibility loci. However, recent successes in transgenic mice may provide an alternative to EAE, regarded by some as a poor model of MS.

CONCLUSION

The first whole genome search for a common human disease was performed over five years ago, and it is now clear, from the lack success in this field, that the genetic complexity of these traits has been underestimated. The genome-wide searches for MS susceptibility genes have suffered from insufficient statistical power, which has probably been compounded by disease and genetic heterogeneity. Studies in isolated populations and better laboratory and clinical definitions of disease are both steps in the right direction to solving these problems. Not withstanding the negative effects of genetic heterogeneity, pooling of resources for meta-analyses may provide the increase in statistical power required for detection of loci that exert a moderate or small effect on disease predisposition.

A score test for the linkage analysis of qualitative and quantitative traits based on identity by descent data from sib-pairs

We propose a general likelihood-based approach to the linkage analysis of qualitative and quantitative traits using identity by descent (IBD) data from sib-pairs. We consider the likelihood of IBD data conditional on phenotypes and test the null hypothesis of no linkage between a marker locus and a gene influencing the trait using a score test in the recombination fraction theta between the two loci. This method unifies the linkage analysis of qualitative and quantitative traits into a single inferential framework, yielding a simple and intuitive test statistic. Conditioning on phenotypes avoids unrealistic random sampling assumptions and allows sib-pairs from differing ascertainment mechanisms to be incorporated into a single likelihood analysis. In particular, it allows the selection of sib-pairs based on their trait values and the analysis of only those pairs having the most informative phenotypes. The score test is based on the full likelihood, i.e. the likelihood based on all phenotype data rather than just differences of sib-pair phenotypes. Considering only phenotype differences, as in Haseman and Elston (1972) and Kruglyak and Lander (1995), may result in important losses in power. The linkage score test is derived under general genetic models for the trait, which may include multiple unlinked genes. Population genetic assumptions, such as random mating or linkage equilibrium at the trait loci, are not required. This score test is thus particularly promising for the analysis of complex human traits. The score statistic readily extends to accommodate incomplete IBD data at the test locus, by using the hidden Markov model implemented in the programs MAPMAKER/SIBS and GENEHUNTER (Kruglyak and Lander, 1995; Kruglyak et al., 1996). Preliminary simulation studies indicate that the linkage score test generally matches or outperforms the Haseman-Elston test, the largest gains in power being for selected samples of sib-pairs with extreme phenotypes.

Chromosomes X, 9, and the H2 locus interact epistatically to control Leishmania major infection

As in other infectious diseases, the outcome of a Leishmania major infection is closely tied to the T helper cell response type; progressive disease is associated with a predominant Th2 lymphocyte response, healing with a Th1 response. In mice, susceptibility is genetically con trolled, with BALB/c (C) mice being susceptible and C57BL/6 (B) mice being resistant. Using a genome-wide scan on two large populations of F2 mice created from these strains, we have shown previously that susceptibility to infection with L. major is controlled by two autosomal loci: lmr1 at the H2 locus, and lmr2 on chromosome 9. Employing a strategy to identify loci that interact, we show here that lmr1 and lmr2 interact synergistically, and we describe a new locus lmr3, lying on the X chromosome, whose effect depends on a specific lmr1 haplotype.

An estimate of the crosstalk matrix in four-dye fluorescence-based DNA sequencing

Color separation is an essential step of the data processing in the four-dye fluorescence detection strategy used in automated DNA sequencing. In this paper, we propose a model to describe the crosstalk phenomenon, and show how the assumptions of the model are supported by experimental data. The crosstalk matrix is estimated via a reparameterization based on a mapping between the distribution of fluorescence intensities and that of dye concentrations. An iterative algorithm is designed to implement the estimation. To evaluate the color-correction quality of a crosstalk matrix, we propose a quantitative measure based on the distribution of the color-corrected data. We illustrate this method by applying it to a sequencing trace of slab gel electrophoresis obtained at the Human Genome Center at Lawrence Berkeley National Laboratory, and that of capillary electrophoresis provided by the Department of Chemistry at UC, Berkeley. The accuracy of this method is also assessed by the bootstrap method.

Relative efficiencies of the chi-square recombination models for gene mapping with human pedigree data

In the paper by Goldstein et al. (Genomics, 1995), the authors carried out a simulation study to investigate the relative efficiencies of a no interference linkage analysis to an analysis with certain models that allow for interference. They showed that, for completely informative and independent recombination data, the analysis with the no interference model was inefficient, in the present of interference. In practice, the assumption of completely informative markers is unrealistic with data from human pedigrees. We report the results of a study investigating whether this conclusion still holds for gametes arising within pedigrees. We consider the same two mapping problems as Goldstein et al.: exclusion mapping and gene ordering. The results obtained were consistent with their findings, although the efficiency gains for analyses using the chi-square model were not as great in some cases. This is not unexpected with less than fully informative data. These results point to the need for research of developing new statistical and computational methods to incorporate interference into multipoint linkage mapping using pedigree data. This would make efficient use of available, but sometimes scarce data, especially in disease gene mapping.

Statistical analysis of half-tetrads

Half-tetrads, where two meiotic products from a single meiosis are recovered together, arise in different forms in a variety of organisms. Closely related to ordered tetrads, half-tetrads yield information on chromatid interference, chiasma interference, and centromere positions. In this article, for different half-tetrad types and different marker configurations, we derive the relations between multilocus half-tetrad probabilities and multilocus ordered tetrad probabilities. These relations are used to obtain equality and inequality constraints among multilocus half-tetrad probabilities that are imposed by the assumption of no chromatid interference. We illustrate how to apply these results to study chiasma interference and to map centromeres using multilocus half-tetrad data.

Statistical analysis of ordered tetrads

Ordered tetrad data yield information on chromatid interference, chiasma interference, and centromere locations. In this article, we show that the assumption of no chromatid interference imposes certain constraints on multilocus ordered tetrad probabilities. Assuming no chromatid interference, these constraints can be used to order markers under general chiasma processes. We also derive multilocus tetrad probabilities under a class of chiasma interference models, the chi-square models. Finally, we compare centromere map functions under the chi-square models with map functions proposed in the literature. Results in this article can be applied to order genetic markers and map centromeres using multilocus ordered tetrad data.

A three-wavelength labeling approach for DNA sequencing using energy transfer primers and capillary electrophoresis

Capillary electrophoresis DNA sequencing has been accomplished by using four different energy transfer primers and three fluorescence detection channels. Methods have also been developed to deconvolve the three-color data into the four base concentrations. The nonnegative least squares and model selection method resulted in the best accuracy. The three-color data were compared to sequencing data obtained using four detection channels and four energy transfer primers. The average accuracy rates obtained over three 500 base M13mp18 runs using three-color coding were 96% including 18 uncallable compressions and 99.6% if these compressions are excluded. The average accuracy rate obtained using four-color coding was 96.3% including 18 uncallable compressions and 99.9% if these compressions are excluded. Although it is unlikely that three-color schemes will replace four-color sequencing, these methods have exposed basic concepts that will be useful in the development of higher-order multiplex coding methods for DNA analysis.

An algorithm for haplotype analysis

This paper proposes an algorithm for haplotype analysis based on a Monte Carlo method. Haplotype configurations are generated according to the distribution of joint haplotypes of individuals in a pedigree given their phenotype data, via a Markov chain Monte Carlo algorithm. The haplotype configuration which maximizes this conditional probability distribution can thus be estimated. In addition, the set of haplotype configurations with relatively high probabilities can also be estimated as possible alternatives to the most probable one. This flexibility enables geneticists to choose the haplotype configurations which are most reasonable to them, allowing them to include their knowledge of the data under analysis.

The effects of genotyping errors and interference on estimation of genetic distance

Analysis of linkage data has typically been carried out assuming genotyping errors are absent. Recent studies have shown, however, that the impact of ignoring genotyping errors can be great, especially in dense marker maps [Buetow, Am J Hum Genet 1991; 49:985-994; Lincoln and Lander, Genomics 1992; 14:604-610]. Because most organisms exhibit positive chiasma interference, we use the chi 2 model [Foss et al., Genetics 1993; 144:681-691] to examine the role interference plays in the estimation of genetic distance in the presence of genotyping errors. For simplicity, we confine our analyses to samples of 1,000 fully informative gametes. Our results support previous findings that ignoring errors inflates distance estimates. The larger the error rate, the greater the inflation. For a given error rate, the relative error in estimated genetic distance is greatest when interference is known to be weak or absent. An approximation to relative error which quantifies the relation to distance, error rate, and interference is provided. Robustness of estimation to error misspecification is also investigated. When the assumed error rate is too low, distance is overestimated while interference is underestimated. The situation is reversed when too large an error rate is assumed (interference is overestimated, and distance underestimated). Unfortunately, the joint estimation of distance and interference is not very robust to error misspecification.

The limits of random fingerprinting

Various random fingerprinting methods are sometimes used to detect overlap between pairs of clones as a first step toward producing a minimal tiling path of clones for subsequent mapping and sequencing efforts. This paper evaluates and compares various statistical procedures for detecting pairwise overlap between clones when the fingerprints arise from any random process meeting simple, plausible assumptions about the relationship between overlap and the resulting fingerprint. Examples of such random processes include, but are not limited to, large-scale hybridization procedures designed to prepare tiling paths of clones for subsequent large-scale genomic sequencing. Our goals are to assess how well random fingerprinting can possibly detect overlap, to assess the effects of inevitable fingerprinting errors on statistical detection, to determine how one can make the best use of the data random fingerprinting provides, and to evaluate how well simple, heuristic techniques for overlap detection compare to more complex, likelihood-based approaches. The paper provides a quantitative assessment of the ability of any random fingerprinting procedure to detect various proportions of clonal overlap and shows the extent to which a small amount of experimental error will vitiate the performance of such techniques. The paper outlines a simple approximation method for constructing Bayesian overlap detectors, while concluding that detectors constructed from linear combinations of fingerprint data can be designed that will perform nearly as well as more complex, likelihood-based approaches.

Incorporating crossover interference into pedigree analysis using the chi 2 model

Crossover interference is now known to exist in humans but to date has been ignored in routine genetic mapping because of the computational burden involved. In a recent paper by Weeks et al. [Hum Hered 1993;43:86-97], interference was accounted for by the use of a variety of multilocus feasible map functions and a crossover model of Goldgar and Fain [Am J Hum Genet 1988;43:38-45]. In this paper, we present an alternative approach to incorporating crossover interference into multilocus likelihood computation, by modelling the underlying chiasma process directly using the chi 2 model, supplemented by an assumption of no chromatid interference. This procedure was applied to the same CEPH consortium chromosome 10 data set that was analyzed by Weeks et al. A fit to the data was achieved which was significantly better than that offered by the no-interference model, and comparable to the best of the alternatives considered by Weeks et al. We briefly discuss the relative merits of the different models for interference.

A note on the combination of estimates of a recombination fraction

A number of ways of combining two or more independent estimates of the same recombination fraction can be found in the literature. We revisit this topic in the context of human gene mapping, and explore the value of transforming the recombination fraction to a new parameter whose log-likelihood function is more nearly quadratic. It is shown that the arcsine of the cube-root is one such function. These observations lead naturally to a way of summarizing and combining the summarized set of log-likelihood functions of a common recombination fraction. This idea is illustrated using pedigree data concerning six loci on chromosome 10 from the CEPH consortium. A comparison is also made with the method of summarizing and combining using 'equivalent numbers' of recombinants and informative meioses.

On genetic map functions

Various genetic map functions have been proposed to infer the unobservable genetic distance between two loci from the observable recombination fraction between them. Some map functions were found to fit data better than others. When there are more than three markers, multilocus recombination probabilities cannot be uniquely determined by the defining property of map functions, and different methods have been proposed to permit the use of map functions to analyze multilocus data. If for a given map function, there is a probability model for recombination that can give rise to it, then joint recombination probabilities can be deduced from this model. This provides another way to use map functions in multilocus analysis. In this paper we show that stationary renewal processes give rise to most of the map functions in the literature. Furthermore, we show that the interevent distributions of these renewal processes can all be approximated quite well by gamma distributions.

Over- and underrepresentation of short DNA words in herpesvirus genomes

The relative abundance and rarity of DNA words have been recognized in previous biological studies to have implications for the regulation, repair, and evolutionary mechanisms of a genome. In this paper, we review several different measures of abundance and rarity of DNA words, including z-scores, representation ratios, and cross-ratios, that have appeared in the recent literature, and examine the concordance among them using the human cytomegalovirus genome sequence. We then rank all words of length k = 2, ..., 5 of seven herpesvirus genomes according to their abundance, as measured by one of the z-scores based upon a stationary Markov model of order k-2. Using a simple metric on the ranks of 2-words of the seven herpesvirus sequences, we construct an evolutionary tree. Several 3-words are observed to be consistently over- or underrepresented in all seven herpesviruses. Furthermore, clusters of some of the most over- and underrepresented 4- and 5-words in the genomes are identified with functional sites such as the origins of replication and regulatory signals of individual viruses.

Alveolar lining layer is thin and continuous: low-temperature scanning electron microscopy of rat lung

The low-temperature electron microscope, which preserves aqueous structures as solid water at liquid nitrogen temperature, was used to image the alveolar lining layer, including surfactant and its aqueous subphase, of air-filled lungs frozen in anesthetized rats at 15-cmH2O transpulmonary pressure. Lining layer thickness was measured on cross fractures of walls of the outermost subpleural alveoli that could be solidified with metal mirror cryofixation at rates sufficient to limit ice crystal growth to 10 nm and prevent appreciable water movement. The thickness of the liquid layer averaged 0.14 micron over relatively flat portions of the alveolar walls, 0.89 micron at the alveolar wall junctions, and 0.09 micron over the protruding features (9 rats, 20 walls, 16 junctions, and 146 areas), for an area-weighted average thickness of 0.2 micron. The alveolar lining layer appears continuous, submerging epithelial cell microvilli and intercellular junctional ridges; varies from a few nanometers to several micrometers in thickness, and serves to smooth the alveolar air-liquid interface in lungs inflated to zone 1 or 2 conditions.

Tests of random mating for a highly polymorphic locus: application to HLA data

Testing for random mating at an HLA locus is a difficult problem because of the highly polymorphic nature of the HLA loci. We discuss some methodological issues and propose several tests. A simulation study is conducted to evaluate these tests. The single allele test and the shared allele test deal with small sample sizes by aggregating the data in different ways. The shared allele test is found to be a more powerful method of detecting non-random mating patterns involving a deficiency or an excess of similar genotypes than the single allele test. We show that random mating of couple at the genotype level implies the random mating of couple at the allele level. Several multi-allele approaches are proposed for large population-based data sets. Among them, the corrected allele-table test performs better than the generalized Wald test in terms of power and size. These methods are then applied to an HLA data set of Caucasian couples, and no solid evidence for non-random mating at the HLA A, B, and DR loci is found.

Reproductive failure and the major histocompatibility complex

The association between HLA sharing and recurrent spontaneous abortion (RSA) was tested in 123 couples and the association between HLA sharing, and the outcome of treatment for unexplained infertility by in vitro fertilization (IVF) was tested in 76 couples, by using a new shared-allele test in order to identify more precisely the region of the major histocompatibility complex (MHC) influencing these reproductive defects. The shared-allele test circumvents the problem of rare alleles at HLA loci and at the same time provides a substantial gain in power over the simple chi 2 test. Two statistical methods, a corrected homogeneity test and a bootstrap approach, were developed to compare the allele frequencies at each of the HLA-A, HLA-B, HLA-DR, and HLA-DQ loci; they were not statistically different among the three patient groups and the control group. There was a significant excess of HLA-DR sharing in couples with RSA and a significant excess of HLA-DQ sharing in couples with unexplained infertility who failed treatment by IVF. These findings indicate that genes located in different parts of the class II region of the MHC affect different aspects of reproduction and strongly suggest that the sharing of HLA antigens per se is not the mechanism involved in the reproductive defects. The segment of the MHC that has genes affecting reproduction also has genes associated with different autoimmune diseases, and this juxtaposition may explain the association between reproductive defects and autoimmune diseases.

Relative efficiencies of chi 2 models of recombination for exclusion mapping and gene ordering

In multilocus linkage analysis, it is common to assume chiasma interference is absent. While this assumption provides mathematical tractability, there is substantial biological evidence contradicting it, particularly when the loci are closely spaced. The chi 2 class of recombination models, recently described by Foss et al. (Genetics 133: 681-691, 1993), has a plausible biological basis and provides a dramatically improved fit over virtually all other models currently in use. Here, a simulation study is performed to assess the relative efficiency of a no interference model analysis to an analysis with a chi 2 model which allows for interference. The results presented show that analysis with the no interference model is inefficient in the presence of interference.

Modeling interference in genetic recombination

In analyzing genetic linkage data it is common to assume that the locations of crossovers along a chromosome follow a Poisson process, whereas it has long been known that this assumption does not fit the data. In many organisms it appears that the presence of a crossover inhibits the formation of another nearby, a phenomenon known as "interference." We discuss several point process models for recombination that incorporate position interference but assume no chromatid interference. Using stochastic simulation, we are able to fit the models to a multilocus Drosophila dataset by the method of maximum likelihood. We find that some biologically inspired point process models incorporating one or two additional parameters provide a dramatically better fit to the data than the usual "no-interference" Poisson model.

Statistical analysis of chromatid interference

The nonrandom occurrence of crossovers along a single strand during meiosis can be caused by either chromatid interference, crossover interference or both. Although crossover interference has been consistently observed in almost all organisms since the time of the first linkage studies, chromatid interference has not been as thoroughly discussed in the literature, and the evidence provided for it is inconsistent. In this paper with virtually no restrictions on the nature of crossover interference, we describe the constraints that follow from the assumption of no chromatid interference for single spore data. These constraints are necessary consequences of the assumption of no chromatid interference, but their satisfaction is not sufficient to guarantee no chromatid interference. Models can be constructed in which chromatid interference clearly exists but is not detectable with single spore data. We then extend our analysis to cover tetrad data, which permits more powerful tests of no chromatid interference. We note that the traditional test of no chromatid interference based on tetrad data does not make full use of the information provided by the data, and we offer a statistical procedure for testing the no chromatid interference constraints that does make full use of the data. The procedure is then applied to data from several organisms. Although no strong evidence of chromatid interference is found, we do observe an excess of two-strand double recombinations, i.e., negative chromatid interference.