Accuracy of evolutionary analysis of populations within a species requires the testing of a large number of genetic polymorphisms belonging to many loci. We report here a reconstruction of human differentiation based on 100 DNA polymorphisms tested in five populations from four continents. The results agree with earlier conclusions based on other classes of genetic markers but reveal that Europeans do not fit a simple model of independently evolving populations with equal evolutionary rates. Evolutionary models involving early admixture are compatible with the data. Taking one such model into account, we examined through simulation whether random genetic drift alone might explain the variation among gene frequencies across populations and genes. A measure of variation among populations was calculated for each polymorphism, and its distribution for the 100 polymorphisms was compared with that expected for a drift-only hypothesis. At least two-thirds of the polymorphisms appear to be selectively neutral, but there are significant deviations at the two ends of the observed distribution of the measure of variation: a slight excess of polymorphisms with low variation and a greater excess with high variation. This indicates that a few DNA polymorphisms are affected by natural selection, rarely heterotic, and more often disruptive, while most are selectively neutral. This paper presents results of the first phase of our study of human evolution based on nuclear DNA variation in five populations studied for a selected set of 100 DNA polymorphisms. At this stage we have chosen to focus on a large number of markers typed on each of a small number of defined populations because the reliability of conclusions is so dependent on the number of independent markers (1-3). Using these data (presented in detail in ref. 38), we examine the shape of the tree and compare it with earlier trees, investigate an observed anomaly of evolutionary rates, and study variation of different genes and its bearing on the theory of neutral evolution.Prior studies of human populations using DNA markers have been limited almost entirely to mitochondrial DNA (4,

In population studies on aging, the data on genetic markers are often collected for individuals from different age groups. The purpose of such studies is to identify, by comparison of the frequencies of selected genotypes, "longevity" or "frailty" genes in the oldest and in younger groups of individuals. To address questions about more-complicated aspects of genetic influence on longevity, additional information must be used. In this article, we show that the use of demographic information, together with data on genetic markers, allows us to calculate hazard rates, relative risks, and survival functions for respective genes or genotypes. New methods of combining genetic and demographic information are discussed. These methods are tested on simulated data and then are applied to the analysis of data on genetic markers for two haplogroups of human mtDNA. The approaches suggested in this article provide a powerful tool for analyzing the influence of candidate genes on longevity and survival. We also show how factors such as changes in the initial frequencies of candidate genes in subsequent cohorts, or secular trends in cohort mortality, may influence the results of an analysis.

The possibility that four loci (REN, THO, PARP, SOD2) are associated with longevity was explored by comparing the genotypic pools of subjects older than 100 years with those of younger subjects matched for sex and geographic area (northern and southern Italy). The markers (all located within the respective gene) were HUMREN4; HUMTHO1; HUMPARP (gt)845nt; SOD2(C/T)401nt. In order to reduce the number of genotypes, multiallelic polymorphisms were recoded as diallelic according to allele size and frequency patterns (small: S, and large: L, alleles). A significant loss of LL homozygous genotypes was found at the THO locus in male but not in female centenarians with respect to matched controls. On the other hand no significant difference was found between case/control genotypic frequencies at REN, PARP, SOD2 loci. The latter loci therefore do not affect inter-individual variability in life expectancy (at least in terms of qualitative variants associated with the tested markers). However, the data is consistent with an association between the THO locus and longevity.

In population studies of aging, the data on genetic markers are often collected for individuals from different age groups. The idea of such studies is to identify "longevity" or "frailty" genes by comparing the frequencies of genotypes in the oldest and in the younger groups of individuals. In this paper we discuss a new approach to the analysis of such data. This approach, based on the maximum likelihood method, combines data on genetic markers with survival information obtained from standard demographic life tables. This method allows us to evaluate survival characteristics for individuals carrying respective candidate genes. It can also be used in the estimation of the effects of allele-area and allele-allele interaction, either in the presence or absence of hidden heterogeneity. We apply this method to the analysis of Italian data on genetic markers for five autosomal loci and mitochondrial genomes. Then we discuss basic assumptions used in this analysis and directions of further research.

SummaryThe genes coding for apolipoprotein A1 (APOA1 ), apolipoprotein C3 (APOC3 ) and apolipoprotein A4 (APOA4 ) are tandemly organised within a short region on chromosome 11q23-q24. Polymorphisms of these genes have been extensively investigated in lipoprotein disorders and cardiovascular diseases, but poorly investigated in healthy ageing. The aim of this study was to describe possible modifications of the APOA1, APOC3, and APOA4 gene pool by cross-sectional studies carried out in a healthy ageing population whose ages ranged from 18 to 109 years (800 subjects, 327 males and 473 females, free of clinically manifested disease, and with emato-chemical parameters in the norm). APOA1-MspI-RFLP (−75 nt from the transcription starting site), APOC3-SstI-RFLP (3 UTR, 3238 nt), and APOA4-HincII-RFLP (Asp 127 /Ser 127 ) were analysed according to age and sex. A significant agerelated variation of the APOA1 gene pool was observed in males. An analysis of the allele average effect exerted by APOA1-MspI-RFLP A/P alleles (Absence/Presence of the restriction site) on lipidemic parameters in 46-80 year old males showed that allele A decreased, while allele P significantly increased, serum LDL-cholesterol. Unexpectedly, the P allele was over-represented in the group of the oldest old subjects, thus giving evidence of another "genetic paradox of centenarians".