Population structure in the Mediterranean basin: a Y chromosome perspective.

The Mediterranean region has been characterised by a number of pre-historical and historical demographic events whose legacy on the current genetic landscape is still a matter of debate. In order to investigate the degree of population structure across the Mediterranean, we have investigated Y chromosome variation in a large dataset of Mediterranean populations, 11 of which are first described here. Our analyses identify four main clusters in the Mediterranean that can be labelled as North Africa, Arab, Central-East and West Mediterranean. In particular, Near Eastern samples tend to separate according to the presence of Arab Y chromosome lineages, suggesting that the Arab expansion played a major role in shaping the current genetic structuring within the Fertile Crescent.

Y chromosomal haplogroup J as a signature of the post-neolithic colonization of Europe.

In order to attain a finer reconstruction of the peopling of southern and central-eastern Europe from the Levant, we determined the frequencies of eight lineages internal to the Y chromosomal haplogroup J, defined by biallelic markers, in 22 population samples obtained with a fine-grained sampling scheme. Our results partially resolve a major multifurcation of lineages within the haplogroup. Analyses of molecular variance show that the area covered by haplogroup J dispersal is characterized by a significant degree of molecular radiation for unique event polymorphisms within the haplogroup, with a higher incidence of the most derived sub-haplogroups on the northern Mediterranean coast, from Turkey westward; here, J diversity is not simply a subset of that present in the area in which this haplogroup first originated. Dating estimates, based on simple tandem repeat loci (STR) diversity within each lineage, confirmed the presence of a major population structuring at the time of spread of haplogroup J in Europe and a punctuation in the peopling of this continent in the post-Neolithic, compatible with the expansion of the Greek world. We also present here, for the first time, a novel method for comparative dating of lineages, free of assumptions of STR mutation rates.

Clinal patterns of human Y chromosomal diversity in continental Italy and Greece are dominated by drift and founder effects.

We explored the spatial distribution of human Y chromosomal diversity on a microgeographic scale, by typing 30 population samples from closely spaced locations in Italy and Greece for 9 haplogroups and their internal microsatellite variation. We confirm a significant difference in the composition of the Y chromosomal gene pools of the two countries. However, within each country, heterogeneity is not organized along the lines of clinal variation deduced from studies on larger spatial scales. Microsatellite data indicate that local increases of haplogroup frequencies can be often explained by a limited number of founders. We conclude that local founder or drift effects are the main determinants in shaping the microgeographic Y chromosomal diversity.

A multistep process for the dispersal of a Y chromosomal lineage in the Mediterranean area.

In this work we focus on a microsatellite-defined Y-chromosomal lineage (network 1.2) identified by us and reported in previous studies, whose geographic distribution and antiquity appear to be compatible with the Neolithic spread of farmers. Here, we set network 1.2 in the Y-chromosomal phylogenetic tree, date it with respect to other lineages associated with the same movements by other authors, examine its diversity by means of tri- and tetranucleotide loci and discuss the implications in reconstructing the spread of this group of chromosomes in the Mediterranean area. Our results define a tripartite phylogeny within HG 9 (Rosser et al. 2000), with the deepest branching defined by alleles T (Haplogroup Eu10) or G (Haplogroup Eu9) at M172 (Semino et al. 2000), and a subsequent branching within Eu9 defined by network 1.2. Population distributions of HG 9 and network 1.2 show that their occurrence in the surveyed area is not due to the spread of people from a single parental population but, rather, to a process punctuated by at least two phases. Our data identify the wide area of the Balkans, Aegean and Anatolia as the possible homeland harbouring the largest variation within network 1.2. The use of recently proposed tests based on the stepwise mutation model suggests that its spread was associated to a population expansion, with a high rate of male gene flow in the Turkish-Greek area.

Y chromosome analysis reveals a sharp genetic boundary in the Carpathian region.

Nine single nucleotide (SNP) or indel binary polymorphisms were used to determine the frequencies and phylogenetic relationships of 12 Y chromosomal haplogroups in 289 males from Romania and the Republic of Moldova. Our data indicated a low but not null rate of the homoplasic appearance of the DYZ3 (-) allelic state. All other markers confirmed the previously proposed phylogeny. Based on the affinities between populations in terms of haplogroup frequencies, this work identified the geographical region of the Carpathians as a break point in the gene geography of Eastern Central Europe, providing a finer definition of one of the possible sharp genetic changes between Western and Eastern Europe.

Patterns of male-specific inter-population divergence in Europe, West Asia and North Africa.

We typed 1801 males from 55 locations for the Y-specific binary markers YAP, DYZ3, SRY10831 and the (CA)n microsatellites YCAII and DYS413. Phylogenetic relationships of chromosomes with the same binary haplotype were condensed in seven large one-step networks, which accounted for 95% of all chromosomes. Their coalescence ages were estimated based on microsatellite diversity. The three largest and oldest networks undergo sharp frequency changes in three areas. The more recent network 3.1A clearly discriminates between Western and Eastern European populations. Pairwise Fst showed an overall increment with increasing geographic distance but with a slope greatly reduced when compared to previous reports. By sectioning the entire data set according to geographic and linguistic criteria, we found higher Fst-on-distance slopes within Europe than in West Asia or across the two continents.

Network analyses of Y-chromosomal types in Europe, northern Africa, and western Asia reveal specific patterns of geographic distribution.

In a study of 908 males from Europe, northern Africa, and western Asia, the variation of four Y-linked dinucleotide microsatellites was analyzed within three "frames" that are defined by mutations that are nonrecurrent, or nearly so. The rapid generation and extinction of new dinucleotide length variants causes the haplotypes within each lineage to diverge from one another. We constructed networks of "adjacent" haplotypes within each frame, by assuming changes of a single dinucleotide unit. Two small and six large networks were obtained, the latter including 94.9% of the sampled Y chromosomes. We show that the phenetic relationships among haplotypes, represented as a network, result largely from common descent and subsequent molecular radiation. The grouping of haplotypes of the same network thus fits an evolutionarily relevant criterion. Notably, this method allows the total diversity within a sample to be partitioned. Networks can be considered optimal markers for population studies, because reliable frequency estimates can be obtained in small samples. We present synthetic maps describing the incidence of different Y-chromosomal lineages in the extant human populations of the surveyed areas. Dinucleotide diversity also was used to infer time intervals for the coalescence of each network.

Differential structuring of human populations for homologous X and Y microsatellite loci.

The global pattern of variation at the homologous microsatellite loci DYS413 (Yq11) and DXS8174 and DXS8175 (Xp22) was analyzed by examination of 30 world populations from four continents, accounting for more than 1,100 chromosomes per locus. The data showed discordant patterns of among- and within-population gene diversity for the Y-linked and the X-linked microsatellites. For the Y-linked polymorphism, all groups of populations displayed high FST values (the correlation between random haplotypes within subpopulations, relative to haplotypes of the total population) and showed a general trend for the haplotypes to cluster in a population-specific way. This was especially true for sub-Saharan African populations. The data also indicated that a large fraction of the variation among populations was due to the accumulation of new variants associated with the radiation process. Europeans exhibited the highest level of within-population haplotype diversity, whereas sub-Saharan Africans showed the lowest. In contrast, data for the two X-linked polymorphisms were concordant in showing lower FST values, as compared with those for DYS413, but higher within-population variances, for African versus non-African populations. Whereas the results for the X-linked loci agreed with a model of greater antiquity for the African populations, those for DYS413 showed a confounding pattern that is apparently at odds with such a model. Possible factors involved in this differential structuring for homologous X and Y microsatellite polymorphisms are discussed.

Characterization of a small family (CAIII) of microsatellite-containing sequences with X-Y homology.

Four X-linked loci showing homology with a previously described Y-linked polymorphic locus (DYS413) were identified and characterized. By fluorescent in situ hybridization (FISH), somatic cell hybrids, and YAC screening, the X-linked members of this small family of sequences (CAIII) all map in Xp22, while the Y members map in Yq11. These loci contribute to the overall similarity of the two genomic regions. All of the CAIII loci contain an internal microsatellite of the (CA)n type. The microsatellites display extensive length polymorphism in two of the X-linked members as well as in the Y members. In addition, common sequence variants are found in the portions flanking the microsatellites in two of the X-linked members. Our results indicate that, during the evolution of this family, length variation on the Y chromosome was accumulated at a rate not slower than that on the X chromosome. Finally, these sequences represent a model system with which to analyze human populations for similar X- and Y-linked polymorphisms.

Population variation analysis at nine loci containing expressed trinucleotide repeats.

The polymorphisms of nine loci containing reiterated CAG repeats were examined in four populations from three continents. Their normal variation was analysed across populations or in subsets of loci grouped according to either the presence/absence of disease-associated expansions or CAG interruptions. A unifying feature of the allele distributions of all loci in all populations was the marked non-normality. Significantly larger numbers of alleles, average lengths, length ranges and variances in repeat number were observed in loci with vs. without known expansions. Significantly longer alleles were found at loci with vs. without interruption of the (CAG)n motif. The nine loci detected levels of inter-population variability comparable to other loci. Altogether the data are at odds with a model assuming that autosomal expressed trinucleotides accumulate variation exclusively by insertion/deletion of a single unit.

European Gene Mapping Project (EUROGEM): breakpoint panels for human chromosomes based on the CEPH reference families. Centre d'Etude du Polymorphisme Humain.

Meiotic breakpoint panels for human chromosomes 2, 3, 4, 5, 6, 7, 8, 9, 10, 13, 14, 15, 17, 18, 20 and X were constructed from genotypes from the CEPH reference families. Each recombinant chromosome included has a breakpoint well-supported with reference to defined quantitative criteria. The panels were constructed at both a low-resolution, useful for a first-pass localization, and high-resolution, for a more precise placement. The availability of such panels will reduce the number of genotyping experiments necessary to order new polymorphisms with respect to existing genetic markers. This paper shows only a representative sample of the breakpoints detected. The complete data are available on the World Wide Web (URL http:/(/)www.icnet.uk/axp/hgr/eurogem++ +/HTML/data.html) or by anonymous ftp (ftp.gene.ucl.ac.uk in/pub/eurogem/maps/breakpoints).

Genetic fitness in Huntington's Disease and Spinocerebellar Ataxia 1: a population genetics model for CAG repeat expansions.

An analysis of genetic fitness was performed in Huntington's Disease (HD) and Spinocerebellar Ataxia 1 (SCA1) families. Two partially overlapping samples were used: clinically defined HD and SCA1 patients from families ascertained in definite geographical areas, and molecularly typed carriers of HD and SCA1 mutations (CAG trinucleotide expansions). In both cases, a control group of normal relatives was used. HD and SCA1 patients born before 1915-20 had more children than normal controls. Carriers of HD and SCA1 mutations, all in the low/medium expansion range (37-49 and 47-54 CAG repeats respectively), had a higher number of children than controls up to more recent times (1935-1950). The reproduction of heterozygotes for large expansions could be analysed only in subjects born after 1950 and provided indirect evidence of a lower than normal number of children. The above results fit a model based on a differential fitness according to the degree of expansion. Such a model predicts that 1) up to relatively recently the frequency of alleles in the low/medium range has been maintained or even increased by the increased fitness of their carriers, as well as by new mutations, and 2) the frequency of large expansions, part of which are lost at each generation, is maintained through further expansions of alleles in the low/medium expansion range. The implications of such a model on linkage disequilibrium and the possible spread of these diseases in future generations are discussed.

Effect of trinucleotide repeat length and parental sex on phenotypic variation in spinocerebellar ataxia I.

Trinucleotide repeat expansion has been found in 64 subjects from 19 families: 57 patients with SCA1 and 7 subjects predicted, by haplotype analysis, to carry the mutation. Comparison with a large set of normal chromosomes shows two distinct distributions, with a much wider variation among expanded chromosomes. The sex of transmitting parent plays a major role in the size distribution of expanded alleles, those with > 54 repeats being transmitted by affected fathers exclusively. Our data suggest that alleles with > 54 repeats have a reduced chance of survival; these appear to be replaced in each generation by further expansion of alleles in the low- to medium-expanded repeat range, preferentially in male transmissions. Detailed clinical follow-up of a subset of our patients demonstrates significant relationships between increasing repeat number on expanded chromosomes and earlier age at onset, faster progression of the disease, and earlier age at death.

Autosomal dominant pure cerebellar ataxia. Neurological and genetic study.

A family with late-onset autosomal dominant pure cerebellar ataxia was studied both neurologically and genetically. Neuroimaging and electrophysiological results were in agreement with the clinical evidence showing involvement of the cerebellar system only, even many years after onset. No atrophy of inferior olives was observed by magnetic resonance imaging, while cerebellar atrophy was extremely marked. A very slow disease progression was observed in all patients. The disease can be differentiated from autosomal dominant olivo-ponto-cerebellar atrophies, and in particular from spinocerebellar ataxia type 1 mapping on chromosome 6p, which shows an early multisystemic involvement and a more rapid progression toward inability. A genetic study of the family with the 6p DNA marker D6S89 closely linked to the spinocerebellar ataxia type 1 locus was performed. Results showed significant exclusion of a linkage between the disease and the marker within a distance of 8.5% recombination, indicating that genetic heterogeneity underlies phenotypic differences.

Disequilibrium of multiple DNA markers on the human Y chromosome.

We characterized four DNA polymorphisms on the Y chromosomes of 123 males from five Caucasian populations. Three markers on the male specific portion of the chromosome varied appreciably in frequency among the populations. When combined, these markers define a limited number of haplotypes compared with the maximum expected on the basis of random association. The associations found in the five groups are qualitatively similar and are thus considered to be relatively stable on an evolutionary time-scale and possibly to predate the divergence of Caucasian populations. However, the haplotype frequencies varied markedly among populations, even between weakly isolated areas such as northern vs. southern Sardinia. This may indicate rapid progression towards fixation of alternative types of Y chromosomes. We also report data suggesting that the same associations no longer hold when examining a marker as close as 275 bp from the boundary of the pseudoautosomal region on the Y chromosome.

Individual variability in the translational regulation of ribosomal protein synthesis in Xenopus laevis.

Ribosomal protein synthesis is regulated by controlling the fraction of mRNA associated with polysomes. It is known that this value changes in different developmental stages during Xenopus embryogenesis or, more generally, with changing cell growth conditions. We present here an analysis of the proportion of mRNA loaded on polysomes, carried out with probes for five different ribosomal proteins on several batches of Xenopus embryos obtained from different individuals. The results obtained indicate the existence of probe-dependent and individual differences, which reflect genetic variations in the cis- and trans-acting regulatory elements responsible for translational regulation. The fraction of ribosomal protein mRNA loaded onto polysomes can be used as an index of an individual's capacity for ribosome production.

Natural selection associated with birth weight. VI. Towards the end of the stabilizing component.

The secular trend for stabilizing selection on birth weight has been analysed in Italy from 1954 to 1985 in order to study changes in the forces of natural selection which have occurred as a consequence of progress in health care. In previous papers we demonstrated a very rapid relaxation of stabilizing selection on birth weight. In this paper we show that in the last few years this kind of selection has been coming to an end for the vast majority of Italian newborns.