Clinal patterns of autosomal genetic diversity within Europe have been interpreted in previous studies in terms of a Neolithic demic diffusion model for the spread of agriculture; in contrast, studies using mtDNA have traced many founding lineages to the Paleolithic and have not shown strongly clinal variation. We have used 11 human Y-chromosomal biallelic polymorphisms, defining 10 haplogroups, to analyze a sample of 3,616 Y chromosomes belonging to 47 European and circum-European populations. Patterns of geographic differentiation are highly nonrandom, and, when they are assessed using spatial autocorrelation analysis, they show significant clines for five of six haplogroups analyzed. Clines for two haplogroups, representing 45% of the chromosomes, are continentwide and consistent with the demic diffusion hypothesis. Clines for three other haplogroups each have different foci and are more regionally restricted and are likely to reflect distinct population movements, including one from north of the Black Sea. Principal-components analysis suggests that populations are related primarily on the basis of geography, rather than on the basis of linguistic affinity. This is confirmed in Mantel tests, which show a strong and highly significant partial correlation between genetics and geography but a low, nonsignificant partial correlation between genetics and language. Genetic-barrier analysis also indicates the primacy of geography in the shaping of patterns of variation. These patterns retain a strong signal of expansion from the Near East but also suggest that the demographic history of Europe has been complex and influenced by other major population movements, as well as by linguistic and geographic heterogeneities and the effects of drift.

We present here a new interesting feature of the human tumor suppressor gene p53: a very pronounced ethnic and clinal variation of polymorphic codon 72 alleles. The frequency of the A1 (Pro) allele showed a north-south cline from 17% in Swedish Saamis to 63% in African Blacks (Nigerians), and there was a significant (p < 0.001) correlation (r = 0.95) between the A2 frequency and latitude. In the Finnish and Swedish populations no significant differences were found with respect to the genotype and allele distributions in spontaneously aborted fetuses and liveborn children, which makes differential intrauterine selection unlikely. However, the ethnic and clinal variations suggest that the codon 72 polymorphism is balanced and maintained by natural selection.

Three polymorphisms in the human tumor suppressor gene p53 (BstUI and MspI RFLPs in exon 4 and intron 6 respectively and a 16 bp duplication in intron 3) and their haplotype combinations were studied in patients with breast cancer and controls. A significant increase in the codon 72 BstUI A1 (pro) allele frequency (P = 0.016) and of individuals carrying the pro allele (pro/pro and pro/arg) (OR, 1.47; P = 0.01 4; 95 % CI, 1.08-2.00) was observed in breast cancer. This increase was most pronounced in highly differentiated breast cancer. Significant associations were found only in BstUI and haplotypes containing this polymorphism, which indicates that the codon 72 pro allele may be functionally involved in low malignancy breast cancer. The distributions of genotypic combinations in breast cancer patients and controls were significantly different (P = 0.005). Two BstUI-16 bp-MspI combinations were significantly overrepresented; 2-1, 1-1, 2-2 (OR, 1.61; 95% CI, 1.13-2.30) and 1-1, 2-1, 2-1 (OR, 2.94; 95% CI, 1.37-6.27).

An association between the BstU I 1-1 (Pro-Pro) genotype of the p53 codon 72 polymorphism and lung cancer has previously been reported by Kawajiri et al. A reanalysis of the data by Kawajiri et al. revealed no significant difference between patients and controls with respect to allele frequencies, and the increased frequency of BstU I 1-1 homozygotes was mostly ascribable to a deviation from the Hardy-Weinberg equilibrium. In an attempt to replicate the results by Kawajiri et al. we have studied three p53 polymorphisms (BstU I and Msp I RFLPs in exon 4 and intron 6 respectively and a 16 bp duplication in intron 3) and their haplotypes in Swedish lung cancer patients and controls. The results concerning the codon 72 polymorphism were largely negative. Thus there was no significant association between lung cancer and the BstU I 1-1 type, and only a marginal difference (P = 0.044) with respect to the BstU I allele frequency when lung cancer patients were compared with patients with chronic obstructive pulmonary disease (COPD). However, when the analysis was based on haplotype frequencies larger differences appeared and it was found that only BstU I 1 (pro) alleles linked to 16 bp 1 alleles were associated with lung cancer. Pro alleles linked to the 16 bp duplication appeared instead to confer some protection against cancer. Thus the codon 72 alleles need not be functionally involved in lung cancer, but may rather be markers in linkage disequilibrium with other cancer susceptibility sites on p53.

Three p53 DNA polymorphisms (BstU I and Msp I restriction fragment length polymorphisms (RFLPs) in exon 4 and intron 6 respectively, and a 16 bp duplication in intron 3) and their haplotype combinations were studied in patients with colorectal cancer and compared with patients with ulcerative colitis and healthy controls. There were only minor differences between patients with ulcerative colitis and controls, the only significant difference was observed in the distribution of BstU I-Msp I haplotypes. When single polymorphisms were studied, a significantly lower frequency of the 16 bp duplication was found in patients with colorectal cancer. The protective effect of the 16 bp duplication was more pronounced in haplotype combinations with the BstU I A1 and Msp I A1 alleles, whereas these alleles in combination with the 16 bp A1 allele (no duplication) were associated with an increased risk for colorectal cancer. The genotypic combination BstU I 2-1, 16 bp 1-I, Msp I 2-1 was found in 8.4% of cases among patients with colorectal cancer and 0.5% of cases in the controls (odds ratio = 18.8). The extended haplotype responsible for the high cancer risk of this genotype appears to be BstU I A1-16 bp A1-Msp I A1. The results of this study indicate that the haplotype approach to the identification of p53 germ line alleles associated with increased susceptibility to cancer is far more powerful than the analysis of single polymorphisms, since the capacity to identify germ line alleles predisposing to cancer should increase with the number of polymorphic sites included in the analysis.