The molecular dissection of mtDNA haplogroup H confirms that the Franco-Cantabrian glacial refuge was a major source for the European gene pool

Abstract

Complete sequencing of 62 mitochondrial DNAs (mtDNAs) belonging (or very closely related) to haplogroup H revealed that this mtDNA haplogroup--by far the most common in Europe--is subdivided into numerous subhaplogroups, with at least 15 of them (H1-H15) identifiable by characteristic mutations. All the haplogroup H mtDNAs found in 5,743 subjects from 43 populations were then screened for diagnostic markers of subhaplogroups H1 and H3. This survey showed that both subhaplogroups display frequency peaks, centered in Iberia and surrounding areas, with distributions declining toward the northeast and southeast--a pattern extremely similar to that previously reported for mtDNA haplogroup V. Furthermore, the coalescence ages of H1 and H3 (~11,000 years) are close to that previously reported for V. These findings have major implications for the origin of Europeans, since they attest that the Franco-Cantabrian refuge area was indeed the source of late-glacial expansions of hunter-gatherers that repopulated much of Central and Northern Europe from ~15,000 years ago. This has also some implications for disease studies. For instance, the high occurrence of H1 and H3 in Iberia led us to re-evaluate the haplogroup distribution in 50 Spanish families affected by nonsyndromic sensorineural deafness due to the A1555G mutation. The survey revealed that the previously reported excess of H among these families is caused entirely by H3 and is due to a major, probably nonrecent, founder event.

Figure 1

Most-parsimonious tree of complete (pre-HV)1, HV*, HV1, pre-V, and H mtDNA sequences. The tree, rooted in haplogroup R, includes 62 mtDNAs (1–62) sequenced in this study and illustrates subhaplogroup affiliations. Phylogeny construction was performed by hand, following a parsimony approach, and was confirmed by use of the program Network 4.0. We have applied the reduced median algorithm (r=2) (Bandelt et al. 1995), followed by the median-joining algorithm (ɛ=0) (Bandelt et al. 1999) and the MP (maximum parsimony) calculation option, as explained at the Fluxus Engineering Web site. For the phylogeny construction, half weight was assigned to the control-region positions, and the pathological mutations 1555, 3460, and 13513 (shown in italics) were excluded. Mutations are shown on the branches; they are transitions, unless the base change is explicitly indicated. Deletions are indicated by a “d” preceding the deleted nucleotides. Insertions are indicated by a “+” preceding the inserted nucleotide(s). Heteroplasmy is indicated by an “h” following the nucleotide position. Underlining indicates recurrent mutations, whereas mutations in boldface are diagnostic of the haplogroup/subhaplogroup. The asterisk (*) indicates the most recent common ancestor of the H mtDNAs. This differs from the reference sequence (Andrews et al. 1999), which belongs to H2, by mutations at the following positions: 263, 315+C, 750, 1438, 4769, 8860, and 15326. The variation in number of Cs at np 309 was not included in the phylogeny (mtDNAs 1, 4, 8, 12, 15, 24, 26, 29, 31–32, 34–35, 37, 39, 41, 45–48, 56, and 58–62 harbored 309+C, whereas the mtDNAs 2–3, 5–7, 14, 17–18, 20, 42, 53, and 57 harbored 309+CC). mtDNAs 2, 4–8, 10–21, 23, 25–31, 33, 35–37, 39–44, 46–47, 49, 51, and 54–62 are from Italian subjects; 9, 24, 45, 48, 50, and 52 are from Spanish subjects; 32 and 53 are from Georgian subjects; 22 and 38 are from Iraqi subjects; 1 is from a Sindhi (Pakistan); 3 is from a Druze subject; and 34 is from a Berber of Egypt.

Figure 2

Geographical locations of populations surveyed for haplogroup H (top) and its spatial frequency distribution (bottom). Frequency values for populations 1–43 are from table 1, whereas those for populations 44–63 are from the literature, as follows: 44–46 from Helgason et al. ; 47–49 and 53 from Richards et al. ; 50 from Richards et al. and Passarino et al. ; 51 from Finnilä et al. ; 52 from Torroni et al. and Richards et al. ; 54 from Baasner et al. , Lutz et al. , Pfeiffer et al. and , and Richards et al. ; 55–56 from Malyarchuk et al. ; and 57–63 from Quintana-Murci et al. . The frequency map was obtained using Surfer version 6.04 (Golden Software), with the Kriging procedure, and estimates at each grid node were inferred by consideration of the entire data set.

Figure 3

Spatial frequency distributions of subhaplogroups H1 and H3. Frequency values are from table 1. Maps were obtained as in figure 2.