Introduction

The ABO blood group is one of the classical genetic markers that are useful for studying the genetic relationships between populations. In the ABO blood group system discovered by Landsteiner,1 determinants of A and B antigens are A and B transferases that are encoded by A and B alleles, respectively, at the single ABO blood group locus on human chromosome 9,2 whereas the product of the O allele has neither A nor B transferase activities.3

Among various local populations in Japan and neighboring regions, different frequencies of the ABO blood group have been reported by many studies.4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 In these studies, however, allele frequencies of the ABO blood group gene were inferred by the maximum-likelihood method on the basis of immunological data, because the molecular genetic basis of ABO blood types was unknown at that time.

Recent studies have revealed that the ABO blood group gene possesses a high degree of polymorphisms at the molecular level19 and that allele frequencies of the ABO gene are different among various populations.20, 21, 22, 23, 24, 25, 26 Therefore, allele frequencies and distributions of the ABO gene can be appropriate indices for investigating the genetic relationships between populations. In this study, to further understand the history and genetic features of human populations in Hokkaido, northern Japan, polymorphisms of the ABO gene in the Jomon/Epi-Jomon and Okhotsk people who lived in Hokkaido were investigated.

On the Hokkaido island, cultures different from those of Honshu (mainland of the Japanese archipelago) have developed after the Jomon period (about 12 000–2400 years BP).27 The Yayoi culture with the rice culturing was not introduced to Hokkaido, and the Epi-Jomon (third century BC to seventh century AD), Satsumon (eighth to thirteenth centuries) and Ainu cultures (since the fourteenth century) developed in Hokkaido.27 In addition, the Okhotsk culture was developed around coastal regions of the Okhotsk Sea from the Epi-Jomon period to the Satsumon period.28

Recent analyses of ancient DNA revealed that genetic features of the Jomon/Epi-Jomon people of Hokkaido were considerably different from those of the Okhotsk people. Adachi et al.29, 30 reported that mitochondrial DNA (mtDNA) haplogroups N9b and D1 were major in the Jomon people of Hokkaido and that they had a close genetic affinity to the Udegey of the continental Far East. By contrast, Sato et al.31, 32 demonstrated that mtDNA haplogroups Y, G1b and N9b were major in the Okhotsk people and that they were genetically closely related to modern local populations in Sakhalin and lower regions of the Amur River, including the Ulchi, Nivkhi and Negidal. Moreover, Sato et al.33 revealed that allele frequencies of the ABCC11 gene, which is an autosomal gene associated with earwax phenotypes (wet/dry), were significantly different between the Jomon/Epi-Jomon and Okhotsk people. The Jomon people had a relatively higher frequency of the wet allele, showing genetic characteristics similar to those of the Ainu. On the other hand, the frequency of the wet allele in the Okhotsk people is relatively low, indicating genetic features similar to modern populations of northeastern Asia.

In this study, we examined polymorphisms of the ABO blood group gene using an amplified product-length polymorphism (APLP) analysis of ancient DNA to further understand genetic features of the Jomon/Epi-Jomon people and the Okhotsk people in Hokkaido. On the basis of comparisons of the allele frequencies among ancient and modern populations, we discuss the genetic features and history of ancient populations of Hokkaido.

Materials and methods

Sample collection

To determine genotypes of the ABO blood group gene in the ancient populations in Hokkaido, 81 skeletal remains of the Jomon and Epi-Jomon people and 50 skeletal remains of the Okhotsk people, excavated from 26 archeological sites in Hokkaido and Sakhalin (Figure 1), were analyzed. Each site was classified into Jomon, Epi-Jomon or Okhotsk culture by using archeological analyses based on earthenware, stone implements, traces of houses and strata. The skeletal remains were preserved at the Hokkaido University Museum, Sapporo Medical University and Tohoku University School of Medicine. To avoid a duplicate analysis of skeletal remains from single individuals, parts in the same positions of bones or bones from different graves within one archeological site were sampled.

Figure 1
figure 1

Geographical locations of archaeological sites of the Jomon/Epi-Jomon culture (open circles) and Okhotsk culture (closed circles) from which specimens analyzed in this study were excavated.

Full size image

Contamination precautions

The following standard contamination precautions were taken: separation of pre- and post-PCR experimental areas, use of gloves, face masks and laboratory coats, use of disposable filter-plugged pipette tips and disposable tubes, treatment with DNA-AWAY (Molecular BioProducts, San Diego, CA, USA), ultraviolet irradiation of equipments and bench, negative extraction controls and negative PCR controls.

DNA extraction

Total DNA was extracted from femurs, ribs, coxal bones, skulls, sacrums or teeth of ancient skeletal remains sampled, according to the following methods.31, 32, 33, 34 To avoid surface contamination of external DNA, each bone or tooth was soaked in sodium hypochlorite solution (Nacalai Tesque, Kyoto, Japan) for 5 min, rinsed with DNase-/RNase-free distilled water and air-dried. Samples were then powdered with a standard dental drill. Approximately 0.2–0.5 g of bone powder per specimen was decalcified with 30 ml of 0.5 M ethylenediaminetetraacetic acid in a 50-ml plastic tube, with rotating at room temperature for 24 h. Decalcified pellets were suspended in 5 ml of 0.5 M ethylenediaminetetraacetic acid containing 100 μl of proteinase K concentrated at 10 mg ml−1 and incubated overnight at 37 °C with rotation. The solution was extracted using the phenol–chloroform extraction method (phenol–chloroform–isoamyl alcohol 25:24:1).35 The extracts were concentrated into approximately 100 μl of TE buffer using VivaSpin 6 Concentrators (Sartorius Stedim Biotech, Goettingen, Germany) and used as a template for subsequent PCR.

APLP and PCR-restriction fragment length polymorphism analyses

To analyze polymorphisms of the ABO blood group gene of ancient specimens, APLP analysis36 was conducted. For high accuracy of PCR amplification, APLP analysis was separated into three PCR reactions (set A using primers 1–4, set B using primers 5–7 and set C using primers 8–10). The PCR amplification was carried out in a reaction mixture of 20 μl containing reagents of the Multiplex PCR Kit (Qiagen, Hilden, Germany), optimum concentrations of each primer36 and 0.4 μg μl−1 of bovine serum albumin (20 mg ml−1; Roche, Basel, Switzerland). The PCR conditions were 95 °C for 15 min, followed by 40 cycles of 94 °C for 30 s, 54 °C for 3 min, 72 °C for 90 s and the final extension at 72 °C for 10 min.

An aliquot of 10 μl of the PCR product was separated by electrophoresis in a 13 cm native polyacrylamide gel (10% T, 5% C, T=acrylamide+methylenebisacrylamide/acrylamide+methylenebisarcylamide+water, C=methylenebisacrylamide/ acrylamide+methylenebisacrylamide) containing 375 mM of Tris-NaOH buffer (pH 8.9) with running buffer (12.5 mM Tris, 96 mM glycine; pH 8.3). The DNA bands were detected with an ultraviolet illuminator after staining with ethidium bromide.

A new allele that has not been reported in the previous studies22, 23, 25, 36, 37 might be generated by the recombination. In this study, however, we surmised that no new alleles were generated by recombination and we used only the results from the APLP genotyping36 (Figure 2a).

Figure 2
figure 2

Electrophoresis images of the amplified product length polymorphism (APLP) (a) and restriction fragment length polymorphism (RFLP) (b) analyses. (a) Sets A, B and C are products of PCR reactions with primers 1–4, 5–6 and 8–10,36 respectively. (b) Lanes 1 and 3 show PCR products before digestion with NaeI, whereas lanes 2 and 4 indicate PCR products after digestion with NaeI. Sample A possesses allele ABO*A101 and sample B has allele ABO*A102.

Full size image

To distinguish allele ABO*A101 from ABO*A102, PCR-restriction fragment length polymorphism (RFLP) analysis was carried out for specimens having A1 allele because only ABO*A102 does not possess a NaeI restriction site in exon 7 (Figure 2b). The DNA fragment including a single-nucleotide polymorphism, 467C → T, defining allele ABO*A102 was amplified with the two primers newly designed in this study: ABO-RFLP1 (5′-AAGCACTTCATGGTGGGCCA-3′) and ABO-RFLP2 (5′-GACAGCTGCCGACCGGTC-3′). After PCR amplification, an aliquot of 8.8 μl of PCR products was digested with NaeI in the reaction mixture containing 1 × buffer and 2 units of restriction enzyme at 37 °C for 3 h. The digested DNA was electrophoresed on the native polyacrylamide gel and detected by the same method as that of the APLP analysis (Figure 2b).

The APLP and PCR-RFLP analyses for each bone sample were carried out three times to assess the reproducibility of the analytical results. Samples from which the reproducibility was not obtained were excluded from the subsequent statistical analysis.

Statistical analysis

To check for any possibilities of external DNA contamination, the Hardy–Weinberg equilibrium test for genotypic data of the Jomon/Epi-Jomon people and the Okhotsk people that were obtained in this study was conducted using GENEPOP version 4.0 (http://genepop.curtin.edu.au/).38, 39 Allele frequencies of the ABO gene locus were estimated by direct counting, and compared with previously reported data from modern populations such as East Asia, Europe and Oceania: 340 Japanese,22 43 Han, 37 Kazakh, 59 Uygur,23 57 Balopa, 69 Gidra, 34 Tongan25 and 169 German.37 On the basis of ABO allele frequencies obtained, principal component (PC) analysis was also performed to examine the genetic affinities among 10 populations using STATISTICA version 06J (Statsoft Japan).

Results

The DNA fragments including the single-nucleotide polymorphisms in the ABO gene exons 6 and 7 were successfully PCR-amplified, and genotyped from 8 of 42 Jomon, 3 of 39 Epi-Jomon and 13 of 50 Okhotsk specimens (Table 1). In this study, because the numbers of successfully genotyped individuals was small in the Jomon and Epi-Jomon specimens, these were regarded together as the Jomon/Epi-Jomon people. No successful results were obtained from the remaining samples because of possible DNA degradation. Departure from Hardy–Weinberg equilibrium was not observed in either Jomon/Epi-Jomon or Okhotsk people (P>0.05).

Table 1 Genotypes of the ABO blood group gene of the Jomon/Epi-Jomon and Okhotsk specimens
Full size table

Table 2 shows allele frequencies of the ABO gene in the Jomon/Epi-Jomon and Okhotsk people obtained in this study, compared with those of other populations cited from previous studies. The Jomon/Epi-Jomon and Okhotsk people shared five alleles (Tables 1 and 2), all of which were previously reported to be present in modern Japanese people.23, 36 The predominant allele in the Jomon/Epi-Jomon people was ABO*O102, whereas that in the Okhotsk people was ABO*O101. Allele frequencies of the A and B groups in the Jomon/Epi-Jomon people were slightly higher than those in modern Japanese, and that of the O group in the Jomon/Epi-Jomon people was slightly lower than that of modern Japanese (Table 2). Allele ABO*A102, which has been detected as a predominant A allele in East Asian populations (0.197 in Japanese22 and 0.186 in Han23), was observed in both Jomon/Epi-Jomon (0.182) and Okhotsk people (0.038), whereas the frequency of ABO*A101 in Jomon/Epi-Jomon people was higher than that of modern East Asian populations such as Japanese and Han (Table 2). The frequencies of subtypes of the O group in the Jomon/Epi-Jomon people were considerably different from those of modern Japanese. Within the Jomon/Epi-Jomon people, the frequency of ABO*O101 (0.136) was significantly lower than that of ABO*O102 (0.318). Frequencies of ABO*O101 and ABO*O102 among O alleles (individual frequency against 0.454 as total ‘O’ frequency) in Jomon/Epi-Jomon people were 30.0% (0.136/0.454) and 70.0% (0.318/0.454), respectively (see Table 2). By contrast, those in modern Japanese were 54.3% (0.293/0.540) and 45.7% (0.247/0.540), respectively. In the Okhotsk people, 76.9% of ABO alleles consisted of ABO*O101 and ABO*O102, indicating that the allele frequencies were specifically different from those of other East Asian populations.

Table 2 Allele frequencies of the ABO blood group gene in ancient populations of Hokkaido, and Asian, European and Oceanic populations
Full size table

Figure 3 shows the result of the PC analysis for the 10 populations. Analyzed populations were roughly separated into three groups: Asian, European and Oceanic groups, and the Jomon/Epi-Jomon people and Okhotsk people were included in the Asian group. The modern Japanese population was located nearer Han rather than near the Jomon/Epi-Jomon people (Figure 3). Although the Jomon/Epi-Jomon and Okhotsk people were distributed in Hokkaido, they were located far from each other in the Asian group (Figure 3). Allele ABO*B101 showed the highest contribution rate to the PC 1 (24.6%) and allele ABO*O201 showed the highest contribution rate to the PC 2 (23.4%) through the 10 populations.

Figure 3
figure 3

Genetic relationships among the 10 populations determined on the basis of principal component analysis of ABO allele frequencies. Data of Japanese,22 Han,23 Kazakh,23 Uygur,23 Balopa,25 Gidra,25 Tongan25 and German37 populations are cited from previous studies.

Full size image

Discussion

This study first revealed polymorphisms of the ABO blood group gene in ancient populations of Hokkaido. All five ABO alleles detected in this study were observed in both Jomon/Epi-Jomon and Okhotsk people (Tables 1 and 2). Allele ABO*A102 is the predominant allele among the A group in East Asian populations. Frequencies of A101 and A102 are 2.3% and 18.6% in Han and 6.9% and 19.7% in Japanese, respectively.22, 23 ABO*A102, which is the most common allele in East Asia, was observed in both the Jomon/Epi-Jomon and Okhotsk people analyzed in this study. Results indicate that it is reasonable that the Jomon/Epi-Jomon and Okhotsk people had the characteristics of the ABO blood group gene that is common to Asian populations.

Among O alleles of the Jomon/Epi-Jomon people, the frequency (0.318) of ABO*O102 included in the OG group, which possesses the A → G mutation in nucleotide position 297, was remarkably higher than that (0.136) of ABO*O101 included in the OA group, which does not possess the mutation in that position (Table 2). Nakamura40 reported that a gentle gradation of O allele frequencies was seen across the Japanese archipelago: the higher frequency of the OA group was observed in central parts of Japan such as Gifu and Mie Prefectures, whereas the higher frequency of the OG group was observed in northern and southern parts of Japan, such as Akita, Okinawa and Kagoshima Prefectures. It is generally considered that the modern Japanese population was formed by two distinct populations (Neolithic Jomon people and Aeneolithic Yayoi immigrants), and gene flow from the Asian continent by migration of the Yayoi people strongly influenced the native population in the central part of the Japanese archipelago, and hardly influenced those in northern and southern parts of the Japanese archipelago such as Hokkaido and Okinawa.41 These findings suggest that the OG group is the predominant O allele in the Neolithic Jomon people who were native to the Japanese archipelago, and that the OA group was the predominant O allele in the Aeneolithic Yayoi people who were immigrants to the Japanese archipelago from the Asian continent between the third century BC and the third century AD. To verify this hypothesis, molecular genetic information on polymorphisms of the ABO gene for the Jomon and Yayoi people in Honshu and Ainu in Hokkaido is required.

Among the Okhotsk people, 76.9% of ABO alleles consisted of members of the O group. On the basis of morphological42, 43, 44, 45 and mtDNA data,31, 32 the Okhotsk people could have originated from the Sakhalin Island and from lower regions of the Amur River in East Siberia. This indicates that the frequency of O alleles among ancestors of those populations might have been high. Otherwise, it is likely that the high frequency of O alleles detected in the Okhotsk people was caused by founder effects through the migration to northeastern coastal regions of Hokkaido. The future investigation on polymorphisms of the ABO gene among modern populations native to Sakhalin and lower regions of the Amur River could provide an insight into unveiling the history of establishment of people around the Okhotsk Sea.

The PC analysis of allele frequencies (Figure 3) indicated that the ancient populations of Hokkaido analyzed in this study were apparently included into the Asian group. In addition, it showed that the population genetically closest to modern Japanese was Han, and not the Jomon/Epi-Jomon people. This suggests that the allele frequencies of the ABO gene in modern Japanese were strongly influenced by gene flow from the Asian continent caused by immigration of Aeneolithic Yayoi people. In this study, however, only the Jomon/Epi-Jomon people in Hokkaido were examined. Future analyses of the ABO gene for the Jomon people in Honshu would clarify the effects of gene flow of the Aeneolithic Yayoi people.

As most of the mtDNA haplogroups identified in the Jomon/Epi-Jomon people of Hokkaido were common to those of eastern Siberian people, the Jomon/Epi-Jomon people of Hokkaido might have been strongly affected by gene flow from northeastern Asian populations.29, 30 Similarly, morphological and genetic studies indicate that the Okhotsk people could have originated from eastern Siberia as well.31, 32, 42, 43, 44, 45 The Jomon/Epi-Jomon and Okhotsk people were, however, located far from each other, as shown in the PC analysis (Figure 3). Although the Jomon/Epi-Jomon people and Okhotsk people shared the five alleles, the frequencies were considerably different from each other. The mtDNA haplogroup distributions of the Jomon/Epi-Jomon29, 30 and Okhotsk people32 are also different from each other. In addition, allele frequencies of the earwax gene ABCC11 in those people were significantly different from each other.33 These findings strongly show that Jomon/Epi-Jomon and Okhotsk people have characteristics that are different from each other in mitochondrial (maternal) and autosomal (biparental) gene pools, in addition to the classical morphological differences.

Although the Ainu (modern indigenous people in Hokkaido) could have been established by genetic mixtures between direct descendants of the Neolithic Jomon and Okhotsk people,31, 32, 33 gene flow from the Okhotsk people to the Ainu was not tested in this study because the previous data of the ABO blood group for the Ainu are of an immunological nature and show only frequencies of the A, B and O groups; that is, frequencies of subtypes of the A, B and O groups for the Ainu are not available. To further understand the history of populations in Hokkaido, it is important to clarify the polymorphisms of the ABO blood group gene in the Ainu for comparison with those of the Jomon/Epi-Jomon and Okhotsk people.