Population genetic relationships between Mediterranean populations determined by HLA allele distribution and a historic perspective

HLA genes allele distribution has been studied in Mediterranean and sub-Saharan populations. Their relatedness has been tested by genetic distances, neighbour-joining dendrograms and correspondence analyses. The population genetic relationships have been compared with the history of the classical populations living in the area. A revision of the historic postulates would have to be undertaken, particularly in the cases when genetics and history are overtly discordant. HLA genomics shows that: 1) Greeks share an important part of their genetic pool with sub-Saharan Africans (Ethiopians and west Africans) also supported by Chr 7 Markers. The gene flow from Black Africa to Greece may have occurred in Pharaonic times or when Saharan people emigrated after the present hyperarid conditions were established (5000 years B.C.). 2) Turks (Anatolians) do not significantly differ from other Mediterraneans, indicating that while the Asians Turks carried out an invasion with cultural significance (language), it is not genetically detectable. 3) Kurds and Armenians are genetically very close to Turks and other Middle East populations. 4) There is no HLA genetic trace of the so called Aryan invasion, which has only been defined on doubtful linguistic bases. 5) Iberians, including Basques, are related to north-African Berbers. 6) Present-day Algerian and Moroccan urban and country people show an indistinguishable Berber HLA profile.     

Mixtec Mexican Amerindians: an HLA Alleles Study for America Peopling,Pharmacogenomics and Transplantation

HLA-A, -B and -DRB1 alleles have been studied in a Mixtec Mexican Amerindian population by indirect DNA sequencing. HLA relatedness has been tested by comparing results with other Amerindians and worldwide populations; a total of 15,681 chromosomes have been used. Genetic distances between populations, Neighbour Joining (NJ) dendrograms and correspondence analyses have been carried out. Conclusions are: 1) Our Mixtec sample from Oaxaca Coastal Mexican area shows an HLA profile different to that of Oaxaca Central Mountains area showing that genes and languages do not correlate which is inferred both by plane genetic distances and NJ dendrograms and correspondence analyses. 2) Genetic distances and NJ dendrograms join together Mazatecan Mexican Amerindians with our studied Coastal Mixtec group; it fits with the historical relationship between Mixtec and Mazatecans. 3) A*24:02-B*35:14-DRB1*04:11, A*02:01-B*15:15-DRB1*04:11 and A*68:03-B*39:08-DRB1*08:02 extended HLA haplotypes have been “de novo” found in our Mixtec Coastal sample. 4) Shared HLA alleles are found between our Pacific Coast Mixtec Amerindians and Pacific Islanders. 5) These results are useful for establishing a future area transplantation waiting list, for the study of HLA linked diseases epidemiology and for pharmacogenomics in certain drug therapy.     

The origin of Cretan populations as determined by characterization of HLA alleles

The Cretan HLA gene profile has been compared with those of other Mediterranean populations in order to provide additional information regarding the history of their origins. The allele frequencies, genetic distances between populations, relatedness dendrograms and correspondence analyses were calculated. Our results indicate that the Indoeuropean Greeks may be considered as a Mediterranean population of a more recent origin (after 2000 B.C.), while all other studied Mediterraneans (including Cretans) belong to an older substratum which was present in the area since pre-Neolithic times. A significant Turkish gene flow has not been detected in the Greek or Cretan populations, although Greeks and Turks have two high frequency HLA-DRB-DQB haplotypes in common. It is proposed that Imazighen (Caucasoid Berbers living at present in the North African coast and Saharan areas) are the remains of pre-Neolithic Saharan populations which could emigrate northwards between about 8000-6000 B.C., when desert desiccation began. They also could be part of the stock that gave rise to Sumerians, Cretans and Iberians; this is supported by both linguistic and HLA genetic data.     

The correlation between languages and genes: the Usko-Mediterranean peoples.

The usko-Mediterraneans peoples are defined as ancient and present day populations that have lived in the Mediterranean/Middle-East/Caucasus area and have spoken a Basque related language. The present day existing populations show an HLA genetic relatedness which is more or less close according to geographical distance. The Greek sample is an outlying in all genetic analyses, because Greeks have a significant genetic input from sub-Saharan Ethiopians and Blacks. This probably occurred in Pharaonic times. Present day comparisons between genes and languages show a lack of correlation: Macedonian, Palestinians, Kurds, part of Berbers, Armenians, and Turks belong to the old Mediterranean substratum, but they do not speak a language included in the old Mediterranean Dene-Caucasian group. This is due to an "elite"-imposed culture and language. Other ethnic groups speak an "old Mediterranean language" or "usko-Mediterranean language" modified by Roman Latin (i.e., Spanish, Italians), or by other not fully explained processes (Jews). Therefore, the correlation between genes and languages may exist at a macrogeographical level, but not when more precise microgeographical studies are done, as shown in the present "usko-Mediterranean" peoples model.     

HLA class I and class II polymorphisms in Tunisian Berbers

Background: The HLA polymorphism is a powerful genetic tool to study population origins. By analysing allele frequencies and haplotypes in different populations, it is possible to identify ethnic groups and establish the genetic relationships among them.

Aim: The Berber (endogenous Tunisians) HLA class I and class II genotypes were analysed and compared with those of Mediterranean and Sub-Saharan African communities using genetic distances, Neighbour-Joining dendrograms, correspondence and haplotype analysis.

Subjects and methods: One hundred and five unrelated Berbers were typed for HLA class I (A, B) and class II (DRB1, DQB1) gene alleles using reverse dot-blot hybridization.

Results: High frequencies of A*0201 (24.76%), A*3402 (22.38%) and B*44 (32.85%) alleles were recorded for Berbers, the highest recorded for Mediterranean and North African populations. This study shows a close relatedness of Tunisian Berbers to other Tunisians, North Africans and Iberians.

Conclusion: The apparent relatedness of Tunisian Berbers to present-day (North African) Tunisians, Algerians and Moroccans suggests that the Arab invasion of North Africa (7^th–11^th centuries AD) did not significantly impact the genetic makeup of North Africans. Furthermore, Tunisian Berbers appear to be closely related to Iberians (Spaniards and Basques), indicating that the 7^th century AD gene flow of invaders was low in Iberians and that the main part of their genetic pool came after the Northward Saharan migration, when hyper-arid conditions were established in Sahara (before 6000 BC). Other studied populations belong to the old Mediterranean substratum, which has been present in the area since pre-Neolithic times. This study indicates a higher proportion of Iberian than Arab ancestry in Tunisian Berbers, which is of value in evaluating the evolutionary history of present-day Tunisians. Greeks seem to share genetic HLA features (Chr 6) with Sub-Saharans. The relatedness of Greeks to Sub-Saharans has been confirmed by other studies based on chromosome 7 genetic markers.     

HLA polymorphism in Bulgarians defined by high-resolution typing methods in comparison with other populations

In the present study we analyzed for the first time HLA class I and class II polymorphisms defined by high-resolution typing methods in the Bulgarian population. Comparisons with other populations of common historical background were performed. Most HLA-A, -B, -DRB alleles and haplotypes observed in the Bulgarian population are also common in Europe. Alleles and haplotypes considered as Mediterranean are relatively frequent in the Bulgarian population. Observation of Oriental alleles confirms the contribution of Asians to the genetic diversity of Bulgarians. The use of high-resolution typing methods allowed to identify allele variants rare for Europeans that were correlated to specific population groups. Phylogenetic and correspondence analyses showed that Bulgarians are more closely related to Macedonians, Greeks, and Romanians than to other European populations and Middle Eastern people living near the Mediterranean. The HLA-A,-B,-DRB1 allele and haplotype diversity defined by high-resolution DNA methods confirm that the Bulgarian population is characterized by features of southern European anthropological type with some influence of additional ethnic groups. Implementation of high-resolution typing methods allows a significantly wider spectrum of HLA variation to be detected, including rare alleles and haplotypes, and further clarifies the origin of Bulgarians.     

HLA alleles and haplotypes in the Turkish population: relatedness to Kurds, Armenians and other Mediterraneans

Turkish and Kurdish HLA profiles are studied for the first time. The comparative study of their allele frequencies, characteristic haplotypes, genetic distances with other Mediterraneans is complemented by neighbor-joining dendrograms and correspondence analyses. Turks, Kurds, Armenians, Iranians, Jews, Lebanese and other (Eastern and Western) Mediterranean groups seem to share a common ancestry: the older "Mediterranean" substratum. No sign of the postulated Indo-European (Aryan) invasion (1200 B.C.) is detected by our genetic analysis. It is concluded that this invasion, if occurred, had a relatively few invaders in comparison to the already settled populations, i.e. Anatolian Hittite and Hurrian groups (older than 2000 B.C.). These may have given rise to present-day Kurdish, Armenian and Turkish populations.     

The contribution of HLA class I and II alleles and haplotypes to the investigation of the evolutionary history of Tunisians

The frequencies of HLA class I and class II alleles and haplotypes of 104 healthy unrelated Tunisians were analyzed by high-resolution PCR-reverse dot blot hybridization, and was compared with other Mediterranean and Sub-Saharan Africans using genetic distances measurements, Neighbor-joining dendrograms, correspondence, and extended haplotypes analysis. The most frequent HLA class I A alleles were A*02, A*24, and A*30, while the most frequent B alleles were B*44, followed by B*50, B*51, and B*07. Among HLA class II DRB alleles analyzed, the most frequent were DRB1*0301, DRB1*0701, DRB1*1501, followed by DRB1*1303 and DRB1*0102; for DQB1, they were DQB1*0301 and DQB1*0201. Three-locus haplotype analysis revealed that A*03-B*07-DRB1*1503 and A*02-B*44-DRB1*0402 were the most common HLA class I and II haplotypes in this population. Compared with other communities, our result indicate that Tunisians are very related to North Africans and Western Europeans, particularly Iberians, and that Tunisians, Algerians, and Moroccans are close to Berbers suggesting little genetic contribution of Arabs who populated the area in 7th to 8th century AD. The similarities and differences between Tunisians and neighboring and related communities in HLA genotype distribution provide basic information for further studies of the MHC heterogeneity among Mediterranean and North African countries, and as reference for further anthropological studies.     

HLA class I and class II polymorphism in a population from south-eastern Tunisia (Gabes Area)

The gene frequencies of HLA class I and class II alleles were investigated in 95 healthy Tunisian individuals from Gabes. Our aim was to compare the genetic relationship between Gabesians and Mediterraneans and sub-Sahara Africans using genetic distances, Neighbour-Joining dendrograms, correspondence and haplotypes analysis, thereby providing additional information about evolutionary history of modern-day Tunisians. Subjects were unrelated and of both genders, and HLA class I and class II genes were genotyped using the polymerase chain reaction-sequence specific primer (PCR-SSO) technique. Our data show that south-eastern Tunisians (Gabes area) are related to present-day North Africans (Algerians, Moroccans, Tunisians) and Iberians (Spaniards, Basques), and along with other North Africans, appear to be genetically related to Berbers, an indication that the Arab invasion (7th-11th centuries) of North Africa had minimal contribution on the HLA makeup of North Africans. On the other hand, Iberians including Spaniards and Basques show relatedness to (native Tunisian) Berbers, suggesting that the gene flow of 7th century AD invaders was also low in Iberians. In conclusion, the successive invasions of North Africa in general, and Tunisia in particular, did not modify markedly the genetic makeup of present-day Tunisians. With the exception of Greeks who have a sub-Saharan genetic profile, all Mediterranean populations depict a typical mediterranean substratum.     

High-resolution typing of HLA-DRB1 locus in the Macedonian population

The Macedonian population is of special interest for HLA anthropological study in the light of unanswered questions regarding its origin and relationship with other populations, especially the neighbouring Balkanians. Two studies have been performed to examine HLA molecular polymorphism in the Macedonian population, so far. The present study is the first to be performed in Macedonia using high-resolution sequence-based method for direct HLA typing. The study included 158 unrelated healthy volunteers of Macedonian origin and nationality, having a Christian Orthodox religion. After the simultaneous amplification of exon-2 on both HLA-DRB1 alleles, DNA sequencing was used for genotype assignment. In the 158 samples analysed, all 316 alleles were typed and a total of 29 different DRB1 alleles were detected, with DRB1*1601 being the most frequent allele (14.9%), followed by DRB1*1104 (13.9%). A phylogenetic tree constructed on the basis of the high-resolution data deriving from other populations revealed the clustering of Macedonians together with other Balkan populations (Greeks, Croats, Turks and Romanians) and Sardinians, close to another "European" cluster consisting of the Italian, French, Danish, Polish and Spanish populations. The included African populations grouped on the opposite side of the tree.     

HLA-DR, DQ nucleotide sequence polymorphisms in the Pasiegos (Pas valleys, Northern Spain) and comparison of the allelic and haplotypic frequencies with those of other European populations

In general, Northern Spain has remained geographically isolated from neighboring Spanish regions for centuries: steep mountains create small isolated and inbred population groups with their own characteristic cultures and unique gene pools. The Pasiego region forms an area of distinctive characteristics among the people living in Northern Spain, although the origin of the inhabitants of the Pas valleys (Pasiegos) is not clearly defined. We have studied the MHC class II alleles in a large sample of unrelated individuals living in the Pas valleys. Allelic and haplotypic frequencies, population distances and their corresponding dendrogram, using the N-J method, were used to study the relationships between populations. The closest is observed between Pasiegos and Danes, followed by other European people in the following decreasing order: Poles, Germans, non-Pasiego Cantabrians, Belgians, Basques, French, other Spaniards from Madrid, Italians, Finns, Croatians, Welsh, Ashkenazi Jews and other Mediterranean populations (Greeks, Hungarians, Sardinians and Bulgarians). Particular characteristic Northern European alleles are observed with high frequency in the Pasiegos and non-Pasiego Cantabrians (DRB1*1501-DQA1*0102-DQB1*0602). The second most frequent three-locus haplotype in both populations is DRB1*0701-DQA1*0201-DQB1*0201. These observations suggest an important mixture of alleles from geographically distinct areas. In conclusion, the Pasiegos are typical examples of isolated genetic pools in the Iberian Peninsula and allow one to suggest that what we call the "Pasiego cluster" can be considered, in many ways, as another example of the few deviant groups (e.g. Lapps, Basques and Sardinians) having preserved their genetic, social and ethnographic characteristics and, in some cases, their ancestral language.     

HLA genes in Cubans and the detection of Amerindian alleles

Caribbean Islands including Cuba were first inhabited by Meso-American and later by Arawak-speaking Amerindians from nowadays Venezuela. Spanish invaders brought to almost extinction to the Amerindian population after 1492. Black slaves from West Africa were taken into Cuba by Europeans. The degree of admixture among populations is approached. HLA alleles were studied by DNA techniques. Comparison with other worldwide populations (a total of 14.094 chromosomes) included genetic distances, Neighbour-Joining dendrograms, correspondence analyses and calculation of extended haplotypes. While African-European HLA features were clearly found, Amerindian HLA characteristics are less evident, indicating that Amerindian devastation was particularly marked after 1492 AD. However, typical Amerindian alleles have been found in our Cuban sample, i.e. DRB1*0403, DRB1*0404, DRB1*0407, BRB1*0411, DRB1*0802 and DRB1*0809. The presence of Amerindian alleles in Cubans may have a bear in the making up of transplantation registries (both for bone marrow and solid organ transplantation) at the regional level and also be important for epidemiological studies of diseases linked to HLA.     

Gorgan (Turkmen in Iran) HLA genetics: transplantation,pharmacogenomics and anthropology

HLA class I and II alleles have been studied in a population from Gorgan (North East Iranian city bordering Turkmenistan). This population is composed of mainly Turkmen who speak Oghuz Turkish language. Comparison of Gorgan people HLA profile has been carried out with about 7984 HLA chromosomes from other worldwide populations; extended haplotypes and three dimension genetic distances have been calculated by using neighbor-joining and correspondence relatedness analyses. Most frequent extended HLA haplotypes show a Siberian/Mediterranean admixture and closest populations are Chuvashians (North Caspian Sea, Russia) and other geographically close populations like Siberian Mansi, Buryats and other Iranians. New extended HLA haplotypes have been found, such as: A*31:01-B*35:01-DRB1*15:01-DQB1*03:01, A*01:01-B*35:01-DRB1*03:01-DQB1*02:01. Relationships of Turkmen with Kurgan (Gorgan) archaeological mounds, Scythians and Sarmatians are discussed. This study is also useful for a future transplantation Gorgan waiting list, Gorgan HLA and disease epidemiology and HLA pharmacogenomics.     

HLA genes in Uros from Titikaka Lake,Peru: origin and relationship with other Amerindians and worldwide populations

Uros population from the Titikaka Lake live in about 42 floating reed (‘totora’) islands in front of Puno City (Peru) at a 4000 m high altiplano. They present both an mtDNA and a human leucocyte antigen (HLA) profile different from the surrounding populations: mtDNA A2 haplogroup is common to Uros and Amazon forest lowland Amerindians. HLA genetic distances between populations have been calculated and neighbour‐joining dendrograms and correspondence analyses were carried out. Approximately 15 006 HLA chromosomes from worldwide populations have been used for comparisons. Only eight HLA‐A alleles have been found, three of them accounting for most of the frequencies. The same phenomenon is seen for HLA‐B, HLA‐DRB1 and HLA‐DQB1 alleles: a few alleles (3, 4 and 3, respectively) are present in most individuals. The presence of HLA‐B*4801 and HLA‐DRB1*0901 alleles in a relatively high frequency (although not the most frequent alleles found) is a characteristic shared with Asians and some populations from the Andean altiplano. Three specific Uros haplotypes have been found among the most frequent ones: HLA‐A*680102‐B*3505‐DRB1*0403‐DQB1*0302; HLA‐A*2402‐B*1504‐DRB1*1402‐DQB1*0301; and HLA‐A*2402‐B*4801‐DRB1*0403‐DQB1*0302. The present study suggests that Uros may have been one of the first populations from the shores of the Titikaka Lake coming from the Amazonian forest, which might have given rise to other later differentiated ethnic group (i.e. Aymaras). Uros HLA profile is also useful to study genetic epidemiology of diseases linked to HLA and to construct a future transplant waiting list by adding up regional lists in order to get a bigger pool for transplanting with better HLA matching.     

Human leukocyte antigen-A, -B and -C alleles and human leukocyte antigen haplotypes in Turkey: relationship to other populations

In this study, we present, for the first time, human leukocyte antigen (HLA) class I allele and haplotype frequencies at the DNA level in a sample of 142 donors from Turkey. HLA typing was performed by medium-to-high resolution polymerase chain reaction sequence-specific oligonucleotide probes method. The most frequent HLA alleles at class I locus were A*0201(0.257), -B*35(0.204) and -Cw*04(0.173). A*0201-B*35-Cw*04(0.056) was the most common three-locus haplotype. Allele and haplotype frequency comparisons and neighbour-joining dendrograms, constructed using DA genetic distances and correspondence analysis using HLA-A, -B and -C, and -DRB1 allele frequencies, revealed similarities with other Mediterranean and European populations, but not with Mongol populations. These results agree with previous studies and confirm that the present day Turkish population is genetically more similar to its geographic neighbours than its historical neighbours in central Asia. The comprehensive HLA data on the Turkish population at the DNA level including up to six-locus putative haplotypes generated in this study will be useful for further studies.     

HLA genes in Southern Tunisians (Ghannouch area) and their relationship with other Mediterraneans

South Tunisian HLA gene profile has studied for the first time. HLA-A, -B, -DRB1 and -DQB1 allele frequencies of Ghannouch have been compared with those of neighboring populations, other Mediterraneans and Sub-Saharans. Their relatedness has been tested by genetic distances, Neighbor-Joining dendrograms and correspondence analyses. Our HLA data show that both southern from Ghannouch and northern Tunisians are of a Berber substratum in spite of the successive incursions (particularly, the 7th-8th century A.D. Arab invasion) occurred in Tunisia. It is also the case of other North Africans and Iberians. This present study confirms the relatedness of Greeks to Sub-Saharan populations. This suggests that there was an admixture between the Greeks and Sub-Saharans probably during Pharaonic period or after natural catastrophes (dryness) occurred in Sahara.     

Origin of Mexican Nahuas (Aztecs) according to HLA genes and their relationships with worldwide populations

A Nahua Aztec isolated group from Morelos State (Mexico) was studied for their HLA profile. The relationship with other Amerindians and worldwide populations was studied by using 13,818 chromosomes and calculating Nei's chord genetic distances (DA), neighbor-joining dendrograms and correspondence multidimensional values. Three new HLA extended haplotypes were found in our group: A*30-B*49-DRB1*1001-DQB1*0501 (the most frequent one in this population), A*02-B*52-DRB1*1402-DQB1*0301 and A*68-B*61-DRB1*1602-DQB1*0303. Both genetic distances and correspondence analyses clearly show that our Nahua isolated group is genetically close to some of the most ancient groups living in Mexico (Mayans, Zapotecans, Mixtecans). This suggests that Nahua language (Nahuatl) may have been imposed to scattered groups throughout Mexico; otherwise Aztecs may have been living in Mexico long before their postulated immigration in the XII century AD.     

Origin of Bolivian Quechua Amerindians: their relationship with other American Indians and Asians according to HLA genes

The Incas were Quechua-speaking people who settled down near Cuzco (Peru). They had an empire ranging from Ecuador to Chile, when Spanish conquerors seized their kingdom around 1532 AD. Nowadays, Quechua-speaking people inhabits Colombia, Ecuador, Bolivia, Peru and Argentina; however, Quechua language was imposed by both Incas and Spaniards to many non-Quechua speaking communities. We have taken a sample of Quechuan Bolivian blood donors from La Paz (Titicaca Lake region) where Inca-Quechuas themselves believed that came from. This group was compared with 6892 individuals from 68 different world populations regarding HLA/DNA allele frequencies distribution. Genetic distances, dendrograms and correspondence analyses were carried out in order to establish relationships among populations. The main conclusions are: (1) DRB1 and -DQB1 haplotypes shared with Asians are found in Quechuas and are not observed in other (Mesoamerican) Amerindians. (2) Aymara-speaking people from the same Titicaca Lake (La Paz) area shows close genetic distances with Quechuas in one dimension results (genetic distances); however, their HLA gene frequency distribution differs according to Neighbor-Joining (NJ) trees and correspondence analysis (multidimensional and more reliable analyses). Also, the common high frequency Asian and Athabascan HLA-DRB1*0901 allele is found in Quechuas in a significant frequency. Quechuas are clearly included within the Amerindian group.     

HLA polymorphism in the Murcia population (Spain): in the cradle of the archaeologic Iberians.

Human leukocyte antigen (HLA) study in Murcian individuals was performed in order to provide information of their historical origins and relationships with other Iberian and Mediterranean populations. HLA class I and class II alleles were determined in 173 unrelated Caucasoid donors from Murcia Region in the Southeast of Spain by serologic and DNA based polymerase chain reaction (PCR) typing. Class I antigen and class II allele frequencies of our series were not very different to those found in Spaniards. The analysis of extended haplotypes showed that the three haplotypes most frequent in our population were respectively, A29-B44-Cwb-DRB1*0701-DRB4*0101-DQA1*0201-DQB1*0202, A1-B8-Cw7-DRB1*0301-DRB3*0101-DQA1*0501-DQB1*0201 and A30-B18-Cw5-DRB1*0301-DRB3*0101-DQA1*0501-DQB1*0201. They were followed by A26-B38-Cwb-DRB1*1301-DRB3*0202-DQA1*0103-DQB1*0603, which could point to an ancestral relationship between Murcian and Portuguese Iberian populations, and by A2-B7-Cw7-DRB1*1501-DRB5*0101-DQA1*0102-DQB1*0602 also present in all Iberian Peninsula populations. Allelic frequencies, populations distance dendrogram and correspondence analysis were used to study the relationships between Murcian and other populations. The closest relation was observed with Spaniards and Portuguese, followed in decreasing order by French, Italians, Algerians, Germans, Catalans, Basques, Cretans, Sardinians, and Greeks. Thus, Murcian population seems to belong to the European genetic pool, revealing a lesser genetic distance with the North Africans and the rest of populations from the Iberian Peninsula.     

The HLA class I and class II allele frequencies studied at the DNA level in the Svanetian population (Upper Caucasus) and their relationships to Western European populations

The Caucasus and the Iberian peninsula have been connected from a linguistic (Basque and Kvartelian languages), toponimic and historic perspectives. They also represent places (e.g. Dmanisi in Georgia and Atapuerca in Northern Spain) where the oldest hominoid remains in Europe are being discovered and studied. These circumstances prompted us to study the genetic background of the Svans (living on the southern slopes of the Greater Caucasus in the Republic of Georgia) in comparison with Basques from the semi-isolated Arratia valley as well with other Northern Spanish and Western European populations. DRB1*1101-DQA1*0501-DQB1*0301 and DRB1*1301-DQA1*0103-DQB1*0603 haplotypes were found in Svans at the highest frequency. The second most frequent three-locus haplotypes in this population were DRB1*0701-DQA1*0201-DQB1*0201 and DRB1*1301-DQA1*0103-DQB1*0602. Furthermore, the following 5-locus extended haplotypes were not found in other populations: A3-B8-DRB1*11-DQA1*0501-DQB1*0301, A2-B8-DRB1*13-DQA1*0103-DQB1*0603, A2-B40-DRB1*14-DQA1*0104-DQB1*0501, A2-B51-DRB1*08-DQA1*0401-DQB1*0402, A3-B7-DRB1*03-DQA1*0501-DQB1*0201 and A24-B39-DRB1*08-DQA1*0401-DQB1*0402. Other haplotypes present in Svans were also frequently observed in Northern Spain and in other Western European countries. However, haplotypes reported as characteristic for Basques were not found in the Svans. A dendrogram using HLA class II alleles places the closest genetic distance observed between Svans and Czechs, whereas Slovenes and other Mediterranean populations (Jews, Hungarians, Frenchmen, Sardinians and Greeks) have the greatest genetic distance. When both HLA class I and class II alleles from 17 populations were compared, the smallest genetic distances were with Rumanians, Czechs and Armenians. Northern Spanish populations were placed closer to each other and clearly separated from Svans. In conclusion, the Svan population shows considerable polymorphism. These observations suggest a mixture of alleles in Svans from geographically distinct areas, and probably do not support a common ancestor for these Caucasian inhabitants and people from Northern Spain.