HLA Class II Profile and Distribution of HLA-DRB1 and HLA-DQB1 Alleles and Haplotypes among Lebanese and Bahraini Arabs

Abstract

The gene frequencies of HLA class II alleles were studied in 95 healthy Lebanese Arab and 72 healthy Bahraini Arab subjects. Our aim was to establish the genetic relationship between Bahraini and Lebanese Arabs in terms of HLA class II gene and haplotype frequencies and to compare these results with frequencies for other countries with populations of Caucasian and non-Caucasian descent. Subjects were unrelated and of both sexes, and HLA-DRB1 and -DQB1 genotyping was done by the PCR sequence-specific primer technique. Comparative analysis of the HLA-DR and -DQ alleles revealed differences in the allelic distribution among Bahraini and Lebanese subjects. Analysis of the 25 HLA-DRB1 alleles that have been investigated showed that the DRB1*040101 and DRB1*110101 alleles were more frequent among Lebanese, whereas DRB1*030101 and DRB1*160101 alleles were more frequent among Bahrainis. Similarly, of the seven HLA-DQB1 alleles analyzed, the presence of DQB1*0201 was more frequent among Bahrainis, whereas DQB1*030101 was more frequent among Lebanese. The DRB1*160101-DQB1*050101 (0.1318 versus 0.0379%) and DRB1*030101-DQB1*0201 (0.1202 versus 0.0321%) haplotypes were more frequent among Bahrainis, while the DRB1*110101-DQB1*030101 (0.3142 versus 0.1198%) and DRB1*040101-DQB1*0302 (0.1416 versus 0.0278%) haplotypes were more frequent in Lebanese subjects. Furthermore, a high prevalence of the DRB1*040101-DRB1*110101-DQB1*0302-DQB1*030101 (12.63 versus 1.35%, P = 0.015) and the homozygous DRB1*110101-DRB1*110101-DQB1*030101-DQB1*030101 (7.37 versus 0.00%, P = 0.046) genotypes was seen among Lebanese, and DRB1*070101-DRB1*160101-DQB1*0201-DQB1*050101 (6.76 versus 0.00%, P = 0.034) was seen more frequently among Bahraini subjects. Our results underline significant differences between these two populations in HLA class II distribution, provide basic information for further studies of major histocompatibility complex heterogeneity among Arabic-speaking countries, and serve as a reference for further anthropological studies.

The major histocompatibility complex (MHC) plays a key role in self-nonself recognition and is broadly divided into highly polymorphic class I (HLA-A, -B, and -C) and class II (HLA-DP, -DQ, and -DR) loci, which are subdivided into distinct regions (refer to the IMGT/MLA database for up-to-date listings of these genes [23]). For example, HLA-DRB1, with more than 315 known alleles, and DPB1, with more than 99 alleles (19), represent the most polymorphic protein-encoding regions of the human genome (19, 23). In addition to their role in conferring immune identity, genetic distances and correspondence analysis demonstrated that patterns of allele and haplotype distribution of class I and class II loci are racially and geographically restricted, thereby allowing their use for population and evolution studies (8, 21).

Comparative analysis of the HLA class II DR and DQ alleles and haplotypes has revealed differences in the distribution of these alleles among different ethnic groups and reflected ancient and recent admixtures between different populations (9, 21, 25). This has significant implications for analyzing human migration trends throughout history (14). Studies of the HLA origin of Arabs are scanty, but studies of related eastern and western Mediterranean groups including Turks, Iranians, Jews, Lebanese, and others suggest a common ancestry (6, 8, 24). Variation in HLA class II distribution among these populations can be explained by the admixture of the ancient population with invading and migratory societies, which may have given rise to present-day racial and ethnic groups (8).

In view of the heterogeneity of the Arab population, which comprises people of distinct ethnic backgrounds and whose origins can be classified according to their area of habitation (North Africa, Arabian peninsula, and eastern Mediterranean), the present study was aimed at elucidating the diversity in HLA class II allele distribution among two distinct Arab communities, Bahrain and Lebanon. Bahrain is an island located in the Arabian Gulf whose inhabitants derive their origin from three major roots: Jaafari Arabs, Sunni Arabs, and Iranians (4). Lebanon is located in the eastern Mediterranean, and its population consists of Christians (Catholic, Maronite, Greek Orthodox, and Coptic) and Moslems (Sunni, Shiite, and Druze). This study provides basic information for further studies of the MHC differences between Arabs of distinct origins and will serve as a reference for further anthropological studies, as well as for studies of associations between HLA and disease.

MATERIALS AND METHODS

Subjects.

The samples used in this study included 95 healthy unrelated Lebanese subjects, with a mean age of 29.8 ± 12.1 years (range, 3 to 56 years), of whom 50 were females and 45 were males. Lebanese participants were of an Arab origin (no Armenian or similar non-Arab Lebanese were included) and from the five provinces of the country, representing the major religious groups and sects. In addition, 72 healthy unrelated Bahrainis, with a mean age (± standard deviation) of 28.4 ± 13.4 years (range, 10 to 64 years), of whom 39 were females and 33 were males, were included in the study. Biochemical profiles (serum glucose, cholesterol, high-density lipoprotein-cholesterol, low-density lipoprotein-cholesterol, and triglycerides) were similar between Bahraini and Lebanese subjects (P = not significant), and none of the participants had a personal or family history of autoimmunity (in particular diabetes), hyperlipidemia, or cardiovascular disorders, thereby avoiding bias in selection of study participants. Participants (or guardians for subjects of <16 years of age) were asked to sign a consent form according to the protocol of the study, and all institutional ethics requirements were met.

HLA genotyping.

Total genomic DNA was extracted from the peripheral blood mononuclear leukocytes from study participants by the phenol-chloroform method, as is standard, and was used for PCR analysis. HLA-DRB1 and -DQB1 gene alleles were analyzed using the PCR sequence-specific priming (SSP) technique, using the SSP2L HLA class II (DRB-DQB) genotyping kit according to the specifications of the manufacturer (One Lambda, Thousand Oaks, Calif.). PCR products were analyzed on a 2.5% (wt/vol) agarose gel stained with ethidium bromide (0.5 μg/ml).

Data analysis.

Allele frequencies were determined as described previously (31) with the HLAStat 2000 software, which also computed the P values and odds ratios (OR). Linkage analysis, the nonrandom association between DRB1 and DQB1 alleles as defined by the delta (D′) coefficient, was calculated as described previously (26) also by using the HLAStat 2000 software. The frequencies of the most frequent haplotypes were determined by the maximum likelihood method, with use of the Arlequin (v. 2000) population genetics data analysis software (S. Schneider, D. Roessli, and L. Excoffier, University of Geneva, Geneva, Switzerland, 2000; http://anthro.unige.ch/arlequin).P values were corrected for the number of different alleles tested (Pc) by the Bonferroni inequality method (10), and significance was determined at P < 0.05. Wright's F_ST analogue (30) was used to assess population differentiation, calculations being made using GenAlEx software (R. Peakall and P. E. Smouse, Australian National University, Canberra, Australia, 2001; http://www.anu.edu.au/BoZo/GenAlEx/), which was also used to calculate Nei genetic distance and identity. Additional statistical analysis was performed with the SPSS version 11.5 for Windows statistical package.

RESULTS

Frequencies of the HLA-DRB1 alleles.

We found significant DRB1 allelic differences between Lebanese and Bahraini Arabs, 7 of 25 loci being significantly different (P < 0.05). When we applied the Bonferroni correction, differences were significant for four loci. These comprised DRB1*030101 (0.139 versus 0.032%; Pc = 0.033) and DRB1*160101 (0.139 versus 0.037%; Pc = 0.01), which were significantly more frequent in Bahraini than in Lebanese subjects (Table 1), and DRB1*040101 (0.179 versus 0.049%; Pc = 0.002) and DRB1*110101 (0.368 versus 0.160%; Pc = 0.003), which were more frequent among Lebanese than among Bahraini subjects (Table 1).

TABLE 1.

HLA-DRB1* allele distribution among Lebanese and Bahraini Arabs

DRB1 allelea	Bahrainib		Lebaneseb		Pc	Pcd	OR
	Allele frequency	SE	Allele frequency	SE
010101	0.014	0.010	0.058	0.017	0.055	1.371	0.243
0103	0.000	0.000	0.005	0.005	1.000	25.000	0.000
030101	0.139	0.029	0.032	0.013	0.001	0.033	4.585
030201	0.007	0.007	0.021	0.010	0.289	7.233	0.320
0310	0.021	0.012	0.005	0.005	0.192	4.811	4.087
040101	0.049	0.018	0.179	0.028	<0.001	0.002	0.171
040302	0.000	0.000	0.005	0.005	1.000	25.000	0.000
0412	0.007	0.007	0.000	0.000	1.000	25.000
070101	0.090	0.024	0.063	0.018	0.331	8.264	1.524
080101	0.035	0.015	0.011	0.007	0.122	3.056	3.470
080201	0.007	0.007	0.011	0.007	0.730	18.249	0.655
0820	0.007	0.007	0.000	0.000	1.000	25.000
090102	0.014	0.010	0.005	0.005	0.406	10.146	2.686
100101	0.076	0.022	0.032	0.013	0.029	0.736	3.245
110101	0.160	0.031	0.368	0.035	<0.001	0.003	0.286
120101	0.007	0.007	0.032	0.013	0.116	2.890	0.209
130101	0.021	0.012	0.037	0.014	0.388	9.695	0.547
130701	0.042	0.017	0.005	0.005	0.020	0.502	8.546
1327	0.035	0.015	0.000	0.000	0.009	0.228
140101	0.042	0.017	0.037	0.014	0.818	20.443	1.143
1402	0.007	0.007	0.000	0.000	1.000	25.000
1415	0.007	0.007	0.005	0.005	0.843	21.079	1.324
1416	0.000	0.000	0.005	0.005	1.000	25.000	0.000
150101	0.076	0.022	0.047	0.015	0.103	2.565	2.267
160101	0.139	0.029	0.037	0.014	<0.001	0.010	4.835

^aDRB1* alleles were assessed by PCR-SSP.

^bA total of 95 Lebanese and 72 Bahraini subjects were analyzed. Allele frequency was determined as frequency = number of an allele/total number of alleles per group.

^cDetermined by Fisher's exact test.

^dPc = corrected P value for the number of alleles tested (n = 25), calculated using the Bonferroni method.

Frequencies of the HLA-DQB1 alleles.

Similarly, significant DQB1 allelic differences were seen between Lebanese and Bahraini Arabs, four of seven loci being significantly different (P < 0.05). With the Bonferroni correction, differences were significant for only two loci. These comprised DQB1*0201 (0.257 versus 0.158%; Pc < 0.001; OR = 4.267), which was more frequent among Bahraini than among Lebanese subjects (Table 2), and DQB1*030101 (0.379 versus 0.160%; Pc < 0.001; OR = 0.245), which was more frequent among Lebanese than among Bahraini subjects (Table 2).

TABLE 2.

HLA-DQB1* allele distribution among Lebanese and Bahraini Arabs

DQB1 allelea	Bahrainib		Lebaneseb		Pc	Pcd	OR
	Allele frequency	SE	Allele frequency	SE
0201	0.257	0.036	0.158	0.084	<0.001	<0.001	4.267
030101	0.160	0.031	0.379	0.035	<0.001	<0.001	0.245
0302	0.097	0.025	0.168	0.027	0.034	0.239	0.455
03032	0.028	0.014	0.026	0.012	0.934	6.538	1.059
0401	0.035	0.015	0.053	0.016	0.423	2.959	0.634
050101	0.292	0.038	0.205	0.029	0.012	0.084	2.220
060101	0.132	0.028	0.084	0.020	0.144	1.009	1.788

^aDQB1* alleles were assessed by PCR-SSP.

^bA total of 95 Lebanese and 72 Bahraini subjects were analyzed. Allele frequency was determined as frequency = number of an allele/total number of alleles per group.

^cDetermined by Fisher's exact test.

^dPc = corrected P value for the number of alleles tested (n = 7), calculated using the Bonferroni inequality method.

Linkage analysis.

While the complete results for the two-locus analysis could not be presented due to space constraints, linkage analysis revealed that DRB1*010101 was found with DQB1*0201 among Lebanese but not Bahraini subjects (Table 3). DQB1*050101 was in linkage with DRB1*010101, -040101, -070101, and -140101 among Bahraini subjects, as were DRB1*160101-DQB1*030101 and DRB1*110101-DQB1*060101 (Table 3).

TABLE 3.

DRB1*-DQB1* linkage analysis^a

DRB1 allele	DQB1 allele	Lebaneseb			Bahrainib
		Delta	Pc	χ2	Delta	Pc	χ2
010101	050101	0.0095	0.1871	15.06	0.0372	<0.001	1.741
030101	0201	0.0845	<0.001	34.16	0.0295	<0.001	22.240
040101	0302	0.0231	0.0012	49.00	0.1115	<0.001	10.450
	050101	−0.0246	0.0542	22.42	−0.0907	<0.001	3.708
070101	0201	0.0706	<0.001	26.74	0.0367	<0.001	19.830
	030101	−0.0279	0.0481	9.189	−0.0462	0.0024	3.908
	03032	−0.0057	0.3340	10.73	0.0136	0.0011	0.933
	050101	0.0218	0.2285	7.311	−0.0296	0.0069	1.450
100101	050101	0.0435	0.0079	9.916	0.0215	0.0016	7.060
110101	0201	−0.0748	0.0016	1.804	−0.0220	0.1792	10.020
	030101	0.0943	<0.001	44.47	0.1746	<0.001	29.94
	060101	−0.0042	0.7978	11.54	−0.0554	<0.001	0.066
130101	060101	0.0184	0.0015	30.3	0.0290	<0.001	10.13
140101	050101	−0.0246	0.0542	14.2	0.0303	<0.001	3.708
150101	060101	0.0371	<0.001	30.0	0.0290	<0.001	11.53
160101	030101	−0.0110	0.5257	8.196	−0.0329	0.0042	0.403
	050101	0.0913	<0.001	14.2	0.0303	<0.001	18.6

^aAnalyzed using HLAStat 2000 software. This is a partial listing showing significant DRB1-DQB1 pairs.

^bA total of 95 Lebanese and 72 Bahraini individuals were analyzed.

^cDetermined by Fisher's exact test.

Frequencies of DRB1-DQB1 haplotypes.

The distribution of the HLA DRB1-DQB1 haplotypes in healthy Bahraini and Lebanese participants was next assessed. The frequencies of DRB1*030101-DQB1*0201 (0.1202 versus 0.0321%; P = 0.004; OR = 4.11) and DRB1*160101-DQB1*050101 (0.1318 versus 0.0379%; P = 0.007, OR = 3.97) were higher among Bahraini than among Lebanese subjects (Table 4). In contrast, the frequencies of DRB1*040101-DQB1*0302 (0.1416 versus 0.0278%; P < 0.001; OR = 0.17) and DRB1*110101-DQB1*030101 (0.3142 versus 0.1198%; P < 0.001; OR = 0.29) were higher among Lebanese than among Bahraini subjects (Table 4).

TABLE 4.

DRB1*-DQB1* haplotype distribution among Lebanese and Bahraini Arabs

DRB1/DQB1a	Lebaneseb	Bahrainb	Pc	χ2	OR	95% CId
010101/050101	0.0490	0.0135	0.165	1.928	0.28	0.08-1.37
030101/0201	0.0321	0.1202	0.004	8.253	4.11	1.54-9.86
040101/0302	0.1416	0.0278	<0.001	11.395	0.17	0.07-0.53
070101/0201	0.0420	0.0938	0.074	3.198	2.45	0.99-5.73
100101/050101	0.0280	0.0658	0.174	1.846	2.47	0.81-6.92
110101/030101	0.3142	0.1198	<0.001	16.958	0.29	0.16-0.53
130101/060101	0.0321	0.0211	0.795	0.067	0.65	0.19-2.62
150101/060101	0.0329	0.0472	0.609	0.262	1.57	0.53-4.53
160101/050101	0.0379	0.1318	0.003	9.039	3.97	1.59-9.09

^aDRB1* and DQB1* alleles were assessed by PCR-SSP, and haplotype frequencies were determined by the maximum-likelihood method.

^bA total of 95 Lebanese and 72 Bahraini subjects were analyzed.

^cDetermined by Fisher's exact test.

^dCI, confidence interval.

DRB1-DQB1 genotype distribution.

Genotype analysis revealed that the DRB1*040101/110101-DQB1*0302/030101 (12.63 versus 1.35%; P = 0.015; OR = 10.55) and the homozygous DRB1*110101/110101-DQB1*030101/030101 (7.37 versus 0.00%; P = 0.046) genotypes were observed more frequently in Lebanese than in Bahraini subjects (Table 5). Among the Bahraini subjects tested, extensive heterogeneity was seen in genotype distribution, with DRB1*070101/160101-DQB1*0201/050101 (6.76 versus 0.00%; P = 0.034) appearing to be most commonly present (Table 5).

TABLE 5.

Genotype distribution among Lebanese and Bahraini Arabs

DRB1*/DRB1*/DQB1*/DQB1*	No. (% of total)		Pa	χ2	OR (95% CIb)
	Bahraini	Lebanese
030101/110101/0201/030101	3 (4.05)	6 (6.32)	0.761	0.093	1.60 (0.39-5.64)
030101/160101/0201/050101	3 (4.05)	0 (0.00)	0.164	1.941	NAc
040101/070101/0201/0302	0 (0.00)	3 (3.16)	0.34	0.913	NA
040101/110101/0302/030101	1 (1.35)	12 (12.63)	0.015	5.950	10.55 (1.31-41.06)
040101/110101/0401/030101	0 (0.00)	5 (5.26)	0.122	2.390	NA
070101/160101/0201/050101	5 (6.76)	0 (0.00)	0.034	4.471	NA
110101/110101/030101/030101	0 (0.00)	7 (7.37)	0.046	3.984	NA
110101/140101/030101/050101	0 (0.00)	6 (6.32)	0.075	3.177	NA

^aDetermined by Fisher's exact test.

^bCI, confidence interval.

^cNA, not applicable.

HLA class II diversity.

The differences at the HLA class II DRB and DQB loci were further investigated using F_ST and Nei genetic distance and identity measurements. When class II alleles from Bahraini and Lebanese Arabs were compared, overall average genetic differentiation between populations (estimated with the F_ST, a measure of genetic differentiation over subpopulations) was 0.026 for DRB1 and 0.029 for DQB1 (Table 6). Nei genetic distance and identity were 0.287 and 0.751, respectively. Collectively, this suggested that the identity between Bahraini and Lebanese Arabs was not high.

TABLE 6.

Statistical analysis of HLA class II diversity among Bahraini and Lebanese Arabs^a

Population	No.	DRB1b			DQB1c
		Ho	He	F	Ho	He	F
Bahraini	72	0.944	0.906	−0.042	0.833	0.795	−0.049
Lebanese	95	0.832	0.815	−0.021	0.779	0.768	−0.014

^aAbbreviations: Ho, observed heterozygosity; He, expected heterozygosity; F, fixation index. The analysis of molecular variance showed a value of 0.0596 (P = 0.01).

^bF_IS, F_IT, and F_ST were −0.032, −0.005, and 0.026, respectively. F_ST was calculated with GenAlEx software as F_ST = (F_IT − F_IS)/(1 − F_IS).

^cF_IS, F_IT, and F_ST were −0.032, −0.002, and 0.029, respectively.

DISCUSSION

Analysis of HLA genes has provided invaluable tools for anthropological studies, and analysis of HLA loci and haplotypes, first by serology and later by molecular tools, has been reported for many societies. Insofar as Arabs represent a heterogeneous mixture of populations extending from the Arabian (Persian) Gulf to the Atlantic Ocean, and in view of the admixture brought about by the transmigration which trans-Arabia has witnessed throughout its history, this study was undertaken to analyze HLA class II usage among Bahraini and Lebanese Arab communities, the former situated in the eastern Arabian peninsula and the latter located in the eastern Mediterranean.

As expected, marked heterogeneity in HLA class II DR-DQ alleles and haplotypes was seen, with DRB1*030101 and DQB1*0201 being more prevalent among Bahrainis while DRB1*110101/040101 and DQB1*030101 were more prevalent among Lebanese Arabs, and also by genotype analysis, which clearly showed that the DRB1*040101-DRB1*110101-DQB1*0302-DQB1*030101 and DRB1*110101-DRB1*110101-DQB1*030101-DQB1*030101 genotypes were more frequent among Lebanese, while DRB1*070101-DRB1*160101-DQB1*0201-DQB1*050101 was more common among Bahrainis. This heterogeneity in HLA class II expression was subsequently confirmed by analysis of molecular variance and F_ST measurements and also by calculating Nei genetic distance and identity, all of which indicated that the identity between these two populations was not high.

Among the DRB1 alleles analyzed we noted that DRB1*110101, a common allele of eastern Mediterranean countries including Greece, Bulgaria (16), Turkey (25) and Slovakia (12), had the highest frequency among Lebanese. In addition to the DRB1*110101 allele which they shared with the Lebanese, Bahrainis also had unique DRB1 alleles including DRB1*030101 and DRB1*160101, which were present at significantly lower frequencies among healthy Lebanese. The similarity of the Lebanese alleles to those of the above-mentioned eastern Mediterranean countries as well as other countries, including Algeria (13) and Hungary (29), may be explained by the admixture brought about by the migration of Phoenicians (ancestors of present-day Lebanese) to North Africa and southern Europe.

Analysis of the Bahraini HLA class II alleles and haplotypes revealed some relatedness to their Lebanese counterparts and also to those of neighboring and distant ethnic groups, as exemplified by DRB1*1101 and DRB1*16011 (6, 27), which suggests the influence of these populations on this distinct Arab population. The strongest similarity that the Bahrainis had was with their related Arabian Gulf neighbors, as exemplified with DRB1*030101, which is also highly frequent among healthy Saudi (2) and Kuwaiti (3, 15) Arabs. Interestingly, the same alleles were highly frequent among Iranians (5), Indians (20), and Thais (28), an indication that Bahraini Arabs share some HLA components with populations of the Arabian Peninsula and the Indian subcontinent.

Perhaps the main feature of Arabs is the high diversity of DR-DQ associations. Most of the DRB-DQB associations reported here for both Bahraini and Lebanese Arabs were also seen in other regions, though at different frequencies. These included DRB1*0701-DQB1*0201, which was common among Yemenite Jews (17), Moroccans (17), and Iranians (5); DRB1*0401-DQB1*0302, reported for Jordanian Arabs (24); and DRB1*1501-DQB1*06, which was reported for Tunisians (1). There were also certain haplotypes which were completely missing among these two Arabic-speaking populations but reported for other ethnic groups, including DRB1*0803-DQB1*0601 for Filipinos (11), DRB1*0401-DQB1*0401 for Koreans (22), and DRB1*16-DQB1*0602 for Japanese (7).

Data presented here, in defining the HLA profile among (eastern Mediterranean) Lebanese and (Arabian Peninsula) Bahraini Arabs, point to differences in the origins of these two distinct Arabic-speaking communities, brought about possibly by the admixture of the original inhabitants with neighboring and distant populations. These results provide information that can be used for future anthropological studies and also in the analysis of disease susceptibility and organ transplantation (18, 22).