Eastern Mediterranean Health Journal | Past issues | Volume 3, 1997 | Volume 3, issue 1 | Major histocompatibility class I antigens in the Lebanese population

Major histocompatibility class I antigens in the Lebanese population


PDF version

Nuha Nuwayri-Salti and Monique Shaya

ABSTRACT: Except for two reports on Lebanese immigrants, there have been no studies on the major histocompatibility (MHC) antigens in the Lebanese population. We describe the frequency and distribution of MHC class I antigens present in the A, B and C loci based on data obtained from 200 healthy unrelated individuals from different parts of Lebanon. The highest gene frequencies were as follows: A2 (24.8%), B35 (17.9%) and Cw4 (18.6%), making this haplotype the commonest. Comparison of genetic distances revealed a pattern closer to the Caucasoid population than to the Mongoloid, Oriental or Black populations.

Les antigènes de la classe I du complexe majeur d'histocompatibilité dans la population libanaise

RESUME: A l'exception de deux rapports sur des immigrants libanais, il n'y a eu aucune étude sur les antigènes du complexe majeur d'histocompatibilité dans la population libanaise. Nous décrivons la fréquence et la répartition des antigènes de la classe I du complexe majeur d'histocompatibilité présents dans les loci A, B et C en fonction des données obtenues auprès de 200 sujets sains n'ayant aucun lien de parenté et originaires de différentes parties du Liban. Les fréquences géniques les plus élevées étaient les suivantes: A2 (24,8%), B35 (17,9%) et Cw4 (18,6%), ce qui rend cet haplotype le plus courant. La comparaison des distances génétiques a mis en évidence un schéma plus proche de la population caucasienne que des populations mongoliennes, orientales ou afro-américaines.


Lebanon and its neighbouring Arab countries are situated at a crossroads between Europe, Africa and Asia, and historically have been invaded and populated by different peoples originating from these three continents. Although the predominant racial constitution is Caucasoid, it is still poorly defined as there have been very few anthropomorphic population studies on this region. Another feature of the population in this part of the world is the high frequency of consanguineous marriages, which continue to be common. Although polygamy occurred in the past, it has become rare in the past half century.

Data on histocompatibility antigens in patients and in population groups are becoming increasingly important, not only in the ever-expanding field of organ transplantation, but also in the area of human histocompatibility leukocyte antigen (HLA) association with a number of diseases. The latter vary in pathophysiology, from the different types of autoimmune disorders (e.g. ankylosing spondylitis, insulin-dependent diabetes mellitus, Graves disease) to some heritable disorders such as 21-hydroxylase deficiency. Information on HLA antigens can be utilized to assess genetic risk in such disorders.

Only two studies have been reported on Lebanese expatriates living in either Australia or the United States of America [1,2], and there have been a few reports about different Arab communities, mainly in Egypt [3-6] and Israel [7]. Beside these, there have been no published studies on the HLA profiles of the Lebanese and their immediate Arab neighbours. In this work, we describe antigen and gene frequencies of the A, B and C loci of the major histocompatibility (MHC) class I gene in a sample of 200 healthy Lebanese individuals. The results, including the haplotypes and linkage disequilibria, are compared to patterns reported for other populations in the world.

Subjects and methods

The population sample studied was chosen randomly from healthy individuals from different parts of Lebanon belonging to the two major religions. Their ages ranged from 20 years to 40 years and there were 100 males and 100 females.

Ten millilitres of venous blood were collected into tubes containing 10 000 IU of preservative-free heparin and, within 15 minutes, isolation of the lymphocytes was started using Ficoll-Hypaque density gradient as described by Boyum [8]. HLA class I antigens were determined immediately using commercially available trays with mono- and polyclonal antisera purchased from Hoechst Laboratories (Behring, Behringwerke AG, Marburg, Germany). The usual grading scale of 1 to 8 was used, in which 8 represented the positive control with nearly total cell kill and 1 was the negative control with the lowest number of dead cells (normally not more than 1%-2%). For the C locus, the antibodies included on the trays were restricted to the first five only; the rest were not available.

All analyses were made using the SPSS program with Quatro-Pro spreadsheet software. Phenotype frequencies were calculated by direct count. Gene frequencies were calculated using the formula

Gf = 1 - √(1-Af )

where Gf is the gene frequency and Af is the antigen frequency, based on the assumption of Hardy-Weinberg equilibrium and as described by Ishawar et al. [9].

The differences in phenotype frequencies were tested by contingency c2 and, wherever possible, by Fisher exact probability. The linkage disequilibrium parameters between the three loci were calculated using the following formula given by Piazza [10]:

D(i,j,k) = p(i,j,k) - p(i)p(j)p(k) -

p(i)D(j,k) - p(j)D(i,k) - p(k)D(i,j)

where p(i,j,k) is the haplotype frequency of alleles i, j and k.

Genetic distances between our population sample and four other populations were computed using the following formula given by Cavalli-Sforza et al. [11] and Bodmer et al. [12]:


where P(i,j) and P(i,k) are the gene frequencies in the populations of alleles j and k respectively.


The phenotype and gene frequency distribution in the study sample are summarized in Table 1. In Table 2, comparison is made of the results on the Lebanese population with data published on Caucasoid, Black and Oriental populations [13]. Table 3 lists the genetic distances of the different antigens between the Lebanese sample and the Caucasoid, Black, Indian and Mongoloid populations.

The haplotype frequency and linkage disequilibria expressed by delta values [10] are shown in Table 4. A comparison of our findings with those published on Lebanese immigrants [1,2] and on other Arab populations [7] is shown in Table 5


The results show that there was no antigen whose frequency was high enough to be considered a distinguishing feature of the Lebanese population (Table 1). Comparison with other populations revealed that, for the antigens coded by the A locus gene, there was no significant difference. As for the common antigens at the B locus, with the exception of antigens B35 and B41, which had the highest frequency in our sample, they were generally closer to the reported pattern of HLA class I markers in the Caucasoid population than to the reported frequencies in the Black and Oriental populations (Table 2). This conclusion is also supported by the data on genetic distance (Table 3).

On the other hand, a review of the antigens which were rare or of zero frequency in the Lebanese people (Table 2) revealed that while many of them are universally of low prevalence, the frequency of some seem, at face value, to be lower than that found in the rest of the groups reported. However, this apparent difference in the frequency between our population and the populations shown in Table 2 is misleading since the values reported for several loci such as A9, B5, B12, B16 and B21 for the Caucasoids, Blacks and Orientals represent the sum of the frequencies of their individual epitope components. In our study, we distinguished between phenotypes exhibiting portions of an antigen (epitope) and others expressing the entire antigen molecule. As an illustration, the frequency of A9 for the Caucasoid population is reported as (frequency of A23 + frequency of A24 = 22.4). In our study, the value considered for A9 represents situations where A9 alone was present (to the exclusion of both A23 and A24 epitopes). Whether or not this frequency represents the third epitope for A9 (A2403) is not yet clear. Hence, if we were to follow the same procedure, then our list of rare antigens would be much shorter and would be restricted to B15 and B40 (including epitopes B62, B63 and B60, B61 respectively).

These findings may prove to be of particular relevance to some HLA-associated diseases. In fact, determination of the class I antigens in patients with Graves disease did reveal a bias in favour of one of these rare antigens (B39), which in these patients exhibited an odds ratio of 9.6 [14].

The prevalent HLA haplotypes in the Lebanese population (Table 4) do not show any unusual combinations.

Some minor differences in the A and B loci were observed between our results and those previously reported on Lebanese immigrants [1,2] and other Arab communities [7] (Table 5). These could be partly due to a possibly higher rate of consanguinity in the immigrant communities, which tend to be more close-knit.


The authors would like to thank Ms Samia Mroueh Abul-Hassan and Ms Fatmeh Jaafar for their technical assistance, Ms Wissal Asaad and Mr Shady Hawwa for their computational work and Ms Dina Tannir for revising the manuscript. This work was supported by a generous grant from the Ousseimi Foundation and the Diana Tamari Sabbagh Foundation.


  1. Bashir HV et al. A study of the HLA system in 100 expatriate Lebanese domiciled in Sydney. In: Dausset J, Colombani J, eds. Histocompatibility testing. Copenhagen, Munksgaard, 1972:133-8.
  2. Noreen H, Furcht L, Yunis EJ. HLA antigens of four different North American ethnic groups. In: Kissmeyer-Nielsen F, ed. Histocompatibility testing. Copenhagen, Munksgaard, 1975:193.
  3. Kamel MA et al. Population analysis of HLA-A and B genetic markers in one thousand Egyptians. Medical journal of Cairo University, 1984, 52(3):507-10.
  4. Hafez M, El-Shennawy FA. HLA-antigens in the Egyptian population. Forensic science international, 1986, 31(4):241-6.
  5. Ghaly I, Kamel M, Anwar O. HLA antigens in families of children with insulin- dependent diabetes mellitus. Egyptian journal of pediatrics, 1985, 2(1-2):41-58.
  6. El-Tayeb et al. HLA and hemophilia `A' in Cairo citizens. Egyptian journal of hematology, 1981, 6(3-4):315-22.
  7. Zamir R et al. HLA antigens in the Arab population of the Gaza strip. Israeli medical journal, 1979, 15:14-6.
  8. Boyum A. Isolation of mononuclear cells and granulocytes from human blood. Isolation of mononuclear cells by one centrifugation, and of granulocytes by combining centrifugation and sedimentation at 1 g. Scandinavian journal of clinical laboratory investigation supplement, 1968, 97:77-89.
  9. Ishawar V, Mathews AR, Chaudhary SS. Histocompatibility antigens (A, B and C loci) in Indians in Delhi. Indian journal of medical research, 1983, 78:233-8.
  10. Piazza A. Haplotypes and linkage disequilibrium from three-locus phenotypes. In: Kissmeyer-Nielsen F, ed. Histocompatibility testing. Copenhagen, Munksgaard, 1975:923.
  11. Cavalli-Sforza LL, Bodmer WF. The genetics of human populations. San Francisco, Freeman, 1970.
  12. Bodmer WF, Pickbourne P, Piazza A. Population analysis. In: Histocompatibilty testing, Copenhagen, Munksgaard, 1978:259.
  13. Schiebl B. HLA-typing: problems and solutions. In: Biotest Mitteilungen, special edition No. 7. Dreieich, Germany, Biotest, 1987:12-13.
  14. Nuwayri-Salti N et al. Class I immunogenetic markers in Lebanese patients with Graves disease. Human immunology,1996, 47:33.