Human leukocyte antigen ( HLA ) polymorphism and type 1 diabetes in the Moroccan population

Insulin-dependent diabetes mellitus or type 1 diabetes is an autoimmune multifactorial disease which has a great socio-economic impact. In Morocco, less is known about the contribution of Human leukocyte antigen (HLA) alleles to type 1 diabetes susceptibility. Our study focused on evaluating the distribution of class II HLA genes among Moroccan patients presenting type 1 diabetes coming from different regions of the country. A total of 76 patients diagnosed with type I diabetes were compared to a group of 248 healthy controls matched by age, sex, and ethnic origin. The HLA class II (DR and DQ) typing of patients and controls was performed using polymerase chain reaction sequence-specific primers (PCR-SSP). Ambiguous typings were reanalyzed by Luminex technology. We found a significant association between HLA-DRB1*03 (p = 0.0001), DRB1*04 (p = 0.0001), DQB1*02 (p = 0.0001, OR = 4.4) and DQB1*03 (p = 0.002) alleles frequencies and the susceptibility to diabetes. The HLADQB1*06 allele was significantly decreased in patients (p = 0.008). DRB1*03-DQB1*02, DRB1*04DQB1*03 and DRB1*04-DQB1*02 haplotypes induce susceptibility to type 1 diabetes while DRB1*13DQB1*06 and DRB1*15-DQB1*06 haplotypes seems to confer a protective effect in our population. These data highlights that Moroccans diabetic patients do share a series of traits with diabetics of other origins but also have specific characteristics.


INTRODUCTION
Insulin-dependent diabetes mellitus or type 1 diabetes (T1D) is the third most common disorder of childhood, affecting between 1 and 4 in 10 000 individuals before the age of 30 with an incidence rate that appears to be increasing in all studied countries (Balafrej, 2003;Behrooz-Alizadeh et al., 2008;Patterson et al., 2009).
T1D is a multifactorial autoimmune disease resulting from the destruction of insulin-producing beta cells in the pancreas by autoreactive T cells.Such destruction results in clinically insufficient insulin production and in a dysregulation of glucose homeostasis (Atkinson and Maclaren, 1994).Symptoms accompanying the disease include high level of glucose in patient's blood and urine, T cell infiltration into the pancreas and the presence of autoantibody against insulin and beta cell antigens (Parry and Brooks, 2008).
Genetically, the occurrence of diabetes mellitus depends on several genes and mainly on the one situated at the level of the major histo-compatibility complex (MHC) (Behrooz-Alizadeh et al., 2008).Crossethnic studies have now confirmed that three Human Leucocytes Antigen (HLA) class II genes, DRB1, DQA1 and DQB1, contribute directly to T1D (Abid Kamoun et al., 2002;Nepom and Erlich, 1991;Parry and Brooks, 2008).
The role of HLA in T1D was first indicated by associations in Caucasians with the HLA-DR3 and -DR4 antigens encoded at the DRB1 locus.Subsequently, DQB1 and DQA1 genes were shown to be more strongly associated with T1D, and the DQ molecule was considered as the primary susceptibility factor.Several studies have clearly demonstrated that both DQ and DR molecules influence T1D susceptibility (Park et al., 1998).A protective effect of HLA class II molecules was also demonstrated.Transracial studies have reported that the alleles involved in T1D appear to be different in various populations (Aribi et al., 2004;Chen et al., 1999;Platz et al., 1981;Sheely, 1992;Thomson et al., 1988;Valdheim et al., 1989).
In Morocco, very few investigations have been undertaken to study the impact of genetic background on the risk to develop T1D when the average prevalence is about 2% (Belkhadir and El Alaoui, 1993).Until now, available studies have been conducted in isolated populations from particular origins like Berber and Jewry (Izaabel et al., 1996;Witt and al., 1994), whereas the present population comprises different ethnic groups (Arabic, Berber and African).Morbidity and mortality are increased because of late diagnosis and lack of regular checks.
The aim of this study was to investigate the relationship between T1D and HLA in a sample of Moroccan population, from diverse regions of the country, and to determine the predisposing and the protective alleles to the disease.

Patients and controls
A total of 76 patients consecutively diagnosed (before fifteen years old) with T1D were recruited from the department of endocrinology of Ibn Sina University Hospital of Rabat during the period 2005 to 2011.
The sample consisted of 37 women and 39 men between 2 and 64 years (mean age 21 years).The patients were compared with 248 healthy unrelated control subjects and matched by age, sex and ethnic origin.The mean age of the control population was 29 years and the sex ratio (males/females) was 127/ 121.These controls were either volunteers as bone marrow-or kidney-donors or just volunteers taking part in the study.
The populations compared in the present study were from diverse regions of Morocco.

Human leukocyte antigen (HLA) typing
Blood samples were collected and analyzed at the Ibn Sina University Hospital of Rabat and analyzed in the laboratory of Immunology of this hospital.
DNA was first extracted from the buffy coat fraction of blood samples using a commercial kit (Qiagen).The HLA typing of the HLA class II alleles (-DRB1 and -DQB1) was performed by polymerase chain reaction-sequence specific primers (PCR-SSP) technology using micro generic HLA DNA typing trays (One Lambda) in accordance to the manufacturer's protocol.Ambiguous typings (blanks, not clearly identified alleles) were reanalyzed by Benseffaj et al. 16127 Luminex technology polymerase chain reaction sequence-specific oligonucleotide (PCR-SSO).

Statistical analysis
Allele, genotype and haplotype frequencies, p values, odds ratio (OR) and 95% confidence interval (CI) were calculated using the commercial SPSS software package for Windows, version 10.0.The frequencies in patients and controls were compared with Fisher exact test.A p value less than 0.05 was considered statistically significant.Corrected p values (pcorr) were calculated using the Bonferoni inequality method.In all cases, pcorr refers to corrected p values which were obtained by multiplying crude p values by the number of comparisons (Millot, 2009).
The distribution of the DQB1 alleles reveals that the frequencies of DQB1*02 and DQB1*03 in T1D patients were significantly higher than that in the control group with a respective Odds ratio of 4.4 (p < 0.0001) and 2.4 (p = 0.002).Conversely, the frequency of the HLA-DQB1*06 allele was significantly decreased in the controls compared to diabetes patients (p < 0.0001) and DQB1*05 frequency was also reduced by half in patients.

DISCUSSION
Morocco is a Maghrebian country situated in North Africa, between the Mediterranean world, the middle and near east and Black Africa.Thanks to its strategic location, Morocco has been a "converging meeting land" since antiquity.The present Moroccan population comprises different ethnic groups (Arabic, Berber, Africa) with a strong influence of various immigrations, mainly  Mediterranean.All these populations probably contributed to the Moroccans genetic pool (Canosi et al., 2010) The HLA alleles and haplotypes distribution among Moroccan T1D population is not well defined because it has not been extensively studied.In this report, we analysed HLA class II polymorphism in Moroccan patients affected by T1D from diverse region of Morocco and we compared them to healthy controls of the same ethnic origin.
Distribution of alleles and haplotypes observed in our controls are in accordance with previous Moroccan studies (Brick et al., 2006;Oumhani et al., 2002;Canosi at al., 2010).Our data confirm the well-known positive association between T1D and HLA-DRB1*03 and -DRB*04.Furthermore, the high frequency of the heterozygous combination DRB1*03/DRB1*04 in patients (30.3%) indicates a synergistic effect on the risk to develop the disease.Indeed, reviews carried out during the last years revealed that 80 to 95% of the Caucasian insulin-dependent diabetics belong to the group DRB1*03 and/or DRB*04 against 40 to 50% from the control group (Abid Kamoun et al., 2002;Atkinson and Maclaren, 1994;Bouhours-Nouet and Coutant, 2005;Chen et al., 1999;Thomson, 2001).It is interesting to note that the DRB1*04 frequency observed in our study was significantly associated to type 1 diabetes (p < 0.0001) even though this allele was highly represented in the Moroccan healthy population.
The associations characterizing Caucasian populations (DR3, DRB4) have neither been found in Africans whose susceptibility to T1D is associated with DR7 and DR9, nor was it found among Asian populations (DR4 and DR9) (Danze et al., 1997).HLA-DRB1*09 was low in both healthy controls and patients.This allele is very rare in North African populations but its association with T1D susceptibility has previously been reported in Moroccan diabetics all originating from the Souss region and mostly of Berber origin (Izaabel et al., 1996).In contrast, DRB1*09 is associated with the disease in Japanese and Korean and the haplotype DRB1*09/DRB1*09 is a very high risk genotype in Asian populations (Ikegami et all 2007;Kanga et al., 2004;Park et al., 1998).The differences among alleles associated with T1D in Moroccan and Asian populations can be explained by the difference in the presence or absence of alleles in each population (Aribi et al., 2004).
The comparison of the allele frequencies in Moroccan T1D and controls revealed a protective role for DQB1*06, DRB1*15 and DQB1*05 (to lesser degree) since their frequency were higher in the controls than in the patients.A similar observation had already been reported (Bouhours-Nouet and Coutant, 2005;Chen et al., 1999;Caillat-Zucman, 2001).
Interestingly, our study revealed that the presence of DQB1*06 on DQB1*02/DQB1*06 and DQB1*03/DQB1*06 genotypes has a protective effect, even in the presence of the predisposing DQB1*02 and DQB1*03 alleles.This confirms the finding of previous studies (Parry and Brooks, 2008) in which DQB1*06 conferred a strong dominant protection even when combined with alleles with high risk.
Studies conducted on different populations showed that several HLA-DRB1-DQB1 haplotypes are associated with a spectrum of different disease risks, ranging from strong susceptibility to almost complete protection (Behrooz Alizadeh and Bobby, 2008).Our results shows that the haplotypes DRB1*04-DQB1*03 and DRB1*03-DQB1*02 increase the susceptibility to T1D while the haplotype DRB1*15-DQB1*06 protect against the disease.These results were close to those reported in most Caucasian, Asian and Black populations (Abid Kamoun et al., 2002;Behrooz-Alizadeh and Bobby Koeleman, 2008;Park et al., 1998).In our series, DRB1*13-DQB1*06 haplotype protects against the disease and seems to be a feature of the Moroccan population.The protective effect of the DRB1*11-DQB1*03 haplotype was not shown in our population contrasting with Tunisian, French, Greek and Japanese populations (Abid Kamoun et al., 2002).
It would be interesting to confirm the study by increasing the sample sizes.Moreover, these results may serve as a reference for future studies to identify specific alleles of predisposition to the disease in our population.

Table 1 .
HLA-DRB1 and allele frequencies in Moroccan type 1 diabetes patients and healthy controls.
AF = Allele frequency; pcorr = P value corrected for the number of comparisons; OR = odds ratio (only statistically significant OR are presented ) CI = confidence interval.

Table 2 .
The DRB1* genotypes of type 1diabetes patients and healthy controls.

Table 3 .
The DQB1*genotypes of type 1diabetes patients and healthy controls.

Table 4 .
Most frequent HLA DR-DQ haplotypes in type 1 diabetes patients and healthy controls.
HF, Haplotype frequencies ; OR , odds ratio (only statistically significant OR are presented) ; CI , confidence interval.