Human immunodeficiency virus (HIV) infection remains a significant public health concern due to its high incidence and mortality. According to World Health Organization in 2022, there were 1.3 million new infections recorded, and 630,000 deaths due to opportunistic infections among the 39 million people living with HIV [1] .

Viral infections lead to a natural selection of genes crucial for species survival. This selection predominantly influences the genes related to the immune system, particularly the Human Leukocyte Antigen (HLA) system and the immunoglobulin-type receptors of killer cells (KIRs), which are among the most numerous and diverse genetic loci in the human genome [2] [3] . The HLA region is known for its high polymorphism, hosting a large number of allelic forms [4] . In December 2023, the Immuno Polymorphism Database (IPD)-international ImMunoGeneTics (IMGT)/HLA database had registered 38,909 HLA alleles and related genes [5] . Several mechanisms have been proposed to explain the generation and maintenance of HLA diversity, emphasizing heterozygosity as the primary mechanism and rare alleles as an alternative means [6] [7] . Heterozygous individuals at HLA loci can present a broader range of peptides derived from pathogens, resulting in a more diverse repertoire of cytotoxic T lymphocytes (CTL) capable of resisting a wide variety of infectious pathogens [2] [7] . Consequently, the ineffective T-cell response in some viral infections, such as HIV, suggests that certain HLA alleles may present viral epitopes more effectively to T-cells than others [8] .

Among people infected by HIV-1 there are long-term non-progressors (LTNP) who can be divided into elite controllers and virus controllers [9] - [11] . It has been shown in this LTNP that HLA-B*57 and B*27 are involved in infection control and reduced risk of transmission [11] . The dominant HIV-1 subtypes and HLA genotype distribution are different all over the world, and the HIV-1 and host HLA interaction could be specific to individual areas [12] . Thus, for ethnicities in which HLA-B*57 and HLA-B*27 are uncommon, other HLA-B alleles play protective roles. For example, HLA-B*6701 and HLA-B*52:01 are protective alleles found in Japanese individuals with HIV infection, CTLs restricted by these alleles control HIV-1 infection progress [13] . Similarly, HLA-B*58:01 and HLA-B*35:05 mediate protective effects in Africans and Thai patients infected with subtype CRF01 AE [14] . Furthermore, several HLA alleles have been identified as either resistance or susceptibility alleles to HIV-1 infection in black population (African American, Botswana/Zimbabwe and South African populations). Protective alleles include HLA-B*13:02, B*42:01, B*44:03, B*57:02, B*57:03, B*58:01, B*81:01, A*74/74:01 [15] - [21] . The susceptibility alleles encountered in these populations are HLA-B*53:01, B*08/08:01, B*18/18:01, B*45/45:01, B*51:01, B*58:02 [15] [17] [19] - [21] . However, most of the data described above on HLA association with disease progression were derived from studies in Caucasian and Southern African populations, where HIV-1 subtype B or C are predominant. In addition, most studies have focused only on class I HLA polymorphism. Nevertheless, some evidence suggests cross-presentation of class II HLA molecules in resistance or susceptibility to HIV-1 infection. For example, HLA-DRB1*13:03, DRB1*01:01 and DRB1*15:01 were associated with reduced burden in HIV-1 infection by clades B and C.

In West Africa, particularly in Ivory Coast, data on HLA polymorphism and association with a few pathologies, such as HIV-1 where the subtype CRF02 AG is predominant, are very limited. This is evidenced by the few studies that have been conducted in the general population.

An HLA polymorphism study by Ellis et al. in general population was limited to the HLA-B locus [22] . This study showed that the most frequent alleles were HLA-B*53:01 (22.7%), B*45:01 (9.1%), B*15:03 (8.0%), B*07:05 (5.7%), B*15:10 (5.7%) and B*35:01 (5.7%).

Jennes et al. identified HLA-B and HLA-C polymorphism in sex workers and showed that KIR/HLA interactions also influence resistance to HIV infection [23] .

The purposes of our study were to determine the class I and II HLA polymorphism in an Ivorian population and to evaluate whether alleles are associated to resistance or susceptibility to HIV-1 infection.

The main function of the HLA system is the presentation of extra- or intracellular peptides to T lymphocytes for the control and/or elimination of a range of pathogens. HLA system genes are associated with protection or predisposition to many diseases, including autoimmune diseases such as acute rheumatic arthritis, ankylosing spondylitis, juvenile diabetes, and viral diseases like HIV-1. Data on HIV-1 interaction and the HLA system have been accumulated in populations where HIV-1 subtypes B and C predominate, but have not been sufficiently collected in West Africa, where the HIV-1 subtype CRF02 AG is predominant [30] [31] . Also, since the polymorphism of these HLA genes differs according to the ethnic composition of the population, we carried out this study in Ivory Coast using a new generation sequencing technique. The purposes of this study were to determine HLA polymorphism in our population and to evaluate whether alleles are associated to resistance or susceptibility to HIV-1 infection.

This study allowed us to determine not only the HLA polymorphism class I but also for the first time in Ivory Coast that of HLA class II.

We detected 20 HLA-A alleles, 32 HLA-B alleles and 21 HLA-C alleles on respectively 3094 HLA-A alleles, 3866 HLA-B alleles and 2616 HLA-C alleles identified to date. Overall, the alleles identified were found in both sub-Saharan and African-American populations with intermediate frequencies. However, some African population-specific alleles such as HLA-A*74:01 and HLA-B*53:01 but also HLA-C*04:01 were much more represented confirming the other data according to which the frequency of these alleles is significantly correlated with the prevalence of Plasmodium falciparum. Using several approaches, including linear modelling on various genetic, geographical and environmental parameters, Sánchez-Mazas et al. have shown that HLA-A*74 and HLA-B*53 are associated with malaria protection [32] . Ivory Coast is malaria endemic area, our data are in agreement with these observations. As regards to HLA class II, very few studies have been listed in sub-Saharan Africa. We identified 20 on 1719 DRB1 alleles, 10/54 DQA1 alleles, 15 DQB1/777 alleles, 6 DPA1/39 alleles and 16 DPB1/520 alleles found in The Allele Frequency Net data base. Overall, the HLA class II alleles observed in our study were also found in the sub-Saharan and African-American population with relatively low frequencies and overlapping according to populations. Although HLA-DPA1*01:104, HLA-DPA1*02:12 HLA-DPB1*417:01 were not found in the sub-Saharan and African-American population of the IMGT database, it should be noted that the alleles DPB1*417:01 and HLA-DPA1*02:12 were identified among African-Americans in 2018 [33] [34] .

The most frequently observed alleles in HLA Class II were HLA-DRB1*07:01, DPA1*02:01, DPB1*01:01, DQA1*01:02, DQB1*02:02. These alleles are also found in the sub-Saharan and African-American population [35] . In Gambia, the DRB1*13:04 variant, observed at a very low frequency in our population, is the most common DRB1 allele [36] . In Kenya, the frequently identified Class II alleles were DRB1*11:01 (0.1165), DPA1*01:03 and DPB1*01:01 (0.2345), DQA1*01:02 (0.3103), DQB1 *03:01 (0.2179) [37] . In South Africa DRB1*15:03 (0.081), DQA1*05:02 (0.258) and DQB1 *03:19 (0.0087) are the most frequent class II HLA alleles [38] . These observations confirm the genetic diversity in Africans with different allelic frequencies according to populations and ethnic or racial origin [39] .

The optional and exclusive alleles DRB3, DRB4 and DRB5 were also encountered. Among the most expressed alleles DRB3*02:02 is known to be a risk factor in autoimmune diseases such as type 1 diabetes. Associated with DRB1*15:01, allele not identified in our study, DRB3*02:02 is a risk factor for idiopathic membrane nephropathy [40] . Also, the relatively high frequencies of these exclusive alleles could pose difficulties in finding a compatible donor.

Susceptibility and resistance to different pathologies are related to HLA polymorphism. In view of the distribution of this polymorphism in our study population, we then searched for the association between HLA and disease with the particular case of HIV. Polymorphism analysis by serological HIV-1 status did not identify any statistically significant alleles associated to HIV-1 infection. This result may be related to the size of our sample, which is one of the limitations of this study. Indeed, the sample size was small and did not provide sufficient statistical power to detect all truly significant associations. We observed alleles present only in subjects living with HIV. Like the literature data, these alleles could be considered as susceptibility alleles. However, it is described in the literature of controversies as to the designation of an allele as predisposing or not. Thus, several authors have defined HLA-B*35 alleles as being associated with susceptibility and rapid progression. This allele is relatively predominant at the HLA-B locus in the West African population [41] , including our cohort, and associated in Caucasians and African Americans with higher viral loads and accelerated disease progression [8] . A study showed a differential association of clade-specific B*35 with the course of HIV-1 disease. In addition, an association of B*35 with higher viral loads was not observed in African cohorts infected with HIV-1 clade C [12] . In a Ghanaian cohort mainly infected with the subtype CRF02 AG of HIV-1, no significant association between HLA-B*35 and higher viral loads or lower CD4 levels was observed either [31] .

In order to identify the possible presence of resistance alleles specific to our population, we analyzed HLA polymorphism in serodiscordant couples. This polymorphism analysis did not identify any statistically significant resistance alleles. However, we found specific alleles exclusively in HIV-negative. Among them, we have alleles described as protective, including HLA-B*57:03 associated with HIV control in African patients in several studies [11] [42] [43] . These studies showed that HLA-B57-mediated HIV control is associated with strong CD8+ T-cell responses at the onset of HIV infection, to the expansion of immunodominant T-cell clones and recognition of preserved HIV Gag epitopes. In addition to the T-cell mediated response, HLA-B57 may also present HIV peptides to natural killer (NK) cells [9] . Other protective alleles were also observed, including HLA-A*02:05, HLA-B*27:05 and HLA-B*44:10 [44] [45] . These observations confirm that some Class I HLA associations with control of HIV-1 infection exceed the limits of race and viral subtype while others appear to be confined within either of these limits [44]

Another particularity among HIV-negative partners was the presence of HLA-B*82:01. This black population-specific allele, poorly reported in studies [35] and not belonging to any of the nine serotypes defined by Sidney et al. [46] could be associated with protection by the rare allele mechanism. It is also important to note that HLA-C*07:02 was associated with increased susceptibility to HIV-1 in the Pumwani sex worker cohort in Kenya and remained HIV-free for over 30 years despite high-risk sex work [37] . In contrast, this allele was observed exclusively in HIV-negative exposed partners in our study. This difference may be related either to a various sensitivity of ethnic groups in relation to environmental factors that can reorient the expression of certain alleles and allow the adaptation of the immune system to specific pathogens.

There is very little data on polymorphism of HLA class II alleles in West Africa. In the Pumwani cohort, a strong protective effect against HIV seroconversion was observed with HLA-DRB1*01, and in particular DRB1*0102, suggesting that restricted CD4+ DRB1 cells may play a role in HIV protection [44] . In our study, DRB1*11:01, DRB1*12:01, DRB1*13:03; DQB1*03:09, DPB1*04:01, DPB1*417:01 and DPB1*584:01 identified only in seronegative partners could have a protective effect. Indeed, the protective activity of HLA-DRB1*13:03 in HIV-1 infection has been elucidated in a cohort of 426 antiretroviral therapy-naive, HIV-1 clade C–infected black female South Africans [45] . On the other hand, we observed alleles at low frequency, like HLA-DRB1*13:04 and HLA-DPB1*11:01, only present in HIV-1 positive partners suggesting their role in susceptibility to HIV infection.

In our study, we found alleles that have not yet been described as involved in resistance or susceptibility to HIV-1 infection. However, some of these alleles play a protective role against certain autoimmune diseases and in effective responses to other pathogens [47] [48] .

This study presents for the first time high-resolution typing data of class I and II HLA alleles in Ivory Coast.

The data collected provides insight into HLA Class I and II diversity in the Ivorian population and contributes to ongoing efforts to fully understand HLA diversity among Africans and to create a resource for future studies. Also, we found alleles only in HIV-positive people who could be considered as susceptibility alleles and alleles present only in serodiscordant couples exposed to HIV but who do not make the disease that could be considered as resistance alleles.

Data will be made available on request.

The authors thank all technical staff of CeDReS, CNTS and EFS Marseille for their support during experimental work. Furthermore, we would like to thank the couples for their participation.

Studies involving human participants were reviewed and approved by the National Research Ethics Committee of Ivory Coast