Celiac disease (CD) is an autoimmune disorder, in which the key genetic factors (human leukocyte antigen [HLA]-DQ2 and HLA-DQ8), the autoantigen involved (tissue transglutaminase [tTG]), and the environmental trigger (gluten) have been well-defined.1 The condition is characterized clinically, by different intestinal and extraintestinal manifestations; histologically, by damage to the intestinal mucosa; and serologically, by the presence of anti-tissue transglutaminase 2 (anti-tTG2), anti-endomysial, and/or deamidated gliadin peptide antibodies.2

As is well-known, the main genetic determinant in CD involves HLA molecules, specifically the HLA-DQ2 and/or HLA-DQ8 heterodimers.3 The HLA-DQ2 heterodimer is composed of an α subunit and a β subunit, encoded by the HLA-DQA1*05 and HLA-DQB1*02 alleles, respectively, differentiating two types of DQ2 heterodimers: DQ2.5 (DQA1*0501/DQB1*0201) and DQ2.2 (DQA1*0201/DQB1*0202). The HLA-DQ8 heterodimer is composed of an α chain and a β chain, encoded by the HLA-DQA1*0301 and HLA-DQB1*0302 alleles, respectively.4 Those HLA-DQ molecules play a key role in CD due to their physicochemical properties and the binding of specific peptides that have been deamidated by the tissue transglutaminase 2 enzyme (tTG2), triggering an innate and adaptive immune response against gluten. Neutrophils, eosinophils, mastocytes, and complement proteins are activated in the innate immune response, potentially contributing to the pathogenesis of the disease. In the adaptive immune response, peptides derived from gluten and modified by tTG2 are presented by antigen-presenting cells in the mesenteric lymph nodes to CD4 + T cells, in the context of HLA-DQ2 and/or DQ8. The resulting Th1-type response leads to IFN-γ production and intestinal inflammation. Chronic inflammation conduces to the expansion and persistence of γδ T cells, which also contribute to IFN-γ production. In addition, gluten peptides induce IL-15 and stress molecule expression on enterocytes. The elevated IL-15 levels promote a NK-like phenotype in CD8 + T cells, directly contributing to enterocyte death. Furthermore, plasma cells are also abundant in the lesion, some of which express the immunodominant gluten peptide and are induced to secrete antibodies that bind to tTG2 and other targets. However, other less well-characterized mechanisms may participate in lesion development.5 Considering that CD is a complex disorder that involves both genetic and environmental factors, as well as the deregulation of the immune response, the aim of our study was to evaluate the frequency of the HLA-DQB1 alleles in individuals with symptoms suggestive of CD and define the diagnostic markers of the condition in a Venezuelan population.

The present study is the first to evaluate the frequency of HLA-DQB1 alleles in Venezuelan individuals with symptoms suggestive of CD.

The main findings included a positive association between the risk alleles (HLA-DQB1 *0201, *0302) and the presence of lesions in the intestinal mucosa, regardless of lesion type (infiltrative, hyperplastic, or destructive). Some studies have suggested that the HLA-DQB1*02 allele that encodes the DQ2 molecule determines the severity of damage to the intestinal mucosa, as well as its progression to possible complications. Specifically, patients with DQB1*02 homozygosity, not only present with more severe clinical symptoms, in particular, diarrhea and anemia, but also develop more severe intestinal lesions that take longer to heal, in the absence of a GFD.10 In our study, we found that 74% of the patients with villous atrophy had at least one DQB1*02 allele and 26% a DQB1*0302 allele, suggesting that the presence of the DQB1*02 and DQB1*0302 alleles, is associated not only with a predisposition to develop CD, but also with the presence of lesions in the intestinal mucosa. Nevertheless, the patients diagnosed with CD that followed a GFD showed clinical improvement and none of them presented with intestinal lymphoma or refractory CD.

Utilizing the criteria of the “four-out-of-five rule”, which includes clinical manifestations, serology, risk genotype, intestinal biopsy results, and clinical response to a GFD, we confirmed the diagnosis of CD in 40 of the individuals with symptoms suggestive of CD. By comparing the frequencies of the HLA-DQB1 alleles in the patients with CD diagnosis and the frequencies described in celiac patients from other populations, we found an increased frequency of the DQB1*0201 and DQB1*0302 risk alleles, similar to that described in a European population11 and in Latin American mestizo populations, such as those in Argentina,12,13 Brazil,14–17 Chile,18,19 Uruguay,20 and Cuba,21 populations characterized by having a White component above 50%. On the other hand, we found a lower frequency of the HLA-DQB1*0302 allele, with respect to the frequency described in populations with a higher prevalence of Amerindian genes.22,23 Recently, Mejía-León and Calderón-de la Barca24 highlighted the fact that, in the Mexican population, the DQ2 and DQ8 molecules contributed to the genetic risk for CD, showing that the risk was not exclusively due to DQ2, as in Europeans, or to DQ8, as seen in populations with a greater prevalence of Amerindian genes.

In individuals with no risk genotypes and with symptoms suggestive of CD, the presence of intestinal lesions characteristic of CD could be explained by the existence of other causes of intestinal villous atrophy, including autoimmune enteropathy, variable common immunodeficiency, enteropathy associated with olmesartan, lymphoproliferative disorders of the small bowel, and atrophy of infectious origin,25 as well as the presence of other non-HLA variants associated with CD. Association studies on the complete genome have identified more than 40 non-HLA genes associated with CD,26,27 that are relevant, even if they confer a limited genetic risk, because they enable the discovery of potential key pathways involved in the pathogenesis of the disease.

One percent of the individuals with symptoms suggestive of CD, that carried a risk genotype, but did not undergo intestinal biopsy, presented with anti-gliadin IgA antibodies and anti-tissue transglutaminase IgA antibodies, as well as with clinical improvement upon following a GFD. That group was made up only of children, with a mean age of 3.6 ± 2.8 years, and could be classified as patients with a presumptive diagnosis of CD, despite not having undergone intestinal biopsy. Even though there is an ongoing debate for defining whether or not noninvasive tests can be used as the only tests for diagnosing CD, positivity for two different serologic tests and the presence of risk genotypes (HLA-DQ2 and/or HLA-DQ8) in children has been suggested as a possible test in the diagnostic algorithm for CD, thus avoiding the performance of gastrointestinal endoscopy with biopsy.3

Finally, individuals with risk genotypes and the presence of anti-tissue transglutaminase antibodies (IgA-IgG) are 4.7-times more prone to develop CD, confirmed upon observing an increased frequency of individuals diagnosed with CD, in the group of patients with risk genotype and seropositivity for anti-tissue transglutaminase antibodies. Even though CD-specific antibodies are utilized for the presumptive diagnosis of CD, it has been suggested that those antibodies carry out important functions in relation to disease progression, at the level of the small bowel mucosa, and in the development of extraintestinal manifestations related to CD.

There is scant research on CD in the Venezuelan population, and in that sense, one of the main contributions of the present study is that the results support the role of the DQB1*02 and DQB1*0302 alleles in disease susceptibility and in the histologic alterations of CD in that population. Therefore, the genotyping of those specific genes in the Venezuelan population would be useful for screening subjects at risk for CD, distinguishing individuals with a low probability for developing the disease from individuals that should be monitored for possible future development of the disorder.

No financial support was received in relation to this study/article.

The authors declare that there is no conflict of interest.