Solving the Conundrum: Immunogenetics of Sarcoidosis

Despite decades of research, sarcoidosis is still among the most enigmatic lung diseases, with poorly understood etiology, course, and outcome. Recent clinical and genetic observations have significantly highlighted the importance of immune mediators and immunogenetics in determining the phenotypic behavior of the disease, and potentially of its pathogenesis. In the context of the paper in this issue of the Journal by Fischer and colleagues (pp. 727–736) (1), the association of the major histocompatibility complex class II allele DRB1*03 with decreased risk for persistent disease in patients with sarcoidosis who are of European and African descent (2, 3); the associations of declines in CD4⁺, CD8⁺, and B cell counts with severe sarcoidosis phenotypes (4); and the peripheral blood expression changes in T-cell receptor signaling, Janus kinase–signal transducer and activator of transcription signaling and cytokine-cytokine receptor signaling pathways (5, 6) are of most relevance. These clinical and genomic observations suggest that although sarcoidosis pathogenesis may be related to aberrant patterns of immune activation, it is unlikely to be caused by a single major gene or molecular pathway but, more likely, to a complex interaction of multiple genes/pathways with contributing environmental factors that determine not only the presenceof the disease but also its unique phenotypic manifestations.

Aiming to address this complex interaction, Fischer and colleagues (1) use an approach applied successfully in other inflammatory disorders. Building on findings from previous genome-wide association studies that most consistently identified loci in BTNL2, ANXA11, and HLA genes as disease susceptibility genes, but also suggested overlap with other autoimmune and inflammatory disorders, they applied the Illumina Immunochip Single Nucleotide Polymorphism (SNP) array, designed for genotyping and fine mapping of gene variants known to associate with immune diseases, to a European cohort of 1,726 patients with sarcoidosis and 5,482 healthy controls. They replicated their results in multiple additional cohorts that included several European cohorts and one African American cohort, using other methods. Overall, they analyzed more than 19,000 individuals, making this the largest sarcoidosis case control study ever performed. This unique design allowed the authors to identify new disease susceptibility loci while fine-mapping previously identified loci. For example, in addition to confirming the association of the BTLN2 region with a predisposition to sarcoidosis, they now delineate that the most strongly associated variants in the promoter region are predicted to reside in transcription factor binding sites, a finding with significant functional implications. Similarly, in addition to finding associations with the known HLA variants, they identify an independent variant located in the 3′ untranslated region of HLA-DPB1, suggesting an effect on microRNA binding.

The novel disease susceptibility loci the authors identified are near inflammatory regulators or cytokines such as 12q24.12 (ATXN2/SH2B3), 5q33.3 (near IL12B), 4q24 (MANBA/NFKB1), 2q33.2 (FAM117B), and 1p31.3 (IL23R). In fact, some overlap with known loci for psoriasis and inflammatory bowel disease (7–9), findings consistent with the finding of multiple aberrations in adaptive immunity in patients with sarcoidosis (10–12). Others have previously demonstrated the significance of reduced levels of nuclear factor-κB (NF-κB)/P65 to sarcoidosis disease severity (12), as well as dysregulation within IL-12 and IL-23 pathways within cutaneous sarcoidosis specimens, using microarray analysis (13). Reversal of global CD4⁺ subset dysfunction is present during spontaneous clinical resolution of pulmonary sarcoidosis. Anergic T-cell responses correlate with diminished expression of NF-κB and sarcoidosis disease progression (14). It is currently unknown whether any of these patients carry the risk variants identified in the current study.

One of the obvious limitations of this study is inherent with the use of Immunochip assays. For example, the authors identified no significant SNP associated with patients with sarcoidosis of African descent, despite inclusion of a large cohort. Immunochip assays were designed on the basis of SNPs identified in individuals of European origin; variation present in non–European-origin populations is underrepresented (9). Moreover, although the chip has dense coverage of the major histocompatibility complex and killer immunoglobulin-like receptor loci, it still misses new loci. Also, those identified after the February 2010 release of the early 1,000 Genomes Pilot data are not densely covered.

However, to the authors of this editorial, it is possible that results would not have been much more impressive, even if the authors performed whole-genome sequencing on their cohorts. The reason for that lies in the intriguing results of the cumulative heritability the authors performed. Their results point to the most significant challenge when trying to assess the risk for sarcoidosis: the interaction between genetic predisposition and environmental exposures. Using sophisticated analyses, the authors define 11 independent sarcoidosis risk variants; however, when they assess their cumulative heritability, they come up with a dismal number of less than 3%, which is a very low number for a very large study. Considering that none of these risk variants was replicated in African Americans, the heritability may be even lower for the general global population of subjects with sarcoidosis, and especially those expressing some of the more severe and life-threatening forms of the disease. Thus, application of rare allele polymorphisms in well-defined phenotypes and clarification of environmental factors, such as in fire fighters, health care workers, or agricultural exposure, or on changes in the microbiome may provide important context regarding gene–environment interactions. The authors’ identification of an independent variant located in the 3′ untranslated region of HLA-DPB1 suggests epigenetic modifications may contribute to sarcoidosis clinical outcomes and may point toward mechanisms influencing relevant gene–environment interactions.

Does this mean genetic variants do not play a major role in sarcoidosis? Our answer is that we believe they do; however, the data to date have not revealed significant global regulators of disease predisposition. Considering the clinical variability of the sarcoidosis, this should not come as a surprise. We should not be looking for genetic variants predisposing to a disease that is so variable in presentation, course, and outcome. We should look for variants explaining distinct phenotypic aspects of the disease. If this is true, one would argue that before more investigations of sarcoidosis genetics are performed, better phenotyping of patients should be conducted, using not only clinical characterizations but also advanced immunophenotyping technologies that could provide functional context to the clinical findings.

Thus, we believe, there is critical importance of a multidisciplinary teams in studying sarcoidosis: clinicians, geneticists, immunologists, microbiome experts, as well as environmental and occupational medicine, should work together to redefine the disease and identify the genetic variants that underlie the molecular underpinning of the disease. The impressive results reported by Fischer and colleagues in this issue of the Journal should encourage funding agencies and the research community to undertake such an effort, because based on these results, we can state that solving the sarcoidosis conundrum is possible.