Skip to main content
Premium Trial:

Request an Annual Quote

Gene Expression Study Finds Patterns of Variability in Immune Response

NEW YORK (GenomeWeb) – Through a gene expression-based analysis, a Broad Institute-led research team uncovered patterns of variability in T lymphocyte activation and immune response.

T lymphocytes are activated as part of the adaptive immune response when they recognize an antigen presented by the major histocompatibility complex. This sets off a chain of events that leads to the amplification of rare T cells specific for the antigen as well as the differentiation of T cells into cytotoxic or helper T cells, including CD4+ T cells, that play various roles in immune response.

By analyzing gene expression of CD4+ T cells collected from some 350 people, the researchers led by the Broad Institute's Christophe Benoist found inter-individual variability, including in cytokine transcripts, as they reported yesterday in Science. Ancestry, they further noted, appeared to influence that variability.

"The study uncovered an unexpected, but substantial population differentiation in overall immune response and in the expression of genes encoding key cytokines, with a trend toward higher activation in cells from donors of African ancestry," Benoist and his colleagues reported in their paper.

Such inter-individual variability, the researchers added, could influence immunopathologies and autoimmune disorders.

As part of the Boston PhenoGenetic Study, the researchers collected primary T cells from 348 healthy people, including 91 individuals of African, 74 of Asian, and 183 of European ancestry.

Following a step-wise approach, Benoist and his colleagues isolated and activated the CD4+ T cells they collected into four conditions: activated CD4+ T cells at the four-hour time point, activated CD4+ T cells at the 48-hour time point, IFN-β stimulated cells, and TH17 cells.

The researchers conducted genome-wide expression profiling on a subset of samples to determine the time points during activation and the transcripts that provided the best picture of the activation process and variation among the research participants.

Based on those findings, Benoist and his colleagues used NanoString Technologies' direct molecular counting approach to measure the selected 236-transcript gene set in the four sets of CD4+ T cells.

Activation of the cells, the researchers reported, induced the expression of some 1,750 genes, while repressing 456 genes. Some 290 transcripts were affected in the TH17 cells, but not in the activated CD4+ T cells. The IFN-β-stimulated cells similarly had 270 transcripts affected.

Cytokine transcripts within those various T cells, the researchers reported, hewed to a generally consistent pattern of response across the individuals, though still had variable response.

By focusing on the expression levels of 51 transcripts that are induced during activation, the researchers calculated a response index to gauge the general intensity of T cell response.

Individuals, they found, had complex response patterns. For instance, they found that the levels of IL17F and IL4 were largely independent, but that the levels of IL12F and IFNG were positively correlated at 48 hours.

"This complexity in individual responses is exemplified by a cluster of donors characterized by higher IL4 and IL9 expression, but which included some high IL17F or IFNG responders as well," they said.

For each condition, the researchers also examined whether certain pairs of cytokines were co-expressed. The clusters, they noted, became more defined after 48 hours, and while some of the clusters were expected — like IL17A/F and IL22 — others — like IFNG and IL21 — were not. Additionally, some early transcripts were linked to the expression of later transcripts.

Variation in immune response was only lowly linked to physiological traits like gender, height, weight, and blood pressure in the European subset of their cohort, but was more significantly associated with sex and age. Physiological covariates, they noted, accounted for some 4 percent of variability while cis-genetic effects accounted for a quarter of variability. The remainder, they noted, is likely due to environmental effects, immunological history, or trans-genetic effects.

Across the full cohort of 248 donors, though, ancestry differentiated expression in some 41 percent of the genes in the 48-hour conditions and explained some 7 percent of the variation in expression. Further, they noted a trend of overexpression within donors of African ancestry, lower expression in donors of European ancestry, and mixed response in donors of Asian ancestry.

To pin down the genetic variants underlying these expression differences, the researchers associated some 10 million imputed genotypes from their cohort of 348 donors. From this, they identified 38 cis-eQTLs that were state-specific.

One low-expression variant — rs12251836 — was rare in African and Asian genomes, but common in European ones.
By superimposing these eQTls onto a catalog of GWAS results, the researchers found that disease-associated variants have been identified near some of these loci, including ones near IL23R and IL2RA, cytokine receptors that have both been linked to autoimmune disorders.

The rs12251836 variant, which was the most significant one linked to IL2RA expression in activated cells at 48 hours, was not linked to rs12722495, a variant that had been found to mark a protective type 1 diabetes allele in IL2RA. But by conditioning their data on rs1225183, the researchers were able to see a secondary association between rs12722495 and IL2RA expression in activated cells.

Additionally, Benoist and his colleagues examined the functional effects of five candidate SNPs near IL2RA. The section housing rs12251836, they found, appears to boost enhancer activity when there's a single base-change there. Indeed, they reported that YY1 would bind only one rs12251836 allele, while other enhancers will bind to that region, no matter the allele.

Because the diversity in transcript variation follows a complex pattern, the researchers noted that pegging disease or therapeutic outcome to the levels of one cytokine might not be sufficient. Instead, they argued that individuals might have to be mapped to a multi-dimensional outcome.

"It is possible that some differences in responsiveness reflect disparate nutritional or environmental influences in addition to the genetic component, but many involve cis genetic differences, for instance, the rs12251836 variant that controls IL2RA through differential binding of YY1," Benoist and his colleagues noted.

"The low responder allele is a derived allele, present almost exclusively in donors of European descent, but rare in African and Asian descendants," they added. "The lower responses to activation and the overrepresentation of the corresponding alleles in European-derived populations suggest that tolerance mechanisms to avoid immunopathology resulting from exuberant T cell responses may also have led to the evolution of protection from autoimmunity."