Computational identification of gene-social environment interaction at the human IL6 locus

Abstract

To identify genetic factors that interact with social environments to impact human health, we used a bioinformatic strategy that couples expression array-based detection of environmentally responsive transcription factors with in silico discovery of regulatory polymorphisms to predict genetic loci that modulate transcriptional responses to stressful environments. Tests of one predicted interaction locus in the human IL6 promoter (SNP rs1800795) verified that it modulates transcriptional response to beta-adrenergic activation of the GATA1 transcription factor in vitro. In vivo validation studies confirmed links between adverse social conditions and increased transcription of GATA1 target genes in primary neural, immune, and cancer cells. Epidemiologic analyses verified the health significance of those molecular interactions by documenting increased 10-year mortality risk associated with late-life depressive symptoms that occurred solely for homozygous carriers of the GATA1-sensitive G allele of rs1800795. Gating of depression-related mortality risk by IL6 genotype pertained only to inflammation-related causes of death and was associated with increased chronic inflammation as indexed by plasma C-reactive protein. Computational modeling of molecular interactions, in vitro biochemical analyses, in vivo animal modeling, and human molecular epidemiologic analyses thus converge in identifying beta-adrenergic activation of GATA1 as a molecular pathway by which social adversity can alter human health risk selectively depending on individual genetic status at the IL6 locus.

Fig. 1.

(A) Computational modeling of the ancestral IL6 promoter sequence (−174G) identified a high-affinity GATA1-binding motif (TRANSFAC V$GATA1_01 mat_sim > 0.90) that was predicted to be abrogated by the rs1800795 G/C transversion (mat_sim < 0.75). (B) RT-PCR detection of GATA factor mRNA in IL6-producing cell types including macrophages, B lymphocytes, adipocytes, ovarian carcinoma cells, and B lymphoid cells latently infected with human herpesvirus 8 (Kaposi's sarcoma–associated herpesvirus). Data represent the mean fold-increase above negative controls in triplicate determinations. (C) Luciferase reporter assays assessed the capacity of the sympathetic neurotransmitter norepinephrine (NE) to activate GATA1, GATA2, or GATA3 transcriptional activity. Data represent the mean ± SE of three independent experiments using reporter constructs specifically responsive to individual GATA factors, with GATA1 showing greatest NE-induced activation in each cell type studied (P < 0.01). (D) NE induction of IL6 gene transcription was confirmed by RT-PCR in primary macrophages homozygous for the G allele of rs1800795. Data are mean ± SE of three replicate determinations in one experiment, with results representative of three independent experiments. (E) NE-induced binding of nuclear transcription factors to the IL6 promoter sequence (−187/−163) bearing either the −174G allele (NE induction, P < 0.001) or the −174C allele (NE induction, P = 0.46). Specificity of effects to the NE-activated PKA signaling pathway was tested by parallel PMA stimulation of PKC. Analyses also verified decreased binding of transcription factors to the −174C sequence under basal conditions (difference from −174G, P = 0.023). Data are representative of results from three independent experiments. (F) Allele-specific chromatin immunoprecipitation assays were preformed in rs1800795 heterozygous cells (primary macrophages shown) to confirm NE induction of GATA1 binding to the −174G allele of the IL6 promoter, but not to the −174C allele. GATA1-bound IL6 promoter DNA was quantified as a fraction of total IL6 promoter DNA. Data represent mean ± SE of three independent experiments.

Fig. 2.

(A) Luciferase reporter assays gauged NE-induced activation of human IL6 −174G vs. C allele promoters (data represent mean ± SE from three replicate determinations in ovarian carcinoma cells; difference in NE-mediated induction, P < 0.001). (B) The role of beta-adrenergic / PKA signaling in mediating NE induction of the IL6 −174G promoter was tested by pre-exposing cells to the beta-adrenergic antagonist propranolol or the PKA antagonist KT5720 before NE exposure. Sufficiency of PKA activation alone to induce the IL6 −174G promoter was tested using the pharmacologic PKA agonist db-cAMP. All data represent the mean ± SE from three replicate determinations. (C) The role of GATA1 in PKA-induced IL6 −174G promoter activation was tested by GATA1-targeted siRNA inhibition. Specificity of GATA1’s effect was tested by parallel siRNA inhibition of other GATA family members (e.g., GATA4 shown). Data represent mean ± SE of three replicate determinations. (D) In vivo activation of GATA-mediated gene transcription was assessed by TELiS promoter-based bioinformatic analysis of genome-wide transcriptional profiles in CD11b+ myeloid spleen cells harvested from adult male C57BL/6 mice after six daily cycles of 2-h exposure to social threat (n = 10) vs. control home caged animals (n = 10). Data represent the average (± SE) prevalence of GATA1 transcription factor-binding motifs in promoters of 100 genes showing the greatest magnitude of up-regulation following social threat relative to 100 genes showing the greatest magnitude of up-regulation in controls. (E) Parallel analyses of GATA1 transcription factor activity in brain prefrontal cortex from the same n = 10 control and n = 10 socially threatened animals as in B. (F) Differential prevalence of GATA1 transcription factor-binding motifs in the promoters of 220 human genes up-regulated ≥ 50% in ovarian carcinoma cells from 10 women experiencing adverse social conditions (high depressive symptoms and low social support) vs. 46 genes up-regulated in grade- and stage-matched ovarian carcinomas from 10 women experiencing favorable social conditions (low depressive symptoms and high social support).

Fig. 3.

Modification of late life socio-environmental mortality risk by rs1800795 genotype. (A) Relationship between depressive symptoms and subsequent all-cause mortality risk for IL6 −174G homozygotes (Right) vs. carriers of one or more C allele (Left). Data represent estimated survival functions from Cox proportional hazards regression analyses controlling for sex and age at study entry (range 70–80 years) in 184 initially healthy Caucasian participants in the MacArthur Study of Successful Aging. Open circles represent mortality risk given cohort average levels of depressive symptoms at study entry (CES-D = 4); filled circles represent mortality risk given high depressive symptoms at study entry (CES-D = 16). (B) Effect of controlling for chronic inflammation (plasma C-reactive protein ≥3 mg/L) on rs1800795 genotype modification of relative mortality hazard associated with low vs. high depressive symptoms at study entry (same cohort of n = 184 MacArthur Study participants as in A). Values represent point estimate (± 95% confidence interval) of relative hazard associated with 75th vs. 25th percentile values of cohort CES-D distribution.