The Inflammatory Transcription Factors NFκB, STAT1 and STAT3 Drive Age-Associated Transcriptional Changes in the Human Kidney

Abstract

Human kidney function declines with age, accompanied by stereotyped changes in gene expression and histopathology, but the mechanisms underlying these changes are largely unknown. To identify potential regulators of kidney aging, we compared age-associated transcriptional changes in the human kidney with genome-wide maps of transcription factor occupancy from ChIP-seq datasets in human cells. The strongest candidates were the inflammation-associated transcription factors NFκB, STAT1 and STAT3, the activities of which increase with age in epithelial compartments of the renal cortex. Stimulation of renal tubular epithelial cells with the inflammatory cytokines IL-6 (a STAT3 activator), IFNγ (a STAT1 activator), or TNFα (an NFκB activator) recapitulated age-associated gene expression changes. We show that common DNA variants in RELA and NFKB1, the two genes encoding subunits of the NFκB transcription factor, associate with kidney function and chronic kidney disease in gene association studies, providing the first evidence that genetic variation in NFκB contributes to renal aging phenotypes. Our results suggest that NFκB, STAT1 and STAT3 underlie transcriptional changes and chronic inflammation in the aging human kidney.

Fig 1. Genomic search for candidate regulators of the kidney aging transcriptome.

A. Schematic of the genomic pipeline used to identify associations between transcription factor ChIP-seq binding targets (ENCODE) and kidney age-related genes. The list of age-related genes was overlapped with the ChIP-seq binding targets of 161 transcription factors from 961 ChIP-seq experiments (see Methods). The overlap between the target genes assigned to each transcription factor with age-related genes was determined for each ChIP-seq dataset. B. The histogram shows the fold-enrichments of seven transcription factors that were >1.5 fold enriched and statistically significant after Bonferroni correction (p < 5 x 10⁻⁵) for binding the kidney age-related genes (see Methods). The three inflammation-associated transcription factors STAT1, NFκB (RelA) and STAT3 showed the strongest ChIP-seq enrichments for binding kidney age-related genes.

Fig 2. Increased activity of STAT1, STAT3 and NFκB transcription factors during kidney aging.

A. Boxplots showing significantly increased mRNA expression of STAT1 and STAT3 in old (> 65 years, n = 42) versus young (< 45 years, n = 10) renal cortex samples. There was not a significant difference in RELA mRNA expression levels between old and young individuals. The normalized gene expression data were from a previously published microarray study [1]. B. Left panels: Images of IHC showing immunoreactivity for STAT1, pSTAT1, STAT3, pSTAT3, RelA and pRelA in sections of renal cortex from young (<40 years) and old (>65 years) individuals. Positive signal is represented by brown color. Right panels: Boxplots showing IHC immunoreactivity scores for STAT1 (n = 10 young, n = 9 old), nuclear pSTAT1 (n = 7 young, n = 10 old), STAT3 (n = 6 young, old n = 8), nuclear pSTAT3 (n = 10 young, n = 10 old), RelA (n = 6 young, n = 6 old) and nuclear pRelA (n = 10 young, n = 10 old). The boxes indicate boundaries for the 25th and 75th percentile, and lines indicate maximum, median and minimum values. ** Indicates p < 0.01, * p < 0.05 (one-sided Mann-Whitney U test).

Fig 3. Activation of STAT1, STAT3 or NFκB by inflammatory cytokines recapitulates kidney aging-related gene expression patterns in human renal epithelial cells.

A. The left column of the heat map shows the log₂ fold-changes of 40 direct targets of STAT1 following IFNγ stimulation of HK-2 cells from microarray expression profiling experiments. The right column of the heat map shows the corresponding log₂-adjusted beta coefficient (age-slope) for these STAT1 direct targets during kidney aging. B. The left column of heat map shows the log₂ fold-changes of 43 direct targets of STAT3 following IL-6 stimulation of HK-2 cells from microarray expression profiling experiments. The middle column shows changes in expression following treatment with the STAT3 inhibitor S3I-201. The right column shows the corresponding log₂-adjusted beta coefficient (age-slope) during kidney aging. Aging gene expression data are from [1]. C. The left column of the heat map shows the log₂ fold-changes of 43 direct targets of NFκB following TNFα stimulation of HK-2 cells from microarray expression profiling experiments. The right column of the heat map shows the corresponding log₂-adjusted beta coefficient (age-slope) for these NFκB direct targets during kidney aging. Yellow indicates increased gene expression (positive fold-change or age-slope) and blue indicates decreased gene expression (negative-fold change or age-slope).

Fig 4. TNFα induces a mesenchymal transition and fibrogenic gene expression in HK-2 cells.

A. Immunofluorescence images of HK-2 cells treated with vehicle control (left) or TNFα (right); Cells were co-stained with anti-vimentin antibodies (green, top panels) and pRelA antibodies (red, bottom panels). Increased nuclear pRelA staining in TNFα-treated cells indicates NFκB activation. Induction of vimentin in TNFα-stimulated cells marks the cell fate transition to a mesenchymal-like phenotype. B. Micrograph of HK-2 cells treated with vehicle control (left) or TNFα (right). The TNFα-stimulated cells display an elongated fibroblastic morphology, characteristic of mesenchymal cells. C. The boxplot shows significantly increased levels of vimentin mRNA expression during kidney aging, using expression data from Rodwell et al. 2004. D. Collagen genes are induced by TNFα stimulation and tend to increase expression during aging. The left column of the heat map shows the log₂-fold-changes in expression of 14 collagen transcripts following stimulation of HK-2 cells with TNFα. The right column of the heat map shows the age-related slope for these genes during kidney aging.

Fig 5. Correlations of the activities of STAT1, STAT3, NFκB and macrophage abundance in individual kidneys.

A. In this heat map the rows indicate estimated macrophage abundance based on macrophage-specific transcript expression, or level of activation of each transcription factor, based on the averaged expression of transcription factor direct target genes (see Methods). The colors indicate relative levels of macrophage abundance or transcription factor target gene expression compared to other individuals of that age; red columns represent individuals with higher expression of macrophage-specific transcripts or transcription factor target genes for their age (older than expected at the gene expression level), and blue columns represent individuals with lower expression levels of macrophage markers of transcription factor target genes for their age (more youthful at the gene expression level). The columns are clustered such that individuals with values that are high or low for their age appear together. B. Macrophage infiltration in the kidney increases with age: images of renal cortex samples showing CD163 staining (green) with DAPI counterstaining (blue). Shown is a representative example of a young renal cortex (left) and an old renal cortex (right). The boxplot quantifies macrophage infiltration. The relative abundance of macrophages was defined as the fraction of CD163⁺ cells/all DAPI⁺ cells The boxes indicate 25th and 75th percentiles for the group, and the lines indicate maximum, median and minimum values. **P < 0.01 (Student’s t test, one-sided). C. Macrophage abundance correlates with levels of activation of STAT3 and NFκB in individual kidneys. Human renal cortex stained with anti-CD163 antibodies (macrophage marker), co-stained with either pRelA (top panels) or pSTAT3 antibodies (bottom panels). Right scatterplots show the correlation between overall macrophage abundance and pSTAT3 nuclear immunoreactivity (top) or pRelA nuclear immunoreactivity (bottom) in epithelial compartments of individual renal cortex tissues. Red diamonds represent old individuals (ages 66–85 years) and blue diamonds represent young individuals (ages 25–44 years). The x-axis indicates the fraction of cells that are CD163⁺ and the y-axis indicates for the fraction of cells stained positive for pRelA or pSTAT3. D. Macrophage-conditioned media responsive genes are induced during kidney aging. The left column of the heat map shows 77 differentially regulated genes in response to macrophage-conditioned medium (S7 Table) from two microarray studies in different human cell lines [38,39]. The right column of the heat map shows the changes in gene expression for these genes during kidney aging (age-related slope).

Fig 6. Genetic variation in NFκB genes associates with kidney function and chronic kidney disease.

A. The regional association plot shows each SNP tested in the RELA gene arranged by its position on chromosome 11 (x-axis) with the–log₁₀(p-value) for association with eGFR on the y-axis. The purple diamond represents the lead SNP within the RELA gene rs11820062. The colors of flanking SNPs in RELA represent their linkage disequilibrium (R²) with the lead SNP. The green dot represents rs4014195, an intergenic SNP that was previously associated with kidney function at genome-wide significance [40]. rs4014195 is moderately linked to rs11820062 in RELA (R² = 0.44). Dotted line indicates Bonferroni significance level. B. The regional association plot shows each SNP tested in the NFKB1 gene arranged by its position on chromosome 4 (x-axis) and the log₁₀(p-value) for their association with eGFR on the y-axis. The purple diamond represents the lead SNP rs12509403. The color of dots representing flanking SNPs in the NFKB1 gene indicates their linkage disequilibrium (R²) with the lead SNP as indicated in the heat map color key. Red dotted line indicates Bonferroni significance level. C. The schematic illustrates how rs12509403 genotype influences both NFKB1 gene expression and kidney function. Rs12509403 correlates with NFKB1 gene expression such that the T allele is associated with higher NFKB1 mRNA expression and higher eGFR, while the C allele is associated with reduced mRNA expression and lower eGFR.