Tissue-Specific Differential Expression of Novel Genes and Long Intergenic Noncoding RNAs in Humans With Extreme Response to Evoked Endotoxemia

Abstract

Background: Cytokine responses to activation of innate immunity differ between individuals, yet the genomic and tissue-specific transcriptomic determinants of inflammatory responsiveness are not well understood. We hypothesized that tissue-specific mRNA and long intergenic noncoding RNA (lincRNA) induction differs between individuals with divergent evoked inflammatory responses.

Methods: In the GENE Study (Genetics of Evoked Response to Niacin and Endotoxemia), we performed an inpatient endotoxin challenge (1 ng/kg lipopolysaccharide [LPS]) in healthy humans. We selected individuals in the top (high responders) and bottom (low responders) extremes of inflammatory responses and applied RNA sequencing to CD14 monocytes (N=15) and adipose tissue (N=25) before and after LPS administration.

Results: Although only a small number of genes were differentially expressed at baseline, there were clear differences in the magnitude of the transcriptional response post-LPS between high and low responders, with a far greater number of genes differentially expressed by endotoxemia in high responders. Furthermore, tissue responses differed during inflammation, and we found a number of tissue-specific differentially expressed lincRNAs post-LPS, which we validated. Relative to nondifferentially expressed lincRNAs, differentially expressed lincRNAs were equally likely to be nonconserved as conserved between human and mouse, indicating that conservation is not a predictor of lincRNAs associated with human inflammatory pathophysiology. Differentially expressed genes also were enriched for signals with inflammatory and cardiometabolic disease in published genome-wide association studies. CTB-41I6.2 ( AC002091.1), a nonconserved human-specific lincRNA, is one of the top lincRNAs regulated by endotoxemia in monocytes, but not in adipose tissue. Knockdown experiments in THP-1 monocytes suggest that this lincRNA enhances LPS-induced interleukin 6 ( IL6) expression in monocytes, and we now refer to this as monocyte LPS-induced lincRNA regulator of IL6 ( MOLRIL6).

Conclusions: We highlight mRNAs and lincRNAs that represent novel candidates for modulation of innate immune and metabolic responses in humans.

Clinical trial registration: URL: https://www.clinicaltrials.gov . Unique identifier: NCT00953667.

Keywords: adipose tissue; endotoxemia; inflammation; monocyte; response syndrome; systemic inflammatory.

Figures

Figure 1.. Number of differentially expressed mRNAs and lincRNAs after LPS in (A) CD14 monocytes (2 hours post-LPS) of high and low responders; (B) adipose tissue (4 hours post-LPS) of high- and low-responders; and (C) the overlap of differentially expressed mRNAs and lincRNAs between CD14 monocytes and adipose tissue in high-responders.

Numbers in figures represent mRNAs and lincRNAs, differentially expressed in first tissue/group (red text), differentially expressed in second tissue/group (blue text), and the overlap (purple text). Differentially expressed genes are defined as those with fold change ≥ 2 and a FDR-adjusted P

Figure 2.. Differentially-expressed lincRNAs identified and validated in CD14 monocytes of endotoxemia high-responders.

Data shown for gene expression estimates from original RNA-seq (“Discovery”, N=8) and qPCR (“Validation”, N=12) 2 hours after LPS. Validation (qPCR) of CD14 monocyte samples included six of the original RNA-seq high-responder subjects, and six new GENE Study high-responders. All ten candidate lincRNAs were DE in the same direction in validation samples, with eight reaching statistical significance. Differentially expressed genes are defined as those with fold change ≥ 2 and a FDR-adjusted P

Figure 3.. Differentially-expressed lincRNAs identified and validated in adipose tissue of endotoxemia high-responders.

Data shown for gene expression estimates from original RNA-seq (“Discovery”, N=13) and qPCR (“Validation”, N=15) 4 hours after LPS. Validation (qPCR) adipose tissue samples included eight of the original RNA-seq high-responder subjects, and seven new GENE Study high-responders. Four of the five candidate lincRNAs were confirmed as differentially expressed in validation samples. Differentially expressed genes are defined as those with fold change ≥ 2 and a FDR-adjusted P

Figure 4A.. A novel inflammatory-modulated lincRNA with tissue-specific activity in high-responders.

Data shown for gene expression estimates from original RNA-seq (“Discovery”; CD14 Monocyte N=8, 2 hours post-LPS; Adipose N=13, 4 hours post-LPS) and qPCR (“Validation”; CD14 Monocyte N=12, 2 hours post-LPS; Adipose N=15; 4 hours post-LPS) in high-responder subjects. RP11–362F19.1 was significantly down-regulated in CD14 monocytes and significantly up-regulated in adipose tissue during endotoxemia, both in the RNA-seq discovery analyses (see details in Table 2) and in the qPCR validation. * P

Figure 4B.. Divergent tissue-specific inflammatory-modulated mRNAs discovered by RNA-seq and validated by qPCR in high-responders.

Data shown for gene expression estimates from original RNA-seq (“Discovery”; CD14 Monocyte N=8, 2 hours post-LPS; Adipose N=13, 4 hours post-LPS) and qPCR (“Validation”; CD14 Monocyte N=12, 2 hours post-LPS; Adipose N=15, 4 hours post-LPS). In both discovery and independent validation samples, these mRNAs were regulated in the opposite directions in CD14 monocytes compared to adipose tissue during endotoxemia. * P

Figure 5.. MOLRIL6 is a novel human lincRNA that regulates LPS-induced IL6 expression

(A) We performed an inpatient endotoxin challenge (1 ng/kg LPS) in healthy humans to induce inflammation. We examined the transcriptome change of CD14+ monocytes (n=15) and adipose tissue (n=25) via RNA-seq. MOLRIL6 (CTB-41I6.2) is induced by LPS in monocytes but not in adipose. MOLRIL6 expression was more abundant in monocytes than in adipose or human adipose stromal cells (ASC)-differentiated adipocytes. (B) Genome browser view shows the genomic location of MOLRIL6 (CTB-41I6.2). Histone modification of H3K4me3 (promoter mark) and H3K4me1 (enhancer mark) around the transcription start site of MOLRIL6 supports an enhancer-driven transcript signature. RNA-seq track of one representative subject shows monocyte MOLRIL6 expression before and 2-hours after endotoxin administration but no basal or endotoxin-induced expression in adipose tissue. (C) MOLRIL6 is expressed in both nuclear and cytoplasmic fractions of THP-1 monocytes. (D) The knockdown of MOLRIL6 by ASO did not affect the expression of MOLRIL6 neighboring genes PIK3R5 and NTN1. (E-G) Knockdown MOLRIL6 in THP-1 monocytes reduced LPS-induced IL6 expression, and also reduced markedly CD86 mRNA at baseline prior to LPS stimulation. Data are show as mean±SEM, n = 3 independent experiments with triplicate. The ΔCTs represent the mean cycle threshold for the target genes relative to human ACTB mRNA as the reference in each sample. n.s. not statistically significant, * P<0.05; ** P<0.005; *** P<0.001