Interleukin-1β and Tumor Necrosis Factor α play related, but distinct, roles in immunity and disease. Our study revealed major mechanistic distinctions in the Toll-like receptor (TLR) signaling-dependent induction for the rapidly expressed genes (IL1B and TNF) coding for these two cytokines. Prior to induction, TNF exhibited pre-bound TATA Binding Protein (TBP) and paused RNA Polymerase II (Pol II), hallmarks of poised immediate-early (IE) genes. In contrast, unstimulated IL1B displayed very low levels of both TBP and paused Pol II, requiring the lineage-specific Spi-1/PU.1 (Spi1) transcription factor as an anchor for induction-dependent interaction with two TLR-activated transcription factors, C/EBPβ and NF-κB. Activation and DNA binding of these two pre-expressed factors resulted in de novo recruitment of TBP and Pol II to IL1B in concert with a permissive state for elongation mediated by the recruitment of elongation factor P-TEFb. This Spi1-dependent mechanism for IL1B transcription, which is unique for a rapidly-induced/poised IE gene, was more dependent upon P-TEFb than was the case for the TNF gene. Furthermore, the dependence on phosphoinositide 3-kinase for P-TEFb recruitment to IL1B paralleled a greater sensitivity to the metabolic state of the cell and a lower sensitivity to the phenomenon of endotoxin tolerance than was evident for TNF. Such differences in induction mechanisms argue against the prevailing paradigm that all IE genes possess paused Pol II and may further delineate the specific roles played by each of these rapidly expressed immune modulators.

Rapid induction of transcription is known to be mediated by factors which bind DNA following post-translational modification. We report here that non-tyrosine phosphorylated (NTP)-Stat1 is involved in a cooperative interaction with Spi-1/PU.1 and IRF8 to form a pre-associated, poised complex for IL1B gene induction. A double point mutation at a putative STAT binding site, which overlaps this composite Spi-1 x IRF8 site located in the LPS and IL-1 response element (LILRE), inhibited human IL1B LPS-dependent reporter activity to about 10 percent of the control wild type vector. Chromatin immunoprecipitation revealed stimulation-independent constitutive binding of IRF8, Spi-1 and NTP-Stat1 at the LILRE, while binding of C/EBP beta was activated at an adjacent C/EBP beta site after LPS stimulation. In contrast to Stat1, IRF8 was tyrosine phosphorylated following LPS treatment. Supporting the involvement of NTP-Stat1, LPS-induced IL1B reporter activity in monocytes was enhanced by ectopic expression of NTP-Stat1 Y701F. In contrast, co-expression of a Y211F IRF8 mutein functioned as a dominant-negative inhibitor of LPS-induced IL1B reporter activity. In vitro DNA binding using extracts from LPS-treated monocytes confirmed that the LILRE enhancer constitutively binds a trimolecular complex containing IRF8, Spi-1 and NTP-Stat1. Binding studies using in vitro-expressed proteins revealed that NTP-Stat1 enhanced the binding of Spi-1 and IRF8 to the LILRE. Co-expression of TRAF6, an LPS surrogate, with Spi-1 and IRF8 enhanced IL1B reporter activity in HEK293R cells, which was dramatically reduced when Y211F IRF8 was co-expressed. These results suggest that the rapid transcriptional induction of an important inflammatory gene is dependent upon constitutive cooperative binding of a Spi-1 x IRF8 x NTP-Stat1 complex to the LILRE, which primes the gene for immediate induction following IRF8 phosphorylation. Phosphorylation of chromatin pre-associated factors like IRF8 may be an important strategy for the rapid transcriptional activation of genes involved in innate immunity.

Endogenous damage associated molecular pattern molecules (DAMPs) released from necrotic, damaged or stressed cells are associated with an inflammatory response. Whether the microRNA (miR) expression signature of this response is different from that of a pathogen associated molecular pattern (PAMP)-stimulated inflammatory response is unknown. We report here that miR-34c and miR-214 are significantly expressed in fresh human peripheral blood mononuclear cells (PBMCs) exposed to DAMP-containing freeze-thaw lysates, or to conditioned media from serum-starved and glucose-deprived cells (p<6×10−4 and p<3.7×10−3), respectively. Interestingly, only miR-34c expression was differentially expressed in PBMCs exposed to freeze-thaw lysates or conditioned media from wildtype High Mobility Group B1 (HMGB1+/+) mouse embryonic fibroblast (MEF) cells, when compared to cultures exposed to lysates or conditioned media from HMGB1−/− MEFs. miR-155 expression in these cultures was negligible, but was significantly expressed in PBMCs stimulated with Lipopolysaccahride (LPS) or most other Toll-like receptor (TLR) ligands, making it the prototypic “PAMPmiR”. Exposure to a damaged human colorectal carcinoma cell line lysate (HCT116) similarly resulted in increased miR-34c and miR-214 levels. When PBMCs were pre-transfected with anti-miR-34c and then exposed to lysate, expression levels of IKKγ mRNA, a putative target of miR-34c, increased, while protein levels of IKKγ in cultures transfected with a pre-miR-34c were abrogated. Levels of miR-34c expression (as well as pro-inflammatory cytokines, IL-1β and TNFα) decreased when PBMC cultures were briefly pre-incubated with the K+ channel (inflammasome) inhibitor, glybenclamide, suggesting that inflammasome activation is upstream of miR-34c expression in response to DAMPs. Our findings demonstrate that a specific microRNA expression signature is associated with the inflammatory response to damaged/injured cells and carries implications for many acute and chronic inflammatory disorders.

Tumor necrosis factor α converting enzyme (TACE) is a sheddase overexpressed in cancers that generates cancer cell growth and survival factors, and is implicated in carcinogenesis and tumor growth. This indicates that TACE could be a potentially important cancer biomarker. Unexpectedly, TACE expression in cancer tissues does not correlate with cancer stage or invasiveness. Although TACE sheddase activity is a more direct and potentially better indicator of TACE biology and might be a better cancer biomarker than TACE expression, it has not been studied in cancer tissues. In the present study, we developed a reliable specific assay for quantification of TACE sheddase activity, investigated TACE activity and TACE protein expression in head and neck cancer (HNC) tissues, and examined the correlation of the results with HNC clinical stages and likelihood to recur.We found that HNC cell lines and tissues contained remarkably higher quantities of TACE activity and TACE protein than normal keratinocytes or oral mucosa. siR-NA silencing of TACE resulted in the inhibition of release of the tumorogenic factors amphiregulin and transforming growth factor α, and tumor protective factors tumor necrosis factor receptors from HNC cells. Importantly, TACE activity, but not TACE protein expression, was significantly higher in large, T3/T4, primary tumors relative to small, T1/T2, primary tumors, and especially in primary tumors likely to recur relative to those unlikely to recur. These data show that increased TACE activity in cancer is biologically and clinically relevant, and indicate that TACE activity could be a significant biomarker of cancer prognosis. (Cancer Epidemiol Biomarkers Prev 2009;18(11):2913-22)