HSF1-Activated Non-Coding Stress Response: Satellite lncRNAs and Beyond, an Emerging Story with a Complex Scenario

Abstract

In eukaryotes, the heat shock response is orchestrated by a transcription factor named Heat Shock Factor 1 (HSF1). HSF1 is mostly characterized for its role in activating the expression of a repertoire of protein-coding genes, including the heat shock protein (HSP) genes. Remarkably, a growing set of reports indicate that, upon heat shock, HSF1 also targets various non-coding regions of the genome. Focusing primarily on mammals, this review aims at reporting the identity of the non-coding genomic sites directly bound by HSF1, and at describing the molecular function of the long non-coding RNAs (lncRNAs) produced in response to HSF1 binding. The described non-coding genomic targets of HSF1 are pericentric Satellite DNA repeats, (sub)telomeric DNA repeats, Short Interspersed Nuclear Element (SINE) repeats, transcriptionally active enhancers and the NEAT1 gene. This diverse set of non-coding genomic sites, which already appears to be an integral part of the cellular response to stress, may only represent the first of many. Thus, the study of the evolutionary conserved heat stress response has the potential to emerge as a powerful cellular context to study lncRNAs, produced from repeated or unique DNA regions, with a regulatory function that is often well-documented but a mode of action that remains largely unknown.

Keywords: HSF1; NEAT1; SATIII; SINE; TERRA; eRNA; lncRNA.

Figure 1

Genomic targets of the HSF1-mediated heat-shock response. (A) Several types of genomic sequences are targeted and transcribed by HSF1. They include protein-coding genes, such as HSP genes, and non(protein)-coding sequences. This later group includes non-repetitive elements present at a single locus (e.g., NEAT1, distal Transcription Regulatory elements (dTREs) transcribed into eRNAs) and repetitive interspersed elements, present as a thousand copies within gene-rich regions (e.g., SINEs). Finally, non-coding sequences also include clusters of highly repetitive elements in tandem such as SATIII and Telomeric repeats (transcribed into TERRA), located at pericentric and telomeric regions. (B) Distribution of non-coding sequences along human chromosomes. SATIII lncRNA and TERRA are transcribed from large regions located at pericentric and telomeric regions, respectively (primarily from chromosome 9 in the case of SATIII lncRNA). SINE sequences are mainly present in the gene-rich and GC-rich regions (R-bands). eRNAs are transcribed from the dTREs present upstream of the promoter and gene transcription start sites (TSS) and are, therefore, likely to be widespread and possibly more abundant in gene-rich regions. The NEAT1 gene is present at a single locus.

Figure 2

Molecular roles assigned to the non(protein)-coding RNAs in the heat shock response. (A) The ncRNAs produced at pericentric regions are thought to play a role in cis through the recruitment, and possible sequestration, of transcriptional repressive complexes. Transcriptional and co-transcriptional regulators as well as RNA-binding factors such as splicing factors accumulate at SATIII transcribed genomic sites upon HS. The resulting transient accumulation of transcription and splicing factors at these sites is thought to possibly form membrane-less compartments and cause a rapid and reversible depletion/concentration of these factors from the rest of the nucleus. Likewise, the nuclear accumulation of NEAT1 lncRNA allows the formation of membrane-less nuclear structures known as paraspeckles, with a role in microRNA processing. (B) HSF1-dependent bi-directional transcription of dTREs into eRNAs may promote or repress transcription by RNAPII of protein-coding genes, upon heat shock. (C) SINE RNAs are thought to repress transcription by disrupting contacts between RNAPII and promoter DNA. In addition, specific SINE RNA-binding sites have been identified at many genic and intergenic targets proximal to RNAPII pausing, where they may prevent transcriptional elongation. (D) The lncRNAs produced at pericentric regions and telomeric regions are thought to play a role in cis through the recruitment, and possible sequestration, of transcriptional repressive complexes. (E) HSF1-dependent transcriptional activation of SINE sequences in an “antisense” orientation may also cause gene repression of the genes transcribed in a “sense” orientation.