A guide to heat shock factors as multifunctional transcriptional regulators

Abstract

The heat shock factors (HSFs) form a family of transcription factors, which are evolutionarily conserved in eukaryotes. They are best known as transcriptional regulators of molecular chaperone genes, including those encoding heat shock proteins, in response to heat shock and other protein-damaging stresses. Since the discovery of the first HSF and its eponymous role in the heat shock response four decades ago, the currently known HSFs in vertebrates, that is, HSF1-5, HSFX, and HSFY, have been implicated in a wide array of physiological and pathological processes, including organismal development and cancer progression. To date, most studies have focused on individual HSFs, but it is becoming increasingly evident that the role of multiple HSFs and their potential crosstalk should be considered. In this review, we provide a comprehensive overview of the structures, functions, and regulation of the mammalian HSF family members and explore their interplay in biological processes. We highlight recent advancements regarding the roles of HSF family members in viral infection, cell adhesion, and spermatogenesis, and discuss the key questions to be addressed by forthcoming studies in HSF biology.

Keywords: HSE; HSF; HSP; HSR; adhesion; cancer; development; spermatogenesis; stress; transcription.

Fig. 1

Phylogram based on HSF protein sequences from vertebrates, invertebrates, and fungi, including human (Homo sapiens), cattle (Bos taurus), mouse (Mus musculus), rat (Rattus norvegicus), chicken (Gallus gallus), western clawed frog (Xenopus tropicalis), zebrafish (Danio rerio), fruit fly (Drosophila melanogaster), fungus (Cryptococcus neoformans), and baker's yeast (Saccharomyces cerevisiae). Constructed using multiple sequence alignment with ClustalO, phylogenetic tree generation with Simple Phylogeny through EMBL‐EBI Job Dispatcher, and formatted with iTOL [194, 195]. Amino acid sequences were retrieved from the NCBI RefSeq (for C. neoformans) or UniProt databases. Tree scale indicates distance; the number of substitutions in proportion to the alignment length, excluding gaps.

Fig. 2

The domain structure of the HSF family is conserved. (A) Comparison of the domain structure of HSFs in human (Homo sapiens), mouse (Mus musculus), chicken (Gallus gallus), zebrafish (Danio rerio), fruit fly (Drosophila melanogaster) and baker's yeast (Saccharomyces cerevisiae). The number of amino acids in each HSF is indicated. AD, transactivation domain; DBD, DNA‐binding domain; HR‐A/B, oligomerization domain; HR‐C, C‐terminal heptad repeat domain; RD, regulatory domain. (B) Sequence identity comparison of human HSFs, determined by multiple‐sequence alignment (MSA) of full‐length HSFs, DBDs, or HR‐domains. EMBL‐EBI Job Dispatcher was used for alignment and percent identity matrices [195]. UniProt accession numbers for the aligned amino acid sequences: Q00613 (HSF1), Q03933 (HSF2), Q9ULV5 (HSF4), Q4G112 (HSF5), Q9UBD0 (HSFX), Q96LI6 (HSFY).

Fig. 3

Overview of domains, posttranslational modifications (PTMs) and predicted disorder in human HSFs. The predicted disorder values were determined with AIUpred, with values above 0.5 indicating a likely disordered region [196]. The databases PhosphoSitePlus, iPTMnet, and GlyGen, which include modifications from high‐throughput screens, were used as sources for PTMs [66, 67, 68]. The PTMs of human HSFs are indicated, but additional PTMs have been detected on conserved amino acids of murine HSFs. Acetylation sites on human HSF2 with experimental support have been added [92]. The number of amino acids in each HSF is indicated. AD, transactivation domain; DBD, DNA‐binding domain; HR‐A/B, oligomerization domain; HR‐C, C‐terminal heptad repeat domain; RD, regulatory domain; α, region specific to the α‐isoform; β, region specific to the β‐isoform.

Fig. 4

HSFs are involved in a diverse range of physiological and pathological processes. Filled boxes indicate HSFs currently known to be associated with each process. ECM, extracellular matrix. References for HSF functions not mentioned in the main text: Cell cycle and HSF1 [197], HSF2 [190], HSF4 [198], HSF5 [26]; Metabolism and HSF2 [124, 199], HSF4 [138]; Circadian rhythm and HSF1 [200]; Chromatin remodeling and HSF1 [201], HSF2 [202], HSF5 [24]; DNA repair and HSF1 [203]; ECM remodeling and HSF2 [105]; Olfactory neurogenesis and HSF1 [204]; Placental development and HSF1 [136]; Spermatogenesis and HSF4 [183]; Aging and HSF1 [3], HSF4 [205]; Fetal alcohol spectrum disorder and HSF2 [188]; Ischemic reperfusion and HSF1 [206, 207].

Fig. 5

HSFs in spermatogenesis. (A) Expression patterns of mouse HSF1, HSF2, HSF4, HSF5, and HSFY in spermatogonia, spermatocytes, round spermatids, elongated spermatids, and spermatozoa [27, 111, 177, 178]. The expression pattern of HSFX corresponds to human, from Human Protein Atlas proteinatlas.org [81], since mouse data is not currently available. (B) Summary of phenotypes corresponding to knockout HSF1, HSF2, HSF1 & HSF2, and HSF5 knockout mice [27, 85, 136, 180]. Created in BioRender [208].