Regulation and function of the human HSP90AA1 gene

Abstract

Heat shock protein 90α (Hsp90α), encoded by the HSP90AA1 gene, is the stress inducible isoform of the molecular chaperone Hsp90. Hsp90α is regulated differently and has different functions when compared to the constitutively expressed Hsp90β isoform, despite high amino acid sequence identity between the two proteins. These differences are likely due to variations in nucleotide sequence within non-coding regions, which allows for specific regulation through interaction with particular transcription factors, and to subtle changes in amino acid sequence that allow for unique post-translational modifications. This article will specifically focus on the expression, function and regulation of Hsp90α.

Keywords: Cancer; Drug target; Gene ontology; Gene promoter; Heat shock element; Heat shock protein; Interactome; Molecular chaperone; Post-translational modification; Stress response.

Figure 1

Schematic of the human HSP90AA1 gene structure. Two transcript variants of Hsp90α protein are encoded on the complement strand of Chromosome 14. Information was obtained from NCBI Gene. Untranslated Region (UTR); ORF, Open Reading Frame (ORF).

Figure 2

Depiction of the HSP90AA1 transcription start sites. RNA polymerase II binds at three different sites on the HSP90AA1 gene; two sites associated with Transcript Variant 2 and one site associated with Transcript Variant 1. Information was obtained from UCSC Genome Browser and Encode.

Figure 3

Gene Ontology (GO) comparisons of the HSP90AA1 POLR2A binding site profiles: GO terms were computed/gathered by compiling all the transcription factors (TFs) that were found within the POLR2A binding site footprints. The lists of TFs were then searched on Amigo2 Term Enrichment Service powered by PANTHER with the biological process and H. Sapiens parameters selected (Carbon et al. 2009). The resulting list of GO terms was ranked by p-value. GO terms that scored p <0.05 were accepted as significant. The resulting lists were then compared using VENNY. The top 10 unique GO terms for each POLR2A sites are listed.

Figure 4

Interactomes and Gene Ontology comparisons of HSP90AA1 and HSP90AB1. Venn diagram of the number of overlapping and unique protein interactors and gene ontology terms associated with each Hsp90 isoform is shown. Information was obtained from Hsp90Int.DB, Amigo2 and Venny. These values are expected to change over time as the interactomes of each Hsp90 isoforms are further defined.

Figure 5

Hsp90 post-translational modification sites. A. Crystal structures of Hsp90 were obtained from RCSB protein databank; 1YET: human Hsp90α N-terminal domain bound to geldanamycin; 2CG9: yeast Hsp90 middle and C-terminal domains from full length structure bound to Sba1. Post-translational modification sites found on the PhosphoSite Plus website (http://www.phosphosite.org/homeAction.do;jsessionid=75252660BD94E928CEA59D49CC039052) are labeled in red using the PyMOL program. B. Display of Hsp90 post-translational sites. Unique phosphorylation sites for Hsp90α are in red.

Figure 6

Alterations of the HSP90AA1 gene across cancer. Data accumulated from the genomic characterization and analysis of tumor types according to The Cancer Genome Atlas. Chart was generated by cBioPrtal for Cancer Genomics.

Figure 7

Hsp90 N-terminal Inhibitors. Hsp90 N-terminal domain interaction with ATP is compared to interaction with Hsp90 inhibitors geldanamycin and radicicol using the PyMol program. 1YET.pdb (adapted from Roe et al. 1999).