Resorcylate Aminopyrazole Hsp90 Inhibitors

[mattodd]

Tweet from Chris Southan suggesting Resorcylate Aminopyrazole Hsp90 Inhibitors (paper). Toronto and Boston University. Anyone want to reach out to see if samples could be secured?

[wwjvdsande]

Good suggestion. Hsp90 is one of the stress response proteins which becomes upregulated under antifungal pressure. For other fungi it has been demonstrated combining an Hsp90 inhibitor with an antifungal agent results in synergy in vitro and enhanced survival in vivo. We did this in Aspergillus too. Unfortunatley most of the already available Hsp90 inhibitors such as 17-AAG are also reactive towards the human enzyme so focussing on structures more specific for the fungal Hsp90 could certainly be a way forward. From our proteomics data in the larvae we found M. mycetomatis Hsp90 present in the grain.

[cdsouthan]

There is an Hsp90 sequence from the Madurella mycetomatis genome https://www.ncbi.nlm.nih.gov/protein/KXX74442.1?report=genbank&log$=protalign&blast_rank=1&RID=T4F2E6JP015

The Candida albicans homologue https://www.uniprot.org/uniprot/P46598 has 74% identity by BLASTP.

[MFernflower]

BLASTP reports 86.34% similarity between hsp90 proteins from Madurella mycetomatis and Aspergillus niger

@wwjvdsande @cdsouthan

[mattodd]

@MFernflower data/source?

[cdsouthan]

As we know, for Hsp90 (and most other targets in fact) multiple alignments with phylogenetic clustering by % identity are relatively straightforward. As usual, the challenge is to find out which species/strains have had any chemistry directed specifically against them as targets, by which assay methods and whether any of the compounds can be sourced (e,g, the Candidia ones above). Ideally, we'd want above 75% identity but this would enough justify x-screening in the first instance (n.b. this was just a Swiss-Prot search, using the non-redundant proteins option will include many more genomic hits) @wwjvdsande @mattodd (just added PDB allighment and note the human identity is 65%)

[wwjvdsande]

In the past they also developed an antibody specific against candida HSP90 called mycograb developed by NeuTec Pharma. They actually did trials with it, but EMA refused marketing authorisation and after that not much was published on it. However they published the specific epitope on which the antibody was directed. It was KILKVIRK. Looking at the comparison above, this epitope is seen around Sbjt 385 (Candida) and is KIMVIKK in Madurella. I don't know if this is the same site where the molecules from the publication bind but it might be something worthwhile to focus on. At least the antibody did not crossreact with the human HSP90

[cdsouthan]

Good orthogonal thinking @wwjvdsande :) The PDBs should give at least a model of that epitope and make it worth a try in vitro? Putative small mol binding sites are a different matter but there may be some human target data

[MFernflower]

@wwjvdsande @cdsouthan Would be interesting if we can get the antibody AlexaFluor tagged

[wwjvdsande]

@MFernflower it might be interesting but I'm not sure if the antibody is still retrievable. Not much has been reported in literature anymore on this.