tag:blogger.com,1999:blog-1136153439451224584.post2324419418544781435..comments2024-03-27T06:45:59.174-07:00Comments on Practical Fragments: Secondary ligand binding sites are commonDr. Teddy Zhttp://www.blogger.com/profile/07288045760981372367noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-1136153439451224584.post-19442897530678500542016-01-08T06:08:04.721-08:002016-01-08T06:08:04.721-08:00Hi, Fred here (one of the authors on the paper). T...Hi, Fred here (one of the authors on the paper). Thanks Dan for the review of the paper - very exciting to get mentioned on your blog! And thanks to bernatv for the comments. Just to clear one thing up: When we talk about sequence conservation in this paper we’re referring to conservation between orthologous proteins e.g. between Human CDK2 and Mouse CDK2, rather than conservation between different human proteins. <br /><br />(We try to be consistent in our use of the terms homolog, paralog and ortholog: human CDK2 and CDK5 are paralogs. Human CDK2 and mouse CDK2 are orthologs. Both are examples of homologs) <br /><br />When you talk about drug selectivity (infectious diseases aside) you’re generally only worried about the conservation between paralogs, e.g. how to generate selectivity for one particular human kinase vs another human kinase. The idea that allosteric/regulatory sites are less conserved than active sites when you compare different proteins <b>in the same species</b> seems pretty much accepted and we’re not arguing against this. This is of course one of the reasons that allosteric sites are interesting – they’re another way to achieve selectivity. <br /><br />To rephrase what we were trying to say in the paper:<br />We’re comparing the “same” protein in different animals. E.g. Human CDK2 vs Mouse CDK2. We would expect these to have similar regulatory mechanisms, given they evolved from some common ancestor however many tens of millions of years ago. If those regulatory mechanisms involve binding at various alternative sites around the protein we’d expect those sites to be preferentially conserved. <br /><br />During the x million years since mice and humans diverged from some common ancestor, their CDK2 genes have accumulated mutations. You can calculate the global sequence identity between mouse and human CDK2 and the sequence identity within specific regions (the primary and secondary sites). <br /><br />We observe that the residues in both primary and secondary sites are conserved more than the rest of the protein, suggesting some evolutionary pressure to conserve those regions. Of course this isn’t by itself proof of biological function, you’d need to look at each pocket in much more depth and do a lot of biological validation, but the point we’re trying to make is that these things occur very frequently (more than half of the targets we look at) and many of them pass some of the checks you’d want to do before investing in the biological validation. <br /><br />I hope that clears it up a bit, rather than just adding to the confusion!Fred Ludlowhttps://www.blogger.com/profile/03115822351404837657noreply@blogger.comtag:blogger.com,1999:blog-1136153439451224584.post-46246051190931218642016-01-07T20:11:00.736-08:002016-01-07T20:11:00.736-08:00Thanks bernatv for the Allosite analysis - very in...Thanks bernatv for the Allosite analysis - very interesting.<br /><br />As to what the conservation of allosteric sites means for developing selective inhibitors, I guess I'm less concerned. Just because the pockets are somewhat conserved doesn't mean they are absolutely conserved, and it only takes one mutation to block binding. After all, the ATP binding sites of kinases are functionally and structurally conserved, yet it is still possible to develop extremely selective inhibitors. This will be all the more true for allosteric sites where the conservation is less.Dan Erlansonhttps://www.blogger.com/profile/07927082337051189270noreply@blogger.comtag:blogger.com,1999:blog-1136153439451224584.post-2221623127404253562016-01-07T05:26:53.280-08:002016-01-07T05:26:53.280-08:00I checked the correlation of the results in the pa...I checked the correlation of the results in the paper and Allosite onlinte tool for two proteins and it seems like they align pretty well (except those shallow binding pockets on the periphery of proteins). https://bernatv.wordpress.com/2016/01/02/finding-new-pockets-with-fragments/<br />What bothers me the most is that argument that allosteric sites are conserved almost as much as orthosteric ones. All that I've learned up to now was based on the opposite assumption.Anonymousnoreply@blogger.com