we also do structure-activity relationship and rational drug design. And I agree with Christian: never rely on one single method and try to include a homogenous assay, such as ITC or FT. you mention a tyrosine involved in the binding pocket. Did you try to track the Tyr in FT?
best, matthias ________________________________ From: CCP4 bulletin board <[email protected]> on behalf of Christian Roth <[email protected]> Sent: Friday, April 27, 2018 9:00:03 AM To: [email protected] Subject: Re: [ccp4bb] Compound with flexible conformation but nM Kd Anectodal evidence I have heard from colleagues working with things, which are immobilized is that the measured Kd value on the surface can be wildly different from what is measured in solution. A superbinder on a surface might not be as good in solution. There seems still a lot of debate why that is. Cheers Christian On Fri, Apr 27, 2018 at 5:07 AM, WENHE ZHONG <[email protected]<mailto:[email protected]>> wrote: Hi Philippe, The affinity was measured by SPR where we immobilized the protein on the chip. One thing I forgot to mention is that the association rate (kon) shown in SPR experiment for this compound is faster (>10-fold faster) compared to other analogues with similar koff. There is a pi-pi interaction between the scaffold structure and the protein (tyrosine ring). Is it possible that the hydrophobic substituent could facilitate the formation of this pi-pi interaction but not necessary to involve in the interaction? Thanks. Kind regards, Wenhe On Apr 27, 2018, at 1:50 AM, DUMAS Philippe (IGBMC) <[email protected]<mailto:[email protected]>> wrote: Le Jeudi 26 Avril 2018 16:50 CEST, WENHE ZHONG <[email protected]<mailto:[email protected]>> a écrit: Just to be sure: how was the nM affinity evaluated ? By in vitro measurements, or by obtaining an IC50 by tests on cells ? Of course, if you are mentioning an IC50, you may have a measurement of the efficacy of drug entrance in the cells, not just of specific binding to your protein target. Philippe D. Dear Community, A little bit out of topic here. We are applying the structure-based approach to design compounds that can bind our protein target. We have synthesized a series of analogues based on the same scaffold with different substituents at one particular site. The most potent analogue (nM Kd) has a long alkyl chain substituent. We thought this hydrophobic substituent should have strong interactions with the target protein leading to nM range affinity. However, crystal structures show very weak densities for this substituent and no obvious interaction between the substituent and the target protein, suggesting that this long alkyl chain substituent is flexible without binding to the protein. This binding site is relatively negative charged according to the electrostatic potential analysis. So it is a puzzle to me that how this dynamic and hydrophobic alkyl chain substituent can lead the compound to achieve nM affinity (>10-fold better than any other substituent) — in particular the binding site is not hydrophobic and no interaction is found between the substituent and the protein. Anything I have miss here that can increase the binding affinity without interacting with the target? Thanks. Kind regards, Wenhe
