Docking Experiments and Comparison of CA Homology Model with Known hCAII Crystal Structure

[Skame161]

@mattodd @Sarahlischen @holeung @drc007

Greetings, all.

I ran a couple of docking experiments with the CA homology model and the hCAII crystal structure #24 to compare the interaction of OSM-S-106 and a couple of its methylated analogues (image below) that we know showed a complete loss of potency (this page) to see how/if substituting the nitrogen of the sulfonamide affects binding. I ran hCAII twice, one with ALL the waters removed and one with water retained in the active site, just to be on the safe side, but it doesn't seem to have made any noticeable difference to the results overall.

As far as I can see, the data indicates that increasing the substitution on the nitrogen significantly impacts the ability of the nitrogen to interact with the Zinc in the active site. Across all three docking runs, both TF-18-1 (the dimethyl analogue) and TF-16-1 (cyclo analogue) showed effectively a total loss of N-Zn interaction and TF-17-1 (the methyl analogue) had noticeably less poses that retained the N-Zn interaction when compared to OSM-S-106.

IF the N-Zn interaction in pfCA is important for potency then this would seem to suggest a possible explanation for the loss of activity in these three analogues at least.

The ELN can be viewed here

[MFernflower]

@Skame161 might not be a bad idea to propose a few that you think should be potent

[Skame161]

@mattodd @Sarahlischen @drc007 @holeung

Greetings, all, (again)

Here is a (edited) summary of the docking experiments that I ran this week (ELN here) with some thoughts on the results. The following image is a summary of the analogues investigated in these docking experiments and the file linked here is a tabulated summary of the key results from docking into the homology model.

The minimised affinity on redocking OSM-S-106 into HM was -7.0 and almost all the analogues docked here were within 10% of this. The actual range observed was -6.2 to -8.1; however, only one analogue was observed to have a binding affinity < -7.6 (OSM-S-137 at -8.1) with all the others lying between -6.2 and -7.6 (-6.9 +/- 10%). The lower value observed with OSM-S-137 could be attributed to the molecule's significantly larger size compared to the other analogues (due to the bulky group added to the structure) and this might suggest that the minimised affinities are probably not significantly different between the different analogues.

The RF Score for redocking OSM-S-106 was 6.1 and all the analogues (including OSM-S-137) had an RF Score +/- 0.1 which indicates that the RF Scores are not significantly different between the analogues.

The analogues showed the same N-Zn distance of 2.7A as OSM-S-106 with the exceptions of TF-3-1, TF-4-1 and PT-22 where the N-Zn distance was 3.0A +/- 0.1A.

The analogues appear to adopt poses that are comparatively similar and in relatively the same space as OSM-S-106 (see this file) except for OSM-S-129. TF-3-1, TF-4-1, PT-22 appear slightly "elevated" in the active site which is consistent with the increased N-Zn distance.

TF-3-1, TF-4-1 and PT-22 were originally synthesised and evaluated to investigate if adding methyl groups (see figure above) and deplanarising the molecule would impact the activity. The data appears to indicate that the addition of these methyl groups does indeed impact the ability of the compund to form the typically expected N-Zn interaction (almost a 15% increase in the distance between the N and the Zn and well outside the usually reported range for N-Zn interaction) which is the most likely explanation for the complete loss of activity in these analogues.

OSM-S-129 was originally synthesised and evaluated to investigate if relocating the sulfonamide from the meta position to the para position (see figure above) would affect the activity and the compound was inactive when biologically evaluated. The data suggests that the more linear arrangement of this compound changes its orientation in the active site and limits the N-Zn interaction. Although some interaction is still possible (and that interaction appears to remain similar to the original compound), it appears that it is insufficient to retain biological activity.

This indicates that the planar orientation of the compound and the meta location of the sulfonamide are both important for binding and altering these will negatively impact any biological activity that can be attributed to interaction with carbonic anhydrase (CA). It also indicates that an analogue of OSM-S-106 that retains these key features should bind to the CA and would be expected to show some activity during biological evaluation IF binding/inhibition of CA is the MoA.

The lack of activity in almost all of the compounds investigated here suggests that inhibition of CA may not be the MoA!

The following should be noted however.

1) It is previously suggested that CA is a mechanism by which OSM-S-106 either accumulates in the host/parasite or acts as a transport mechanism to a site where the antiparasitic activity occures. The results from these docking experiments do not contradict this proposal.

2) OSM-S-137 appeared to be quite potent when evaluated biologically. A related analogue, OSM-S-136 (not shown here), also appeared to show some potency. The structure of these two analogues is quite similar to known kinase inhibitors (see this file); therefore, it is proposed that the activity of OSM-S-137 (and OSM-S-136) is an off-target effect from the inhibition of one or more kinases and not CA.

3) TF-8-1 showed limited activity in vitro against pf yet the data suggests that if TF-8-1 is active then TF-7-1 should also be active (TF-6-1 and TF-9-1 have not been synthesised not biologically evaluated). Could this have been a false positive or action through some other mechanism that has not yet been considered? Further investigation may be necessary.

4) We are still working with a homology model of pfCA rather than an actual crystal structure and the homology model may have defects in it that cannot be known without comparison to an actual crystal structure. It would be good practice to rerun these docking experiments should the crystal structure of pfCA become available.

5) A crystal structure of hCAII with a known hCAII inhibitor (sulpiride) bound to it has been obtained from Claudiu T Supuran (his original paper here) to investigate if redocking a known hCAII inhibitor produces produces a similar minimised affinity, RF Score and N-Zn distance.

PS. I updated the title of this issue to reflect that I'm mainly posting the results of docking experiments in here.

PPS. Does anyone know a good way to add images to a comment that allows me to resize them so they're not massive or so that I don't to spend ages resizing/redrawing things before I add them to a post/comment? (Answered by @drc007 )

[Sarahlischen]

Hi @Skame161, the sulfonamide-N-Zn distance in the crystal structure is 1.9Å. Hope that helps :)

[drc007]

@Skame161

• replace![image](https://your-image-url.type) with <img src="https://your-image-url.type" width="100" height="100">

Edit the size parameters to suit. You don't have to provide both width and height, only providing one dimension means you don't have to worry about the image becoming distorted.

[Skame161]

@Sarahlischen @drc007

That's great! Thank you both!