Structure of human CAII with OSM-S-106

[Sarahlischen]

Hi everyone,

I’m Sarah and a second year PhD student at the Griffith Institute of Drug Discovery. My project focuses on different proteins that are interesting targets for fragment screening using native state mass spec and crystallography. I work in the lab of Sally-Ann Poulsen and with the Halili-Martin group in Brisbane, but at the moment I am based in Melbourne at CSIRO to learn more about crystallography under supervision of Tom Peat. And I did my first crystallography experiments on the human CAII. Nice to meet you all 😃

Recently, we were able solve a crystal structure of hCAII with the compound OSM-S-106 #14 OSM-S-106 is bound to hCAII with the nitrogen of the sulfonamide coordinated to the zinc ion in the active site of the enzyme as usually reported for sulfonamides. In addition, the sulfonamide oxygen participates to a hydrogen bond with the amide nitrogen of Thr199. There seems to be a potential interaction between the Thr200 and the aromatic ring system of the compound. The scaffold of the inhibitor is oriented towards the hydrophobic part of the active site.

Fig 1 The residues with potential binding interactions and the Zn(II) ion (purple sphere) with the three coordinating histidines His94, His96 and His 119 are shown. There is a possible interaction between the Thr200 side chain hydroxyl (3.00Å) and the aromatic ring system of the compound. The Fo-Fc (difference) electron density map for the compound is represented at a 3 σ contour level.

Fig 2 Binding interactions between OSM-S-106 (coloured in orange) and residues (coloured in grey) of the active site of hCAII. Two hydrogen bonds (indicated in yellow) are formed between the Thr199 backbone nitrogen (3.0Å) and the side chain hydroxyl (2.8Å) with the sulfonamide. There are also polar interactions between Gln92 and water molecules (red) with the scaffold of the compound. The residues Ile91, Val121, Phe131, Leu198 and Trp209 form potential hydrophobic interactions with the compound at 3.5 – 4.0Å

Fig 3 Surface of the enzyme active site with OSM-S-106. Hydrophobic residues are coloured in blue and hydrophilic residues and coloured in red.

At the moment I'm trying to express and purify the PfηCA in the hope of crystallising it, but it is not that easy. The protein doesn't seem to be very stable and/or the expression is very poor. Compared to the human isoform, the Plasmodium variant is much larger (600aa vs ~260aa) and has lysine rich regions in the sequence. There are indications that it is a new class of CA (η-class), as it is hypothesised that the zinc ion is not coordinated by 3 histidines, but by 2 histidines and one glutamine, which still needs to be confirmed. But that's another chapter... In any case, I will keep you posted ☺️

[holeung]

Great work! Can the binding mode help you create a homology model for Pf CA?

[mattodd]

Great post @Sarahlischen about crucial new results. @holeung yes, @drc007 started on this at #21 . I was talking with Reuben (@Skame161), who's doing a student project on OSM-S-106 for a few months, about comparing the two binding modes to check that they are similar, or at least that the interactions are similar. We can now do this!

[Skame161]

Greetings @Sarahlischen,

I'm Reuben. I'm doing my Master's in Drug Discovery at UCL, @mattodd is my research project supervisor, and like he said, I'm doing some work on OSM-S-106.

Great to see the results! Do you have a pdb file of the hCAII with OSM-S-106 bound that we could access? We would love to compare this to the homology model of CA #21 and, if we have time, run docking experiments with some OSM-S-106 analogues.

Thanks in advance and good luck with PfηCA. Hope you manage to get it to work for you.

[Sarahlischen]

@Skame161 here is the pdb file of the structure, good luck with it :)

hCAII_OSM-S-106.pdb.zip

[drc007]

@Sarahlischen @Skame161 @mattodd @holeung A while back I created a homology model of PfηCA (https://github.com/OpenSourceMalaria/Series3/issues/21) based on PDB code 1Y7W a Halotolerant alpha-type carbonic anhydrase (DCA II) as the closest match (<20% sequence homology).

I thought it might be interesting to compare the homology model of PfηCA with crystal structure of hCAII with the compound OSM-S-106.

The crystal structure of hCAII is shown in light blue and the homology model of PfηCA in orange. Overall things look pretty good, however the overall shape of all CAs is pretty well conserved despite very little sequence homology (Perhaps an example of convergent evolution).

Looking at the active site in particular.

I made the decision to coordinate 3 histidines with Zn in the homology model, this might not be correct but it looks believable. I then docked the compound OSM-S-106, and the result is shown above. As with the crystal structure of hCAII the nitrogen of the sulphonamide interacts with the Zinc atom and hydrogen bonds are formed between the sulphonamide and Thr217 (corresponds to Thr199 in human CA). The aromatic system of OSM-S-106 appears to be occupying a different space, this could be an artifact of the docking or since we know there is good selectivity between hCAII and PfηCA it might reflect a different binding mode.

It would be fantastic to get a crystal structure to confirm.

[Skame161]

I just finished looking at a similar comparison of the homology model and the hCAII on BIOVIA Discovery Studio Visualiser to get a 2D comparison of the residue interactions.

This first one is from the homology model...

...and this second one is from the hCAII crystal structure.

The interactions of the sulfonamide with the Zn and Thr residue show up here too. It looks like the other residues involved are different and it's also interesting to see that in the hCAII the NH on the thienopyrimidine core appears to be interacting with water molecule present in the active site.

[Skame161]

Greetings, all.

I thought it might be interesting to do a 3D comparison of the two models while I was running some docking experiments #25 and see what differences there are around the active site. The first image here is the homology model active site surface in 3D and the second is the same from the hCAII crystal structure.

The 3D shape of the active site is observably different between the two models and probably explains why OSM-S-106 appears to sit in a different orientation when the two enzymes are overlaid. If the inhibition of CA is the MoA of OSM-S-106 then this difference in 3D structure may explain why the compound is selective for pfCA over hCAII.

Of course, there is the caveat that we are working with a homology model of pfCA rather than an actual crystal structure and the results of the docking experiments need to be considered before making any conclusions.

[Skame161]

@Sarahlischen

Greetings,

Am I understanding correctly that the current challenge with the PfηCA is adequate expression rather than crystallisation? Was just wondering if the enzyme is truncating at a consistent point in the sequence and if there might be any feasability/advantage to producing sequences of a known length that could subsequently be clipped together?

Apologies if I'm being a little naive. Protein expression was never a strong point of mine.

Hi everyone,

I’m Sarah and a second year PhD student at the Griffith Institute of Drug Discovery. My project focuses on different proteins that are interesting targets for fragment screening using native state mass spec and crystallography. I work in the lab of Sally-Ann Poulsen and with the Halili-Martin group in Brisbane, but at the moment I am based in Melbourne at CSIRO to learn more about crystallography under supervision of Tom Peat. And I did my first crystallography experiments on the human CAII. Nice to meet you all 😃

Recently, we were able solve a crystal structure of hCAII with the compound OSM-S-106 #14 OSM-S-106 is bound to hCAII with the nitrogen of the sulfonamide coordinated to the zinc ion in the active site of the enzyme as usually reported for sulfonamides. In addition, the sulfonamide oxygen participates to a hydrogen bond with the amide nitrogen of Thr199. There seems to be a potential interaction between the Thr200 and the aromatic ring system of the compound. The scaffold of the inhibitor is oriented towards the hydrophobic part of the active site.

Fig 1 The residues with potential binding interactions and the Zn(II) ion (purple sphere) with the three coordinating histidines His94, His96 and His 119 are shown. There is a possible interaction between the Thr200 side chain hydroxyl (3.00Å) and the aromatic ring system of the compound. The Fo-Fc (difference) electron density map for the compound is represented at a 3 σ contour level.

Fig 2 Binding interactions between OSM-S-106 (coloured in orange) and residues (coloured in grey) of the active site of hCAII. Two hydrogen bonds (indicated in yellow) are formed between the Thr199 backbone nitrogen (3.0Å) and the side chain hydroxyl (2.8Å) with the sulfonamide. There are also polar interactions between Gln92 and water molecules (red) with the scaffold of the compound. The residues Ile91, Val121, Phe131, Leu198 and Trp209 form potential hydrophobic interactions with the compound at 3.5 – 4.0Å

Fig 3 Surface of the enzyme active site with OSM-S-106. Hydrophobic residues are coloured in blue and hydrophilic residues and coloured in red.

At the moment I'm trying to express and purify the PfηCA in the hope of crystallising it, but it is not that easy. The protein doesn't seem to be very stable and/or the expression is very poor. Compared to the human isoform, the Plasmodium variant is much larger (600aa vs ~260aa) and has lysine rich regions in the sequence. There are indications that it is a new class of CA (η-class), as it is hypothesised that the zinc ion is not coordinated by 3 histidines, but by 2 histidines and one glutamine, which still needs to be confirmed. But that's another chapter... In any case, I will keep you posted ☺️

[gchoy-specs]

@Sarahlischen

I am trying to get hold of compound OSM-S-106 for a non governmental organization that also wants to work on this compound. Could you let me know how you got the compound? If it was synthesized, can you share the synthesis route or can you share the supplier contact details where you got it from?

[MFernflower]

@mattodd

Message ID: @.***>

[gchoy-specs]

I am trying to get hold of compound OSM-S-106 for a non governmental organization that also wants to work on this compound. Could you let me know how you got the compound? If it was synthesized, can you share the synthesis route or can you share the supplier contact details where you got it from?

Please contact george.choy@specs.net

[mbhebhe]

Hi @gchoy-specs , we made this compound and the synthetic route can be found here. You are based in the Netherlands right? If there is some in London, the compound could be sent to you. How much do you need?

[gchoy-specs]

Hi @mbhebhe, thank you the synthetic route. Our non governmental organization customer requested 25 mg, so if you have that, that would be much easier than conducting a whole new synthesis. How much would you charge for 25 mg? And do you also have the spectral data (LC/MS and/or H-NMR) available?

[gchoy-specs]

and we are indeed based in the Netherlands