search

opensourceantibiotics / murligase

Everything to do with the Mur Ligase Project
29 stars 6 forks source link

AZ Compound Binding Mode Justification issue #114

Open Yuhang-CADD opened 3 months ago

Yuhang-CADD commented 3 months ago

Correction here

C-terminus was truncated, not. central domain. check Joe's reply in https://github.com/opensourceantibiotics/murligase/issues/36 but does not influence the logic of the idea below

Dear team,

As the project progress, I realised one issue that cannot be properly explained at the moment and may require further investigations into: the binding mode of AZ compounds in Pae MurC ligase is so far not fully justified; they were captured to bind like what we saw in PDBs in a truncated protein and the interactions between the AZ hits and the central domain were omitted throughout the project design

Let me take AZ5595 as an example: What we saw:

Screenshot 2024-08-05 at 15 57 44

Since the central domain was truncated, we cannot justify this compound would bind this way.

If the protein was further engineered with N or C -terminus truncated (with AZ5595 soaked in the pocket), we might observe the other potent interactions (see as follows) which can still explain the nano molar potency in vitro.

Screenshot 2024-08-05 at 16 01 37

What we want to see:

Screenshot 2024-08-05 at 16 03 40

However, we were having difficulties in obtaining the crystal of full length Pae MurC protein (if I remember it right).

So, as a chemist I might not be able to bring in too much of an insight on how to do the protein work but I still would like to raise some shallow ideas:

  1. if the full-length protein was stable but just could not crystallise, can we potentially use CryoEM to obtain the structure?
  2. if the protein was unstable, can we try building up engineered Pae MurC structures with N or C terminus truncated and then soaked with AZ5595 or AZ compound 4?
  3. After obtaining the PDB of AZ compound 4 in Pae MurD, our understanding can be further justified? Screenshot 2024-08-05 at 16 10 24
Yuhang-CADD commented 3 months ago

Pae MurC very stable, go Xstallography But AZ5595 (WYH25) has DMSO that subtracts water to form Ice send AZ5595 and other compounds in solids not DMSO stocks worst scenario go CryoEM

Laura (AZ5595 in Pae MurC) & Scott (AZ compound4/WYH9 in Pae MurD)

Yuhang-CADD commented 3 months ago

Additional note here: AZ compound 4 was selected for Pae MurD co-crystalisation for two reasons:

  1. potentially can validate the ubiquitous binding mode of AZ compounds with Murs
  2. have a better multi-targeting data (see wiki)
Yuhang-CADD commented 2 months ago

Alphafold3 trying to provide full-length MurC structure (inputing truncated MurC Xstallography data to locate the coordinates and sequential data to generate the truncated central domain) @Yuhang-CADD try this