Feb 2022 Update on Recent Mur Ligase Crystal Structures

[mattodd]

As will be discussed in #66 there have been two new structures deposited:

Nov 2021: Pseudomonas aeruginosa MurD, SSGCID code PsaeA.17938.a, PDB 7SY9. Jan 2022: Acinetobacter baumannii MurD in complex with ADP, SSG code AcbaC.17938.a, PDB 7TI7.

There was subsequent email correspondence, summarised here:

Nov 2021 @eyermanncj had asked Jan Abendroth:

1) "Is there any possibility of getting a Pae murD structure with ADP? It seems that murD requires a ligand present to obtain the closed conformation. The closed conformation is what would be most informative for new design work." 2) Best case would be to get Pae murD with the same substrates bound as in Eco murD (PDB: 2UAG) 3) The nano-DSF results for the AZ compounds were less than encouraging, but it would still be good to try to co-crystallize one or more of the four AZ compounds in Pae murD to give us an example of binding of the same compound to both Pae murC and murD. Nano-DSF suggests a long shot, but the value of such a structure pair is very high for the project.

Dec 2021 Jan Abendroth provided an update:

1) MurD, ADP The co-crystallization of MurD from A. baumanii with ADP was successful. Yesterday, I collected a 2.3Å in-house data set that clearly has ADP bound (first attached picture, below).

More crystals are on their way to the synchrotron for possibly better diffraction on Thursday. The data set at hand will be good enough for the refinement of the structure. There definitely is a closing of the structure compared with the apo structure, see second attached screenshot, below: green for the ADP-bound structure, blue for the apo structure, RMSD is 3Å.

No crystals in parallel set-ups with P. aeruginosa MurD.

2) MurD + AZ compounds We can certainly set up trays with MurD and AZ compounds, despite a shift in binding assays. Since we have limited supplies of the protein, I just want to be cautious and set up crystallization experiments that are more likely to succeed. Given the conformational change that we see between the apo and the ADP-bound MurD structure, do we know which form the AZ compounds bind to? Would it make sense to run the binding assay in the presence of ADP again?

@eyermanncj replied "I vote for trying the nano-DSF in the presence of ADP as they should bind the closed form of murD. Hopefully they can displace the ADP in the binding assay. Does that make sense? I am not sure if it matters which AZ compound to try in the nano-DSF assay with ADP. Maybe just try one AZ compound to conserve murD protein."

3) For the 7TI7 structure, Jan Abendroth pointed out "there are distinct conformational changes between the apo and the ADP-bound structure. RMSD is at 3Å, and especially the C-term is hard to overlay. A curious observation is the carbamylation / carboxylation of Lys 198, now named KCX 198. I have seen this modified Lys in only a few structures within SSGCID. It typically serves as an ‘elongated Glu’ that tends to ligate active site waters or ions."

Frank von Delft asked: Is that carbamoylated lysine pointing anywhere important?

Jan responded: "This is a clear observation based on very nice data. The tip of the side chain ligates a Magnesium that is in close proximity to the beta phosphate of ADP. I just wanted to make sure that this is known and considered. The equivalent residue in the current P. aeruginosa structure (Lys 202) is likely unmodified. Better data might give us a confirmation."

Laura: The modification seems to be transient and important for the reaction (Mg2+ binding). This is discussed in this paper. None of my EcMurD structures (crystals grown as apo protein) present the carbamoylated lysine.

@LauraDS1 and @eyermanncj: "we are hopeful that soaking of AZ compounds into ADP bound murD ligase might yield a x-ray structure of an AZ compound bound to murD."

Jan was (1/22) collecting a data set on a Psae MurD crystal grown in the presence of ADP. Clearly have ADP bound:

Other misc notes

From Jan (13/1/22): "Laura, to an earlier question on crystallization of Psae and Acba with ADP:

Acba-MurD cystallized under a number of conditions: The 1.5Å data set came from a crystal from Morpheus A6. Other, similar looking blocks appeared under MCSG1 B2, D4, F11 and JCSG+ H8, H9.

I just found crystals of Psae-MurD in JCSG Top96 a6. I will screen these crystals in-house tomorrow. I’ll report back if we get data.

Overall, we found that the protein concentration for both Acba and PsaeA MurD needed to be quite high to get crystals. We use ~40mg/ml. That might be worthwhile trying for the E.coli protein."

From Laura (14/1/22) "In the case of our EcMurD, 10 mg/mL did the trick. Also, the untagged protein hasn’t produced crystals, only the His-tagged form. This is consistent with the published structures, as well as the crystallisation condition. I need to confirm diffraction and I will be able to do so as soon as Diamond is up and running. In the meantime, I’m trying to improve reproducibility."

[mattodd]

Update received Feb 8th 2022 from Jan Abendroth (ppt files uploaded here): "we have a structure for P. aeruginosa MurD + ADP. Diffraction was very noisy, however, the density for ADP is convincing. RMSDs with apo P. aeuruginosa MurD are around 3Å, with ADP-bound A. baumanii MurD around 2Å. See slides. We will give it one round of crystal optimization, worst case we can finish up with the data we have."

[mattodd]

Further update from Jan Abendroth (Feb 28th 2022): "Please find the just deposited structure of Pseudomonas MurD in complex with ADP. We got significantly cleaner diffraction and data to 1.95Å resolution just a good week ago. There are three copies of the protein in the asymmetric unit. The N-terminal domain of chain C is only poorly ordered." Relevant files posted here.

Posted online here.

[eyermanncj]

@mattodd @Yuhang-CADD @LauraDS1 @chrisdowson1 @drc007 @loriferrins @bartrum @Rebecca-Steventon @LizbeK

Thanks Jan Abendroth!!!!

Here are some figures of Pae murC with the AZ inhibitor and the Pae murD with ADP bound. It seems that when an inhibitor/substrate is bound to Pae murD there is substantial movement of the murD protein (as previosuly established for Eco murD) and a long helix will prevent the AZ compound from binding due ot steric clashes with the benzyl group of the AZ inhibitor. So it seems further design work is needed for the AZ series to achieve a dual Pae murC/D inhibitor.

Pae murC AZ active site

Pae murD ADP active site

Pae murC AZ (gold) versus Pae murD ADP (purple) top view

Pae murC AZ (gold) versus Pae murD ADP (purple) side view

[eyermanncj]

There is a pyMol file with the Pae murC and murD x-ray structures overlaid here

https://github.com/opensourceantibiotics/murligase/tree/master/FIle%20Diary/2022/

[phraenquex]

Zounds, that's a lovely clean result from a structure, for once! @eyermanncj is that benzyl group in the AZ compound doing any useful work in murC? (I didn't find it easy to see.)

[mattodd]

@eyermanncj additionally shared by email (9/3/22) a paper on the dynamics of Acinetobacter baumannii murE. This is relevant to the discussion about full-length vs truncated protein that was had in #70.