Mechanism of action - Githubissues

danaklug commented 4 years ago

This Issue is for discussion and tracking of mechanism of action experiments for series 2. We are currently in the process of planning the details of the first experiments and will update accordingly.

mattodd commented 3 years ago

We have data from Lee Graves and his group (Thomas Sterns Karim Gilbert)! I've posted it here.

@edwintse @danaklug could one of you please post a copy of the data to an ELN so we're sure we have it securely?

Lee says (by email): "Here is the data you have been waiting for. I hope there are some clues in here. I had Karim look at all proteins that were bound- not sure if any of these are kinases. What I tried to do was to look for things that showed compound #28-dependent changes, but not 26. There are some. I also tried to look for dose-dependent changes and again there are some. For example some that showed increased binding after treatment include ATPF, FEMB, MCSA, DLTD, ESAA, LYRA; down includes AZO1, CVFB, CRR, HISS IDH2, TRMB. Please don’t take my word for it- go over the data and we can talk."

Most interesting. Now we need to understand what it means.

danaklug commented 3 years ago

I’ve gone through the results documents and below is a summary of my interpretation, open to any feedback, corrections, or additional insights!

Summary of experimental procedure

I think the below image (from DOI: 10.1074/jbc.RA119.011083) sums up the experimental design nicely. In short, kinase inhibitors are immobilized on beads and cell lysate constitutes the mobile phase. The degree of protein binding to the beads is quantified and compared between a DMSO control and compound of interest. Decreased binding in the presence of compound indicates that the compound is competing for the binding site of the kinase inhibitor.

Data interpretation

From the “Data Interpretation” slide of file “20201022_PC781_Wilson_MIBs_Analysis”:

A log2 fold change (FC) of each sample compared to DMSO was calculated. A log2 FC ≤ -0.5 can be considered a kinase with decreased MIB binding/abundance. A log2 FC ≥ 0.5 can be considered a kinase with increased MIB binding/abundance. A more stringent cutoff is log2 FC -1 or lower/1 or higher. With a competition assay, the most compelling data will often be the protein groups that respond in a dose-dependent manner in either direction.

Experimental conditions

Active: OSA_000822/ALM-DAI-28 (“28” in “PC781_Wilson_MIBsComeptition”) Inactive: OSA_000820/ALM-DAI-26 (“26” in “PC781_Wilson_MIBsComeptition”) Cell type: MRSA

Follow-up

Kinases most meriting follow-up are those that show significant dose-dependent changes in MIB binding when treated with OSA_822, but not OSA_820. I went through the Excel file and filtered out kinases in the following order:

Those showing -0.5 < log2 FC < 0.5 at 50 uM OSA_822 (28)
Those that did not show a dose-dependent change when treated with OSA_822
Those that showed dose-dependent changes when treated with both OSA_822 and OSA_820

The following kinases remained, which seem to be good candidates for follow-up: sodA, mqo1, MW0361, SAR1965, purD, saeR, SACOL271, msrA2, purQ, esxD, SA2102, lyrA, femB, pyrR, lipA, mcsA

Next steps would be to find out what is known about these proteins and their role in MRSA biology, and potentially look into running some biochemical assays with our compounds.

Outstanding questions

What is the mechanism by which MIB binding/abundance would increase in the presence of compound? Upregulation perhaps? What does this tell us about the MoA?
From “20201022_PC781_Wilson_MIBs_Analysis” – what is LFQ?
Are the targets identified human kinases or MRSA kinases? (I assume MRSA based on Sheet3 of Excel document – to be confirmed.)

Giada-chem commented 3 years ago

I've done some reseaches about the kinases that @danaklug found as result of the proteomic data. I used NCBI and UNIPROT databases. Hopefully this will clarify the role of these enzymes!

SA2102 Gene - Putative formate dehydrogenase : https://www.uniprot.org/uniprot/Q99RW4

LyrA Gene - Lysostaphin resistance protein A: https://www.uniprot.org/uniprot/Q6GEA0

mattodd commented 3 years ago

An essential to do item for the above list is to figure out what's known about their essentiality. i.e. if inhibiting them would lead to cell death. Another to do item: do any of the promising targets have labs isolating and running assays against them? i.e. are they known targets? Also wondering about resistance experiments. Could we in theory develop bacterial resistance to these compounds and see whether changes map to the same proteins? From the meeting on Jan 29th (opensourceantibiotics/Series-2-Diarylimidazoles#53) it seemed as though the negative changes for mqo1 and PurD were most interesting.

danaklug commented 3 years ago

Thought this might be helpful - this is a screenshot of the targets that I pulled out from my analysis with the actual data associated.

MFernflower commented 3 years ago

One target one paper? @Giada-chem @mattodd @danaklug

MQO1 - this seems to be involved with metabolic adaptation and is a virulence factor - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4894658/ - Staphylococcus aureus Lactate- and Malate-quinone Oxidoreductases Contribute to Nitric Oxide Resistance and Virulence ------ no inhibitors seem to be known and no PDB entry - so if our compound hits it this could be groundbreaking.
PURD seems to be effected by various antibiotic compounds so this could be a red herring - see this https://www.biorxiv.org/content/biorxiv/early/2020/03/12/2020.03.10.986687.full.pdf

"In addition, genes involved in purine biosynthesis pathway (purC, purD, purF, purH, purK, purL, purM, purN, purQ) and pyrimidine biosynthesis pathway (pyrB, pyrC, pyrE, pyrP) were downregulated under IBG treatment"

Giada-chem commented 3 years ago

It seems that all our "selected kinases" are involved in saeR regulation : Rochat T, Bohn C, Morvan C, et al. The conserved regulatory RNA RsaE down-regulates the arginine degradation pathway in Staphylococcus aureus. Nucleic Acids Res. 2018;46(17):8803-8816. doi:10.1093/nar/gky584

" RsaE is a global and highly conserved regulator of central metabolic functions, including the TCA cycle, the glycine cleavage pathway, tetrahydrofolate metabolism and the catabolism of several amino acids including arginine" PurD is involved in purine synthesis ( from glycine cleavage system) and mqo1 is involved in TCA cycle, then we have also kinases involved in formate's, argine's, glycine's and amino acid's metabolism. Moreover saeR GENE-SAR0759: Response regulator SaeR has been detected.

MFernflower commented 3 years ago

@Giada-chem @danaklug @drc007 @mattodd

A partial x-ray crystallography file exists for SAER - but only covers positions 123-228 in the protein https://www.ebi.ac.uk/pdbe/entry/pdb/4QWQ

MQO1 doesn't seem to have structure data publically available

mattodd commented 3 years ago

Just on the potential connectivity issue: there's no reason why any of the potential targets should be functionally linked since this is a lysate experiment, correct? The molecule is added to the lysate, not to the MRSA prior to lysis?

It still seems to me to be important that we try to find any groups working on these proteins already: MQO1, MW0361, SAR1965, purD. Can anyone find anything recent, e.g. an enzymatic assay (in a perfect world!).

For transparency I asked a few questions of Lee by email (since he's not on Github) and will ping him again:

"Hi Lee,

Thanks. We just had a quick catch up and wondered about these things:

1) Hot targets identified from your experiments. So Dana was looking at the proteins that were shifted down in a dose dependent manner (here's the thread - scroll down) and it seemed to us that we wanted to look at things like sodA and purD. Are we reading it wrongly?

2) You said you wanted to repeat the experiment with more protein - that would certainly help our joint paper, great. Do you have the funds for this, and do you need any more compound?

3) The unbiased proteome-disturbance experiment you mentioned - same question. Is that doable, and do you have enough 26/28?

Giada's going to do some digging on whether any of these proteins are known antimicrobial drug targets, and/or whether there's anyone working on them already. And whether they're known to be essential.

Quick question, arising from my spending too long on malaria projects. Can we grow resistant bugs then sequence them to find out genetic changes associated with resistance? If there's a correlation then that's pretty strong evidence, obviously.

M"

Giada-chem commented 3 years ago

PurD (purine biosynthesis pathway) is involved indirectly in many drugs' mechanism of action:

I could't find anything about Staphylococcus aureus's mqo1 (putative Malate: Quinone Oxidoreductase) but I found a paper regarding mqo1 of Mycobateria:

Polymyxin B Identified as an Inhibitor of Alternative NADH Dehydrogenase and Malate: Quinone Oxidoreductase from the Gram-positive Bacterium Mycobacterium smegmatis

one paper regarding the aetiological Malaria agents:

Identification of Plasmodium falciparum Mitochondrial Malate: Quinone Oxidoreductase Inhibitors from the Pathogen Box

one paper regarding Pseudomonas taetrolens:

Purification and Characterization of a Malate:Quinone Oxidoreductase from Pseudomonas taetrolens Capable of Producing Valuable Lactobionic Acid

Importance of mqo gene:

The PEP-pyruvate-oxaloacetate node: variation at the heart of metabolism "However, no crystal structure is available at the moment, and details about the precise reaction mechanism are lacking."
Identification of a Lactate-Quinone Oxidoreductase in Staphylococcus aureus that is Essential for Virulence "Lqo and Mqo are differentially required for utilization of various carbon sources. Neither Mqo nor Lqo are essential for growth in chemically defined media using glucose (0.5%) as a primary carbon source. While Mqo is dispensable for growth on l-lactate (Figure 2) it is essential for maximal growth in chemically defined media using amino acids (0.5% Cas Amino Acids) as a primary carbon source."
Staphylococcus aureus lactate- and malate-quinone oxidoreductases contribute to nitric oxide resistance and virulence> "Nitric oxide (NO·) is an important effector of the innate immune response with both immunomodulatory and antibacterial activities. Inflammatory NO· is required for the clearance of a variety of pathogens. NO· is generated by activated leukocytes through an inducible NO·-Synthase (iNOS) and can act directly on invading organisms or indirectly through oxidation to reactive nitrogen species (RNS)." "Thus, both Lqo and Mqo are required for growth on peptides and L-lactate during moderate NO·-stress. These enzymes are highly similar (50% identity) and thus represent an attractive “double hit” target for anti S. aureus therapeutic development. "

here a paper regarding Saer Inhibitor: