Guillaume Médard

Thanks #ChembioParis2025 for organising this nice cruise through Paris yesterday evening, following an inspiring conference day! Looking forward to another day of great #chembio today.

2/2
Thanks to Le Studium, Loire Valley Institute for Advanced Studies, for funding this exciting consortium with great collaborators! Honoured to be involved for proteomics and chemoproteomics projects.

Back to Greece after a few intense scientific days in Tours, France, in the frame of the « Pharmacological Inhibition of Cathepsin C in Neutrophil-Mediated Inflammatory Diseases » consortium coordinated by Brice Korkmaz at the Centre of studies for Respiratory Pathologies (CEPR). 1/2

Finally the paper to convince you to toss that nanoLC off a roof? If you've ever been to this blog  before, you might have encountered the fact that I despite nanoflow liquid chromatography.  These author...

proteomicsnews.blogspot.com/2025/09/fina...

Serendipitous and Systematic Chemoproteomic Discovery of MBLAC2, HINT1, and NME1-4 Inhibitors from Histone Deacetylase-Targeting Pharmacophores Metalloenzyme inhibitors often incorporate a hydroxamic acid moiety to bind the bivalent metal ion cofactor within the enzyme’s active site. Recently, inhibitors of Zn2+-dependent histone deacetylases...

doi.org/10.1021/acsc...

Target deconvolution of HDAC pharmacopoeia reveals MBLAC2 as common off-target - Nature Chemical Biology Lechner et al. used an affinity-based chemical proteomics approach to investigate the target landscape of HDAC inhibitors, and identified an extracellular vesicle regulator MBLAC2 as a universal off-t...

doi.org/10.1038/s415...

« Médard et al. used a chemoproteomics approach to investigate the off-target effects of tubacin. They identified MBLAC2 as a novel target, and its inhibition or silencing in HEK293 cells increased EV accumulation, suggesting a regulatory role in EV secretion » .

FEBS Press The 2nd MObility for Vesicle research in Europe (MOVE) Symposium in Belgrade brought over 280 attendees from 28 countries to advance extracellular vesicle (EV) research. Featuring keynotes, presentat...

Nice to be mentioned in the recently published account of the conference!
Kudos to @sevlechner.bsky.social
#EV #chemoproteomics

febs.onlinelibrary.wiley.com/doi/full/10....

Stirring Bars are Superstition?

Do you really need to stir your reactions? How do you know?

go.bsky.app/JKJ1ZHt

Starter pack: go.bsky.app/2vBi9Ya
Let’s get this list complete !

#ChemBio #chemoproteomics @kusterlab.bsky.social

Miniaturisation of the assay allowed by mind-blowing progress in mass-spectrometry equipment and proteomics workflows is a story for another day …

For the mixing, we have to consider that less is sometimes more. Since there is little chance of achieving ideal complementary of probes with no overlaps, kinases enriched by many probes will dilute the ones enriched by only one in the mass-spec readout, eventually leading to poor quantification.

Full proteomes and transcriptomics help select biological material with good kinome diversity for our purpose … to a certain extent only, because many kinases have low abundance and because transcripts and protein levels are not always proportional.

Is kinase X nowhere in the data because no probe can fish it, or because this bugger is not expressed decently in any of the biological material we use? Do we need to test more probes or more kinome-sources? Indeed, all kinases are not expressed everywhere and all the time. Proteomes are dynamic.

Then comes the testing of individual probes. Once covalently attached to beads, they need to robustly enrich many native kinases of the human (or other) kinome in so-called pulldown experiments. And here comes the challenge and its vicious circle:

Or we can also be super pragmatic, avoid synthesis, and limit ourselves to commercial synthetic intermediates and inhibitors which already have an immobilisation handle at the right spot.

If the final molecules are already quite unselective, that is great for us. If the same molecular scaffold is used for a range of inhibitors relatively selective towards different kinase, great again. And, with those data, we can prioritise the molecules we put synthetic efforts in.

...“best” as a drug i.e. sub-optimal for us, since they achieve a certain level of selectivity. This criterium of selectivity is generally assessed for all (at best) molecules described against only one or two selected other kinases than the intended target.

We want a promiscuous probe though … do we have selectivity data in the literature? There will, in general, be panel selectivity data for one or two of the best inhibitors of a class …

Looking at X-ray co-crystal structures or docking experiments, we can see where the solvent-pointing part of the molecule is. Can we design a molecule with an immobilisation handle on this spot, that can be easily synthesized? Great. Here we go, this could be an affinity probe.

The latter are here to achieve a certain level of selectivity towards one kinase or favourable pharmaco-kinetics/dynamics properties, so we want to remove them. Basically, we mentally do the opposite of a medicinal chemist job: we trim down instead of slowly crafting.

We rely on the structure-activity relationships data tables of each of the many reported molecules to delineate what is the pharmacophore (the part of the molecule we need to absolutely keep for potency) and what are the fine-tuning elements that were introduced by the chemists on the pharmacophore.

In practice, now. For the selection of a probe for a sub-kinome, we mine the scientific literature and patents for molecules that are designed to hit various kinase families.

Quite simply, conceptually: we select potent, yet promiscuous pharmacophores that can be immobilised on beads without losing their potency and which are complementary to each other. We make the molecules, prepare the beads, mix them. Bob’s your uncle.

Cellzome published a mixture of immobilised kinase inhibitors in 2007, that constituted the first Kinobeads. I joined the Kuster lab in 2011 as a group leader with the clear mission of making Kinobeads even better. How do we do that rationally?

It is also a good opportunity to reflect on the Kinobeads chemoproteomics assay development over 20 years. Particularly, on the Kinobeads affinity matrix itself.

The “target landscape of clinical kinase drugs” (2017) has just gotten its 900th citation! There is no shame, I believe, to feel proud of having been part of this effort of the Kuster lab to provide such an - apparently very useful - interactive resource to the scientific community.

@medard-chemprot.bsky.social in the chembio starter pack for tools (chemoprotemics affinity matrices for kinases and HDACs and target deconvolution/profiling)?