Oxamides, anyone?

Recently, I stumbled on a BMCL 2020 article from a team at BMS. In their quest for a new generation of HIV integrates inhibitors, they optimised a series of pretty sweet-looking (technical assessment) bridged tricyclic compounds . Actually, what really caught my eye was the oxamide functionality on those molecules.

I've always associated anything with adjacent carbonyl groups as being reactive and only useful in the context of reversible covalent inhibitors (for a recent review with a section on keto-amides see MedChemRes2023). I was wrong, and when dealing with oxamides, their mode of inhibition is good old fashion binding with no associated covalency.

What really gets exciting in this functional group is the shape and the number of interactions available to it. Unfortunately, there are not many examples of these in the CSD (small molecule X-Ray database with >1,000,000 entries) or in the COD (open small molecule X-Ray database with >500,000 entries). A couple of representative examples are depicted below.

What seems pretty clear is that mono-substituted oxamides are planar with the carbonyl facing opposite directions. As substitution increases, the planarity can be lost (presumably due to sterics) and a perpendicular conformation becomes either the lowest energy or at least readily accessible to the molecules.

Using those 2 conformational options and some really, really crude 3D overlays, it does look like oxamides could mimic sulfonamides.

(One pretty big caveat being the mismatch in exit vector... meaning that oxamides could never mimic the kink a central sulfonamide linker always gives to a molecule)

I tried to find examples from the literature to back this up. A striking example comes from J&J scientists (BMCL 2010) in the Cathepsin S literature. In their report, they investigated groups to replace the methyl sulfonamide group. Following an extensive exploration (I am only showing 2 cpmds below) they identified the oxamide as a suitable replacement. Co-crystal X-ray of close analogs illustrate that the oxamide can mimic the interactions made by the sulfonamide with the protein.

Mining the PDB confirmed oxamides often make multiple hydrogen bonds with the protein from either the carbonyls or N-Hs, as depicted on the left.

Given their versatility, it is not surprising to find oxamides being used as fragments in screening campaigns. Amos Smith et al identified oxamides as HIV-1 entry inhibitor starting points (JMedChem2012, ACSMedChemLett2021) and Merck/Edelris identified some as novel Cyclophilin D inhibitors (BMCL2019).

When "hunting" for drugs, however, finding ligands is only a part of the equation. The bigger questions for oxamides are how these would behave in vivo and whether they could actually end up in drugs.

It turns out that they are quite drug-like, as demonstrated by the HIV-integrase literature. Merck scientists worked extensively on this target and identified multiple well-behaved oxamide compounds (BMCL2006, JMedChem2008). This is best illustrated in the table to the right. In the JMedChem account of the discovery of the HIV-integrase drug Raltegravir, the authors also profiled an oxamide analog (cmpd 11). In the end, the compounds were quite similar from a PK perspective and the more active oxadiazole compound was selected for development.

Coming back to the original point around ketoamides/glyoxamides being used as reversible covalent inhibitors, I also realised that this was not always the case. There are 2 notorious examples where glyoxamides are found in drugs or drug-like compounds, while not relying on covalent binding.

Temsavir (for a page-turning account on its discovery, see MedChemRes2021) and Talmapimod (unfortunately, I was not able to find a MedChem account of its discovery but a story around a PET version can be found here: ACSChemNeuroscience 2023).

The same perpendicular orientation of the carbonyl groups can be found in both cases and as far as I can tell, the reactivity of the ketone is minimal (either because of stereoelectronic effects or the because it is more a vinyloguous amide than a ketone).

To conclude, oxamides were certainly never on my radar, but given their drug-like nature, good in-vivo behaviour and the number of possible H-bonds they can make, I now think they should join amides, sulfonamides, ureas in our library explorations.