Nirmatrelvir

Pharmaceutical[edit]

Coronaviral proteases cleave multiple sites in the viral polyprotein, usually after there are glutamine residues. Early work on related human rhinoviruses showed that the flexible glutamine side chain in inhibitors could be replaced by a rigid pyrrolidone.^[9][10] These drugs had been further developed prior to the COVID-19 pandemic for other diseases including SARS.^[11] The utility of targeting the 3CL protease in a real world setting was first demonstrated in 2018 when GC376 (a prodrug of GC373) was used to treat the previously 100% lethal cat coronavirus disease, feline infectious peritonitis, caused by feline coronavirus.^[12] Nirmatrelvir and GC373 are both peptidomimetics, share the aforementioned pyrrolidone in P1 position and are competitive inhibitors. They use a nitrile and an aldehyde respectively to bind the catalytic cysteine.^[13][14] Pfizer investigated two series of compounds, with nitrile and benzothiazol-2-yl ketone as the reactive group, respectively, and in the end settled on using nitrile.^[15]


Nirmatrelvir was developed by modification of the earlier clinical candidate lufotrelvir,^[16][17] which is also a covalent protease inhibitor but its active element is a phosphate prodrug of a hydroxyketone. Lufotrelvir needs to be administered intravenously limiting its use to a hospital setting. Stepwise modification of the tripeptide peptidomimetic led to nirmatrelvir, which is suitable for oral administration.^[3] Key changes include a reduction in the number of hydrogen bond donors, and the number of rotatable bonds by introducing a rigid bicyclic non-canonical amino acid (specifically, a "fused cyclopropyl ring with two methyl groups"^[15]), which mimics the leucine residue found in earlier inhibitors. This residue had previously been used in the synthesis of boceprevir.^[18] Tert-leucine (abbreviation: Tle) used in the P3 position of nirmatrelvir was identified first as optimal non-canonical amino acid in potential drug targeting SARS-CoV-2 3C-like protease using combinatorial chemistry (hybrid combinatorial substrate library technology).^[19][20]


The leucine-like residue resulted in loss of a nearby contact with a glutamine on the 3C-like protease.^[15] To compensate Pfizer tried adding methane sulfonamide, acetamide, and trifluoroacetamide and discovered that of the three, trifluoroacetamide resulted in superior oral bioavailability.^[15]