The β-lactams remain the most important antibacterials, but their use is increasingly compromised by resistance, importantly by β-lactamases. Although β-lactam and non-β-lactam inhibitors forming stable acyl-enzyme complexes with nucleophilic serine β-lactamases (SBLs) are widely used, these are increasingly susceptible to evolved SBLs and do not inhibit metallo-β-lactamases (MBLs). Boronic acids and boronate esters, especially cyclic ones, can potently inhibit both SBLs and MBLs. Vaborbactam, a monocyclic boronate, is approved for clinical use, but its β-lactamase coverage is limited. Bicyclic boronates rapidly react with SBLs and MBLs forming stable enzyme-inhibitor complexes that mimic the common anionic high-energy tetrahedral intermediates in SBL/MBL catalysis, as revealed by crystallography. The ability of boronic acids to 'morph' between sp2 and sp3 hybridisation states may help enable potent inhibition. There is limited structure-activity relationship information on the (bi)cyclic boronate inhibitors compared to β-lactams, hence scope for creativity towards new boron-based β-lactamase inhibitors/antibacterials.

Resistance to β-lactam antibacterials, importantly via production of β-lactamases, threatens their widespread use. Bicyclic boronates show promise as clinically useful, dual-action inhibitors of both serine- (SBL) and metallo- (MBL) β-lactamases. In combination with cefepime, the bicyclic boronate taniborbactam is in phase 3 clinical trials for treatment of complicated urinary tract infections. We report kinetic and crystallographic studies on the inhibition of AmpC, the class C β-lactamase from Escherichia coli, by bicyclic boronates, including taniborbactam, with different C-3 side chains. The combined studies reveal that an acylamino side chain is not essential for potent AmpC inhibition by active site binding bicyclic boronates. The tricyclic form of taniborbactam was observed bound to the surface of crystalline AmpC, but not at the active site, where the bicyclic form was observed. Structural comparisons reveal insights into why active site binding of a tricyclic form has been observed with the NDM-1 MBL, but not with other studied β-lactamases. Together with reported studies on the structural basis of inhibition of class A, B and D β-lactamases, our data support the proposal that bicyclic boronates are broad-spectrum β-lactamase inhibitors that work by mimicking a high energy ‘tetrahedral’ intermediate. These results suggest further SAR guided development could improve the breadth of clinically useful β-lactamase inhibition.