Understanding the Structural and Dynamic Foundations of β-Lactam Resistance in PBP2a of methicillin-resistant Staphylococcus aureus

Antibiotic resistance is a worldwide calamity, set to become the greatest single cause of human mortality by 2050. β-lactams account for approximately 60% of the total antibiotics used in the clinic however, their overuse has seen their efficacy diminished by rapidly rising resistance rates. Methicillin-resistant Staphylococcus aureus (MRSA) remains a major global health threat, owing to its expression of penicillin binding protein 2a (PBP2a) – an enzyme that sustains cell wall synthesis despite the presence of β-lactam antibiotics. PBP2a
achieves this via an evolved allosteric domain (AD) regulates a 60 Å distal transpeptidase (TP) containing the active site, wherein AD activation increases the volume of the TP domain which has insufficient occupancy in the apo state. Critically, PBP2a’s canonical substrate peptidoglycan (PG) but not β-lactams activate the AD, thus enabling the selective of inhibitory agents whilst sustaining the crosslinking of PG necessary for cell survival. Only two β-lactams retain inhibitory efficacy against PBP2a – the fifth generation cephalosporins ceftaroline (CFT) and ceftobiprole (CFB). However, while substantial data evidence has linked CFT’s inhibitory capacity to AD activation (as in PG), less is known about the mechanistic details of CFB, where it remains unclear if allostery plays a role and to what extent. The structural and dynamic mechanisms by which PBP2a resists inhibition, and how these may be overcome, remain incompletely understood. This thesis integrates structural biology and NMR spectroscopy to investigate binding of antibiotics to PBP2a and how PBP2a’s conformational dynamics govern antibiotic resistance and inhibition. It also investigates whether rationally de novo designed mini-proteins can resensitise PBP2a to β-lactam antibiotics as a potential route to antibiotic adjuvant development. A selective methyl-labelling strategy was used to enable residue-level characterisation of PBP2a – previously intractable to conventional NMR assignment strategies due to its size and intrinsic flexibility. NMR relaxation dispersion analyses suggest that the apo enzyme samples a broad dynamic landscape, with coupled motions between its allosteric and transpeptidase domains that regulate active-site accessibility. Binding studies showed that ceftobiprole inhibits PBP2a via a direct, non-allosteric mechanism, engaging the transpeptidase domain and dampening global motions while preserving local flexibility at catalytic loops. This contrasts with ceftaroline, which requires allosteric activation, consistent with the observation that ceftaroline-resistant MRSA isolates carry mutations in both domains, whereas ceftobiprole resistance arises almost exclusively from substitutions in the catalytic region. Building on these mechanistic insights, de novo designed mini-protein binders were created using deep-learning-based protein design tools. One binder was experimentally validated by chemical shift perturbations to engage PBP2a across both domains and – critically – restore susceptibility of MRSA to oxacillin at 8 µg/mL, showing that such binders are a potential source of antibiotic adjuvant to circumvent this resistance mechanism. Together, these findings show that the structurally related ceftaroline and ceftobiprole inhibit PBP2a by different mechanisms and demonstrate that
resistance can be functionally reversed by rationally designed protein ligands, opening new avenues for re-sensitising MRSA to existing β-lactam antibiotics and redefining strategies for targeting antibiotic-resistant pathogens.