Antibiotics can damage the gut microbiome leading to opportunistic infections and the emergence of antibiotic resistance. Microbiome protection via antibiotic inactivation in the gastrointestinal (GI) tract represents a strategy to limit antibiotic exposure of the colonic microbiota. Proof of concept for this approach was achieved with an orally-administered beta-lactamase enzyme, SYN-004 (ribaxamase), that was demonstrated to degrade ceftriaxone excreted into the GI tract and protect the gut microbiome from antibiotic-mediated dysbiosis. Ribaxamase efficiently degrades penicillin and cephalosporin beta-lactam antibiotics, but is not active against carbapenems. To expand this microbiome protection strategy to include all classes of beta-lactams, three distinct carbapenemases were evaluated for manufacturability, antibiotic degradation spectrum, and stability in human intestinal fluid. E. coli production strains were generated for P2A, a novel metallo-enzyme isolated from B. cereus, New Delhi metallo-beta-lactamase (NDM), and Klebsiella pneumoniae carbapenemase (KPC). While all three enzymes effectively inactivated a broad range of antibiotics, including penicillins, most cephalosporins, and carbapenems in vitro, only P2A retained biological activity when incubated with human chyme. As functional stability in the intestinal tract is a key requirement for an orally-delivered enzyme, P2A was chosen as a potential clinical candidate. An enteric formulation of P2A was developed, called SYN-006, that was inert under high acid conditions, with enzyme dissolution occurring at pH > 5.5. SYN-006 has the potential to expand microbiome protection via antibiotic inactivation to include all classes of beta-lactam antibiotics.