This PhD thesis describes research into the structure and function of lysostaphin (EC 3.4.24.75), a glycylglycine endopeptidase secreted by Staphylococcus simulans biovar staphylolyticus ATCC 1362. Lysostaphin is a member of the M23/M37 zinc metalloprotease family and is a pre-pro-enzyme. The mature form (after removal of the pro-region) contains two distinct domains, the C-terminal cell wall targeting domain of lysostaphin (termed LssTdom in the thesis) facilitates binding to Staphylococcus aureus cells. The endopeptidase domain (termed LssEdom in the thesis) cleaves the pentaglycine crosslinks in the peptidoglycan resulting in cell death through cell rupture of S. aureus. Lysostaphin is a potential therapeutic antibiotic for Methicillin-resistant Staphylococcus aureus (MRSA) for which new antibacterials are required owing to the widespread occurrence of multi-drug resistant strains. To date, the structural requirements for enzymatic activity and the target in the cell wall for lysostaphin have not been fully elucidated. Thus a structure might enable rationally guided design of lysostaphin variants for the generation of new enzymes which would also cleave the non-canonical crosslinks and would thus overcome bacterial resistance. Thus, one approach was to obtain the lysostaphin structure using homology modelling. Lysostaphin shares significant homology with the ALE-1 (83 % identity) and LytM (48 % identity) bacteriocins, and modelling the structure of lysostaphin was achieved using the recently released structures of LytM and ALE-1 derived from X-ray crystallography as templates. In addition, we report the successful production of active recombinant lysostaphin (as well as the endopeptidase and the targeting domains) and initial characterisation of their secondary structure by Fourier-transform infrared spectroscopy. Initially, advanced spectroscopic techniques including X-ray and NMR methods were investigated for molecular interaction studies between Lss and its putative ligands. Significant problems were encountered with these methods and mass spectrometry studies proved more amenable. Lysostaphin targeting domain-ligand complexes have been identified, along with their stoichiometry. The strength of the protein-ligand interactions has also been quantified. Lysostaphin was shown to bind in vitro Gly5 (mimicking the pentaglycine cross-bridge) and Lys-D-Ala-D-Ala (mimicking the stem peptide) with low affinity, but not NAM-L-Ala-D-iGln-Lys. It was also shown that lysostaphin targeting domain affinity for Gly5 was significantly reduced by addition of Gly-Gly-Ser-Gly-Ser (found in the host bacteria resistant to lysostaphin action – S. simulans) in solution. From these studies it could be concluded that resistance due to the incorporation of serine residues in the crossbridge were a result of the endopeptidase domain being unable to cleave this sequence and not due to the targeting domain being unable to bind it.