In all forms of life, DNA ligase enzymes are essential for joining strands of DNA when new DNA is being made or when damaged DNA is being repaired. These enzymes fall into two classes dependent on their requirement for either ATP or NAD+ as a co-factor. Most bacteriophage genomes sequenced to date do not encode a DNA ligase; these phages rely on the activity of the host cell ligase for lytic infection. The T5-like phages (Demerecviridae) in contrast encode a ligase that is essential for phage infection. This NAD+ -dependent ligase has a highly unusual structure: the enzyme is split into two separate polypeptides encoded by distinct non-overlapping ORFs. This structural organisation - which is unprecedented in DNA ligase biology - is an evolutionarily conserved feature of all T5-like phages, implying that having a split DNA ligase is important for phage fitness. We have initiated a study of the T5-like split DNA ligase, aiming to better understand its structure, its biological function, and the evolutionary rationale for splitting the enzyme into two parts. In particular, we aim to determine whether the existence of the split ligase could be related to the presence of single-strand DNA nicks in the packaged phage genome. We have determined the crystal structure of the split ligase encoded by Providencia rettgeri phage vB_PreS_PR1 (PR1) bound to nicked substrate DNA, offering insight into how the two subunits come together to form an active enzyme, and have embarked on a programme of genome engineering of phage T5 using CRISPR/Cas methodologies to ask whether the split is essential for ligase function. The results of these studies will be described.