DNA polymerase I has four domains. Due to its appearance as a right-hand, the different domains are referred to as the different anatomical parts of the hand (fingers, thumb, palm, …).
The palm domain binds to the DNA. On DNA binding, a conformational change occurs in the protein, causing the fingers and palm to curl in. This activates the catalytic activity of the polymerase, allowing nucleotides to be added to the daughter strand.
Magnesium ions help orientate the DNA correctly, ensuring the incoming nucleotides are in the correct position for efficient polymerisation.
A fourth domain, a 3’ to 5’ exonuclease, is also found. This is towards the ‘bottom’ of the polymerase, a distance of some 70 angstroms, allowing error correction where the incorrect base is incorporated. Through a mechanism that isn’t fully understood, the DNA strand is unwound, and the daughter strand moved to the exo domain. The erroneous nucleotide is removed, and the strand returned to the palm-fingers-thumb domain. This allows obvious errors to be corrected, with the normal DNA repair pathways used for more complicated errors, such as thymine dimers.