Cellular DNA is spontaneously and continuously damaged by environmental and internal factors such as X-rays, UV light and agents such as the antitumor drugs bleomycin and neocarzinostatin or those that generate oxygen radicals. Apurinic/apyrimidinic (AP) sites form both spontaneously and as highly cytotoxic intermediates in the removal of the damaged base by the base excision repair (BER) pathway. DNA repair at the AP sites is initiated by specific endonuclease cleavage of the phosphodiester backbone. Such endonucleases are also generally capable of removing blocking groups from the 3'terminus of DNA strand breaks.

AP endonucleases can be classified into two families on the basis of sequence similarity and structure (cf. family 2 <PDOC00599>). What we call family 1 groups the enzymes listed below [1].

• Escherichia coli exonuclease III (gene xthA) (EC 3.1.11.2).
• Streptococcus pneumoniae and Bacillus subtilis exonuclease A (gene exoA) (EC=3.1.11.2).
• Mammalian AP endonuclease 1 (AP1) (EC 4.2.99.18).
• Drosophila recombination repair protein 1 (gene Rrp1) (EC=4.2.99.18).
• Arabidopsis thaliana apurinic endonuclease-redox protein (gene arp) (EC=4.2.99.18).
• Dictyostelium DNA-(apurinic or apyrimidinic site) lyase (gene apeA) (EC=4.2.99.18).

Except for Rrp1 and arp, these enzymes are proteins of about 300 amino-acid residues. Rrp1 and arp both contain additional and unrelated sequences in their N-terminal section (about 400 residues for Rrp1 and 270 for arp).

The structures of bacterial exonuclease III and mammalian AP endonuclease 1 show an α/β-sandwich structure (see <PDB:1HD7; A>) with a fold similar to that of DNase I (see <PDOC00711>). One or two divalent metal ions such as magnesium or manganese can bind in the active site [2].

We developed three signature patterns and a profile for this family of enzymes. The first pattern contains a glutamate which has been shown [3], in the Escherichia coli enzyme to bind a divalent metal ion such as magnesium or manganese. The patterns are based on the most conserved regions [4]. We also developed a profile that spans the entire AP endonucleases family 1 structure.