TI - RESULTS Release of a 3-meT-induced replication block by treatment with AlkB proteins . AB - 3-meT lesions in DNA have previously been shown to block primer extension by DNA polymerases ( 25 ) , demonstrating that 3-meT is a replication blocking lesion . To test whether AlkB proteins can repair 3-meT lesions in DNA , we studied their ability to rescue a 3-meT -induced replication block . We performed T7 DNA polymerase mediated extension of a 32P-end-labelled 15mer primer on a 24mer oligonucleotide template containing a single 3-meT residue in position 19 from the 3' end ( Figure 2A ) . Primer extension yielded exclusively an 18mer extension product ( Figure 2B , lane 1 ) , indicating that the T7 DNA polymerase is not able to incorporate a nucleoside opposite the 3-meT residue . Extension on an otherwise identical template , where the 3-meT residue had been replaced by thymine , yielded a mixture of 24 and 25 nt products ( Figure 2D , lane 1 ) . ( The 25mer product results from the ability of T7 DNA polymerase to add an additional residue after completing template-directed synthesis ) . However , when the 3-meT containing template had been incubated with human or bacterial AlkB proteins as well as the cofactors Fe2+ and 2-oxoglutarate prior to primer extension , a partial rescue of the primer extension block was observed ( Figure 2B ) , i.e.an additional 25mer primer extension product was observed . However , even at the highest enzyme concentrations , we were not able to observe complete repair , as indicated by the presence of the 18mer extension product . Further increasing the enzyme concentrations did not increase the repair efficiency but rather caused adverse effects , such as SUBstrate degradation and inhibition of primer extension ( data not shown ) . Normalized for the amounts of enzyme used , the relative repair efficiency of the three enzymes can be summarized as follows : AlkB ~ hABH2 gt hABH3 (Figure 2C) . In contrast , when we previously introduced 1-meA and 3-meC residues by [3H] MNU treatment of the corresponding 3-meT-free oligonucleotide , the relative repair efficiencies were AlkB gt hABH2 ~ hABH3 ( 23 ) . To more directly compare the activity of these enzymes on 1-meA and 3-meC with that observed for 3-meT , the 3-meT-free oligonucleotide was treated with DMS to introduce 1-meA and 3-meC lesions , and similar primer extension experiments were performed . Clearly , the DMS treatment inhibited the primer extension , leading to a substantial reduction in the amount of the 24mer/25mer extension products (Figure 2D) . The pattern of observed stop sites corresponded reasonably well to the positions of A and C bases in the template , indicating that the extension block is caused by the introduction of 1-meA and 3-meC . As observed with the 3-meT containing oligonucleotide , incubation of the DMS-treated oligonucleotide with AlkB , hABH2 or hABH3 partially or completely reversed the primer extension block , thus demonstrating repair ( Figure 2E ) . The relative repair efficiencies of these three enzymes were similar to those previously obtained when using the [3H] MNU-treated SUBstrate ( 23 ) , i.e . AlkB gt hABH2 ~ hABH3 (Figure 2F) , but somewhat different from that obtained when using the 3-meT containing oligonucleotide ( AlkB ~ hABH2 gt hABH3 ) . In summary , these experiments indicate that 3-meT lesions can be reversed by AlkB , hABH2 and hABH3 , but that the relative activity of the enzymes on this lesion is different from that observed with 1-meA and 3-meC .