While apparent that His Artemis fractionated over a HAP colu
While apparent that [His]6-Artemis fractionated over a HAP column is devoid of 5′–3′ exonuclease activity but still retains DNA-PK dependent hairpin-opening activity, we sought to further assess DNA-PK dependent Artemis overhang cleavage activity to ensure all the in vivo, intrinsic enzymatic activities are retained in this purified preparation of protein. We therefore prepared a 5′-radiolabeled DNA substrate with a 3′ single-strand overhang . This substrate was efficiently cleaved in close proximity to the single-strand/double-strand junction by both the nickel and HAP flow-through pools of protein when incubated with DNA-PK and ATP (Fig. 6B, lanes 4, 5 and 6 and lanes 7, 8 and 9, respectively). As expected, the endonucleolytic product is completely dependent on DNA-PK and ATP. Although the presence of duplex DNA has been reported to stimulate DNA-PK dependent Artemis endonuclease activity , the addition of a 30-mer double-stranded cold DNA substrate did not increase endonuclease cleavage on the hairpin radiolabeled substrate (Fig. 6A, lane 7) or the 3′ overhang substrate (Fig. 6B, lanes 5 and 9). Again, the 5′ NMP is apparent in reactions performed with the nickel pool of protein but is reduced to background levels in reactions containing the HAP flow-through fraction. Finally, a DNA substrate with a 5′ single-strand overhang and 3′ [α-32P] dCMP label was used to assay endonuclease activity on a 5′ overhang . DNA-PK dependent Artemis-catalyzed endonuclease activity on this substrate is anticipated to result in a product of approximately 26 GSK-923295 following cleavage at the single-strand/double-strand junction. In the presence of DNA-PK and ATP, both the nickel pool of protein and the HAP flow-through endonucleolytically cleaved the 3′ overhang to generate a 26 nucleotide product (Fig. 6C, lanes 5 and 7). An exonuclease-like product appears at the bottom of the gel, but as it only appears in lanes that have DNA-PK, including but not limited to the lane that contains DNA-PK alone (Fig. 6C, lane 4), it is attributed to a contaminating 3′ exonuclease in our preparation of DNA-PK from Hela cells. The data presented in Fig. 5, Fig. 6 demonstrate that both pools of protein, from the nickel and the HAP FT column, maintain hairpin-opening and endonuclease activity that has previously been reported to be intrinsic to Artemis. Importantly, the HAP flow-through pool of protein, containing [His]6-Artemis, no longer exhibits single-strand exonuclease activity but retains DNA-PK dependent endonucleolytic activities. This data demonstrates that the exonuclease activity of Artemis previously reported is not an intrinsic component of the Artemis polypeptide. It is possible that the exonuclease activity is another enzyme that is associated with Artemis and plays a physiological role with Artemis in the cell, but it is equally possible that the exonuclease activity is simply a contaminating protein that is difficult to separate from the pool of Artemis protein during a purification procedure. To finally ensure that there are no other alternatives to the separation of exonuclease activity from Artemis we quantified the endonuclease and exonuclease activities throughout the purification and the results are presented in Table 1. These results clearly demonstrate that following fractionation on a HAP column, 100% of the endonuclease activity is recovered in the HAP FT resulting in nearly a 30-fold purification while only 2.5% of the exonuclease activity is retained in that fraction with no increase in specific activity despite the loss of 96% of the total protein. Consistent with this data, analysis of the Artemis polypeptide as determined by western blotting in conjunction with analysis of exonuclease activity was performed on the HAP FT and elution pools of protein from a separate preparation. The percent of total antibody reactivity or exonuclease activity resolved in the two pools of protein was determined and the results demonstrate that greater than 90% of Artemis protein loaded onto the HAP column was recovered in the HAP flow-through material, while less than 10% was recovered in the gradient elution pool while greater than 90% of the exonuclease was identified in the elution pool and less than 10% in the FT pool of protein (Supplemental Fig. 3). These results indicate that the exonuclease found to co-purify with [His]6-Artemis under certain conditions can be separated away from [His]6-Artemis under other conditions (as described above), and therefore is probably a prominent exonuclease that has a similar affinity as Artemis for certain column conditions, but is not intrinsic to the Artemis polypeptide.