Does human attractin have DP4 activity?

Mutations in the mouse ATRN gene, which encodes attractin, offer links between this protein and pigmentation, metabolism, immune status and neurodegeneration. However, the mechanisms of attractin action are not understood. The protein was first identified in humans in a circulating form in serum. A protease activity was postulated similar to the membrane-bound ectoenzyme DP4/CD26. In the last decade, both DP4/CD26 and attractin were controversially described to be the major source of human serum DP4 activity. We purified attractin from human plasma, and found that the DP4-like activity of the preparation shows nearly identical kinetic properties to that of recombinant human DP4. In contrast, the native electrophoretic behavior of this activity is clearly different from human and porcine DP4, but co-migrates with the protein band identified as attractin by Western blotting and N-terminal sequencing. Nevertheless, a DP4 impurity could be demonstrated in purified plasma attractin and the activity could be removed by ADA affinity chromatography, resulting in a homogenous attractin preparation without DP4 activity. These results are substantiated by expression of different attractin isoforms, in which no DP4 activity was found either. This indicates that the multidomain protein attractin acts as a receptor or adhesion protein rather than a protease.