Exploration of subsite binding specificity of human cathepsin D through kinetics and rule-based molecular modeling.

The family of aspartic proteinases includes several human enzymes that may play roles in both physiological and pathophysiological processes. The human lysosomal aspartic proteinase cathepsin D is thought to function in the normal degradation of intracellular and endocytosed proteins but has also emerged as a prognostic indicator of breast tumor invasiveness. Presented here are results from a continuing effort to elucidate the factors that contribute to specificity of ligand binding at individual subsites within the cathepsin D active site. The synthetic peptide Lys-Pro-Ile-Glu-Phe*Nph-Arg-Leu has proven to be an excellent chromogenic substrate for cathepsin D yielding a value of kcat/Km = 0.92 x 10(-6) s-1 M-1 for enzyme isolated from human placenta. In contrast, the peptide Lys-Pro-Ala-Lys-Phe*Nph-Arg-Leu and all derivatives with Ala-Lys in the P3-P2 positions are either not cleaved at all or cleaved with extremely poor efficiency. To explore the binding requirements of the S3 and S2 subsites of cathepsin D, a series of synthetic peptides was prepared with systematic replacements at the P2 position fixing either Ile or Ala in P3. Kinetic parameters were determined using both human placenta cathepsin D and recombinant human fibroblast cathepsin D expressed in Escherichia coli. A rule-based structural model of human cathepsin D, constructed on the basis of known three-dimensional structures of other aspartic proteinases, was utilized in an effort to rationalize the observed substrate selectivity.

Structural requirements of procathepsin D activation and maturation.

Cathepsin D biosynthesis involves several proteolytic events; however, the enzymology and sequence of these events are not known. Procathepsin D undergoes a pH-dependent, intramolecular proteolysis in vitro which removes 26 residues yielding an active form that is intermediate in size between procathepsin D and single-chain cathepsin D. This form, designated pseudocathepsin D, has not been shown to be an in vivo intermediate. The N-terminal sequence of the light chain of cathepsin D, isolated from human placenta, showed that 42 residues were removed as compared with 44 residues predicted by comparison with porcine cathepsin D. Site-directed mutations were generated at both processing sites within the propeptide of procathepsin D. Mutation at the autocatalytic site prevented in vitro autoactivation, but, after transfection of mouse Ltk- cells, the mutant procathepsin D was transported to the lysosome and processed normally to the mature enzyme despite its inability to autoactivate in vitro. Mutation at the mature N terminus of cathepsin D prevented in vivo formation of the single-chain form of the enzyme; however, the protein was still processed to the two-chain form of human cathepsin D. This change at the mature N terminus did not prevent in vitro autoactivation. Procathepsin D with mutations at both cleavage sites was processed to the two-chain form despite the inability to undergo removal of the propeptide. These results indicated that stepwise autoactivation and propeptide removal were not necessary for later processing or delivery of human cathepsin D to the lysosome. The results also suggested that pseudocathepsin D was not a normal intermediate of procathepsin D processing in vivo.