Dipeptidyl peptidase (DPP) IV has roles in T-cell costimulation, chemokine biology, type-II diabetes and tumor biology. Fibroblast activation protein (FAP) has been implicated in tumor growth and cirrhosis. Here we describe DPP8, a novel human postproline dipeptidyl aminopeptidase that is homologous to DPPIV and FAP. Northern-blot hybridization showed that the tissue expression of DPP8 mRNA is ubiquitous, similar to that of DPPIV. The DPP8 gene was localized to chromosome 15q22, distinct from a closely related gene at 19p13.3 which we named DPP9. The full-length DPP8 cDNA codes for an 882-amino-acid protein that has about 27% identity and 51% similarity to DPPIV and FAP, but no transmembrane domain and no N-linked or O-linked glycosylation. Western blots and confocal microscopy of transfected COS-7 cells showed DPP8 to be a 100-kDa monomeric protein expressed in the cytoplasm. Purified recombinant DPP8 hydrolyzed the DPPIV substrates Ala-Pro, Arg-Pro and Gly-Pro. Thus recombinant DPP8 shares a postproline dipeptidyl aminopeptidase activity with DPPIV and FAP. DPP8 enzyme activity had a neutral pH optimum consistent with it being nonlysosomal. The similarities between DPP8 and DPPIV in tissue expression pattern and substrates suggests a potential role for DPP8 in T-cell activation and immune function.

Dipeptidyl peptidase (DPP) IV has roles in T-cell costimulation, chemokine biology, type-II diabetes and tumor biology. Fibroblast activation protein (FAP) has been implicated in tumor growth and cirrhosis. Here we describe DPP8, a novel human postproline dipeptidyl aminopeptidase that is homologous to DPPIV and FAP. Northern-blot hybridization showed that the tissue expression of DPP8 mRNA is ubiquitous, similar to that of DPPIV. The DPP8 gene was localized to chromosome 15q22, distinct from a closely related gene at 19p13.3 which we named DPP9. The full-length DPP8 cDNA codes for an 882-amino-acid protein that has about 27% identity and 51% similarity to DPPIV and FAP, but no transmembrane domain and no N-linked or O-linked glycosylation. Western blots and confocal microscopy of transfected COS-7 cells showed DPP8 to be a 100-kDa monomeric protein expressed in the cytoplasm. Purified recombinant DPP8 hydrolyzed the DPPIV substrates Ala-Pro, Arg-Pro and Gly-Pro. Thus recombinant DPP8 shares a postproline dipeptidyl aminopeptidase activity with DPPIV and FAP. DPP8 enzyme activity had a neutral pH optimum consistent with it being nonlysosomal. The similarities between DPP8 and DPPIV in tissue expression pattern and substrates suggests a potential role for DPP8 in T-cell activation and immune function.

Two dipeptidyl peptidase IV (DPPIV, DPP4)-related proteins, DPP8 and DPP9, have been identified recently [Abbott, Yu, Woollatt, Sutherland, McCaughan, and Gorrell (2000) Eur. J. Biochem. 267, 6140-6150; Olsen and Wagtmann (2002) Gene 299, 185-193; Qi, Akinsanya, Riviere, and Junien (2002) Patent application WO0231134]. In the present study, we describe the cloning of DPP10, a novel 796-amino-acid protein, with significant sequence identity to DPP4 (32%) and DPP6 (51%) respectively. We propose that DPP10 is a new member of the S9B serine proteases subfamily. The DPP10 gene is located on the long arm of chromosome 2 (2q12.3-2q14.2), close to the DPP4 (2q24.3) and FAP (2q23) genes. The active-site serine residue is replaced by a glycine residue in DPP10, resulting in the loss of DPP activity. The serine residue is also replaced in DPP6, which lacks peptidase activity. DPP8 and DPP9 share an identical active site with DPP4 (Gly-Trp-Ser-Tyr-Gly). In contrast with the previous results suggesting that DPP9 is inactive, we show that DPP9 is a DPP, hydrolysing Ala-Pro-(7-amino-4-methyl-coumarin) with similar pH-specificity and protease-inhibitor-sensitivity to those of DPP4 and DPP8. Northern-blot analysis shows that whereas DPP8 and DPP9 are widely expressed, DPP10 is expressed mainly in the brain and pancreas. DPP6, which has the highest amino acid identity with DPP10, has been shown previously [Nadal, Ozaita, Amarillo, de Miera, Ma, Mo, Goldberg, Misumi, Ikehara, Neubert et al. (2003) Neuron 37, 449-461] to associate with A-type K(+) channel subunits, modulating their transport and function in somatodendritic compartments of neurons. It is possible that DPP10 is involved in similar functions in the brain. Elucidation of the physiological or pathophysiological role of DPP8, DPP9 and DPP10 and characterization of their structure-function relationships will add impetus to the development of inhibitor molecules for pharmacological or therapeutic use.

Dipeptidyl peptidase IV (DP-IV/CD26), fibroblast activation protein (FAP), DP-like 1 (DPL1), DP8, DP9, and DPL2 comprise the CD26 gene family. CD26/DP-IV has roles in liver disease, T cell costimulation, chemokine biology, type II diabetes, and tumor biology. DPIV substrates include the glucagonlike peptides, neuropeptide Y, and the chemokines CCL3, CCL5, CCL11, CCL22, and CXCL12. We have proposed that the extracellular region of CD26 is analogous to prolyl oligopeptidase in consisting of an alpha/beta hydrolase domain contributed by both N- and C-terminal portions of the polypeptide and a seven-blade beta-propeller domain. Replacing the C-terminal portion of the predicted alpha/beta hydrolase domain of CD26 (residues 501-766) with the homologous portion of DP8 or DP9 produced intact proteins. However, these chimeric proteins lacked dimerization and peptidase activity, suggesting that CD26 dimerization requires the C-terminal portion of the alpha/beta hydrolase domain. Deleting some N-terminal residues of the alpha/beta hydrolase domain of CD26 ablated peptidase activity and greatly diminished cell surface expression. Together with previous data that CD26 peptidase activity requires the C-terminal 20 residues, this suggests that peptidase activity requires the entire alpha/beta hydrolase domain. The catalytic triad of DP8 was shown to be Ser(739)-Asp (817)-His(849). Glu(259) of DP8, a residue distant from the catalytic triad yet greatly conserved in the CD26 gene family, was shown to be required for peptidase activity. These data concord with our predicted CD26 structure, indicate that biosynthesis of a functional fragment of CD26 is difficult, and confirm the functional homology of DP8 with CD26.

DPP8 is a prolyl dipeptidase homologous to DPP-IV, which is a drug target for Type II diabetes. The biological function of DPP8 is not known. To identify potent and selective chemical compounds against DPP8, we have synthesized a series of isoquinoline and isoindoline derivatives and have tested their inhibitory activity against DPP8, DPP-IV and DPP-II. Isoindoline derivatives were found to be more potent DPP8 inhibitors than isoquinoline derivatives. Isoindoline with a 1-(4,4'-difluor-benzhydryl)-piperazine group at the P2 site was observed to be a very potent DPP8 inhibitor, having an IC(50) value of 14nM with at least a 2500-fold selectivity over either DPP-IV or DPP-II. From SAR results, we speculate that the S1 site of DPP8 may be larger than that of DPP-IV, which would allow the accommodation of larger C-terminal residues, such as isoquinoline or isoindoline.

Dipeptidyl peptidase (DPP) IV has roles in T-cell costimulation, chemokine biology, type-II diabetes and tumor biology. Fibroblast activation protein (FAP) has been implicated in tumor growth and cirrhosis. Here we describe DPP8, a novel human postproline dipeptidyl aminopeptidase that is homologous to DPPIV and FAP. Northern-blot hybridization showed that the tissue expression of DPP8 mRNA is ubiquitous, similar to that of DPPIV. The DPP8 gene was localized to chromosome 15q22, distinct from a closely related gene at 19p13.3 which we named DPP9. The full-length DPP8 cDNA codes for an 882-amino-acid protein that has about 27% identity and 51% similarity to DPPIV and FAP, but no transmembrane domain and no N-linked or O-linked glycosylation. Western blots and confocal microscopy of transfected COS-7 cells showed DPP8 to be a 100-kDa monomeric protein expressed in the cytoplasm. Purified recombinant DPP8 hydrolyzed the DPPIV substrates Ala-Pro, Arg-Pro and Gly-Pro. Thus recombinant DPP8 shares a postproline dipeptidyl aminopeptidase activity with DPPIV and FAP. DPP8 enzyme activity had a neutral pH optimum consistent with it being nonlysosomal. The similarities between DPP8 and DPPIV in tissue expression pattern and substrates suggests a potential role for DPP8 in T-cell activation and immune function.

Two dipeptidyl peptidase IV (DPPIV, DPP4)-related proteins, DPP8 and DPP9, have been identified recently [Abbott, Yu, Woollatt, Sutherland, McCaughan, and Gorrell (2000) Eur. J. Biochem. 267, 6140-6150; Olsen and Wagtmann (2002) Gene 299, 185-193; Qi, Akinsanya, Riviere, and Junien (2002) Patent application WO0231134]. In the present study, we describe the cloning of DPP10, a novel 796-amino-acid protein, with significant sequence identity to DPP4 (32%) and DPP6 (51%) respectively. We propose that DPP10 is a new member of the S9B serine proteases subfamily. The DPP10 gene is located on the long arm of chromosome 2 (2q12.3-2q14.2), close to the DPP4 (2q24.3) and FAP (2q23) genes. The active-site serine residue is replaced by a glycine residue in DPP10, resulting in the loss of DPP activity. The serine residue is also replaced in DPP6, which lacks peptidase activity. DPP8 and DPP9 share an identical active site with DPP4 (Gly-Trp-Ser-Tyr-Gly). In contrast with the previous results suggesting that DPP9 is inactive, we show that DPP9 is a DPP, hydrolysing Ala-Pro-(7-amino-4-methyl-coumarin) with similar pH-specificity and protease-inhibitor-sensitivity to those of DPP4 and DPP8. Northern-blot analysis shows that whereas DPP8 and DPP9 are widely expressed, DPP10 is expressed mainly in the brain and pancreas. DPP6, which has the highest amino acid identity with DPP10, has been shown previously [Nadal, Ozaita, Amarillo, de Miera, Ma, Mo, Goldberg, Misumi, Ikehara, Neubert et al. (2003) Neuron 37, 449-461] to associate with A-type K(+) channel subunits, modulating their transport and function in somatodendritic compartments of neurons. It is possible that DPP10 is involved in similar functions in the brain. Elucidation of the physiological or pathophysiological role of DPP8, DPP9 and DPP10 and characterization of their structure-function relationships will add impetus to the development of inhibitor molecules for pharmacological or therapeutic use.