Cloning and sequence analysis of cDNA encoding rat carboxypeptidase D

Identification of mouse CPX-2, a novel member of the metallocarboxypeptidase gene family: cDNA cloning, mRNA distribution, and protein expression and characterization

A novel member of the metallocarboxypeptidase gene family was identified from its homology with carboxypeptidase E and has been designated CPX-2. The cDNA of 2500 nucleotides encodes a protein of 764 amino acids that contains an N-terminal signal peptide-like sequence, a 158-residue discoidin domain, and a 400-residue carboxypeptidase domain. The 400-residue metallocarboxypeptidase domain has 59% amino acid identity with a protein designated AEBP-1; 44% to 46% identity with carboxypeptidases E, N, and Z; and lower homology with other members of the metallocarboxypeptidase gene family. The discoidin domain of CPX-2 has 22% amino acid identity with the carbohydrate-binding domain of discoideum-I, 29% to 34% identity with the phospholipid-binding domain of human factors V and VIII, and 59% identity with the discoidin-like domain on AEBP-1. CPX-2 is missing several of the predicted active-site residues that are conserved in most other members of the metallocarboxypeptidase gene family and which are thought to be required for enzyme activity. Expression of CPX-2 using the baculovirus system produced several forms of protein, from 80 to 105 kDa, but no detectable activity toward a variety of carboxypeptidase substrates. A shorter 50-kDa form of CPX-2, which contains the carboxypeptidase domain but not the discoidin domain, was also inactive when expressed in the baculovirus system. CPX-2 is able to bind to Sepharose-Arg; this binding is blocked by 10 mM Arg. Northern blot analysis showed CPX-2 mRNA in mouse brain, liver, kidney, and lung. In situ hybridization analysis of brain revealed a broad distribution. Areas that are enriched in CPX-2 include the hippocampus, cerebral cortex, median eminence, and choroid plexus. Taken together, these data suggest a widespread function for CPX-2, possibly as a binding protein rather than an active carboxypeptidase.

Carboxypeptidase D (gp180), a Golgi-resident protein, functions in the attachment and entry of avian hepatitis B viruses

Carboxypeptidase D (gp180), one of many candidate receptors proposed for hepatitis B viruses (HBVs), was examined and found to be the actual cellular receptor for avian HBVs. This conclusion was based on the following observations: (i) gp180 was the only host protein that bound with high affinity to the pre-S ectodomain of the large duck hepatitis B virus (DHBV) envelope protein, which is known to be essential for virus infection; (ii) a pre-S subdomain which determines physical binding to gp180 was found to coincide with a domain functionally defined in infection competition experiments as a receptor binding domain; (iii) soluble gp180, lacking the membrane anchor, efficiently inhibited DHBV infection; (iv) efficient interspecies gp180-pre-S interaction was limited to the natural hosts of avian hepadnaviruses; and (v) expression of gp180 in a heterologous hepatoma cell line mediated cellular attachment and subsequent internalization of fluorescently labeled viral particles into vesicular structures. However, gp180 expression did not render transfected heterologous cells permissive for productive infection, suggesting that a species-specific coreceptor is required for fusion to complete viral entry. In contrast to the case for known virus receptors, gp180 was not detected on the hepatocyte cell surface but was found to be concentrated in the Golgi apparatus, from where it functions by cycling to and from the plasma membrane.

Altered processing of procholecystokinin in carboxypeptidase E-deficient fat mice: differential synthesis in neurons and endocrine cells

The fat mouse strain exhibits a late-onset obesity syndrome associated with a mutation in the gene encoding carboxypeptidase E (CPE). CPE plays a central role in the biosynthesis of many regulatory peptides. Therefore, we examined the processing of procholecystokinin (proCCK) in the brain (neurons) and small intestine (endocrine cells) of fat/fat mice. In the brain, bioactive CCK was markedly reduced (7.9+/-1.0 pmol/g in fat/fat mice vs. 82.5+/-11.2 pmol/g in controls), but the concentration of the CPE substrate, glycylarginine-extended CCK, was elevated 105-fold. In contrast, the concentration of bioactive CCK in intestinal endocrine cells was unaffected. Endocrine cell processing was, nevertheless, altered with a 33-fold increase in glycyl-arginine-extended CCK. Interestingly, although total proCCK products were normal in the brain they were elevated 3-fold in the intestine, indicating that biosynthesis is upregulated in endocrine cells but not neurons to compensate for the processing defect. These results demonstrate that the CPE mutation differentially affects CCK processing in these two cell types. Intestinal CCK synthesis more closely resembles progastrin processing, suggesting the presence of an endocrine-specific biosynthetic regulatory mechanism not present in neurons.

Regulation of endosome sorting by a specific PP2A isoform

The regulated sorting of proteins within the trans-Golgi network (TGN)/endosomal system is a key determinant of their biological activity in vivo. For example, the endoprotease furin activates of a wide range of proproteins in multiple compartments within the TGN/endosomal system. Phosphorylation of its cytosolic domain by casein kinase II (CKII) promotes the localization of furin to the TGN and early endosomes whereas dephosphorylation is required for efficient transport between these compartments (Jones, B.G., L. Thomas, S.S. Molloy, C.D. Thulin, M.D. Fry, K.A. Walsh, and G. Thomas. 1995. EMBO [Eur. Mol. Biol. Organ.] J. 14:5869-5883). Here we show that phosphorylated furin molecules internalized from the cell surface are retained in a local cycling loop between early endosomes and the plasma membrane. This cycling loop requires the phosphorylation state-dependent furin-sorting protein PACS-1, and mirrors the trafficking pathway described recently for the TGN localization of furin (Wan, L., S.S. Molloy, L. Thomas, G. Liu, Y. Xiang, S.L. Ryback, and G. Thomas. 1998. Cell. 94:205-216). We also demonstrate a novel role for protein phosphatase 2A (PP2A) in regulating protein localization in the TGN/endosomal system. Using baculovirus recombinants expressing individual PP2A subunits, we show that the dephosphorylation of furin in vitro requires heterotrimeric phosphatase containing B family regulatory subunits. The importance of this PP2A isoform in directing the routing of furin from early endosomes to the TGN was established using SV-40 small t antigen as a diagnostic tool in vivo. The role of both CKII and PP2A in controlling multiple sorting steps in the TGN/endosomal system indicates that the distribution of itinerant membrane proteins may be acutely regulated via signal transduction pathways.

PACS-1 defines a novel gene family of cytosolic sorting proteins required for trans-Golgi network localization

We report the role of one member of a novel gene family, PACS-1, in the localization of trans-Golgi network (TGN) membrane proteins. PACS-1 directs the TGN localization of furin by binding to the protease's phosphorylated cytosolic domain. Antisense studies show TGN localization of furin and mannose-6-phosphate receptor, but not TGN46, is strictly dependent on PACS-1. Analyses in vitro and in vivo show PACS-1 has properties of a coat protein and connects furin to components of the clathrin-sorting machinery. Cell-free assays indicate TGN localization of furin is directed by a PACS-1-mediated retrieval step. Together, these findings explain a mechanism by which membrane proteins in mammalian cells are localized to the TGN.

Chromosomal localization of the genes for human carboxypeptidase D (CPD) and the active 50-kilodalton subunit of human carboxypeptidase N (CPN1)

Human carboxypeptidase N is a 280-kDa tetrameric enzyme consisting of two 83-kDa regulatory subunits and two catalytic 50-kDa subunits. The 83-kDa subunit is a member of the leucine-rich repeat family of proteins and has been localized to chromosome 8p22-p23. The 50-kDa subunit is a member of the regulatory B-type carboxypeptidase family, which includes carboxypeptidases M, E/H, AEBP1, and a newly described member, carboxypeptidase D, which has three tandem active site domains. The human genes for carboxypeptidase D (HGMW-approved symbol CPD) and the 50-kDa subunit of carboxypeptidase N (HGMW-approved symbol CPN1) were localized to chromosomes 17 and 10, respectively, using the polymerase chain reaction with gene-specific primers and DNAs derived from somatic cell hybrids. The carboxypeptidase D gene was further localized to the centromeric region 17p11.1-q11.1/11.2 by use of a regional mapping panel derived from somatic cell hybrids containing different portions of chromosome 17.

gp180, a protein that binds duck hepatitis B virus particles, has metallocarboxypeptidase D-like enzymatic activity

Duck gp180 was previously identified by its ability to bind to the preS envelope protein of duck hepatitis B virus particles (Kuroki, K. , Cheung, R., Marion, P. L., and Ganem, D. (1994) J. Virol. 68, 2091-2096). Cloning and sequencing of gp180 cDNA revealed that it is a polyprotein with three carboxypeptidase-like domains (Kuroki, K., Eng, F., Ishikawa, T., Turck, C., Harada, F., and Ganem, D. (1995) J. Biol. Chem. 270, 15022-15028). To evaluate enzymatic properties of this protein, a soluble 170-kDa form of the protein (gp170) lacking the C-terminal transmembrane domain and cytoplasmic tail was expressed in a baculovirus system. The purified 170-kDa protein cleaved 5-dimethylaminonaphthalene-1-sulfonyl (dansyl)-Phe-Ala-Arg with a pH optimum of 5.5-6.5. With this substrate at pH 5.5, the 170-kDa protein displayed a Km of 12 microM and a Kcat of 57 s-1. Dansyl-Pro-Ala-Arg and dansyl-Phe-Phe-Arg were cleaved with Km values of 17 and 21 microM, and Kcat values of 57 and 17 s-1, respectively. Constructs containing only the first or second carboxypeptidase domains also showed enzymatic activity. The effects of inhibitors and ions on enzyme activity of gp170 were generally similar to the effects of these compounds on purified bovine carboxypeptidase D. To evaluate the regions within gp180 necessary for binding preS, a series of deletion mutants were expressed in the 293T human kidney cell line. Deletions of the first and second domains, leaving the third domain intact, eliminated carboxypeptidase activity but retained preS binding. Deletion of the third domain eliminated preS binding but not carboxypeptidase activity. These results indicate that the third domain is responsible for preS binding, and this binding does not require carboxypeptidase activity.

Intracellular trafficking of metallocarboxypeptidase D in AtT-20 cells: localization to the trans-Golgi network and recycling from the cell surface

Carboxypeptidase D (CPD) is a recently discovered membrane-bound metallocarboxypeptidase that has been proposed to be involved in the post-translational processing of peptides and proteins that transit the secretory pathway. In the present study, the intracellular distribution of CPD was examined in AtT-20 cells, a mouse anterior pituitary-derived corticotroph. Antisera to CPD stain the same intracellular structures as those labeled with furin and wheat germ agglutinin. This distribution is distinct from carboxypeptidase E, which is localized to the secretory vesicles in the cell processes. The perinuclear distribution of CPD is detected even when the AtT-20 cells are treated with brefeldin A for 1–30 minutes, suggesting that CPD is present in the trans-Golgi network (TGN). Although CPD is predominantly found in the TGN, an antiserum to the full length protein is internalized within 15–30 minutes of incubation at 37 degrees C. In contrast, an antiserum raised against the C-terminal region of CPD does not become internalized, suggesting that this domain is cytosolic. The antiserum to the full length CPD is internalized to a structure that co-stains with furin and wheat germ agglutinin, but is distinct from transferrin recycling endosomes. The internalization of CPD is not substantially affected by treatment of the AtT-20 cells with brefeldin A. These data are consistent with the cycling of CPD to the cell surface and back to the TGN. The TGN localization of CPD raises the possibility of a role for this enzyme in the processing of proteins that transit the secretory pathway.

Cloning, sequence analysis, and distribution of rat metallocarboxypeptidase Z

A cDNA encoding human carboxypeptidase Z (CPZ), a novel metallocarboxypeptidase, was recently cloned (Song and Fricker, J. Biol. Chem., 272, 1054, 1997). In the present study, a cDNA encoding the rat homolog of CPZ was identified. As with the human form, rat CPZ contains an N-terminal domain of 120 amino acids that has 20% to 30% amino acid identity with the "frizzled" domain found on proteins that interact with Wnt, a protein involved in tissue polarity in early embryogenesis. Sequence analysis showed rat and human CPZ to be highly conserved within the frizzled domain (77% amino acid identity), the carboxypeptidase domain (91%), and the C-terminal 28 residues (78%). The entire rat CPZ protein has high sequence similarity with human CPZ (81% amino acid identity), moderate sequence similarity to human carboxypeptidase N (45%), human carboxypeptidase E (41%), and human carboxypeptidase M (33%), and less sequence similarity with other metallocarboxypeptidases. Northern blot analysis showed rat CPZ mRNA to be abundant in the placenta, with low to moderate levels in the brain, lung, thymus, and kidney. The BRL3A rat liver cell line and the PC12 rat adrenal cell line express high levels of CPZ mRNA. In situ hybridization analysis indicated that CPZ is expressed only in specific cell types. For example, in the brain, CPZ mRNA is present in leptomeningeal cells, but not in the majority of other cell types. This distribution in leptomeningeal cells is shared by AEBP1, a recently reported member of the metallocarboxypeptidase gene family. However, the distribution of CPZ and AEBP1 differ in pituitary and thyroid. Taken together, these studies suggest that CPZ functions in a range of cell types.

Cellular carboxypeptidases

This article focuses on four human carboxypeptidases (CPs): two metallo-CPs and two serine CPs. The metallo-CPs are members of the so-called B-type regulatory CP family, as they cleave only the C-terminal basic amino acids Arg or Lys. The plasma membrane-bound CPM and the mainly, but not exclusively, intracellular CPD are surveyed from this group of enzymes. These enzymes can regulate peptide hormone activity at the cell surface and possibly intracellularly after receptor-mediated endocytosis and may also participate in peptide hormone processing. The serine CPs, as their name indicates, contain a serine residue in the active center essential for catalytic activity that reacts with organophosphorus inhibitors. Prolylcarboxypeptidase (PRCP) (angiotensinase C) and deamidase (cathepsin A, lysosomal protective protein) are discussed here. These two enzymes are highly concentrated in lysosomes; however, they may also be active extracellularly after their release from lysosomes in soluble form or in a plasma membrane-bound complex. Whereas deamidase cleaves a variety of peptides with C-terminal or penultimate hydrophobic residues (e.g. substance P, angiotensin I, bradykinin, endothelin, fMet-Leu-Phe). PRCP cleaves only peptides with a penultimate Pro residue (e.g. des-Arg9-bradykinin, angiotensin II). These enzymes may also be involved in terminating signal transduction by inactivating peptide ligands after receptor endocytosis.

Prodynorphin processing by proprotein convertase 2. Cleavage at single basic residues and enhanced processing in the presence of carboxypeptidase activity

Endoproteolytic processing of the 26-kDa protein precursor prodynorphin (proDyn) at paired and single basic residues is most likely carried out by the proprotein convertases (PCs); however, the role of PCs at single basic residues is unclear. In previous studies we showed that limited proDyn processing by PC1/PC3 at both paired and single basic residues resulted in the formation of 8- and 10-kDa intermediates. Because PC2 is colocalized with proDyn, we examined the potential role of this convertase in cleaving proDyn. PC2 cleaved proDyn to produce dynorphin (Dyn) A 1-17, Dyn B 1-13, and alpha-neo-endorphin, without a previous requirement for PC1/PC3. PC2 also cleaved at single basic residues, resulting in the formation of the C-peptide and Dyn A 1-8. Only PC2, but not furin or PC1/PC3, could cleave the Arg-Pro bond to yield Dyn 1-8. Structure-activity studies with Dyn A 1-17 showed that a P4 Arg residue is important for single basic cleavage by PC2 and that the P1' Pro residue impedes processing. Conversion of Dyn A 1-17 or Dyn B 1-13 into leucine-enkephalin (Leu-Enk) by PC2 was never observed; however, Dyn AB 1-32 cleavage yielded small amounts of Leu-Enk, suggesting that Leu-Enk can be generated from the proDyn precursor only through a specific pathway. Finally, PC2 cleavages at single and paired basic residues were enhanced when carried out in the presence of carboxypeptidase (CP) E. Enhancement was blocked by GEMSA, a specific inhibitor of CPE activity, and could be duplicated by other carboxypeptidases, including CPD, CPB, or CPM. Our data suggest that carboxypeptidase activity enhances PC2 processing by the elimination of product inhibition caused by basic residue-extended peptides.

Beta-cell lines derived from transgenic Cpe(fat)/Cpe(fat) mice are defective in carboxypeptidase E and proinsulin processing

A spontaneous point mutation in the coding region of the carboxypeptidase E (CPE) gene in Cpe(fat)/Cpe(fat) mice affects proinsulin processing. Cell lines derived from the pancreatic beta-cells of Cpe(fat)/Cpe(fat) mice were generated by crossing C57BLKS/J-Cpe(fat)/+ mice with NOD mice expressing the simian virus 40 large T oncogene under the control of the rat insulin II promoter. Two cell lines, designated NIT-2 and NIT-3, were cultured from adenomatous islets obtained from F2 littermates and were compared with the NIT-1 cell line previously developed from mice with wild-type CPE. Electron microscopy of the cultured NIT-2 and -3 cells showed increased numbers of enlarged and electron-lucent granules compared with NIT-1 cells. Pro-CPE, but not the mature form of CPE, is present in NIT-2 and -3 cells, and neither pro-CPE nor CPE are secreted into the medium. Immunocytochemistry shows the pro-CPE to be localized to an endoplasmic reticulum-like structure in NIT-3 cells. Proinsulin is less extensively processed in NIT-2 and -3 cells than in NIT-1 cells, indicating that the Cpe(fat) mutation affects both the endopeptidase and carboxypeptidase reactions. The secretion of insulin/proinsulin from NIT-2 and -3 cells is significantly elevated by secretagogues, indicating that CPE is not required for sorting proinsulin into the regulated pathway.