Hyperproinsulinaemia in obese fat/fat mice associated with a carboxypeptidase E mutation which reduces enzyme activity.

Mice homozygous for the fat mutation develop obesity and hyperglycaemia that can be suppressed by treatment with exogenous insulin. The fat mutation maps to mouse chromosome 8, very close to the gene for carboxypeptidase E (Cpe), which encodes an enzyme (CPE) that processes prohormone intermediates such as proinsulin. We now demonstrate a defect in proinsulin processing associated with the virtual absence of CPE activity in extracts of fat/fat pancreatic islets and pituitaries. A single Ser202Pro mutation distinguishes the mutant Cpe allele, and abolishes enzymatic activity in vitro. Thus, the fat mutation represents the first demonstration of an obesity-diabetes syndrome elicited by a genetic defect in a prohormone processing pathway.

Peptides, enzymes and obesity: new insights from a 'dead' enzyme.

The identification of the fat mutation, which causes obesity in mice, as a defect in carboxypeptidase E (CPE) has raised more questions than answers. CPE is required for the processing of numerous neuroendocrine peptides and a mutation that inactivates CPE was predicted to be lethal. However, Cpe(fat) mutated mice live and become obese. So, why are mice with the Cpe(fat) mutation viable, and why does obesity develop as a consequence of the pleiotropic effects of this mutant allele? Recently, several new members of the carboxypeptidase family have been discovered, of which at least one, CPD, can partially compensate by contributing to neuroendocrine peptide processing. Obesity due to the Cpe(fat) mutation is not caused by increased food consumption but, rather, is a result of defective nutrient partitioning, the exact mechanism of which remains to be elucidated.

Sequences within the cytoplasmic domain of gp180/carboxypeptidase D mediate localization to the trans-Golgi network.

Gp180, a duck protein that was proposed to be a cell surface receptor for duck hepatitis B virus, is the homolog of metallocarboxypeptidase D, a mammalian protein thought to function in the trans-Golgi network (TGN) in the processing of proteins that transit the secretory pathway. Both gp180 and mammalian metallocarboxypeptidase D are type I integral membrane proteins that contain a 58-residue cytosolic C-terminal tail that is highly conserved between duck and rat. To investigate the regions of the gp180 tail involved with TGN retention and intracellular trafficking, gp180 and various deletion and point mutations were expressed in the AtT-20 mouse pituitary corticotroph cell line. Full length gp180 is enriched in the TGN and also cycles to the cell surface. Truncation of the C-terminal 56 residues of the cytosolic tail eliminates the enrichment in the TGN and the retrieval from the cell surface. Truncation of 12-43 residues of the tail reduced retention in the TGN and greatly accelerated the turnover of the protein. In contrast, deletion of the C-terminal 45 residues, which truncates a potential YxxL-like sequence (FxxL), reduced the protein turnover and caused accumulation of the protein on the cell surface. A point mutation of the FxxL sequence to AxxL slowed internalization, showing that this element is important for retrieval from the cell surface. Mutation of a pair of casein kinase II sites within an acidic cluster showed that they are also important for trafficking. The present study demonstrates that multiple sequence elements within the cytoplasmic tail of gp180 participate in TGN localization.

Identification and characterization of proSAAS, a granin-like neuroendocrine peptide precursor that inhibits prohormone processing.

Five novel peptides were identified in the brains of mice lacking active carboxypeptidase E, a neuropeptide-processing enzyme. These peptides are produced from a single precursor, termed proSAAS, which is present in human, mouse, and rat. ProSAAS mRNA is expressed primarily in brain and other neuroendocrine tissues (pituitary, adrenal, pancreas); within brain, the mRNA is broadly distributed among neurons. When expressed in AtT-20 cells, proSAAS is secreted via the regulated pathway and is also processed at paired-basic cleavage sites into smaller peptides. Overexpression of proSAAS in the AtT-20 cells substantially reduces the rate of processing of the endogenous prohormone proopiomelanocortin. Purified proSAAS inhibits prohormone convertase 1 activity with an IC(50) of 590 nM but does not inhibit prohormone convertase 2. Taken together, proSAAS may represent an endogenous inhibitor of prohormone convertase 1.

Metallocarboxypeptidase Z is dynamically expressed in mouse development.

Metallocarboxypeptidase Z (CPZ), a new member of the regulatory metallocarboxypeptidases, contains a 120-residue cysteine-rich region that has 20-35% amino acid sequence identity to Drosophila and mammalian frizzled proteins. In order to gain insights into the function of CPZ, we have examined the distribution of the protein by immunohistochemistry throughout mouse development. The expression of CPZ peaks at E9-E12, decreases in late gestation and falls further in adult tissues. CPZ expression in amnion cells, cochlear epithelial cells and surrounding mesenchyme, ventricular lining cells in the brain and cartilagenous condensations and surrounding connective tissue in ribs remains at high levels throughout mouse gestation. The expression pattern of CPZ overlaps with the expression pattern of several Wnt genes, consistent with the putative role of CPZ in Wnt signaling.

Structural characterization of the rat carboxypeptidase-E gene.

Several genomic clones encoding carboxypeptidase-E (CPE) have been isolated and partially sequenced. Southern blot analysis indicates that a single copy of this gene is present in the rat genome. The entire gene spans approximately 50 kilobases and consists of nine exons, each of which contains protein-coding regions. Only one of the exon/intron junctions of the rat CPE gene is present in a comparable position within the genes for carboxypeptidase-A and -B, both of which are only 17-21% homologous to CPE at the amino acid level. Nuclease protection analysis shows that alternative splicing of exons 7, 8, and 9 does not occur, indicating that the heterogeneity of the C-terminal region of CPE is due to posttranslational processing. Primer extension and nuclease protection analyses have identified the 5' end of CPE mRNA to be 105 nucleotides up-stream from the ATG used for protein translation. The 5' flanking region does not contain TATA and/or CCAAT boxes in the near vicinity of the transcription initiation site. The 5' flanking region is GC rich, containing 70% GC residues over nucleotides -1 to -150 (relative to the transcription initiation site). Putative consensus sites for the enhancer elements SP-1, NF-1, Pan-1, and AP-2 are present in the region from -60 to -330. Since this report describes the first neuropeptide-processing enzyme gene to be partially sequenced, it is not possible to compare the sequence with those of other processing enzymes that show similar tissue-specific expression. However, comparison of the CPE sequence with 5' flanking regions of other neuroendocrine genes has revealed a short region (12-18 nucleotides) that is highly conserved among CPE, neuropeptide-Y, oxytocin, insulin, and tyrosine hydroxylase genes.

Expression of the rat carboxypeptidase-E gene in neuroendocrine and nonneuroendocrine cell lines.

To identify cis-acting elements involved with the expression of the rat carboxypeptidase-E (CPE) gene, constructs containing various regions of the 5'-flanking region of the CPE gene attached to the luciferase reporter gene were transiently expressed in cell lines derived from pituitary (AtT-20 and GH4C1), liver (SK-HEP-1), and kidney (HEK293 and COS1). Regions of the CPE gene spanning the major transcription initiation site (-12 to 47) are sufficient for low levels of transcription. Activity is enhanced 3- to 15-fold by sequences present between -12 and -395 in all cell lines examined. Sequences between -395 and -3081 influenced transcription activity up to 5-fold in some, but not all, cell lines. There was no correlation between the transcription activities of the various constructs and the level of endogenous CPE mRNA in the cell lines, indicating that the tissue-specific elements responsible for the large variations in endogenous CPE mRNA levels are not present within -3081 to 47. The region between -395 and 45 was examined in greater detail using transient expression assays and DNase-I protection analysis. Transcription activity is enhanced in GH4C1 and HEK293 cells by sequence present between -12 and -84; this region contains a potential GC box, which binds factors present in GH4C1 nuclear extracts. Other regions between -340 and 80 that bind proteins in the GH4C1 nuclear extracts include the major transcription initiation site, which has homology to the initiator sequence; the pituitary-specific transcription initiation sites (-101 and -105); and sequences with homology to NF-1, Pan-1, simian virus-40 enhancer core, and AP-2-binding sites. Taken together, these results suggest that basal expression of the CPE gene from its major transcription initiation site, which does not contain an up-stream TATA box, is primarily under the control of an initiator-like element together with an upstream GC box.

Isolation and sequence analysis of cDNA for rat carboxypeptidase E [EC 3.4.17.10], a neuropeptide processing enzyme.

Carboxypeptidase E (CPE) is the carboxypeptidase B-like enzyme associated with the biosynthesis of numerous peptide hormones and neurotransmitters. This enzyme has been previously purified to homogeneity from bovine tissues, and cDNA clones (non-full length) isolated from a bovine pituitary cDNA library. In the present study, cDNA encoding full-length rat CPE has been isolated and sequenced. Both the nucleotide and amino acid sequences of rat CPE show substantial homology with the bovine sequences. The bovine and rat nucleotide sequences are homologous within the entire coding region, as well as within several portions of the 3'-untranslated region. The predicted amino acid sequence of rat CPE is greater than 90% homologous with the bovine enzyme. Northern blot analyses indicate a single species of CPE mRNA approximately 2100 nucleotides in length to be present in many neural and endocrine tissues. High levels of CPE mRNA are present in rat hypothalamus, hippocampus, midbrain, striatum, and cerebral cortex; and moderate levels are present in the brain stem, cerebellum, heart, adrenal, and eye. Low levels are detected in testis and duodenum, but not in liver or thymus. This tissue-specific expression of CPE mRNA is consistent with the proposed role for this enzyme in the production of numerous peptide hormones and neurotransmitters.

Missense polymorphism in the human carboxypeptidase E gene alters enzymatic activity.

Carboxypeptidase E (CPE) is involved in the biosynthesis of peptide hormones and neurotransmitters, including insulin. One of the features of type 2 diabetes mellitus (T2DM) is an elevation in the proinsulin level and/or proinsulin/insulin molar ratio, suggesting that mutations in proinsulin processing enzymes may contribute to the development of T2DM. We scanned CPE for mutations in a collection of Ashkenazi T2DM families and identified five novel single nucleotide polymorphisms (SNPs). An SNP in the 283(rd) codon, c.847C>T, changes arginine to tryptophan (R283W). The residue Arg283 is conserved among CPE orthologs as well as most enzymatically active metallocarboxypeptidases. Of the 272 Ashkenazi T2DM pedigrees screened, we found four families segregating R283W. Within these four families, patients who inherited one copy of this variant had much earlier age of onset for T2DM. The R283W CPE protein cleaves peptide substrates with substantially lower efficiencies and is less stable at elevated temperature. In addition, the R283W CPE variant has a narrower pH optimum and is much less active at pH 6.0-6.5, indicating that the R283W CPE variant would be substantially less active than wild type CPE in the trans-Golgi network and immature secretory vesicles where the enzyme functions in vivo. To summarize, we uncovered a rare non-conservative missense mutation in CPE and demonstrated that the mutant protein has altered enzymatic properties. We predict that this mutant could cause hyperproinsulinism and diabetes in the homozygous state.

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.

Effect of carboxypeptidase E deficiency on progastrin processing and gastrin messenger ribonucleic acid expression in mice with the fat mutation.

Proforms of gastrointestinal peptides are cleaved at paired basic residues into intermediate forms. Paired basic residues at the C-terminal then are excised by carboxypeptidases before the peptide is amidated. An obese mouse, called Cpe(fat)/Cpe(fat), has a missense mutation in carboxypeptidase E (CPE) with no pancreatic CPE activity and a reduced processing of pancreatic proinsulin to insulin. The purpose of this study was 1) to look for the presence of CPE in the antrum of the stomach, duodenum, and colon in the Cpe(fat)/Cpe(fat) mouse; 2) to determine whether CPE is involved in the processing of progastrin (Pro-G) to its carboxyl-terminal amidated form; and 3) to determine whether a decrease in amidated gastrin results in an up-regulation of stomach gastrin messenger RNA (mRNA) levels. In Cpe(fat)/Cpe(fat) mice, CPE activity was absent in the antrum and colon. In Cpe(fat)/Cpe(fat) mice, amidated gastrin levels were reduced significantly. Levels of the precursor for amidated gastrin (gastrin-Gly-Arg-Arg) were markedly elevated. Gastrin mRNA levels were increased approximately 2-fold over the levels in Cpe(fat)/Cpe(fat) mice. These results indicate that CPE is needed for processing progastrin to gastrin in the stomach and that amidated gastrin exerts an inhibitory feedback effect on gastrin mRNA levels.

Characterization of Aplysia carboxypeptidase E.

Carboxypeptidase E (CPE) is involved in the biosynthesis of peptide hormones and neurotransmitters. To determine whether a recently reported Aplysia californica cDNA encodes a CPE-like enzyme, this cDNA was expressed in the baculovirus system. The Aplysia CPE is optimal at pH 5.5-6.5 and is inhibited by chelating agents and by the sulfhydryl reagent p-chloromercuriphenyl sulfonate. The effect of divalent cations and active site-directed inhibitors on enzyme activity are generally similar for Aplysia and rat CPE. Western blot analysis using antisera to the N- and C-terminal regions of the Aplysia CPE show that the Aplysia CPE is present in atrial glands and ovotestis. This Aplysia CPE is purified on a p-aminobenzoyl-Arg Sepharose affinity column under conditions that selectively purify rat CPE. Taken together, these results suggest that the previously cloned cDNA represents a CPE-like enzyme that is expressed in Aplysia tissue.

Expression of the carboxypeptidase E gene: characterization of the initiator-binding proteins.

Several of the genes for enzymes involved in peptide hormone processing, such as carboxypeptidase E (CPE), do not contain a TATA box. The region surrounding the major transcription initiation site of the CPE gene has sequence homology with the 'initiator' (Inr) elements of the TATA-less terminal deoxynucleotidyltransferase (TdT) gene, and the adenovirus major late (AdML) and other promoters. To investigate the promoter region of the CPE gene, GH4C1 cells were transiently transfected with constructs containing the luciferase reporter gene attached to various portions of the rat CPE gene (-395 to +45). Positive regulator elements were detected in positions -84 to -12 and +30 to +47. However, the Inr-like element of the CPE gene (-12 to +20) produced detectable luciferase activity in the absence of upstream and downstream sequences. This region of the CPE gene was much more active when expressed in the normal (sense) orientation than when expressed in the antisense orientation. A mutation within the consensus sequence between CPE and other Inr elements was much less active than the wild-type sequence. Interestingly, deletion of the Inr and surrounding sequences produced a large increase in the transcription from upstream sites, suggesting that proteins which bind at, or near, the Inr sequence suppress transcription from other sites. To characterize GH4C1 nuclear proteins which bind to the CPE gene, Southwestern blotting, UV cross-linking, and gel shift analyses were performed. The Southwestern analysis showed that the CPE and AdML Inr sequences labeled several proteins of similar sizes which are distinct from the transcription factor USF; this factor has been previously reported to bind to the AdML Inr sequence. A CPE Inr-binding protein co-purifies with an AdML Inr-binding protein on a CPE Inr affinity column. Gel shift assays showed that with some binding conditions, the proteins that bind to the CPE sequence also bind to the TdT and AdML Inr elements. Taken together, these results indicate that the -12 to +20 region of the CPE gene has the properties of an Inr element which binds some, but not all, of the factors which bind to other Inr elements.

Distribution of proSAAS-derived peptides in rat neuroendocrine tissues.

Using a technique to identify substrates of the peptide processing enzyme carboxypeptidase E (CPE), several novel peptides were detected in the brain and pituitary of Cpe(fat)/Cpe(fat) mice and found to be derived from a single precursor, termed proSAAS. In order to gain further information regarding the potential physiological roles of these peptides, we have examined the distribution of two proSAAS-derived peptides, ARPVKEPRSLSAASAPLAETSTPLRL (SAAS) and LENSSPQAPARRLLPP (LEN), in rat neuroendocrine tissues using immunohistochemistry. Both peptides are detected throughout the brain, with the highest concentrations of SAAS peptide in the hypothalamus. In the hippocampus, both peptides are co-localized with prohormone convertase 1 in the dentate gyrus and CA1-3 region. In cerebellum, SAAS peptide is co-localized with prohormone convertase 1 in Purkinje and granular cells, whereas LEN is much more abundant in the Purkinje cells relative to the granular cells. Similarly, SAAS and prohormone convertase 1 are co-localized in the dorsal horn of the spinal cord, while LEN is mainly restricted to fibers of the white matter. In the pituitary, SAAS, LEN, and prohormone convertase 1 are detected in all three lobes. In the pancreas, SAAS, LEN, and prohormone convertase 1 are only detected in the islets, although the peptides are enriched in the peripheral cells (alpha and/or delta) while prohormone convertase 1 is only expressed in the inner cells (beta). Both SAAS and LEN are present in the adrenal medulla along with prohormone convertase 1. Taken together, these data are consistent with the proposed role for proSAAS as an endogenous inhibitor of prohormone convertase 1 in many, but not all cell types. However, the broader localization of the peptides allows for the possibility that they perform additional functions.

Carboxypeptidase D is a potential candidate to carry out redundant processing functions of carboxypeptidase E based on comparative distribution studies in the rat central nervous system.

Post-translational processing is essential for the biological activation of many proteins and peptides. After precursor cleavage at specific single residues or pairs of basic residues by the proprotein convertases, the C-terminal basic residues are removed. Carboxypeptidase E was thought to be the only enzyme responsible. Recent studies with carboxypeptidase E-deficient mice, Cpe(fat)/Cpe(fat), indicated the existence of carboxypeptidase E-like carboxypeptidases, such as carboxypeptidase D. In order to define potential redundant functions in vivo, we compared the distributions of both carboxypeptidases in the rat central nervous system and selected endocrine tissues. Carboxypeptidase D messenger RNA was abundantly expressed in glial cells in the gray and white matter, while neurons in several brain regions, such as the piriform cortex, basolateral amygdala and hippocampus, also expressed high levels of carboxypeptidase D messenger RNA. Co-localization of carboxypeptidases E and D messenger RNAs was observed in many brain regions, the spinal cord and endocrine tissues. Immunohistochemistry showed the intracellular distribution of carboxypeptidase D with a perinuclear pattern. The extensive distribution of carboxypeptidase D in both glial and neuronal cells indicates the important role of carboxypeptidase D in peptide processing, possibly working together with furin, a ubiquitously expressed proprotein convertase. The co-localization of carboxypeptidases D and E suggests that carboxypeptidase D may, at least partially, compensate for carboxypeptidase E processing functions in Cpe(fat)/Cpe(fat) mice.

Immunohistochemical localization of carboxypeptidases E and D in the human placenta and umbilical cord.

Carboxypeptidase E (CPE) is highly concentrated in neuroendocrine tissues and is the only carboxypeptidase detected in mature secretory vesicles. Carboxypeptidase D (CPD), a carboxypeptidase with CPE-like activity, is widely distributed in tissues and is present in the trans-Golgi network. Previous work had shown that both CPE and CPD are expressed in the human placenta and that CPD is expressed at much higher levels than CPE. The present work provides evidence for the co-localization of CPE and CPD to basal plate extravillous trophoblasts and maternal uteroplacental vascular endothelial cells, chorionic villous endothelial cells, amnionic epithelial cells, and umbilical venous and arterial smooth muscle cells. Whereas the intensity of CPD immunostaining is similar in the placenta and umbilical cord, CPE staining in the placenta is much weaker than in the umbilical cord, suggesting that CPD plays a more important role in the processing of placental peptides. Immunoelectron microscopy of umbilical venous smooth muscle cells shows subcellular localization of both enzymes to the rough endoplasmic reticulum. In addition, CPE is present just subjacent to the cell membrane. The difference in cellular and subcellular localization between the two enzymes indicates that they perform distinct functions in the processing of placental peptides and proteins.

Processing of prodynorphin in BRL-3A cells, a rat liver-derived cell line: implications for the specificity of neuropeptide-processing enzymes.

Prodynorphin is post-translationally processed into dynorphin B-13 and other peptides by the action of endopeptidases that cleave at pairs of basic amino acids and at single basic residues, followed by a carboxypeptidase that removes the C-terminal basic residues. To evaluate the specificity of neuropeptide processing enzymes, rat prodynorphin was transfected into BRL-3A cells, a rat liver-derived cell line which produces insulin-like growth factor II, but does not normally express prodynorphin. The transfected prodynorphin was post-translationally processed at both monobasic and dibasic cleavage sites, with the formation of dynorphin B-13 and other peptides. This finding indicates that BRL-3A cells express prodynorphin-processing enzymes. These cells were found to secrete two enzyme activities previously implicated in the processing of dynorphin, a monobasic cleaving 'dynorphin converting enzyme' and 'carboxypeptidase E', based on inhibitor sensitivities and pH optima. The dynorphin converting enzyme secreted from BRL-3A cells elutes from an anion exchange column under the same conditions as the enzyme secreted from pituitary-derived cell lines (AtT-20, GH4C1). Northern blot analysis indicates that BRL-3A cells express carboxypeptidase E mRNA in addition to mRNA encoding furin, a prohormone-processing endopeptidase. The mRNAs for two other related endopeptidases, prohormone convertase 1 and 2, were not detected on Northern blots, suggesting that these enzymes are not required for the processing of prodynorphin. The expression of carboxypeptidase E, furin, and dynorphin converting enzyme in BRL-3A cells suggests that these peptide processing enzymes are not specific for neuropeptides, but are also present in cells which process peptide growth factors.

Dopamine antagonist haloperidol increases carboxypeptidase E mRNA in rat neurointermediate pituitary but not in various other rat tissues.

Carboxypeptidase E (CPE) is involved with the biosynthesis of many neuropeptides, including several whose genes are regulated by haloperidol treatments. In this study, we examined whether haloperidol alters CPE mRNA levels in a variety of tissues. Rats were treated for either 1, 3, 7, 14, or 21 days with 2 mg/kg haloperidol, and then Northern blot analysis performed on RNA from neurointermediate pituitary, anterior pituitary, hypothalamus, striatum, cerebellum, and adrenal. The 14 and 21 day treatments produced a significant 90-110% elevation of CPE mRNA in neurointermediate pituitary. However, the levels of CPE mRNA in the other tissues were not significantly influenced by the haloperidol treatments. This finding indicates that CPE is not co-regulated with peptide hormone mRNAs in all tissues.

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.

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.