A meta-analysis of peripheral blood nerve growth factor levels in patients with schizophrenia.

Neurotrophins particularly brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) are crucial modulators in the neurodevelopment and maintenance of central and peripheral nervous systems. Neurotrophin hypothesis of schizophrenia (SCZ) postulated that the changes in the brains of SCZ patients are the result of disturbances of developing processes involving neurotrophic factors. This hypothesis was mainly supported by the abnormal regulation of BDNF in SCZ, especially the decreased peripheral blood BDNF levels in SCZ patients validated by several meta-analyses. However, the regulation of NGF in SCZ remains unclear because of the inconsistent findings from the clinical studies. Therefore, we undertook, to the best of our knowledge, the first systematic review with a meta-analysis to quantitatively summarize the peripheral blood NGF data in SCZ patients compared with healthy control (HC) subjects. A systematic search of Pubmed, PsycINFO and Web of Science identified 13 articles encompassing a sample of 1693 individuals for the meta-analysis. Random-effects meta-analysis showed that patients with SCZ had significantly decreased peripheral blood levels of NGF when compared with the HC subjects (Hedges's g=-0.633, 95% confidence interval (CI)=-0.948 to -0.318, P<0.001). Subgroup analyses revealed reduced NGF levels both in serum (Hedges's g=-0.671, 95% CI=-1.259 to -0.084, P=0.025) and plasma (Hedges's g=-0.621, 95% CI=-0.980 to -0.261, P<0.001) of the patients, and in drug-free (Hedges's g=-0.670, 95% CI=-1.118 to -0.222, P=0.003) and medicated (Hedges's g=-0.357, 95% CI=-0.592 to -0.123, P=0.003) patients with SCZ. Furthermore, meta-regression analyses showed that age, gender and sample size had no moderating effects on the outcome of the meta-analysis, whereas disease severity might be a confounding factor for the meta-analysis. These results demonstrated that patients with SCZ are accompanied by the decreased peripheral blood NGF levels, strengthening the clinical evidence of an abnormal neurotrophin profile in the patients with SCZ.

Decreased peripheral brain-derived neurotrophic factor levels in Alzheimer's disease: a meta-analysis study (N=7277).

Studies suggest that dysfunction of brain-derived neurotrophic factor (BDNF) is a possible contributor to the pathology and symptoms of Alzheimer's disease (AD). Several studies report reduced peripheral blood levels of BDNF in AD, but findings are inconsistent. This study sought to quantitatively summarize the clinical BDNF data in patients with AD and mild cognitive impairment (MCI, a prodromal stage of AD) with a meta-analytical technique. A systematic search of Pubmed, PsycINFO and the Cochrane Library identified 29 articles for inclusion in the meta-analysis. Random-effects meta-analysis showed that patients with AD had significantly decreased baseline peripheral blood levels of BDNF compared with healthy control (HC) subjects (24 studies, Hedges' g=-0.339, 95% confidence interval (CI)=-0.572 to -0.106, P=0.004). MCI subjects showed a trend for decreased BDNF levels compared with HC subjects (14 studies, Hedges' g=-0.201, 95% CI=-0.413 to 0.010, P=0.062). No differences were found between AD and MCI subjects in BDNF levels (11 studies, Hedges' g=0.058, 95% CI=-0.120 to 0.236, P=0.522). Interestingly, the effective sizes and statistical significance improved after excluding studies with reported medication in patients (between AD and HC: 18 studies, Hedges' g=-0.492, P<0.001; between MCI and HC: 11 studies, Hedges' g=-0.339, P=0.003). These results strengthen the clinical evidence that AD or MCI is accompanied by reduced peripheral blood BDNF levels, supporting an association between the decreasing levels of BDNF and the progression of AD.

The Chromogranin A-derived sympathomimetic serpinin depresses myocardial performance in teleost and amphibian hearts.

Chromogranin A (CgA) is an acidic protein co-stored with catecholamines, hormones and neuropeptides in the secretory granules of endocrine, neuronal and other cell types (including cardiomyocytes). Proteolytic cleavage in the C terminus of CgA generates a 2.9kDa peptide named serpinin (Serp; Ala26Leu) that can be modified at its N terminus to form a pyroglutamate residue (pGlu-Serp). In the rat heart, both peptides increase contractility and relaxation through a ß-adrenergic-like action mechanism. Accordingly, Serp and pGlu-Serp were proposed as novel myocardial sympatho-adrenergic modulators in mammals. On a comparative basis, here we report the actions of Serp and pGlu-Serp on myocardial contractility in three poikilotherm vertebrate species: the eel (Anguilla anguilla), the goldfish (Carassius auratus) and the frog (Rana esculenta). Using isolated working heart preparations, we show that pGlu-Serp reduces stroke volume in all species tested, while Serp reduces contractility in the frog heart, but is uneffective in eel and goldfish hearts. In the goldfish and frog hearts, pGlu-Serp activates the Nitric Oxide/cGMP pathway involving Endothelin-1 B receptors (frog) and ß3 adrenergic receptors (goldfish). pGlu-Serp-treated hearts from goldfish and frog show increased cGMP content. Moreover, the exposure of the frog heart to pGlu-Serp is accompanied by an increased expression of activated eNOS and Akt. In conclusion, this first report showing that pGlu-Serp inhibits mechanical cardiac performance in teleost and amphibians supports an evolutionary role of the CgA system, and particularly its serpinin component, in the sympatho-adrenergic control of the vertebrate heart.

Neurotrophic factor-α1 prevents stress-induced depression through enhancement of neurogenesis and is activated by rosiglitazone.

Major depressive disorder is often linked to stress. Although short-term stress is without effect in mice, prolonged stress leads to depressive-like behavior, indicating that an allostatic mechanism exists in this difference. Here we demonstrate that mice after short-term (1 h per day for 7 days) chronic restraint stress (CRS), do not display depressive-like behavior. Analysis of the hippocampus of these mice showed increased levels of neurotrophic factor-α1 (NF-α1; also known as carboxypeptidase E, CPE), concomitant with enhanced fibroblast growth factor 2 (FGF2) expression, and an increase in neurogenesis in the dentate gyrus. In contrast, after prolonged (6 h per day for 21 days) CRS, mice show decreased hippocampal NF-α1 and FGF2 levels and depressive-like responses. In NF-α1-knockout mice, hippocampal FGF2 levels and neurogenesis are reduced. These mice exhibit depressive-like behavior that is reversed by FGF2 administration. Indeed, studies in cultured hippocampal neurons reveal that NF-α1 treatment directly upregulates FGF2 expression through extracellular signal-regulated kinase-Sp1 signaling. Thus, during short-term CRS, hippocampal NF-α1 expression is upregulated and has a key role in preventing the onset of depressive-like behavior through enhanced FGF2-mediated neurogenesis. To evaluate the therapeutic potential of this pathway, we examined, rosiglitazone (Rosi), a PPARγ agonist, which has been shown to have antidepressant activity in rodents and humans. Rosi upregulates FGF2 expression in a NF-α1-dependent manner in hippocampal neurons. Mice fed Rosi show increased hippocampal NF-α1 levels and neurogenesis compared with controls, thereby indicating the antidepressant action of this drug. Development of drugs that activate the NF-α1/FGF2/neurogenesis pathway can offer a new approach to depression therapy.

Novel biomarker SYT12 may contribute to predicting papillary thyroid cancer outcomes.

AIM: To investigate biomarkers for predicting papillary thyroid cancer outcomes. MATERIALS & METHODS: The expression of biomarkers (ITGA2, SYT12 and CDH3) was studied in a prospective cohort of patients with papillary thyroid cancer. Three outcomes of initial metastases, baseline status and longitudinal status were analyzed and correlated with the biomarkers. RESULTS: SYT12 provided the best prediction of initial metastasis (sensitivity: 72%; specificity: 54%). SYT12 had the highest accuracy for predicting longitudinal status (sensitivity: 100%; specificity: 47%). The best performance for longitudinal status resulted from combining SYT12 with American Thyroid Association risk stratification, with sensitivity and specificity of 88 and 73%, respectively. CONCLUSION: SYT12 has some prognostic significance in papillary thyroid cancer. Further validation studies in larger populations are warranted.

Carboxypeptidase E (CPE) inhibits the secretion and activity of Wnt3a.

The Wnt pathway has essential roles in cell proliferation, cell fate determination and tumorigenesis by regulating the expression of a wide range of target genes. As a core signaling cascade, the canonical Wnt pathway is regulated at different levels by numerous proteins. We have previously shown that carboxypeptidase E (CPE) is a novel regulator of the canonical Wnt signaling pathway. Here, we show that CPE and the Wnt3a ligand are co-secreted from cells. We show that although the C'-terminal Lys residue of Wnt3a is critical for its activity and is important for the effect of CPE on the Wnt pathway, CPE does not execute its effect by removing this Wnt3a residue. Interestingly, CPE through its N'-terminal sequence, forms aggregates with Wnt3a and possible endoplasmic reticulum (ER) stress leading to its loss of function. Together, our current results provide a mechanistic insight into the way CPE regulates the canonical Wnt signaling pathway.

A human carboxypeptidase E/NF-α1 gene mutation in an Alzheimer's disease patient leads to dementia and depression in mice.

Patients with Alzheimer's disease (AD), a common dementia among the aging population, often also suffer from depression. This comorbidity is poorly understood. Although most forms of AD are not genetically inherited, we have identified a new human mutation in the carboxypeptidase E (CPE)/neurotrophic factor-α1 (NF-α1) gene from an AD patient that caused memory deficit and depressive-like behavior in transgenic mice. This mutation consists of three adenosine inserts, introducing nine amino acids, including two glutamines into the mutant protein, herein called CPE-QQ. Expression of CPE-QQ in Neuro2a cells demonstrated that it was not secreted, but accumulated in the endoplasmic reticulum and was subsequently degraded by proteasomes. Expression of CPE-QQ in rat hippocampal neurons resulted in cell death, through increased ER stress and decreased expression of pro-survival protein, BCL-2. Transgenic mice expressing CPE-QQ did not show any difference in the processing enzyme activity of CPE compared with wild-type mice. However, the transgenic mice exhibited poor memory, depressive-like behavior, severely decreased dendrites in the hippocampal CA3 region and medial prefrontal cortex indicative of neurodegeneration, hyperphosphorylation of tau at Ser396, and diminished neurogenesis in the dentate gyrus at 50 weeks old. All these pathologies are associated with AD and the latter with depression and were observed in 50-week-old mice. Interestingly, the younger CPE-QQ mice (11 weeks old) did not show deficits in dendrite outgrowth and neurogenesis. This study has uncovered a human CPE/NF-α1 gene mutation that could lead to comorbidity of dementia and depression, emphasizing the importance of this gene in cognitive function.

Biosynthesis, processing, and control of release of melanotropic peptides in the neurointermediate lobe of Xenopus laevis.

The neurointermediate lobes of dark-adapted toads Xenopus laevis were incubated for 30 min in [3H]arginine and then "chased" for various time periods. By use of this pulse-chase paradigm there were detected 10 trichloroacetic acid (TCA)-precipitable peptides separated on acid-urea polyacrylamide gels and one TCA-soluble peptide separated by high-voltage electrophoresis (pH 4.9) with melanotropic activity. Each of these peptides had a different degree of melanocyte stimulating hormone (MSH) activity as revealed by the Anolis skin bioassay. Three of these TCA-precipitable peptides comigrated with ACTH, beta-lipotrophin, and alpha-MSH on acid-urea gels. Evidence suggesting a precursor-product mode of biosynthesis of the melanotropic peptides is presented. 7 of the 10 TCA-precipitable peptides and the one TCA-soluble peptide with melanotropic activity were released into the medium. The half-time of release of the TCA-precipitable peptides was about 2 h, whereas the half-time of TCA-soluble peptide release was about 30 min. The release of these peptides was inhibited by 5 X 10(-5) M dopamine. Dopamine inhibition of release did not appear to affect the biosynthesis of the melanotropic peptides, but did appear to enhance the degradation of the newly synthesized TCA-soluble peptide in the tissue. White adaptation of the toads greatly decreased the biosynthesis of all of the TCA-precipitable melanotropic peptides.

Receptor-mediated targeting of hormones to secretory granules: role of carboxypeptidase E.

Peptide hormones, neuropeptides, and other molecules such as the granins are specifically packaged into granules of the regulated secretory pathway and released in a calcium-dependent manner upon stimulation. Many of these molecules are synthesized as larger precursors (prohormones) that are processed to biologically active products within the granules. It has now become apparent that prohormones, proneuropeptides, and the granins contain conformation-dependent sorting signal motifs that facilitate their specific sorting and packaging into regulated secretory granules. Recently, a receptor to which these sorting signals bind has been identified as the membrane form of carboxypeptidase E (CPE) and localized to the Golgi apparatus, where sorting occurs, specifically at the trans-Golgi network. In this article, we review the evidence for a sorting signal-receptor-mediated mechanism for routing peptide hormones and prohormones to the regulated secretory granules. We also describe a mouse model, Cpe(fat), which has the CPE gene naturally mutated. Pituitary hormones were misrouted and secreted in an unregulated manner via the constitutive pathway in these Cpe(fat) mice, leading to endocrine disorders.

Identification of a novel prohormone sorting signal-binding site on carboxypeptidase E, a regulated secretory pathway-sorting receptor.

Sorting of the prohormone POMC to the regulated secretory pathway necessitates the binding of a sorting signal to a sorting receptor, identified as membrane carboxypeptidase E (CPE). The sorting signal, located at the N terminus of POMC consists of two acidic (Asp10, Glu14) and two hydrophobic (Leu11, Leu18) residues exposed on the surface of an amphipathic loop. In this study, molecular modeling of CPE predicted that the acidic residues in the POMC-sorting signal bind specifically to two basic residues, Arg255 and Lys260, present in a loop unique to CPE, compared with other carboxypeptidases. To test the model, these two residues on CPE were mutated to Ser or Ala, followed by baculovirus expression of the mutant CPEs in Sf9 cells. Sf9 cell membranes containing CPE mutants with either Arg255 or Lys260, or both residues substituted, showed no binding of [125I]N-POMC1-26 (which contains the POMC-sorting signal motif), proinsulin, or proenkephalin. In contrast, substitution of an Arg147 to Ala147 at a substrate-binding site, Arg259 to Ala259 and Ser202 to Pro202, in CPE did not affect the level of [125I]N-POMC1-26 binding when compared with-wild type CPE. Furthermore, mutation of the POMC-sorting signal motif (Asp10, Leu11, Glu14, Leu18) eliminated binding to wild-type CPE. These results indicate that the sorting signal of POMC, proinsulin, and proenkephalin specifically interacts with Arg255 and Lys260 at a novel binding site, independent of the active site on CPE.

pGlu-serpinin protects the normotensive and hypertensive heart from ischemic injury.

Serpinin peptides derive from proteolytic cleavage of Chromogranin-A at C-terminus. Serpinin and the more potent pyroglutaminated-serpinin (pGlu-Serp) are positive cardiac ß-adrenergic-like modulators, acting through ß1-AR/AC/cAMP/PKA pathway. Because in some conditions this pathway and/or other pro-survival pathways, activated by other Chromogranin-A fragments, may cross-talk and may be protective, here we explored whether pGlu-Serp cardioprotects against ischemia/reperfusion injury under normotensive and hypertensive conditions. In the latter condition, cardioprotection is often blunted because of the limitations on pro-survival Reperfusion Injury Salvage Kinases (RISK) pathway activation. The effects of pGlu-Serp were evaluated on infarct size (IS) and cardiac function by using the isolated and Langendorff perfused heart of normotensive (Wistar Kyoto, WKY) and spontaneously hypertensive (SHR) rats exposed to ischemic pre-conditioning (PreC) and post-conditioning (PostC). In both WKY and SHR rat, pGlu-Serp induced mild cardioprotection in both PreC and PostC. pGlu-Serp administered at the reperfusion (Serp-PostC) significantly reduced IS, being more protective in SHR than in WKY. Conversely, left ventricular developed pressure (LVDevP) post-ischemic recovery was greater in WKY than in SHR. pGlu-Serp-PostC reduced contracture in both strains. Co-infusion with specific RISK inhibitors (PI3K/Akt, MitoKATP channels and PKC) blocked the pGlu-Serp-PostC protective effects. To show direct effect on cardiomyocytes, we pre-treated H9c2 cells with pGlu-Serp, which were thus protected against hypoxia/reoxygenation. These results suggest pGlu-Serp as a potential modulatory agent implicated in the protective processes that can limit infarct size and overcome the hypertension-induced failure of PostC.

Evidence for intragranular processing of pro-opiocortin in the mouse pituitary intermediate lobe.

Mouse pituitary neurointermediate lobes were pulse-incubated in [3H] arginine or [3H] lysine for 10 min and then chase-incubated for periods 0 to 4h. The labeled peptides from the lobes were analysed by immunoprecipitation with specific antisera, and thereafter, by acid-urea polyacrylamide gel electrophoresis. Using this paradigm, the synthesis of a prohormone common to adrenocorticotropin (ACTH) and endorphin was detected in 10 min pulse labeled lobes. Following a chase period, processing of the prohormone to several forms of ACTH (mol. wt. 25000, 23000, and 13000), beta-lipotropin and beta-endorphin was observed. To determine the intracellular site of processing of the prohormone, subcellular fractionation studies of labeled lobes were carried out. Analysis of the fractions from the pulse-labeled lobes revealed that the newly synthesized labeled prohormone was primarily localized in a granule-enriched fraction. In lobes that were pulsed and then chase-incubated for 1 h, there was a decrease in the amount of prohormone and an appearance of processed products in the granule-enriched fraction. In another paradigm, where the secretory granule-fraction was isolated from pulse-labeled lobes and then incubated in vitro for 6 h at pH 5.5, processing of the endogenous labeled prohormone within the isolated granule fraction was observed. These data suggest, that in the mouse neurointermediate lobe, the ACTH/endorphin prohormone (pro-opiocortin) is rapidly packaged into secretory granules after synthesis and processed intragranularly.

The amino-terminal sequence of pro-opiomelanocortin directs intracellular targeting to the regulated secretory pathway.

The molecular signal which targets the pro-opiomelanocortin (POMC) prohormone into the regulated secretory pathway was investigated. DNA sequences encoding the first 10, 26, 50, and 101 N-terminal amino acids of mouse POMC were fused in frame to the chloramphenicol acetyltransferase (CAT) gene and expressed in AtT-20 cells. Immunofluorescence microscopy using antibody directed against CAT indicated that fusion proteins carrying 26, 50 and 101 amino acids of N-POMC were directed to secretory granules. This finding was confirmed by secretion studies in which 1 microM forskolin stimulated the release of fusion proteins carrying 26 and 101 amino acids of N-POMC, whereas no regulated secretion was observed with the shortest fusion protein. Subcellular fractionation studies also indicated the presence of the fusion proteins with 26 and 101 amino acids of N-POMC in secretory granules. These results provide evidence that the signal directing POMC to secretory granules is contained within the N-terminus of the prohormone, with the first 26 amino acids being sufficient for targeting. Binding studies showed that N-POMC1-76 bound to secretory granule membranes specifically on the luminal side and in a pH-sensitive manner. Only N-POMC1-76 bound optimally to secretory granule membranes at pH 5 to 6.5, but not the ACTH1-39 (mid), corticotropin-like intermediate lobe peptide (CLIP) and beta-lipotropin (C-terminal) domains of POMC. Such binding may be involved in the mechanism of sorting POMC to the regulated secretory pathway.

Effect of salt loading on proopiomelanocortin (POMC) messenger ribonucleic acid levels, POMC biosynthesis, and secretion of POMC products in the mouse pituitary gland.

The effect of salt loading on POMC-derived peptide secretion, POMC mRNA levels, and POMC biosynthesis in the mouse pituitary gland was investigated. Plasma alpha MSH levels decreased to 56.9% of the control value after 2 days of salt loading. Concomitantly, POMC mRNA levels and POMC biosynthesis in the intermediate lobe decreased to 47.7% and 66.4% of the control value, respectively. After 4 days of salt loading, plasma alpha MSH levels, POMC mRNA, and POMC biosynthesis returned to control levels. ACTH secretion increased to 142.2% of the control value after 2 days of salt loading, but decreased to 73.4% of the control levels after 4 days. After 9 days, ACTH secretion was still low; however, by 12 days, plasma ACTH levels in salt-loaded mice were not significantly different from control levels. POMC mRNA levels in the anterior lobe increased to 205.0% of the control value after 2 days of salt loading, continued to be high after 4 days, but returned to control levels by 9 days. However, POMC biosynthesis was not altered under these conditions during the entire period of salt loading. Thus, POMC mRNA levels and biosynthesis were modulated differently in the intermediate and anterior lobes of the mouse pituitary gland during salt loading.

In vitro processing of proopiocortin by membrane-associated and soluble converting enzyme activities from rat intermediate lobe secretory granules.

Secretory granules (SGs) from rat intermediate lobes (IL) were isolated in a highly purified form by differential centrifugation, followed by sucrose density gradient centrifugation. The purified IL-SGs were lysed by freezing and thawing. The granule lysate was then centrifuged to generate membrane and soluble fractions. Proopiocortin -converting enzyme (PCE) activity was assayed by incubation of [3H]arginine- or [3H] phenylalanine-labeled toad proopiocortin with the total granule lysate, the membrane, or the soluble fraction at pH 5.0. The processed products were identified by immunoprecipitation with ACTH and beta-endorphin antisera, followed by acid-urea-gel electrophoresis. The PCE activity in rat IL-SG lysate cleaved proopiocortin to 21,000 mol wt ACTH, 21,000 mol wt ACTH/lipotropin (LPH), 13,000 mol wt ACTH, beta LPH, beta-endorphin-like peptides, and alpha MSH-like peptides, similar to those synthesized by the toad intermediate lobe in situ. Treatment of the PCE cleavage products with carboxypeptidase B resulted in the liberation of free arginine. This observation together with the nature of the products formed suggest that the PCE activity cleaved at pairs of basic residues of proopiocortin , yielding one or more products that terminated with an arginine or an arginine-lysine. PCE activity was found in membrane and soluble granule fractions, and both activities were inhibited by leupeptin, p-chloromercuribenzoate, dithiodipyridine, and pepstatin A, but not by chloroquine or N-alpha-p-tosyl-L-lysine-chloromethylketone HCl. Diisopropyl fluorophosphate and other thiol protease reagents (p-chloromercuriphenyl sulfonic acid, iodoacetic acid, and HgCl2) had a small inhibitory effect. The products formed by PCE activities in the membrane and soluble fractions were similar to those cleaved by the total granule lysate. The membrane fraction primarily cleaved proopiocortin between ACTH and beta LPH to form 21,000 (21 K) mol wt ACTH and beta-LPH, similar to the first processing step in the IL in situ. The soluble fraction, however, showed a greater tendency to cleave proopiocortin between the 16 K N-terminal glycopeptide and ACTH, to yield twice as much 21 K ACTH/LPH product as the membrane fraction. The membrane-associated PCE activity was found to be easily solubilized by extraction with high salt (1 M NaCl), suggesting that it is not an integral granule membrane protein.

The role of the carbohydrate in the stabilization, processing, and packaging of the glycosylated adrenocorticotropin-endorphin common precursor in toad pituitaries.

The neurointermediate lobes of dark adapted toads, Xenopus laevis, were incubated for 30 min in [3H]arginine, [3H]arginine plus [14C]glucosamine, or [3H]glucosamine and then chased for various time periods ranging from 1--3 h. The labeled polypeptides synthesized and secreted by the lobes were analyzed by acid-urea polyacrylamide gel electrophoresis. A glycosylated ACTH-endorphin precursor (32,000 mol wt) was synthesized during the pulse and identified by immunoprecipitation by ACTH-(11--24) antiserum. During the chase, this precursor was processed to various glycopeptides and peptides, including ACTH, beta-lipotropin, and alpha-MSH, which were subsequently secreted into the medium. An immunoprecipitable ACTH-related glycoprotein (approximately 150,000 mol wt) and other nonimmunoprecipitable glycoproteins (approximately 80,000--100,000 mol wt) were also synthesized and secreted by the neurointermediate lobe. The secretion of these glycoproteins and peptides was inhibited by dopamine. The significance of glycosylation of the precursor for the biosynthesis, processing, and secretion of the ACTH, beta-lipotropin-, and MSH-related peptides was examined by using a specific inhibitor of glycosylation, tunicamycin. Tunicamycin treatment did not affect the synthesis of the 32,000 mol wt ACTH-endorphin precursor but did prevent its glycosylation. The absence of carbohydrate on the precursor resulted in its rapid intracellular degradation. Precursors that escaped degradation were processed incompletely, leading to the formation and secretion of an unglycosylated intermediate and various other abnormal peptides. The data indicate that glycosylation of the ACTH-endorphin precursor may not be involved in the processes of intracellular transport, packaging, and secretion per se but, rather, may provide specific conformational stability to the precursor as a signal for directed limited proteolysis.

Depletion of carboxypeptidase E, a regulated secretory pathway sorting receptor, causes misrouting and constitutive secretion of proinsulin and proenkephalin, but not chromogranin A.

Previous studies have shown that the prohormone POMC is sorted to the regulated secretory pathway (RSP), at the trans-Golgi network, by binding of a conformation-dependent sorting signal to a sorting receptor, identified as membrane-bound carboxypeptidase E (CPE) (Cool et al., 1997, Cell, 88:73-83). In this study, the role of CPE as a sorting receptor for other RSP proteins that contain sorting signals (proinsulin, proenkephalin, and chromogranin A) was investigated in neuroendocrine cells (Neuro-2a) stably expressing CPE antisense RNA. Whereas these cells were depleted of CPE by greater than 85%, electron microscopy showed that they contain dense core secretory granules identical to wild-type Neuro-2a cells, indicating that CPE is not essential for granulogenesis. Secretion and immunocytochemical studies showed that, in wild-type Neuro-2a cells, endogenous proenkephalin and transfected proinsulin/insulin were localized to punctate secretory granules and were released via the RSP. However, in CPE-depleted cells, these two prohormones were released constitutively and had a Golgi-like distribution but were not localized to punctate secretory granules. In contrast, chromogranin A was present in punctate secretory granules and released via the RSP, in wild-type and CPE-depleted Neuro-2a cells. Thus, the sorting of proinsulin and proenkephalin to the RSP, like POMC, necessitates binding to CPE, and hence, CPE acts as a common sorting receptor for targeting these prohormones to the RSP. In contrast, the sorting signal of chromogranin A does not use CPE as a sorting receptor, suggesting the existence of other sorting receptors for the RSP.

Characterization of proopiocortin converting activity in rat anterior pituitary secretory granules.

Lysates from purified secretory granules of rat anterior pituitary glands were incubated with [3H]phenylalanine or [3H]arginine-labeled toad proopiocortin. The processed products formed were identified by immunoprecipitation with ACTH and beta-endorphin antisera, and by comigration with known markers on acid-urea polyacrylamide gels. Proopiocortin was cleaved by the secretory granule lysate primarily to 21,000 mol wt ACTH, 13,000 mol wt ACTH, 16,000 mol wt NH2-terminal glycopeptide, beta-lipotropin, a beta-endorphin-like peptide, and beta-endorphin. Characterization of the anterior pituitary proopicortin-converting activity shows that it: (1) cleaves specifically at the peptide bond on the carboxy side of the lysine-arginine residues of proopiocortin, (2) has a pH optimum in the acidic range, (3) is present in membrane and soluble fractions of the granule lysate, and (4) is inhibited by leupeptin, pepstatin A, and 2,2' dithiodipyridine, but not by p-chloromercuribenzoate, diisopropyl fluorophosphate, N alpha-p-tosyl-L-lysine chloromethyl ketone hydrochloride, chloroquine, L-1-tosylamide-2-phenylethyl-chloromethyl ketone, or EDTA.

Immunological identification and localization of yeast aspartic protease 3-like prohormone-processing enzymes in mammalian brain and pituitary.

The novel aspartic proteases, yeast aspartic protease 3 and the mammalian POMC-converting enzyme (PCE), can process prohormones at specific basic residue cleavage sites. We show that an antibody against yeast aspartic protease 3 (YAP3p) cross-reacted with purified bovine PCE on Western blot, indicating structural homology between these two enzymes, but not with other aspartic proteases, such as renin or cathepsin D. A PCE-sized anti-YAP3p-immunoreactive band was detected on Western blots of bovine intermediate lobe where PCE activity has been found. YAP3p antiserum also cross-reacted with a protein of approximately 90 kDa from mouse hypothalamus and anterior pituitary, and bovine anterior pituitary secretory granules. Distribution studies showed the presence of anti-YAP3p-immunopositive cells in bovine pituitary and peptide-rich brain regions, including the mouse arcuate nucleus and hippocampus and the rat supraoptic nucleus, paraventricular nucleus, cortex, striatum, and reticular nucleus. In the bovine intermediate pituitary, a subpopulation of cells was intensely stained with the YAP3p antiserum, and in combination with in situ hybridization, these cells were shown to contain POMC messenger RNA (mRNA). Only a subpopulation of cells was immunopositive for anti-YAP3p in bovine anterior pituitary, and most of these cells were identified by double immunostaining with ACTH antiserum as corticotrophs. In situ hybridization in combination with immunocytochemistry provided evidence for the localization of arginine vasopressin mRNA in YAP3p-immunopositive neurons in the rat supraoptic nucleus, whereas cholecystokinin mRNA was detected in YAP3p-immunopositive cells in the rat cortex and hippocampus. These results support the hypothesis that YAP3p-like aspartic proteases, including PCE, play a role in prohormone processing in endocrine/neuroendocrine cells in vivo.

Yeast aspartic protease 3 is sorted to secretory granules and activated to process proopiomelanocortin in PC12 cells.

The subcellular localization and functionality of transfected yeast aspartic protease 3 (YAP3p) in a mammalian cell line were investigated. The complementary DNAs encoding the prohormone-processing enzyme (YAP3p) and a prohormone, bovine POMC, were cotransfected into PC12 (rat pheochromocytoma) cells. Immunocytochemical analysis of the cells using a YAP3p antibody showed a perinuclear punctate distribution of YAP3p in the cell body as well as immunostaining in the tips of the neurites. This pattern of immunostaining indicates localization of YAP3p in secretory granules. Analysis of the processing of POMC showed that in cells transfected with the POMC complementary DNA alone, only POMC was found, indicating a lack of processing of the prohormone. However, in cells coexpressing YAP3p, the POMC was completely processed to yield ACTH-(1-39) and ACTH-(1-14), consistent with the specificity of YAP3p found in vitro. Pulse-chase studies showed that POMC was processed after 20 min of chase, suggesting that processing occurred in the late Golgi network and continued in the secretory granules. Western blot analysis determined that YAP3p was secreted from the cells in a regulated manner. This study provides the first demonstration that a yeast prohormone-processing enzyme (YAP3p) of the aspartic protease class can be sorted correctly to secretory granules and activated to process a prohormone (POMC) in a highly efficient manner in mammalian cells.