Skip Internal Navigation
We study the cell biology of endocrine and neuroendocrine cells. Our focus is three-fold: (i) investigate the mechanisms of biosynthesis and intracellular trafficking of peptide hormones and neuropeptides and their processing enzymes; (ii) uncover mechanisms involved in the regulation of dense-core secretory granule biogenesis, transport, and exocytosis; and (iii) determine the physiological and pathological roles of the prohormone-processing enzyme carboxypeptidase E (CPE). Our work has led to the discovery of novel molecular mechanisms of protein trafficking to the regulated secretory pathway and identified players and mechanisms that control secretory granule biogenesis and transport in endocrine and neuroendocrine cells as well as uncovered new roles of carboxypeptidase E gene in neuroprotection, dendritic pruning, and cancer. Using cell lines, primary cell cultures, and mouse models, such studies have provided a better understanding of diseases related to defects in hormone and neuropeptide targeting, synaptic transmission, neurodegeneration, memory, learning, diabetes, obesity, and metastatic disease.
Mechanism of sorting pro-neuropeptides, neurotrophins, and their processing enzymes to the regulated secretory pathway
The intracellular sorting of pro-neuropeptides and neurotrophins to the regulated secretory pathway (RSP) is essential for processing, storage, and release of active proteins and peptides in the neuroendocrine cell. We investigated the sorting of pro-opiomelanocortin (POMC, pro-ACTH/endorphin), proinsulin, and brain-derived neurotrophic factor (BDNF) to the RSP. Our studies showed that, as a concentration step, these pro-proteins undergo homotypic oligomerization as they traverse the cell from the site of synthesis in the endoplasmic reticulum (ER) to the trans-Golgi network (TGN). In the TGN, they are sorted into dense-core granules of the RSP for processing by prohormone convertases and CPE and then secreted. We showed that the sorting of prohormones to the RSP occurs by a receptor-mediated mechanism. Site-direct mutagenesis studies have identified a 3-D consensus sorting motif consisting of two acidic residues found in POMC, proinsulin, and BDNF. We identified an RSP sorting receptor that is specific for the sorting signal of these proproteins as the transmembrane form of CPE.
We investigated the role of membrane CPE and secretogranin III as sorting receptors for targeting POMC to the RSP. Using our CPE knockout (KO) mouse, we showed that 50% of newly synthesized POMC in primary cultures of the pituitary anterior lobe cells was degraded, which suggests that, in the absence of efficient sorting to the granules of the RSP owing to the lack of CPE, POMC was targeted for degradation. However, some of the remaining POMC was sorted into the RSP. A candidate for a compensatory sorting receptor is Secretogranin III (SgIII), which has been shown to bind to POMC in precipitation assays. SgIII is a member of the granins that are found in neuroendocrine cells and is involved in trafficking of chromogranin A (CgA) to the RSP. We used RNA interference (siRNA) to knock down SgIII and CPE expression in AtT20 cells and demonstrated in both cases that POMC secretion via the constitutive secretory pathway was elevated. However, we only observed increased constitutive secretion of CgA in the SgIII knock-down cells. In double CPE-SgIII knock-down cells, we observed elevated constitutive secretion of POMC and that stimulated secretion of ACTH was perturbed. The results demonstrate that CPE is involved in the trafficking of POMC to the RSP and that SgIII may play a compensatory role for CPE in the sorting of POMC to the RSP, in addition to its more general role in the RSP trafficking process.
In collaboration with Bruce Baum, we also investigated the secretory behavior of peptide hormones in the exocrine salivary gland. Given that the salivary gland secretes proteins into the upper GI tract via the RSP and into the circulation via the constitutive secretory pathway, it is a target tissue for the expression of proteins for gene therapy. Glucagon-like peptide 1 (GLP-1) is an incretin peptide hormone synthesized in intestinal L cells of the gut. It functions to stimulate insulin secretion in a glucose-dependent manner, it slows gastric emptying, and increases β-cell mass. We engineered an adenovirus expressing GLP-1 to transduce murine submandibular salivary gland cells. The transgenic GLP-1 was expressed and secreted into the circulation, where it functioned in an animal model of induced diabetes; mice transduced with the Ad-GLP-1 were protected against the onset of diabetes whereas mice treated with an Ad-Luciferase control virus were not. Thus, this method of gene therapy for the production of GLP-1 may, in the future, be useful for the treatment of Type 2 diabetes.
Identification of a novel trans-Golgi network (TGN) sub-compartment for synaptic vesicle and large dense core vesicle protein sorting
Neuroendocrine cells and peptidergic neurons contain synaptic vesicles, which include such proteins as vesicular acetylcholine transporter (VAChT) and synaptophysin (SYN), and large dense-core vesicles (LDCVs) vesicles, which include neuropeptides. Proteins destined for these different vesicles are synthesized in the ER and trafficked to the TGN, where they are differentially sorted into specific transport vesicles. The small transport vesicles are delivered constitutively to the active zone at the presynaptic terminus or plasma membrane (PM) and recycled for the production of synaptic vesicles. Neuropeptides, however, are sorted to LDCVs, which are also delivered to the release site but are not fused to the presynaptic membrane or PM under resting conditions. A long-standing question was whether synaptic vesicle and LDCV proteins use clearly distinct sorting compartments as well as transport routes from the Golgi complex to the presynaptic membrane or PM.
In earlier work, we found that depletion by antisense RNA of CgA (a cargo protein of LDCVs) from PC12 cells resulted in significantly reduced levels of VAChT, a cargo protein of synaptic-like microvesicles (SLMV). Given that we had previously shown that CgA mediates LDCV biogenesis and had subsequently identified the responsible fragment of CgA, we hypothesized that a common pathway beyond the TGN existed between the biogenesis of VAChT–containing SLMVs and LDCVs. We investigated the subcellular trafficking pathway and biogenesis of VAChT– and SYN–containing SLMVs from the TGN and their relationship with CgA–containing LDCVs. This was accomplished by transfecting PC12 cells with CgA conjugated to EGFP or VAChT and SYN each conjugated to mRFP (1). We blocked post-Golgi trafficking at 20oC overnight and then released it by raising the temperature to 37oC. We analyzed subsequent movement of fluorescent CgA and VAChT or SYN, both by fixed immunocytochemistry, and live-cell confocal microscopy. VAChT and CgA initially co-localized in the Golgi compartments identified by p115 marker staining. At 15 min, VAChT was localized to a large, immobile, p115–negative compartment containing Golgin97 and TGN46 but not TGN38, from which smaller vesicles positive for VAChT, but not for CgA, budded. At 45 min after release from the 20oC block, this compartment persisted, but many smaller vesicles were evident in the cytoplasm. Additionally, at the PM and in highly mobile vesicles, VAChT was primarily present in vesicles separate from those bearing CgA. These data indicate that synaptic-vesicle and LDCV proteins share a common, previously uncharacterized, sub-compartment of the TGN, giving rise to a common trafficking pathway for SLMV and LDCV in PC12 cells. Thus, VAChT appears to be trafficked and sorted from the TGN into this new compartment, where it is actively sorted into smaller vesicles (possibly constitutive-like vesicles). The vesicles then fuse to the active zone in the synaptic terminal, and VAChT is subsequently internalized into SLMV.
Role of CPE in obesity, bone remodeling, neuroprotection, and stress
CPE plays a significant role in obesity, and recently the gene has been coined an obesity-susceptibility gene. We showed that CPE–KO mice were not able to process pro-CART to CART and therefore lacked this anorexigenic neuropeptide in the hypothalamus. The animals over-eat and become obese, thus providing further evidence linking loss of this neuropeptide to the cause of obesity. Additionally, in collaboration with the Accili group, we found that the transcription factor FoxO1 negatively regulates CPE gene expression. Normally insulin binds to insulin receptors in POMC neurons, which leads to nuclear signaling, nuclear exclusion, and inactivation of FoxO1. To model this physiological event, we deleted FoxO1 in the POMC neurons in the arcuate nucleus of the hypothalamus in mice, which resulted in increased CPE levels, increased alpha-MSH—an anorexigenic neuropeptide derived from POMC—and reduced food intake without change in energy expenditure. These findings raise the possibility of targeting CPE to develop weight-loss medications. We also showed that extremely obese CPE–KO mice have low bone mineral density and concluded that the lack of CART, which promotes bone formation, plays an important role in the poor bone density of the mice.
CPE–KO mice exhibit nervous system deficiencies. Morris–water maze and object-preference tests indicate defects in learning and memory. We showed that in 6- to14-week-old CPE–KO mice, dendritic pruning was poor in cortical and hippocampal neurons, which would affect synaptogeneis. Additionally, electrophysiological measurements revealed a defect in the generation of long-term potentiation (LTP) in hippocampal slices of these mice. A major cause for this defect was the loss of neurons in the CA3 region of the hippocampus of CPE–KO animals observed at 4 weeks of age and older. These neurons, which are normally enriched in CPE, were normal at 3 weeks of age just before the animals were weaned. Interestingly, when weaning was delayed a week, this degeneration was not observed until postnatal week 5 in the CPE–KO mice. The results suggest that the degeneration is correlated with the stress of weaning and maternal separation and that CPE is important for the survival of CA3 neurons during that period. Indeed, we showed that, when CPE was overexpressed in hippocampal neurons in culture, the neurons were protected from apoptosis after inducing oxidative stress with hydrogen peroxide. Thus, CPE has a novel neuroprotective role in hippocampal neurons.
CPE mediates peptidergic vesicle transport and synaptic vesicle localization to the pre-active zone for secretion.
Transport of vesicles containing hormone or BDNF to the plasma membrane for activity-dependent secretion is critical for endocrine function and synaptic plasticity. We showed that the cytoplasmic tail of a transmembrane form of CPE in hormone- or BDNF–containing dense-core secretory vesicles plays an important role in their transport to the vesicles' release site. To compete with the endogenous tail, we overexpressed the CPE cytoplasmic tail in the cytoplasm, which diminished localization of endogenous POMC, BDNF, and fluorescence-tagged CPE in the processes of an endocrine cell line, AtT20, and of hippocampal neurons. In hippocampal neurons, primary pituitary, and AtT20 cells, overexpression of the CPE tail inhibited the movement of BDNF– and POMC/CPE–containing vesicles to the processes, respectively. S-tagged CPE tail pulled down the microtubule-based motors dynactin (p150), dynein, and KIF1A/KIF3A from cytosol of AtT20 and brain cells. Finally, overexpression of the CPE tail inhibited the regulated secretion of ACTH from AtT20 cells. We also showed that the CPE tail interacted with the C-terminus of gamma-adducin, a component of the cytoskeleton that binds and stabilizes F-actin. Overexpression of the C-terminal 38 amino acids of gamma-adducin inhibited the transport of POMC vesicles out of the cell body into the processes of AtT-20 cells. Our studies demonstrate that the vesicular CPE cytoplasmic tail plays a novel mechanistic role in anchoring regulated secretory vesicles containing POMC/ACTH and BDNF to actin via gamma-adducin for actin-based movement, immediately after budding from the TGN. The vesicles subsequently move to the microtubule-based motor system for transport along the processes to the plasma membrane for activity-dependent secretion in endocrine cells and neurons.
We recently found that transmembrane CPE is not only associated with LDCVs but also with glutamate-containing synaptic vesicles (SVs) in the mouse hypothalamus and with synaptic-like microvesicles in PC12 cells. High K+–stimulated release of glutamate from hypothalamic neurons was diminished in CPE-KO mice. Electron microscopy revealed that the number of SVs located in the pre-active zone (within 200 nm of the plasma membrane at the active zone) of synapses was significantly lower in hypothalamic neurons of CPE-KO mice than in wild-type mice. Total internal reflective fluorescence (TIRF) microscopy using PC12 cells as a model showed that overexpression of the CPE cytoplasmic tail reduced the steady-state level of synaptophysin-containing synaptic-like microvesicles that accumulated in the area within 200 nm from the sub-plasma membrane (TIRF zone). Our findings show that the CPE cytoplasmic tail, which interacts with gamma adduccin and actin, is a newly discovered mediator for the localization of SVs in the actin-rich pre-active zone in hypothalamic neurons and the TIRF zone of PC12 cells.
Regulation of secretory granule biogenesis by chromogranin A and serpinin
Our recent studies in pituitary AtT-20 cells provided evidence for an autocrine mechanism for up-regulating LDCV biogenesis to replenish LDCVs following stimulated exocytosis of the vesicles. The autocrine signal was identified as serpinin, a novel 26 amino-acid CgA–derived peptide cleaved from the C-terminal of CgA (2). Serpinin was first isolated from AtT20 cell–conditioned medium and demonstrated to be released in an activity-dependent manner from LDCVs. Subsequently, secreted serpinin was found to activate adenyl cyclase, to raise cAMP levels, and protein kinase A in the cell. Activation led to the translocation of the transcription factor sp1 from the cytoplasm into the nucleus and enhanced transcription of a protease inhibitor, protease nexin 1 (PN-1), which then inhibited granule protein degradation in the Golgi complex. Stabilization of the granule proteins increased their levels in the Golgi, resulting in significantly enhanced LDCV formation. We also identified, by mass spectroscopy, a modified form of serpinin, pyroglutamate-serpinin (pGlu-serpinin), in secretion medium of pituitary AtT20 cells and rat heart tissue (3). pGlu-serpinin is synthesized and stored in secretory granules and secreted in an activity-dependent manner from AtT20 cells. pGlu-serpinin immunostaining has been observed in nerve terminals of neurites in mouse brain and found to exhibit neuroprotective activity against oxidative stress in AtT20 cells and against low K+–induced apoptosis in rat cortical neurons.
A splice isoform of carboxypeptidase E is a tumor inducer and biomarker for predicting future metastasis.
Despite the numerous biomarkers reported, few are useful for predicting metastasis. In our recent studies (5), we discovered a novel splice isoform of the prohormone processing enzyme carboxypeptidase E (CPE-deltaN) that is elevated in metastatic hepatocellular, colon, breast, head, and neck carcinoma cells. CPE-deltaN lacks the N-terminus normally present in secretory granule wild-type CPE and is localized to the nucleus of metastatic cancer cells. Overexpression of CPE-deltaN in hepatocellular carcinoma (HCC) cells promoted their proliferation and migration by up-regulating expression of a metastasis gene via epigenetic mechanisms. SiRNA knockdown of CPE-deltaN expression in highly metastatic HCC cells inhibited their growth and metastasis in nude mice. In retrospective clinical studies, CPE-deltaN mRNA quantification in primary HCC from patients established a cut-off level; levels above this predicted future metastasis within 2 years with high sensitivity and specificity and independent of cancer stage. Furthermore, in a prospective clinical study on patients with pheochromocytoma/paragangliomas, we were able to predict with high accuracy from the mRNA copy numbers of CPE-deltaN in the resected tumors those patients who would develop future metastasis, although they were diagnosed with benign tumors at the time of surgery. Additionally, we showed that CPE-delta N is an excellent biomarker for diagnosis of metastatic colorectal cancer. Thus, CPE-deltaN is a novel tumor inducer and a powerful prognostic marker for predicting future metastasis in various cancers, superior to histopathological diagnosis.