Neuropeptides and Carboxypeptidase E/ Neurotrophic Factor-alpha1 in Neural and Cognitive Functions

Eunice Kennedy Shriver National Institute Of Child Health & Human Development

Investigators

Linked publications, trials & patents

Abstract

Our current focus is to study the neurotrophic functions of CPE/NF-alpha1. Humans with null and mis-sense CPE mutations have been reported to have severe learning disability in early childhood, obesity, diabetes and infertility, due to lack of CPE. We have identified a CPE mutation in an Alzheimer Disease (AD) patient which results in a CPE mutant protein with an additional nine amino acids, we named CPE-QQ. When expressed in Neuro2a cells, CPE-QQ was not secreted but degraded by proteosomes. Immunocytochemical studies showed CPE-QQ localized to the endoplasmic reticulum (ER) and overexpression in hippocampal neurons increased ER stress and decreased levels of pro-survival protein, BCL-2, resulting in neuronal cell death. Transgenic mice overexpressing CPE-QQ exhibited memory deficits in the Morris water maze test, but their spatial learning ability was unimpaired. These mice showed depressive-like behavior by the forced swim test. They had fewer neurites in the hippocampus and medial prefrontal cortex, indicative of neurodegeneration. They showed diminished neurogenesis in the sub-granular zone and hyperphosphorylation of tau, a hallmark of AD. Thus, this human mutation in CPE/NF-alpha1 is neurotoxic, leading to neurodegeneration and cognitive decline. Our in vitro studies showed that CPE/NF-alpha1 acts extracellularly, independent of its enzymatic activity to protect hippocampal neurons against oxidative stress via activation of the ERK- or AKT- pathways to up-regulate BCL-2 expression. The neuroprotective effect of CPE/NF-alpha1 was demonstrated in vivo using a transgenic knock-in mouse model expressing a non-enzymatic form of CPE/NF-alpha1, CPE-E342Q, but no WT form. CPE knock-out mice lacking CPE/NF-alpha1 showed complete degeneration of hippocampal CA3 neurons and cognitive dysfunction after social (maternal separation) and physical stress (ear tagging and tail clipping) associated with the weaning paradigm, despite having WT levels of other growth factors such as BDNF, GDNF, NGF and NT3. Treatment of WT mice with ANA12, an inhibitor of TrkB, the BDNF receptor did not result in neurodegeneration of the CA3 region after weaning stress. Most importantly, CPE-E342Q mice, although lacking WT-CPE and had endocrinological deficits, showed no hippocampal degeneration or cognitive dysfunction after the weaning paradigm, indicating that CPE/NF-alpha1 is critical, but not BDNF, in preventing stress-induced hippocampal cell death, independent of its enzymatic activity. Recently, we made a conditional CPE knock-out mouse using the Cam2a-cre system to specifically knock out expression of CPE in brain. Compared to WT mice, CPEflox/flox mice showed impaired learning and memory based on object recognition, Y-Maze and fear conditioning tests. Body weight and glucose were normal in these CPEflox/flox mice. However, they exhibited complete degeneration of the subiculum region, unlike the CPE full knockout mice, which exhibit hippocampal CA3 region neurodegeneration. In addition, reduced doublecortin immunostaining in the dentate gyrus of the hippocampus suggested decreased neurogenesis in CPEflox/flox mice. Thus, specific neuronal CPE knockout leads to central nervous system dysfunction in mice. To understand the mechanism of the trophic action of CPE/NF-a1, we searched for a membrane receptor for this factor. Screening a human G-protein coupled receptor (GPCR) library, we found a serotonin receptor, HTR1E with no known function that interacted with CPE/NF-alpha1. This interaction was confirmed by co-immunoprecipitation and pulldown assays. Binding studies revealed a Kd=13.82nM. Molecular dynamics studies indicated that CPE/NF-a1 interacts with HTR1E via 4 salt-bridges stabilized by several hydrogen bonds and is independent of the serotonin binding pocket. This theoretical model was recently confirmed by experimental data. Immunohistochemistry revealed co-localization of HTR1E and CPE/NF-a1 on the surface of hippocampal neurons. Signal transduction studies showed that HTR1E-CPE/NF-alpha1 interaction activated the Erk1/2-CREB pathway via recruitment of beta-arrestin. This in turn activated the BCL2 pro-survival pathway. We showed that HTR1E-CPE/NF-alpha1 interaction mediated neuroprotection of human primary neurons against H2O2 -induced cytotoxicity and glutamate-induced neurotoxicity [1]. These findings indicate that CPE/NF-alpha1 interacts with HTR1E to promote neuronal survival. Unlike CPE/NF-1, serotonin binds to a different site on 5-HTR1E and the serotonin-5-HTR1E complex activates ERK pathway via classical Gi and PKA to regulate specific biological functions, such as cell cycle and survival. Based on these studies, we have designed small molecules that mimic CPE and demonstrated that they have neuroprotective effects [2]. We have examined the role of CPE/NF-a1 in preventing restraint stress-induced depression. Prolonged (6h/d for 21 days), but not short-term (1h/d for 7d) restraint stress reduced fibroblast growth factor 2 (FGF2) in the hippocampus, leading to depressive-like behavior in mice. Mice after short-term restraint stress increased hippocampal NF-a1, FGF2 and doublecortin, a marker for immature neurons, suggesting increased neurogenesis. NF-a1 added to cultured hippocampal neurons, increased FGF2 expression. Moreover, CPE/NF-alpha1-KO mice exhibited severely reduced hippocampal FGF2 levels and immature neuron numbers in the sub-granular zone. These mice displayed depressive-like behavior that was rescued by FGF2 administration. Thus, CPE/NF-alpha1 prevents stress-induced depression by up-regulating hippocampal FGF2 expression which leads to enhanced neurogenesis and anti-depressant activity. Given CPE-NF-alpha1's role in preventing cell death, we investigated its use as a therapeutic agent for Alzheimer Disease (AD). We delivered viral-(AAV)-NF-al/CPE) gene into the hippocampus of 3xTg-AD mice , an AD model, at an early pre-symptomatic age ( 2 months) and found that it prevented later development of cognitive dysfunction at 7-8 months of age , as assessed by Morris water maze and novel object recognition assays. Neurodegeneration and tau hyperphosphorylation were prevented in AAV-NF-a1/CPE treated 3xTg-AD mice. Additionally, amyloid precursor protein (APP) expression was reduced to near non-AD levels, and insoluble Abeta1-42 was reduced significantly. Pro-survival proteins: mitochondrial Bcl2 and Serpina3g were increased; and mitophagy inhibitor Plin4 and pro-inflammatory protein Card14 were decreased in AAV-NF-a1/CPE treated AD mice. Autophagy markers indicate that autophagy was impaired in 3xTg-AD mice but was restored with AAV-NF-a1/CPE treatment. Thus NF-a1/CPE gene therapy targets many regulatory components to prevent development of cognitive deficits and AD pathology in 3xTg-AD mice [3]. Using a more severe AD mouse model, 5xFAD, we found that AAV-NF-a1/CPE treatment of these mice post-symptomatically resulted in marked decrease in amyloid plaque accumulation across both early and late stages of the disease (5, 7, and 9 months) and in a dose-dependent manner [4]. The overexpression of NF-α1/CPE mitigated spatial memory deficits and normalized hippocampal synaptogenesis and microglial cell anomalies in the 5xFAD mice. Taken together, our findings from both the 3xTg-AD and the 5xFAD mice implicate NF-a1/CPE gene therapy as a promising new approach for the treatment of AD. We also investigated the role of CPE/NFa-1 in embryonic neurodevelopment. Addition of CPE/NF-a1 to E13.5 neocortex-derived neurospheres, which contains stem cells and neuroprogenitors, resulted in reduced proliferation of the neurospheres without causing cell death. These CPE/NF-a1 treated neurospheres showed down-regulation of the wnt-beta catenin pathway known to promote proliferation. Differentiation studies using neurospheres dissociated into single cells showed an increase in astrocytes in the presence of NF-a1, without altering the percentage of neuronal and oligodendrocyte populations. Interestingly, dissociated cells from neurospheres derived from NF-a1-KO mouse embryos showed decreased astrocytes and increased neurons. NF-alpha1 KO mice had 49% fewer GFAP+ astrocytes in P1 brain. Thus NF-a1 plays a critical role in differentiating neural stem cells into astrocytes.