Huaxi Xu, Ph.D. - US grants

PROJECT SUMMARY SNX27 is a member of a large family of PX-domain proteins which mediate protein trafficking and sorting from the endosome. Highly enriched in human brain, SNX27 facilitates endosome to cell surface trafficking of numerous transmembrane signaling components; our work shows that SNX27 mediated surface distribution of the ionotropic AMPA receptor subunit GluR1 is impaired in Down?s syndrome (DS) through triplication of the miR-155 locus on human Chromosome 21 which attenuates SNX27 expression through a C/EBP?-dependent mechanism. Our work also indicates that SNX27 function is fundamental to proper ciliated ependymal layer formation within lateral brain ventricles, where deletion of SNX27 results in early-stage hydrocephalus. Interestingly, our preliminary data demonstrates that SNX27 signaling is critical in Alzheimer?s disease (AD) through the regulation of amyloidogenic processing of the Amyloid Precursor Protein (APP): SNX27 limits A? generation through impairment of the ?-secretase complex, and through enhancing endosome to cell surface trafficking of a SORLA/APP complex. Together, these results demonstrate that SNX27 is a critical signaling junction which can affect developmental and pathological outcome in various neurological disorders. As SNX27 function is impaired in these disorders, enhancing SNX27 function may be an effective means to restore brain dysfunction in DS, hydrocephalus and AD. The focus of this study is to characterize and exploit SNX27-dependent pathways to restore and repair dysfunctional SNX27 signaling/trafficking nodes in DS and AD. Similar to SNX27 deletion strains, DS mouse models also feature hydrocephalus (DS-hydrocephalus) and ventriculomegaly phenotypes. We will use DS- hydrocephalus as a model system to characterize SNX27-dependent hydrocephalus pathways, and determine whether SNX27 modulation can affect DS-hydrocephalus phenotypes. As we have previously shown that SNX27 haploinsufficiency induces A? accumulation in AD cell and animal models, we will determine whether SNX27 overexpression can have restorative effects in reducing amyloidogenic A? generation, restore synaptic/cognitive deficits in an AD background and DS background, and determine whether these effects are dependent on SORLA trafficking. As our preliminary results indicate that impairment of SNX27 function drives certain aspects of DS and AD pathogenesis, enhancing SNX27/target interactions will likely restore SNX27-associated impairment in DS and AD. We will therefore explore molecular mechanisms of SNX27 interactions with downstream pathological targets such as GluR1 and SORLA, and screen for small molecules that may enhance SNX27 interaction and trafficking of its targets. As SNX27-associated impairment may span a variety of neurological disorders (DS, hydrocephalus and AD), enhancers of SNX27 function may have restorative effects in a variety of neurodegenerative contexts. Results from this study may yield mechanistic insight into how SNX27 defects may influence DS and AD, and define strategies to restore SNX27 function in these disorders.

DESCRIPTION (provided by applicant): Dysregulated trafficking of the ?-amyloid precursor protein (APP) and glutamate receptors contributes to the pathogenesis/pathology of Alzheimer's disease (AD). The SorL1 protein interacts with APP and changes in SorL1 expression or function affects the subcellular distribution and thus the processing of APP. The retromer complex is composed of VPS35, VPS29, VPS26 and Sorting Nexins 1 and 2, and regulates retrograde sorting from the endosome to the trans-Golgi network (TGN). Deficits in retromer transport are implicated in sporadic AD but most studies propose a link between the retromer complex and ?-secretase (BACE1) transport. Although VPS35 and VPS26 interact with SorL1, how retromer regulates APP trafficking and processing is largely unknown. Sorting Nexin 27 (SNX27) belongs to the Sorting Nexin family whose members regulate protein endocytosis and recycling. Our recent findings published in Nature Medicine demonstrate that SNX27 directly mediates glutamate receptor sorting to the plasma membrane, and attenuated SNX27 expression in Down syndrome (DS) brain contributes to synaptic dysfunction and neurodegeneration. Significantly, we find that SNX27 re-expression can ameliorate neuropathological and behavioral phenotypes in a DS mouse model. We attempt to further establish molecular mechanisms underlying SNX27-mediated neurodegeneration through the trafficking and homeostasis of glutamate receptors and APP. Our preliminary results indicate that SNX27 interacts with VPS26 and VPS35, components of the retromer complex. In addition, we find that SNX27 regulated cell surface glutamate receptor levels in a retromer-dependent manner and downregulation of VPS26 or VPS35 abolished the effect of SNX27 overexpression on promoting cell surface levels of glutamate receptors. Additionally, we find that overexpression and downregulation of SNX27 reduced and increased A? levels, respectively, further suggesting that SNX27 can also regulate trafficking and processing of APP. Our results also show that SNX27-mediated APP trafficking potentially occurs through direct interactions with the APP trafficking receptor SorL1 since overexpression and downregulation of SNX27 can increase and decrease SorL1 levels, respectively. Therefore, we hypothesize that interactions between SNX27, SorL1 and the retromer complex play important roles in AD by coordinating APP endocytic and golgi retromer trafficking pathways that attenuate A? generation, and by maintaining proper cell surface homeostasis of glutamate receptors. In this proposal, we will ascertain whether SNX27, SorL1 and the retromer complex coordinately regulate APP trafficking/processing and cell surface homeostasis of glutamate receptors and synaptic function. Moreover, as deficiencies in VPS35 and SorL1 may accelerate onset of disease-like phenotypes in AD mice, we will study whether overexpression of SNX27 can delay early-onset neuropathological and cognitive phenotypes in AD mice bearing VPS35- or SorL1-haploinsufficiencies.

? DESCRIPTION (provided by applicant): Alzheimer's disease (AD) is defined by two major pathological hallmarks, specifically senile plaques consisting of ?-amyloid (A?) and neurofibrillary tangles composed of abnormally phosphorylated and cleaved tau. A? is neurotoxic and can trigger a cascade of neurodegenerative events in AD. Tau abnormalities also contribute to AD progression. In addition, abnormal tau phosphorylation, cleavage, and mutations in the tau gene, MAPT can induce tau aggregation and consequent toxicity in neurons, leading to several other neurodegenerative tauopathic disorders which include progressive supranuclear palsy (PSP), Pick's disease, corticobasal degeneration, frontotemporal dementia and Parkinsonism linked to chromosome 17 (FTDP-17). Hence, identification of new factors mediating neurotoxicity caused by A? and/or tau is important for disease intervention. We recently identified a novel pro-apoptotic protein, appoptosin, and demonstrated that overexpression of appoptosion results in caspase-dependent apoptosis. Importantly, we found that appoptosin levels are elevated in neurons exposed to A? and [excitotoxic] glutamate stimulation, with increased levels in brain samples from AD and PSP patients. Increased appoptosin expression leads to caspase-mediated tau cleavage and concomitant tau aggregation and synaptic dysfunction in neurons. Appoptosin transduction impairs motor function and exacerbates neuropathology in tau Tg mice; whereas reduced expression of appoptosin inhibits tau cleavage and aggregation, and abrogates mitochondrial fragmentation, caspase activation and neuronal death caused by A? insults. Appoptosin+/- mice have normal learning/memory and LTP, but show reduced LTD and slowed memory decay, thereby implicating its involvement in synaptic plasticity. Moreover, a single nucleotide polymorphism (SNP) rs1768208(C/T) near the appoptosin gene was reported as a risk factor for PSP, AD, CBD and frontotemporal dementia and we demonstrate that the T-allele variant occurs much more frequently in PSP and correlates tightly with increased appoptosin expression. Therefore, we hypothesize that an upregulation of appoptosin expression induced by A? in AD, or controlled by the SNP rs1768208 in various tauopathies, plays a central role in inducing neurodegeneration, and that downregulation of appoptosin provides a novel strategy for disease intervention. In this proposal, we will further ascertain that appoptosin SNP and expression are associated with tauopathic diseases, and determine that the SNP rs1768208(C/T) variant regulates appoptosin expression. We will then determine whether caspases, tau and glutamate receptor dynamics are involved in appoptosin-mediated synaptic plasticity. Finally, we will corroborate the role of appoptosin in AD and other tauopathies in vivo by studying whether upregulation of appoptosin leads to tauopathy-related neuropathologies and behavior, and whether a decrease of appoptosin can ameliorate disease-related phenotypes in APP/tau bigenic mice. Results from these studies will establish appoptosin as a novel and important player and therapeutic target in AD and other tauopathies.

PROJECT SUMMARY Alzheimer?s Disease (AD) is the most common form of dementia, characterized by misfolding and aggregation of specific proteins which manifest in pathological features including neuronal loss, cognitive decline and histopathological hallmarks such as the accumulation of amyloid plaques and neurofibrillary tangles in the brain. Amyloid plaques comprise extracellular deposits of amyloid-b (Ab) aggregates, where A? oligomers are thought to be proteotoxic to neuronal function. Numerous studies have demonstrated that AD-associated proteotoxicity triggers an adaptive unfolded protein response (UPR) which attempts to restore proteostatic dysfunction due to accumulation of misfolded proteins in the endoplasmic reticulum (ER). UPR signaling is mediated through PERK/eIF2a, IRE1/XBP1, and ATF6 sensor pathways; relative contributions of these signaling arms to neurodegeneration is complex as they have dual roles in mediating cell survival and cell death. Elevations in PERK/phosphorylated eIF2a, XBP1 mRNA splicing, and increased levels of ER chaperones such as BiP/GRP78, GRP94 and PDI in human AD brain strongly suggests chronic activation of ER stress is evident in human AD pathology. Further, a polymorphism previously linked to bipolar disorders within the XBP1 promoter region was linked to increased AD risk in Chinese populations. Although it has been established that UPR pathways are activated in disorders such as AD, it is not clear whether UPR pathways confer neuroprotective effects, or if their activation can contribute to pathogenesis. Potential neuroprotective effects of the ATF6 pathway in neurodegeneration have remained particularly elusive. ATF6 functions as an ER stress sensor and transcription factor that promotes expression of genes that enhance proper protein folding via increased production of ER chaperones and increased degradation of misfolded proteins. We present new evidence that ATF6 is essential for synaptic function, as Atf6-/- mice display cognitive and behavioral defects associated with reduced synaptic spine density. Further, exogenous expression of the active ATF6 form suppressed accumulation of amyloid fibrils in a murine model of AD. These results provide strong evidence that ATF6 activation may have a physiological role in synaptic activity and cognitive behavior, and acute ATF6 activation can confer neuroprotective effects with AD-associated proteotoxicity. In the proposed study, our efforts will be focused on elucidating potential neuroprotective effects of ATF6 on neuronal and synaptic function, and differentiate potential roles for ATF6 in neurons and microglia. Given the effects of ATF6 on attenuating A? plaque formation, we will characterize proteins that are particularly susceptible to proteostatic dysfunction in AD. We will also characterize the effects of ATF6 mutational variants identified in human Achromatopsia patients on neuronal function, and determine whether pharmacological activation of ATF6 is protective in AD. The sum of these results will implicate a novel role for ATF6 in neuronal/synaptic function, and provide insight into potential strategies to reverse synaptic impairment through enhancing UPR function.

PROJECT SUMMARY Proteostatic regulatory mechanisms in the endoplasmic reticulum (ER) such as the ER-associated degradation (ERAD) and unfolded protein response (UPR) systems are commonly dysregulated in neurodegenerative proteinopathies such as Alzheimer?s disease (AD). We identify a new role for the ER-resident component membralin in ERAD function. Interestingly, membralin polymorphisms and splice variation have been implicated in AD, and our results indicate that membralin expression is reduced in AD brain. We also identify nicastrin as a membralin/ERAD substrate, and membralin downregulation results in elevated nicastrin levels, thereby enhancing ?-secretase activity, A? generation and cognitive impairment. Interestingly, we also identify ERAD components in complex with the ?-secretase activating protein (GSAP) by proteomic analysis and demonstrate that membralin and GSAP interact by coimmunoprecipitation. Although GSAP is thought to be pathogenic due to its role in enhancing A? generation, we unexpectedly find that GSAP KO mice show impaired fear memory and synaptic function. Moreover, synaptic response and long-term potentiation in with GSAP deletion is further impaired with A?, thereby implicating a protective ?-secretase-independent role for GSAP in synaptic function. Together, our results indicate that membralin/ERAD and GSAP physically and functionally interact, and that these two components likely support both ?-secretase dependent and independent modes of neuroprotection. Membralin splice variation has been previously characterized in AD, however, how this potentially impacts membralin/ERAD function is unclear. We will therefore characterize expression of long and short membralin isoforms in human AD brain and determine whether C-terminal truncation can impact membralin-associated ERAD function in the short isoform. Given that homozygous membralin deletion is lethal, while heterozygous membralin deletion yields no visible phenotype, it seems likely that membralin haploinsufficiency may induce compensatory changes in the ERAD complex, or associated UPR system to normalize proteostatic function. We will characterize changes in ERAD/UPR systems with membralin downregulation in the absence and presence of A? and determine whether enhancing these compensatory changes can enhance ERAD function. GSAP has been previously known to be processed into N- and C-terminal fragments (NTF/CTFs) through caspase-3 cleavage; we will determine whether full-length GSAP or its cleaved forms are involved in GSAP-dependent synaptic function by overexpression/complementation in GSAP KO mouse hippocampus. Since our previous results indicate that membralin depletion can induce caspase-3 activation, we will determine whether modulating membralin levels can affect GSAP-dependent synaptic function and cognitive behavior. Lastly, we will establish an AD/membralin haploinsufficiency model to determine whether enhancing membralin/ERAD, UPR, or GSAP pathways can ameliorate cognitive, synaptic and pathological AD defects.