Biology subjects

Small, S. A.

Publications and source records attributed to Small, S. A..

10 recordsLinked to original sources

bioRxiv2026

Disrupting SORL1-retromer endosomal recycling in mice drives the age-dependent neurodegenerative signature of Alzheimer's disease

MRI studies have mapped the atrophic signature that distinguishes Alzheimer's disease (AD), isolating regions differentially vulnerable to its age-dependent neurodegenerative process. An endosomal recycling pathway regulated by the interaction of SORL1 with retromer is pathogenically disrupted in AD. Here we test the hypothesis, that the pathway might drive the neurodegenerative process, via multiple yearlong studies in over 140 wildtype and SORL1 deficient mice, modeling SORL1's causative mutations. MRI studies revealed that SORL1 deficiency induced an age-dependent AD-pattern of atrophy, centered on the trans-entorhinal cortex, and that atrophy was prevented by SORL1 restoration, secondary to overexpressing the retromer protein VPS26b. Postmortem studies validated that the atrophy reflects neuronal loss. Proteomic profiling revealed that the atrophy is associated with reductions in cell death suppressors, one of which was then found deficient in the trans-entorhinal cortex of AD patients. Besides supporting the hypothesis, when taken together with other findings the results suggest that SORL1-retromer disruptions represent a therapeutically-amenable common 'neurodegeneration pathway' in AD.

neuroscience↗

bioRxiv2026

RhoGEF12 regulates endosomal SORL1-retromer and its inhibition is therapeutic in human neuronal models of Alzheimer's disease

The interaction of the endosomal sorting protein SORL1 with the retromer complex at endosomal membranes controls a recycling pathway whose dysfunction is pathogenic in Alzheimers disease (AD) and is linked to other neurodegenerative disorders. To search for novel therapeutic targets, we hypothesize that endosomal SORL1-retromer might be regulated by SORL1s cytoplasmic tail. We begin by completing an in vitro analysis of the tail and show that its phosphorylation by ROCK2 (Rho-associated kinase 2) reduces SORL1s affinity to retromer. Since RhoGEF12 (Rho guanine nucleotide exchange factor 12) is an upstream activator of ROCK2 that is upregulated in AD, we used a RhoGEF12 pharmacological inhibitor to mechanistically and therapeutically validate the findings in neuronal cultures. First, in mouse neurons we confirm that the inhibitor increases endosomal SORL1-retromer. Next, we turned to human iPSC-derived neurons to show that the inhibitor reduces A{beta}40 and A{beta}42, an indicator of pathway upregulation, in a SORL1-dependent manner. Finally, we validate its therapeutic potential by applying the RhoGEF12 inhibitor to human iPSC-derived neurons expressing AD-associated mutations in either APP or SORL1. Collectively, our results identify a novel and therapeutically amenable mechanism that regulates endosomal SORL1-retromer and preclinically validate RhoGEF12 as a therapeutic target for AD and potentially other neurodegenerative disorders. One sentence summaryPharmacological inhibition of RhoGEF12 increases endosomal SORL1-retromer recycling and reduces pathogenic amyloid secretion in human neuronal models, identifying a novel, targetable pathway for treating Alzheimers disease.

neuroscience↗

bioRxiv2021

In vivo evidence that SORL1, encoding the endosomal recycling receptor SORLA, can function as a causal gene in Alzheimer's Disease

The few established causal genes in Alzheimers disease (AD), mutations in APP and PSENs, have been functionally characterized using biomarkers, capturing an in vivo profile reflecting the diseases initial preclinical phase. SORL1, a gene encoding the endosome recycling receptor SORLA, epidemiologically behaves as a causal gene when truncating mutations lead to partial loss of protein function. Here, in an effort to test whether SORL1 can indeed function as an AD causal gene, we used CRISPR-Cas9-based gene editing to develop a novel model of SORL1 haploinsufficiency in Gottingen Minipigs taking advantage of porcine models for biomarker investigations. SORL1 haploinsufficiency in young minipigs was found to phenocopy the preclinical in vivo profile of AD observed with other causal genes, resulting in spinal fluid abnormalities in A{beta} and tau, with no evident neurodegeneration or amyloid plaque formation. These studies provide functional support that SORL1 is a bona fide causal gene in AD, and when taken together with recent insight on other AD-causal genes, support the idea that dysfunctional endosomal recycling is a dominant pathogenic pathway in the disease.

neuroscience↗

bioRxiv2021

Magnetic Resonance Spectroscopy Frequency and Phase Correction Using Convolutional Neural Networks

We introduce DeepSPEC, a novel convolutional neural network (CNN) -based approach for frequency-and-phase correction (FPC) of MRS spectra to achieve fast and accurate FPC of single-voxel PRESS MRS and MEGA-PRESS data. In DeepSPEC, two neural networks, including one for frequency correction and one for phase correction were trained and validated using published simulated and in vivo PRESS and MEGA-PRESS MRS dataset with wide-range artificial frequency and phase offsets applied. DeepSPEC was subsequently tested and compared to the current deep learning solution - a "vanilla" neural network approach using multilayer perceptrons (MLP). Furthermore, random noise was added to the original simulated dataset to further investigate the model performance with noise at varied signal-to-noise (SNR) levels (i.e., 6 dB, 3 dB, and 1.5 dB). The testing showed that DeepSPEC is more robust to noise compared to the MLP-based approach due to having a smaller absolute error in both frequency and phase offset prediction. The DeepSPEC framework was capable of correcting frequency offset with 0.01{+/-}0.01 Hz and phase offset with 0.12{+/-}0.09{degrees} absolute errors on average for unseen simulated data at a high SNR (12 dB) and correcting frequency offset with 0.01{+/-}0.02 Hz and phase offset within -0.07{+/-}0.44{degrees} absolute errors on average at very low SNR (1.5 dB). Furthermore, additional frequency and phase offsets (i.e., small, moderate, large) were applied to the in vivo dataset, and DeepSPEC demonstrated better performance for FPC when compared to the MLP-based approach. Results also show DeepSPEC has superior performance than the model-based SR implementation (mSR) in FPC by having higher accuracy in a wider range of additional offsets. These results represent a proof of concept for the use of CNNs for preprocessing MRS data and demonstrate that DeepSPEC accurately predicts frequency and phase offsets at varying noise levels with state-of-the-art performance.

neuroscience↗

bioRxiv2020

Retromer Combinatorials for Gene-Therapy Across a Spectrum of Neurological Diseases

Endosomal trafficking is a biological pathway implicated in Alzheimers and Parkinsons disease, and a growing number of other neurological disorders. For this category of diseases, the endosomes trafficking complex retromer has emerged as a validated therapeutic target. Retromers core is a heterotrimeric complex composed of the scaffold protein VPS35 to which VPS26 and VPS29 bind. Unless it is deficient, increasing expression of VPS35 by viral vectors has a limited effect on other trimeric members and on retromers overall function. Here we set out to address these constraints and, based on prior insight, hypothesized that co-expressing VPS35 and VPS26 would synergistically interact and elevate retromers trimeric expression and function. Neurons, however, are distinct in expressing two VPS26 paralogs, VPS26a and VPS26b, and so to test the hypothesis we generated three novel AAV9 vectors harboring the VPS35, or VPS26a, or VPS26b transgene. First, we optimized their expression in neuroblastoma cell lines, then, in a comprehensive series of neuronal culture experiments, we expressed VPS35, VPS26a, and VPS26b individually and in all possible combinations. Confirming our hypothesis, expressing individual proteins failed to affect the trimer, while VPS35 and VPS26 combinatorials synergized the trimers expression. In addition, we illustrate functional synergy by showing that only VPS35 and VPS26 combinatorials significantly increase levels of Sorl1, a key retromer-receptor deficient in Alzheimers disease. Collectively, and together with other recent observations, these results suggest a precision-medicine logic when applying retromer gene therapy to a host of neurological disorders, depending on each disorders specific retromer-related molecular and anatomical phenotype.

neuroscience↗

bioRxiv2020

Endosomal trafficking is required for glycosylation and normal maturation of the Alzheimer's-associated protein sorLA

The sorting receptor sorLA encoded by the SORL1 gene is implicated in Alzheimers disease (AD) pathogenesis. Genetic studies have identified AD-associated SORL1 mutations and the expression of sorLA in AD brains is reported to be reduced. SorLA is a receptor of the retromer trafficking complex and functions at the endosome, and deficiency in sorLA phenocopies the endosomal pathologies found in AD. SorLA undergoes posttranslational modifications and maturation with ultimate ectodomain shedding, however knowledge of these processes remains limited. Here we demonstrate that sorLA exists at the cell membrane in two forms, an immature and a mature form, characterized by distinct N-glycosylation profiles. The mature sorLA form has acquired complex type N-glycans and is shed from the cell surface by the TACE juxtmembrane cleavage. The immature form of sorLA present at the cell surface is shown to have immature ER-type N-glycans (high-mannose type susceptible to endo H) and does not undergo shedding, however, upon endocytosis and recycling to the cell surface via endosomal trafficking pathways the immature sorLA form acquires complex-type N-glycans. These results suggest an unusual secretion model for sorLA whereby that immature sorLA first traffics to the cell membrane without acquiring Golgi processing of N-glycans, and only upon retrograde trafficking does sorLA acquire normal Golgi maturation of N-glycans and become susceptible to TACE regulated shedding. Supportive evidence for this model include a sorLA mutant with deficient endosomal trafficking and in vivo studies demonstrating requirement of retromer for sorLA trafficking in the brain of retromer VPS26 deficient mice. Collectively, our study establishes the role endosomal trafficking plays in sorLAs normal maturation, and point to impaired maturation as a signature of AD-associated sorLA dysfunction.

cell biology↗

bioRxiv2020

Depletion of the AD risk gene SORL1 selectively impairs neuronal endosomal traffic independent of amyloidogenic APP processing

The SORL1 gene encodes for the protein SorLA, a sorting receptor involved in retromer-related endosomal traffic. Many SORL1 genetic variants increase Alzheimers disease (AD) risk, and rare loss-of-function truncation mutations have been found to be causal of late-onset AD. SORL1 is expressed in neurons and glia of the central nervous system and loss of SORL1 has been reported in AD tissue. To model the causal loss-of-function mutations, we used CRISPR/Cas9 technology to deplete SORL1 in human induced pluripotent stem cells (hiPSCs) to test the hypothesis that loss of SORL1 contributes to AD pathogenesis by leading to endosome dysfunction. We report that loss of SORL1 in hiPSC-derived neurons leads to early endosome enlargement, a cellular phenotype that is indicative of traffic jams and is now considered a hallmark cytopathology AD. We validate defects in neuronal endosomal traffic by showing decreased localization of amyloid-precursor protein (APP) in the trans-Golgi network (TGN), and increased localization of APP in early endosomes, a site of APP cleavage by the {beta} secretase BACE1. Microglia, immune cells of the CNS, which play a role in AD pathology also express SORL1. We therefore tested and found no effect of SORL1 depletion on endosome size or morphology in hiPSC-derived microglia, suggesting a selective effect on neuronal endosomal trafficking. Finally, because BACE1 dependent APP fragments can cause endosome enlargement, we treated SORL1 deficient hiPSC-derived neurons with BACE1 inhibitors and demonstrate that endosome enlargement occurs independent of amyloidogenic APP fragments. Collectively, these findings clarify where and how SORL1 links to AD. Moreover, our data, together with recent findings, underscores how sporadic AD pathways that regulate endosomal trafficking, and autosomal-dominant AD pathways that regulate APP cleavage, independently converge on the defining cytopathology of AD.

cell biology↗

bioRxiv2019

Detecting prodromal Alzheimer’s disease with MRI through deep learning

Deep learning applied to MRI for Alzheimers classification is hypothesized to improve if the deep learning model implicates diseases pathophysiology. The challenge in testing this hypothesis is that large-scale data are required to train this type of model. Here, we overcome this challenge by using a novel data augmentation strategy and show that our MRI-based deep learning model classifies Alzheimers dementia with high accuracy. Moreover, a class activation map was found dominated by signal from the hippocampal formation, a site where Alzheimers pathophysiology begins. Next, we tested the models performance in prodromal Alzheimers when patients present with mild cognitive impairment (MCI). We retroactively dichotomized a large cohort of MCI patients who were followed for up to 10 years into those with and without prodromal Alzheimers at baseline and used the dementia-derived model to generate individual deep learning MRI scores. We compared the two groups on these scores, and on other biomarkers of amyloid pathology, tau pathology, and neurodegeneration. The deep learning MRI scores outperformed nearly all other biomarkers, including--unexpectedly--biomarkers of amyloid or tau pathology, in classifying prodromal disease and in predicting clinical progression. Providing a mechanistic explanation, the deep learning MRI scores were found to be linked to regional tau pathology, through investigations using cross-sectional, longitudinal, premortem and postmortem data. Our findings validate that a diseases known pathophysiology can improve the design and performance of deep learning models. Moreover, by showing that deep learning can extract useful biomarker information from conventional MRIs, the advantages of this model extend practically, potentially reducing patient burden, risk, and cost.

neuroscience↗

bioRxiv2019

Retromer repletion with AAV9-VPS35 restores endosomal function in the mouse hippocampus

Retromer has emerged as a master conductor of endosomal trafficking, and VPS35 and other retromer-related proteins are found to be deficient in late-onset Alzheimers disease (AD). Depleting VPS35 in neurons impairs retromer function, affecting for example the trafficking of the amyloid-precursor protein (APP) and the glutamate receptor GluA1. Whether VPS35 repletion, after chronic in vivo depletion, can rescue these impairments remains unknown. Here we set out to address this question by using a viral vector approach for VPS35 repletion. First, we completed a series of studies using neuronal cultures in order to optimize AAV9-VPS35 delivery, and to understand how exogenous VPS35 expression affects its endogenous levels as well as its binding to other retromer proteins. Next, we completed a series of studies in wildtype mice to determine the optimum protocol for in vivo delivery of AAV9-VPS35 to the hippocampus. We relied on this information to deliver AAV9-VPS35 to the hippocampus of mice genetically engineered to have chronic, neuronal-selective, VPS35 depletion. VPS35 repletion in the hippocampus was found to normalize APP cleavage and to restore glutamate receptor levels. Unexpectedly, chronic VPS35 depletion in neurons caused glial activation, similar to the pattern observed in AD, which was also partially normalized by VPS35 repletion. Taken together, these studies strengthen the mechanistic link between retromer and AD, and have therapeutic implications.

neuroscience↗

bioRxiv2018

Deep Learning on MRI Affirms the Prominence of the Hippocampal Formation in Alzheimer’s Disease Classification

Deep learning techniques on MRI scans have demonstrated great potential to improve the diagnosis of neurological diseases. Here, we investigate the application of 3D deep convolutional neural networks (CNNs) for classifying Alzheimers disease (AD) based on structural MRI data. In particular, we take on two challenges that are under-explored in the literature on deep learning for neuroimaging. First deep neural networks typically require large-scale data that is not always available in medical studies. Therefore, we explore the use of including longitudinal scans in classification studies, greatly increasing the amount of data for training and improving the generalization performance of our classifiers. Moreover, previous studies applying deep learning to classifying Alzheimers disease from neuroimaging have typically addressed classification based on whole brain volumes but stopped short of performing in-depth regional analyses to localize the most predictive areas. Additionally, we show a deep net trained to distinguish between AD and cognitively normal subjects can be applied to classify mild cognitive impairment patients, with classification scores aligning empirically with the likelihood of progression to AD. Our initial results demonstrate both that we can classify AD with an area under the receiver operator characteristic curve (AUROC) of .990 and that we can predict conversion to AD among patients in the MCI subgroup with an AURUC of 0.787. We then localize the predictive regions, by performing both saliency-based interpretation and rigorous slice and lobar level ablation studies. Interestingly, our regional analyses identified the hippocampal formation, including the entorhinal cortex, to be the most predictive region for our models. This finding adds evidence that the hippocampal formation is an anatomical seat of AD and a prominent feature in its diagnosis. Together, the results of this study further demonstrate the potential of deep learning to impact AD classification and to identify ADs structural neuroimaging signatures. The proposed classification and regional analyses methods constitute a general framework that can easily be applied to other disorders and imaging modalities.\n\nAcronyms

neuroscience↗