Hypoxic Preconditioning Ameliorates Amyloid-β Pathology and Longterm Cognitive Decline in AβPP/PS1 Transgenic Mice

Predicting Alzheimer's disease CSF core biomarkers: a multimodal Machine Learning approach

Introduction

Alzheimer's disease (AD) is a progressive neurodegenerative disorder. Current core cerebrospinal fluid (CSF) AD biomarkers, widely employed for diagnosis, require a lumbar puncture to be performed, making them impractical as screening tools. Considering the role of sleep disturbances in AD, recent research suggests quantitative sleep electroencephalography features as potential non-invasive biomarkers of AD pathology. However, quantitative analysis of comprehensive polysomnography (PSG) signals remains relatively understudied. PSG is a non-invasive test enabling qualitative and quantitative analysis of a wide range of parameters, offering additional insights alongside other biomarkers. Machine Learning (ML) gained interest for its ability to discern intricate patterns within complex datasets, offering promise in AD neuropathology detection. Therefore, this study aims to evaluate the effectiveness of a multimodal ML approach in predicting core AD CSF biomarkers.

Methods

Mild-moderate AD patients were prospectively recruited for PSG, followed by testing of CSF and blood samples for biomarkers. PSG signals underwent preprocessing to extract non-linear, time domain and frequency domain statistics quantitative features. Multiple ML algorithms were trained using four subsets of input features: clinical variables (CLINVAR), conventional PSG parameters (SLEEPVAR), quantitative PSG signal features (PSGVAR) and a combination of all subsets (ALL). Cross-validation techniques were employed to evaluate model performance and ensure generalizability. Regression models were developed to determine the most effective variable combinations for explaining variance in the biomarkers.

Results

On 49 subjects, Gradient Boosting Regressors achieved the best results in estimating biomarkers levels, using different loss functions for each biomarker: least absolute deviation (LAD) for the Aβ42, least squares (LS) for p-tau and Huber for t-tau. The ALL subset demonstrated the lowest training errors for all three biomarkers, albeit with varying test performance. Specifically, the SLEEPVAR subset yielded the best test performance in predicting Aβ42, while the ALL subset most accurately predicted p-tau and t-tau due to the lowest test errors.

Conclusions

Multimodal ML can help predict the outcome of CSF biomarkers in early AD by utilizing non-invasive and economically feasible variables. The integration of computational models into medical practice offers a promising tool for the screening of patients at risk of AD, potentially guiding clinical decisions.

Why Is Iron Deficiency/Anemia Linked to Alzheimer's Disease and Its Comorbidities, and How Is It Prevented?

Impaired iron metabolism has been increasingly observed in many diseases, but a deeper, mechanistic understanding of the cellular impact of altered iron metabolism is still lacking. In addition, deficits in neuronal energy metabolism due to reduced glucose import were described for Alzheimer's disease (AD) and its comorbidities like obesity, depression, cardiovascular disease, and type 2 diabetes mellitus. The aim of this review is to present the molecular link between both observations. Insufficient cellular glucose uptake triggers increased ferritin expression, leading to depletion of the cellular free iron pool and stabilization of the hypoxia-induced factor (HIF) 1α. This transcription factor induces the expression of the glucose transporters (Glut) 1 and 3 and shifts the cellular metabolism towards glycolysis. If this first line of defense is not adequate for sufficient glucose supply, further reduction of the intracellular iron pool affects the enzymes of the mitochondrial electron transport chain and activates the AMP-activated kinase (AMPK). This enzyme triggers the translocation of Glut4 to the plasma membrane as well as the autophagic recycling of cell components in order to mobilize energy resources. Moreover, AMPK activates the autophagic process of ferritinophagy, which provides free iron urgently needed as a cofactor for the synthesis of heme- and iron-sulfur proteins. Excessive activation of this pathway ends in ferroptosis, a special iron-dependent form of cell death, while hampered AMPK activation steadily reduces the iron pools, leading to hypoferremia with iron sequestration in the spleen and liver. Long-lasting iron depletion affects erythropoiesis and results in anemia of chronic disease, a common condition in patients with AD and its comorbidities. Instead of iron supplementation, drugs, diet, or phytochemicals that improve energy supply and cellular glucose uptake should be administered to counteract hypoferremia and anemia of chronic disease.

Chronic hypoxia leads to cognitive impairment by promoting HIF-2α-mediated ceramide catabolism and alpha-synuclein hyperphosphorylation

Chronic hypoxia leads to irreversible cognitive impairment, primarily due to hippocampal neurodegeneration, for which the underlying mechanism remains poorly understood. We administered hypoxia (13%) to C57BL mice for 1-14 days in this study. Chronic hypoxia for 7 or 14 d, but not 1 or 3 d, resulted in alpha-synuclein hyperphosphorylation at serine129 (α-Syn p-S129) and protein aggregation, hippocampal neurodegeneration, and cognitive deficits, whereas the latter could be prevented by alpha-synuclein knockdown or an administered short peptide competing at α-Syn S129. These results suggest that α-Syn p-S129 mediates hippocampal degeneration and cognitive impairment following chronic hypoxia. Furthermore, we found that chronic hypoxia enhanced ceramide catabolism by inducing hypoxia-inducible factor (HIF)-2α and HIF-2α-dependent transcriptional activation of alkaline ceramidase 2 (Acer2). Thus, the enzymatic activity of protein phosphatase 2A (PP2A), a specific phosphatase for α-syn, is inhibited, leading to the sustained induction of α-Syn p-S129. Finally, we found that intermittent hypoxic preconditioning protected against subsequent chronic hypoxia-induced hippocampal neurodegeneration and cognitive impairment by preventing α-Syn p-S129. These results proved the critical role of α-syn pathology in chronic hypoxia-afforded cognitive impairment and revealed a novel mechanism underlying α-syn hyperphosphorylation during chronic hypoxia. The findings bear implications in developing novel therapeutic interventions for chronic hypoxia-related brain disorders.

Hypoxic preconditioning reduces NLRP3 inflammasome expression and protects against cerebral ischemia/reperfusion injury

Hypoxic preconditioning can protect against cerebral ischemia/reperfusion injury. However, the underlying mechanisms that mediate this effect are not completely clear. In this study, mice were pretreated with continuous, intermittent hypoxic preconditioning; 1 hour later, cerebral ischemia/reperfusion models were generated by middle cerebral artery occlusion and reperfusion. Compared with control mice, mice with cerebral ischemia/reperfusion injury showed increased Bederson neurological function scores, significantly increased cerebral infarction volume, obvious pathological damage to the hippocampus, significantly increased apoptosis; upregulated interleukin-1β, interleukin-6, and interleukin-8 levels in brain tissue; and increased expression levels of NOD-like receptor family pyrin domain containing 3 (NLRP3), NLRP inflammasome-related protein caspase-1, and gasdermin D. However, hypoxic preconditioning significantly inhibited the above phenomena. Taken together, these data suggest that hypoxic preconditioning mitigates cerebral ischemia/reperfusion injury in mice by reducing NLRP3 inflammasome expression. This study was approved by the Medical Ethics Committee of the Fourth Hospital of Baotou, China (approval No. DWLL2019001) in November 2019.