The Role of Myelin and Large-Scale Brain Networks in Alzheimer's Disease and Cognition

Thesis Overview: Oligodendrocyte Lineage and Network Dysfunction

• Primary Research Question: This thesis investigates whether oligodendrocyte lineage pathology acts as a cellular substrate for default mode network (DMN) structure-function decoupling in Alzheimer’s Disease (AD), and whether cognitive learning independently drives myelin remodeling in healthy circuits.
• Experimental Frameworks:     - Alzheimer’s Model: Utilized 9.4T resting-state functional Magnetic Resonance Imaging (rs-fMRI) and triple immunofluorescence for myelin basic protein (MBP), PDGFR$\alpha$, and amyloid-beta ($A\beta$) across three stages in $App$ knock-in mouse models ($App^{NL-G-F/NL-G-F}$).     - Healthy Model: Utilized a touchscreen-based Trial-Unique Delayed Non-Matching-to-Location (TUNL) task with a yoked control design and ex vivo quantitative $T1$ mapping to isolate cognitive learning in healthy $C57BL/6$ mice.
• Key Findings in AD Models:     - Amyloid pathology results in a biphasic reorganization of DMN connectivity: early hyper-synchrony transitioning to late-stage hypo-synchrony.     - This is paralleled by region-specific demyelination, loss of oligodendrocyte precursor cells (OPCs), and the arrest of normal age-dependent myelin maturation.     - Myelin coverage positively predicts network amplitude in the anterior cingulate cortex (ACC) and $CA3$ in controls, but this coupling is absent in animals with amyloidopathy.
• Key Findings in Healthy Models:     - Cognitive rule learning produces task-specific $T1$ reductions (indicating increased myelin) in prefrontal and hippocampal circuits.     - Conditional $Myrf$ knockout established that new myelination is a biological requirement for spatial working memory acquisition.

Large-Scale Brain Networks and Connectomics

• Historical Foundations:     - Camillo Golgi (1843–1926): Developed the "reazione nera" (black reaction) silver-chromate impregnation technique in 1873 to visualize individual neurons and glia.     - Santiago Ramón y Cajal: Used Golgi’s method to reveal that the brain is an assembly of discrete cells (synaptic theory) rather than a continuous syncytium.
• Modern Mapping Initiatives:     - Human Connectome Project (HCP): Launched in 2010 to map structural and functional connectivity in healthy humans.     - Allen Mouse Brain Atlas: Spatially resolved map of gene expression and cell types (transcriptomic, morphological, and electrophysiological properties).     - Connectome Definition: A comprehensive map of neural connections. Structural connectivity refers to physical white matter tracts. Functional connectivity refers to statistical dependencies (temporal correlations) between spatially separated regions.
• Functional MRI (fMRI) Principles:     - Neurovascular Coupling: Local neuronal activity increases cerebral blood flow beyond oxygen consumption, reducing the ratio of paramagnetic deoxygenated hemoglobin to oxygenated hemoglobin.     - BOLD Signal: Blood-Oxygen-Level-Dependent signal. Spontaneous low-frequency fluctuations ($< 0.1\,Hz$) characterize resting-state functional connectivity.
• Cross-Species Considerations: Comparing networks like the DMN across species is more reliable than comparing individual regions (e.g., debate over whether rodents possess a true prefrontal cortex comparable to primates).

The Default Mode Network (DMN) and Structure-Function Coupling

• DMN Characteristics: A set of brain regions active during rest/internal thought and suppressed during goal-directed behavior. Key hubs include the medial prefrontal cortex (mPFC), posterior cingulate cortex (PCC), and precuneus.
• Behavioral Relevance: Strong DMN connectivity is associated with working memory and episodic memory. Efficient suppression of the DMN during task performance (negative coupling with the frontoparietal network) predicts sustained attention.
• Structure-Function Coupling (S-F):     - Strength of coupling is highest in primary sensory/motor areas and weakens in transmodal association regions (DMN, frontoparietal).     - Myelin provides the structural basis for synchronization; disruption slows axonal conduction and weakens functional connectivity.     - Anatomical Pathways: The cingulum bundle is the primary structural correlate of DMN connectivity. Higher-order connectivity is also mediated by higher-order thalamic nuclei (pulvinar and mediodorsal nuclei) acting as subcortical intermediaries.

Oligodendrocyte Lineage and Myelin Dynamics

• The Myelin Sheath: A modified plasma membrane that spirals around axons. In the Central Nervous System (CNS), one oligodendrocyte myelinates multiple segments on various axons. In the Peripheral Nervous System (PNS), one Schwann cell forms a single sheath.
• Saltatory Conduction: Discovered by Tasaki (1939) and Huxley & Stämpfli (1949). Action potentials regenerate at nodes of Ranvier. Benefits include conduction velocity up to $100$-fold faster and reduced metabolic cost ($Na^+/K^+-ATPase$ activity restricted to nodes).
• Oligodendrocyte Lineage Stages:     1. Oligodendrocyte Precursor Cells (OPCs): Express $NG2$ and $PDGFR\alpha$. Persist throughout adulthood as a reservoir for repair.     2. Late OPCs / Premyelinating Oligodendrocytes.     3. Mature (Myelinating) Oligodendrocytes: Capable of forming sheaths.
• Mechanisms of Myelination:     - Axon Selection: Biotransport threshold is typically a diameter $> 0.2\,\mu m$.     - Signals: Neuregulin-1 ($NRG1$) promotes differentiation. PSA-NCAM and L1 inhibit ensheathment. Electrical activity triggers ATP release and astrocyte-derived leukemia inhibitory factor ($LIF$) to promote myelination.
• Developmental Trajectory: Human white matter volume peaks in the fifth decade of life. A rapid decline in myelin density occurs after age $60$, contributing to age-related cognitive decline.

Demyelination and Remyelination Processes

• Vulnerability of Oligodendrocytes: Mature cells have high metabolic demands (membrane surface area $100\times$ the cell body). They rely on glucose and lactate (supplied via Monocarboxylate Transporter 1, $MCT1$).
• Causes of Demyelination: Ischemia, hypoxia, mitochondrial dysfunction, ethanol exposure, and heavy metal toxicity (cadmium, lead). Myelin is lipid-rich and susceptible to lipid peroxidation due to low antioxidant capacity (low catalase and glutathione peroxidase).
• Consequences: Conduction failure, dissipation of depolarizing current ("leaky" membrane), and pathological calcium ($Ca^{2+}$) accumulation leading to axonal loss.
• Remyelination: Restorative process primarily mediated by OPC differentiation. Remyelinated sheaths are characteristically thinner and shorter than developmental ones.
• Therapeutic Strategies:     - Clemastine Fumarate: Muscarinic receptor antagonist ($M1/M3$) that promotes OPC differentiation. ReBUILD clinical trial showed improved visual evoked potential latency in MS patients.     - Other Targets: Anti-$LINGO-1$ antibodies ($Opicinumab$), $RXR\gamma$ agonists, and non-invasive brain stimulation (TMS/tDCS).

Alzheimer’s Disease Pathophysiology and Models

• Biomarkers (ATN Framework):     - (A) Amyloid: Reduced $A\beta 42/40$ ratio in CSF; PET visualization with florbetapir.     - (T) Tau: Phos-tau concentrations and PET visualization of neurofibrillary tangles.     - (N) Neurodegeneration: Structural MRI (atrophy), FDG-PET (hypometabolism), and plasma neurofilament light chain ($NfL$).
• Risk Factors: $APOE4$ is the strongest genetic risk for late-onset AD. Facilitating cholesterol transport in $APOE4$ models can restore myelination.
• The $App^{NL-G-F}$ Model: Knock-in model containing Swedish ($NL$), Arctic ($G$), and Iberian ($F$) mutations. Recapitulates amyloid pathology without the artifacts of protein overexpression found in earlier transgenic lines (APP/PS1).
• AD Stages in Mice:     - 3 months: Preclinical/Presymptomatic; plaques detectable but spatial memory remains intact.     - 7 months: MCI to moderate AD transition; near-saturation of plaques, neuroinflammation, and onset of cognitive deficits.     - +1 year: Advanced AD; extensive cognitive impairment and mitochondrial dysfunction.

Chapter 2: Myelin Maturation Arrest and DMN Dysfunction

• Methodology: rs-fMRI at 9.4T and triple immunofluorescence (MBP, PDGFR$\alpha$, $A\beta$) in $App^{NL-G-F/NL-G-F}$ and $App^{NL/NL}$ (control) mice.
• Biphasic DMN Trajectory:     - 3 Months: Hyper-synchrony in ACC, S1, and CA1.     - 7 Months: Transition to hypo-synchrony in ACC and CA1; S1 remains hyper-synchronous.     - +1 Year: Marked attenuation of genotype differences; sparse residual hypo-synchrony.
• Cellular Pathology Results:     - Demyelination: Significant myelin loss in $CA3$ and DG as early as 7 months.     - Myelin Maturation: Normal age-dependent increases in MBP signal ($3\,m$ to $+1\,yr$) observed in control $S1$ and $CA3$ were completely absent in AD mice.     - OPC Loss: Progressive reduction in PDGFR$\alpha$ coverage in $CA3$ at 7 months and +1 year.
• S-F Coupling Breakdown: In healthy animals, MBP coverage positively correlates with DMN amplitude in the ACC ($p = 0.0121, R^2 = 0.395$) and $CA3$ ($p = 0.0001, R^2 = 0.692$). This relationship is statistically absent ($p > 0.05$) in amyloidopathy models.

Chapter 3: Cognitive Learning and De Novo Myelination

• Objective: Determine if cognitive rule learning (non-motor) drives myelin remodeling.
• Task: Touchscreen-based Trial-Unique Delayed Non-Matching-to-Location (TUNL). Yoked controls received rewards with pseudo-randomized rules to control for motor and sensory exposure.
• MRI Results (T1 Mapping):     - Trained animals exhibited significantly shorter $T1$ relaxation times (indicating higher myelin) in the mPFC, dorsal hippocampus, amygdala, VTA, SN, and parietal cortex.     - No changes in primary motor ($M1$) or primary somatosensory ($S1$) cortices.
• Cellular and Causal Evidence:     - Oligodendrogenesis: TUNL training increased newly differentiated $GFP^+$ oligodendrocyte density in mPFC and dorsal hippocampus (confirmed via $Pdgfra$-CreER:Tau-mGFP mice).     - Functional Requirement: Blocking new myelination via inducible $Myrf$ deletion impaired task acquisition ($p = 0.0027$).
• Implications: Adaptive myelination is not just for motor skills; cognitive experience selectively sculpts myelin architecture to optimize conduction timing (e.g., theta-gamma coupling) for memory encoding.

Collective Research Significance

• Oligodendroglial Lineage as a Target: Protecting myelin maturation and promoting repair are promising strategies for maintaining network integrity in Alzheimer's.
• Diagnostic Utility: The hyper-to-hypo-synchrony transition in rs-fMRI may serve as a functional biomarker for the clinical tipping point in AD progression.
• Cognitive Engagement: Cognitive learning stimulates new myelin growth, suggesting that experience-dependent myelination supports long-term brain health and plasticity.