Cognitive dysfunction is a major global burden. Microglia mediate neuroinflammation linking diverse etiologies to neuronal injury, with their functions extending beyond the brain parenchyma to include bidirectional crosstalk with the neurovascular unit. This review examines the dual role of microglia in cognitive disorders and therapeutic implications of their plasticity. Modern transcriptomic studies indicate that microglial states are multidimensional and context-dependent, rather than conforming to a simple M1/M2 dichotomy. Instead, microglia exist along a dynamic continuum encompassing homeostatic, disease-associated, and senescence-associated states. The relative contribution of protective and detrimental programs changes according to disease stage, anatomical location, pathological burden, and local cellular environment, and the two programs may coexist even within the same disease stage. Therapeutic strategies are therefore shifting from broad anti-inflammatory suppression toward precision modulation of microglial states. Their translational readiness varies from preclinical proof-of-concept to Phase II/III evaluation, and target engagement alone does not ensure restoration of physiological microglial function or cognitive benefit. Emerging interventions include TREM2 agonists, CSF1R modulators, PPARγ activators, NLRP3/STING/complement inhibitors, and microglia replacement. Key challenges include brain-targeted delivery and biomarker validation. Understanding microglia as plastic, context-dependent regulators may inform precision immunomodulatory therapies.
Parkinson's disease (PD) is characterized by early alterations in large-scale brain functional connectivity and possible disturbances in excitatory amino acid metabolism. However, it is unknown whether topological changes in functional connectome reflect disruptions in the homeostasis of circulating glutamate and aspartate levels.Resting-state fMRI and High-Performance Liquid Chromatography are used to correlate brain and blood measures in 74 early-stage PD patients (mean age ± s.d. 63.5 ± 8.6 years, 51 males, 37 drug-naïve) and 20 healthy control (HC) volunteers (mean age ± s.d. 60.2 ± 5.6, 11 males). PD-relevant graph-theoretical metrics of fMRI connectomes are co-varied to serum-levels of N-Methyl-d-aspartate receptor (NMDAR)-related (L-/D-) amino acids (including precursors and related ratios).In male PD patients, compared to sex-matched HC subjects, strength of cortical sensorimotor nodes and clustering of subcortical nodes negatively correlate to D−/total aspartate ratio; clustering of temporal nodes positively correlates to l-aspartate; global characteristic path length negatively correlates to L-glutamate. Enlarging the panel of L-amino acids, l-serine markedly predicts metrics in treated male PD patients.Peripheral balance of excitatory amino acids and their precursors appear linked to large-scale brain network (re)organization in early-stage PD, supporting the potential value of these neuroactive biomolecules as circulating markers of functional brain integrity.
Maternal immune activation (MIA) during pregnancy, triggered by infectious or non-infectious factors, is a significant risk factor for neurodevelopmental disorders such as schizophrenia (SZ) and autism spectrum disorder (ASD), which are associated with dysfunctions of the dopaminergic system. Animal models using immune challenges like LPS and Poly(I:C) during gestation recapitulate key behavioral and neurochemical phenotypes observed in these disorders. This review synthesizes current evidence from both human epidemiological studies and animal models, highlighting the links between prenatal inflammation and alterations in dopaminergic circuits. In animal models, MIA causes region-specific changes in dopamine (DA) metabolism, receptor expression, and dopaminergic neuron activity. Notably, hyperdopaminergic states in the striatum and hypodopaminergic states in the cortex have been reported in animal models of MIA, but with some variability across studies. Surprisingly, decreased activity of ventral tegmental area (VTA) dopaminergic neurons in vivo emerges as one of the most robust findings in animal models of MIA. We discuss the possible mechanisms underlying MIA-induced dopaminergic phenotypes, focusing in particular on the endocannabinoid system and on microglia-mediated synaptic refinement and complement system activation. This review underscores the complexity of MIA's impact on brain development and dopaminergic function, emphasizing the need for further research into region-specific mechanisms underlying neuropsychiatric vulnerability, as well as innovative methodologies to investigate how MIA leads to long-term changes in the dynamics of DA release in behaving animals.
Ageing impairs oligodendrocyte precursor cell (OPC) function, but it is unclear whether age-related transcriptional changes in the human cortex mainly reflect shifts in OPC-state composition or remodelling within transcriptionally defined states. We defined a bidirectional 41-gene OPC ageing programme (22 age-down and 19 age-up genes) and tested it in two external cortical single-nucleus datasets that were not used for gene selection. The age-down and age-up components showed the expected associations in each dataset. In paired RNA–chromatin profiles from 349 donors, formal decomposition showed that within-state change accounted for 98.8–99.5% of the programme's adjusted cell-mean contrast, whereas composition accounted for 0.8–1.2% under the primary clustering configuration; composition point estimates remained approximately 2% or less across five clustering resolutions and three representations. The relative occupancy of a commitment-associated OPC state declined with age in both donor sources after adjustment for sampled OPC depth, but removing this state left the broad programme essentially unchanged. Promoter accessibility changed in parallel with the RNA programme and correlated with paired RNA scores. The decomposition applies to the adjusted contrast of this 41-gene programme, and the occupancy estimate describes sampled OPC nuclei rather than absolute tissue abundance. Together, the results distinguish transcriptional remodelling distributed across OPC states from a focused change in commitment-associated state representation.
TAR DNA-binding protein 43 (TDP-43) is a DNA/RNA-binding protein whose pathological aggregation characterizes a spectrum of neurodegenerative disorders known as TDP-43 proteinopathies, including amyotrophic lateral sclerosis (ALS), frontotemporal lobar degeneration (FTLD), and chronic traumatic encephalopathy (CTE). Recent breakthroughs in cryo-electron microscopy (cryo-EM) have revolutionized our understanding of these diseases by resolving the atomic structures of TDP-43 amyloid fibrils. These structural insights have been pivotal in elucidating the transition from functional liquid-liquid phase separation (LLPS) to irreversible solid-state aggregation. This review synthesizes key cryo-EM findings from 2019 to 2026, encompassing fibrils formed by the full-length low-complexity domain, disease-relevant core fragments, and filaments extracted directly from patient tissues, as well as the pathological implications of distinct structural polymorphs. These findings not only deepen our understanding of the brain science aspects of the pathogenesis of TDP-43 proteinopathies, but also provide an important structural basis for the development of specific conformational diagnostic methods and targeted therapeutic drugs.
