Single-Cell Transcriptomics Uncovers Early Cellular Drivers Across Living ALS Disease Models

A comprehensive review of single-cell and single-nucleus RNA sequencing in amyotrophic lateral sclerosis demonstrates that profiling living model systems uncovers early, cell-type-specific pathogenic mechanisms that remain undetectable in end-stage human post-mortem tissue. Amyotrophic lateral sclerosis is a fatal neurodegenerative disorder defined by the progressive degeneration of motor neurons in the cortex, brainstem, and spinal cord. While approximately ninety percent of cases are classified as sporadic, roughly ten percent are familial, linked to well-characterized mutations in genes including C9orf72, SOD1, TARDBP, and FUS. Historically, mechanistic inquiries into human disease relied heavily on bulk transcriptomic profiling of post-mortem tissue samples. However, terminal tissues capture only a static cross-section of surviving cells confounded by post-mortem intervals, cryopreservation damage, and advanced cellular loss. By integrating high-resolution single-cell technologies across fruit flies, transgenic rodents, and human induced pluripotent stem cells, researchers can now track transcriptional cascades across complete disease trajectories, uncovering distinct molecular signatures that precede overt neurodegeneration.

Technical execution of single-cell transcriptomics in neurodegenerative research requires navigating distinct trade-offs between cellular integrity and transcript coverage. Isolating intact single cells from adult central nervous system tissues poses substantial challenges because mechanical and enzymatic dissociation shears fragile neuronal projections and distorts cellular transcriptomes. To circumvent these dissociation artifacts, single-nucleus RNA sequencing has emerged as an indispensable alternative, extracting resilient double-membrane nuclei from complex or frozen tissues using ultracentrifugation alongside fluorescence-activated nuclei sorting or isolation of nuclei tagged in specific cell types. While single-nucleus profiling avoids whole-cell dissociation stress and captures intact nuclear RNA, it necessarily excludes cytoplasmic transcripts, potentially missing localized translational and metabolic processes. Conversely, in vitro models such as human induced pluripotent stem cells allow direct whole-cell capture through microfluidics or flow-activated sorting. Plate-based methods like Smart-seq2 and Fluidigm C1 process fewer cells but yield over 1.5 million reads per cell, facilitating deep splice-variant detection, whereas microdroplet-based platforms like 10X Genomics capture thousands of cells simultaneously at shallower sequencing depth, balancing scale against gene coverage.

In Drosophila melanogaster models, single-cell RNA sequencing has demonstrated how intrinsic baseline transcriptional states dictate neuronal survival. Fruit flies share over seventy-seven percent of human disease genes and offer rapid generation times, compact nervous systems, and powerful genetic screening assays evaluating climbing mobility, eye morphology, and lifespan. In neuronal fly models expressing thirty-six GGGGCC hexanucleotide repeats within C9orf72, researchers utilized microdroplet single-cell transcriptomics across progressive timepoints to resolve differential vulnerability across distinct neuronal subsets. The study revealed that neuronal susceptibility to repeat toxicity is determined primarily by basal expression profiles rather than de novo stress responses mounted after insult. Specifically, vulnerability aligned with the baseline expression of proteostasis networks, wherein genetic cross-comparisons demonstrated that the transcription factor Xbp1 directly modulates resistance to C9orf72-mediated neurotoxicity. These findings demonstrate that innate endoplasmic reticulum proteostatic capacity acts as a crucial checkpoint, showing how genetically tractable invertebrate models can identify functional modifiers of neurodegenerative cascades across whole lifespans.

Transgenic mouse models provide essential mammalian physiology, conserved neuroanatomy, and intact neuroimmune networks to map non-cell-autonomous contributions to motor neuron death. In high-copy SOD1-G93A mice, massively parallel single-cell sequencing of spinal cord tissues revealed a distinct subpopulation of disease-associated microglia that increased by twenty-four percent between early day eighty and late day one hundred thirty-five disease stages. Subsequent longitudinal single-nucleus profiling across the cerebrum, brainstem, and spinal cord resolved the temporal trajectory governing microglial activation. The transition into a hyperactive neuroinflammatory state is driven by the downregulation of homeostatic inhibitory checkpoints, including Cx3cr1 and P2ry12, accompanied by the upregulation of endocytosis, inflammation, and phagocytic machinery. Spatial transcriptomics paired with nuclear sequencing further showed that microglial clustering directly adjacent to motor neurons precipitates the transition of motor neurons into a disease-associated state, driven by the transcription factors CREB3 and ATF3, marked by impaired axonogenesis and synaptic loss.

Single-cell explorations in rodent models have further illuminated shared mechanisms of intrinsic neuronal hyperexcitability and organelle dysfunction across disparate genetic subtypes. Microdroplet profiling of pontine brainstem tissue from SOD1-G93A mice identified distinct ion channel remodeling within motor neurons, characterized by the upregulation of voltage-gated sodium channel genes such as Scn2a and the coordinated downregulation of potassium channel genes, including Kcna1 and Kcnc3. This channel imbalance promotes membrane hyperexcitability and calcium dysregulation that accelerate cell death. Parallel single-nucleus evaluations in Tdp-43 conditional knockout mice demonstrated that cortical TDP43 depletion induces severe loss of GABAergic inhibitory neurons, calcium homeostasis defects, and the activation of microglial inflammatory genes including Oasl2 and Lyz2. Concurrently, nuclear transcriptomics in frontal cortex tissues from C9-BAC and C9orf72 knockout mice revealed that progressive loss of C9orf72 depletes ribosomal structural proteins and biogenesis machinery, triggering an overwhelming proteotoxic cellular stress response.

Human induced pluripotent stem cell platforms offer human genetic context, permitting researchers to model early pathogenic molecular dynamics in patient-derived motor neurons and glia. In early droplet-based studies tracking patient-derived C9orf72 motor neurons throughout in vitro differentiation, single-cell analyses identified early predictive gene expression alterations linked to differentiation, cell-type specification, and axon guidance, including downregulation of ELAVL3, NUAK1, and ADCYAP1. To enhance analytical resolution beyond standard shallow droplet sequencing, researchers deployed automated Fluidigm C1 microfluidics to sequence SOD1-E100G patient-derived motor neurons, achieving over three times higher gene detection sensitivity. Computational regulatory network analysis identified hyperactivation of transforming growth factor beta signaling and downstream SMAD2 nuclear phosphorylation as primary drivers of motor neuron death. Validating these computational predictions, experimental pharmacologic inhibition of transforming growth factor beta restored motor neuron survival in both SOD1-E100G and FUS-P525L cultures, confirming that developmental signaling abnormalities actively drive adult neurodegenerative phenotypes.

Pairing patient-derived neurons with CRISPR-Cas9 isogenic control lines has similarly uncovered convergence upon early mitochondrial and bioenergetic collapse across divergent ALS genotypes. In isogenic motor neurons carrying TARDBP-M337V, FUS-P525L, or FUS-R495X mutations, single-cell profiling revealed that broad transcriptional downregulation of oxidative phosphorylation and electron transport-coupled adenosine triphosphate synthesis occurs prior to the cytoplasmic aggregation of TDP43 or FUS proteins. While TARDBP mutations selectively impaired non-coding RNA nuclear export and FUS mutations disrupted endoplasmic reticulum protein localization, both mutant lines shared severe early bioenergetic failure. Cross-referencing these data with bulk transcriptomics from C9orf72 motor neurons confirmed shared bioenergetic disruption, indicating that mitochondrial collapse operates upstream of overt pathological protein aggregation across multiple familial forms of the disease.

Beyond motor neurons, single-nucleus and single-cell sequencing in stem cell models have transformed our understanding of non-neuronal cells, exposing profound functional divergence across genetic backgrounds. While SOD1 and TARDBP mouse models exhibit pronounced microglial hyperactivation, single-nucleus profiling of patient-derived C9orf72 microglial xenografts in immunodeficient mice revealed an opposing phenotype termed microglial immunoparalysis. Mutant C9orf72 microglia significantly downregulated antigen-presenting human leukocyte antigen genes, such as HLA-DPA1, while aberrantly upregulating homeostatic markers, preventing their transition into an immunocompetent reactive state. In vitro single-cell sequencing of co-cultured motor neurons and microglia stimulated with lipopolysaccharides confirmed this immune paralysis, showing reduced cytokine release, suppressed nuclear factor kappa B signaling, and the complete absence of a lipopolysaccharide-responsive microglial subpopulation, demonstrating that C9orf72 haploinsufficiency directly compromises microglial neuroprotective capacity.

Integrating single-cell datasets from animal and cellular models with human patient profiles bridges experimental tractability and clinical relevance. Utilizing expression-weighted cell-type enrichment, investigators compared single-cell profiles across human and mouse spinal cords, revealing that resistant gamma motor neurons survive in SOD1-G93A mice while vulnerable alpha motor neurons selectively degenerate. This cross-species computational mapping demonstrated that gamma motor neurons selectively associate with causal ALS mutations, such as SOD1 and VCP, whereas alpha motor neurons associate with broad genetic susceptibility loci. Furthermore, cross-species differential gene expression comparisons between human tissues and SOD1 mice identified conserved microglial candidate biomarkers, including CALM1, CALM2, and HSPA1A, while revealing that microglial metabolic decline correlates with disrupted astrocyte interactions and worsening neuroinflammation.

Despite these critical discoveries, current single-cell investigations face significant methodological and biological limitations inherent to existing model platforms. In vivo rodent research remains heavily skewed toward transgenic overexpression lines, particularly high-copy SOD1-G93A animals, which express supra-physiological protein levels that risk generating unrepresentative overexpression artifacts. Conversely, human induced pluripotent stem cell cultures represent simplified in vitro environments lacking mature multi-organ physiology, and their cellular reprogramming erases age-associated epigenetic hallmarks, yielding embryonic-like states that struggle to capture late-onset neurodegenerative kinetics. Additionally, current laboratory models cannot recreate the cumulative decades of environmental exposures, such as pesticides or heavy metals, implicated in sporadic ALS etiology. Resolving these challenges will require expanding single-cell analyses to endogenous knock-in models and underrepresented mutations, while incorporating spatial transcriptomics and single-cell proteomics to validate whether transcriptomic shifts accurately reflect functional protein expression and structural degeneration in human pathology.

Subject of Research: Single-cell transcriptomic profiling of neurodegenerative mechanisms in amyotrophic lateral sclerosis model systems

Article Title: Integrating single-cell transcriptomics into model systems to decode neurodegeneration in amyotrophic lateral sclerosis

Article References: Orsulik, D., & Niccoli, T. (2026). Integrating single-cell transcriptomics into model systems to decode neurodegeneration in amyotrophic lateral sclerosis. Frontiers in Cellular Neuroscience, 20, Article 1911064. https://doi.org/10.3389/fncel.2026.1911064

Image Credits: AI Generated

DOI: 10.3389/fncel.2026.1911064

Keywords: Amyotrophic Lateral Sclerosis, Single-Cell RNA Sequencing, Single-Nucleus RNA Sequencing, Disease Models, Motor Neurons, Microglia, Neuroinflammation, C9orf72, SOD1, TARDBP, FUS, Mitochondrial Dysfunction, Induced Pluripotent Stem Cells, Drosophila melanogaster, Neurodegeneration, Transcriptomics, Cross-Species Integration