Region-resolved myelin and OXPHOS mapping in neurodevelopmental and neuromuscular mouse models

Question: How have spatial transcriptomics studies in other neurodevelopmental or neuromuscular disease mouse models dissected region-specific myelin and oxidative phosphorylation gene changes?

Answer

Within the retrieved corpus, this dissection has been done cleanly exactly once: whole-brain, atlas-registered spatial transcriptomics of mdx and mdx4cv mice, which resolves myelin/oligodendrocyte and oxidative phosphorylation modules region by region and shows that both programs move in region- and isoform-dependent directions rather than uniformly [1]. The other models supply the surrounding logic rather than the same measurement: tract-resolved spatial work in neonatal hypomyelination and cuprizone demyelination shows that white-matter regions diverge from one another and that "region" must be resolved at the level of individual tracts [9] [12]; Complex I deficiency places OXPHOS upstream of region-selective oligodendrocyte and corpus callosum developmental failure [8]; and cross-sectional multi-tissue spatial designs in SMA extend regional analysis beyond the CNS to muscle, bone, and marrow [4]. Across these papers, region specificity is the reproducible finding; the direction of myelin and OXPHOS change is not, and no retrieved study besides the DMD one measures both programs jointly in the same sections.

Only whole-brain atlas-registered profiling has simultaneously resolved myelin and OXPHOS modules across regions

The DMD study performed spatial transcriptomics on brains of mdx (Dp427-deficient), mdx4cv (Dp427+Dp140-deficient), and control mice at two weeks of age, and reported that transcriptional changes in both dystrophic models converged on myelin and oligodendrocyte biology and on oxidative phosphorylation, but with distinct regional patterns [1]. In mdx mice, myelin genes were upregulated in amygdala and hippocampus and downregulated in fiber tract regions, while mitochondrial genes were upregulated in most regions but downregulated in thalamus and isocortex [1]. This is the key methodological point for the field: a single whole-brain section can show the myelin and OXPHOS modules pointing in opposite directions in different regions, so any bulk or single-region assay would average the signal away.

The regional pattern tracks the specific genetic lesion rather than the disease label

Direct comparison of the two dystrophin-deficient models showed that additional loss of Dp140 expanded localized myelin upregulation across the brain and shifted the mitochondrial response toward widespread downregulation, a difference the authors attribute to the specific dystrophin isoform(s) lost [1]. The same principle appears in the OXPHOS direction: Complex I (NDUFS4) deficiency disrupted neural stem cells, their neuronal and oligodendroglial progeny, and lineage progression specifically in the subventricular zone and corpus callosum, leading the authors to propose mitochondrial dysfunction as a neurodevelopmental signature that predates classical Leigh syndrome diagnosis [8]. In SBMA, laser-capture motor neuron profiling of AR100 mice showed disruption of mitochondrial and DNA damage repair pathways progressing from pre-symptomatic to end stage, with shared spliceosome, cell cycle, and mitochondrial dysregulation emerging early in motor neurogenesis [7]. Together these argue that regional myelin/OXPHOS signatures are readouts of which gene product is missing and when, not of clinical category.

Tract-level and compartment-level ROIs show that coarse anatomy hides the signal

CosMx spatial transcriptomics comparing intermittent hypoxia and hypoxia-ischemia in neonatal mice found that the two models affected oligodendrocyte populations differently and that effects varied by region: in the internal capsule, intermittent hypoxia caused earlier and more pronounced changes in oligodendrocyte differentiation gene expression, whereas the corpus callosum was more affected by hypoxia-ischemia, and mature oligodendrocytes in both regions showed reduced myelination-associated gene expression in that model [9]. Integrating single-cell and spatial transcriptomics in the cuprizone model revealed demyelination and neuroinflammation beyond the corpus callosum with region-specific differences, and showed that mature oligodendrocytes nearly reversed their phenotype during remyelination while microglia remained in a demyelination-associated state [12]. Compartment-resolved spatial transcriptome–metabolome integration in an amyloid model similarly separated plaque-rich gray matter, where microglial Pld4 and glycerophospholipid remodeling dominated, from plaque-sparse white matter, where oligodendrocyte Ugt8a loss and altered galactosylceramide metabolism dominated [11]. The consistent lesson is that "region" in these disorders must mean tract, lamina, or niche, not broad structure.

OXPHOS lesions are sufficient to produce region-selective myelin and white-matter developmental defects

NDUFS4 knockout disrupted oligodendroglial lineage progression and produced overt regional differences, with the corpus callosum among the affected structures [8]. This is the closest thing in the corpus to a causal ordering experiment: a primary respiratory-chain lesion generates region-selective glial developmental failure, supporting the interpretation that OXPHOS gene changes are not merely downstream of demyelination. Consistent with a metabolic-to-myelin axis, tissue proteomics in SMA showed heart and muscle converging on metabolic and mitochondrial pathways and persistent dysregulation of these pathways after partial SMN restoration [5]. Imaging transcriptomics in the PLP-Syn multiple system atrophy model linked regional MRI phenotypes to gene programs for oligodendrocyte biology, energy metabolism, and neuroinflammation with modality-specific signatures [2].

Extending the regional logic outside the CNS changes what counts as a region

Spatial transcriptomics of whole lumbar cross-sections in pre-symptomatic SMA preserved spinal cord, dorsal root ganglia, muscle, bone, cartilage, bone marrow, adipose, and connective tissue in native context and found region-restricted programs: neurofilament changes largely confined to motor neurons, broader ventral tubulin dysregulation, fiber-type-specific stress responses in muscle, an osteoclast-dominant bone signature, and inflammatory activation in bone marrow [4]. In mdx nerve, single-nucleus RNA sequencing and structural analysis identified an active denervation program and failed neural repair, with full-length Dp427c normally accumulating at Schmidt-Lanterman incisures and nodes of Ranvier [10]. High-resolution spatial mapping of soleus muscle delineated the neuromuscular junction as a multi-compartment niche with distinct peri-synaptic and myelinating Schwann cell-associated programs, and showed spatially nonuniform damage-response gene activation along individual myofibers after denervation [6].

Conflicts and gaps

The direction of myelin gene change is unresolved: it is upregulated in amygdala and hippocampus of mdx mice, broadly upregulated in mdx4cv, and reduced in mature oligodendrocytes of hypoxia-ischemia animals [1] [9]. Age and stage differ across studies (two weeks in DMD, P12/P20 in hypomyelination, P4 pre-symptomatic in SMA), so developmental and degenerative components cannot be separated [1] [9] [4]. Region definitions are not shared: Allen-atlas structures in the DMD study, named tracts in the hypomyelination study, and tissue types in the SMA study are not mutually mappable [1] [9] [4]. No retrieved mouse study reports a direct spatial correlation between myelin and OXPHOS module scores within the same sections, and none resolves whether the OXPHOS change is oligodendrocyte-intrinsic or secondary to axonal or vascular dysfunction. The strongest regional-contrast spatial evidence in a motor neuron disease is human, where motor cortex and lumbar spinal cord signatures were predominantly non-overlapping [3].

What would change the answer

Running mdx, mdx4cv, NDUFS4 knockout, SMA, SOD1-G93A, and cuprizone tissue on a single platform with a shared myelin/OXPHOS panel (Xenium, CosMx, or MERFISH) and quantifying module scores against one common region ontology would make the directionality question answerable across models. Coupling such panels to in situ oligodendrocyte lineage trajectories and to spatial metabolomics or NAD+/ATP imaging, as done for glial lipid metabolism [11], would test whether myelin and OXPHOS changes co-localize within the same spatial domains or dissociate. Conditional rescue of Complex I or OXPHOS subunits in oligodendrocytes in a dystrophic background would establish causal ordering between the two programs.

References

[1] Mao, Q., Heezen, L. G. M., de Vries, S., Verwoerd, N., Aartsma-Rus, A., van Putten, M., Spitali, P., Mahfouz, A.. "Spatial transcriptomics maps region and isoform specific brain alterations in mdx and mdx4cv models of Duchenne muscular dystrophy". bioRxiv. 2026-10-19T00:00:00.000Z. URL: https://doi.org/10.64898/2026.10.13.758478

[2] Kim, E., Cash, D., Martins, D., Simmons, C., Heras-Garvin, A., Klippel, E., Krismer, F., Mantoan-Ritter, L., Stefanova, N.. "Multiparametric MRI and imaging transcriptomics reveal molecular and cellular correlates of neurodegeneration in experimental multiple system atrophy". bioRxiv. 2026-02-14T00:00:00.000Z. URL: https://doi.org/10.1101/2025.11.12.688033

[3] Barretto, N., Fullerton, B. T., Daly, A. C., Casel, O., Kuksenko, O., Kang, K., Petrescu, J., Leung, M., Khiste, S., Gebremedhin, B., Smith, C., Jackson, C., Phatnani, H.. "A multi-omics characterization reveals distinct molecular signatures in the human motor cortex and lumbar spinal cord in ALS". bioRxiv. 2026-07-28T00:00:00.000Z. URL: https://doi.org/10.64898/2026.07.24.740361

[4] Rietz, A., Kumari, L., Androphy, E. J.. "Spatial transcriptomics identifies dysregulated programs across neural and non-neural tissues in spinal muscular atrophy". bioRxiv. 2026-03-23T00:00:00.000Z. URL: https://doi.org/10.64898/2026.03.20.713154

[5] Vrettou, S., Mueller, S., Wirth, B.. "System-Wide Proteomic Remodeling in Spinal Muscular Atrophy Reveals Tissue-Specific Responses and Partial Rescue by SMN Restoration". bioRxiv. 2026-03-31T00:00:00.000Z. URL: https://doi.org/10.64898/2026.03.30.715402

[6] Hsu, J.-E., Ruiz, L., Hwang, Y., Guzman, S., Cho, C.-S., Cheng, W., Si, Y., Macpherson, P., Anacleto, A., Zhao, Q., Zhao, X., Schrank, M., Jun, G., Kang, H.-M., Kim, M., Brooks, S., Lee, J. H.. "High-Resolution Spatial Transcriptomic Atlas of Mouse Soleus Muscle: Unveiling Single Cell and Subcellular Heterogeneity in Health and Denervation". bioRxiv. 2026-01-06T00:00:00.000Z. URL: https://doi.org/10.1101/2024.02.26.582103

[7] Devine, H., Roberts, M. J., Ziff, O. J., Hanna, M. G., Greensmith, L., Patani, R., Malik, B.. "Conserved Aberrant Developmental Trajectories of Human and Mouse SBMA Motor Neurons". bioRxiv. 2025-09-17T00:00:00.000Z. URL: https://doi.org/10.1101/2025.09.17.674754

[8] Biswas, S. R., Tomsick, P. L., Kelly, C., Lester, B. A., Milner, J. P., Henry, S. N., Soto, Y., Brindley, S., DeFoor, N., Morton, P. D., Pickrell, A. M.. "Impaired Complex I dysregulates neural/glial precursors and corpus callosum development revealing postnatal defects in Leigh Syndrome mice". bioRxiv. 2025-05-16T00:00:00.000Z. URL: https://doi.org/10.1101/2025.05.16.654318

[9] Fabres, R. B., da Silva, V. R. C. T., Zhang, S., Synowiec, S., Takada, S. H., Duan, J., Drobyshevsky, A.. "Comparative Analysis of Neonatal Hypomyelination Models Using Spatial Transcriptomics". bioRxiv. 2025-06-24T00:00:00.000Z. URL: https://doi.org/10.1101/2025.06.23.661209

[10] bourguiba, A., Gelin, m., Fail, A., Saillard, L., Bauche, S., Peccate, C., Meunier, P., Guesmia, Z., Mirabile, L. A., Perronnet, J., Lemaitre, M., Giordani, L., Falcone, S., Gentil, C., Pietri-Rouxel, F.. "Axonopathy in Duchenne Muscular Dystrophy limits microdystrophin gene therapy efficacy". bioRxiv. 2026-08-20T00:00:00.000Z. URL: https://doi.org/10.64898/2026.08.17.744857

[11] Xu, L., Yang, H., Leon, J., Li, X., Oehler, A., Modavi, C., Abate, A. R., Condello, C.. "Spatial transcriptome-metabolome integration reveals region-specific glial lipid dysregulation linked to Alzheimer's pathology". bioRxiv. 2026-10-08T00:00:00.000Z. URL: https://doi.org/10.64898/2026.01.23.701345

[12] Tsai, H.-H., Piya, S., Wang, J., Zhu, J., Hu, W., Gehrke, A. R., Cao, S., Guise, A. J., Chan, S. J., Sheehan, M., Chu, J., Ouyang, Z., Ryals, M., Lee, M., Wang, W., Zhao, E., Cullen, P., Challa, R., Marshall, E., Zeng, W., Kaeser-Woo, Y. J., Ehrenfels, C., Jandreski, L., McLaughlin, H., Carlile, T. M., Gagnon, J., Reynolds, T. L., Li, M., Li, K., Zhang, B.. "Spatial transcriptomics reveals heterogeneous cell-cell interactions among brain regions in a cuprizone model consistent with multiple sclerosis lesions". bioRxiv. 2025-02-08T00:00:00.000Z. URL: https://doi.org/10.1101/2025.02.07.637125