A narrative review published in Frontiers in Cell and Developmental Biology synthesizes what multi-omics research has revealed about stem cell-mediated repair after traumatic spinal cord injury, and it arrives at a sobering conclusion: although transplantation studies show that grafts can protect tissue, add myelin, and communicate with host circuits in selected models, no single study has yet traced the full sequence from a cell product’s molecular features to a patient’s recovery. The review, authored by Haotian Xu, Jun Wu, and Chao Cheng of Wuxi People’s Hospital in China, examines neural stem and progenitor cells alongside neural products derived from pluripotent stem cells, using mesenchymal stromal cells as a comparison for repair driven mainly by secreted factors.
The central argument is that the effect of a graft depends on the starting cell product, how donor cells respond to the hostile lesion environment, and whether they ultimately contribute to useful function. Regional identity and developmental stage matter. Spinal and forebrain neuroepithelial grafts behaved differently in engraftment and repair, and in one rodent model, reversibly suppressing activity in a spinal graft reduced locomotor performance, supporting a relay contribution. A separate study linked spinal specification of human pluripotent stem cell-derived neural stem cells with graft integration. Yet the best regional match for clinical lesions remains unknown.
One of the review’s most striking findings concerns fate redirection. Using donor-specific RiboTag profiling, which captures ribosome-associated RNA from graft-derived cells, researchers showed that neural progenitors acquired a wound-repair astroglial phenotype in mouse lesions, including subacute spinal cord injury. The injured host environment can literally rewrite what transplanted cells become. The authors caution that this does not mean every neural graft turns astroglial, nor that the change improves motor function, but it underscores why molecular profiling before and after transplantation is essential.
Developmental stage presents another concrete variable. Comparing spinal neural progenitor preparations from different stages, one research team found differences in neuronal composition and host axon growth; later-stage grafts worsened thermal hypersensitivity, and transplantation did not significantly improve locomotor recovery relative to comparison groups in that model. Anatomical changes did not amount to overall functional benefit, illustrating why studies aimed at motor repair must assess more than locomotion. Human neural grafts also matured over prolonged periods in primates, meaning product comparisons need follow-up long enough to distinguish persistent developmental differences from temporary delays.
The lesion environment itself is a dynamic participant. Single-cell studies identify immune, astroglial, fibrotic, and vascular cell populations after injury, and their locations and changes over time provide a baseline for examining donor survival and host responses. Scar-forming astrocytes can support axon growth under growth-stimulating conditions, and microglia at the lesion border help organize a neuroprotective scar, so broadly suppressing these responses could remove tissue-protective functions. In a more direct transplantation study, neural precursor grafts changed phagocyte responses and reduced secondary damage after mouse contusion, suggesting grafts may protect tissue before or while neural replacement occurs.
On the molecular measurement side, the review catalogs what each assay layer can and cannot answer. Single-cell RNA sequencing resolves donor and host populations but reflects RNA abundance, not cellular activity. Spatial transcriptomics locates cells relative to lesion borders and blood vessels while measuring only one molecular layer. Pairing RNA with chromatin accessibility, proteins, or metabolites adds genuinely distinct information. One study of human iPSC-derived spinal neural progenitors profiled the product before transplantation and donor nuclei one week after grafting into injured immunodeficient mice, alongside host protein and tissue measures, but longer follow-up and intact immune systems are needed to assess later maturation.
Metabolomics adds biochemical context that RNA alone cannot provide. Spatial maps of RNA and metabolites identify lesion regions where researchers can test donor survival and maturation, and additional experiments tested copalic acid-associated inflammatory responses and IGF2-associated protection and recovery. However, regional enrichment does not establish the cellular source, and neither measurement alone shows how quickly a pathway operates. Whether altered metabolism helps graft cells survive remains unknown.
Functional evidence is the review’s most carefully parsed territory. Physiological recordings showed activity within graft networks and communication with host circuits. In a mouse contusion study, grafted spinal neurons connected to hindlimb motor circuits, and chemogenetic activation triggered muscle activity in some animals, yet transplantation alone did not improve locomotion. Removing human neural stem-cell grafts abolished locomotor benefit in injured immunodeficient mice, showing graft cells were required in that model, though the experiment could not separate neuronal from glial contributions. These conclusions apply only to the products, lesions, and outcomes tested.
Clinical translation faces steep barriers. The phase I CELLTOP study evaluated an autologous adipose-derived stromal product in ten participants primarily for safety, reporting no serious adverse events over 96 weeks, with neurological outcomes remaining exploratory. In 2026, an open-label study treated four participants with subacute complete cervical injury using iPSC-derived neural progenitors, reporting no tumors or graft-related adverse events during two to four years of follow-up, though graft survival and integration were not directly confirmed, and four participants provide little information about uncommon risks. A safety warning comes from earlier work: one human iPSC-derived neural preparation initially improved function but subsequently formed tumors, with transcriptome analysis implicating mesenchymal transition, demonstrating that early benefit cannot establish long-term safety.
The authors’ roadmap is deliberately restrained. They call for studies that relate the administered product to later donor-cell states, host responses, and functional outcomes within the same experiment, recording both time since injury and time since transplantation. Candidate molecular markers must predict outcomes across independent cell lines, manufacturing batches, and recipients before serving as potency assays. Because animals and donors, rather than individual cells, define independent replication, profiling more cells from one preparation does not add transplantation outcomes. Longitudinal sampling, selective perturbation, and independent replication, they conclude, should take priority over collecting more molecular data without a defined repair question.
Subject of Research: A narrative review of multi-omics evidence on stem cell graft behavior, host niche responses, and functional outcomes after traumatic spinal cord injury.
Article Title: Multi-omics insights into stem cell-mediated regeneration after traumatic spinal cord injury: cell fate, niche remodeling, and functional integration
Article References: Xu, H., Wu, J., & Cheng, C. (2026). Multi-omics insights into stem cell-mediated regeneration after traumatic spinal cord injury: cell fate, niche remodeling, and functional integration. Frontiers in Cell and Developmental Biology, 14, Article 1994054. https://doi.org/10.3389/fcell.2026.1994054
Image Credits: AI Generated
DOI: 10.3389/fcell.2026.1994054
Keywords: spinal cord injury, stem cell transplantation, neural progenitor cells, multi-omics, single-cell transcriptomics, spatial transcriptomics, graft-host interactions, neural regeneration, mesenchymal stromal cells, clinical translation, insights, stem
News Source: Drew Townsend. (October 29, 2026). Mapping How Stem Cell Grafts Behave Inside the Injured Spinal Cord. Scienmag.
