← BioTransfer GEO Dataset Finder
GEO series

Cerebrospinal fluid-linked periventricular gradient pathology in multiple sclerosis associates with ependymal cell reactivity and cell adhesion disruption

GSE283715 Homo sapiens Genome binding/occupancy profiling by high throughput sequencing; Expression profiling by high throughput sequencing 18 samples Submitted 2026/07/10 Platform GPL24676
Summary
Cerebrospinal fluid (CSF)-adjacent brain surfaces – namely the subpial cortical and ependyma-adjacent periventricular regions – are uniquely vulnerable to diffuse pathology in multiple sclerosis (MS). So-called surface-in gradients, which predict disease relapses independent of white matter lesions, are thought to arise from CSF-borne cytotoxic factors and represent a distinct therapeutic target. Yet, how such factors access the brain, particularly at the ventricular border, remains unclear. While ependymal injury in MS has been suggested, a comprehensive assessment of cell health in the disease is lacking. Using ultra-high-field MRI-guided histology, electron microscopy, and single-nucleus multiomics, we deeply profiled human ependymal cells in MS. We found that ependymal pathology correlates with periventricular gradients and involves an immune-reactive state marked by transporter and junctional protein dysregulation. We then further defined the gene regulatory networks underpinning the MS ependymal state, predicted ligands known to be enriched in MS CSF that could drive the emergence of this state, and tested one candidate in vivo. IFNγ increased murine ependymal permeability and conditional knockout of its receptor (Ifngr1) reversed this effect. These findings implicate ependymal dysfunction in periventricular pathology and highlight how CSF ligands modulate ependymal function and regional inflammation.
This dataset
Download

Direct links to NCBI, no account and no request form: the whole study as GSE283715_RAW.tar, processed values as the series matrix, the supplementary file directory, and per-sample supplementary files for any of the 18 samples. Raw sequencing reads are also available from ENA.

Also filed as BioProject PRJNA1195083. Searching any of these in the dataset finder brings you back here.

Study design
18 conditions, each sampled once — no replicated groups

Read from 18 sample titles: 18 distinct titles with little repetition. Check it against the sample list below before relying on it.

Samples in this study
Similar datasets

Search all human ChIP / ATAC / CUT&Tag datasets in GEO →

Share this dataset

Metadata from NCBI GEO, cached and refreshed periodically — the NCBI page above is authoritative. Downloads link straight to NCBI/ENA; nothing is proxied through BioTransfer.