Exosome based analysis for Space Associated Neuro-Ocular Syndrome and health risks in space exploration

The study addresses how exosome-derived RNA from accessible biofluids can be used to longitudinally monitor molecular changes related to spaceflight exposures, with a focus on Spaceflight-Associated Neuro-ocular Syndrome (SANS) and intracranial pressure (ICP)-related conditions. The context is the need for deep molecular characterization in astronauts exposed to microgravity, radiation, altered gases, stress, and isolation. Exosomes, shed by all living cells and present in plasma, urine, and CSF, contain RNA reflective of their tissue of origin. The purpose is to establish an exosome-based transcriptomic platform that can profile whole RNA transcriptomes from multiple biofluids, enabling pathway mapping and risk characterization for SANS. The importance lies in creating minimally invasive, logistically feasible monitoring tools for astronauts and patients with ICP-related disorders, informing risk assessment and potential diagnostics.

Background highlights include: (1) SANS presents with optic disc edema, optic nerve sheath distension, increased retinal and choroidal thickness, cotton wool spots, posterior globe flattening, and decreased near visual acuity, with findings reproduced in head-down tilt bed rest analogs. The role of elevated ICP remains unresolved in astronauts; genetic determinants in one-carbon metabolism may contribute to SANS susceptibility. (2) Prior pilot work showed inflammatory transcriptional signatures in CSF exosomes from idiopathic intracranial hypertension (IIH) patients, suggesting partial pathophysiological overlap with SANS. (3) Most exosome RNA studies focus on small RNAs; here the authors emphasize long RNA diversity comparable to tissue RNA-Seq. (4) Microgravity alters endothelial cell transcriptomes and depresses pro-inflammatory cytokine secretion, supporting the utility of transcriptomic monitoring during spaceflight. These prior findings frame the rationale for broad, long RNA exosome profiling across biofluids to study ICP-related biology relevant to SANS.

Study population and sampling: Neurological patients (n=3) requiring therapeutic lumbar puncture (LP) were recruited under IRB approvals (JSC and Baylor College of Medicine) with informed consent. At Visit 1 (pre-treatment, high ICP): whole blood, plasma, and CSF were collected; ICP opening pressure was measured via manometer over 5 minutes prior to therapeutic CSF drainage. At Visit 2 (post-treatment, approximately 1 month later, reduced ICP): whole blood, plasma, and urine were collected. Additional controls included four pools of normal-ICP CSF from a biobank and six de-identified normal control urine samples. Samples were pre-filtered (0.8 µm), aliquoted, and stored at -80 °C prior to processing. Biofluid volumes processed: 2 ml plasma, 20 ml urine, 3 ml CSF for EV isolation; whole blood collected in PAXgene tubes for intracellular RNA comparison. EV isolation: Exosome Diagnostics performed centralized extraction of extracellular vesicles (EVs, including exosomes), isolating lipid vesicles <800 nm. RNA-Seq library preparation: Total EV RNA and whole-blood RNA were DNase-treated; ERCC synthetic spike-ins were added. Reverse transcription used random hexamers and oligo-dT. Second-strand synthesis and adapter addition were PCR-based. Libraries were purified with AMPureXP, with enzymatic ribosomal cDNA depletion, amplified, quantified (Bioanalyzer High Sensitivity, Qubit), pooled, and sequenced on Illumina NextSeq500 (2×150 bp). An exome hybrid-capture arm (human whole exome plus UTR) was tested for 2 ml plasma and 10 ml urine libraries (1 h hybridization; streptavidin magnetic capture; PCR; quantified; NextSeq550 2×150 bp). Bioinformatics: Reads aligned to hg38 (Gencode v25) using STAR v2.5.2a; transcriptome alignments quantified with Salmon v0.8.1 (bias corrections). Transcript-level TPM and counts were aggregated to gene-level via tximport. Differential expression (DE) used DESeq2 with covariates; p-values adjusted by Benjamini–Hochberg; significance at adjusted p<0.05. Reads <30 nt were not mapped. Gene ontology and GSEA employed WebGestalt and MSigDB. Additional feasibility studies: (1) Comparison of urine collection hardware: pooled healthy urine split into standard cups vs NASA Sarstedt tubes, stored 24 h at 4°C; EV isolation and long RNA-Seq performed; mapping metrics and gene expression compared. (2) Dry preservation of cDNA libraries: libraries desiccated and stored at room temperature for at least 1 month using DNAstable, Whatman FTA elute, or GenTegra-DNA; sequencing metrics pre- vs post-desiccation compared; DNAstable selected for further use.

• EV RNA content and sequencing quality: Majority of reads mapped to the transcriptome across biofluids; whole blood showed higher intronic reads consistent with pre-mRNA; CSF had fewest mappable reads, likely due to fragmentation; ribosomal RNAs were effectively depleted. ERCC spike-ins showed Pearson correlation >0.9 across samples and a dynamic detection range of ~5 orders of magnitude. Plasma and urine EV RNA each detected over 20,000 genes on average at the given depth. Genes with ≥80% exon coverage: urine EV RNA detected 4,448 genes; plasma 2,245; CSF 334. PCA showed clear separation by biofluid type. Exome capture increased proportions mapping to transcriptome and protein-coding genes relative to total RNA-Seq.
• Whole blood vs plasma for biomarkers: In whole blood, HBA1, HBB, and HBA2 comprised nearly 75% of informative transcriptome reads, reducing diversity for biomarker discovery; these genes were much less dominant in plasma EVs. DE between pre- and post-treatment: whole blood had only 3 significant DE genes; plasma EVs had 185 DE genes, predominantly downregulated post-treatment, with patient-specific transcriptome differences evident. GSEA of plasma DE genes indicated upregulated post-treatment sets related to adhesion, extracellular structure, binding, and migration; downregulated sets involved immune signaling and neutrophil-mediated immunity.
• CSF high-ICP vs normal-ICP: CSF EV RNA showed large expression differences. Compared to high-ICP CSF, normal-ICP CSF had 777 genes upregulated and 134 downregulated. GO terms enriched among genes higher in normal-ICP included leukocyte/lymphocyte activation, cytokine production/regulation, lymphoid organ development, adaptive immune response—indicating decreased immune-associated gene expression in high-ICP patient CSF. PCA strongly separated high-ICP from normal-ICP CSF.
• Urine post-treatment vs controls: Only modest differences were observed (120 genes upregulated and 89 downregulated in controls relative to post-treatment patients). No significant GO categories; PCA did not separate groups, consistent with clinical resolution post-treatment.
• NASA urine collection compatibility: Urine collected in NASA Sarstedt tubes yielded similar quality to standard cups: ~75% reads mapped to transcriptome; ~75% of transcriptome reads mapped to protein-coding genes; no significant expression differences—compatible for in-flight use.
• Dry storage of cDNA libraries: Desiccated libraries stored at room temperature for at least one month produced sequencing metrics consistent with pre-desiccation; DNAstable, Whatman FTA elute, and GenTegra-DNA performed similarly, with DNAstable selected for operational use.

The findings support exosome-based whole-transcriptome profiling from plasma, urine, and CSF as a feasible, informative approach to monitor systemic and tissue-enriched molecular changes relevant to ICP-associated conditions and SANS. Plasma EV RNA provided broad, system-wide information with greater biomarker discovery potential than whole blood (which is dominated by globin transcripts). Urine EV RNA offered high-quality gene coverage enriched for urinary tract signals; CSF EV RNA, although lower yield, was enriched for brain and immune-related genes and showed pronounced immune pathway suppression in high-ICP states. These observations align with prior microgravity-induced transcriptional changes in endothelial cells and suggest that exosome RNA profiling can capture relevant pathophysiology noninvasively. Compatibility of NASA's in-flight urine collection hardware and successful dry preservation of cDNA libraries bolster the platform's operational viability for longitudinal astronaut health surveillance, enabling creation of a stable, freezer-independent “medical library” for retrospective and in-mission analyses. Together, the approach advances the capability to study SANS mechanisms, assess ICP-related risks, and potentially develop noninvasive diagnostics for conditions with impaired CSF dynamics.

This proof-of-concept establishes that exosome-derived long-RNA transcriptome profiling from plasma, urine, and CSF is technically robust, biofluid-specific, and suitable for biomarker discovery relevant to SANS and ICP-related disorders. It demonstrates: (1) distinct, high-quality RNA signatures across biofluids, with plasma offering superior systemic biomarker potential; (2) compatibility of NASA’s in-flight urine collection tubes with exosome RNA-Seq workflows; and (3) feasibility of desiccated cDNA library storage at room temperature, enabling a longitudinal genomic/transcriptomic repository for astronaut health surveillance. Future work should include larger cohorts for validation, inclusion of samples at diagnosis (e.g., urine during high-ICP), optimization of neuronal exosome enrichment strategies, extended stability testing for desiccated libraries, and application to longitudinal astronaut studies to elucidate SANS mechanisms and develop actionable, noninvasive diagnostics.

• Small pilot cohort (n=3 patients) limits statistical power and generalizability; GSEA and GO findings require validation in larger studies.
• Sampling constraints: CSF obtained only at high-ICP; urine collected only post-treatment; lack of paired urine at high-ICP limits conclusions about urine biomarkers at diagnosis.
• CSF EV RNA had lower yields and fewer mappable reads, potentially restricting gene detection and coverage.
• Reads shorter than 30 nt were not mapped, potentially excluding some small/fragmented RNAs.
• Exome-capture enhancements were explored but principal analyses focused on total RNA-Seq; further optimization may refine capture of informative transcripts.
• IIH is an imperfect analog for SANS; translation to spaceflight conditions requires astronaut cohort validation.
• Dry-storage stability was demonstrated for at least one month; longer-term shelf life remains under evaluation.