Compartment-specific PLP1, KCNQ2 and OXPHOS signatures in the NASA RR-3 brain dataset: spatial transcriptomic observations (manuscript in preparation)

AWG Projects Brain AWG Open Projects
, ,
1

Hi everyone,

For those who haven’t seen my earlier post: I’ve been working on the VEN Fatigue Hypothesis, the argument that von Economo Neuron circuit degradation may underlie the non-monotonic cognitive trajectory seen in long-duration spaceflight, with early improvement followed by late decline. That work, with @windymc and @nilufarali, is to be submitted to Life Sciences in Space Research.

Building on that, I’ve been exploring the NASA RR-3 spatially-resolved multiomics dataset (Masarapu et al. 2024, Giacomello lab, Nat Commun 15:4778) using a targeted panel of genes associated with VEN biology, cortical projection neurons, myelination and excitability. Three observations emerged from the analysis.

Bidirectional Plp1 across compartments. Plp1 is upregulated in cortical neurons (ST Cluster 9; logFC +0.89, FDR = 0.041) but downregulated in oligodendrocytes (MO Cluster 3; logFC -0.37, FDR = 0.043). Bulk data from previous studies showed an apparent increase, interpreted as compensatory remyelination. The spatial data suggest that bulk-level analyses may mask opposing cell-type-specific responses.

KCNQ2 downregulation in the choroid plexus. Kcnq2 is significantly downregulated in ST Cluster 16, annotated as choroid plexus (logFC -0.65, FDR = 0.016). I haven’t found previous discussion of this observation in the RR-3 paper or related analyses.

A possible upstream connection to the OXPHOS signatures. Masarapu et al. reported severe OXPHOS failure in cortical bottom layers (padj = 3.39 x 10-6). I’m proposing, as a hypothesis, that altered choroid plexus function could influence CSF dynamics and, consequently, glymphatic clearance, which could contribute to metabolic stress in the most energy-demanding cortical neurons. That chain isn’t demonstrated here, it’s a prediction the data suggest and that future work could test.

Together, these observations suggest that spatial transcriptomics can reveal cell-type-specific responses that are difficult or impossible to distinguish in bulk transcriptomic analyses, generating mechanistic hypotheses that bulk data alone cannot resolve.

All reported values were extracted directly from the published supplementary DEG tables.

Resources

I’m writing this up as a manuscript. If you’re working with RR-3, spaceflight neuroscience, spatial transcriptomics, or have alternative interpretations of these observations, I’d be very interested in discussing these observations, alternative interpretations, or related datasets. Feel free to reach out: esila2.keskin@live.uwe.ac.uk

Esila Keskin

2 Likes