Cerebellar pathology in SMA is autonomous — platform expands tissue coverage
The finding
Gerstner et al. (Leipzig, Brain 2025, PMID 40585211) and Cottam et al. (Brain Pathology 2025, 35(6):e70025) converge on one uncomfortable point: cerebellar pathology in SMA is autonomous. It is not rescued by motor-neuron-targeted correction, it contributes to motor AND cognitive deficits, and it has been under the radar because the field's omics queries have been dominated by spinal motor neurons and muscle.
O36 at the 2026 SMA Congress (Budapest) presented Purkinje-cell p53-dependent death and parallel-fiber dysfunction. Cottam showed that motor-neuron rescue does not repair cerebellar defects.
This connects directly into our existing p53 → ROCK work
The p53-dependence of Purkinje cell death is not a new axis — it is a cell-type extension of a mechanistic cascade our platform already captured from SMA motor-neuron literature.
Three existing claims tie SMA p53 biology into one chain:
- Claim #30084 — "Impaired SMN-p53 interaction may lead to apoptosis and thereby explain motor neuron death in SMA" (SMN ↔ TP53 direct interaction).
- Claim #28407 — "Motor neuron apoptosis following nerve injury is p53-dependent and mediated by Fas death receptor activation" (p53 → Fas → Bax → caspase-8).
- Claim #46341 — "Axonal pathology is associated with an abnormal signalling pathway consisting in p38 MAPK activation leading to Mdm2-mediated p53 degradation, followed by ROCK activation" — the critical cascade link. This places ROCK activation downstream of p53, which means ROCK inhibition rescues the same pathology from the opposite end.
Gerstner's Purkinje observation slots into this cascade without modification: SMN loss → impaired SMN-p53 interaction → p38 MAPK activation → p53 / Mdm2 imbalance → ROCK activation → cytoskeletal collapse → cell death. Spinal motor neurons and cerebellar Purkinje cells converge on the same axis.
Our platform already hits this cascade at three points
| Layer | Drug | Target in cascade | In platform |
|---|---|---|---|
| Upstream stress | MW150 (CNS-penetrant p38 MAPK inhibitor) | MAPK14 | /drugs, /pathway MAPK14 node |
| Mid-cascade | Pifithrin-α (p53 inhibitor) | TP53 | /pathway TP53 node — already listed as a repurposing candidate |
| Downstream | Fasudil (ROCK2 inhibitor) | ROCK2 | /pathway, ROCK2 apo 100 ns MD baseline |
This is the kind of cross-connection the platform was built to surface. Three drugs, three layers of the same cascade, now each one is also a candidate for cerebellar rescue — not just spinal motor neurons.
Why this matters operationally
- Zero new compute is required to extend Fasudil evidence into cerebellum — the mechanistic bridge is already published (claim #46341) and only needed a new finding (Gerstner 2025) to trigger the extension.
- Pifithrin-α moves up our priority list: it acts in the middle of the cascade and hits cerebellum and spinal cord at the same layer. Worth a dedicated
/hypothesescard. - MW150 complements Fasudil as an upstream combination partner — relevant both for the 3-mechanism combo memo and for the newly-added NRF2/KEAP1 redox campaign (p38 and NRF2 both sit at the oxidative-stress node).
What we added today
- New pathway node CEREBELLUM on
/pathway, wired to ROCK2, LIMK1, and TP53 with explicit edges. - GEO dataset candidates for ingestion: GSE102204 (SMA whole-brain RNA-seq) and SCP795 (wild-type cerebellum atlas, 611,034 nuclei).
- This news post (cross-connecting Gerstner 2025 into our existing p53/ROCK/p38 platform evidence).
Open question
Are LIMK2 and ROCK2 transcriptionally dysregulated in SMA cerebellum the way they are in SMA spinal cord? A single re-analysis of GSE102204 (or Gerstner's deposited transcriptome when it becomes public) would answer this. If yes, the entire three-drug cascade above becomes a whole-CNS proposition rather than a spinal-cord-only one.