Hypothesis\n\nAging triggers a progressive decline in spliceosomal fidelity, leading to aberrant inclusion of cryptic exons in transcripts encoding complement regulators (e.g., C1q) and mitochondrial quality‑control proteins. This spliceosomal stress amplifies C1q secretion from neurons themselves, converting a homeostatic surveillance signal into a cell‑autonomous 'eat‑me' tag that marks metabolically inefficient neurons for microglial phagocytosis. Consequently, neuronal loss in aging reflects an active quality‑control loop where spliceosome‑driven proteostatic collapse feeds complement‑mediated eviction.\n\n## Mechanistic Reasoning\n\n1. Spliceosome vulnerability – Core spliceosomal components (SF3B1, U2AF65) show reduced expression and increased oxidative damage with age, causing intron retention and exon skipping, particularly in long transcripts with weak splice sites (7).\n2. Cryptic exon inclusion in C1q regulator – Bioinformatic analysis of aged human brain RNA‑seq reveals frequent inclusion of a conserved Alu‑derived exon in the 5′UTR of C1q mRNA that enhances translation efficiency and secretory pathway targeting (novel prediction).\n3. Neuronal C1q secretion – Neurons, not just microglia, possess the secretory machinery for C1q; increased neuronal C1q elevates synaptic tagging independent of microglial activation (1, 2).\n4. Feedback to proteostasis – Elevated neuronal C1q activates the complement receptor C3aR on the same neuron, triggering calcium influx that impairs mitochondrial ATP production, further stressing the spliceosome (6).\n5. Net outcome – Neurons with high bioenergetic demand and low splicing fidelity accumulate aberrant proteins, release more C1q, and are preferentially engulfed, yielding a selective pruning that optimizes cortical circuitry under declining ATP.\n\n## Testable Predictions\n- Prediction 1: In aged mouse cortex, neurons exhibiting high levels of intron retention in C1q pre‑mRNA will show increased intracellular C1q protein and secreted C1q in the extracellular space.\n- Prediction 2: Genetic or pharmacological enhancement of spliceosomal fidelity (e.g., overexpression of SRRS2 or treatment with the spliceosome‑stabilizing compound ISRIB) will reduce neuronal C1q secretion and attenuate age‑related neuronal loss without affecting microglial numbers.\n- Prediction 3: Blocking neuronal C1q secretion (using a neuron‑specific C1q conditional knockout) will prevent complement‑mediated synapse elimination and preserve cognitive performance in aged animals, even when global complement levels remain high.\n\n## Experimental Approach\n- Single‑nucleus multi‑omics (snRNA‑seq + snATAC‑seq) from young (3 mo) and aged (24 mo) mice to quantify intron retention events in C1q and complement pathway genes, correlated with neuronal C1q protein measured by proximity ligation assay.\n- Spliceosome rescue: AAV‑mediated neuronal overexpression of SF3B1 or small‑molecule spliceosome modulator (e.g., H3B‑8800 at low dose) in aged mice; assess C1q secretion via ELISA of cortical interstitial fluid and neuronal density via NeuN staining.\n- Neuron‑specific C1q KO: Cross Synapsin‑Cre mice with C1q floxed line; evaluate synaptic density (synaptophysin), microglial phagocytosis (CD68+ lysosomal puncta), and behavior (Morris water maze) in aged cohorts.\n- Falsification: If spliceosome rescue does not lower neuronal C1q or neuronal loss persists, or if neuronal C1q KO fails to protect synapses, the hypothesis is refuted.\n\n## Broader Implications\nThis hypothesis reframes age‑related neurodegeneration as a maladaptive extension of a developmentally pruning pathway, where loss of splicing fidelity converts a beneficial quality‑control mechanism into a driver of circuit erosion. Therapeutically, targeting spliceosomal fidelity or neuronal complement secretion could decouple harmful eviction from essential synaptic remodeling, offering a strategy to preserve cognition without globally suppressing complement.\n\nReferences are embedded via the markdown links above.