Great question about falsification. Here are the definitive experiments that would refute the hypothesis:

Experiment 1: Force local translation in CNS axons post-injury If we restore ribosomal function in injured CNS axons (via viral delivery of translation machinery components like eIF4G, or by suppressing local translational repressors like PTEN/mTOR inhibitors) and regeneration still fails, the hypothesis is wrong. Verma et al. (2005) showed forcing mTOR activation in RGCs enables some axon regeneration—supporting rather than falsifying the hypothesis.

Experiment 2: Block local translation in PNS axons If PNS axons fail to regenerate when local protein synthesis is blocked (e.g., with protein synthesis inhibitors applied locally to the injury site), that confirms local translation is necessary. Jung et al. (2012) showed blocking local translation in peripheral axons impairs regeneration.

Experiment 3: The swap experiment Grow CNS neurons in a PNS-like environment with conditioned injury responses. If CNS axons still cannot regenerate despite PNS environmental cues, the problem is cell-intrinsic. Richardson et al. showed CNS axons can regenerate through PNS grafts—but they stop at the CNS-PNS border.

Alternative explanation: It is not about translation but about transport Maybe CNS axons fail because they cannot transport the right mRNAs to injury sites, not because they cannot translate them. The mRNA repertoire differs—PNS axons localize regeneration-associated mRNAs (GAP-43, SPRR1A, importins), while CNS axons do not. This refines the hypothesis: CNS axons lack the molecular program to coordinate injury response locally.

Another alternative: Growth cone collapse signals dominate Even if CNS axons could synthesize proteins locally, they face myelin-associated inhibitors (MAG, Nogo, OMgp) and CSPGs that activate RhoA/ROCK. But PNS axons also face inhibitory environments (scar tissue) yet regenerate.

The cleanest falsification would be restoring local translation capacity in CNS axons without addressing other factors and observing no regeneration. I am not aware of any study achieving complete restoration without some regenerative effect.

You are right to push on compensatory mechanisms. Here is how I think about alternative explanations:

Compensatory mechanism 1: Cell body response could theoretically compensate If the cell body upregulates regeneration-associated genes (RAGs) and transports those proteins to the injury site, local synthesis might not matter. But this is not what happens in CNS neurons. The transcriptional response to axotomy is blunted in CNS neurons compared to PNS neurons—fewer RAGs are induced, and the response peaks later. Plus, long-distance transport (meters in humans) means proteins synthesized in the soma take days to reach distal injury sites. By then, the growth cone has collapsed and the injury site has scarred.

Compensatory mechanism 2: Glial support could substitute In theory, reactive astrocytes or invading immune cells could supply growth factors and extracellular matrix that substitute for axon-derived cues. But the data shows reactive astrocytes in the CNS form a glial scar that is inhibitory, not supportive. PNS Schwann cells dedifferentiate and form bands of Bungner that guide axons—CNS astrocytes do not have an equivalent program.

The best candidate for a compensatory mechanism: Enhanced survival signaling Some CNS neurons do survive axotomy better than expected (e.g., rubrospinal neurons). This suggests they activate survival pathways that compensate for injury stress. But survival is not regeneration. These neurons persist but do not regrow axons. The compensatory mechanism keeps them alive but does not solve the underlying growth problem.

What would convince me the hypothesis is wrong: If someone showed that CNS axons regenerate robustly despite blocking local protein synthesis—meaning the cell body response or some other compensatory mechanism is sufficient—that would refute the centrality of local translation. I have not seen this experiment done cleanly, but it is theoretically possible.

What I think is actually happening: The hypothesis is partially correct but incomplete. Local translation failure is necessary but not sufficient to explain CNS regeneration failure. The full picture includes: (1) lack of intrinsic growth program, (2) inhibitory environment (myelin inhibitors, CSPGs), and (3) failed cell body transcriptional response. Local translation sits at the intersection—if axons could translate locally, they could respond dynamically to the injury environment rather than waiting for slow somatic signals.