Alzheimer´s disease (AD) spreads through the brain in a non-random manner indicating propagation along connecting fiber tracts. The molecular mechanisms driving the spread of the pathology remain largely unknown. We have recently discovered an unexpected mode of long-range transmission of the pathological effects of Aβ oligomers from axons to cell bodies that relies on intra-axonal translation of ATF4. These findings support a model in which retrograde transport of locally translated proteins leads to pathological, transcriptional changes in the neuronal cell bodies.

Local translation enables axons to react to extracellular stimuli in an acute manner. Intra-axonal protein synthesis is best understood in the context of neurodevelopment where it plays crucial roles in growth cone behavior, axon pathfinding and retrograde signaling. On the other hand adult axons have long been thought to be translationally inactive. However, high-throughput analyses have revealed that mature axons have a more complex and dynamic transcriptome than expected, especially under pathological conditions, and local translation is required for axonal regeneration upon nerve injury, it improves motor function in a mouse model of spinal muscular atrophy (SMA), and we have recently found that it mediates Aβ-induced neurodegeneration in vitro and in vivo.

ATF4 mediates the cellular responses upon activation of the integrated stress response (ISR) by inducing the transcription of genes involved in cell death or survival, but it can also repress long-term potentiation under normal conditions acting as a CREB-1 antagonist. Our results establish that Aβ application selectively to axons triggers retrograde somatic degeneration through ATF4 axonal translation. Thus axonally-synthesized ATF4 could be targeted in order to prevent or slow down the spread of AD pathology throughout the brain. However, ATF4 is also translated in the neuronal soma of granule cells in the dentate gyrus (DG) following Aβ infusion in the mouse hippocampus in vivo. Atf4 knockdown increases Aβ-mediated neurodegeneration in the DG, suggesting that in this particular case ATF4 would rather be involved in a protective, adaptive response to Aβ exposure. This raises the intriguing possibility that ATF4 elicits distinct responses based on its translation at the subcellular level. Such differential responses could be explained by the availability of potential ATF4 binding partners in axons and cell bodies upon Aβ exposure. Here we characterize the role of ATF4-related transcription regulators in axons, whose interaction with ATF4 could be targeted in order to prevent axon-to-soma degeneration triggered by amyloid peptides.