Axons degenerate when they are cut from their cell body. However, neurons from mice are protected from degeneration when they overexpress either Nmnat1, which synthesizes NAD, or the Wlds fusion protein that contains Nmnat1.
Wang et al. found that overexpression of Wlds or Nmnat1 prevented a decrease of NAD, normally seen in axons after severing, and delayed degeneration. Moreover, addition of NAD to culture media delayed axonal degeneration, even when it was added to an axonal segment already isolated from its soma. This suggests that NAD acts locally rather than via transcription. Because NAD is essential for ATP synthesis, and a drop in ATP parallels the drop in NAD in severed axons, the team asked if exogenous pyruvate could delay degeneration. The energy-rich chemical did protect axons, indicating that NAD may be required to sustain ATP levels necessary to maintain the integrity of the cell boundary.
So what is the difference between this work and the earlier study? It could be that Wang et al. used higher concentrations of NAD and thereby bypassed a need for transcriptional activation of SIRT1. No difference after NAD addition in neurons from wild-type and SIRT1-deficient mice makes that less likely. Instead, Wang et al. point to the contrast between isolated axons in their study and surrounding glial cells in the Araki et al. study.
The real test of NAD's protective abilities will come when Wang et al. start to look at neurons in the brains of animals. Such cells are postmitotic and can't synthesize NAD from tryptophan. If they start to lose NAD for some reason, such as axonal injury, they may become more susceptible to degeneration—and adding it back may be protective.