Research paper

A rise in NAD precursor nicotinamide mononucleotide (NMN) after injury promotes axon degeneration

M Di Stefano, I Nascimento-Ferreira, G Orsomando, V Mori, J Gilley, R Brown, L Janeckova, ME Vargas, LA Worrell, A Loreto, J Tickle, J Patrick, JRM Webster, M Marangoni, FM Carpi, S Pucciarelli, F Rossi, W Meng, A Sagasti, RR Ribchester, G Magni, MP Coleman, L Conforti

Cell Death and Differentiation 2015 Volume 22, Issue 5, pages 731–742

Tier 3 · Preclinical or mechanistic with translational relevanceNeurological

Source identifiers

DOI 10.1038/cdd.2014.164

PubMed PMID 25323584

Funding This work was funded by the Biotechnology and Biological Sciences Research Council Institute Strategic Programme Grant (to MPC, LC, INF), Alzheimer's Research UK grant ART/PG2009/2 (to LC, JP), a Faculty of Medicine and Health Sciences, University of Nottingham, nonclinical senior fellowship (to LC, MDS, WM, MM, FR), a grant from the National Institute of Dental and Craniofacial Research (1R01DE019487) to AS, a Marie Curie Intra European Fellowship within the 7th European Community Framework Program (to MDS, LC) and by the ERASMUS mobility program (AL). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Declared competing interests LC and MPC are inventors on a patent application filed by the Babraham Institute relating to this study. The remaining authors declare no conflict of interest.

Study snapshot

Design: Multi-modal preclinical mechanistic study combining pharmacological inhibition (NAMPT inhibitors FK866 and CHS-828), genetic manipulation (bacterial NMN deamidase expression; drug-resistant NAMPT G217R rescue), in vivo axotomy (mouse sciatic nerve; zebrafish larval sensory axons), ex vivo nerve explants, primary neuronal cultures (SCG, DRG), and nucleotide quantification by HPLC and mass spectrometry.
Population: No human subjects. Models used: mouse (C57BL/6 wild-type; YFP-H transgenic; WldS; Nmnat2gtE/gtE homozygous E18.5 embryos) primary neuronal explants and sciatic nerves; zebrafish larvae (48–54 h postfertilisation); HEK293T and PC12 cell lines for enzyme activity characterisation.
Sample: Varies by experiment. Representative: in vivo sciatic nerve nucleotide time-course n = 4 per time point; zebrafish laser axotomy n = 9–25 per FK866 concentration; SCG explant degeneration index n = 6–18 per group; ex vivo nerve NAD/NMN n = 10; nerve–muscle preparation electrophysiology n = 4–8 per group.
Intervention: Pharmacological: FK866 (1 nM–10 μM) and CHS-828 (NAMPT inhibitors); exogenous NMN (25 μM–1 mM); exogenous nicotinamide riboside (NR) and NAD. Genetic: ectopic expression of E. coli and S. oneidensis NMN deamidase (WT and catalytically inactive mutants); drug-resistant NAMPT G217R rescue; WldS transgenic background; Nmnat2 loss-of-function.
Endpoints: Axon degeneration index; time to axon fragmentation; NMN, NAD, NR, and adenine nucleotide (ATP/ADP/AMP) concentrations by HPLC and LC-MS/MS; Neuromuscular synapse function measured by end-plate potentials (EPPs) and miniature end-plate potentials (mEPPs) in FDB muscle preparations; Percent of surviving axons at defined post-injury time points; Whole-brain nucleotide levels in Nmnat2-null embryos

What the study found, in plain language

This 2015 primary research paper reports a finding that runs counter to the simple "more NMN is better" framing common in the supplement space: after nerve injury, NMN itself accumulates inside axons and actively promotes their degeneration. The paper was published in Cell Death and Differentiation, a peer-reviewed Nature Publishing Group journal, by a multi-institutional team led by groups at the University of Nottingham and the Babraham Institute (Cambridge, UK), with collaborators in Italy and the United States.

The background biology: axons depend on an enzyme called NMNAT2 to convert NMN into NAD+. NMNAT2 is a labile protein that has to be continuously delivered from the cell body down the axon. When an axon is cut, or when NMNAT2 is otherwise depleted, this delivery stops. NAMPT — the enzyme upstream of NMNAT2, which makes NMN from nicotinamide — is far more stable and keeps producing NMN. The result: NMN builds up in the injured axon with nowhere to go, because the enzyme that would convert it to NAD+ is gone.

Using three independent methods — the NAMPT inhibitor FK866, which blocks NMN synthesis; a bacterial NMN deamidase enzyme that scavenges NMN; and the well-established WldS protein — the authors show that preventing this NMN rise protects injured axons and their neuromuscular synapses from degeneration. Conversely, adding exogenous NMN reverses that protection. In injured mouse sciatic nerve in vivo, NMN begins rising within 12 hours of injury and reaches roughly 2.5× normal by 30 hours — approximately 24 hours before axons visibly fragment. The same protective effect of NAMPT inhibition was reproduced in zebrafish larvae in vivo.

The scope of the finding is specific and important to understand precisely. This paper documents a mechanism operating in a defined pathological context: axons that have lost NMNAT2 activity and cannot metabolise NMN. The authors themselves note that in other cellular contexts NMN is cytoprotective, because those cells retain NMNAT to convert NMN to NAD+. The paper's contribution is to identify NMN accumulation as an active driver of a specific type of nerve degeneration (Wallerian and Wallerian-like), and to identify NAMPT as a potential therapeutic target for peripheral neuropathies — including chemotherapy-induced peripheral neuropathy, which is the paper's proposed translational application.

Key findings

NMN rises before axons degenerate, in vivo. In transected mouse sciatic nerve, NMN begins rising within 12 hours of injury and reaches approximately 2.5× baseline by 30 hours — 24 hours before the first axons fragment.
NAMPT inhibition protects injured axons despite lowering NAD. The NAMPT inhibitor FK866 (1–100 nM) potently delayed degeneration of transected superior cervical ganglion (SCG) neurites, with protection optimal when the drug was added within 0–3 hours of axotomy. A second NAMPT inhibitor, CHS-828, showed the same protective phenotype.
Exogenous NMN reverses that protection. Co-administration of 25 μM to 1 mM NMN with FK866 dose-dependently restored rapid axon degeneration; the EC50 at 8 hours was 60 μM.
Genetic scavenging of NMN protects axons independently of NAD. Ectopic expression of E. coli NMN deamidase — which converts NMN to nicotinic acid mononucleotide (NaMN) without generating NAD — strongly delayed degeneration of cut SCG neurites. Catalytically inactive mutants of the enzyme showed correspondingly reduced protection.
NMN accumulation occurs in NMNAT2-null embryos. Whole brains of E18.5 embryos homozygous for a Nmnat2 loss-of-function allele showed clearly elevated NMN levels versus wild-type littermates, confirming the mechanism operates in vivo in the absence of physical injury.
Protection extends to neuromuscular synapses and to chemotherapy-induced neuropathy models. FK866 dose-dependently preserved evoked and spontaneous end-plate potentials in ex vivo tibial nerve/FDB muscle preparations and delayed vincristine-induced neurite degeneration in primary culture.
Protection is reproduced in vivo in zebrafish. In larval zebrafish, FK866 pretreatment delayed laser-induced axon degeneration up to fivefold, with no evidence of cell death or regeneration defects at the doses tested.
ATP and energy charge are preserved throughout the latent phase. Injured sciatic nerves show no early bioenergetic deficit despite falling NAD, arguing against the hypothesis that Wallerian degeneration is triggered by ATP depletion.

What this study can and cannot tell us

Readers should weigh the following when interpreting this paper's relevance to NMN supplementation:

Preclinical only. All findings are in mouse and zebrafish models, plus mammalian cell lines. No human data are reported. Whether the mechanism operates identically in human peripheral nerves at supplement-relevant systemic NMN exposures has not been directly tested.
Scope is a specific pathological context. The mechanism documented depends on NMNAT2 depletion — either through axon injury (Wallerian degeneration) or through genetic loss of NMNAT2. In healthy axons with intact NMNAT2 activity, NMN is converted to NAD+ normally, and the paper's own Discussion contrasts this against earlier reports in which NMN is cytoprotective in cell types that retain NMNAT. Extrapolation to intact healthy tissue is not supported by the paper's data.
Local axonal concentrations differ from systemic exposure. The effective concentrations in this system (EC50 ~60 μM in cultured neurites; ~4 nmol/g in vivo) represent local intra-axonal accumulation of endogenously synthesised NMN, not steady-state plasma NMN from oral supplementation. The paper does not measure plasma NMN or model oral bioavailability.
Author-acknowledged mechanistic uncertainties. The authors state they cannot completely rule out (a) a role for a related, potentially unknown metabolite of NMN, (b) off-target effects of FK866 including at neuromuscular junctions, or (c) a rise in NMN within Schwann cells contributing to the phenotype. The downstream effector — the protein or signal that NMN accumulation actually activates to drive degeneration — is not identified in this paper. Subsequent work from this group and others has implicated SARM1, though that mechanism is beyond this paper's scope.
Population-level implications remain open. The paper does not resolve whether individuals with subclinical peripheral neuropathy, chemotherapy-related nerve injury, hereditary NMNAT2 insufficiency, or other conditions of compromised axonal NAD metabolism could be adversely affected by exogenous NMN. It also does not address the reverse question — whether NMN supplementation could be neuroprotective in intact tissue via NAD+ generation. Both directions warrant human study.
Commercial disclosure. Two of the authors (LC and MPC) are named on a patent application filed by the Babraham Institute relating to this study — specifically to NAMPT inhibition as a strategy for axon protection. Their commercial interest runs in the opposite direction from NMN supplement manufacturers: they benefit if NMN accumulation is shown to be pro-degenerative in nerve injury. This does not invalidate the findings — the mechanistic data are strong and have been reproduced by independent groups — but it is disclosed on the paper and worth noting.