Suppl. Fig. 1. FK866 treated neurites remain healthy for at least three days. (A) Neurites of SCG explants cultured for 7 days, then left untreated (a-c and g-j), or treated with 100nM FK866 (d-f and k-m) for 3 daysor for 4 daysas indicated. Neurites wereimaged at a proximal (a-f) and at a distal (g-m) site from the cell body mass. Neurites started to degenerate 4 days after FK866 addition to the culture medium. This result indicates that 3-4 fold depletion of NAD is not sufficient to induce axon degeneration within this timeframe. 1nM FK866 confers axon protection after injury. (B) Bright field images of SCG untreated (top panels) or treated with 1nM FK866 for 1 day (bottom panels), then cut and imaged at the indicated time points. (C) Bright field images of SCG untreated (top panels) or treated with 1nM FK866 for 3 days (bottom panels), then cut and imaged at the indicated time points showing a still detectable protective effect on neurites 8h after cut. FK866 does not affect NMNAT2 rapid turnover.(D) Western blot of cut neurites, treated with 100nM FK866 or left untreated as indicated, and collected at the indicated time points shows that FK866 fails to prevent NMNAT2 turnover (n=3, one representative experiment shown).
Suppl. Fig. 2. FK866 delays axon degeneration of injured DRG neurites by fourfold. (A) DRG explants were treated for the indicated time with 100nM FK866, then whole explants (left panel), or the cell bodies (middle panel) and neurite fractions (right panel) were separately collected. NAD was determined with an HPLC-based method (see methods; n=3, Mean and SD are shown). (B) DRG neurites were cut and left untreated or treated with 100nM FK866 just after cut. Images were acquired at the indicated timepoints. (C) Degeneration index of experiments in (B) was calculated from 3 independent experiments and three to four fields per experiment were quantified. The results show that there is a time effect (P < 0.0001) and the effect of treatment is highly significant (n = 10, Mean ±SEM, two-way repeated measures ANOVA followed by Bonferroni post-hoc test, ****p<0.0001).
Suppl. Fig. 3. CHS-828 delays axon degeneration of injured SCG neurites. (A) SCG neurites were left untreated or treated with CHS-828 at the indicated concentrations the day before or immediately after transecting them. Images were acquired at the indicated time points after transection. Degeneration index was calculated from three or four fields in two or three different experiments for CHS-828 preincubated (B) or added just after cut (C) (Mean ±SEM, n = 8-14, one-way ANOVA followed by Bonferroni post-hoc test, *, p<0.05, **, p<0.01, ***p<0.001). FK866 and CHS-828 pharmacological target NAMPT is present within axons and it is stable for at least 24hafter injury. The presence of NAMPT in neurites was confirmed both by immunocytochemistry and western blot analysis. (D) NAMPT immunofluorescence (green) of dissociated SCG neurites. Nuclei are stained with DAPI (blue). Scale bar, 10 m. (E) Western blot of wild-type and WldS SCG neurites and cell body fractions collected 0h or 24h after a cut, probed with NAMPT antibody. Wild-type SCG neurites were kept in the presence of FK866 for 24h collection to avoid degeneration. (F) Integrated NAMPT band intensity of SCG neurites normalised to -actin control (Mean and SD shown, n = 3) shows that NAMPT declines by only 40% during 24h of isolation from its site of synthesis in cell bodies.
Suppl. Fig. 4. Blocking the extracellular degradation of NAD and NMN or blocking the intracellular uptake of NR robustly reduces their effect on axon degeneration. (A) NAD and NMN extracellular degradation pathways convert them to NR which is then transported inside the cell and re-converted in NMN by the action of NRK (adapted from(33)). In cut axons NMN cannot be converted to NAD due to the loss of NMNAT2. Some direct uptake of NMN remains possible. (B, C) SCG neurites were left untreated or treated immediately after cut with the compounds indicated. NMN (B) and NAD (C) were used at 1 mM. CMP and UMP at 25 mM, dipyridamole at 50 µM and Ap4A (diadenosine tetraphosphate) at 10 mM. Images were acquired at the indicated time points. Degeneration index was calculated in three fields of three independent experiments (Mean ±SEM, n = 9, 2-way ANOVA followed by Bonferroni post-hoc test, ***p<0.001).
Suppl. Fig. 5. NAD precursors from Na salvage pathway do not affect axon degeneration and maintain cell bodies in the presence of FK866. (A) Dissociated SCG untreated [(a) and (b)], or treated with 100nM FK866 [(c) and (d)], with 100nM FK866 and 1mM NMN [(e) and (f)], or with 100nM FK866 and 1mM NAD [(g) and (h)]. FK866-induced NAD depletion causes neurodegeneration at 5 days but NMN or NAD prevent neurodegeneration, presumably by restoring NAD levels. (B) DRG neurites were separated by their cell bodies and treated with 100nM FK866 and with various NAD precursors (1mM) as indicated. Neurites were imaged at 0h, 24h and 48h as indicated. Cell body explants were imaged at 0, 6 and 13 days. NAD depletion by FK866 causes degeneration of the cell body compartment which is detectable 6 days after FK866 addition and 2 weeks later all cell bodies have degenerated and detached (b). Cell bodies are maintained if NAD synthesis is restored by the addition of NMN (c) or all precursors of the Na salvage pathway (d, e, f) (see also Fig. 1A). FK866 axon protection is affected only by NMN (c) which restores rapid degeneration. In contrast all other precursors which are converted to NAD via the Na salvage (Preiss-Handler) pathway do not affect axon degeneration (see also Fig. 1A).
Suppl. Fig. 6. FK866 strongly delays Vincristine-induced axon degeneration and NMN abolishes FK866 protection. (A) Bright field images of wild-type SCG distal axons cultured for 7 days and then treated with 0.02μM Vincristine, with 0.02μM Vincristine and 100nM FK866 or with 0.02μM Vincristine, 100nM FK866 and 1mM NMN as indicated. Images of an untreated explant are also shown. (B) Degeneration index was calculated from three independent experiments (n=3, single measurements indicated, two-way repeated measures ANOVA followed by Bonferroni post-hoc test, **, p<0.01, ****p<0.0001).
Suppl. Fig. 7.S. oneidensis NMN deamidase confers robust protection to injured SCG axons.Shewanella oneidensis NMN deamidase has a similar activity to the E. coli enzyme despite its divergent amino acid sequence. (A) Western blot of HEK293T cells transfected with S. oneidensis NMN deamidase expression vector and harvested two days after transfection, probed with an anti-his antibody. (B) Extracts of HEK293T cells transfected with NMN deamidase cDNA in pcDNA3.1/HISA or with empty pcDNA3.1/HISA were used for NMN deamidase activity determination (see methods). (C) S. oneidensis NMN deamidase expressing vector or the empty vector was microinjected into SGC nuclei along with fluorescent marker DsRed2 to visualise transfected neurons. Fluorescent neurites were cut and imaged at the indicated time points. SCG neurons expressing S. oneidensis NMN deamidase are resistant to degeneration after cut. NAMPTG217R partially resists FK866 inhibition. (D) Western blot of cell extracts of HEK293T cells transfected with empty vector or with plasmid cDNA vectors encoding NAMPTG217R or NAMPT wild-type (WT) and probed with NAMPT antibody or with a -acting loading control confirms overexpression of NAMPT constructs. (E) Integrated NAMPT band intensity normalised to -actin control (Mean ±SEM, n = 10). (F) HEK293T untransfected or transfected with empty vector or with plasmid cDNA vectors encoding NAMPTG217R or NAMPT WT were homogenised and NAD levels were determined with an HPLC based method as a measure of NAMPT enzymatic activity in the absence or presence of 100nM FK866. NAMPTG217R enzyme activity is in part retained in the presence of the drug (n=3, Mean and SD shown, one-way ANOVA followed by Bonferroni post-hoc test, ***p<0.001).
Suppl. Fig. 8. NR, ATP and other nucleotides are stable in sciatic nerves after a 30h lesion.(A) (a) HPLC chromatograms of pooled sciatic nerves 30h after a lesion and of unlesioned contralateral nerves. The peaks of interest are labelled; among them, NMN and nicotinamide (Nam) are partially overlapped with other interfering peaks or below the detection limit. Wild-type (b) uncut and (c) 30h-cut nerve extracts (see above and “Methods”) were subjected to derivatization with acetophenone(41) and analysed by reverse-phase C18-HPLC. The image shows chromatographic fluorescence profiles of the above nerve extracts after derivatization, before and after addition of 7 pmol NMN spike. Black arrows indicate the position of an NR standard showing absence in both nerve extracts (before and after cutting) under these experimental conditions. (B) NMN and NR levels in subcellular fractions of explanted pooled mouse sciatic nerves after 0h and 30h ex vivo, as determined by liquid chromatography electrospray ionization–tandem mass spectrometry.
Suppl. video 1. Wallerian degeneration in axotomised larval zebrafish neurons treated with vehicle. Time-lapse movie corresponding to Fig. 6C (top panels). Each frame is a projection of a confocal image stack. Frames every 20 minutes.
Suppl. video 2. FK866 robustly delays Wallerian degeneration in axotomised larval zebrafish neurons. Time-lapse movie corresponding to Fig. 6C (centre panels). Each frame is a projection of a confocal image stack. Frames every 22 minutes.