Targeting NMNAT1 to axons and synapses transforms its neuroprotective potency in vivo

Abstract

Axon and synapse degeneration are common components of many neurodegenerative diseases, and their rescue is essential for effective neuroprotection. The chimeric Wallerian degeneration slow protein (Wld(S)) protects axons dose dependently, but its mechanism is still elusive. We recently showed that Wld(S) acts at a non-nuclear location and is present in axons. This and other recent reports support a model in which Wld(S) protects by extranuclear redistribution of its nuclear NMNAT1 portion. However, it remains unclear whether cytoplasmic NMNAT1 acts locally in axons and synapses or at a non-nuclear site within cell bodies. The potency of axon protection by non-nuclear NMNAT1 relative to Wld(S) also needs to be established in vivo. Because the N-terminal portion of Wld(S) (N70) localized to axons, we hypothesized that it mediates the trafficking of the NMNAT1 portion. To test this, we substituted N70 with an axonal targeting peptide derived from amyloid precursor protein, and fused this to NMNAT1 with disrupted nuclear targeting. In transgenic mice, this transformed NMNAT1 from a molecule unable to inhibit Wallerian degeneration, even at high expression levels, into a protein more potent than Wld(S), able to preserve injured axons for several weeks at undetectable expression levels. Preventing NMNAT1 axonal delivery abolished its protective effect. Axonally targeted NMNAT1 localized to vesicular structures, colocalizing with extranuclear Wld(S), and was cotransported at least partially with mitochondria. We conclude that axonal targeting of NMNAT activity is both necessary and sufficient to delay Wallerian degeneration, and that promoting axonal and synaptic delivery greatly enhances the effectiveness.

Figure 1.

Confocal z-projections showing the targeting of NMNAT1 to axons. A, Immunostained hippocampal neurons transfected with FLAG-tagged native NMNAT1 (NMNAT1-FLAG), FLAG-tagged NMNAT1 with disrupted nuclear localization (ΔNLS NMNAT1-FLAG), and FLAG-tagged ΔNLS NMNAT1 N-terminally fused to a peptide from exon 7 of SMN protein (Exon7 SMN-ΔNLS NMNAT1-FLAG) or to a peptide from the AICD of APP (AICD-ΔNLS NMNAT1-FLAG). Anti-FLAG immunocytochemistry labeled the overexpressed NMNAT1 variants (red). On the right, 20× images show that AICD-ΔNLS NMNAT1-FLAG redistributes to distal axons most efficiently. Cell body insets are shown at 63× magnification on the left. Anti-ankyrin G identified proximal axons (green), and anti-MAP-2 marked dendrites (pink); nuclei were counterstained with DAPI (blue). Superimposed images of overexpressed NMNAT1 variants and ankyrin G are shown in the merge panel. Arrows point to axons. B, The AICD-ΔNLS NMNAT1-FLAG construct was named Ax-NMNAT1 (axonally targeted NMNAT1) and selected for the generation of transgenic mice. In the N-terminal axonal targeting sequence, evolutionarily conserved amino acids are highlighted in yellow.

Figure 2.

Axonally targeted NMNAT1 levels in brains of Ax-NMNAT1 mice from line 1 and 2. A, Western blotting of Ax-NMNAT1 line 2 mouse brain homogenates showing approximately twofold higher expression level of Ax-NMNAT1 protein in homozygotes than in hemizygotes, by anti-FLAG antibody detection (top) and anti-NMNAT1 antibody detection (bottom). B, Representative anti-NMNAT1 (antibody 183) Western blot of total brain homogenates of Ax-NMNAT1 mice. Anti-NMNAT1 antibody 183 detects both NMNAT1 and Wld^S. Axonally targeted NMNAT1 is 23 aa longer than murine native NMNAT1 and therefore migrates more slowly on SDS-polyacrylamide gel. Transfected HEK lysates indicate the molecular size of NMNAT1 and Ax-NMNAT1. Ax-NMNAT1-transfected HEK lysate is 10-fold diluted; thus, its loading control band (β-actin) is not visible at this exposure time. In A and B, # indicates nonspecific band. C, Confirmation that axonally targeted NMNAT1 is enzymatically active in brain homogenates and that activity levels correlate with protein levels. Levels of NMNAT specific enzyme activity are higher in Ax-NMNAT1 mice from line 2 than in negative control littermates (WT). NMNAT enzyme activity of Ax-NMNAT1 mice from line 1 is similar to the WT one. Both Ax-NMNAT1 lines show levels lower than the ones from all the other strains. **p = 0.0037.

Figure 3.

Comparison of delay of Wallerian degeneration in Ax-NMNAT1 mice and in mice with the indicated genotypes expressing YFP in a representative subset of neurons. Longitudinal imaging (confocal z-projections) of sciatic and tibial nerves 7, 14, and 35 d after transection, and quantification of percentages of intact axons in distal tibial nerves 35 d after lesion. Nerves from NMNAT1-overexpressing mice (TgNMNAT1) fragment with a time course similar to WT nerves, whereas 35 d lesioned nerves from hemizygous Ax-NMNAT1 mice still show intact axons. Thirty-five days after axotomy, the number of unfragmented axons in Ax-NMNAT1 samples is significantly higher than the one in Wld^S samples and similar to the one in ΔNLS Wld^S samples. (**p = 0.003, one-sample t test; NS: p = 0.399).

Figure 4.

Neurite preservation after axotomy in DRG explant cultures from hemizygous Ax-NMNAT1 embryos. A, Phase-contrast 0.3 mm × 0.3 mm images of the same field at the time of transection, and 1, 3, and 6 d after. Neurites from Ax-NMNAT1 explant cultures remain unfragmented for at least 6 d similar to neurites from spontaneous mutant Wld^S and transgenic ΔNLS Wld^S explant cultures. Neurites from homozygous NMNAT1-overexpressing (TgNMNAT1) embryos fragment within the first 24 h. Scale bar, 0.1 mm. B, Quantification of neurite continuity by assessment of the PI plotted against time after cutting. C, Western blot probed with anti-NMNAT1 antibody 183, showing the respective transgenic protein expression in homozygous DRGs explant cultures. β-Actin is the loading control.

Figure 5.

Cytoplasmic redistribution of axonally targeted NMNAT1. Representative Western blot of nuclear and postnuclear (cytoplasmic) fractions from brains of Ax-NMNAT1 mice in comparison to native Wld^S mice, mice expressing the nonprotective variant of Wld^S lacking the first 16 N-terminal amino acids (ΔN16 Wld^S hemi), and mice expressing the enhanced protective extranuclear variant (ΔNLS Wld^S L3 hemi) (n = 4). Nuclear and cytoplasmic factions, blotted with anti-NMNAT1 antibody 183 (top left and right), are shown at different exposure times to optimize the visualization of the faint bands (Ax-NMNAT1, Wld^S, and ΔN16Wld^S). Sp1 is the loading control for the nuclear fraction. The densitometry (bottom) shows the intensity of the cytoplasmic bands, normalized to β-actin, and is expressed as percentage of ΔNLS Wld^S levels for comparative analysis. The axonally targeted NMNAT1 protein shows the lowest levels if compared to Wld^S and variants in the other strains. ^#Nonspecific band.

Figure 6.

Neuromuscular junctions in Ax-NMNAT1 mice remain functional for at least 6 d after axotomy in young (2 months old) and adult (>6 months old) mice. A, Confocal projections showing alignment of vital staining of presynaptic motor nerve terminals (AM1-44, green) with postsynaptic acetylcholine receptors (TRITC-BTX, red) in lumbrical and FDB muscles of mice of the indicated genotype and age. Protection was weak at this time point in young Wld^S homozygotes and absent in older (6 months) Wld^S homozygotes but appeared equally strong at both ages in Ax-NMNAT1 mice. B, Intracellular recordings showing evoked (I–III) and spontaneous (IV) responses in 4 different FDB muscle fibers from 2-month-old Ax-NMNAT1 hemizygotes (line 2), 6 d after axotomy. The peaks of evoked action potential are clipped in BI and BII. BI, Response to single tibial nerve stimulus; BII, superimposed traces at 1 Hz stimulation showing EPPs with reduced quantal content near the firing threshold; BIII, subthreshold EPPs; BIV, spontaneous MEPPs recorded in several superimposed sweeps on a free-running time base. C, Intracellular recordings of synaptic potentials recorded from FDB muscle fibers of 6-month-old Ax-NMNAT1 mice (I, II) and of 5.5-month-old Wld^S homozygotes (III) 6 d after axotomy, and from unlesioned contralateral controls (IV). CI, Superimposed subthreshold EPPs in axotomized Ax-NMNAT1 FDB muscle. CII, Subthreshold EPPs and a MEPP in axotomized Ax-NMNAT1 FDB muscle. CIII, No evoked action potential visible in Wld^S muscle. CIV, Expected evoked action potentials in an unlesioned fiber.

Figure 7.

Blockage of axonal entry abolishes Ax-NMNAT1 protective capacity. A, Diagram illustrating vincristine and nocodazole application (gray bars) at the indicated times to DRG neurons coexpressing EGFP and Ax-NMNAT1-mCherry. Vincristine was applied at 0.04 μm (i and ii), and nocodazole at 5 or 20 μg/ml (iii and iv), at the time of transfection (ii and iv), or 24 h later at the time of axotomy (i and iii). B, Representative EGFP epifluorescence images of DRGs expressing Ax-NMNAT1-mCherry, at the time of axotomy (I) and 48 h later (II), treated with 0.04 μm vincristine (i and ii), or with 20 μg/ml nocodazole (iii and iv), at the times indicated in A. Dashed line marks the cut location. C, Quantification of axonal preservation 48 h after axotomy (imaging II) in DRGs cultured and treated as explained in A. The diagram shows the percentage of axons (±SD) that did not develop swelling or fragmentation between imaging I and imaging II. D, Diagram illustrating 0.04 μm vincristine, and 20 μg/ml nocodazole treatments on DRG explant cultures from wild-type mice and Ax-NMNAT1 homozygotes from line 1. Treatments started 6 d after plating, 24 h before axotomy. E, Phase-contrast images (0.3 mm × 0.3 mm) of Ax-NMNAT1 or wild-type explants, as indicated, at the time of cut (I) and 6 d later (II), treated with vincristine as illustrated in D. F, Phase-contrast images (0.3 mm × 0.3 mm) of Ax-NMNAT1 or wild-type, as indicated, at the time of cut (I) and 6 d later (II), treated with nocodazole as illustrated in D. Scale bars: B, E, F, 0.1 mm.

Figure 8.

Localization of Ax-NMNAT1 protein in subcellular fractions. A, Brain subcellular fractions from hemizygous Ax-NMNAT1 mice, and variant and native Wld^S mice (blots representative of 3 independent experiments). Membranes from negative littermate (WT) and Ax-NMNAT1 mice were probed with anti-NMNAT1 antibody 183 (left two panels) and membrane from ΔNLS Wld^S and Wld^S mice were probed with Wld18 antibody (right two panels). Additional probing with the organelle markers validates fraction purity: anti-adaptin γ for Golgi-enriched fractions, anti-p38 SNPH for synaptic vesicles-enriched fractions, anti-calnexin for ER-enriched fractions, anti-Sp1 for nuclear-enriched fractions, and anti-prohibitin for mitochondrial-enriched fractions. H, Total homogenate; N, nuclear-enriched fraction; PN, postnuclear fraction; M, mitochondrial-enriched fraction; Vo, V1, V2, vesicle fractions; C, cytosol. Ponceau S indicates the amount of loaded protein. B, Densitometry of Ax-NMNAT1, ΔNLS Wld^S, and Wld^S band intensity in the respective organelle-enriched fractions. Because in each fraction the quantified protein is depleted or concentrated from the total homogenate, the intensities are expressed as percentage of the band intensity in the respective H fraction. Ax-NMNAT1 and (variant) Wld^S protein appear concentrated in vesicle fractions with intensity correlating with the p38 SNPH intensity, and Ax-NMNAT1 is particularly concentrated also in the mitochondrial fraction.

Figure 9.

Localization of Ax-NMNAT1 protein in transfected primary cultures. A, Confocal z-projection of the subcellular distribution of Ax-NMNAT1-mCherry in transfected hippocampal neuron. B, Single confocal plane of the axonal inset in A. C, Confocal z-projection of Ax-NMNAT1-mCherry and ΔNLS Wld^S-EGFP cotransfected hippocampal neuron. D, Single confocal plane of the soma and axon insets in C. E, F, Confocal z-projection of the subcellular distribution of Ax-NMNAT1-mCherry in transfected DRG and its growth cone respectively. G, Single confocal plane of the inset in F; arrowheads point to Ax-NMNAT1-mCherry puncta at the growth cone. H, I, Single confocal plane of soma and axon, respectively, of an Ax-NMNAT1-mCherry and ΔNLS Wld^S-EGFP cotransfected DRG. In D, H, and I, arrows indicate partial colocalization of Ax-NMNAT1 with ΔNLS Wld^S. Scale bars: A, C, E, F, 20 μm; B, D, G–I, 2 μm. In A–I, blue is DAPI staining, red is Ax-NMNAT1-mCherry fluorescence, and green is ΔNLS Wld^S-EGFP fluorescence. J, Single confocal plane showing colocalization (arrows) of Ax-NMNAT1-mCherry particles and mitochondria (MitoTracker Green) in a transfected DRG axon. Scale bar, 2 μm. K, Time-lapse imaging of the experiment in J, showing coordinated retrograde movement of an Ax-NMNAT1-mCherry particle (arrow, left) with a mitochondrion (arrow, right). Images were zoomed from supplemental Movie 1 (available at www.jneurosci.org as supplemental material).