The authors have declared that no competing interests exist.
The author(s) have made the following declarations about their contributions: Conceived and designed the experiments: SM JG MPC. Performed the experiments: SM. Analyzed the data: SM. Wrote the paper: SM JG MPC.
Modulation of the subcellular localization of the endogenous axon survival factor Nmnat2 boosts its axon protective capacity, suggesting a novel approach to delaying axon degeneration in neurodegenerative disease.
Axons require a constant supply of the labile axon survival factor Nmnat2 from their cell bodies to avoid spontaneous axon degeneration. Here we investigate the mechanism of fast axonal transport of Nmnat2 and its site of action for axon maintenance. Using dual-colour live-cell imaging of axonal transport in SCG primary culture neurons, we find that Nmnat2 is bidirectionally trafficked in axons together with markers of the trans-Golgi network and synaptic vesicles. In contrast, there is little co-migration with mitochondria, lysosomes, and active zone precursor vesicles. Residues encoded by the small, centrally located exon 6 are necessary and sufficient for stable membrane association and vesicular axonal transport of Nmnat2. Within this sequence, a double cysteine palmitoylation motif shared with GAP43 and surrounding basic residues are all required for efficient palmitoylation and stable association with axonal transport vesicles. Interestingly, however, disrupting this membrane association increases the ability of axonally localized Nmnat2 to preserve transected neurites in primary culture, while re-targeting the strongly protective cytosolic mutants back to membranes abolishes this increase. Larger deletions within the central domain including exon 6 further enhance Nmnat2 axon protective capacity to levels that exceed that of the slow Wallerian degeneration protein, WldS. The mechanism underlying the increase in axon protection appears to involve an increased half-life of the cytosolic forms, suggesting a role for palmitoylation and membrane attachment in Nmnat2 turnover. We conclude that Nmnat2 activity supports axon survival through a site of action distinct from Nmnat2 transport vesicles and that protein stability, a key determinant of axon protection, is enhanced by mutations that disrupt palmitoylation and dissociate Nmnat2 from these vesicles.
Neurons are polarized cells that rely on bidirectional transport to deliver thousands of cargos between the cell body and the most distal ends of their axons. One cargo that is of particular importance is the NAD-synthesising enzyme Nmnat2. This surprisingly unstable protein is produced in the cell body and its constant supply into axons is required to keep them alive. If this supply is interrupted, Nmnat2 levels in the distal axon drop below a critical threshold, leading to axon degeneration. The rapid turnover of Nmnat2 contributes critically to the time course of axon degeneration. If its half-life could be extended, axons may be able to survive transient interruptions of its supply. In this study, we find that disruption of Nmnat2 localization to axonal transport vesicles increases both its half-life and its capacity to protect injured neurites. Specifically, association of Nmnat2 with transport vesicles reduces it stability by making it vulnerable to ubiquitination and proteasome-mediated degradation. These findings suggest that modulation of the subcellular localization of Nmnat2 on transport vesicles could serve as a potential avenue for therapeutic treatment of axon degeneration.
The chimeric fusion protein WldS (Entrez Gene ID 22406) affords robust protection of injured axons in vitro and in vivo
Supporting a role of Nmnat2 in axon survival, strong overexpression of Nmnat2 delays Wallerian degeneration in vitro, and this protective effect is dependent on its enzymatic activity
Neurons are extremely polarized cells with processes extending up to centimetres or even meters beyond the cell body. Moreover, in some neurons the axon constitutes over 99% of total cytoplasmic volume
Given that Nmnat2 is essential for axon maintenance
Nmnat2 localizes to vesicular structures and undergoes fast axonal transport in the neurites of primary culture neurons
Here we report that the small, centrally located exon 6 is both necessary and sufficient for palmitoylation, stable membrane association, and vesicle-mediated delivery of Nmnat2 into axons. By manipulating its localization, we then test the hypotheses that these transport vesicles are the sites of Nmnat2 axon-protective action and that vesicular Nmnat2 is somehow protected from rapid turnover, enabling it to reach the ends of long axons before being degraded. Surprisingly, we find that a diffuse, nonvesicular localization enhances Nmnat2 axon protection through increased protein stability. Our results support a model in which Nmnat2 subcellular localization regulates its turnover and protective capacity and suggest a site of action distinct from its transport vesicles.
Previously, we reported that, in primary culture neurons, Nmnat2-EGFP is transported in particulate structures with an anterograde bias and at velocities in the range of fast axonal transport
Co-migration was analysed using kymographs (time-distance graphs) obtained from live-cell imaging of neurites from primary culture SCG neurons. Representative kymographs from neurites co-labelled with Nmnat2-EGFP and (A) mito-tagRFP, (B) lamp1-RFP, or Nmnat2-mCherry and (C) LmanI-EGFP, (D) Golga2-EGFP, (E) Syntaxin6-EGFP, (F) TGN38-EGFP, (G) GFP-Bassoon, (H) GAP43-EGFP, (I) SNAP25b-EGFP, (J) Synaptophysin-EGFP, and (K) SynaptotagminI-EGFP. (L) Quantification of co-migration. The quantification shown for each construct represents the percentage of moving Nmnat2-labelled vesicles that were also labelled by the relevant marker. Error bars indicate SEM.
Next, we tested whether the reported mechanism of Golgi-targeting in HeLa cells
(A) Use of photoactivatable GFP (PA_GFP) fusion proteins to study Nmnat2 membrane association. Time course of a representative cell body of an SCG primary culture neuron injected with each indicated construct is shown. The region of activation is marked by an orange circle in each image. (B) Quantification of protein mobility. Shown is the percentage of fluorescence that remains in the originally activated area compared to a non activated area of equal size elsewhere in the cell body. Error bars indicate SEM.
y∞ | k | |
PA_GFP | 0.51 | 0.56 (0.42 |
Nmnat2-PA_GFP | 0.88 | 0.23 |
Exon6-PA_GFP | 0.88 | 0.23 |
Nmnat2ΔPS-PA_GFP | 0.53 | 0.18 |
Nmnat2ΔBR-PA_GFP | 0.69 | 0.22 |
Nmnat2ΔPSΔBR-PA_GFP | 0.51 | 0.25 |
k-value adjusted for difference in molecular size between PA_GFP and Nmnat2-PA_GFP
The C164/165 palmitoylation site is located at the centre of the 27 amino acids encoded by exon 6 of Nmnat2 (see
Representative kymographs from neurites co-labelled with Nmnat2-mCherry and (A) EGFP, (B) Nmnat2-EGFP, (C) exon6-EGFP, (D) Nmnat2ΔPS-EGFP, (E) Nmnat2ΔBR-EGFP, (F) Nmnat2ΔPSΔBR-EGFP, and (G) Nmnat2Δex6-EGFP. (H) Quantification of co-migration. The quantification shown for each construct represents the percentage of moving wild-type Nmnat2-labelled vesicles that were also labelled by the relevant marker. Error bars indicate SEM. * and ** indicate statistically significant difference compared to EGFP (*
We then confirmed the requirement for the C164/165 palmitoylation site for membrane association in neurons, using a C164S/C165S construct (Nmnat2ΔPS-PA_GFP; see
GAP43 (Entrez Gene ID 14432), which associates with membranes through palmitoylation of a similar double-cysteine motif, also requires an adjacent group of basic residues for efficient and stable membrane association
To confirm that these changes in membrane association reflect the degree of palmitoylation of Nmnat2, we used radiolabelling to measure palmitate incorporation into wild-type and mutant Nmnat2 (
Next, we sought to test the effect of exon 6 mutations on axonal transport of Nmnat2. As expected, mutation of the palmitoylation site in Nmnat2ΔPS-EGFP led to a diffuse, nonvesicular distribution in neurites. We detected little co-migration with Nmnat2-mCherry (
We next tested the hypothesis that vesicle targeting is necessary for Nmnat2-mediated axon protection. We injected wild-type or variant Nmnat2-EGFP together with a DsRed2 fluorescent marker into SCG cell bodies and transected their neurites 48 h later. Due to its short half-life, Nmnat2-EGFP protects neurites for 24 h after transection only if strongly overexpressed. Surprisingly, however, both Nmnat2ΔPS-EGFP and Nmnat2ΔBR-EGFP preserved transected neurites significantly more strongly when only 0.001 µg/µl of DNA was injected (
(A) Representative fields of view of distal primary culture SCG neurites 0 and 24 h after neurite cut, labelled by dsRed2 expression and injected with 0.001 µg/µl EGFP or the relevant Nmnat2-EGFP variant. (B) Representative fields of view of primary culture SCG neurites 0 and 72 h after neurite cut, labelled by dsRed2 expression and injected with 0.002 µg/µl EGFP or the relevant Nmnat2-EGFP variant. (C, D) Quantification of experiments in (A) and (B), respectively. Error bars indicate SEM. * and ** indicate statistically significant difference compared to wild-type Nmnat2 (*
These surprising findings prompted us to investigate the effects on axon protection of deletions within and around exon 6. In particular, three deletion mutants were recently reported to retain enzyme activity (Nmnat2Δ32-EGFP, Nmnat2Δ43-EGFP, Nmnat2Δ69-EGFP
(A) Representative fields of view of distal primary culture SCG neurites 0 and 72 h after neurite cut, labelled by dsRed2 expression and injected with 0.0005 µg/µl EGFP or the relevant Nmnat2-EGFP variant or WldS-EGFP. (B) Quantification of experiment shown in (A). Error bars indicate SEM. * and *** indicate statistically significant difference compared to wild-type Nmnat2 (*
Next, we sought to identify the mechanism by which these mutations increase axon protective capacity. As the short half-life of Nmnat2 limits survival of injured axons
(A) Representative Western blots of HEK293 cells co-transfected with FLAG-WldS and the indicated FLAG-Nmnat2 variant. Twenty-four hours after transfection, cells were treated with 10 µM emetine for the amount of time indicated after which samples were processed for SDS-PAGE and Western blot using anti-FLAG antibody. (B) Quantification of Nmnat2 turnover after emetine treatment. For each sample and time point, the amount of FLAG-Nmnat2 remaining was normalised to FLAG-WldS as an internal control. Error bars indicate SEM. (C) Representative Western blots of GST-Dsk2 pulldown assay. HEK293 cells expressing a FLAG-Nmnat2 variant were lysed (inp. – total input) and ubiquitinated proteins were immunoprecipitated using GST-Dsk2 bound to glutathione beads (ubiq.). GST-fused mutant Dsk2 was used for control pulldown (cont.). Eluted proteins were processed for SDS-PAGE and analysed by Western blot using anti-FLAG antibody. (D) Quantification of ubiquitination assay. For each Nmnat2 variant, the total amount of ubiquitinated FLAG-Nmnat2 was normalised to total input. Error bars indicate SEM. ** and *** indicate statistically significant difference compared to wild-type Nmnat2 (**
t½ (h) | |
FLAG-Nmnat2 | 0.6 |
FLAG-Nmnat2 + 2-BP | 1.1 |
FLAG-Nmnat2ΔPS | 1.3 |
FLAG-Nmnat2ΔBR | 1.8 |
FLAG-Nmnat2ΔPSΔBR | 3.6 |
FLAG-Nmnat2ΔPSΔBR-NterTGN | 1.0 |
FLAG-Nmnat2ΔPSΔBR-NterMOM | 0.9 |
FLAG-Nmnat2Δex6 | 3.4 |
FLAG-Nmnat2Δex6-NterTGN | 0.9 |
As Nmnat2 degradation is blocked by proteasome inhibitor MG132
Based on these results, we hypothesized that palmitoylation and vesicle association cause wild-type Nmnat2 to become destabilised through increased levels of ubiquitination. In contrast, the nonvesicular, cytosolic location of the Nmnat2 mutants reduces ubiquitination and increases protein stability and protective capacity. The Nmnat2ΔPS data, and strongly enhanced protective capacity of Nmnat2Δex6 over Nmnat2ΔPSΔBR, indicate that the observed effects do not just reflect removal of lysine residues. In agreement with this model, inhibiting palmitoylation directly with 2-BP resulted in reduced levels of ubiquitination on FLAG-Nmnat2 (
(A) Representative fields of view of distal primary culture SCG neurites 0 and 72 h after neurite cut, labelled by dsRed2 fluorescence and injected with 0.001 µg/µl of Nmnat2ΔPSΔBR-PA_GFP or one of its N-terminally membrane targeted variants (Nmnat2ΔPSΔBR-Nterex6-PA_GFP (Nterex6), Nmnat2ΔPSΔBR-NterTGN-PA_GFP (NterTGN), and Nmnat2ΔPSΔBR-NterMOM-PA_GFP (NterMOM). (B) Quantification of experiment shown in (A). Error bars indicate SEM. ** and *** indicate statistically significant difference compared to Nmnat2ΔPSΔBR (**
While these results suggest that re-targeting cytosolic Nmnat2 to membranes reverts its stability and protective capacity to lower levels as expected, we cannot rule out the possibility that the N-terminal sequences have a direct effect on Nmnat2 stability. To address this issue, we used a commercially available heterodimerisation system (iDimerize, Clontech), in which two proteins tagged with DmrC and DmrA domains, respectively, undergo heterodimerisation after addition of a soluble “A/C heterodimeriser” compound
(A) Representative fields of view of distal primary culture SCG neurites 0 and 24 h after neurite cut, labelled by dsRed2 fluorescence and injected with 0.002 µg/µl of DmrA-Nmnat2ΔPSΔBR-PA_GFP plus 0.01 µg/µl TGN38-DmrC-HA or 0.001 µg/µl of DmrA-Nmnat2ΔPSΔBR-PA_GFP without any TGN38-DmrC-HA present. Eight hours before neurite cut, 500 nM of A/C heterodimeriser was added to relevant cultures. (B) Quantification of experiment shown in (A). Error bars indicate SEM. ** indicates statistically significant difference compared to control (**
At this point, it is interesting to ask whether the observed changes after Nmnat2 re-targeting arise from a special property of the vesicle membranes that Nmnat2 exon 6 and TGN38 target to, or whether they reflect a more general effect of Nmnat2 membrane association. To test this, we attached the mitochondrial outer membrane anchor (a.a. 1–37) of TOM20 to the N-terminus of Nmnat2ΔPSΔBR (Nmnat2ΔPSΔBR-NterMOM). Note that, as with Nmnat2ΔPSΔBR-NterTGN, the Nmnat2 portion of this construct faces the cytosol. The NterMOM tag led to stable membrane association in the photoactivation assay (
As described above, deletion of exon 6 led to a very strong increase in Nmnat2 protective capacity without any further changes in stability with respect to Nmnat2ΔPSΔBR. To further explore this dissociation between protective capacity and protein stability, we re-targeted Nmnat2Δex6 to vesicle membranes using the N-terminal TGN38 tag. Nmnat2Δex6-NterTGN was stably targeted to membranes in the photoactivation assay (
(A) Representative fields of view of distal primary culture SCG neurites 0 and 72 h after neurite cut, labelled by dsRed2 expression and injected with 0.0005 µg/µl Nmnat2Δex6-EGFP or Nmnat2Δex6-NterTGN-EGFP. (B) Quantification of experiment shown in (A). Error bars indicate SEM. (C) Representative Western blot of GST-Dsk2 pulldown assay. HEK293 cells expressing FLAG-Nmnat2Δex6 or FLAG-Nmnat2Δex6-NterTGN were lysed (inp. – total input), and ubiquitinated proteins were immunoprecipitated using GST-Dsk2 bound to glutathione beads (ubiq.). GST-fused mutant Dsk2 was used for control pulldown (cont.). Eluted proteins were processed for SDS-PAGE and analysed by Western blot using anti-FLAG antibody. (D) Representative Western blot of HEK293 cells co-transfected with FLAG-WldS and FLAG-Nmnat2Δex6 or FLAG-Nmnat2Δex6-NterTGN. Twenty-four hours after transfection, cells were treated with 10 µM emetine for the amount of time indicated after which samples were processed for SDS-PAGE and Western blot using anti-FLAG antibody. (E) Quantification of Nmnat2Δex6 turnover after emetine treatment. For each sample and time point, the amount of FLAG-Nmnat2Δex6 remaining was normalised to FLAG-WldS as an internal control. Error bars indicate SEM.
Nmnat2 is required for axon survival and is the only confirmed endogenous Nmnat isoform in axons. However, its ability to promote axon survival is limited by its short half-life. We have identified a series of mutations that extend Nmnat2 half-life without disrupting enzyme activity and which significantly increase axon protection. For deletion mutants lacking exon 6 the efficacy even surpasses that of WldS. Surprisingly, these changes arise when Nmnat2 targeting to a population of post-Golgi axonal transport vesicles is disrupted and are reversed when vesicle targeting is restored, indicating a nonvesicular site for the axon survival function of Nmnat2. The more stable and protective variants are less prone to ubiquitination through a mechanism likely to involve subcellular targeting and not just lysine availability. We identify cysteine-linked palmitoylation as the vesicle targeting mechanism and propose modulation of this targeting as a promising, novel therapeutic strategy for axonopathies.
We show that Nmnat2 axonal transport vesicles carry Golgi markers as well as synaptic vesicle markers. In contrast, we find no evidence of Nmnat2 undergoing co-transport with mitochondria. This suggests that Nmnat2 is involved in the regulation of cytosolic and not mitochondrial NAD metabolism in the axon, especially since mitochondria are not thought to take up cytosolic NAD under normal conditions
Given the importance of axonal transport, surprisingly little is known about sequences targeting proteins to axons. The mechanism of Nmnat2 membrane association appears very similar to that reported for another axonally transported protein, GAP43. Both proteins lack a transmembrane domain or alternative membrane targeting structures and depend fully on palmitoylation of a double-cysteine motif for membrane association
Palmitoylation regulates the axonal transport and subcellular sorting of several axonally delivered proteins
We also found support for the hypothesis that increased protein stability underlies the mechanism by which diffuse, cytosolic Nmnat2 becomes more highly protective. Wild-type Nmnat2, which is mainly vesicle-bound, is very short-lived both in cell lines and in the neurites of primary culture neurons
Interestingly, this destabilising effect of palmitoylation-mediated membrane attachment contrasts with findings for several palmitoylated transmembrane domain proteins, including cell-surface receptors
Furthermore, our results indicate that subcellular localization and protein stability are not the only determinants of Nmnat2 axon protective capacity. Deletion of exon 6 dramatically increases Nmnat2-mediated neurite protection without any further increase in protein stability. Our findings with the enzyme-dead exon 6 deletion mutant suggest that this strong increase in protective capacity depends on Nmnat2 enzymatic activity. However, the reduced stability of this mutant means we cannot completely rule out the possibility that other, nonenzymatic mechanisms contribute to the rise in protective capacity. Interestingly, deletion of exon 6 overcomes the reduction in Nmnat2 protective capacity upon membrane re-targeting that was observed for point mutants. This rescue occurred despite the destabilising effects of membrane attachment, which were unchanged by the removal of exon 6. This suggests that exon 6 regulates Nmnat2 axon protective function through various mechanisms, which could include protein-protein interactions or additional posttranslational modifications.
In summary, we show that Nmnat2, normally the least axon protective of the three endogenous Nmnat isoforms due to its short half-life, can be converted to a highly protective molecule by disrupting its targeting to axonal transport vesicles. While the importance of these vesicles for long-range axonal trafficking is clear, we suggest that Nmnat2 must dissociate to carry out its axon survival function optimally. We also propose that cytosolic NAD metabolism is central to the axon survival mechanism. Our data establish the principle that Nmnat2 can be modified to promote axon survival and highlight modulation of its palmitoylation state as a route to achieve this. Unlike WldS or other Nmnats, this approach utilizes a protein already identified in wild-type axons, raising the attractive prospect of converting an endogenous axonal protein into one with a protective capacity that matches or even exceeds that of WldS.
Nmnat2-EGFP, FLAG-Nmnat2, and FLAG-WldS constructs were described previously
For organelle markers, the following accession numbers were used for PCR primer design. Constructs were amplified from mouse brain cDNA and inserted into the MCS of pEGFP-NI (Clontech) or ptagRFP-NI (Evrogen) vectors. TGN38-EGFP, NM_009443; Syntaxin6-EGFP, NM_021433; Synaptophysin-EGFP, NM_009305; mito-tagRFP, AK003116 (bp 1–72); LmanI-EGFP, AK011495; Golga2-EGFP, NM_133852; GAP43-EGFP, BC080758; SynaptotagminI-EGFP, NM_001252341. All constructs were verified by DNA sequencing (Beckman Coulter Genomics).
Nmnat2 deletion mutants (Δ32, Δ43, and Δ69) were a gift from Prof. Giulio Magni (Ancona, Italy). GST-Dsk2 UBA was kindly provided by Dr. Simon Cook (Cambridge, UK). The SNAP25-EGFP construct was a gift from Dr. Luke Chamberlain (Glasgow, UK). GFP-Bassoon was a gift from Prof. Eckart Gundelfinger (Magdeburg, Germany). Lamp1-RFP
All animal work was carried out in accordance with the Animals (Scientific Procedures) Act, 1986, under Project License 80/2254. C57BL/6JOlaHsd mice were obtained from Harlan UK (Bicester, UK).
Dissociated superior cervical ganglia cultures were prepared and maintained in culture as described previously
DNA microinjections into the nuclei of primary culture SCG neurons were performed as described
Time-lapse imaging of axonal transport was performed on an Olympus CellR imaging system (IX81 microscope, Hamamatsu ORCA ER camera, 100×1.45 NA apochromat objective, 485 and 561 nm laser excitation). During imaging, cell cultures were maintained at 37°C in an environment chamber (Solent Scientific). Images were captured at 4 (single-color) or 2.5 (dual-colour) frames per second for 1–2 min.
The extent of axonal co-migration of two fluorescent protein markers was analysed in time-lapse recordings of individual neurites. Using ImageJ software version 1.44 (Rasband, W.S., ImageJ, NIH, Bethesda, Maryland, USA,
Parameters of axonal transport of fluorescently labelled proteins were determined from straightened time-lapse recordings of individual neurites using the Difference Tracker ImageJ software plugin
Photoactivation imaging was carried out on an Olympus FV1000 point scanning confocal microscope system (IX81 microscope, 60×1.35 NA plan super apochromat objective, 488 and 561 nm laser excitation). Microinjected cell bodies were identified based on their mCherry fluorescence. Imaging settings were adjusted to standard settings (5× zoom, scan rate 8 µm/s, frame rate 3 s/frame). After taking a pre-activation image, PA_GFP was activated by a 100 ms pulse of a 405 nm laser at 50% intensity in a 100 pixel region of interest in the cell body. Images were then taken every 3 s for a total of 5 min. For analysis, two circular regions of interest of identical size (50 pixel diameter) were selected in the cell body. One was placed in the originally activated area, while the other one was placed 10–20 µm away in an area that was not activated by the original laser pulse. For quantification, the percentage of combined fluorescence in these two areas that remained in the originally activated area was determined for each time point. Data were fitted for exponential decay, and decay constant (k) was calculated using GraphPad Prism 5.04.
Degeneration of ds-Red2 labelled neurites was determined for the same field of view at indicated time points after neurite transection. The percentage of neurites remaining continuous and morphologically normal compared to the initial time point was scored for each field. For experiments involving heterodimerisation, A/C heterodimeriser (Clontech) was added to the relevant dishes 8 h before cut at a final concentration of 500 nM. Fresh medium (with heterodimeriser where appropriate) was added 24 and 48 h after cut.
HEK 293 cells were maintained in culture as described
For ubiquitination experiments, HEK293 cells in 10 cm dishes were transfected with the appropriate FLAG-Nmnat2 construct and, for re-targeting experiments, with empty pCMV-Tag4A or TGN38-DmrC-HA constructs. Twenty-four hours after transfection, 20 µM MG132 (Sigma) was added to the medium. Six hours later, cells were lysed and subjected to GST-Dsk2 UBA (wild-type or mutant) pulldown assay as described
SDS-PAGE analysis and Western blotting analysis were performed as described
HEK293 cells in six-well dishes were transfected with the appropriate FLAG-Nmnat2 construct. Twenty-four hours after transfection, 0.5 mCi/ml [9,103H]-palmitate (Perkin Elmer) was added to the medium. After 6 h, cells were washed in PBS, lysed in 500 µl lysis buffer (20 mM Tris pH 7.5, 137 mM NaCl, 1 mM EGTA, 1% TritonX-100, 10% glycerol, 1.5 mM MgCl2, 50 mM NaF, 1 mM Na3VO4, and protease inhibitor mix (Roche); all chemicals AnalaR unless stated otherwise). The lysate was centrifuged for 10 min, 13,000 rpm. Following overnight incubation of the lysate with 5 µg of anti-FLAG antibody (Sigma), 50 µl of washed Sepharose beads (GE Healthcare) was added and mixed for another 3 h at 4°C. Beads were washed thrice in lysis buffer and twice in wash buffer (50 mM Tris, pH 8.0). Bound protein was eluted with Laemmli sample buffer (BioRad) and boiling for 5 min and processed for SDS-PAGE. After transfer to PVDF membrane, blots were dried and radiolabel was detected by exposure on Tritium phosphor screen (Fuji) for 14 d.
Statistical analyses and graph fitting were performed using GraphPad Prism 5.04 (GraphPad Software Inc.) and SPSS Statistics 19 (IBM).
Nmnat2 primary structure and mutants. (A) Primary structure of
(TIF)
Palmitate labelling of Nmnat2 mutants. (A) Palmitate label and Western blot of wild-type and mutant Nmnat2. HEK293 cells expressing FLAG-Nmnat2 or one of its mutants were labelled with 3H palmitate, subjected to FLAG-immunoprecipitation, and processed for Phosphor Imaging and Western blot (see
(TIF)
Increased protection by untagged cytosolic Nmnat2 mutants. (A) Representative fields of view of primary culture SCG neurites 0 and 24 h after neurite cut, labelled by dsRed2 expression and injected with 0.01 µg/µl empty vector or the relevant untagged Nmnat2 variant. (B) Quantification of experiment in (A). Error bars indicate SEM. * indicates statistically significant difference compared to wild-type Nmnat2 (*
(TIF)
Untagged Nmnat2 deletion mutants are able to strongly preserve neurites. (A) Representative fields of view of distal primary culture SCG neurites 0 and 72 h after neurite cut, labelled by dsRed2 expression and injected with 0.0005 µg/µl empty vector or the relevant unlabelled Nmnat2 variant. (B) Quantification of experiment shown in (A). Error bars indicate SEM. *, **, and *** indicate statistically significant difference compared to wild-type Nmnat2 (*
(TIF)
Enzymatic activity is required for protection by Nmnat2Δex6. (A) Representative fields of view of distal primary culture SCG neurites 0 and 72 h after neurite cut, labelled by dsRed2 expression and injected with 0.0005 µg/µl Nmnat2Δex6-EGFP or enzyme-dead Nmnat2Δex6H24D. (B) Quantification of experiment shown in (A). Error bars indicate SEM. ** indicates statistically significant difference compared to Nmnat2Δex6 (**
(TIF)
2-Bromopalmitate treatment impairs Nmnat2 membrane targeting and extends Nmnat2 half-life. (A) Individual frames from photoactivation assay of SCG primary culture neurons expressing Nmnat2-PA_GFP in absence or presence of 40 µM 2-BP. (B) Quantification of protein mobility in (A). Error bars indicate SEM. *** indicates statistically significant difference compared to control (***
(TIF)
Successful re-targeting to membranes of Nmnat2ΔPSΔBR mutant by N-terminal tags. (A) Individual frames from photoactivation assay of SCG primary culture neurons expressing Nmnat2ΔPSΔBR-PA_GFP or one of its N-terminally membrane targeted variants (Nmnat2ΔPSΔBR-Nterex6-PA_GFP, Nmnat2ΔPSΔBR-NterTGN-PA_GFP, and Nmnat2ΔPSΔBR-NterMOM-PA_GFP). The region of activation is indicated by an orange circle in each image. (B) Quantification of protein mobility in (A). Error bars indicate SEM. (C) Representative images of cell bodies of primary culture SCG neurons expressing Nmnat2ΔPSΔBR-NterMOM-PA_GFP. The whole cell body was subjected to a 405 nm laser pulse in order to activate the entire pool of PA_GFP to enable co-localization analysis. Cells were then stained with MitoTracker dye.
(TIF)
Successful membrane re-targeting of Nmnat2ΔPSΔBR mutant by heterodimerisation. (A) Photoactivation assay of SCG primary culture cell bodies co-expressing DmrA-Nmnat2ΔPSΔBR-PA_GFP and TGN38-DmrC-HA in the absence (control) or presence of 500 µM A/C heterodimeriser for 8 h before imaging. The region of activation is indicated by an orange circle in each image. (B) Quantification of protein mobility in (A). Error bars indicate SEM.
(TIF)
Confirmation of re-targeting to membranes of Nmnat2Δex6 by N-terminal TGN38 tag. (A) Individual frames from photoactivation assay of SCG primary culture neurons expressing Nmnat2Δex6-PA_GFP or Nmnat2Δex6-NterTGN-PA_GFP. The region of activation is indicated by an orange circle in each image. (B) Quantification of protein mobility in (A). Error bars indicate SEM.
(TIF)
We thank Robert Adalbert, Laura Conforti, Simon Cook, Giulio Magni, Michele Di Stefano, and Matthias Ziegler for helpful discussion; Simon Walker for help with live imaging, and the following for donation of plasmid constructs: Giulio Magni, Simon Cook, Luke Chamberlain, and Eckart Gundelfinger.
2-bromopalmitate
enhanced green fluorescent protein
ER–Golgi intermediate compartment
Growth Associated Protein 43
isoform-specific targeting and interaction domain
nicotinamide adenine dinucleotide
Nicotinamide mononucleotide adenlylyltransferase
photoactivatable green fluorescent protein
superior cervical ganglion
slow Wallerian degeneration protein