Wld ^S but not Nmnat1 protects dopaminergic neurites from MPP⁺neurotoxicity

Previously we have shown that dopaminergic terminal fields but not cell bodies of Wld ^S mice are protected against MPTP injury [23]. To confirm and extend these observations in a more tractable system, we utilized dissociated cultures of midbrain neurons in which 1-5% of the total cells plated are dopaminergic [33]. Results show that cultures from Wld ^S mice exhibited significant protection of neurites not seen in wild type cultures after MPP⁺ treatment (Figure 1A, C). Moreover, dopaminergic cell death from MPP⁺ treatment was also attenuated in Wld ^S cultures, unlike those seen in vivo (Figure 1A, B). Thus Wld ^S can effectively protect neurites (dendrites and axons) as well as cell bodies from known PD mimetics in vitro.

Recent studies have reported that the localization of Wld ^S influences its neuroprotective effect. Babetto et. al. have reported that a cytoplasmic version of the Wld ^S protein (cyto Wld ^S ) confers a higher level of protection than the native form of Wld ^S [34]. In addition, Sasaki et. al. have reported that cytoplasmic Nmnat1 (cyto Nmnat1) and Nmnat3, which is primarily localized in the mitochondria, also confer a higher level of protection than Nmnat1 [35, 36]. To test whether cytoplasmic localization of Wld ^S rescued or changed the level of protection seen with nuclear Wld ^S , primary dopaminergic neurons were prepared from wild type and cyto Wld ^S mice, treated with MPP⁺, and analyzed as described. Results show that cyto Wld ^S mice exhibited a similar level of protection of dopaminergic cell bodies and neurites as seen in Wld ^S mice (Figure 2). Therefore, as reported for peripheral model systems and certain CNS paradigms [37], redirecting most of Wld^S from the nucleus to the cytoplasm protects processes equally well.

Many studies, especially in peripheral model systems, have shown that Nmnat1 can at least partially mimic the effects of Wld ^S [26, 27]. To determine whether this is true in dopaminergic neurons, we transduced primary midbrain cultures from wild type animals with either GFP, Nmnat1, the 70 amino acid fragment of Ube4b encoded within the Wld ^S gene, or the entire Wld ^S coding region using lentiviral vectors expressing GFP [27] (Figure 3). We also tested the effects of Nmnat3, cytoplasmic Nmnat1, and enzymatically inactive Nmnat1 (Nmnat1 (W170A)) [28, 38] (Figure 3D, E and Additional File 1 Figure S1C, D). Immunofluorescence and western blotting was done to confirm that transductions led to similar expression levels in dissociated cultures (Figure 3B, C, Additional File 1 Figure S1B, D-E). Despite equivalent levels of transgene expression, only neurites transduced with the entire coding sequence of Wld ^S were protected from MPP⁺ injury (Figure 3D, E).

Because many studies have suggested that Nmnat and in particular cyto Nmnat or axonally targeted Nmnat can be as effective as Wld ^S in protecting axons from mechanical or toxic insults, we used DRG cultures as a positive control [34, 39]. Consistent with those studies, both Wld ^S and cyto Nmnat rescued DRG neurites from the neurotoxin, vincristine, whereas the GFP-only and inactive Wld ^S virus did not (Figure 4). Taken together, these data confirm previous results showing that cyto Nmnat is necessary and sufficient to save DRG neurites. In contrast, only Wld ^S but not cyto Nmnat, Nmnat1, or Nmnat3 was able to protect dopaminergic neurons from the neurotoxic effects of MPP⁺.

To corroborate the hypothesis that Nmnat1 does not protect dopaminergic neurons from MPP⁺, we transduced dissociated primary midbrain cultures with enzymatically inactive Wld ^S (W258A; [27]). In contrast to our own results in DRG cultures (Figure 4) as well as results published by others using this same construct [27], the inactive Wld ^S plasmid was as effective as NAD⁺-synthesizing Wld ^S animals in protecting dopaminergic cell bodies and neurites against MPP⁺ injury (Figure 5). Therefore, the entire Wld ^S chimeric protein, but not its NAD⁺-synthesizing activity, is required for neuroprotection of dopaminergic neurons.

Previous studies have shown that NAD⁺ itself can be neuroprotective [27]. Although Nmnat1 by itself did not recapitulate the neuroprotective effect of Wld ^S on dopaminergic neurons, we tested whether NAD⁺ or one of its precursors (Figure 6A) rescued cell bodies or neurites from MPP⁺ treatment. Therefore, dissociated dopaminergic wild type cultures were pretreated with either 1 mM of NAD⁺, nicotinamide mononucleotide (NMN), or nicotinic acid mononucleotide (NaMN) 24 hours before MPP⁺ treatment. Both NAD⁺ and NMN but not NaMN protected cell bodies and neurites against MPP⁺ toxicity (Figure 6B, C). These findings together with the results showing that catalytically-inactive Wld ^S was able to protect dopamine neurons (Figure 5) but catalytically active Nmnat did not (Figure 3F) suggest that different pathways are being invoked in response to MPP⁺ toxicity.

Previous studies in DRG neurons have attributed the protective phenotype of Wld ^S to its action on the Nmnat1-NAD⁺-Sirt1 pathway [27]. To test the involvement of Sirt1, we prepared dissociated dopaminergic cultures from Sirt1 knockout mice. Following 24 hour pretreatment with 1 mM NAD⁺ or vehicle control, cultures were exposed to MPP⁺. Consistent with the notion that NAD⁺ is not acting through Sirt1 but rather through a different mechanism, NAD⁺ protected cell bodies and neurites in Sirt1 knockout cultures from MPP⁺ toxicity (Figure 7).

The data described above suggest that Wld ^S is acting through a separate possibly parallel pathway from that of NAD⁺ in dopaminergic neurons. If so, then adding NAD⁺ to Wld ^S cultures will enhance the neuroprotective phenotype of Wld ^S . To see whether the NAD⁺ effect overlapped with Wld ^S or was additive, dissociated dopaminergic cultures were prepared from wild type and Wld ^S mice and pre-treated with and without NAD⁺ as previously described. Both NAD⁺ and Wld ^S alone exhibited similar levels of cell body and neurite protection (Figure 8). However, NAD⁺ together with Wld ^S generated significantly higher levels of protection suggesting this is an additive process (Figure 8). To test whether these effects were maximal, additional cultures were treated with 5 mM NAD⁺; no significant differences in neuroprotection were observed when compared with the lower dose of 1 mM NAD⁺ (Figure 8D). These findings demonstrate that NAD⁺ and Wld ^S are additive in the MPP⁺ model suggesting that they are acting through separate and possibly parallel neuroprotective mechanisms.

For primary midbrain cultures, the ventral mesencephalon was removed from embryonic day 14 (E14) murine embryos as previously described [33, 72]. Wild-type (C57/Bl6) and Wld ^S (C57Bl/OlaHsd-WldS) mice were ordered from Harlan (Bichester, UK). Sirt1 knockout mice were obtained from Dr. Christian Sheline (Louisiana State University - Health Science Center, New Orleans, LA). Cyto Wld ^S mice were obtained from Dr. Michael Coleman (Babraham Institute, UK) [73]. Animals were treated in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All procedures were approved by the Washington University School of Medicine animal experimentation committee. Plates were pre-coated overnight with 0.2 mg/ml poly-D-lysine (Sigma-Aldrich, St. Louis, MO). Cells were plated at a density of approximately 125,000 cells/cm² and maintained in serum-free Neurobasal medium (Invitrogen, Carlsbad, CA) supplemented with 1× B27 supplement (Invitrogen), 0.5 mM L-glutamine (Sigma-Aldrich), and 0.01 μg/ml streptomycin plus 100 U penicillin. Half of the culture medium was replaced with fresh Neurobasal medium after 5 days in vitro (DIV). Cultures were pretreated with 1 mM NAD⁺, 1 mM NMN, 1 mM nicotinic acid mononucleotide (NaMN), or a comparable volume of vehicle 24 hours before toxin treatment. Cultures were treated with either 1 μM 1-methyl-4-phenylpyridinium (MPP+), the active metabolite of MPTP or vehicle on DIV 7. Dorsal root ganglion (DRG) cells were obtained from E14 murine embryos as previously described [74]. Cells were plated on coverslips precoated with 0.1 mg/ml poly-L-ornithine (Invitrogen) and 32 μg/ml laminin-1 (Invitrogen) and maintained in DRG media which consisted of Eagle Minimal Essential Media (Invitrogen) supplemented with chick embryo extract (Invitrogen), 10% fetal calf serum (Invitrogen), 50 ng/ml Nerve Growth Factor (Harlan Biosciences, Madison, WI) and 50 U/ml penicillin-50 g/ml streptomycin. Half of the culture medium was replaced with fresh DRG medium after DIV 5. After transduction with lentivirus on DIV 2, DRG cultures were treated with 0.4 μM vincristine or vehicle on DIV 7. NAD⁺, NMN, NaMN, MPP⁺, and vincristine were all obtained from Sigma-Aldrich.

The lentiviral expression plasmids FUGW, FCIV-Wld^S, FCIV-Nmnat1, FCIV-Ube4b, FCIV-Nmnat3, FCIV- Nmnat1(W170A), FCIV-cytNmnat1, and FCIV-Wld^S(W258A) were obtained from Dr. Jeffrey Milbrandt (Washington University, Saint Louis). Lentiviruses expressing transgenes were generated by the Hope Center for Neurological Disorders Viral Core (Washington University, Saint Louis). For infection of DRG and primary midbrain neurons, 50 μl lentivirus (10⁵ infectious units/μl) was added to the well of a 7-mm dish containing approximately 70,000 neurons on DIV 2. Transduced primary midbrain and DRG neurons were treated with MPP⁺ and vincristine, respectively, on DIV 7. Viral infection and transgene expression was monitored using the GFP reporter via fluorescent microscopy.

Primary dopaminergic cultures and DRGs were plated in 7 mm microwell plates (MatTek Corp., Ashland, MA). Cells treated with MPP⁺ were fixed with 4% paraformaldehyde (PFA) in PBS after 48 hours. Cultures were stained with sheep polyclonal anti-tyrosine hydroxylase (TH) (Novus Biologicals, Littleton, CO) and Cy3 α-sheep (Molecular Probes, Carlsbad, CA). Localization of cytoplasmic Wld ^S was confirmed using rabbit Wld ^S antibody (gift of M.P. Coleman) and Alexa488 α-rabbit (Molecular Probes). TH⁺ cells and neurites were counted using unbiased stereological methods (Stereo Investigator, MicroBrightField, Williston, VT). DRG cultures treated with vincristine were subsequently stained with mouse acetylated tubulin (Sigma-Aldrich) and Cy3 α-mouse (Molecular Probes). Neurites were counted as described above. All images were acquired by confocal microscopy (Olympus Fluoview 500, Olympus, Center Valley, PA) and processed in ImageJ (NIH).

Primary midbrain cultures were plated in 48-well plates and transduced with the transgene of interest as described above. Lysates were collected in RIPA buffer (150 mM NaCl, 1% Nonidet P-40, 0.5% NaDoc, 0.1% SDS, 50 mM Tris pH 8.0) with protease inhibitor mixture (Roche, Mannheim, Germany) and incubated on ice for 30 minutes. Insoluble cell debris was removed by centrifugation and the protein concentration of each cell lysate was determined by Bradford protein assay (BioRad, Hercules, CA). Equal amounts of protein were run on SDS-polyacrylamide gels and transferred to polyvinylidene diflouride (PVDF) membranes (BioRad). PVDF membranes were probed with either rabbit Wld ^S antibody or chicken polyclonal anti-GFP antibody (Aves Labs, Tigard, OR). As a control, PVDF membranes were also probed with goat polyclonal anti-HRP60 antibody (Santa Cruz Biotechnology, Santa Cruz, CA). The secondary antibodies used were either a HRP-linked rabbit antibody or HRP-linked anti-chicken antibody and a HRP-linked anti-goat antibody (Jackson Immunoresearch, West Grove, PA). Membranes were developed with enhanced chemiluminescence (Amersham Biosciences), imaged with either a Storm PhosphorImager (Molecular Dynamics) or a ChemiDoc XRS System (Bio-Rad, Hercules, CA) and band intensities were determined using ImageQuant software (Amersham Biosciences).

TH⁺ cells and neurites were counted using unbiased stereological methods [75] (Stereo Investigator (MicroBrightfield, Williston, VT), in combination with a Zeiss Axioplan2 microscope (Thornwood, NY) and an Optronics Microfire camera. The number of counting sites necessary to achieve a coefficient of error < 0.1 was determined by preliminary experiments. The total number of TH⁺ cell bodies was calculated using the Fractionator function on Stereo Investigator by dividing the estimated number of cells by the estimated volume (μm³) of the dish sampled. Using the Petrimetrics function on Stereo Investigator, TH⁺ neurites intersecting the boundary of the Petrimetric probe were counted. Neurite length was derived by dividing the total estimated neurite length (μm) by the estimated volume (μm³) of the dish sampled

GraphPad Prism software (San Diego, CA) was used for statistical analysis. All data was collected from a minimum of three independent experiments done in triplicate. The significance of effects between control and experimental conditions was determined by a Student t-test or one-way ANOVA with Bonferroni Multiple Comparisons tests.