Experimental Neurology 251 (2014) 91–100

Contents lists available at ScienceDirect

Experimental Neurology

j ourna l homepage: www.e lsev ie r .com/ locate /yexnr

Resveratrol delays Wallerian degeneration in a NAD+ and DBC1dependent manner

Aldo Calliari a,b,⁎, Natalia Bobba b, Carlos Escande c, Eduardo N. Chini c,⁎⁎a Department of Molecular and Cellular Biology, School of Veterinary-UdelaR., Av. A. Lasplaces 1550, CP 11600, Montevideo, Uruguayb Department of Protein and Nucleic Acids, IIBCE-MEC, Av. Italia 3318, CP 11600, Montevideo, Uruguayc Laboratory of Signal Transduction, Department of Anesthesiology and Kogod Center on Aging, Mayo Clinic College of Medicine, Rochester, Minnesota 55905, USA

⁎ Correspondence to: A. Calliari, Department of Moleof Veterinary, University of the Republic, Av. A. LasplaUruguay.⁎⁎ Correspondence to: E. N. Chini, Department of AnesthMedicine, 200 First St. SW, Rochester, MN 55902, USA. Fa

E-mail addresses: acalliari@iibce.edu.uy (A. Calliari), ch(E.N. Chini).

0014-4886/$ – see front matter © 2013 Elsevier Inc. All rihttp://dx.doi.org/10.1016/j.expneurol.2013.11.013

a b s t r a c t

a r t i c l e i n f o

Article history:Received 12 April 2013Revised 6 November 2013Accepted 10 November 2013Available online 16 November 2013

Keywords:Axonal degenerationWallerian degenerationDBC1SIRT1Resveratrol

Axonal degeneration is a central process in the pathogenesis of several neurodegenerative diseases. Understandingthe molecular mechanisms that are involved in axonal degeneration is crucial to developing new therapiesagainst diseases involving neuronal damage. Resveratrol is a putative SIRT1 activator that has been shown todelay neurodegenerative diseases, including Amyotrophic Lateral Sclerosis, Alzheimer, andHuntington's disease.However, the effect of resveratrol on axonal degeneration is still controversial. Using an in vitro model ofWallerian degeneration based on cultures of explants of the dorsal root ganglia (DRG), we showed that resveratrolproduces a delay in axonal degeneration. Furthermore, the effect of resveratrol on Wallerian degenerationwas lost when SIRT1 was pharmacologically inhibited. Interestingly, we found that knocking out Deleted inBreast Cancer-1 (DBC1), an endogenous SIRT1 inhibitor, restores the neuroprotective effect of resveratrol.However, resveratrol did not have an additive protective effect in DBC1 knockout-derived DRGs, suggestingthat resveratrol and DBC1 are working through the same signaling pathway. We found biochemical evidencesuggesting that resveratrol protects against Wallerian degeneration by promoting the dissociation of SIRT1and DBC1 in cultured ganglia. Finally, we demonstrated that resveratrol can delay degeneration of crushednerves in vivo. We propose that resveratrol protects againstWallerian degeneration by activating SIRT1 throughdissociation from its inhibitor DBC1.

© 2013 Elsevier Inc. All rights reserved.

Introduction

The risk of developing neurodegenerative diseases, includingAlzheimer, Parkinson and diabetic neuropathy, increases dramaticallywith age and the onset of metabolic disorders (Mayeux, 2003). Sincelife expectancy is increasing worldwide and metabolic disorders, likeobesity and type II diabetes, are growing health problems, it is expectedthat the incidence of neurodegenerative diseases will keep increasing.Finding new therapies to prevent and ameliorate these diseases mayhave a tremendous impact in the upcoming years.

Axonal degeneration is a common feature of many neurodegenera-tive diseases, being central in the development of Amyothrophic LateralSclerosis, Charcot Marie-Tooth, Spinal Muscular Atrophy and diabeticneuropathy, just to name a few (see recent reviews by Adalbert andColeman, 2013; Coleman, 2013). For many of these pathologies it has

cular and Cell Biology, Schoolces 1550, 11600, Montevideo,

esiology, Mayo Clinic College ofx: +1 507 255 7300.ini.eduardo@mayo.edu

ghts reserved.

been proposed that axonal degeneration precedes degeneration ofthe soma, highlighting the importance of the axon in the pathogenesisof neural diseases. Understanding the molecular pathways involvedin the maintenance of axonal function is imperative for developingstrategies to prevent or slow down axonal degeneration. In this regard,Wallerian degeneration (WD) is a widely accepted model foraxonopathies (Beirowski et al., 2005). The interest in the molecularmechanisms underlying WD has grown dramatically since the discoveryof the mutantWldS mouse strain, in which WD is greatly delayed (Lunnet al., 1989). This finding challenged the accepted concept that WD wasa passive process of degeneration, caused by the interruption of materialdelivery from the neuronal cell body to the distal part of the axon. In-stead, it is now believed thatWD is an active, tightly controlled processof axon self-destruction (Tsao et al., 1999). It was shown that theWldS

mutation also protects central nervous system axons from Walleriandegeneration-like processes in several animal models of human neuro-degenerative diseases (Ferri et al., 2003; Fischer et al., 2005; Gillingwateret al., 2004, 2006; Sajadi et al., 2004). Thus, understanding the molecularmechanisms thatmodulateWDmay contribute to the understanding andtreatment of axonopathies.

Resveratrol, a polyphenol that exhibits beneficial effects in differenttype of cancers, cardiovascular, and autoimmune diseases (Baur &Sinclair, 2006; Fulda and Debatin, 2006), has also been reported to

92 A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

exert neuroprotective effects in pathological situations includingcerebral ischemia and neurodegenerative diseases (Zhang et al. 2010).However, there are contradictory reports on the effect of resveratrolon in vitro models of WD, with different studies calling for neuroprotec-tive effects, null, or even deleterious (Araki et al., 2004; Conforti et al.,2007; Suzuki and Koike, 2007). Since the influence of resveratrol on thecourse of axonal degeneration is controversial, new studies are neededin order to clarify this issue.

Interestingly, some of the effects produced by resveratrol are mim-icked by the activation of SIRT1, a NAD+ dependent histone deacetylasethat regulates the acetylation of several proteins. In particular, SIRT1 hasbeen implicated in the control of histones (Imai et al. 2000; Vaqueroet al., 2004), as well as non-histone proteins such as PGC1α (Rodgerset al., 2005), p53; (Vaziri et al., 2001), the forkhead transcription factors(Nakae et al., 2006); NfκΒ (Yeung et al., 2004); Ku 70 (Cohen et al.,2004); andMyoD (Fulco et al., 2003). As a result of its deacetylase activ-ity, SIRT1 regulates gene silencing, apoptosis, stress resistance, circadiancycle, and metabolism (Kim et al., 2008; Kim and Um, 2008; Nakahataet al., 2008). SIRT1 has gained a lot of attention due to its beneficialeffects on a wide array of neurodegenerative conditions. In fact, pharma-cological activation or overexpressing SIRT1 was reported to delay theonset of specific alterations or modulate the time course of neuropatho-logical conditions in animal models of Alzheimer's, ALS, Huntington's,Parkinson, Multiple Sclerosis, and other diseases (Zhang et al., 2011).

We have recently found that Deleted in Breast Cancer-1 (DBC1) is animportant regulator of SIRT1 and HDAC3 protein deacetylases (Chiniet al., 2010; Escande et al., 2010; Nin et al., 2012). DBC1 acts as anendogenous inhibitor of SIRT1 by binding to the catalytic site of SIRT1and blocking its catalytic activity (Kim et al. 2008). We have alsoshown that AMPK activation by resveratrol leads to SIRT1 activationby dissociation from DBC1 (Nin et al. 2012). These findings promptedus to reevaluate the effect of resveratrol in Wallerian degenerationand explore the role of DBC1 in WD.

In agreement with a previous report (Araki et al., 2004), we foundthat resveratrol delaysWallerian degeneration. The effect of resveratrolon axonal degeneration is modulated by NAD+ and is exerted locally inthe axons, since it also occurs when the neuronal soma is removedfrom the culture before the treatment. The protective effect of resvera-trol was lost by the addition of suramin, a SIRT1 inhibitor (Trapp et al.,2007). Parallel to its neuroprotective effect, we found that resveratrolpromotes dissociation of the SIRT1/DBC1 complex in neurons. UsingMouse Embryonic Fibroblasts (MEFs) and HEK293T cells, we saw thatthe dissociation of SIRT1 and DBC1 induced by resveratrol is simulta-neous with an increase in SIRT1 activity. Importantly, maximal SIRT1activitywas dependent on DBC1. Very interestingly, the effect of resver-atrol on Wallerian degeneration was mimicked in DRGs derived fromDBC1 knockout mice, consistent with the fact that DBC1 KO mice haveincreased SIRT1 activity in tissues and cells (Escande et al., 2010).Resveratrol had no additive protective effect againstWalleriandegener-ation in DBC1 KO-derived DRGs, suggesting that resveratrol exerts itseffects in the SIRT1-DBC1 pathway. Finally, we found that chronicresveratrol treatment protects against nerve degeneration in vivo.Chronic treatment of mice with resveratrol preserves excitability ofinjured sciatic nerves. In summary our data strongly suggest thatresveratrol delays the speed of axonal degeneration by a mechanisminvolving SIRT1 and DBC1. This highlights the putative role of SIRT1-DBC1 complex as a pharmacological target to treat axonopathy-relatedneurodegenerative diseases.

Material and methods

Animals

All mice used in this study were maintained in the Mayo Clinic Ani-mal Breeding facility. All experimental protocols were approved by theInstitutional Animal Care and Use Committee at Mayo Clinic (protocol

no. A33209), and all studies were performed according to the methodsapproved in the protocol. DBC1 KO mice were generated as previouslydescribed (Escande et al., 2010).

Culture of cells, culture of explants of dorsal root ganglia and pharmacolog-ical treatments

MEFs and HEK 293T cells were cultured in Dulbecco's modifiedEagle's medium (5 g/L glucose) supplemented with 10% FBS and peni-cillin/streptomycin (Invitrogen). Dorsal root ganglia from new bornmice (P0-7) were plated onto poly L-lysine coated 35 mm dishes andcultured in NeurobasalMedium plus supplements (B27 and Glutamax),penicillin-streptomycin and Nerve Growth Factor (50 ng/ml) for2–3 days. To obtain dissociated DRG neurons, ganglia were collectedin Hank's Salt Balanced Solution (HBSS) and treated with collagenase0.25% w/v for 45 min at 37 °C. Following collagenase digestion, DRGswere washed with HBSS free of calcium and magnesium and treatedwith trypsin 0.25% w/v for 15 min at 37 °C. Mechanical dissociation ofneurons was achieved by continuously pipetting using a fire polishedPasteur pipette until a homogeneous cell suspension was obtained.The cells were applied onto a 30% Percoll solution in HBSS, in a 15 mlconical centrifuge tube and centrifuged at 1200 rpm for 10 min. Neu-rons were recovered from the bottom of the tube, washed with supple-mented Neurobasal and plated onto poly-L-lysine coated coverslips.When NAD+ was included in the culture media (1 mM or 10 mM), itwas added 24 h prior to axotomy. Resveratrol (Enzo Life Sciences Inter-national) was tested both at 50 μM and 100 μM in the culture media. Inone set of experiments, resveratrol was added 4 h before axotomy intwo steps: for 2 h and then replacedwith newmedia containing freshlyprepared resveratrol for twomore hours. When added after axotomy, itwas done as a single pulse. Suramin (Biomol International) was used at100 μM and added 4 h before axotomy, together with resveratrol. Oncethe DRGs had developed an axonal field (~2 mm in diameter, typicallyafter 2–3 days in culture), we proceeded to axotomize the axons usinga surgical blade scalpel number 15.

Immunofluorescent labeling

DRG and dissociated neurons were fixed by immersion in 4% para-formaldehyde in phosphate buffer saline, pH 7.2, for 15 min at 37 °C.They were washed in PBS and 0.01% Triton X 100, pH 7.2, three timesfor 5 min each and then immersed in an immunoblocking solution,composed of 5% normal goat serum in PBS, 5 mM NaN3 for 30 min. In-cubation with primary antibodies was done overnight. Coverslipswere washed three times with PBS and incubated for 60 min withsecondary antibodies conjugated to Alexa fluorophores having an exci-tation maximum at either 488 nm or 546 nm. The immunostainedspecimens were washed three times for 5 min each before beingmounted in Pro Long Gold anti fade reagent (Invitrogen). Control exper-iments were done in parallel with secondary antibodies alone.

The following antibodies were used: (1) polyclonal antibody againstDBC1 protein (Bethyl, cat IHC00135-1 for immunohystochemistry andBethyl, cat A300-432A for Western blot); (2) monoclonal anti phos-phorylated 200 kDa neurofilament (clone SMI 31, from Covance);(3) polyclonal anti SIRT1 (Cell Signaling, cat 2028); (4) goat anti mouseIgG Alexa 546 conjugate and goat anti rabbit Alexa 488 conjugate(Invitrogen).

Microscopy

The confocal microscope used consisted of a confocal moduleFluoview FV 300 equippedwith a diode emitting at 405 nm and a kryp-ton 488 and a He-Ne 543 lasers in combination with an upright Olym-pus BX61 microscope. Images were obtained with a Plan Apox60 oil,1.42 NA lens. The acquisition parameters (laser intensity, gain, offsetand photomultiplier intensity) used to capture images of SIRT1 and

93A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

DBC1 distribution in cultured cells were the same as those used in neg-ative control experiments. Imageswere processedwith ImageJ software(NIH).

Transfections

Transient overexpression of SIRT1 and DBC1 was performed in HEK293T cells as previously described (Kang et al., 2009). Transient overex-pression was performed using Lipofectamine 2000 (Invitrogen) for48 h, following the manufacturer's instructions.

Quantification of axonal degeneration in vitro

The protective effect of 10 mM NAD+ against Wallerian degenera-tion has been described previously (Araki et al., 2004; Wang et al.,2005). We assessed axonal degeneration using two alternative criteria:a) counting the number of axons immediately after the axotomy andexpressing the subsequent degeneration as the percentage of axons re-maining attached to the culture plate without evidence of fragmenta-tion, as has been done previously (Araki et al., 2004; Wang et al.,2005). b) Scoring the appearance of beads along a group of individual-ized axons and expressing the degeneration as the number of beadswithin an axonal length of 100 μm. The kinetics of degeneration wasfollowed from the moment of axotomy (time 0) and every 12 h for atotal of 36 h. To do this, we recorded the same microscopy field ateach time point so we could follow the time course of degeneration ofindividual axons. Using both techniques, we obtained a series of fourscores corresponding to degeneration at 0, 12, 24 and 36 h afteraxotomy.

Fig. 1. Comparative approaches to evaluate in vitro axonal degeneration. A: Successive photoggenerate by axotomy. The comparison was carried out using by two different experimental cBar is 50 μm. B: Kinetics of axonal degeneration expressed as the percentage of axons remaininresentative axons scored at different time points. C: The axonal degeneration in the same micreach 12 h and expressing it as the number of beads present each 100 μm of axon length. Bars

Measurement of SIRT1 deacetylase activity

SIRT1 activitywasmeasuredwith afluorimetric assay (ENZO catalognumber BML-AK555-0001), as previously described (Escande et al.,2010; Nin et al., 2012). Cells were lysed and proteins extracted withNETN buffer as described in the following section. Protein concentrationwas measured by Bradford and protein concentrations were equalizedwith deacetylase buffer (50 mM Tris–HCl pH 8; 137 mM NaCl;2.7 mM KCl; 1 mM MgCl2; 1 mg/ml BSA). Samples were incubated10 min at 30 °C to allow for NAD+ degradation, and incubated for10 min with DTT 2 μM., Sample containing 30–50 mg of total proteinwas transferred to 6wells of a 96well plate and a solution of deacetylasebuffer containing 100 mM of substrate and 5 μM trichostatin A wasadded to the wells. 100 μM NAD+ was added to half the wells. Thereaction proceeded for 2 h at RT and then a developer was added for1 h. Fluorescence was measured with an excitation of 360 nm andemission at 460 nm. SIRT1 activity was calculated as NAD+ dependentfluorescence.

Immunoprecipitation assays and Western Blotting

Cultured cells (MEFs, HEK 293T) and dorsal root ganglia were lysedin NETN buffer (20 mM Tris–HCl, pH 8.0; 100 mM NaCl; 1 mM EDTA;and 0.5% NP-40), 5 mM NaF, 50 mM 2-glycerophosphate, and a prote-ase inhibitor cocktail (Roche). Homogenates were incubated at 4 °Cfor 30 min under constant agitation, and then centrifuged at 10,000 gfor 10 min at 4 °C. For immunoprecipitation 1 mg protein was incubat-ed with 20 μl Protein A-G (Santa Cruz Biotechnology Inc.) and 1 μg an-tibody (anti SIRT1, Abcam cat 12193 and anti DBC1, Bethyl Laboratories

raphs of the same microscopy field were taken to register a group of axons pruned to de-onditions: degeneration under control conditions and in the presence of 10 mM NAD+.g attached to the plate. Each point in the graph represents a mean value of a group of rep-oscopy field but followed by scoring the appearance of beads in a group of selected axonsin B and C are SEM.

94 A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

cat A300-432) for 1 h at 4 °C under constant agitation. Nonspecific IgG(Santa Cruz Biotechnology Inc.) was used as control. Finally, immuno-precipitates were washed two times with cold NETN before additionof 2× Laemmli buffer. SIRT1 and DBC1 were detected by Western blotwith anti SIRT1 antibody (Cell Signaling Technology, cat 2028) andanti DBC1 (Bethyl Laboratories, cat A300-432). Changes in phosphoryla-tion of AMPK were assessed by Western blot with antibodies againstphosphorylated AMPK (Thr172) and (pan, α-subunit, β-subunit)AMPK antibodies (Cell Signaling Technology). Western blots were de-veloped using secondary antibodies or protein A-HRP and SuperSignalWest Pico chemiluminescent substrate (Pierce). Quantification wasperformed using ImageJ software (NIH).

Sciatic nerve degeneration

Two-month-old mice were treated daily for a total of 3 weeks withresveratrol (intraperitoneally, 20 mg/kg) dissolved in DMSO. Controlanimals were injected with the vehicle. On the last day of treatment,

Fig. 2. Resveratrol delays axonal degeneration in vitro. A: Both NAD+ and resveratrol (50 μM)noticeable than that of 1 mM NAD+ (p b 0.001). On the other hand, the contribution of resNAD+ concentrations (non significant differences between 10 mM NAD+ alone and 10 mM Nthe drugwas added to the culturemediumbefore or just after axotomy (non significant differenceexerted by resveratrol. A series of representative results obtained in axotomized axons exposed toNAD+ plus resveratrol and 100 μM suramin. The samemicroscopy field was registered immediatgions to depict axonal beads (thick arrows) and fragmentations (thin arrow) developed in axons pafter axotomy. Note that the axon resulted better preserved in the presence of resveratrol. Bar isdifferent animals) yielded the results summarized in D. Bars are SEM.

the animals were anesthetized with ketamine (85 mg/kg) and xilacine(10 mg/kg) and bipolar cuff electrodes were implanted in the rightsciatic nerve for electrical stimulation. Wallerian degeneration was in-duced by crushing the nerve for 30 s with a number 5 Dumont forceps,proximal to the cuff. Nerve degeneration was evaluated by recordingthe excitability as the threshold motor response of the right foot, usingcombinations of time and intensity of electrical stimulation (0.1–4 msand 1–100 μA, respectively). A series of strength–duration curveswere obtained from each experimental group immediately before(time 0) and 18 h after nerve crushing. The curves were used to calcu-late two classic parameters of axonal excitability: the rheobase andstrength–duration time constant (chronaxie). While rheobase repre-sents the minimal current input of infinite duration needed to reachthe depolarization threshold, the chronaxie represents the minimaltime required for an electrical current (whose magnitude is two timesthe rheobase), to be applied to trigger an axon potential. Chronaxieand rheobase were estimated from a linear transformation based onthe strength–duration function (Lapicque's equation) as follows:I = R + A(1/D), where I is the strength of the stimulus (measured in

can modify the course of axonal degeneration. The effect of 10 mM NAD+ is much moreveratrol to the overall effect is clear at low NAD+ concentrations but marginal at highAD+ plus resveratrol). B: Axonal protection exerted by resveratrol was unchanged whens between groups RSV pre and RSVpost axotomy). C–D: Suramin reverts the neuroprotectiondifferent culture conditions: 1 mMNAD+as control, 1 mMNAD+plus resveratrol and 1 mMely after axonal injury and 12 h later. Insets in C (upper row) aremagnification of framed re-runed to degenerate. Insets in C (middle row) aremagnifications of the same axonbefore and50 μm. The quantification of 10 to 16 different axonal fields for each condition (from 4 to 6

Fig. 3. Presence and localization of SIRT1 and DBC1 proteins in cultured DRG neurons. A: A neuron double staining for SIRT1 and 200 kDa neurofilament proteins. Note that while anti NFantibody labels the pericarion and neurite extensions, SIRT1 is heavily labeled in cell nucleus (arrow), soma and growth cone and to a lesser extent, in neurites. The inset is amagnificationof the framed areawhere thepunctate nature of SIRT1 label is better appreciated. B: Single confocal Z plane of axons grown fromaDRGexplant. Insets aremagnifications of the same imageto illustrate the fine granular labeling of axons in greater detail. Arrows point to a fibroblast nucleus showing the standard distribution of nuclear SIRT1. C: Negative control experiment(no primary antibody). The arrow points to the nuclear region. Weak and diffuse label of axons is depicted in the inset. D: SIRT1was detected inWestern blots of proteins extracted fromDRG explants and inmouse brain (used as a positive control). E: A neuron in culture double labeled for DBC1 and 200 kDa neurofilament proteins. Note thatwhile anti NF antibody labelsthe pericarion and neurite extensions, DBC1 is heavily labeled in cell nucleus, and to a lesser extent, in the soma and in neurites. The framed area (axon)wasmagnified to illustrate the finegranular nature of DBC1 reaction. F: Single confocal Z plane of axons grown from aDRG explant. The framed areawasmagnified to illustrate theDBC1presence and distribution in axons ingreater detail. The arrow points to a nucleus of a fibroblast showing the standard distribution of nuclear DBC1. G: Negative control experiment with secondary antibody alone where nu-clear labeling is absent and axons developed a tiny and diffuse signal. H: DBC1 was detected in Western blots of proteins extracted from DRG explants and in mouse brain (as a positivecontrol). Bars are 20 μm.

95A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

amperes), R is the rheobase value of the nerve, D is the duration of thestimulus (in seconds) and A is the product of 2R by the value of thechronaxie.

Statistics

Each experimental condition of the in vitro experiments comprisedat least five different independent experiments. Each animal yieldedat least 10 axonal fields per experimental condition. The number ofbeads was recorded in the axons that met the following criteria:a) axons that could be traced all along its length and longer than70 μm; b) individual axons that could be identified in frames corre-sponding both to 0, 12 and 24 h. The statistical analysis was assessed

by two way ANOVA or two-tailed Student's t test, as indicated. Ap-value less than 0.05 was considered significant.

In vivo studies of nerve degeneration implied two groups of fiveanimals each (five sciatic nerves for each experimental condition).

All values concerning quantification are presented as mean ± SEM.

Results

Quantification of the axonal degeneration. Validation of the method

The known protective effect of NAD+ at high concentration (10 mM)againstWallerian degeneration (Araki et al., 2004;Wang et al., 2005)wasused as experimental condition todevelop an alternativemethodused forquantification of early degeneration events (see Material and methods).

Fig. 4. Resveratrol triggers the dissociation of SIRT1 from DBC1, induces AMPK phosphorylation and reverts SIRT1 inhibition by DBC1. A:Western blot showing that the amount of SIRT1detected in extracts and immunoprecipitates of DRGs (inputs) does not vary with the experimental conditions (1 mM NAD+ and 1 mM NAD+ plus resveratrol 100 μM); however theamount of DBC1 co-purified with SIRT1 antibodies depends on resveratrol treatment. B: Densitometric analysis of protein bands corresponding to SIRT1 and DBC1. The amount ofDBC1 dissociated from SIRT1 is expressed as a ratio DBC1/SIRT1 × 100. The data represents the average of 5 independent experiments, p b 0.05 (t-test). Bars are SEM. C: SIRT1/DBC1association evaluated by co-immunoprecipitation in MEFs cells. Cells were treated with resveratrol (100 μM) for 2 h before performing immunoprecipitation for DBC1. Proteins wereimmunoblotted for SIRT1 and DBC1. D: Phosphorylation of AMPK at Thr172 and it correlation with SIRT1 activity (E) measured in MEFs cells extracts and expressed as Arbitrary Unitsof Fluorescence upon treatment with 100 μM resveratrol. Note that both SIRT1 activity and AMPK phosphorylation decreased after 2 h of incubation with resveratrol. Bars are SEM. F:SIRT1 activity measured in 293T cells transfected with FLAG-SIRT1, FLAG-SIRT1 + Myc-DBC1, or FLAG-SIRT1 + ΔLZ Myc-DBC1, a mutant DBC1 that does not have the LZ domain anddoes not bind to SIRT1. SIRT1 activity was measured and shown as fold change respect to the control. *Indicates p b 0.05, ANOVA test, n = 3.

96 A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

Fig. 1A shows the morphological changes of a group of axons duringWD(control) and the neuroprotective effect of NAD+ at high concentration.Figs. 1B and 1C show that there is correlation between the axonaldegeneration valuesmeasuredwith bothmethods, the previously report-ed byAraki et al. (2004), and the newmethodpresented here. Indeed, thedifference between NAD+-treated and control axons 12 h after axotomywas more evident when the new quantitative method of scoring(counting the number of beads) was used. We conclude that this ap-proach is more suitable to measure early degenerative events (i.e. upto 12 h).

Effects of resveratrol on the kinetics of axonal degeneration

Next, we tested the effect of resveratrol on the time course ofWallerian degeneration and how this effect was modulated by NAD+.Fig. 2A shows the degree of axonal degeneration 12 h after axotomy.There is no effect of NAD+ at a concentration of 1 mM, whereasthe 10 mM NAD+ concentration induces a neuroprotective effect(p b 0.001). On the other hand, resveratrol protects against WD in-dependent of NAD+. There was a trend for resveratrol to furtherimprove the protection promoted by 10 mM NAD+. Interestingly, wefound that resveratrol protects against WDwhen the axons were treated

before transection, but also when the axons were treated immediatelyafter transection (Fig. 2B). This suggests that the effect of resveratrolon Wallerian degeneration is a local axonal event, independent of itsbiological action in the soma.

Since resveratrol is a putative SIRT1 activator (Wood et al., 2004), wedecided to investigate the role of SIRT1 in the protective effect observed.For this, we followed the course of WD in the presence of resveratroland suramin, a SIRT1 inhibitor (Trapp et al., 2007). Suramin (100 μM)reverted the effect of resveratrol to levels similar to those of controlaxons (Figs. 2C and D), suggesting a role of SIRT1 in the molecularmechanism behind the protection against Wallerian degeneration.

SIRT1 and DBC1 are present in axons of DRG neurons in culture

The effect of resveratrol on Wallerian degeneration is mainly localsince it is also observed when resveratrol is added to the media afteraxotomy (Fig. 2B). We have recently proposed that SIRT1 activationby resveratrol involves dissociation of SIRT1 from its endogenous inhib-itor DBC1 (Nin et al., 2012). Therefore, we decided to investigatewheth-er SIRT1 and DBC1 are present in axons. Using high resolution confocalmicroscopy in dissociatedDRGneurons andDRG explants in culture,weobserved that both SIRT1 and DBC1 compartmentalize in soma and

97A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

neurites. We used the nuclear labeling in fibroblasts present in the cul-ture as positive controls of antibody specificity. In fibroblasts, anti SIRT1and anti DBC1 antibodies yielded a strong nuclear signal, mostly on thechromatin and excluded from nucleolus (arrows in Figs. 3B and F), rem-iniscent of the pattern already described in cultured cells (Langley et al.,2002; Sundararajan et al., 2005). A punctuated signal can be observed inthe cytoplasm of fibroblasts, consistent with previous reports showingthe cytoplasmic localization for SIRT1 (Aquilano et al., 2010). On theother hand, the SIRT1 distribution in DRG neurons was different fromfibroblasts. In neurons, the cytoplasm showed denser labeling (Fig. 3A)than the fibroblasts. Interestingly, we found that SIRT1 and DBC1 areexpressed in axons (insets in Figs. 3A and B; insets in 3E and F), wherethey show a punctuated and less dense signal than observed in thesoma. The presence of SIRT1 in axons also reached growth cones(Fig. 3A), covering the central region and extending to more peripheralareas of the cone, beyond the region occupied by neurofilaments(Fig. 3A). The same antibody used for immunofluorescence of SIRT1 inneurons was used to detect the protein in Western blot of extracts ofDRG explants grown in vitro. The single band detected just below120 kDa, suggests that the antibody used for immunolocalization isspecific and reacts only against SIRT1, at least by Western blot analysis(Fig. 3D).

Resveratrol dissociates SIRT1 from DBC1 in DRGs in culture

Next, we investigated if resveratrol induces dissociation of SIRT1from DBC1 in neurons, as a putative mechanism for the protection

Fig. 5. Genetic deletion of DBC1 recapitulates the effect of resveratrol onWallerian degenerationerate by axotomy.Note that after 12 h of axotomy, the structure andmorphology are better presby the axotomy are more resistant toWDwhen 1 mMNAD+ is added to the culture media (p b

further improve their performance by the pre-treatment with resveratrol. C: Time course of de

against Wallerian degeneration. DRGs in culture were treated withresveratrol and later SIRT1 was immunoprecipitated. Fig. 4A showsthat DBC1 co-immunoprecipitates with SIRT1 in untreated cultures, in-dicating that in DRG neurons, these proteins form a macromolecularcomplex. After 2 h of treatment with resveratrol (50 μM resveratroland 1 mM NAD+), the amount of co-immunoporecipitated DBC1was reduced, suggesting the disassembling of the SIRT1-DBC1 com-plex. We measured the amount of dissociation and found thatresveratrol treatment induces a 30% decrease in SIRT1/DBC1 associ-ation (Fig. 4B).

Resveratrol triggers the dissociation of DBC1 from SIRT1, induces AMPKphosphorylation, and SIRT1 activation in cells

It has been reported that resveratrol is an activator of AMPK in sev-eral models including DRG neurons (Dasgupta and Milbrandt, 2007).Since AMPK activation increases SIRT1 activity (Canto et al., 2009; Ninet al. 2012), we investigated if the same pharmacological treatmentused for DRGs activate AMPK, leading to dissociation of the SIRT1/DBC1 complex and SIRT1 activation in mouse embryonic fibroblasts(MEFs). First, we evaluated if resveratrol treatment induces dissociationof SIRT1 fromDBC1 inMEFs.We found that 100 μMresveratrol promot-ed complete dissociation of the SIRT1/DBC1 complex (Fig. 4C). Next, weinvestigated the relationship between the phosphorylation of AMPK atThr172 and the enzymatic activity of SIRT1 under the same experimen-tal conditions.

inWT axons. A:WT and DBC1 KO axons treated with NAD+ 1 mM and pruned to degen-erved inDBC1KOaxons (insets). Bar is 50 μm. B:DBC1KOaxons committed to degenerate0.01 between 0 mMNAD+ and 1 mMNAD+ DBC1 KO). Note that DBC1 KO axons do notgeneration for both groups in the presence of 1 mM NAD+. Bars are SEM.

Fig. 6. Effect of resveratrol on sciatic nerve degeneration. A: Strength–duration curves incontrol and resveratrol-treated mice nerves, before and 18 h after crushing. B: Rheobasecalculated from the strength duration curves indicates that resveratrol induces a differ-ence in the excitability of intact nerves. The differences between both groups are presentfrom the beginning (time 0 h, p = 0.05) and increase as nerves degenerate (time 18 h,p b 0.001). C: The values in chronaxie between control and resveratrol-treated micenerves showed no significant differences either at 0 h or at 18 h after injury.

98 A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

SIRT1 activity was determined in cell extracts using a SIRT1 fluo-rometric kit (Enzo Life Sciences) and expressed as Arbitrary FluorescentUnits (AFU). This assay uses a small peptide derived from p53 thatincludes an acetylated lysine (K382), that is a target for deacetylationby SIRT1. Subsequently, the deacetylated substrate is transformed in afluorescent compound whose accumulation correlates with the enzy-matic activity. Determination of cellular SIRT1 activity using this meth-od depends on the addition of exogenous NAD+ (see Material andmethods), and the activity is inhibited when the cellular extracts arealso incubated with nicotinamide, suramin, or EX527, three inhibitorsof SIRT1 (Aksoy et al., 2006; Escande et al., 2010; Nin et al., 2012). Wefound that treatment of cells with resveratrol induces a sustainedphosphorylation of AMPK (Fig. 4D), which correlates with an increasein SIRT1 activity (Fig. 4E). These results suggest that DBC1 may berequired to increase SIRT1 activity. To further study this possibility, wetransfected 293T cells with SIRT1 and DBC1, or with ΔLZ-DBC1 (a mu-tant DBC1 lacking the LZ Sirt1-bindingdomain). As previously described(Nin et al. 2012), we observed that co-expression of SIRT1 with DBC1decreases SIRT1 activity. In contrast, deacetylase activity was notdecreased in cells co-transfected with the truncated form of DBC1(ΔLZ-DBC1) that does not bind to SIRT1. In the condition where DBC1was not bound to SIRT1, resveratrol failed to further activate SIRT1,showing that SIRT1 activation by resveratrol involves the release of in-hibition by DBC1 (Fig. 4F).

DBC1 regulates Wallerian degeneration

Next, we studied if the neuroprotective effect of resveratrol could berecapitulated by deletion of the SIRT1 inhibitor DBC1. We have shownpreviously that DBC1 KO mice have constitutively high SIRT1 activity(Escande et al., 2010). We observed that in the absence of NAD+, theaxons belonging to DRGs derived fromDBC1 knockoutmice degenerateas fast as WT controls. However, when 1 mM NAD+ was added to theculture medium, DBC1 KO axons significantly improved their survivalrate compared to WT axons (Figs. 5A–C). Moreover, in contrast toaxons fromWT mice, the degeneration of DBC1 KO axons was insensi-tive to the addition of resveratrol (Fig. 5B). In fact, the degeneration ofDBC1 KO axons was similar to that observed in WT DRGs treated withresveratrol. These experiments suggest that DBC1 participates in thesame molecular pathway involved in the neuroprotection exerted byresveratrol.

Resveratrol modulates Wallerian degeneration in vivo

Finally, we investigated the effect of chronic resveratrol treatmenton axonal degeneration in vivo. We treated mice with resveratrol(20 mg/kg daily) or vehicle for 3 weeks. At the end of the treatment,we performed sciatic nerve crushing and followed the course of nervedegeneration by electrophysiology. To evaluate the functionality of sci-atic nerves from control and resveratrol-treated mice, we constructedstrength–duration curves (Fig. 6A).

Prior to the lesion, we observed a difference in the electrophysiologi-cal performance between control and resveratrol-treated nerves. Infact, sciatic nerves belonging to resveratrol-treated animals appear to bemore excitable than controls (rheobase values of 3.5 and 6.8 μA,respectively; p = 0.05; see Fig. 6B). On the other hand, the differencesin chronaxie (that measures how rapidly the membrane potential canchange in response to a current) resulted not significant. After 18 h ofcrushing, the rheobase of control nerves increased much more than theresveratrol-treated nerves (12.6 and 6.2 μA, respectively, p b 0.001; seeFigs. 6A and B), indicating that the excitability of control nerves decaysfaster than resveratrol-treated ones. At this time point, chronaxie valuestend to decay in regard to those observed before crushing; howevernon statistical differences were found in the same experimental groupbefore and after crushing, neither between groups at any time point(Fig. 6C).

Discussion

The mechanisms that control axonal degeneration are still poorlyunderstood. However, it has become clear that axonal dysfunction

99A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

plays a central role in several neurodegenerative diseases (Adalbert andColeman, 2013; Coleman, 2013). Thus, in order to develop new treat-ments for these diseases, it is crucial to understand the molecular path-ways underlying axonopathies. The appearance of the Wlds mutationand its protective effects on axons suggest the existence of commonmo-lecularmechanisms that lead to axonal degeneration, despite the natureof the initial insult. In this regard it has been shown that the Wlds pro-tein delays axonal degeneration in several animal models of central orperipheral human neurodegenerative diseases (Ferri et al., 2003;Fischer et al., 2005; Gillingwater et al., 2004, 2006; Sajadi et al., 2004).Araki et al. (2004) first postulated that SIRT1 is the main enzyme re-sponsible for the phenotype observed in Wlds mutants. However, Wanget al. (2005) challenged this idea, and proposed thatmost of the observedprotective effect in the Wlds mutant correlate with the intra axonal con-centration of NAD+. Using essentially the same experimental model tostudy WD, Conforti et al. (2007) did not find effects of either 1 mMNAD+ or 0.1 mM resveratrol on the time course of axonal degeneration.

In our hands, the treatment of DRGs with 0.05 to 0.1 mM of resver-atrol produced a consistent delay on axonal degeneration. In agreementwith the postulated involvement of SIRT1 in the resveratrol protectiveeffect against axonal degeneration (Araki et al., 2004), we found thatpharmacological inhibition of SIRT1 reverts the protective effect of res-veratrol (Figs. 2C and D). Interestingly, the effect of resveratrol on axo-nal degeneration was local, since it was also present when the drugwas added just after axotomy. In agreement with the local effect of res-veratrol on axonal degeneration, we found that SIRT1 is present inaxons and particularly in growth cones (Figs. 3A and B). Of note, it hasrecently been reported that SIRT1 is up regulated inmouse sensory neu-rons during axonal regeneration, in vivo and in vitro, and that this in-creased expression is essential for axonal regrowth (Liu et al. 2013). Inaddition, Li et al. (2013) described the presence of endogenous SIRT1in axons of hippocampal neuron growth in vitro. We also found thatDBC1 is present in the axonal territory (Figs. 3E and F) and that it canbe immunoprecipitated as a protein complex from cultured DRGs(Fig. 4A). These findings suggest that SIRT1 activitymight be locally reg-ulated by DBC1 in the axonal territory. Based on that, we evaluated theinvolvement of this protein duringWD.We found that deletion of DBC1recapitulates the neuroprotective phenotype produced by resveratrol inwild type axons. We propose that DBC1 KO axons are protected againstWallerian degeneration due to the increased basal SIRT1 activity(Escande et al. 2010). SIRT1 activation through dissociation from DBC1has been previously shown by us (Escande et al., 2010) Indeed, wehave recently shown that cells exposed to AMPK and PKA activatorshave decreased DBC1 binding to SIRT1 and increased SIRT1 activity bya mechanism that is AMPK dependent (Nin et al., 2012). Because it isextremely difficult to measure NAD+-dependent deacetylase activityin culturedDRGs,we investigated the effect produced by the samephar-macological treatment but using mouse fibroblasts. As expected, incu-bation of cells with 100 μM resveratrol for 2 h induced an increase inSIRT1 activity (Fig. 4E) as well as a sustained AMPK phosphorylation(Fig. 4D). Additionally, we observed that SIRT1 dissociation from DBC1is necessary to increase SIRT1 enzymatic activity (Fig. 4F). We suggestthat under the presence of resveratrol, SIRT1 activity is likely to be in-creased also in DRGs.

Different from the effect observed by resveratrol incubation, whichprotects against axonal degeneration even in the absence of NAD+, wefound that the protective effect of DBC1 KO on axonal degeneration isNAD+-dependent (Fig. 2A). One possible explanation for this differenceis that resveratrol activates AMPK, which in turn not only activatesSIRT1, but also promotes an increase in intracellular NAD+ levels(Canto et al., 2009). If this is also happening in DRG neurons, thenAMPK activation by resveratrol would be promoting SIRT1/DBC1 disso-ciation and also an increase in NAD+ levels, thus leading to full SIRT1 ac-tivation. On the other hand, in the case of DBC1 KO axons, althoughSIRT1 is constitutively active, NAD+ still needs to be added in order toget fully active SIRT1.

Our results go against what was shown by Conforti et al. (2007),who did not observe any effect of 0.1 mM resveratrol on axons incubat-ed 24 h before axotomy, or those of Suzuki and Koike (2007) usingcerebellar granule cells and colchicine treatment as a model to studyWD. Moreover, these authors showed that 0.1 mM resveratrol acceler-ated degeneration after 24 h of incubation. On the other hand, it hasbeen reported that resveratrol can produce opposite effects dependingon the dose used. In this regard, it has been shown that 0.1 μM resvera-trol induces growth and replication of mesenchymal cells in culture,whereas 0.5 μM promotes cellular senescence in the same cells (Peltz,2012). A retrospective analysis of these data suggests that the effectsof resveratrol are highly dependent on the conditions in which it isused (mainly the period of time and concentrations to which the cellsare exposed). This fact could provide an explanation for the apparentlycontradictory results obtained using similar, although not identical, ex-perimental conditions. Furthermore, in our hands axonal degenerationwas more accurately evaluated when it was recorded at early stagesand using a qualitative approach as the appearance of axonal beads(Kilinc et al., 2008) rather than a categorical one (see Figs. 1B and C).These observations, together with the fact that NAD+ plays an impor-tant role by modulating SIRT1 deacetylase activity, probably furthercontribute to explain thementioned contradictory results.We concludethat both the effect of resveratrol as well as the involvement of DBC1in Wallerian degeneration justifies the re-evaluation of the putativeinvolvement of SIRT1 in axonal metabolism.

As reported by Moldovan et al. (2009) we found an impairment ofexcitability of sciatic nerves pruned to degenerate by crushing at least14 h after injury. At this point, more current was required to evoke afoot motor response. This impaired excitability was evidenced as anincreased rheobase. A delay in the development of this phenotype wasobserved in resveratrol treated nerves. Interestingly, differences inrheobase between control and resveratrol-treated nerveswere detectedeven before crushing, which suggests that the differences observedbetween both groups during degeneration may be a consequence of apre-existing difference of nerve physiology induced by the chronictreatment with resveratrol. Rheobase is a measure of excitability thatreflects the passive properties of the membrane. In this regard, a de-creased sodium conductance of the membrane has been postulatedas a mechanism to explain the impaired excitability observed indegenerating nerves (Moldovan et al., 2009). Thus, it is conceivable thatRSVmay be affectingmembrane properties. At this point, it is hard to es-tablish a correlation between the effects of resveratrol in vitro andin vivo. More work and complementary experimental approachesare necessary to expand these results, which could lead to a bettercomprehension of the mechanism underlying the effects of resveratrolon axon pathology.

Acknowledgments

The authors gratefully acknowledge Dr. Angel Caputi (Departmentof Integrative and Computational Neuroscience, IIBCE) for help withthe electrophysiological recordings and data analysis, and Berthil FFClasen (Institute of Clinical Medicine, Aarhus University Hospital,Aarhus, Denmark) for manuscript editing. This research was supportedin part by the Human Resources Program of the University of theRepublic (CSIC-UdelaR, Uruguay) and by the Mayo Clinic.

References

Adalbert, R., Coleman, M.P., 2013. Axon pathology in age-related neurodegenerativedisorders. Neuropathol. Appl. Neurobiol. 39, 90–108.

Aksoy, P., Escande, C., White, T.A., Thompson, M., Soares, S., Benech, J.C., Chini, E.N., 2006.Regulation of SIRT 1 mediated NAD dependent deacetylation: a novel role for themultifunctional enzyme CD38. Biochem. Biophys. Res. Commun. 349 (1), 353–359(13).

Aquilano, K., Vigilanza, P., Baldelli, S., Pagliei, B., Rotilio, G., Ciriolo, M.R., 2010. Peroxisomeproliferator-activated receptor gamma co-activator 1alpha (PGC-1alpha) and sirtuin 1

100 A. Calliari et al. / Experimental Neurology 251 (2014) 91–100

(SIRT1) reside in mitochondria: possible direct function in mitochondrial biogenesis.J. Biol. Chem. 285 (28), 21590–21599 (9).

Araki, T., Sasaki, Y., Milbrandt, J., 2004. Increased nuclear NAD biosynthesis and SIRT1activation prevent axonal degeneration. Science 305 (5686), 1010–1013.

Baur, J.A., Sinclair, D.A., 2006. Therapeutic potential of resveratrol: the in vivo evidence.Nat. Rev. Drug Discov. 5 (6), 493–506.

Beirowski, B., Adalbert, R., Wagner, D., Grumme, D.S., Addicks, K., Ribchester, R.R.,Coleman, M.P., 2005. The progressive nature of Wallerian degeneration in wild-type and slow Wallerian degeneration (WldS) nerves. BMC Neurosci. 6 (6) (1).

Canto, C., Gerhart-Hines, Z., Feige, J.N., Lagouge, M., Noriega, L., Milne, J.C., Elliott, P.J.,Puigserver, P., Auwerx, J., 2009. AMPK regulates energy expenditure by modulatingNAD+ metabolism and SIRT1 activity. Nature 458, 1056–1060.

Chini, C.C.S., Escande, C., Nin, V., Chini, E.N., 2010. HDAC3 is negatively regulated by thenuclear protein DBC1. J. Biol. Chem. 285 (52), 40830–40837.

Cohen, H.Y., Miller, C., Bitterman, K.J., Wall, N.R., Hekking, B., Kessler, B., Howitz, K.T.,Gorospe, M., de Cabo, R., Sinclair, D.A., 2004. Calorie restriction promotes mam-malian cell survival by inducing the SIRT1 deacetylase. Science 305 (5682),390–392 (16).

Coleman, M.P., 2013. The challenges of axon survival: introduction to the special issue onaxonal degeneration. Exp. Neurol. 246, 1–5.

Conforti, L., Fang, G., Beirowski, B., Wang, M.S., Sorci, L., Asress, S., Adalbert, R., Silva, A.,Bridge, K., Huang, X.P., Magni, G., Glass, J.D., Coleman, M.P., 2007. NAD(+) andaxon degeneration revisited: Nmnat1 cannot substitute forWld(S) to delayWalleriandegeneration. Cell Death Differ. 14 (1), 116–127.

Dasgupta, B., Milbrandt, J., 2007. Resveratrol stimulates AMP kinase activity in neurons.104 (17), 7217–7222.

Escande, C., Chini, C.C., Nin, V., Dykhouse, K.M., Novak, C.M., Levine, J., van Deursen, J.,Gores, G.J., Chen, J., Lou, Z., et al., 2010. Deleted in breast cancer-1 regulates SIRT1activity and contributes to high-fat diet-induced liver steatosis in mice. J. Clin. Invest.120, 545–558.

Ferri, A., Sanes, J.R., Coleman, M.P., Cunningham, J.M., Kato, A.C., 2003. Inhibiting axon de-generation and synapse loss attenuates apoptosis and disease progression in a mousemodel of motoneuron disease. Curr. Biol. 13 (8), 669–673 (15).

Fischer, L.R., Culver, D.G., Davis, A.A., Tennant, P., Wang, M., Coleman, M.P., Asress, S.,Adalbert, R., Alexander, G.M., Glass, J.D., 2005. The Wld(S) gene modestly prolongssurvival in the SOD1(G93A) fALS mouse. Neurobiol. Dis. 19, 293–300.

Fulco, M., Schiltz, R.L., Iezzi, S., King, M.T., Zhao, P., Kashiwaya, Y., Hoffman, E., Veech, R.L.,Sartorelli, V., 2003. Sir2 regulates skeletal muscle differentiation as a potentialsensor of the redox state. Mol. Cell 12 (1), 51–62 (Erratum in: Mol Cell.11;20(3):491).

Fulda, S., Debatin, K.M., 2006. Targeting inhibitor of apoptosis proteins (IAPs) for diag-nosis and treatment of human diseases. Recent Pat. Anticancer Drug Discov. 1 (1),81–89.

Gillingwater, T.H., Haley, J.E., Ribchester, R.R., Horsburgh, K., 2004. Neuroprotection aftertransient global cerebral ischemia in Wld(s) mutant mice. J. Cereb. Blood FlowMetab. 24 (1), 62–66.

Gillingwater, T.H., Ingham, C.A., Parry, K.E., Wright, A.K., Haley, J.E., Wishart, T.M.,Arbuthnott, G.W., Ribchester, R.R., 2006. Delayed synaptic degeneration in the CNSof Wlds mice after cortical lesion. Brain 129 (Pt 6), 1546–1556.

Imai, S., Armstrong, C.M., Kaeberlein, M., Guarente, L., 2000. Transcriptional silencing andlongevity protein Sir2 is an NAD-dependent histone deacetylase. Nature 403 (6771),795–800 (17).

Kang, H., Jung, J.W., Kim, M.K., Chung, J.H., 2009. CK2 is the regulator of SIRT1 substrate-binding affinity, deacetylase activity and cellular response to DNA-damage. PLoS One4 (8), e6611 (14).

Kilinc, D., Gallo, G., Barbee, K., 2008. Mechanically-induced membrane poration causesaxonal beading and localized cytoskeletal damage. Exp. Neurol. 212 (2), 422–430.

Kim, E.J., Um, S.J., 2008. SIRT1: roles in aging and cancer. BMB Rep. 41 (11), 751–756 (30).Kim, J.E., Chen, J., Lou, Z., 2008. DBC1 is a negative regulator of SIRT1. Nature 451,

583–586.Langley, E., Pearson, M., Faretta, M., Bauer, U.M., Frye, R.A., Minucci, S., Pelicci, P.G.,

Kouzarides, T., 2002. Human SIR2 deacetylates p53 and antagonizes PML/p53-induced cellular senescence. EMBO J. 21 (10), 2383–2396 (15).

Li, X.H., Chen, C., Tu, Y., Sun, H.T., Zhao, M.L., Cheng, S.X., Qu, Y., Zhang, S., 2013. Sirt1 pro-motes axonogenesis by deacetylation of Akt and inactivation of GSK3.Mol. Neurobiol.48 (3), 490–499.

Liu, C.M., Wang, R.Y., Saijilafu, Jiao, Z.X., Zhang, B.Y., Zhou, F.Q., 2013. MicroRNA-138 andSIRT1 form amutual negative feedback loop to regulatemammalian axon regeneration.Genes Dev. 27 (13), 1473–1483 (1).

Lunn, E.R., Perry, V.H., Brown, M.C., Rosen, H., Gordon, S., 1989. Absence of Walleriandegeneration does not hinder regeneration in peripheral nerve. Eur. J. Neurosci. 1(1), 27–33.

Mayeux, R., 2003. Epidemiology of neurodegeneration. Annu. Rev. Neurosci. 26, 81–104.Moldovan, M., Alvarez, S., Krarup, C., 2009. Motor axon excitability during Wallerian

degeneration. Brain 132 (Pt 2), 511–523.Nakae, J., Cao, Y., Daitoku, H., Fukamizu, A., Ogawa, W., Yano, Y., Hayashi, Y., 2006.

The LXXLL motif of murine forkhead transcription factor FoxO1 mediates Sirt1-dependent transcriptional activity. J. Clin. Invest. 116 (9), 2473–2483.

Nakahata, Y., Kaluzova, M., Grimaldi, B., Sahar, S., Hirayama, J., Chen, D., Guarente,L.P., Sassone-Corsi, P., 2008. The NAD+-dependent deacetylase SIRT1 modulatesCLOCK-mediated chromatin remodeling and circadian control. Cell 134 (2),329–340 (25).

Nin, V., Escande, C., Chini, C.C., Giri, S., Camacho-Pereira, J., Matalonga, J., Lou, Z., Chini,E.N., 2012. Role of deleted in breast cancer 1 (DBC1) protein in SIRT1 deacetylaseactivation induced by protein kinase A and AMP-activated protein kinase. J. Biol.Chem. 287 (28), 23489–23501.

Peltz, L., Gomez, J., Marquez, M., Alencastro, F., Atashpanjeh, N., Luang, T., Bach, T., Zhao, Y.,2012. Resveratrol exerts dosage anddurationdependent effect onhumanmesenchymalstem cell development. PLoS One 7 (5), e37162.

Rodgers, J.T., Lerin, C., Haas, W., Gygi, S.P., Spiegelman, B.M., Puigserver, P., 2005. Nutrientcontrol of glucose homeostasis through a complex of PGC-1alpha and SIRT1. Nature434, 113–118.

Sajadi, A., Schneider, B.L., Aebischer, P., 2004. Wlds-mediated protection of dopaminergicfibers in an animal model of Parkinson disease. Curr. Biol. 14 (4), 326–330 (17).

Sundararajan, R., Chen, G., Mukherjee, C., White, E., 2005. Caspase-dependent processingactivates the proapoptotic activity of deleted in breast cancer-1 during tumor necrosisfactor-alpha-mediated death signaling. Oncogene 24 (31), 4908–4920 (21).

Suzuki, K., Koike, T., 2007. Resveratrol abolishes resistance to axonal degeneration in slowWallerian degeneration (WldS) mice: activation of SIRT2, an NAD-dependent tubulindeacetylase. Biochem. Biophys. Res. Commun. 359 (3), 665–671 (3).

Trapp, J., Meier, R., Hongwiset, D., Kassack, M.U., Sippl, W., Jung, M., 2007. Structure-activity studies on suramin analogues as inhibitors of NAD+-dependent histonedeacetylases (sirtuins). ChemMedChem (10), 1419–1431.

Tsao, J.W., George, E.B., Griffin, J.W., 1999. Temperature modulation reveals three distinctstages of Wallerian degeneration. J. Neurosci. 19 (12), 4718–4726 (1999 Jun 15).

Vaquero, A., Scher, M., Lee, D., Erdjument-Bromage, H., Tempst, P., Reinberg, D., 2004.Human SirT1 interacts with histone H1 and promotes formation of facultative hetero-chromatin. Cell 16 (1), 93–105.

Vaziri, H., Dessain, S.K., Ng Eaton, E., Imai, S.I., Frye, R.A., Pandita, T.K., Guarente, L.,Weinberg, R.A., 2001. hSIR2(SIRT1) functions as an NAD- dependent p53 deacetylase.Cell 107, 149–159.

Wang, J., Zhai, Q., Chen, Y., Lin, E., Gu, W., McBurney, M.W., He, Z., 2005. A local mecha-nism mediates NAD-dependent protection of axon degeneration. J. Cell Biol. 170(3), 349–355.

Wood, J.G., Rogina, B., Lavu, S., Howitz, K., Helfand, S.L., Tatar, M., Sinclair, D., 2004. Sirtuinactivators mimic caloric restriction and delay ageing in metazoans. Nature 430(7000), 686–689 (5, Erratum in Nature. 2;431(7004):107).

Yeung, F., Hoberg, J.E., Ramsey, C.S., Keller, M.D., Jones, D.R., Frye, R.A., Mayo, M.W., 2004.Modulation of NF-kappaB-dependent transcription and cell survival by the SIRT1deacetylase. EMBO J. 23 (12), 2369–2380 (16).

Zhang, F., Liu, J., Shi, J.S., 2010. Anti-inflammatory activities of resveratrol in thebrain: role of resveratrol in microglial activation. Eur. J. Pharmacol. 636 (1–3), 1–7(25, Review).

Zhang, F., Wang, S., Gan, L., Vosler, P.S., Gao, Y., Zigmond, M.J., Chen, J., 2011. Protectiveeffects and mechanisms of sirtuins in the nervous system. Prog. Neurobiol. 95 (3),373–395.