Theranostics 2020, Vol. 10, Issue 11 5154 Ivyspring International Publisher Theranostics 2020; 10(11): 5154-5168. doi: 10.7150/thno.43765 Research Paper Novel neuroprotective therapy with NeuroHeal by autophagy induction for damaged neonatal motoneurons David Romeo-Guitart, César Marcos-DeJuana, Sara Marmolejo-Martínez-Artesero, Xavier Navarro, and Caty Casas  Institut de Neurociències (INc) and Department of Cell Biology, Physiology and Immunology, Universitat Autònoma de Barcelona (UAB), & Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Bellaterra, Barcelona, Spain.  Corresponding author: Caty Casas Louzao, Unitat de Fisiologia Mèdica, Facultat de Medicina, Universitat Autònoma de Barcelona, E-08193 Bellaterra, Barcelona, Spain. Tel: +34-935811324, Fax: +34-935812986, E-mail: Caty.Casas@uab.cat © The author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/). See http://ivyspring.com/terms for full terms and conditions. Received: 2020.01.08; Accepted: 2020.03.20; Published: 2020.04.06 Abstract Rationale: Protective mechanisms allow healthy neurons to cope with diverse stresses. Excessive damage as well as aging can lead to defective functioning of these mechanisms. We recently designed NeuroHeal using artificial intelligence with the goal of bolstering endogenous neuroprotective mechanisms. Understanding the key nodes involved in neuroprotection will allow us to identify even more effective strategies for treatment of neurodegenerative diseases. Methods: We used a model of peripheral nerve axotomy in rat pups, that induces retrograde apoptotic death of motoneurons. Nourishing mothers received treatment with vehicle, NeuroHeal or NeuroHeal plus nicotinamide, an inhibitor of sirtuins, and analysis of the pups were performed by immunohistochemistry, electron microscopy, and immunoblotting. In vitro, the post-translational status of proteins of interest was detailed using organotypic spinal cord cultures and genetic modifications in cell lines to unravel the neuroprotective mechanisms involved. Results: We found that the concomitant activation of the NAD+-dependent deacetylase SIRT1 and the PI3K/AKT signaling pathway converge to increase the presence of deacetylated and phosphorylated FOXO3a, a transcription factor, in the nucleus. This favors the activation of autophagy, a pro-survival process, and prevents pro-apoptotic PARP1/2 cleavage. Major conclusion: NeuroHeal is a neuroprotective agent for neonatal motoneurons that fine-tunes autophagy on by converging SIRT1/AKT/FOXO3a axis. NeuroHeal is a combo of repurposed drugs that allow its readiness for prospective pediatric use. Key words: neonatal motoneurons, neuroprotection, autophagy, SIRT1/AKT/FOXO3a, NeuroHeal Introduction Neurons, like other cells, responds to stress by autophagy pathway, the ubiquitin-proteasome engaging both pro-survival subprograms, which we system, chaperone systems, the endoplasmic collectively term endogenous mechanisms of reticulum-associated degradation machinery, and the neuroprotection, and cell death subprograms. antioxidant defense. Excess damage as well as aging Depending on the intensity of injury, the balance result in defective functioning of one or more of these between these subprograms results in either neuronal subprograms. We reasoned that favor their function survival or death [1]. These mechanisms of in pathological conditions may be a successful neuroprotection include the unfolded protein strategy for neuroprotection. With this idea in mind, response (UPR), the heat-shock response, the we discovered NeuroHeal, a new neuroprotective http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5155 agent for adult axotomized spinal motoneurons /AKT axis, which suppresses apoptosis [19] although (MNs), using a top-down strategy based on a it may inhibit autophagy [20]. We aimed to test network-centric approach, the use of artificial whether NeuroHeal treatment may exert intelligence, and proper preclinical animal models [2]. neuroprotection to axotomized neonatal MNs and to For this discovery, we took advantage of the unravel the mechanism involved, which possibly successful engagement of endogenous mechanisms implies the interplay between SIRT1 and PI3K/AKT. by MNs that allow their survival after distal axotomy (e.g., following distal nerve transection and suture of Materials and Methods the sciatic nerve) [3–5], in contrast to the Surgical procedures neurodegenerative process engaged after proximal axotomy (e.g., after nerve root avulsion) [6–8]. All the procedures involving animals were Nerve injuries, such as facial palsy or brachial approved by the Ethics Committee of the Universitat plexus avulsion during obstetric interventions, Autònoma de Barcelona and Generalitat de Catalunya and provoke axonotmesis or neurotmesis depending, for followed the European Community Council Directive instance, whether the injury is preganglionic or post- 2010/63/EU. Pregnant rats were maintained under ganglionic. Motor function recovery after pregangli- standard conditions of light and temperature and fed onic avulsion injuries does not occur spontaneously, ad libitum. Once born, the litter was individualized for and hence surgical reconstruction is needed. each rat and maintained with under the same Functional recovery is incomplete so that substantial conditions. For surgical intervention, we deeply chronic impairment will affect the whole life of the anesthetized 4-day-old Sprague Dawley rat pups by infant [9]. The proper early timing of the surgical hypothermia and dissected out the right sciatic nerve. intervention for functional recovery in children is The nerve was crushed twice with a fine forceps crucial. The critical dependency of the developing (Dumont #5) for 30 s. We closed the wound with motoneurons (MNs) on their interaction with their sutures, and animals were allowed to recover in a target muscle [3,4] makes them quickly succumb after warm environment and returned to the mothers. axotomy. Without survived MNs, surgical Drug treatments reimplantation of nerves is averted and justify the NeuroHeal is a combination of acamprosate and urgent need for pediatric neuroprotective agents. ribavirin [2]. For treatment with NeuroHeal, we Hence, we sought to test whether NeuroHeal is ground acamprosate (Merck) and ribavirin (Normon) capable to rescue target-deprived neonatal MNs in a into fine powders and added to drinking water at model of nerve injury in newborn rats as well as to final concentrations of 2.2 mM and 1 mM, identify key nodes for neonatal neuroprotection. In respectively. NeuroHeal and 5 mM nicotinamide this model, in contrast to what happens to adult MNs, (NAM) (Sigma-Aldrich) were added to drinking distal axotomy performed during the first two weeks water given to the mothers on the day prior to injury of life lead to massive MN death [10,11]. The death of to ensure that the pups began to receive drug on the axotomized neonatal motoneurons is linked to day of the injury. Drug or vehicle (water) treatments excitotoxicity [12] and occurs by apoptosis [13–15] In were maintained by refreshing water every 3 days. contrast, adult ventral horn spinal MNs presented abnormal response to endoplasmic reticulum stress Histology and immunohistochemistry and autophagy flux blockage as key events in their At 10 dpi, 4 animals per group the animals were non-apoptotic neurodegenerative process [6–8]. We euthanized with dolethal (60 mg/kg, intraperitoneal) found that the neuroprotective effect of NeuroHeal and were perfused with a transcardial infusion of a for adult MNs depends on the activation of the saline solution containing heparin (10 U/mL) NAD+-dependent deacetylase sirtuin 1 (SIRT1) [2,16]. followed by 4% paraformaldehyde in 0.1 M PBS. We SIRT1 has emerged as an important player in collected the L4-L5 spinal cord segments, post-fixed neurodegenerative diseases, after stroke, and in for 2 h with 4% paraformaldehyde in 0.1 M PBS and peripheral and cranial nerve injuries [17] and is introduced into a 30% sucrose solution for tightly linked with cellular metabolic status and cryopreservation at 4 ºC. Tissue-tek embedded spinal longevity. SIRT1 has a widely range of targets in both cord samples were serially cut (20-µm sections) in a the cytosol and the nucleus, and modulates different cryotome (Leica) and were kept at -20 ºC until functions such as inflammation, apoptosis, and auto- analysis. For immunolabelling, we washed samples phagy [18]. Thus, we started to focus on this molecule with Tris-buffered saline (TBS), with 0.1 mM glycine as one of the key nodes in the action of NeuroHeal, as in TBS, and with blocking solution (0.3% Triton-X-100 well as other possible pro-survival nodes such as the and 10% donkey serum in TBS) for 1 h at room phosphatidylinositol-4,5-bisphosphate 3-kinase(PI3K) temperature. Samples were then incubated overnight http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5156 at 4 ºC with primary antibodies (see Table S1). After Spinal organotipics cultures several washes with 0.1% Tween-20 in TBS, we added For spinal cord organotypic cultures (SOCs), we donkey-Cy3 or Alexa488 against the primary anti- removed the spinal cords from 7-day-old Sprague body (1:200; Jackson Immunoresearch) and incubated Dawley rats and placed them in cold 30% glucose for 1 h and 30 min at room temperature. Counter- Gey’s Balanced Salt Solution (Sigma-Aldrich). We staining was performed with fluorescent green or blue removed meninges and cut spinal cords into 350-μm NeuroTracer Nissl Stain (Molecular Probes) and/or thick slices, which were placed onto Millicell-CM of DAPI (Sigma) and mounting with Fluoromount-G 30-mm diameter (0.4 μm, Millipore) within wells of mounting medium (SouthernBiotech). Samples from 6-well plates (Thermo Fisher Scientific) containing 1 groups to be compared were processed and analyzed mL of culture medium (50% (v/v) minimal essential together on the same slide. Confocal images were medium (MEM), 2 mM glutamine, 25 (v/v) Hank’s systematically acquired using a confocal Laser Balanced Salt Solution (Sigma-Aldrich) supplemented Scanning Microscope (Zeiss LSM 700) using three with 25.6 mg/mL glucose and 25 mM HEPES, pH 7.2 separate photomultiplier channels with a 1.4 as described previously [21,22]. Cultures were main- numerical aperture objective of 20X under the same tained in a 5% CO2, humidified environment at 37 ºC. conditions of sensitivity, resolution, and exposure The following day, we changed the medium and time for each analyzed marker. Images were added drug or vehicle: Neuro, NeuroHeal+Ex-527, separately projected and merged using a pseudocolor NeuroHeal+NAM, NeuroHeal+3-MA, or NeuroHeal display. Signal intensity was analyzed using ImageJ +LY294, water or DMSO. Medium was changed twice software (National Institutes of Health; available at per week. After 2 weeks of treatment, we removed the http://rsb.info.nih.gov/ij/). The Nissl or DAPI medium, post-fixed the spinal cord slices with cold labelling were used to identify cytosol or nucleus, 4% paraformaldehyde (PFA) solution in 0.1 M PBS, respectively, and the integrated densities within these pH 7.2, for 1 h, washed them with TBS several times, regions of interest were obtained for at least 15 MNs and incubated for 48 h with primary antibodies extracted from three different sections (separated by combined with mouse-anti SMI32 (1:1000; Biolegend) at least 100 μm) per animal for each marker. at 4 ºC. Confocal microphotographs of a predefined In vitro cultures Z-stack were taken covering ventral horn of each SOC, and MN presence was assessed by counting For in vitro studies, acamprosate, ribavirin, SMI32-positive neurons for each SOC hemisection Ex-527, NAM, 3MA, LY-294-002 hydrochloride, and (n=8-12 slices/condition). Staurosporine (St) (Sigma-Aldrich) and Bafilomycin A1 (Santa Cruz Biotechnologies) were diluted into Motor neuron counting water, DMSO or EtOH according to manufacturer’s We stained 20 slides (separated by 100 µm) instructions. Concentrations used in cell culture were covering all the L4-L5 medullar segments of each 55 µM, 1 µM, 10 µM, 5 mM, 10 µM, 10 µM, and 50 nM animal (4 animals per group) with FluoroNissl green respectively. NSC-34 cells were cultured in modified (Life Technologies) for 20 min following the manu- Eagle’s medium high-glucose (DMEM, Biochrom) facturer’s protocol. We took sequential microphoto- supplemented with 10% fetal bovine serum (Sigma- graphs from at least 20 slices of the lateral funiculus Aldrich), 100 U/mL penicillin (Sigma-Aldrich), and from contra- and ipsilateral sides of each animal with 0.5 X penicillin/streptomycin solution (Sigma- the aid of a digital camera (Olympus DP76) attached Aldrich) and maintained with a humidified incubator to a microscope (Olympus BX51) at 20X. Only those at 37 ºC under 5% CO2, essentially as described in a MNs localized in the more lateral neuron pool of the previous report [6]. For nucleotransfection, we used grey matter with a prominent nucleus and soma of at the Amaxa cell line Nucleofector II Kit R (Lonza) and least 900 µm2 were counted as MNs. The percentage of the Nucleofactor V Kit (Lonza) following the surviving MNs was calculated as the number of manufacturer’s recommendations. After 24 h of drug surviving MNs at the ipsilateral side with respect to treatment with St (10, 20, and 50 nM) with and with- the contralateral side for each animal. out NeuroHeal, cells were incubated with 4 mg/mL 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium Western blotting and immunoprecipitation bromide (MTT) solution for 1 h, medium was For immunoblotting, 4 pooled SOC sections for removed, and the resulting formazan salts were each condition after 3 DIV (n=4 different treatments), dissolved in DMSO. Absorbance was measured with a pup spinal cord L4-L6 segments at 3 dpi (n=4), or microplate reader (Bio-tek, Elx800) at 570 nm, and the NSC34 cell extracts obtained at 6 h post-treatments percentage of cell survival was calculated relative to (n=4) were added to lysis buffer (50 mM Tris, pH 6.8, the control on the same plate. 2 mM EDTA, 0.5% Triton-X-100, and a cocktail of http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5157 proteases (Sigma-Aldrich) and phosphatase inhibitors (w/v) PFA in 0.1 M PBS, pH 7.4 were incubated for 2 (Roche)). Samples were homogenized with a Pellet h at room temperature on a rocking platform. Samples pestle (Sigma-Aldrich) on ice, and sonicated with a were fixed in 1% (w/v) PFA and subsequently Ultrasonic homogenizer (Model 3000, Biologics Inc.). post-fixed with 1% (w/v) osmium tetroxide (TAAB Then they were centrifuged for 10 min at 13000 g at 4 Laboratories) containing 0.08% (w/v) potassium ºC, and protein in supernatants were quantified by hexacyanoferrate (Sigma-Aldrich) in PBS, pH 7.4 for 2 BCA assay (Pierce Chemical Co.). Proteins (10 h at 4 ºC. After four washes with deionized water, we µg/well) were resolved by SDS-PAGE and dehydrated samples with sequential washes of transferred to a PVDF membrane in a BioRad cuvette acetone. We embedded samples in EPON resin and system in 25 mM Tris, pH 8.4, 192 mM glycine, and polymerized at 60 ºC for 48 h and then cut into semi- 20% (v/v) methanol. We blocked the membrane with thin sections (1 µm) with a Leica ultracut UCT micro- 5% milk solution in 0.1% Tween-TBS (TBS-T) for 1 h at tome (Leica Microsystems). Sections were stained room temperature and incubated overnight with with 1% (w/v) aqueous toluidine blue solution and primary antibodies (see Table S2). After washing in examined with a light microscope to identify the TBS-T, the membrane was incubated with an ventral horn area. Ultra-thin sections (70 nm) were appropriate secondary antibody conjugated with then cut with a diamond knife, placed on coated grids, horseradish peroxidase (1:5000, Vector) for 1 h. The and contrasted with conventional uranyl acetate and membrane was visualized using a chemiluminescent Reynolds lead citrate solutions. Finally, we observed method (ECL Clarity Kit, Bio-Rad Laboratories), and the sections with a JEM-1400 transmission electron the images were captured and quantified with Image microscope (Jeol) equipped with a Gatan Ultrascan Lab Software (Bio-Rad Laboratories). ES1000CCD camera. We analyzed three to four MNs For immunoprecipitation, we followed the per animal (n=4 per group). manufacturer’s protocol (Life Technologies). Briefly, we linked the anti-acetylated lysine antibodies (1:200; Statistical analysis Sigma) to magnetic particles by incubation on a Data are presented as means ± standard error of rotatory wheel for 10 min at room temperature, the mean (SEM). Statistical analyses were conducted washed, and incubated with 20 µg of protein extract using GraphPad Prism 5 software. We performed an by incubation for 10 min at room temperature on a unpaired t-test to compare two groups and one-way rotatory wheel. After elution, proteins were analysis of variance (ANOVA) to compare three or denatured, resolved by SDS-PAGE, transferred to a more groups followed by Bonferroni’s multiple PVDF membrane, and incubated with primary and comparison test. Differences were assumed to be secondary antibody as described above. significant for p < 0.05. Real-time PCR Results Pups at 3 dpi (n=4) were euthanized, and the SIRT1 activation is necessary for survival of lumbar segment L4-L5 was removed. Total RNA was neonatal axotomized MNs extracted using the RNeasy Plus Mini Kit (QIAGEN) according to manufacturer’s instructions. RNA We first established the neuroprotective effect of concentrations and quality were measured with a NeuroHeal treatment in a model of neonatal axotomy. NanoDrop 2000 (Thermo Fisher Scientific). Reverse It is known that neonatal axotomy of the sciatic nerve transcription was carried out using iSCript cDNA on or before postnatal day 4 triggers retrograde MN Synthesis Kit (BioRad), and the reaction was apoptosis [23–25]. We treated mothers of the pups performed on a SureCycler 8000 Thermal Cycler (Agi- with either vehicle or NeuroHeal through drinking lent). Finally, we performed PCR using cDNA diluted water until sacrifice of pups according to workflow in sterile water as template for primers listed in planning (Figure 1A). Since one of the main Neuro- supplemental Table S3 using the brilliant III Ultra-Fast Heal targets is SIRT1, a group of mothers was treated SYBR Green QPCR Master Mix (Agilent Technolo- with NeuroHeal plus nicotinamide (NAM) as NAM is gies). Samples were plated in duplicate in a 96-well also an inhibitor of SIRT1 [26]. Axotomy resulted in PCR plate and were analyzed using the MyiQ Single death of 38% of L4-L5 ipsilateral anterior MNs, Colour Real-Time PCR Detection System (BioRad) measured relative to the contralateral no- lesioned and the Bio iQ5 Optical System Software (BioRad). side at 10 days post injury (dpi), as expected (Figure 1B). NeuroHeal treatment significantly increased the Transmission electron microscopy number of axotomized MN survival, but NAM co- Perfused spinal cords in a fixative solution of treatment blocked this neuroprotective effect (Figure 2.5% (v/v) glutaraldehyde (EM grade, Merck) and 2% 1B). We confirmed that SIRT1 was activated by Neu- http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5158 roHeal by analysis of SIRT1 substrates such as histone PARP1/2 levels, hardly detected only in the nucleus H3 acetylated at Lys9 (H3K9Ac) and p53 acetylated at of MNs and other surrounding cells, as in the control Lys373 (p53Ac) [27,28]. Using immunohistochemical animals. In contrast, NAM co-treatment abolished the analysis, we observed that axotomy resulted in SIRT1 effect produced by NeuroHeal with MNs saturated of nuclear accumulation and an accumulation of both PARP1/2 in both nuclei and cytosol (Figure S1). This H3K9Ac and p53Ac (Figure 1C). Treatment with result suggested that NeuroHeal might prevent late NeuroHeal normalized the immunofluorescence apoptotic events somehow, and considering that intensity within MNs of these three proteins but in PARP1/2 can be degraded by autophagy [32] we pups co-treated with NAM, the levels were similar to explored the existence of these process further down. those in untreated animals (Figure 1C). To confirm these results in vitro, we used a At 10 dpi, we observed copious staining of previously validated model of neonatal axotomy cleaved caspase 3 (Casp3) in nuclei of damaged MNs, based on spinal organotypic culture (SOC) [21]. In and other cells, possible glial cells, at the ipsilateral spinal cord slices devoid of roots, around 25% of MNs ventral horn, as expected [25]. The presence of cleaved died by 14 days of in vitro culture (DIV) compared to Casp3 in glial cells has been reported by other authors the number of neurons present by 1 DIV (Figure 2D). previously in a model of intracortical neuronal We observed that treatment with NH during 14 days excitotoxicity [29] or more recently due to exercise sustained neuronal population. This neuroprotective [30], claiming not necessarily link to apoptosis in these effect was abolished with the addition of NAM to the cells. Surprisingly, in axotomized MNs, we observed culture suggesting that SIRT1 activation is necessary. persistent cleaved Casp3 pattern in the group treated (Figure 2D; p<0.05). Accordingly, axotomy provoked with NeuroHeal (Figure 2A). Accordingly, we accumulation of the 89-kDa PARP1/2 fragment, and observed similarly by immunoblotting, with no this was attenuated by NeuroHeal treatment (Figure differences in the ratio between cleaved Casp3 versus 2E; p<0.05). Hence, the in vitro model of axotomy total Casp3 among groups, including the group reproduced the phenomena observed in vivo. treated with NeuroHeal plus NAM despite higher levels of each of these forms (Figure 2B). We did not NeuroHeal-induced neuroprotection sustain explore other casp3-independent apoptotic both SIRT1 and AKT activation and mechanisms since it was well established the downstream FOXO3a modifications predominant role of this caspase in the apoptotic In addition to SIRT1, we previously observed neuronal demise by axotomy in pups [15]. Hence, we that NeuroHeal may modulate the PI3K/AKT hypothesized that NeuroHeal may modulate factors pathway in some conditions in adult MNs [2]. We downstream of Casp3, such as PARP1/2, to exert the wanted to analyze whether this pathway is involved neuroprotective effect [31]. Cleavage of PARP1/2 by in NH-induced neuroprotection in axotomized caspases typically yields 89-kDa fragments that are neonatal motoneurons as well. We found that considered hallmarks of apoptosis. In injured spinal neonatal axotomy provoked an increase of pAKT that cord L4-L5 extracts there was an increase in the was sustained by NeuroHeal treatment alone or in amount of 89-kDa PARP1/2 fragment compared to combination with NAM (Figure 3A; p<0.05). This levels in extracts of uninjured samples as expected result suggested that any effect on the PI3K/AKT (Figure 2C). NeuroHeal treatment reduced levels of pathway was independent of the SIRT1 activation full-length PARP1/2 (115-kDa band) and also the induced by NeuroHeal. To confirm the activation of abundance of the 89-kDa form (Figure 2C). In the PI3K/AKT axis, we analyzed further down. A agreement with this, immunohistochemical analysis main target of Akt is tuberous sclerosis 2 (TSC2) of PARP1/2 localization revealed higher levels within which forms heterodimeric complex with TSC1 and the nuclei and the cytosol of axotomized MNs inhibits mTORC1, part of the mTOR complex. compared to uninjured control where slighter levels Phosphorylation of TSC2 by Akt blocks its activity were observed in the nucleus (Figure S1). Other nuclei and therefore allow activation of mTORC1which were also stained in surrounding cells, possibly glial directly phosphorylate P70S6k (pP70S6k) [33]. By cells. The raised of PARP1/2 after damage is expected immunohistochemical analysis we observed that the as well as the nuclear and cytosolic presence. We have intensity of pP70S6k within MNs raised notably after used a pan antibody, the same as for immunoblotting, injury compared to control (Figure 3B; p<0.01). The that did not distinguish both forms by immuno- average intensity in axotomized MNs when animals histochemistry. However, full-length PARP1/2 is were treated with NeuroHeal diminished respect to predominantly nuclear, but the 89kDa-fragment is vehicle treated group (Figure 3B; p<0.01), what was liberated from the nucleus into the cytosol [31]. abolished by co-treatment with NAM (Figure 3B; NeuroHeal treatment drastically reduced overall p<0.05). Altogether, these observations suggested that http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5159 axotomy caused activation of PI3K/AKT/mTOR axis. NeuroHeal treatment, downstream SIRT1 activation Although PI3K/AKT activation is sustained by by this drug attenuates mTOR activation. Figure 1. SIRT1 activation mediates neuroprotection of neonatal axotomized motoneurons by NeuroHeal treatment. (A) Schematic experimental workflow: sciatic nerve crush was performed on pups of 4 days old which were returned to mothers thereafter. Mothers were orally treated with either vehicle (Veh), NeuroHeal (NH), or NH + nicotinamide (NAM) for up to 3 or 10 days post injury (dpi) when pups were collected to perform transmission electron microscopy (TEM), Real-Time PCR (RT-PCR), western blot (WB), or immunohistochemical analyses, respectively. (B) Left, Representative microphotographs of FluoroNissl-stained (green) spinal cord sections from non-injured control (Ctrl) or injured vehicle-treated (Veh) animals at 10 dpi. Scale bar 100 μm at all images. Right, Graph of the percentage of MNs found at the ipsilateral (crushed) with respect to the contralateral anterior spinal cord sides, from vehicle-treated (Veh), NH and NH+NAM groups (n=4 per group, ANOVA, post hoc Bonferroni *p<0.05). (C) Top, Confocal images of MNs immunolabelled to reveal SIRT1 (red), and its downstream substrates: H3K9Ac (red), and p53Ac (red), and counterstained with FluoroNissl (green) and DAPI (blue), from the different animal groups at 10 dpi. At each condition, panels a-c are zoom images: a and c panels are merged microphotographs, and b panels are the amplified imaged of the corresponding single labeling in red. Scale bar 50 μm and identical for all corresponding microphotographs as represented in the first image panel, the Ctrl/SIRT1 condition. Bottom, Graph of the means of the immunofluorescence density (ID) for each marker inside the nuclei of MNs (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05). http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5160 Figure 2. NeuroHeal mediates neuroprotection from apoptosis at PARP1/2 level. (A) Left, Representative confocal images of neonatal axotomized MNs immunolabelled to reveal active cleaved caspase 3 (Casp3) (red), FluoroNissl (green), and DAPI (blue), from injured animals treated with vehicle (Veh) or NeuroHeal (NH) at 10 dpi. Scale bar 100 µm high magnification, and 20 µm low magnification. (B) Immunoblotting and corresponding graphs of the analysis of the ratio of cleaved Casp3 vs the total form in the different groups. (C) Representative western blot membrane and bar graph showing the protein fold change analysis of PARP1/2 levels in uninjured (Ctrl) and injured animals treated with and without NH at 3 dpi (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05). (D) Top, Representative microphotographs of SMI-32-labelled MNs at the whole spinal cord slice (in 1day-in-vitro culture (DIV)) or at their ventral horns at 14 DIV from vehicle-, or NH-treated spinal cord organotypic cultures (SOCs). Scale bar 100 µm. Bottom, Bar graphs showing the mean number (± SEM) of SMI-32-positive neurons in the ventral horn of each hemisection of the spinal cord slice (n=8-12 complete spinal cord slices, ANOVA, post hoc Bonferroni *p<0.05). (E) Western blot and associated bar graph showing the analysis of PARP1/2 protein levels from different treatment conditions of 14DIV SOCs (untreated Ctrl, or treated with vehicle or NH) (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05). http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5161 Figure 3. NeuroHeal modulates FOXO3a activity. (A) Immunoblots and bar graphs showing the analysis of pAKT related to total AKT levels, and pFOXO3a related to its total form in the different experimental groups (uninjured (Ctrl), vehicle (Veh), NeuroHeal (NH) or NH plus nicotinamide (NAM) treated animals) at 3 dpi. (n=4 animals, ANOVA, post hoc Bonferroni, *p<0.05). (B) Confocal microphotographs and graphs showing the immunofluorescent density of pP70S6k (red) co-labelled with FluoroNissl (green) from different experimental groups. Scale bar 100 µm. (n=4 animals, ANOVA, post hoc Bonferroni, *p<0.05, **p<0.01) (C) Left, Microphotographs showing the localization and levels of total FOXO3a (red), co-labelled with FluoroNissl (green) in different experimental groups at 10 dpi. Scale bar 100 µm (FOXO3a) and 10 µm (in zoomed merged image). Right, Bar graph showing the percentage of nuclear respect to whole cell immunofluorescence of FOXO3a within MNs at the ventral horns from the different experimental conditions (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05, **p<0.01). (D) Western blots and graphs showing the analysis of pFOXO3a respect to total FOXO3a in SOCs treated as indicated (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05). (E) Western blots and graphs showing the analysis of immunoprecipitated acetylated pFOXO3a (Ac-pFOXO3a) levels in different experimental groups at 3 dpi from injured pups (left) or SOCs (right) (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05, **p<0.01). (F) Bar graph of mean values (mean ± SEM) of transcript fold change obtained by quantitative real-time PCR for Bax, Bnip3, Gadd45, LC3, and Beclin-1 genes in different experimental groups at 3 dpi (n=4, ANOVA, post hoc Bonferroni *p<0.05). http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5162 Considering a possible key point of convergence alter the NeuroHeal treatment-induced up-regulation between AKT and SIRT1 in the action of NeuroHeal, of pro-autophagy genes LC3 and Beclin1. we next analyzed FOXO3a, a member of the O Altogether, these results suggested that the subclass of the forkhead family of transcription concurrent activation of SIRT1 and AKT pathways factors, which is both an AKT substrate [34] and it can when treatment with NeuroHeal induces nuclear be deacetylated by SIRT1 [35]. It has been reported localization of acetylated pFOXO3a, which selectively that FOXO3a may drive cell survival or death [36,37]. reduces the expression of some FOXO3a-downstream In the neonatal in vivo model, injury caused an pro-apoptotic genes. increase in the isoform of FOXO3a phosphorylated at Ser 253 relative to total FOXO3a (pFOXO3a/FOXO3a Concomitant deacetylation and ratio), which was sustained by NeuroHeal treatment. phosphorylation of FOXO3a is key to The co-administration of NAM tended to normalize neuroprotection the relative levels of pFOXO3a although there were no In order to establish the role of deacetylated differences with the treatment of NeuroHeal alone pFOXO3a in the neuroprotective effect mediated by (Figure 3A; p<0.05). Despite immunoblotting analysis NeuroHeal against apoptosis, we used an in vitro did not reveal quantitative differences, qualitative model based on the use of staurosporine (St) as a differences were observed by immunohistochemistry pro-apoptotic agent [42]. We confirmed that St since the levels of FOXO3a within axotomized MN decreased survival of NSC-34 MN-like cells in a nuclei were drastically reduced by NeuroHeal dose-responsive manner and we found that treatment and they were restored by co-treatment NeuroHeal was neuroprotective at the highest with NAM (Figure 3C; p<0.01). concentration of St (Figure 4A; p<0.01). In contrast to what observed in pups, in SOCs, Immunoblotting analysis showed an increase we observed an increase of the pFOXO3a/FOXO3a pFOXO3a/FOXO3a ratio by NeuroHeal co-treatment ratio upon NeuroHeal treatment respect to control respect to St treatment alone (Figure 4B; p<0.05). and vehicle, that was abolished when cultures were To further establish the importance of pFOXO3a co-treated with either NAM or EX-527, an specific in survival of apoptotic neurons, we transfected NSC- SIRT1 inhibitor (Figure 3D; p<0.05). 34 cells with: a vector for overexpression of either the Regarding the differences in localization wild-type (WT) FOXO3a, a dominant negative mutant observed in the pups and considering that not only (FOXO3a-DN), or a triple mutant form (FOXO3a-TM) phosphorylation but also acetylation influences the (Figure 4C). FOXO3a-DN, constructed by deletion of activity and nuclear localization of FOXO3a [36], we the transactivation domain from the C-terminus, investigated the extent of acetylation by competes with wild-type FOXO3a for binding to immunoblotting with anti-pFOXO3a after performing DNA [43]. The FOXO3a-TM yield a immunoprecipitation against anti-acetyl-lysine non-phosphorylatable (with alanine residues at proteins. The level of acetylated pFOXO3a positions Thr32, Ser253 and Ser315) and, hence, (AC-pFOXO3a) was significantly higher after injury constitutively inactive form of FOXO3a (TM-FOXO3a) with respect to control, but NeuroHeal treatment [43]. We verified the expected effect of mutants normalized its levels in vivo (Figure 3E; p<0.05). analyzing FOXO3a protein size and subcellular Similarly, in SOCs NeuroHeal treatment drastically location by immunoblot. As shown in Figure S2A, reduces the levels of AC-pFOXO3a (Figure 3E; FOXO3a was detected in both cytoplasmic and p<0.01). nuclear fractions in transfected cells with either WT or Since the presence or absence of FOXO3a within GFP plasmids at the expected size (around 80 KDa). In the nucleus may alter downstream gene expression, FOXO3a-DN cells, which yield a deleted product, the we subsequently analyzed the expression of several band was smaller size as expected in both fractions genes known to be regulated by this transcription [43]. We also verified reduced levels of BIM protein, factor such as Bax [38], Bnip3 [39], Gadd45 [40], Lc3, depending on FOXO3a activity, in DN and drastically, and Beclin1 [41] in the in vivo model. Neonatal in TM mutants. Transfection of the cells with these axotomy increased levels of almost all of them except vectors did not affect survival when compared to GFP Beclin1 for which the trend did not achieve statistical group (Figure S2B). Overexpression of wild-type significance (Figure 3F; p<0.05). NeuroHeal treatment FOXO3a promoted survival itself after St significantly reduced levels of Bax and Bnip3, genes administration similar to NeuroHeal treatment, related to apoptosis, and this reduction was blocked whereas survival was drastically reduced in cells by NAM co-treatment. Sustained expression of overexpressing FOXO3a-DN or FOXO3a-TM in all Gadd45 by NeuroHeal treatment was also blocked by treatment conditions (Figure 4C; p<0.001). These NAM co-administration. In contrast, NAM did not results suggested that the levels of phosphorylated http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5163 form of FOXO3a and its activity are important for to influence the autophagy/apoptosis balance [44]. neuroprotection mediated by NeuroHeal in the Since FOXO3a favored the expression of autophagy- staurosporine-induced apoptosis paradigm. related genes, we analyzed for activation of autophagy after neonatal injury. In cells undergoing autophagy, phagophore formation initiates after the Unc-51 like autophagy activating kinase 1 (ULK1) activation, and its elongation is regulated by two ubiquitin-like reactions: the first leading to the formation of the complex ATG12-ATG5-ATG16L1; and the second, involves the conjugation of the microtubule-associated protein light chain 3 (MAP- LC3/ATG8/LC3) to phosphatidylethanolamine at the autophagosome membrane to form autophagosome- associated LC3-II. Once, the autophagosome is formed, it acquires the ability to bind autophagic substrates and/or proteins that mediate cargo selectivity (including sequestosome 1 (p62/ SQSTM1)). Autophagosomes mobilize toward lysosomes along microtubules; then, the outer membrane of the autophagosome fuses with the lysosome, thanks to lysosomal proteins like LAMP1, to form an autolysosome [45]. We evaluated autophagy markers Ulk1 phosphorylated at Ser 555, conjugated ATG5-ATG12, and LC3II, as previously described by immunoblotting spinal cord extracts at 3 dpi and perform quantitative western analysis. Neonatal axotomy did not affect the levels of these proteins, whereas there were a significant increase in all of them in the NeuroHeal treated group compared to vehicle. The addition of NAM abolished all these effects observed in the NeuroHeal group (Figure 5A; p<0.01). Accordingly, by immunohistochemical analysis we observed abundant dot-like structures which were positive for LAMP1 within the cytosol of axotomized MNs preferently in the NeuroHeal group and the overall immunofluorescence intensity was higher in this group than in the vehicle and control ones (Figure S3 & 5B). Although the increase in LAMP1 might be related also to lysosomal biogenesis, we observed similar increase and dot-like pattern for ATG5 staining in NeuroHeal-treated animals (Figure 5B). Correspondingly, by electron microscopy we Figure 4. FOXO3a mediates the neuroprotective effect of NeuroHeal. (A) confirmed the presence of autophagosomes and Bar graph showing the percentage of NSC-34 cell survival (mean ± SEM) after autolysosomes in MNs only from injured animals treatment with 10, 20, or 50 nM staurosporin (St) and with or without NeuroHeal (NH) (n=6-10, ANOVA, post hoc Bonferroni *p<0.05). (B) Western blots and bar treated with NeuroHeal (Figure 5C; p<0.005). graph showing the analysis of pFOXO3a respect to total FOXO3a in NSC-34 cell In SOCs, we observed similar effects of Neuro- cultures (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.0). (C) Bar graph showing the percentage of NSC-34 cell survival (mean ± SEM) relative to Heal treatment on LC3II, pUlk1, and ATG5-ATG12 vehicle-treated cells after treatment with 50 nM St with or without NH (n=4 per levels (Figure 6A). The increases due to NeuroHeal group, ANOVA, post hoc Bonferroni *p<0.05, **p<0.01, ***p<0.005, ****p<0.001 ). treatment were abolished when using the inhibitors of phosphatidylinositol 3 kinases such as 3-methyl- adenine (3MA) or LY294, or with SIRT1 inhibitors SIRT1 and AKT activations are necessary for EX-527 or NAM (Figure 6A). Lastly, we observed a neuroprotective autophagy significant reduction in the levels of p62/SQSTM1 The apoptosis/autophagy balance may be (selectively degraded by autophagy) suggesting determinant of cell fate and several factors are known absence of autophagic flux blockage. Both SIRT1 http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5164 inhibitors increased the acetylated forms of p53 and sown in Figure S5, only cells treated with NeuroHeal H3 as expected (Figure S4; p<0.05). Moreover, the had higher cell viability than those treated with St, NeuroHeal-induced neuroprotection was abolished what was abolished when Baf was added too. These when cultures were co-treated with any of these observations corroborate the importance of auto- inhibitors (Figure 6B; p<0.05). We also verified the phagy induction by NeuroHeal for neuroprotection. importance of autophagy-mediated induction by These results indicate that both SIRT1 and PI3K/ Neuroheal in neuroprotection against apoptosis by AKT activation are required by NeuroHeal-mediated treating NSC34 cells with bafilomycin (Baf). Baf is neuroprotection and the induced autophagy- potent (V)-ATPase inhibitor that interrupt the dependent process. autophagosome-lysosome fusion step autophagy. As Figure 5. NeuroHeal activates SIRT1-dependent autophagy. (A) Western blots and bar graphs showing the analysis of pUlk1, ATG5-ATG12, and LC3II levels in different experimental groups at 3 dpi (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05, **p<0.01). (B) Left, Bar graphs of the mean immunofluorescence intensity for each marker within the MNs (n=4 animals per group, t-test, *p<0.05, **p<0.01). Right, Representative confocal microphotographs of samples ATG5 (red), and counterstained with FluoroNissl (Blue) from contralateral side (Ctrl) and injured side of vehicle-treated (Veh) and NH-treated animals. Scale bar 25 µm. (C) Left, Representative transmission electronic images within MNs form the different animal groups at 3 dpi. Note that only autophagosomes* or autolysosomes**, are detected in NH-treated MNs. Scale bar: a= 1 µM, b & c= 500 µM, d= 1000 µM. Right, Bar graph of the number of autophagosomes and autolysosomes found within MNs at each group per area (t-test, ***p<0.005). http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5165 Figure 6. Neuroprotective autophagy triggered by NeuroHeal depends on co-activation of AKT/SIRT1 axes. (A) Western blots and bar graph showing the analysis of pUlk1, ATG5-ATG12, and LC3II protein levels from differentiated SOCs after 2 days of treatment with vehicle or NeuroHeal (NH) alone or in combination with SIRT1 or PI3K inhibitors (3MA and LY294) (n=4 per group, ANOVA, post hoc Bonferroni, *p<0.05). (B) Representative microphotographs of SMI32 immunolabeled differentiated motoneurons at the ventral horn of the SOCs treated for 2 days at the different conditions. Scale bar 100 µm. Bar graphs showing the numbers (± SEM) of SMI-32-positive cells (green) in the ventral horns of each hemi-section of the spinal cord slice at the different conditions (n=8-12 complete spinal cord slices, ANOVA, post hoc Bonferroni *p<0.05). Scale bar 100 µm. Discussion action involved. We discovered that NeuroHeal induces fine-tuned autophagy and concomitant Obstetric complications like brachial plexus activation of SIRT1 and AKT/PI3K axes converging avulsion during birth lead to axotomy of MNs and onto FOXO3a, as a necessary node for neuro- consequent mobility problems for life. Nerve surgical protection. repair may palliate the consequences only if We previously found that NeuroHeal is a performed early enough to avoid the rapid decline of neuroprotective agent for disconnected MNs in disconnected MNs. The immature neuronal system different models of nerve injury in adult rodents does not behave identically at the molecular level as [2,46]. Adult spinal MNs that suffered from root the mature system. Because of that, neuroprotective avulsion or proximal axotomy suffer a slow neuro- agents well performing in the adult should prove to degenerative process marked by a non-canonical UPR be effective also during development if advisable for response and defective autophagy flux [6,8]. In this pediatric use. We aimed to test whether NeuroHeal context, we observed that NeuroHeal activates SIRT1, treatment may exert neuroprotection to axotomized and increased integrins and cytoskeletal and neonatal MNs that succumb by apoptosis, mainly motor-related proteins such as dynactin and Kinesin dependent on Casp 3, and elucidate the mechanism of http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5166 5c [2].Herein, we observed that NeuroHeal exerted necessary for the neuroprotection established by also neuroprotection to apoptotic MNs. The NeuroHeal against apoptosis. mechanism involved may interfere with the final Finally, a question that remains is how substrates of Casp3 since persistent activation of this autophagy induction may exert neuroprotection in caspase was observed event after NeuroHeal treat- this model. Whether autophagy is a cell death or a ment despite neuroprotection. So, in which manner pro-survival mechanism is still controversial. Some can NeuroHeal arrest apoptosis or drive the cell authors working in the field of hypoxia-ischemia (HI) towards other pro-survival signals? in neonates have observed that autophagy is strongly We observed that Neuroheal fine-tunes auto- induced after damage, and consequently, blocking it phagy induction which was necessary to exert its yield neuroprotection [58,59]. Although some claimed neuroprotection and it is conducted by opposite that autophagy pathways are disturbed in regional- forces. NeuroHeal activated SIRT1 and SIRT1- and sex-specific patterns in the rat brain following dependent autophagy [47–49]. On the other hand, neonatal HI reinforcing the idea that many factors NeuroHeal sustained AKT activation after injury influence [60]. Indeed, it appears that autophagy which normally inhibits autophagy primarily by apparition and modulation by any agent would activating mTOR [20]. Surprisingly, NeuroHeal depend on the brain cell type and/or nutrient treatment attenuated mTOR activation in axotomized availability, i.e., the activation status of autophagy neonatal MNs what would indeed favor the (basal vs. induced). Besides, neonatal HI triggers both autophagy induction that we observed. How might caspase-dependent and independent cell-death [61] this be? Several reports have described that SIRT1 and hence may be different than what accounts to activation may interfere with PARP1 and mTOR axotomized neonatal MNs. Peripheral axotomy in activities that altogether seem to form a tripartite key neonates induces excitotoxicity and oxidative stress node to orchestrate cell destiny [50]. SIRT1 leading to DNA damage, mitochondria function fail deacetylates PARP1 blocking its catalytic activity and energy depletion on damaged MNs. Although it [51,52]. SIRT1 blocks mTOR activity indirectly, is speculative, the induction of mild autophagy as probably by deacetylation and activation of TSC2, an with NeuroHeal may proportionate energy from inhibitor of mTOR [53]. It may happen that activation enhancing the clearance of dysfunctional proteins and of SIRT1 by NeuroHeal blocks both PARP1/2 and organelles like mitochondria or diminish mTOR, and consequently facilitating cell survival, excitotoxicity-induced DNA damage, as it has been since overactivation of PARP1/2 can lead to cell death described in several models of neurodegenerative [54], and pro-survival autophagy [33], respectively. diseases [62]. Besides, PARP1/2 can be degraded by autophagy A recent study, it has been observed that [32]. However, further experiments should be excitotoxicity blocks autophagy triggered in cells performed to confirm all these details that may be previously to the insult by starvation. These authors involved. Nevertheless, we have validate previous in concluded that appears that autophagy modulation silico prediction about NeuroHeal claiming PARP1/2 by gluta- mate might be more complex then as one key node possible target mechanism of action previously anticipated, including both stimulatory [2]. and inhibitory effects depending on the brain cell type Both factors, SIRT1 and AKT, interact with and/or nutrient availability, i.e. the activation status FOXO3a, a transcription factor whose regulation of autophagy (basal vs. induced). We can add to this depends on multiple post-translational modifications the condition of immaturity of neonatal neurons were that determine both the subcellular localization and the channels receptors and signaling machinery are activity [55]. We established herein that concomitant not exactly as the mature neuron. phosphorylation and deacetylation of FOXO3a in axotomized neurons drive a “FOXO code” that Conclusions downregulates apoptotic genes while preserving the We conclude that NeuroHeal is a neuropro- expression of autophagy-related ones. Our results are tective agent that sustains the activation of both SIRT1 in agreement with a study in hematopoietic stem cells and PI3K/AKT converging to promote a specific that indicated that FOXO3a regulates a genetic coded FOXO3a function to avoid neuronal demise. In program that results in activation of autophagy under light to overall results, we propose NeuroHeal as a stress conditions to allow survival [56]. Our data also neuroprotective agent for pediatric use. agree with reports that FOXO3a deacetylation by SIRT1 leads to the expression of particular genes, like Abbreviations Gadd45 [57], but not others, such as Bim [37]. We also AC-pFOXO3a: acetylated pFOXO3a; DIV: days established that the phospho-form of FOXO was of in vitro culture; dpi: days post injury; FOXO3a-DN: http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5167 dominant negative mutant; FOXO3a-TM: triple 8. Penas C, Font-Nieves M, Forés J, et al. Autophagy, and BiP level decrease are early key events in retrograde degeneration of motoneurons. Cell Death mutant form; H3K9Ac: histone H3 acetylated at Lys9; Differ. 2011; 18: 1617–27. 9. Pondaag W, A Malessy MJ, Gert van Dijk J, W M Thomeer RT. Natural history MN: motoneuron; MTT: 3-(4,5-dimethylthiazol-2- of obstetric brachial plexus palsy: a systematic review. Dev Med Child Neurol yl)-2, 5-diphenyltetrazolium bromide; NAM: nicoti- 2004; 46: 138–44. 10. Soreide AJ. Variations in the Axon Reaction in Animals of Different Ages. namide; PI3K: phosphatidylinositol-4,5-bisphosphate Cells Tissues Organs. 1981; 110: 40–7. 3-kinase; p53Ac: p53 acetylated at Lys373; SIRT1: 11. Lowrie MB, Vrbová G. Dependence of postnatal motoneurones on their targets: review and hypothesis. Trends Neurosci. 1992; 15: 80–4. sirtuin 1; SOC: spinal organotypic culture; St: 12. Mentis GZ, Greensmith L, Vrbová G. Motoneurons destined to die are rescued staurosporine; TBS: Tris-buffered saline; TSC2: by blocking n-methyl-d-aspartate receptors by MK-801. Neuroscience. 1993; 54: 283–5. tuberous sclerosis 2; UPR: unfolded protein response. 13. De Bilbao F, Dubois-Dauphin M. Time course of axotomy-induced apoptotic cell death in facial motoneurons of neonatal wild type and bcl-2 transgenic Supplementary Material 14. mice. Neuroscience. 1996; 71: 1111–9. Rossiter JP, Riopelle RJ, Bisby MA. Axotomy-Induced Apoptotic Cell Death of Neonatal Rat Facial Motoneurons: Time Course Analysis and Relation to Supplementary figures and tables. NADPH-Diaphorase Activity. Exp Neurol. 1996; 138: 33–44. http://www.thno.org/v10p5154s1.pdf 15. Vanderluit JL, McPhail LT, Fernandes KJL, et al. Caspase-3 is activated following axotomy of neonatal facial motoneurons and caspase-3 gene deletion delays axotomy-induced cell death in rodents. Eur J Neurosci. 2000; Acknowledgements 12: 3469–80. 16. Romeo-Guitart D, Forés J, Navarro X, Casas C. Boosted Regeneration and We thank Ariadna Arbat from our laboratory for Reduced Denervated Muscle Atrophy by NeuroHeal in a Pre-clinical Model of help with the neonatal axotomy model, Eeva-Liisa Lumbar Root Avulsion with Delayed Reimplantation. Sci Rep. 2017; 7: 12028. 17. Langley B, Sauve A. Sirtuin Deacetylases as Therapeutic Targets in the Eskelinen from the University of Helsinki for help in Nervous System. Neurotherapeutics. 2013; 10: 605–20. the interpretation of ME results, and Anne Brunet 18. Donmez G, Outeiro TF. SIRT1 and SIRT2: emerging targets in neurodegeneration. EMBO Mol Med. 2013; 344–52. from Stanford University for FOXO mutant 19. Simon DJ, Pitts J, Hertz NT, et al. Axon Degeneration Gated by Retrograde Activation of Somatic Pro-apoptotic Signaling. Cell. 2016; 164: 1031–45. expression vectors. This work was mainly supported 20. Noguchi M, Hirata N, Suizu F. The links between AKT and two intracellular by the Ministerio de Economía y Competitividad of proteolytic cascades: Ubiquitination and autophagy. Biochim Biophys Acta - Rev Cancer. 2014; 1846: 342–52. Spain (#SAF 2014-59701). We are also grateful for 21. Guzmán-Lenis M-S, Navarro X, Casas C. Drug screening of neuroprotective support from CIBERNED and TERCEL. agents on an organotypic-based model of spinal cord excitotoxic damage. Restor Neurol Neurosci. 2009; 27: 335–49. 22. Herrando-Grabulosa M, Casas C, Aguilera J. The C-terminal domain of Author Contributions tetanus toxin protects motoneurons against acute excitotoxic damage on spinal cord organotypic cultures. J Neurochem. 2013; 124: 36–44. DRG performed the experiments, analyzed the 23. Deckwerth TL, Elliott JL, Knudson CM, Johnson EM, Snider WD, Korsmeyer results, and wrote part of the paper. CMJ performed SJ. BAX is required for neuronal death after trophic factor deprivation and during development. Neuron. 1996; 17: 401–11. in vitro experiments with mutants. SMAM performed 24. Chan YM, Wu W, Yip HK, So KF, Oppenheim RW. Caspase inhibitors Casp3 WB analysis and in vitro experiments with Baf. promote the survival of avulsed spinal motoneurons in neonatal rats. Neuroreport. 2001; 12: 541–5. XN conceived experiment with this model, helped 25. Sun W, Oppenheim RW. Response of motoneurons to neonatal sciatic nerve with the surgery and contributed to writing of the axotomy in Bax-knockout mice. Mol Cell Neurosci. 2003; 24: 875–86. 26. Bitterman KJ, Anderson RM, Cohen HY, Latorre-Esteves M, Sinclair DA. paper. CC conceived most of the experimental design, Inhibition of silencing and accelerated aging by nicotinamide, a putative supervised the experiments and analysis, and wrote negative regulator of yeast sir2 and human SIRT1. J Biol Chem. 2002; 277: 45099–107. the paper. 27. Imai S, Armstrong CM, Kaeberlein M, Guarente L. Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature. Competing Interests 28. 2000; 403: 795–800. Luo J, Nikolaev AY, Imai S, et al. Negative control of p53 by Sir2alpha promotes cell survival under stress. Cell. 2001; 107: 137–48. The authors have declared that no competing 29. Acarin L, Villapol S, Faiz M, Rohn TT, Castellano B, González B. Caspase-3 interest exists. NeuroHeal is currently under patent activation in astrocytes following postnatal excitotoxic damage correlates with cytoskeletal remodeling but not with cell death or proliferation. Glia. 2007; 55: protection. 954–65. 30. Stevenson ME, Lensmire NA, Swain RA. Astrocytes and radial glia-like cells, References but not neurons, display a nonapoptotic increase in caspase-3 expression following exercise. Brain Behav. 2018; 8: e01110. 1. Fulda S, Gorman AM, Hori O, Samali A. Cellular Stress Responses: Cell 31. Chaitanya G V, Steven AJ, Babu PP. PARP-1 cleavage fragments: signatures of Survival and Cell Death. Int J Cell Biol. 2010; 2010: 1–23. cell-death proteases in neurodegeneration. Cell Commun Signal. 2010; 8: 31. 2. Romeo-Guitart D, Forés J, Herrando-Grabulosa M, et al. Neuroprotective 32. Wang Y, Kuramitsu Y, Tokuda K, et al. Gemcitabine induces poly Drug for Nerve Trauma Revealed Using Artificial Intelligence. Sci Rep. 2018; (ADP-ribose) polymerase-1 (PARP-1) degradation through autophagy in 8: 1879. pancreatic cancer. PLoS One. 2014; 9: e109076. 3. Vanden Noven S, Wallace N, Muccio D, Turtz A, Pinter MJ. Adult spinal 33. Weichhart T. mTOR as Regulator of Lifespan, Aging, and Cellular Senescence: motoneurons remain viable despite prolonged absence of functional synaptic A Mini-Review. Gerontology. 2018; 64: 127–34. contact with muscle. Vol. 123, Experimental Neurology. 1993; 123: 147-156 34. Brunet A, Bonni A, Zigmond MJ, et al. Akt Promotes Cell Survival by 4. Lowrie MB, Lavalette D, Davies CE. Time Course of Motoneurone Death after Phosphorylating and Inhibiting a Forkhead Transcription Factor. Cell. 1999; Neonatal Sciatic Nerve Crush in the Rat. Dev Neurosci. 1994; 16: 279–84. 96: 857–68. 5. Valero-Cabré A, Tsironis K, Skouras E, Navarro X, Neiss WF. Peripheral and 35. Motta MC, Divecha N, Lemieux M, et al. Mammalian SIRT1 Represses spinal motor reorganization after nerve injury and repair. J Neurotrauma. Forkhead Transcription Factors. Cell. 2004; 116: 551–63. 2004; 21: 95–108. 36. van der Horst A, Burgering BMT. Stressing the role of FoxO proteins in 6. Leiva-Rodríguez T, Romeo-Guitart D, Marmolejo-Martínez-Artesero S, et al. lifespan and disease. Nat Rev Mol Cell Biol. 2007; 8: 440–50. ATG5 overexpression is neuroprotective and attenuates cytoskeletal and 37. Brunet A. Stress-Dependent Regulation of FOXO Transcription Factors by the vesicle-Trafficking alterations in axotomized motoneurons article. Cell Death SIRT1 Deacetylase. Science. 2004; 303: 2011–5. Dis. 2018; 9: 626. 38. Das TP, Suman S, Alatassi H, Ankem MK, Damodaran C. Inhibition of AKT 7. Casas C, Isus L, Herrando-Grabulosa M, et al. Network-based proteomic promotes FOXO3a-dependent apoptosis in prostate cancer. Cell Death Dis. approaches reveal the neurodegenerative, neuroprotective and pain-related 2016; 7: e2111–e2111. mechanisms involved after retrograde axonal damage. Sci Rep. 2015; 5: 9185. http://www.thno.org Theranostics 2020, Vol. 10, Issue 11 5168 39. Dijkers PF, Medema† RH, Lammers J-WJ, Koenderman L, Coffer PJ. Expression of the pro-apoptotic Bcl-2 family member Bim is regulated by the forkhead transcription factor FKHR-L1. Curr Biol. 2000; 10: 1201–4. 40. Tran H. DNA Repair Pathway Stimulated by the Forkhead Transcription Factor FOXO3a Through the Gadd45 Protein. Science. 2002; 296: 530–4. 41. Sengupta A, Molkentin JD, Yutzey KE. FoxO transcription factors promote autophagy in cardiomyocytes. J Biol Chem. 2009; 284: 28319–31. 42. Parker GC, Li X, Anguelov RA, Toth G, Cristescu A, Acsadi G. Survival motor Neuron protein regulates apoptosis in anin vitro model of Spinal muscular atrophy. Neurotox Res. 2008; 13: 39–48. 43. Skurk C, Maatz H, Kim H-S, et al. The Akt-regulated Forkhead Transcription Factor FOXO3a Controls Endothelial Cell Viability through Modulation of the Caspase-8 Inhibitor FLIP. J Biol Chem. 2004; 279: 1513–25. 44. Maiuri MC, Zalckvar E, Kimchi A, Kroemer G. Self-eating and self-killing: crosstalk between autophagy and apoptosis. Nat Rev Mol Cell Biol. 2007; 8: 741–52. 45. Klionsky DJ, Abdelmohsen K, Abe A, et al. Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy. 2016; 12: 1-222. 46. Romeo-Guitart D, Leiva-Rodríguez T, Espinosa-Alcantud M, et al. SIRT1 activation with neuroheal is neuroprotective but SIRT2 inhibition with AK7 is detrimental for disconnected motoneurons. Cell Death Dis. 2018; 9: 531. 47. Ng F, Tang BL. Sirtuins’ modulation of autophagy. J Cell Physiol. 2013; 228: 2262–70. 48. Petegnief V, Planas AM. SIRT1 regulation modulates stroke outcome. Transl Stroke Res. 2013; 4: 663–71. 49. Romeo-Guitart D, Leiva-Rodriguez T, Forés J, Casas C. Improved Motor Nerve Regeneration by SIRT1/Hif1a-Mediated Autophagy. Cells. 2019; 8: 1354. 50. Pan YR, Song JY, Fan B, et al. MTOR may interact with PARP-1 to regulate visible light-induced parthanatos in photoreceptors. Cell Commun Signal. 2020; 18: 1–17. 51. Rajamohan SB, Pillai VB, Gupta M, et al. SIRT1 Promotes Cell Survival under Stress by Deacetylation-Dependent Deactivation of Poly(ADP-Ribose) Polymerase 1. Mol Cell Biol. 2009; 29: 4116–29. 52. Cantó C, Sauve AA, Bai P. Crosstalk between poly(ADP-ribose) polymerase and sirtuin enzymes. Molecular Aspects of Medicine. 2013; 34: 1168-1201 53. Ghosh HS, McBurney M, Robbins PD. SIRT1 negatively regulates the mammalian target of rapamycin. PLoS One. 2010; 5: e9199. 54. Fatokun AA, Dawson VL, Dawson TM. Parthanatos: mitochondrial-linked mechanisms and therapeutic opportunities. Br J Pharmacol. 2014; 171: 2000–16. 55. Wang X, Hu S, Liu L. Phosphorylation and acetylation modifications of FOXO3a: Independently or synergistically? Oncol Lett. 2017; 13: 2867–72. 56. Warr MR, Binnewies M, Flach J, et al. FoxO3a Directs a Protective Autophagy Program in Hematopoietic Stem Cells. Nature. 2013; 494: 323–7. 57. Ferguson D, Shao N, Heller E, et al. SIRT1-FOXO3a Regulate Cocaine Actions in the Nucleus Accumbens. J Neurosci. 2015; 35: 3100–11. 58. Descloux C, Ginet V, Rummel C, Truttmann AC, Puyal J. Enhanced autophagy contributes to excitotoxic lesions in a rat model of preterm brain injury. Cell Death Dis. 2018; 9: 853. 59. Puyal J, Vaslin A, Mottier V, Clarke PGH. Postischemic treatment of neonatal cerebral ischemia should target autophagy. Ann Neurol. 2009; 66: 378–89. 60. Weis SN, Toniazzo AP, Ander BP, et al. Autophagy in the brain of neonates following hypoxia–ischemia shows sex- and region-specific effects. Neuroscience. 2014; 256: 201–9. 61. Zhu C, Wang X, Xu F, et al. The influence of age on apoptotic and other mechanisms of cell death after cerebral hypoxia–ischemia. Cell Death Differ. 2005; 12: 162–76. 62. Peker N, Gozuacik D. Autophagy as a Cellular Stress Response Mechanism in the Nervous System. J Mol Biol. 2020; In press. doi: 10.1016/j.jmb.2020.01.017 http://www.thno.org