International Journal of Molecular Sciences Article Human-Induced Neural and Mesenchymal Stem Cell Therapy Combined with a Curcumin Nanoconjugate as a Spinal Cord Injury Treatment Pablo Bonilla 1 , Joaquim Hernandez 2 , Esther Giraldo 1,3 , Miguel A. González-Pérez 1 , Ana Alastrue-Agudo 1 , Hoda Elkhenany 1,4 , María J. Vicent 5 , Xavier Navarro 2 , Michael Edel 6 and Victoria Moreno-Manzano 1, * 1 Neuronal and Tissue Regeneration Laboratory, Centro de Investigación Príncipe Felipe, 46012 Valencia, Spain; pbonilla@cipf.es (P.B.); egiraldo@cipf.es (E.G.); miguelangelgozalezperez14@gmail.com (M.A.G.-P.); aalastrue@cipf.es (A.A.-A.); hoda.atef@alexu.edu.eg (H.E.) 2 Neuroplasticity and Regeneration Group, Department Cell Biology, Physiology and Immunology, Institute of Neurosciences, Universitat Autònoma de Barcelona and CIBERNED, 08193 Bellaterra, Spain; Joaquim.Hernandez@uab.cat (J.H.); Xavier.Navarro@uab.cat (X.N.) 3 Department of Biotechnology, Universitat Politècnica de València, 46022 Valencia, Spain 4 Department of Surgery, Faculty of Veterinary Medicine, Alexandria University, Alexandria 22785, Egypt 5 Polymer Therapeutics Laboratory, Centro de Investigación Príncipe Felipe, 46012 Valencia, Spain; mjvicent@cipf.es  6  Laboratory of Regenerative Medicine, Institut Barraquer, 08021 Barcelona, Spain; edel.michael@gmail.com * Correspondence: vmorenom@cipf.es Citation: Bonilla, P.; Hernandez, J.; Giraldo, E.; González-Pérez, M.A.; Abstract: We currently lack effective treatments for the devastating loss of neural function associated Alastrue-Agudo, A.; Elkhenany, H.; with spinal cord injury (SCI). In this study, we evaluated a combination therapy comprising human Vicent, M.J.; Navarro, X.; Edel, M.; neural stem cells derived from induced pluripotent stem cells (iPSC-NSC), human mesenchymal Moreno-Manzano, V. stem cells (MSC), and a pH-responsive polyacetal–curcumin nanoconjugate (PA-C) that allows the Human-Induced Neural and sustained release of curcumin. In vitro analysis demonstrated that PA-C treatment protected iPSC- Mesenchymal Stem Cell Therapy NSC from oxidative damage in vitro, while MSC co-culture prevented lipopolysaccharide-induced Combined with a Curcumin Nanoconjugate as a Spinal Cord activation of nuclear factor-κB (NF-κB) in iPSC-NSC. Then, we evaluated the combination of PA-C Injury Treatment. Int. J. Mol. Sci. 2021, delivery into the intrathecal space in a rat model of contusive SCI with stem cell transplantation. 22, 5966. https://doi.org/10.3390/ While we failed to observe significant improvements in locomotor function (BBB scale) in treated ijms22115966 animals, histological analysis revealed that PA-C-treated or PA-C and iPSC-NSC + MSC-treated animals displayed significantly smaller scars, while PA-C and iPSC-NSC + MSC treatment induced Academic Editors: Mootaz Salman, the preservation of β-III Tubulin-positive axons. iPSC-NSC + MSC transplantation fostered the Philip Kitchen and Roslyn Bill preservation of motoneurons and myelinated tracts, while PA-C treatment polarized microglia into an anti-inflammatory phenotype. Overall, the combination of stem cell transplantation and PA-C Received: 4 May 2021 treatment confers higher neuroprotective effects compared to individual treatments. Accepted: 29 May 2021 Published: 31 May 2021 Keywords: spinal cord injury; stem cells; curcumin; neuroprotection; polymer–drug conjugate Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- 1. Introduction iations. The physiopathology of spinal cord injury (SCI) involves the disruption of spinal pathways and a primary succession of processes initiated immediately after trauma that prompt rapid and massive cell death within the nervous tissue and the concomitant Copyright: © 2021 by the authors. invasion of ectopic immune and fibrotic cells [1]. During a secondary injury phase, neuronal Licensee MDPI, Basel, Switzerland. death expands to neighboring segments, and neuroinflammation and the associated non- This article is an open access article permissive microenvironment [2] limit the capacity of adult neurons to spontaneously distributed under the terms and regenerate axons [3]. Therefore, a multifactorial approach to SCI treatment addressing the conditions of the Creative Commons broad range of pathological factors involved may have better success than conventional Attribution (CC BY) license (https:// monotherapeutic advances that focus on a single factor. creativecommons.org/licenses/by/ Stem cell transplantation, particularly the use of neural stem cells (NSC) [4], represents 4.0/). a potentially effective treatment approach for SCI. Pre-clinical assays have demonstrated Int. J. Mol. Sci. 2021, 22, 5966. https://doi.org/10.3390/ijms22115966 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2021, 22, 5966 2 of 25 that NSC can replace lost host cells, bridge lesions [5], recover myelin sheaths to reconsti- tute neuronal circuitry and lost connectivity [6], provide trophic support [7], and modulate neuroinflammation [8]. The US Food and Drug Administration (FDA) approved an im- mortalized NSC line (NSI-566), derived from human early fetal spinal cord tissue, to treat chronic SCI in phase I clinical trials reporting no serious adverse effects [9]. The admin- istration of NSI-566 in several pre-clinical models of SCI (rodents [10] and non-human primates [11]) prompted significant improvements to neurological function and suppressed spasticity by supporting extensive axonal sprouting and the development of synaptic con- tacts with host neurons. However, NSI-566 transplantation requires immunosuppression, which is a disadvantage that could be avoided via the use of derivatives of autologous or immune-compatible induced pluripotent stem cells (iPSC). The induced differentiation of human iPSC into NSC (iPSC-NSC) provides a prolifera- tive and broadly expandable in vitro cell source with glial [12] and neuronal differentiation potential [13,14]. In 2019, Okano et al. described the first licensed trial for the clinical evalu- ation of iPSC-NSC in chronic SCI treatment in Japan [15]. Okano’s group had previously employed rodent [13] and primate [16] models to demonstrate how iPSC-NSC significantly improved locomotion after severe to moderate traumatic SCI [17]. Despite the beneficial effects of iPSC-NSC, including their autologous nature that avoids immune rejection [18], their potential tumorigenicity remains a significant impediment to their clinical use [19]. In this regard, we recently generated genetically stable human iPSC using a modified reprogramming procedure employing the transfection of synthetic mRNAs coding for CYCLIN D1 and the OSKL reprogramming factors (OCT3/4, Sex determining Region Y-box 2 (SOX2), Kruppel Like Factor 4 (KLF4), homolog lin-28 (LIN28)), thereby avoiding the use of C-MYC [20], under clinically compatible conditions. iPSC-NSC subsequently derived from these newly generated iPSC displayed reduced genetic instability, reduced cell proliferation in teratoma assays, and efficient survival, engraftment, and differentiation in a hostile SCI microenvironment [20]. Numerous related approaches have also explored the safety and efficacy of mesenchy- mal stem cell (MSC)-based therapies for SCI [21–25] and highlighted MSC transplantation as an interesting means to locomotor recovery in animal models [26]. Our recent studies indicated that the transplantation of allogeneic MSC combined with immunosuppression supports the survival of engrafted cells, which improves functional and morphological outcomes after SCI [27]. The therapeutic benefits of MSC primarily relate to the secre- tion of paracrine acting factors [28] that provide neural support and promote remyelina- tion [29]. In addition, MSC also display immunomodulatory [30], anti-apoptotic [31], and angiogenic [32] abilities but lack tumor-initiating potential [33]. Furthermore, given their accessibility and feasible isolation, MSC can be used in an autologous manner. Given the pathological heterogeneity of SCI, multifaceted combinatorial approaches have been developed that induce improvements regarding cell grafting and survival [11,34], neural differentiation [35], axonal regeneration [35], and the prevention of secondary damage [36]. We previously described a successful combination therapy comprising NSC transplantation and treatment with a water-soluble pH-responsive polyacetal–curcumin nanomedicine incorporating curcumin in the polyacetal mainchain (PA-Curcumin or PA-C), which supports the local sustained release of the active compound curcumin [37]. This synergistic combination led to increased levels of neuroprotection and induced functional recovery in a severe model of chronic SCI in adult rats [37]. In the current study, we report on the development of an enhanced synergistic combi- nation therapy comprising PA-C and iPSC-NSC + MSC treatment as a treatment option for severe sub-acute traumatic SCI. This advanced approach combines the induction of neuroprotection and neuroregeneration to target the secondary pathological mechanisms that occur after SCI. Int. J. Mol. Sci. 2021, 22, 5966 3 of 28 Int. J. Mol. Sci. 2021, 22, 5966 neuroprotection and neuroregeneration to target the secondary pathological mechanisms 3 of 25 that occur after SCI. 2. Results 2. Results 2.1. 2.1. Cultured Cultured iPSC-NSC iPSC-NSC Express Neuronal, Astroglial, Express Neuronal, Astroglial, and and Oligodendrocyte-Specific Oligodendrocyte-Specific Lineage Markers Lineage Markers We firstcharacterized We first characterized thethe heterogeneous heterogeneous iPSC-NSC iPSC-NSC derived derived fromD-reprogrammed from cyclin cyclin D-repro- grammed iPSC [20]. Immunofluorescent evaluation of iPSC [20]. Immunofluorescent evaluation of the three neural lineages performedthe three neural lineagesunder per- formed underculture proliferative proliferative culture conditions conditions (representative (representative images shown images in Figureshown in Figure 1A) 1A) demonstrated demonstrated a significant a significant percentage ofpercentage cells positiveof cells positive for early for early neural neuralmarkers, precursor precursorincluding markers, including SOX2 (75.39 SOX2 (75.39 of ± 3.24% ± 3.24% of totalNESTIN total cells), cells), NESTIN (93.05 ± (93.05 2.69% ± 2.69% of cells), of total total cells), Paired Paired box box protein protein (PAX6) (PAX6) (45.98(45.98 ± 5.01%± 5.01% of totalofcells), total and cells), and neurogenic neurogenic locushomolog locus notch notch homolog protein 1protein (Notch1) 1 (Notch1) (39.76 ±(39.76 5.63%±of5.63% total of totalwhich cells), cells), confirmed which confirmed their lineage their neural neural lineage [38,39] [38,39] (Figure (Figure 1B). Furthermore, 1B). Furthermore, we discovered we discovered that around that around one-third one-third of iPSC-NSC of iPSC-NSC expressedex- pressed ForkheadForkhead Box J1 (FOXJ1), Box J1 (FOXJ1), which is awhich marker is for a marker ciliatedfor ciliated cells cells that that include theinclude the neu- neuroepithelial roepithelial cells lining thecells lining thecanal ependymal ependymal canalcord of the spinal of the spinal [40]. cord [40]. iPSC-NSC also iPSC-NSC also ex- expressed markers of neurons pressed (β-III of markers neurons (β-III±Tubulin—37.05 Tubulin—37.05 4.06% of total cells, ± 4.06% andofdoublecortin total cells, and(DCX)—54.8 doublecortin ± 4.23% of total± cells), (DCX)—54.8 4.23%glia (glialcells), of total fibrillary acidicfibrillary glia (glial protein (GFAP)—48.35 ± 5.97% of total± acidic protein (GFAP)—48.35 cells), and 5.97% oligodendrocytes of total (O4—39.97 ± 8.01% cells), and oligodendrocytes of total± cells) (O4—39.97 8.01%(Figure of total1A,B). cells) We previously (Figure 1A,B). showed We comparable previously showed neuronal comparableand glial cell identity neuronal and glialincelliPSC-NSC identity cultures in iPSC-NSC uponcultures neural induction upon neuralmedium inductionconditions, mediumwith about 60% conditions, withof Tubulin-positive β-III60% about cells and about of β-III Tubulin-positive cells 34% about and of GFAP-positive cells [14]. cells [14]. 34% of GFAP-positive Figure 1. In vitro characterization of cultured iPSC-NSC. (A). Representative immunofluorescence Figure immunofluorescence staining staining images images for the indicated cell indicated cell markers. markers. DAPI DAPI (blue) (blue) employed employed for for nuclei nuclei counterstaining. counterstaining. Scale Scale bar bar == 20 20 µm. µm. (B). (B). Quantitative Quantitative analysis analysis of of the percentage of positive cells for each cell marker. Data expressed as mean ± S.E.M. from three independent cell the percentage of positive cells for each cell marker. Data expressed as mean ± S.E.M. from three independent cell culture culture experiments. experiments. 2.2. 2.2. PA-C TreatmentIncreases PA-C Treatment IncreasesiPSC-NSC iPSC-NSCViability, Viability,Enhances EnhancesNeurite NeuriteElongation, Elongation,but butFails Failstoto Induce Neural and Induce Neural and Glial Glial Differentiation Differentiation We previouslydescribed We previously describedthat thatpolyacetal–curcumin polyacetal–curcuminconjugate conjugate (PA-C, (PA-C, curcumin curcumin load- loading ing 3.8% w/v) displayed significantly lower cytotoxicity than free curcumin (C) 3.8% w/v) displayed significantly lower cytotoxicity than free curcumin (C) in primary cul- in primary cultures of adult tures of adult rat spinal rat spinal cordcord NSCNSC[37]. [37]. To elucidate To elucidate whetherwhether PA-C influences PA-C influences the via- the viability of bility of iPSC-NSC, we performed an MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carbox- iPSC-NSC, we performed an MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl) ymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) -2-(4-sulfophenyl)-2H-tetrazolium) cell viability assaycell viability at 24 h after assay at 24with treatment h after treat- increas- ment with increasing ing concentrations concentrations (up to 20 µM) of (upPA-Cto 20 µM) of PA-C at curcumin at curcumin equivalent equivalent con- concentrations. We centrations. discovered thatWe the discovered evaluatedthat the evaluatedofconcentrations concentrations PA-C failed to of PA-Ccytotoxicity induce failed to induce cy- in iPSC- NSC; moreover, PA-C treatment significantly increased iPSC-NSC metabolic activity in a dose-dependent manner at concentrations higher than 10 µM (Figure 2A). Int. J. Mol. Sci. 2021, 22, 5966 4 of 25 Int. J. Mol. Sci. 2021, 22, 5966 5 of 28 Figure 2. (A). Effect of PA-C treatment on iPSC-NSC viability and neurite outgrowth in vitro. (A). Cell viability evaluations Figure 2. (A). Effect of PA-C treatment on iPSC-NSC viability and neurite outgrowth in vitro. (A). Cell viability evaluations of iPSC-NSC treated with increasing concentrations (5, 10, 12.5, 15, 17.5, and 20 µM) of PA-C for 24 h. Values represented of iPSC-NSC treated with increasing concentrations (5, 10, 12.5, 15, 17.5, and 20 µM) of PA-C for 24 h. Values represented as a percentage relative to vehicle-treated control iPSC-NSC. Data expressed as mean ± S.E.M. of three independent experiments determined by one-way ANOVA with Dunnett’s multiple comparison test (* p < 0.05, ** p < 0.01, *** p < 0.001 vs. control). (B). Immunofluorescence quantification of the percentage of iPSC-NSC positive for astroglial (GFAP; left panel) and neuronal (β-III-Tubulin; right panel) markers after treatment with 10 µM PA-C for 24 h. (C). Representative images of immunofluorescence staining of β-III-Tubulin (green) and nuclear DAPI staining (blue) (scale bar = 150 µM). (D). Quantitative analysis of neurite elongation in iPSC-NSC treated with 10 µM PA-C for 24 h. Data represented as average neurite length (µm) and expressed as mean ± S.E.M. of three independent experiments determined by unpaired Student’s t-test (* p < 0.05 vs. Control). (E). Representative images of immunofluorescence staining of β-III-Tubulin (green) and nuclear DAPI staining (blue) (scale bar = 150 µm). (F). Quantitative analysis of neurite elongation in iPSC-NSC pre-treated with 10 µM PA-C or 50 µM PGA-SS-Fas for 24 h and then treated with LPA for a further 24 h. Data represented as average neurite length (µm). Data expressed as mean ± S.E.M. of three independent experiments determined by one-way ANOVA with Tukey’s multiple comparison test and unpaired Student’s t-test (* p < 0.05, ** p < 0.01, *** p < 0.001 vs. Control; ### p < 0.001 as indicated). Int. J. Mol. Sci. 2021, 22, 5966 5 of 25 Next, we studied whether a 24 h treatment with 10 µM PA-C influenced iPSC-NSC fate determination. Staining for cells positive for GFAP or β-III-Tubulin indicated that PA-C treatment failed to alter the proportion of astrocytes and neurons, respectively, com- pared to control vehicle-treated iPSC-NSC (Figure 2B). We also analyzed gene expression profiles using an RT2 Profiler PCR Array that includes a set of genes related to neural differentiation such as Brain-Derived Neurotrophic Factor (BDNF), Oligodendrocyte tran- scription factor (Olig2), SOX2, Achaete-Scute Family BHLH Transcription Factor 1 (ASCL1), Dopamine receptor D2 (DRD2), and apoptosis-related genes such as Adenosine A1 receptor (ADORA1) and BCL2 (Table S1). PA-C treatment failed to influence the expression of genes involved in apoptosis, which agrees with the results of the cytotoxicity assays (Figure 2A). Furthermore, from this select group of genes, only DRD2 expression became significantly downregulated after treatment with PA-C (* p < 0.05 compared to control vehicle-treated iPSC-NSC; Table S1), indicating that PA-C treatment did not significantly induce neural differentiation of iPSC-NSC. Previous studies established that PA-C treatment induced neurite outgrowth [37]. While the incubation of iPSC-NSC with 10 µM PA-C for 24 h failed to induce significant neural differentiation, we found that PA-C treatment significantly induced neurite elongation in β-III Tubulin-positive cells (Figure 2C,D). We also evaluated the functional relevance of induced neurite elongation by PA-C treatment in an inhibitory environment by treating iPSC-NSC with lysophosphatidic acid (LPA), which activates the Rho/ROCK pathway and induces growth cone retraction and neurite collapse [41,42]. We employed free fasudil (Fas) [43], a well-known Rho kinase inhibitor, and a nanoconjugated form of fasudil (PGA-SS-Fas) [44] as positive controls for neurite elongation. In a similar manner to curcumin, we previously reported that polymer conjugation enhanced the sta- bility of fasudil and supported sustained release, which improved the neuroprotective and regenerative activity of fasudil in SCI models [44]. While treatment with PGA-SS-Fas or Fas (Figure 2E,F) permitted a significant increase in axonal elongation in the presence of LPA, we found no differences following the treatment of iPSC-NSC with PA-C (Figure 2D) or C (Figure 2E,F). Representative images of β-III-Tubulin immunostaining in the absence of LPA treatment show neurite elongation following PA-C treatment (Figure 2C), abundant and long neural extensions following PGA-SS-Fas treatment (Figure 2E,F), and the absence of morphological changes in PA-C-treated iPSC-NSC in the presence of LPA in comparison with vehicle-treated control iPSC-NSC (Figure 2E,F). 2.3. PA-C Protects iPSC-NSC against Hydrogen Peroxide-Induced Toxicity The excessive production and release of reactive oxygen species following SCI con- tribute to the secondary injury phase by exacerbating acute damage to spinal cord-resident cells [45,46] and generating a hostile microenvironment for cell-based therapies. Antioxi- dant treatments have emerged as an effective approach to ameliorate cell death/secondary damage. We evaluated PA-C pre-treatment (5, 10, 12.5, 15, 17.5, or 20 µM) in the prevention of oxidative cell damage following the exposure of iPSC-NSC to high doses of hydrogen peroxide (H2 O2 ) (75 or 100 µM) for 24 h. We determined H2 O2 -induced cytotoxicity by MTS viability assay, discovering that PA-C treatment in the range of 5 to 15 µM prevented 75 µM H2 O2 -induced cytotoxicity closed to the control non-treated condition. Representa- tive phase-contrast images evidenced a cell survival-inducing effect of 10 µM of PA-C in the presence of 75 µM H2 O2 (Figure 3A); however, PA-C pre-treatment failed to protect iPSC-NSC treated with 100 µM H2 O2 (Figure S1). Int. J. Mol. Sci. 2021, 22, 5966 6 of 25 Int. J. Mol. Sci. 2021, 22, 5966 7 of 28 Figure 3.Figure 3. Neuroprotective Neuroprotective effecteffect of PA-C of PA-C pre-treatment pre-treatment followingperoxide-induced following peroxide-induced cytotoxicity in iPSC-NSC. cytotoxicity (A). Effect in iPSC-NSC. (A). Effect of of PA-C pre-treatment (5, 10, 12.5, 15, 17.5, and 20 µM) on iPSC-NSC viability following a 24 h incubation with 75 µM PA-C pre-treatment (5, 10, 12.5, 15, 17.5, and 20 µM) on iPSC-NSC viability following a 24 h incubation with 75 µM H2 O2 . H2O2. MTS viability values represented as a percentage of vehicle-treated control cells. Data expressed as mean ± S.E.M. MTS viability values of three represented independent as a percentage experiments of one-way determined by vehicle-treated ANOVAcontrol multiplecells. Data expressed comparison meanvs.±Control; tests (*** p