Safety and efficacy of human embryonic stem cell-derived astrocytes following intrathecal transplantation in SOD1G93A and NSG animal models

Safety and efficacy of human embryonic stem cell-derived astrocytes following intrathecal... Background: Amyotrophic lateral sclerosis (ALS) is a motor neuron (MN) disease characterized by the loss of MNs in the central nervous system. As MNs die, patients progressively lose their ability to control voluntary movements, become paralyzed and eventually die from respiratory/deglutition failure. Despite the selective MN death in ALS, there is growing evidence that malfunctional astrocytes play a crucial role in disease progression. Thus, transplantation of healthy astrocytes may compensate for the diseased astrocytes. Methods: We developed a good manufacturing practice-grade protocol for generation of astrocytes from human embryonic stem cells (hESCs). The first stage of our protocol is derivation of astrocyte progenitor cells (APCs) from hESCs. These APCs can be expanded in large quantities and stored frozen as cell banks. Further differentiation of the APCs yields an enriched population of astrocytes with more than 90% GFAP expression (hES-AS). hES-AS were G93A injected intrathecally into hSOD1 transgenic mice and rats to evaluate their therapeutic potential. The safety and biodistribution of hES-AS were evaluated in a 9-month study conducted in immunodeficient NSG mice under good laboratory practice conditions. Results: In vitro, hES-AS possess the activities of functional healthy astrocytes, including glutamate uptake, promotion of axon outgrowth and protection of MNs from oxidative stress. A secretome analysis shows that these hES-AS also secrete several inhibitors of metalloproteases as well as a variety of neuroprotective factors (e.g. TIMP-1, G93A TIMP-2, OPN, MIF and Midkine). Intrathecal injections of the hES-AS into transgenic hSOD1 mice and rats significantly delayed disease onset and improved motor performance compared to sham-injected animals. A safety study in immunodeficient mice showed that intrathecal transplantation of hES-AS is safe. Transplanted hES-AS attached to the meninges along the neuroaxis and survived for the entire duration of the study without formation of tumors or teratomas. Cell-injected mice gained similar body weight to the sham-injected group and did not exhibit clinical signs that could be related to the treatment. No differences from the vehicle control were observed in hematological parameters or blood chemistry. Conclusion: Our findings demonstrate the safety and potential therapeutic benefits of intrathecal injection of hES-AS for the treatment of ALS. Keywords: Amyotrophic lateral sclerosis, Astrocytes, Human embryonic stem cells, Superoxide dismutase 1 * Correspondence: m.izrael@kadimastem.com Neurodegenerative Diseases Department at Kadimastem Ltd, Pinchas Sapir 7, Weizmann Science Park, Nes-Ziona, Israel Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 2 of 17 Background Inflammation-mediated neuronal injury is also recog- Amyotrophic lateral sclerosis (ALS) is an adult-onset nized as a major factor to promote ALS disease progres- disease characterized by the loss of both upper and sion and amplifies MN death-inducing processes. The lower motor neurons (MNs). Symptoms include pro- neuroimmune activation is not only a physiological reac- gressive paralysis of MN target muscles. The disease is tion to cell-autonomous death, but also an active compo- incurable, and fatal within 3–5 years of first symptoms, nent of non-autonomous cell death. Astrocytes participate due to respiratory failure when the diaphragm is affected in the cellular response to damage and danger signals by [1]. About 10–15% of cases of ALS are familial, and the releasing inflammation-related molecules like NO, IL-6, other cases are sporadic. Familial ALS includes muta- INF-γ, Prostaglandin D2, TGF-β and TNF-α that can 2+ 2+ tions in Cu /Zn superoxide dismutase-1 (SOD1) [2] induce the apoptosis of neurons observed in ALS disease and in RNA/DNA binding proteins FUS and TAR DNA [19–23]. In both physiological and pathological condi- binding protein-43 [3] However, the most frequent tions, astrocytes secrete a wide range of factors with genetic cause of ALS (40% of familial ALS) is an amplifi- multiple influences on their cellular neighbors. cation of a hexanucleotide in a noncoding region of the In addition, disruption of the astrocytic TNFR1– C9orf72 gene [4]. GDNF axis accelerates MN degeneration and disease The pathological mechanisms for ALS are still not well progression, as the levels of the protective agents for understood and the proposed mechanisms include MNs, glial-derived neurotrophic factor (GDNF), are inflammation, oxidative stress, glutamate cytotoxicity reduced [24]. Astrocytes in the ALS rat model acquire and protein aggregation. Although MNs are the main an accelerated senescent phenotype that shows reduced affected cells in the disease, a growing body of evidence support in MNs, that can be partially reversed by GDNF suggests the involvement of astrocytes in the pathology [25]. Another factor that plays a role in ALS pathology is of ALS in a non cell autonomous pathway. The contri- vascular endothelial growth factor (VEGF), originally bution of astrocytes to the pathology of ALS is probably described as a factor with a regulatory role in vascular a combination of loss of homeostatic functions and/or growth and development but it also directly affects neur- gain of toxic functions. Several mechanisms by which onal cells [26, 27]. Transgenic mice expressing reduced ALS patients’ astrocytes affect ALS pathology include levels of VEGF develop late-onset MN pathology, similar astrocyte toxicity; astrocytes that were isolated from to that of ALS [28, 29]. VEGF is secreted by astrocytes sporadic and familial postmortem ALS patients and and has been shown to protect MNs from excitotoxic astrocytes derived from iPSCs of ALS patients have been death, as occurs in ALS [30]. In line with these results, shown to be toxic to healthy (WT) MNs [5, 6]. Similar low levels of VEGF and GDNF were reported in the results were obtained by primary astrocytes isolated cerebrospinal fluid (CSF) of ALS patients [31]. Other G93A from the hSOD1 mouse model with both WT and mechanisms include activation of necroptosis [32] and MNs derived from ALS [7, 8]. The toxic effect of mitochondrial alterations [33–37]. astrocytes on MNs was also demonstrated by addition of These observations led to the rationale that ALS could astrocyte condition medium [9, 10]. This lead to the be treated by implantation of normal wild-type healthy notion that astrocytes of ALS patients secrete toxic/mu- astrocytes from an external source, to support or replace tated proteins that cause specific death of MNs. This dysfunctional ALS astrocytes [38]. In the present work, hypothesis is also supported by in-vivo studies in the we have used human embryonic stem cells (hESCs) as a G93A hSOD1 high copy number ALS models [11–14]. source for large-scale production of astrocyte progenitor Another proposed mechanism is the reduction of func- cells (APCs), which can be stored as frozen banks. These tional astrocytic glutamate uptake suggested to contrib- APCs can be further expanded and differentiated into an ute to glutamate excitotoxicity found in ALS patients enriched population of young committed astrocytes by [15]. GLT-1, a glutamate transporter (aka EAAT2), was removal of the growth factors for 7 days (hES-AS), found impaired in ALS patients [16, 17]. In-vivo studies which demonstrate functional properties of “healthy” as- have demonstrated that focal loss of GLT-1 in the ven- trocytes in vitro. These properties include: uptake of glu- tral horn of the spinal cord precedes disease onset in a tamate; production and secretion of a wide diversity of transgenic rat model for ALS overexpressing SOD1 [18]. neuroprotective factors, as seen by secretome analysis; Transplantation of SOD1(G93A) glial-restricted precur- promotion of axonal outgrowth; and protection of MNs sor cells–glial progenitors that are capable of differenti- from oxidative stress. In animal ALS models (high-copy G93A ating into astrocytes in the cervical spinal cord of WT number hSOD1 transgenic mice and rats), we show rats induced host MN ubiquitination and death, fore- that intrathecal injection of hES-AS into the CSF of G93A limb motor and respiratory dysfunction, and reactive hSOD1 mice and rats had significant effects on delay- astrocytosis and reduced GLT-1 transporter expression ing disease onset, maintaining motor performances and in WT animals [11]. delayed death. To obtain safety data that are relevant to Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 3 of 17 both hES-AS and to their proposed clinical use, we con- (7-day astrocytes, hES-AS), flow cytometry showed that ducted long-term safety and toxicology studies in NSG the percentages of GLAST, GFAP and AQP-4 astrocytic immune-deficient mice. These studies were designed to markers were increased compared to APCs (Fig. 1d). address key safety aspects associated with direct adminis- Upon differentiation of APCs toward committed young tration of hES-AS into the CSF by intrathecal injection, astrocytes there were no remaining undifferentiated cells, including toxicity, biodistribution, long-term engraftment as shown by the levels of TRA-1-60, SSEA-4 and EPCAM, and formation of tumors. which remained < 0.1% (Fig. 1e), indicating high purity and low risk of teratoma formation [47]. It is important to Results note that only few Ki-67-positive cells were observed in Direct differentiation of hESCs into astrocyte progenitor hES-AS cultures (Fig. 1f), indicating that most hES-AS cells and young astrocytes are post mitotic. Two hESC lines (HADC100 and NCL-14) were used to G93A produce astrocytes for engraftment in hSOD1 ALS Biological functionality of hES-AS animal models. Both hESC lines had a normal karyotype, Glutamate uptake capacity expressed pluripotency markers and were capable of The glutamate uptake capacity of hES-AS was tested by differentiating into all three embryonic germ layers incubating the cells in medium containing 0.5 mM glu- [39, 40]. We modified our previously reported protocol tamate and measuring the remaining concentration of [41] to generate an enriched population of APCs from the neurotransmitter at different times up to 120 min. hESCs, followed by further differentiation of the APCs Astrocytes from human spinal cord served as positive into functional astrocytes (Fig. 1a). The protocol was control and medium without cells as negative control. optimized to include good medical practice (GMP)-grade As shown in Fig. 2a, the hES-AS take up glutamate from materials and factors to be compatible for clinical use. In the medium occurred in a time-dependent manner simi- brief, hESC cultures having at least 70% of pluripotent lar to the control human spinal cord astrocytes. After stem cells expressing the SSEA4, TRA-1-60 and EPCAM 2 h, more than 85% of the glutamate was removed from markers were used as a starting material. The hESCs were the culture media. detached and cultured in suspension with stepwise To investigate whether GLT-1 (EAAT2) participates in changes in media composition (Fig. 1a, b). First, all-trans the glutamate uptake, the same experiment was done in retinoic acid and EGF were added for 7 days. This elicited the presence of either WAY-213,613 (1 μM) or dihydro- increased production of bone morphogenetic factors (i.e. kainic acid (DHK, 500 μM) [48]. With either of these BMP4, BMP6, BMP2, BMP7 and BMP11), which were GLT-1 inhibitors (Fig. 2b) the removal of glutamate in found to be essential for obtaining glial restricted cells, 60 min was inhibited by 60% (from 64.1% removal in the particularly astrocyte lineage cells [41, 42]. The suspension control to 25% with the inhibitors), demonstrating that a culture was continued with EGF resulting in the formation significant part of the glutamate uptake can be attributed of neurospheres, which were seeded in 2D culture on to GLT-1 activity in the hES-AS. laminin. The cells were expanded by successive passages in the presence of growth factors (bFGF and EGF) and human serum with the doubling time being 21 ± 2.6 h. This produced APCs that can be stored as frozen cell Neuroprotective effect against oxidative stress banks. The APC karyotype was tested at different passages Cultures of mouse spinal cord MNs were challenged (up to passage 12) and was found normal (Fig.1c). Flow with 150 μM hydrogen peroxide (H O ). The number of 2 2 cytometry analysis of APCs showed that the levels of apoptotic MNs was measured after staining for activated pluripotent markers, SSEA-4, EPCAM and Tra-1-60, were caspase-3 and the total number of MNs being measured <0.2% (Fig. 1e). Above 90% of APCs were positive for the by staining for tubulin-β3. Using high-content image astrocytic marker CD44 [43](Fig. 1d). The APCs had screening analysis, we calculated the percentage of apop- additional astrocytic markers such as the Glutamate totic MNs (seen as yellow cells, Fig. 3b, left panel). The Aspartate Transporter (GLAST, aka Excitatory Amino results (Fig. 3a) indicate a significant decrease (p < 0.05) Acid Transporter 1 (EAAT1)) [44], glial fibrillary acidic in MN death by adding conditioned medium from the protein (GFAP) [45] and Aquaporin-4 (AQP-4) [46], as hES-AS, as seen by the decrease in caspase-3-positive well as neuroepithelial stem cell markers Nestin, A2B5 cells (Fig. 3b, right panel). When the hES-AS were and CXCR-4 (Fig. 1d). The frozen/thawed APCs were added in coculture with the MNs, there was a greater further expanded for 2–3 weeks and then differentiated decrease in apoptosis resulting from oxidative stress toward committed astrocytes, by removing growth factors (Fig. 3a, p < 0.01) to levels similar to spontaneous apop- EGF and bFGF as well as human serum from the media tosis. These results demonstrate the neuroprotective and adding vitamin C. After 7 days without growth factors effects by hES-AS in vitro. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 4 of 17 d e Fig. 1 Differentiation of human embryonic stem cells into astrocyte progenitor cells and committed astrocytes. a Steps and timeline for differentiation of hESCs first into astrocyte progenitor cells (APCs) which can be stored frozen in APC banks. These APCs are further expanded with growth factors (bFGF, EGF and human serum), and then differentiated into astrocytes (hES-AS) by removal of growth factors for 7 days. b Representative images of different steps from hESCs to APCs (as in a, steps marked by asterisk). Arrows show selected neurospheres. c Representative spectral karyotyping analysis showing normal karyotype of APC cell bank at passage 12. d Flow cytometry analysis on nine batches of APC banks (grown with human serum, bFGF and EGF) versus 13 batches of astrocytes differentiated for 7 days showing expression of astrocytic markers (CD44, GLAST, GFAP, and Aquaporin-4) and neuroepithelial stem cell markers (Nestin, A2B5 and CXCR4). e Flow cytometry analysis of APCs and astrocytes differentiated for 7days(same batchesasin d) showing very low expression of pluripotent markers (below limit of detection, 0.1%). f Representative immunofluorescence images of astrocytes differentiated 7 days, showing expression of astrocyte markers (GFAP, GLAST, S100β and AQP-4) and very low proliferation marker (Ki-67, arrow). Scale bars = 100 μm. Error bars represent SD. hESC human embryonic stem cell, DAPI 4′,6-diamidino-2-phenylindole, GFAP Glial Fibrillary Acidic Protein, GLAST Glutamate Aspartate Transporter, RA Retinoic acid hES-AS stimulate axonal outgrowth in vitro The cultures were labeled by ICF with antibodies against We next assessed the ability of hES-AS to induce axonal axonal neurofilament-160 and GFAP markers. Represen- outgrowth in vitro. Rat primary cortical neurons derived tative images of the five conditions are shown in Fig. 4a. from day 18 embryos were precultured for 2 days in By high-content image screening analysis, the total area of Neurobasal medium (with B27) and then further cul- axons and neurites in the NF160-stained images was deter- tured for 4 more days in either medium alone or mined. A significant increase in axonal outgrowth was seen supplemented with 10 ng/ml Neurotrophin-3 (NT-3, as in the neurons cocultured with hES-AS (Fig. 4b, p < 0.01). positive control), or cocultured with hES-AS (1–2×10 Moreover, addition of the hES-AS conditioned medium cells), or cocultured with hES-AS conditioned medium was found to stimulate axonal outgrowth to a similar (collected from days 5 to 7 of astrocyte differentiation). extent as compared to the cocultures, indicating that Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 5 of 17 a b Fig. 2 hES-AS take up glutamate from medium. a Glutamate concentration measured in solutions with 500 μM glutamate that were incubated for indicated times either alone (black bars 1–2) or with hES-AS differentiated for 28 days (black bars 3–7). Kinetics of glutamate removal by hES-AS similar to that by astrocytes extracted from human spinal cord (gray bars). b Percentage of glutamate uptake after 60 min by hES-AS alone or in presence of inhibitors of glutamate transporter GLT-1, WAY-213,613 (1 μM) and DHK (500 μM). Error bars are SD of triplicates. *p < 0.05. hESC human embryonic stem cell, DHK dihydrokainic acid this neurogenic activity can be attributed to factors se- peptidases. This indicates that there is a complex set of creted by these astrocytes. As expected, GFAP-positive factors secreted by the hES-AS, beyond the classical neuro- cells were observed only in the cocultures, indicating trophic factors. Many of these factors may be responsible that the rat cortical neurons were not contaminated by for the neurogenic and neuroprotective activities observed rat astrocytes. earlier. Examples of the secreted factors with effects on neurons or with antiprotease activity are presented in Neurotrophic factor synthesis and secretion Table 1. Several of these factors may be relevant for poten- We first measured the levels of known neurotrophic tial therapeutic mechanism of action in ALS (e.g. Osteo- factors GDNF, BDNF, VEGF and IGF-I both in hES-AS pontin, tissue inhibitor of metalloproteinase (TIMP)-1 and culture supernatant media and in cell extracts (cell con- TIMP-2,Midkine,MIF;see Discussion). tent). VEGF was found to be secreted from hES-AS that G93A were differentiated without growth factors for 28 days Transplantation of hES-AS in SOD1 mouse and rat (Additional file 1: Figure S1). IGF-1 was also secreted, ALS models G93A whereas GDNF and BDNF were found inside the cells Both SOD1 mouse and rat models present a typical but less was secreted (Additional file 1: Figure S1). The pattern of ALS disease progression, in which onset of levels of these classical neurotrophic factors were in the the disease in hindlimbs precedes that in forelimbs, and range found in human CSF [49, 50]. in which the end stage results from compromised To have a more comprehensive view of the factors respiratory function [18, 51]. A dose of 2 × 10 hES-AS secreted by 7-day and 28-day differentiated hES-AS, we (differentiated for 7 days) were injected into the CSF of G93A carried out secretome analysis. The 48-h conditioned hSOD1 mice through the cisterna magna (CM), medium of replica cultures of hES-AS were analyzed using either once on day 67 ± 2 after birth or twice on days 67 the human Quantibody Kiloplex Array (RayBiotech), cap- ± 2 and 97 ± 2 (Additional file 3: Figure S2). Disease able of detecting 1000 proteins. A total of 220 protein fac- onset was determined by the loss of 3% of maximal body tors were found to be secreted at levels over the threshold weight. Results demonstrate that double transplantation in 7-day hES-AS, about 25% of which being more abun- of the hES-AS significantly delayed disease onset com- dant at 28 days (see Additional file 2:Table S1).Among pared to sham-injected controls (Additional file 3: Figure the highest 120, there were 25 proteins with activities in S2A; median 119 days vs 112 days; p = 0.0012, log-rank), neurogenesis, axon or neurite outgrowth or axon guidance. and was better than with a single injection. Motor per- Interestingly, there were 13 proteins with antiprotease formance, as measured by Rotarod test as well as by neuro- activity. In addition, there were extracellular matrix (ECM) logical scoring, was significantly improved in mice that were components, cell adhesion membrane proteins and a few injected twice with hES-AS, compared to sham-injected Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 6 of 17 Fig. 3 hES-AS protect MNs from oxidative stress. A Mouse motor neurons exposed in 96-well plates to 150 μMH O for6h (bar1)or left untreated 2 2 (bar 4). During H O treatment, neuron cultures supplemented with conditioned medium from hESC-derived astrocytes, differentiated for 28 days 2 2 (ACM, bar 2), or with 20,000 of the same hES-AS (bar 3). After fixation, cells double-stained with anti-tubulin β3 antibody (neuron marker, green) and anti-Caspase-3a (apoptotic marker, red). Percentage of apoptotic neurons (Caspase3a over tubulin β3-positive cells) counted using high-content image screening system (Arrayscan; Cellomics). Results represent average ± SD for 10 wells of 96 well-plate per treatment (for each well, 49 fields were analyzed). *p < 0.05; **p <0.01. b Left panel: representative image of neuron cultures with H O treatment. Apoptotic neuronal cell bodies yellow 2 2 (arrows, due to overlapping of red Caspase-3 staining with green tubulin β3). Right panel: with ACM, much less apoptotic yellow cells are seen. Scale bar: 100 μm. hESC human embryonic stem cell, H O hydrogen peroxide 2 2 mice (Additional file 3:FigureS2D,E; p <0.05). Two injec- curve (AUC) analysis). The disease onset was delayed very tions were better than a single dose. The survival of mice significantly by hES-AS treatment (Fig. 5b, p = 0.0001); injected twice with hES-AS was somewhat prolonged com- Kaplan–Meier analysis showed that 50% of treated rats pared to sham-injected mice (Additional file 3:FigureS2B; lost 3% of their body weight by day 175 compared to day median survival 130 days vs 126.5 days; but p=0.1, 157 in the sham-injected group. The hES-AS-treated rats log-rank). With the double injection there was also a trend maintained their body weight significantly longer (by for longer survival at late times, compared to one injection. about 30 days) than sham-injected rats (Fig. 5c, p =0.007). G93A We then shifted to the rat hSOD1 ALS model, A set of motor tests demonstrated the therapeutic which allows use of intrathecal injection by lumbar punc- effect of the hES-AS treatment. First the overall devel- ture (LP), a route of administration similar to what could opment of clinical symptoms, as evaluated by open field be applied in human patients. The rat model also allowed neurological scoring, was significantly delayed (Fig. 5d, administration of more cells. A total of 6 × 10 hES-AS p < 0.001). The decline of motor functionality, as mea- (differentiated for 7 days) was administered divided into sured by “time to fall” from a Rotarod, was markedly two injections, the first on day 50 ± 2 after birth and the slowed down by hES-AS treatment, the animals main- second on day 70 ± 2. A control group was sham-injected taining normal motor activity for more than 1 month with the vehicle solution. The LP injections were in the longer than the controls (Fig. 5e, p < 0.001). Likewise, subarachnoid space between L5 and L6 vertebra. The the loss of forelimb muscle strength, as measured by median survival of the hES-AS-treated rats was 216 days the grip strength test, was significantly slowed down, compared to 182 days in the sham-injected rats (Fig. 5a); just as the Rotarod performance (p <0.001; data not Kaplan–Meier analysis for the entire experiment showed shown). Other observations were that no tumors were an increased survival trend (p = 0.077 by area under the observed in the animals post mortem. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 7 of 17 a b Fig. 4 hES-AS and their conditioned medium stimulate axonal outgrowth in cortical neurons. a Mouse cortical neurons cocultured with hES-AS 4 4 (7-day differentiated APC) (2 × 10 and 4 × 10 cells), or with neurotrophin 3 (NT3) as positive control, or left untreated (negative control). Last row shows neurons cultured with conditioned medium from same hES-AS (taken after 48 h of culture). Representative images of cells stained with DAPI and by immunofluorescence for neurofilament-160 (NF160) and GFAP shown for each condition. Stimulation of axon and neurite outgrowth seen from NF160 stain and merge of NF160 (green) and GFAP (red). Scale bar = 100 μm b By high-content image screening analysis (Arrayscan; Cellomics), area covered by axon and neurite outgrowth quantified, using 49 fields for each of six replica wells from each experimental conditions. Error bars represent SD. *Student’s t test, p <0.05). DAPI 4′,6-diamidino-2-phenylindole, GFAP Glial Fibrillary Acidic Protein Assessment of safety, tumorigenicity and biodistribution attached to the pia mater. To assess the biodistribution of of hES-AS following a single injection to the cisterna hES-AS outside the CNS, the detection of human cells in magna of NSG mice mouse tissues was performed by quantitative real-time PCR The safety, tumorigenicity and biodistribution phases were (qPCR), targeting the specific sequence of the human Alu conducted in compliance with principles of good laboratory sequence. The detection was performed in nine organs practice (GLP) over a period of up to 9 months. hES-AS, including the spleen, kidney, testis/ovary, liver, heart, bone differentiated for 7 days, were injected intrathecally into the marrow of the femur, lungs, and cervical lymph nodes. The CSF of NSG mice through the CM with 0.4 × 10 cells/ qPCR method was validated prior to the study and both mouse, or with a vehicle. Mice were sacrificed 4, 17 and the limit of detection (LOD) and the limit of quantification 39 weeks post transplantation. No clinical signs were (LOQ) were set at one human cell (DNA equivalent) per attributable to treatment during the monitoring periods. 1 μg of mouse DNA. The PCR results showed no detection Cell-injected mice made similar body weight gain by 4, 17 of human DNA above the LOD in any of the tested organs and 39 weeks post dose to the vehicle control groups. In 4 and 17 weeks after transplantation. addition, there were no differences from the vehicle control We also examined the astrocytic identity of hES-AS in at the hematological and blood chemistry investigations at vivo 2 months after their transplantation in the CSF of 4, 17 and 39 weeks after dose administration (data not immunodeficient mice. Histological sections were shown). Histopathological evaluation of the brain and stained for the general human cytoplasmic specific spinal cord was performed to assess tumorigenicity. No marker Stem121 and for Stem123 (human-specific teratoma or other tumors that could be related to the treat- GFAP antibody) in order to ascertain the presence of ment were seen in the transplanted animals in any of tested human cells. All Stem121-positive cells were positive for time points. In order to evaluate the hES-AS distribution in human GFAP, demonstrating that the transplanted the CNS, the sections were stained using an in-situ hES-AS maintained their astrocytic identity in the CSF hybridization (ISH) technique with a human-specific Alu Y (Fig. 7). Further staining for the cell cycle marker sequence. Cells positive for Alu Y sequences were present Ki67 showed that 0.33 ± 0.15% of Stem121-positive at all levels of the CNS in similar incidences between the cells in the CSF were also positive for Ki67, indicating three study time points. The incidence for the various levels for the very low proliferative capacity of hES-AS in range between 17% (distal areas from injection site) and vivo (Fig. 7g). 80% (at vicinity of the injection site) after 4 weeks, between 13% and 97% after 17 weeks and between 21% and 96% Discussion after 39 weeks (Fig. 6 and Additional file 4:Table S2).The This work describes the derivation of young astrocytes cells were almost uniformly seen along the meninges, from human embryonic stem cells (hES-AS), which have Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 8 of 17 Table 1 hES-AS secrete a variety of factors with effects on multiplicity of mechanisms that underlie MN degener- neurons or with antiprotease activity ation in this disease. Thus, a potential therapy that acts 7-day astrocytes 28-day astrocytes through multiple mechanisms of action to treat the 6 6 (ng/ml/10 cells) (ng/ml/10 cells) broad pathological aspects of the disease is more likely Secreted factors with effects on neurons to be effective. An example for the complexity of the Osteopontin (OPN) 53.1 ± 29 56.8 ± 5.5 disease is the involvement of astrocytes in the degener- ation of MNs [5, 7, 8, 57]. Such noncell autonomous Dickkopf-3 (DKK-3) 43.1 ± 14.2 33.8 ± 1.6 death of MNs caused by ALS-type astrocytes supports Thrombospondin 22.7 ± 11.5 118.9 ± 36.8 the rationale that transplantation of healthy human as- (TSP-1) trocytes into the CNS of ALS patients may compensate Secreted Frizzled 20.8 ± 10.9 41.2 ± 23.0 Protein (sFRP3) for the malfunctional astrocytes and rescue dying MNs (review in [38]). Brevican proteoglycan 15.6 ± 4.9 12.6 ± 3.3 hES-AS exhibit multiple activities that were shown to Tripeptidyl peptidase 11.5 ± 4.2 20.1 ± 11.7 be impaired in ALS-type astrocytes. Astrocytes from (CLN2) ALS transgenic mice express more iNOS/NOS2, leading Clusterin 9.5 ± 3.2 6.5 ± 0.5 to increased release of NO, which exacerbates oxidative Midkine 8.4 ± 3.0 6.1 ± 3.5 stress leading to MN death [58]. We show in our study NSE 3.5 ± 1.8 0.9 ± 0.2 that hES-AS protect in-vitro spinal cord MNs from MIF chemokine 1.8 ± 0.6 0.4 ± 0.1 oxidative stress produced by H O In ALS patients, a 2 2. CXCL16 1.5 ± 0.8 2.1 ± 0.2 decrease of the astroglial GLT-1 glutamate transporter is observed [16], leading to decreased glutamate uptake in Thrombospondin-2 0.85 ± 0.4 2.3 ± 0.4 the synaptic clefts of the spinal cord. Accumulation of GRFα-1 0.45 ± 0.2 1.0 ± 0.6 excitatory glutamate makes MNs in ALS more suscep- VEGF 0.05 ± 0.02 0.23 ± 0.09 tible to excitotoxicity [59]. hES-AS express both glutam- Antiprotease activity ate transporters GLAST and GLT-1 and efficiently Fetuin A 1816.0 ± 677 1404.7 ±+ 129.4 uptake glutamate, which is in part due to their GLT-1 Tissue inhibitor of 16.6 ± 6.8 14.5 ± 0.8 expression, as shown by GLT-1 inhibitors. Another metalloprotease TIMP-2 mechanism by which the diseased astrocytes lead to MN PAI-1 Serpine 1 7.2 ± 6.2 54.9 ± 5.9 death is by a decrease in the secretion of neurotrophic protease inhibitor factors. hES-AS produce and secrete the neurotrophic Tissue inhibitor of 7.0 ± 3.8 6.5 ± 0.8 factors GDNF, BDNF, IGF-1 and VEGF in a comparable metalloprotease TIMP-1 amount to that of endogenous astrocytes. The neuro- Serpin A4 4.3 ± 2.5 4.0 ± 0.4 tropic property of hES-AS was demonstrated by cocul- Results shown as mean ± standard deviation for triplicates of hES-AS differentiated tures of hES-AS with neurons and by hES-AS conditioned for 7 days and duplicates of hES-AS differentiated for 28 days medium alone, indicating activity of soluble secreted Secretome analysis performed on 48-h conditioned media of hES-AS. Listed are factors with activities in neuroprotection, neurogenesis, axon growth or factors. Secreted VEGF is likely to play an important guidance, as well as antiproteases. For relevance to amyotrophic lateral sclerosis, role by protecting neurons in ALS, reducing excitotoxi- see Discussion. Complete secretome list presented in Additional file 2:Table S1 GRF GDNF family receptor, hES-AS human embryonic stem cell-derived city [28, 60], and its concentration is lower in the CSF astrocytes (differentiated from APCs for 7 days), MIF macrophage migration of ALS patients [31]. In addition, GDNF synergizes with inhibitory factor, NSE neuron specific enolase, PAI plasminogen activator VEGF to prolong survival in a murine ALS model [61]. inhibitor, VEGF vascular endothelial growth factor Intrathecal injection of CSF from sporadic ALS patients therapeutic activity in vivo following intrathecal injection to neonatal rats induces selective degeneration of MNs G93A into the CSF of transgenic SOD rats and mice. In [62] and downregulates the levels of both BDNF and addition, we describe the results of a preclinical safety IGF-1 in the spinal cord [63]. Supplementation of study in immunodeficient mice to assess the tumorigen- BDNF reverses the neurodegenerative changes induced icity potential and biodistribution of hES-AS in target by ALS-CSF in MN cultures [64]. and distal organs. The nature of the secreted factors was further investi- To date, two FDA-approved drugs, riluzole and Radi- gated by a secretome analysis, clearly illustrating the pleio- cava, were shown to modestly attenuate motor deterior- tropic activity of the cells. hES-AS secrete many factors ation in ALS patients [52–55]. Still, many late-phase having activities on neurons [65, 66–68] as well as several clinical trials failed to demonstrate a significant improve- antiproteases and factors which could remodel the ECM ment in slowing down disease progression when using (see Table 1). Among the more abundant factors found in single-target drugs [56]. ALS is a multifactorial disease the secretome analysis, several have been linked to ALS, and therapeutic approaches should take into account the thereby shedding new light on the possible mechanisms of Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 9 of 17 ab c e G93A Fig. 5 Effect of hES-AS transplantation on disease onset, motor activity and survival in hSOD1 rat ALS model. hES-AS (APCs differentiated for 7 days) injected intrathecally through lumbar puncture (L5–L6), in two doses of 3 × 10 cells each on days 50 and 70 after birth in hSOD1G93A rats. a Kaplan–Meir survival curves of rats treated with hES-AS (green) show prolongation of median survival compared to sham-injected group (vehicle, red). b Kaplan–Meir plot of disease onset (defined by 3% body weight loss) shows significant delay in hES-AS-treated ALS rats. c Body weight maintained significantly longer in hES-AS-treated ALS rats. d Neurological score. e Significant prolongation of motor performance on Rotarod in hES-AS-treated ALS rats. Same seen by grip strength measurement. c, d Values represent mean ± SEM action underlying the observed therapeutic effect in ALS capacity to save primary MNs from the degeneration models. One of the most abundant factors in the secre- caused by the ALS mutant SOD1 form, probably by acting tome is Osteopontin (OPN/SSP1), which in the mutant as a chaperone [74]. Also secreted is Clusterin, another SOD1 model of ALS is found to be associated with MNs chaperone, promoting axon regeneration, as observed on that are more resistant to degeneration early in the dis- peripheral sensory neurons [71], and increasing neuron ease, but low in the MNs more vulnerable to degeneration survival [75]. Midkine secreted by astrocytes is a known in ALS [69]. Conversely, the vulnerable MNs are high in neurotrophic factor promoting neurite outgrowth and high low matrix metalloproteinase MMP-9 (MMP9 /OPN ), neuron survival (review in [76]). The multiple nature of whereas MMP-9 is low and OPN is high in the the factors secreted by the hES-AS supports a mode of ALS-resistant MNs [69, 70]. Exogenous addition of OPN action much more diversified than merely through the has neurogenic effects, stimulating regeneration of motor classical neurotrophic factors. axons [71] and protecting neurons after ischemia in vitro The efficacy of hES-AS to delay disease onset and to and in vivo [72]. Although MMP9 was not detected in the ameliorate disease progression was evaluated in trans- G93A secretome of our astrocyte cultures, inhibitors of MMP9 genic high copy number SOD1 mouse and rat and other matrix metalloproteases were abundantly se- models, which recapitulate many of the clinical symp- creted, particularly the tissue inhibitors of metallopro- toms of the ALS disease in humans [18, 51, 77]. Intra- teases TIMP-1 and TIMP-2, which play a major role in thecal injection of hES-AS significantly delayed the preventing degradation of ECM components by MMPs or onset of the disease and slowed down the deterioration regulating ECM remodeling (review in [73]). Another of motor function. These effects were more pronounced chemokine found in the secretome is MIF, which has the when the cells were administered twice (3–4 weeks Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 10 of 17 b e Fig. 6 hES-AS distribute throughout CNS after intrathecal injection. hES-AS (400,000 cells) differentiated for 7 days transplanted intrathecally into NSG mice (into CSF through CM). a Illustration of brain and spinal cord sections performed: seven brain sections (L#1–L#7 as in [64]) and four of representative regions of spinal cord. b–d Graphical representation of AstroRx cell presence (as determined by Alu cell staining) and percent incidence of frequency scores ≥ 2 (one to three foci of 10–20 cells per foci) after 4-week (b), 17-week (c) and 39-week (d) follow up. AstroRx Cell presence calculated as incidence (%) from all samples (n) within each group. Frequency of score ≥ 2 calculated as incidence (%) of frequency scores ≥ 2 from only those sections in which AstroRx cells present. e–g Representative images of different sections demonstrating distribution of hES-AS throughout CNS using ISH with and Alu Y probe (human specific) of 17-week cohort. e Sacral region of spinal cord with numerous Alu cells (arrows) along surface of the spinal nerves (asterisks). f Brain, level 5. Arrows indicate cells along meningeal surface at many locations. g Brain, level 6. Arrows indicate Alu cells along meningeal surface along base of medulla at brain level 6. Cells attached to the pia mater (arrows). hES-AS human embryonic stem cell-derived astrocytes (differentiated from APCs for 7 days) apart) than with a single injection. Intrathecal injection A major safety concern associated with pluripotent stem into the CSF is in line with the proposed mode of cell-based therapies is the presence of residual undifferen- action, in which the healthy astrocytes would work at a tiated stem cells that might continue to divide without distance to modify the environment of brain and spinal control or develop teratoma after their transplantation in cord MNs. Indeed, the CSF composition shows several the body [82, 83]. We minimize the possibility of teratoma changes in the course of ALS [78, 79], including an formation by assuring a complete differentiation of hESCs increase in oxidative stress markers, an increase in into committed astrocytes with a normal diploid karyo- glutamate in at least 40% of patients and variations of type and minimal proliferation capacity. Teratoma forma- VEGF concentration correlating with the length of tion from undifferentiated hESCs depends on several survival [80], and other changes including OPN in- factors, among them the site of implantation and number crease [81]. Moreover, the fact that inoculation of CSF of transplanted cells. Several studies reported that undif- from ALS patients to animals is neurotoxic [63]dem- ferentiated hESCs develop teratomas within 6 weeks after onstrates that materials injected into the CSF can transplantation in immunodeficient mice [47, 82, 84, 85]. affect the parenchyma. We previously reported that injection of undifferentiated Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 11 of 17 b c Fig. 7 hES-AS are post mitotic and maintain their astrocytic identity in vivo. a–c High-content analysis of hES-AS cells in vitro displayed homogeneous + + + expression of human GFAP (Stem123). %Ki67 cells calculated as % Ki67 nuclei / total number of nuclei. Ki67 cells rarely found within hES-AS cell population (arrows). d–f Two million hES-AS injected intrathecally into the lumbar region twice, with interval of 21 days. Analysis of graft, 8 weeks post first cell injection, showed transplanted cells were located in subarachnoid space, attached to pia mater of lumbar spinal cord and nerve bundles. Cells + + maintained their astrocytic characters and homogeneously expressed human-origin GFAP. %Ki67 hES-AS cells calculated as % Ki67 nuclei / total + + number of nuclei of Stem123 cells. Ki67 staining very rare among hES-AS cells (arrows), indicating that cells are non-proliferative in vivo. hES-AS human embryonic stem cell-derived astrocytes (differentiated from APCs for 7 days), DAPI 4′,6-diamidino-2-phenylindole hESCs intrathecally into immunodeficient mice results in maintained stable over time, supporting that the cells re- teratoma formation within 5–7 weeks after injection [86]. main quiescent in the CSF. The effective biodistribution In our current study, we evaluated the formation of of hES-AS along the entire CSF supports the clinical G93A teratomas, or any other tumor, by hES-AS up to 39 weeks benefits we observed in SOD1 models. We found an after their intrathecal injection, long enough to allow attenuation in motor activity loss in both lower and development of teratomas. Histology evaluation showed upper limbs and the tail, indicating that the cells exert the cells survived in the CSF for the entire duration of the their activity on multiple regions of the CNS. The pos- study, attached to the pia mater along the neuroaxis, The sible migration of cells to distant organs was evaluated cells uniformly expressed astrocytic markers with very rare by qPCR for amplification of the Alu Y genomic se- coexpression of the cell cycle marker Ki67. Importantly, quence in nine organs. hES-AS were not found in any hES-AS did not develop teratoma or any other tumors in distant organ above the detection limit of the method (1 any of the treated mice. In line with these results, Priest et cell) at 4 and 17 weeks after their intrathecal injection. al. [87] also reported the absence of teratomas in the CNS This confined distribution of the cells to the CNS mini- following intraspinal injection of oligodendrocyte progeni- mizes any possible risk of presence of ectopic glial tissue tors derived from hESCs into the spinal cord of immuno- in nontarget organs outside the CNS. deficient rats. Large quantities of human astrocytes would be needed To access the CNS, we chose the CSF as the injection for the treatment of ALS patients worldwide. As shown site for hES-AS. The circulating CSF helps to distribute here, clinical-grade human ESCs provide a robust and the injected cells throughout the subarachnoid space. In controlled source of cells for mass production of glial addition, injection into the CSF by LP is a common progenitors that can give rise to functional astrocytes. To low-risk medical practice already demonstrated in comply with GMP standards, we adjusted our previous several clinical trials with cell-based therapies [88–91]. A protocol, originally aimed to produce both astrocytes and biodistribution evaluation of hES-AS in the CNS was oligodendrocytes [41], to include only GMP-grade mate- performed by in-situ hybridization of the Alu Y gene at rials. Under this protocol, large amounts of astrocyte pro- 4, 17 or 39 weeks following a single intrathecal injection genitor cells (APCs) are obtained, which can be frozen in of cells into immunodeficient mice. The analysis re- liquid nitrogen for long-term storage [41] as master and vealed the presence of hES-AS cells in the subarachnoid working cell banks for future expansion. Upon thawing of space throughout the entire CNS. Cell numbers were the APC vial, the differentiation into hES-AS is completed Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 12 of 17 within 7 days of culturing. In terms of yield, using our undifferentiated state of the hESCs was routinely protocol we can produce a total of 2 × 10 hES-AS from assessed by flow cytometry analysis of the surface a single batch of hESCs. Hence, the process described here markers SSEA-4, EpCAM and TRA-1-60, and by im- is suitable for mass production of clinical-grade hES-AS munofluorescence staining for the transcription factors per batch, which can potentially treat thousands of pa- NANOG and OCT4. Both lines were propagated in tients [92, 93]. undifferentiated state on a HFF feeder layer (25,000 In recent years, clinical trials of cell therapy in ALS cells/cm ) by passaging every 6–7 days using collage- have mainly used autologous transplantation of mesen- nase in order to detach the whole hESC colonies from chymal stem or stromal cells (MSCs) [89, 94], in which the feeder cell layers. The colonies were mechanically cells are taken from the patients and after in-vitro broken and seeded in a ratio of 1:3–6. The hESCs were culture are returned to the same patient. While giving grown in ES1 media composed of KO-DMEM, 14% (v/v) promising clinical efficacy, these autologous transplan- KO serum replacement, 2 mM glutamine, 1× MEM non- tations have limitations and it would be advantageous essential amino acids, 0.1 mM β-mercaptoethanol and to develop allogeneic cells as a shelf-product that would 25 U/ml penicillin, 25 μg/ml streptomycin (all from Life provide a treatment for all ALS patients. Given that Technologies) and 8 ng/ml bFGF (R&D). Important to intrathecal administration is effective (as seen with the note is that for generation of clinical-grade hESCs, the MSCs), it would be easier than injections in the spinal cells were adapted to feeder free conditions and the media cord anterior horn, which requires major surgery as composition was changed to Essential 8™ (E8) medium done in recent ALS clinical trials with neural stem cells (Thermo Fischer Scientific). taken from human organ donors [95, 96]. Future Formation of neurospheres (NS) was done in suspension clinical trials could use human pluripotent stem cell (3D) cultures. In brief, the harvested hESC colonies were cultures for mass production of neural cells, either transferred into 100-mm ultralow attachment culture plates from human iPSCs [97, 98] or from human ES cell lines (Corning) containing ITTSPP/B27 medium. ITTSPP/B27 is as described here. a mixture of DMEM/F12 containing 1% B27 supplement, 1% Glutamax, 1.5% Hepes at pH 7.4 (all from Thermo Sci- Conclusions entific), 1% penicillin/streptomycin/amphotericin solution Here we describe the derivation of a highly enriched (Biological Industries), 25 μg/ml human insulin (ActRapid; population of functional, clinical-grade, human astrocytes Novo Nordisk), 50 μg/ml human Apo-transferrin (Athens), (hES-AS) from embryonic stem cells. The hES-AS were 6.3 ng/ml progesterone, 10 μg/ml putrescine, 50 ng/ml shown to protect MNs by multiple mechanisms, similarly sodium selenite and 40 ng/ml triiodothyronine (T3) (all to normal astrocytes, including clearance of glutamate, from Sigma). ITTSPP/B27 was supplemented with 20 ng/ secretion of multiple NTFs, neutralization of ROS and ml r-human EGF (R&D Systems). After 2 days, the medium promotion of neural outgrowth. Intrathecal injection of was switched to ITTSPP/B27 supplemented with 20 ng/ml hES-AS to rodent models of ALS delays disease onset, EGF and 10 μM ATRA (Sigma). The culture was continued slows down disease progression and extends life expect- in suspension in the nonadherent plates for 7 days with ancy. A 9-month safety study conducted in an immunode- daily replacement of the medium (stage 2; Fig. 1). During ficient NSG animal model, under GLP conditions, showed the last step, which allows for NS ripening, the culture was that intrathecal transplantation of hES-AS cells to the continued in ITTSPP/B27 medium supplemented with cerebrospinal fluid (CSF) is safe. Thus, these findings 20 ng/ml EGF for 18 days. Medium was replaced every demonstrate the feasibility, safety and potential efficacy of other day (stage 3; Fig. 1). For APC expansion, round yel- intrathecal injections of hES-AS for the treatment of ALS. low NS were manually selected using a stereoscopic micro- The safety and efficacy of hES-AS treatment in ALS scope and transferred into six-well plates coated with patients will be tested in a phase I/IIa clinical trial (Clini- Matrigel or GMP-compliant laminin 521 (from Biolamina) calTrials.gov identifier: NCT03482050). in ITTSPP/B27 supplemented with 20 ng/ml EGF. Medium was replaced every other day for 7–10 days (passage 0). In Methods order to produce a monolayer, the spheres were dissociated Derivation of astrocyte progenitor cells and committed with TryplE (Thermo Scientific) and reseeded on ECM astrocytes from hESCs (passage 1) in N2/B27 medium consisting of DMEM/F12 Two clinical-grade hESC lines, were used: NCL14, licensed with 0.5% (v/v) N supplement, 1% (v/v) B27 supplement, from the University of Newcastle; and HADC100, 1% Glutamax and 1.5% Hepes at pH 7.4 (all from Thermo obtained from the Hadassah Medical Organization Scientific). The growth factors EGF and bFGF (R&D (HMO), Jerusalem (Prof. Benjamin Reubinoff). Master Systems) were added at 10 ng/ml each. The monolayer cells cell banks (MCB) and working cell banks (WCB) of were further passaged weekly until a sufficient number of these hESCs were created at Kadimastem Ltd. The cells was generated. Cells were then frozen in liquid Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 13 of 17 nitrogen andstoredas banks of APCs.ThawedAPCswere Department of Life Sciences Core Facilities, Weizmann further expanded as described earlier for 2–3 weeks. In Institute of Science. order to differentiate the APCs toward astrocytes, EGF and bFGF were removed from the media, 50 μg/ml ascorbic Flow cytometry acid (Sigma) was added and the culture was continued for Cells were analyzed by flow cytometry for identity and pur- 7or28days. ity markers using the following antibodies: anti-A2B5 (1:20; Miltenibiotec), anti-GLAST (1:20; Miltenibiotec), anti-CD44 (1:20; BD Pharmingen), anti-CXCR4 (1:20; Immunocytofluorescence assays Biolegend), anti-TRA-1-60 (1:50; Biolegend), anti-EPCAM Cells were fixed with 4% paraformaldehyde (PFA), washed (1:50; Biolegend), anti-SSEA4 (1:50; Biolegend), anti-GFAP with PBS and kept at 4 °C before staining. Permeabilization (1:2000; Sigma), Nestin (1:500; BD Pharmingen) and was done by 0.5% Triton X-100 in Blocking solution (5% anti-AQP-4 (1:2000; Abcam). The Flow Cytometer FACS BSA; Sigma) and 3% horse serum (w/v in PBS; Biological Canto II (BD) was operated with FACSDIVA software Industries). Incubation in the same blocking solution was (BD). At least 10,000 events were collected per sample. done for 1 h at RT. Primary antibodies, diluted in blocking solution, were as follows: anti-Nanog, anti-Nestin (1:500; Glutamate uptake assay BD Pharmingen), anti-GFAP-cy3 (mouse monoclonal anti- Glutamate uptake capability of the cells was measured in body (Mc), 1:500; Sigma), anti-GLAST (rabbit Mc, 1:100; 28-day differentiated hESC-derived astrocytes. Glutamic Miltenibiotec), anti-S100 (rabbit polyclonal antibody, 1:100; acid (0.5 mM; Sigma) in Hanks’ Balanced Salt Solution DAKO), anti-AQP-4 (rabbit, 1:2000; Mc Abcam) and (Gibco) was added to 1 × 10 cells/ml. After 0, 10, 30, 60 anti-Ki67 (rabbit, 1:50; Mc Cell Marque). After overnight and 120 min, the solution was aspirated and kept at 4 °C incubation at 4 °C, secondary antibody (1:200; Jackson until further testing. Human astrocytes derived from the Immuno Research) was added for 1 h at RT, followed spinal cord (from Thermo Scientific) served as positive by the nuclear fluorescent dye DAPI (0.05 μg/ml; control, while 0.5 mM glutamic acid kept at 37 °C for Sigma). Pictures were taken using Arrayscan VTI 120 min served as negative control. In addition, 0.5 mM (Thermo Scientific, Cellomics). glutamic acid kept at 4 °C for 120 min served as time 0 concentration control. The EnzyChrom™ Glutamate Assay Immunohistochemical staining Kit (BioAssay Systems) was used to measure the concen- Brain and spinal cord tissues were trimmed, decalcified tration of glutamate in the collected samples according to and embedded in paraffin, sectioned at approximately the manufacturer’s protocol and recommendations. The 5 μm thickness and stained with hematoxylin and eosin optical density was read at 565 nm using the iMark (H&E). For immune-cytofluorescence assays, tissues were Microplate reader (Bio Rad). Dihydrokainic acid (DHK, deparaffinized using the following washes: xylene (Sigma), 500 μM; Sigma) or 1 μM WAY-213,613 (Sigma) were used two washes × 5 min; 100% ethanol, two washes × 5 min; as inhibitors of GLT-1. 95% ethanol, one wash × 5 min; 70% ethanol, one wash × 5 min; and cold tap water, two washes × 5 min. Secretome analysis Heat-induced epitope retrieval was performed by boiling In order to promote astrocyte differentiation, APCs the sections in a domestic microwave, twice for 10 min, were deprived from growth factors (bFGF and EGF) using 100× H-3300 citrate-based solution (Vector Labora- and vitamin C was added for 7 days and 28 days. tories). Permeabilization was done by 0.5% Triton X-100 Conditioned media were collected after 48 h from each in blocking solution as described earlier, and incubation experimental well. The number of cells for each well continued in the same blocking solution for 1 h at RT. was counted (at least two replicas per each cell type) Primary mouse Mc antibody Stem123 or Stem121 (1:500; and secretome analysis was performed by multiplex Stem Cells) were added overnight and kept at 4 °C. ELISA using the human quantibody kiloplex Array Secondary antibody goat anti mouse Cy2 or Cy3 (1:200; (Raybiotech). The values obtained in secretome analysis Jackson Immuno Research) were added for 1 h at RT, 6 were normalized to 1 × 10 cells/ml. followed by the nuclear fluorescent dye DAPI (0.05 μg/ml; Sigma). G93A Transplantation of hES-AS in the hSOD1 animal model G93A Karyotype Transgenic hSOD1 mice aged 8–9 weeks of mixed The test was performed using spectral karyotyping gender (B6SJL-Tg(SOD1*G93A)1Gur/J) were purchased analysis (SKY) on cells from two APC banks (passages from The Jackson Laboratory (Bar Harbor, ME, USA; G93A 11 and 12). The analysis was performed by the Stem https://www.jax.org/). Transgenic hSOD1 rats aged G93A Cell Core and Advanced Cell Technologies Unit, 5–6 weeks of mixed gender (NTac:SD-Tg(SOD1 )L26H) Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 14 of 17 were purchased from Taconic Biosciences Inc. (Hudson, and lasting all throughout the duration of the experiment. NY, USA; http://www.taconic.com). Allanimalcareand CellCept was administered orally twice a day at a dose of surgical procedures described here were carried out accord- 15 mg/kg (total daily dose was 30 mg/kg). Dosing started ing to protocols approved by the Israeli National Commit- 3 days prior to the treatment and lasted for a total of 10 tee for Animal Care. The animals were kept in a certified consecutive days. Cohort 3, which was given the treatment animal facility in IVC cages with a light cycle of 12 h and at twice, started receiving CellCept 3 days prior to each treat- temperature of 22 ± 2 °C. Rodent diet and drinking water ment injection for 10 consecutive days. were provided ad libitum. Measurements Intrathecal injection through the cisterna magna Measurement of body weight and all motor tests took Mice were anesthetized with an i.p. injection of keta- place 7–10 days prior to cell transplantation and rou- mine/xylazine (K4138; Sigma) and then mounted on a tinely afterward. Motor function was tested using an stereotaxic frame. The head was then bent, resulting in acceleration Rotarod device (Rotarod 7650; Ugo Basile, nape distention. A midline skin incision was made at the Comerio, Italy) for the duration of 180 s. The time it nape area to expose the sagittal suture of the cranium took each mouse to fall from the rod was recorded. and midline of the nape. Under a dissection microscope, Animals were trained for 1 week prior to conducting the the subcutaneous tissue and muscles were separated by test. Forelimb muscle grip strength was determined blunt dissection with forceps to expose the cleft between using a Grip Strength Meter 47,200 (UGO Basile). Grip the occipital bone and the atlas vertebra. The muscles strength testing was performed by allowing the animals were held apart to expose the dura mater which was to grasp a thin bar attached to the force gauge. This is carefully penetrated using a 29G-gauge 45° beveled done by pulling the animal away from the gauge until needle (Hamilton, Reno, NV, USA) connected to a 10-μl the mice forelimbs released the bar. The procedure Hamilton syringe preloaded with 10 μl of cell suspension provides a value of the force of maximal grip strength. or vehicle (DMEM/F12 medium). Then 2 × 10 hES-AS The force measurements were recorded in three separate (APCs differentiated for 7 days) were injected once on trials, and the averages were used in the statistical day 67 ± 2 (CellsX1 group, n = 14 mice) or twice on day analysis. Neurological scoring was done according to 67 ± 2 and on day 97 ± 2 at interval of 30 days (CellsX2 neurological score on a scale from 0 to 5 [99]. group, n = 13), or injected with DMEM F12 (Sham group, n = 10) into the CSF through the CM. The Statistical analysis G93A syringe was held in position for 3 min before being grad- Kaplan–Meier analysis of the SOD1 mice and rats ually pulled away to avoid liquid outflow along the was conducted using the statistical software Sigmastat needle tract. The skin cut was secured with stainless (SAS Software) to analyze survival, disease onset and dur- steel surgical clips and wiped with 70% ethanol. ation data. Weight, time to fall from the Rotarod, neuro- logical score and grip strength results were analyzed via Injection of the cells by lumbar puncture repeated-measures ANOVA. All data are presented as The rats were anesthetized with ketamine/xylazine. The mean ± SEM, and significance level was set at p ≤ 0.05. lumbar region was shaved, sterilized with iodine and the Statistical analysis was performed by MediStat Ltd, Israel. intervertebral spaces widened by placing the animal on a 15-ml conical plastic tube. The injections were per- Transplantation of hES-AS in NSG mice formed by inserting a 29-gauge 45° beveled needle The mouse was mounted on a stereotaxic frame. A mid- (Hamilton) connected to a 10-μl Hamilton syringe into line skin incision was made at the nape area to expose the the tissues between the dorsal aspects of L5 and L6. sagittal suture of the cranium and midline of the nape. Correct subarachnoid positioning of the tip of the needle The head was then bent, resulting in nape distention. was verified by a tail flick test. A volume of 10 μl Under a dissection microscope, the subcutaneous tissue containing 3 × 10 APCs was injected twice on day 50 ± and muscles were separated by blunt dissection with for- 2 and on day 70 ± 2 (n = 7), or vehicle (DMEM/12 ceps to expose the cleft between the occipital bone and medium, n = 7) was injected. The syringe was held in the atlas vertebra. The muscles were held apart to expose position for 30 s before being progressively pulled away. the dura mater which was penetrated using a 29G needle connected to a Hamilton syringe, preloaded with 10 μlof Immunosuppression 0.4 × 10 hES-AS. The cells were injected within 30 s into Immunosuppression was used only in the transplantation the CSF space. The needle was held for about 30 s G93A experiment in SOD1 mice. In this experiment, Cyclo- after injection and then withdrawn. The skin cut was sporin A was given daily by intraperitoneal injection, at a secured with stainless steel surgical clips and wiped dose of 10 mg/kg, starting 3 days prior to the treatment with polydine solution. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 15 of 17 Additional files Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Additional file 1: Figure S1. hES-AS produce and secrete neurotrophic factors. Conditioned media of 24 h from cultures of hES-AS differentiated Author details for 28 days as well as cell extracts used to measure level of neurotrophic Neurodegenerative Diseases Department at Kadimastem Ltd, Pinchas Sapir factors GDNF, BDNF, VEGF and IGF-1. For each factor, bars show cell content, 7, Weizmann Science Park, Nes-Ziona, Israel. Department of Molecular amount secreted and negative control (medium only), expressed in pg/10 Genetics, Weizmann Institute of Science, 76100 Rehovot, Israel. cells (triplicates ± SD) (PDF 91 kb) Additional file 2: Table S1. Secretome analysis of hES-AS, differentiated for Received: 5 April 2018 Revised: 30 April 2018 7 days or 28 days. The 220 most secreted factors by the 7-day differentiated Accepted: 1 May 2018 hES-AS sorted by mean ng/ml/10 cells ± SD (PDF 140 kb) Additional file 3: Figure S2. Effect of hES-AS transplantation on disease References G93A onset, progression and survival in hSOD1 mice. hES-AS, differentiated 1. Rowland LP, Shneider NA. Amyotrophic lateral sclerosis. N Engl J Med. 2001; G93A for 7 days, transplanted intrathecally through CM of hSOD1 mice. A 344:1688–700. Three experimental groups tested, single injection of 2 × 10 hES-AS on 2. Rosen DR. Mutations in Cu/Zn superoxide dismutase gene are associated day 67 of life (Cellsx1), two injections of 2 × 10 hES-AS each on days 67 with familial amyotrophic lateral sclerosis. Nature. 1993;364:362. and 97 (Cellsx2) and once sham-injected mice (vehicle). Kaplan–Meir plot 3. Lagier-Tourenne C, Cleveland DW. Rethinking ALS: the FUS about TDP-43. of disease onset (measured by 3% body weight loss from maximal weight) Cell. 2009;136:1001–4. showing more delay in twice-injected group. B Kaplan–Meier survival curves 4. Renton AE, Chio A, Traynor BJ. State of play in amyotrophic lateral sclerosis with similar trends. C Body weight maintained longer in hES-AS-treated mice. genetics. Nat Neurosci. 2014;17:17–23. Note that a few days after second injection, day 97, weight loss occurred 5. Haidet-Phillips AM, et al. Astrocytes from familial and sporadic ALS patients related to injection. D Neurological score. E Significant improvement in motor are toxic to motor neurons. Nat Biotechnol. 2011;29:824–8. performance (Rotarod test) for hSOD1 mice transplanted twice with hES-AS. 6. Meyer K, et al. Direct conversion of patient fibroblasts demonstrates non- C, D Values are mean ± SEM (PDF 262 kb) cell autonomous toxicity of astrocytes to motor neurons in familial and Additional file 4: Table S2. Percent of cell presence and percent of sporadic ALS. Proc Natl Acad Sci U S A. 2014;111:829–32. frequency scores greater than, or equal to ‘2’ (one to three foci of 10-20 7. Di Giorgio FP, Carrasco MA, Siao MC, Maniatis T, Eggan K. Non-cell cells per foci) for each follow up time (4, 17 and 39 weeks after hES-AS autonomous effect of glia on motor neurons in an embryonic stem cell- transplantation). Supplementary materials and methods. (ZIP 150 kb) based ALS model. Nat Neurosci. 2007;10:608–14. 8. Nagai M, et al. Astrocytes expressing ALS-linked mutated SOD1 release factors selectively toxic to motor neurons. Nat Neurosci. 2007;10:615–22. Abbreviations 9. Di Giorgio FP, Boulting GL, Bobrowicz S, Eggan KC. Human embryonic stem ALS: Amyotrophic lateral sclerosis; APC: Astrocyte progenitor cell; AQP- cell-derived motor neurons are sensitive to the toxic effect of glial cells 4: Aquaporin-4; CM: Cisterna magna; CNS: Central nervous system; carrying an ALS-causing mutation. Cell Stem Cell. 2008;3:637–48. CSF: Cerebrospinal fluid; G93A mutation: Glycine 93 changed to alanine; 10. Marchetto MC, et al. Non-cell-autonomous effect of human SOD1 G37R GFAP: Glial Fibrillary Acidic Protein; GLP: Good laboratory practice; GLT- astrocytes on motor neurons derived from human embryonic stem cells. 1: Glutamate transporter 1; GMP: Good manufacturing practice; hES- Cell Stem Cell. 2008;3:649–57. AS: Human embryonic stem cell-derived astrocytes (differentiated from APCs 11. Papadeas ST, Kraig SE, O'Banion C, Lepore AC, Maragakis NJ. Astrocytes for 7 days); hESC: Human embryonic stem cell; hSOD1: Human superoxide carrying the superoxide dismutase 1 (SOD1G93A) mutation induce wild- dismutase 1; LOD: Limit of detection; LP: Lumbar puncture; MN: Motor type motor neuron degeneration in vivo. Proc Natl Acad Sci U S A. 2011; neuron; NTF: Neurotrophic factor; SOD1: Superoxide dismutase 1; 108:17803–8. TIMP: Tissue inhibitor of metalloproteinase; VEGF: Vascular endothelial 12. Yamanaka K, et al. Mutant SOD1 in cell types other than motor neurons and growth factor oligodendrocytes accelerates onset of disease in ALS mice. Proc Natl Acad Sci U S A. 2008;105:7594–9. Acknowledgements 13. Tong J, et al. Expression of ALS-linked TDP-43 mutant in astrocytes causes The authors would like to thank Science in Action and Envigo-IL animal non-cell-autonomous motor neuron death in rats. EMBO J. 2013;32:1917–26. facilities for conducting animal studies. 14. Rossi S, Cozzolino M, Carri MT. Old versus new mechanisms in the pathogenesis of ALS. Brain Pathol. 2016;26:276–86. Funding 15. Foran E, Trotti D. Glutamate transporters and the excitotoxic path to motor This work was supported by the Israel Innovation Authority National Natural neuron degeneration in amyotrophic lateral sclerosis. Antioxid Redox Signal. (grant No. 59652). 2009;11:1587–602. 16. Rothstein JD, Van Kammen M, Levey AI, Martin LJ, Kuncl RW. Selective loss Availability of data and materials of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann The data that support the findings of this study are available from the Neurol. 1995;38:73–84. corresponding author upon reasonable request. 17. Lin CL, et al. Aberrant RNA processing in a neurodegenerative disease: the cause for absent EAAT2, a glutamate transporter, in amyotrophic lateral Authors’ contributions sclerosis. Neuron. 1998;20:589–602. RM, IM, SSG, CJ and HA, JI-E conceived and designed the studies. TA, LC, GA, 18. Howland DS, et al. Focal loss of the glutamate transporter EAAT2 in a KPL, LN, SYI, SJL, ZR, ZA and EV performed the experiments. KG was respon- transgenic rat model of SOD1 mutant-mediated amyotrophic lateral sible for quality assurance. IM and SSG analyzed the data. IM, SSG, RM and CJ sclerosis (ALS). Proc Natl Acad Sci U S A. 2002;99:1604–9. wrote the manuscript. SSG and IM interpreted the data and reviewed the 19. Raoul C, et al. Motoneuron death triggered by a specific pathway manuscript. All authors read and approved the final manuscript. downstream of Fas. potentiation by ALS-linked SOD1 mutations. Neuron. 2002;35:1067–83. Ethics approval and consent to participate 20. Barbeito LH, et al. A role for astrocytes in motor neuron loss in amyotrophic This study was performed following an application-form review by the lateral sclerosis. Brain Res Brain Res Rev. 2004;47:263–74. National Council for Animal Experimentation and after receiving approval 21. Endo F, et al. Astrocyte-derived TGF-β1 accelerates disease progression in (No. IL-16-09-279) that the study complies with the rules and regulations ALS mice by interfering with the neuroprotective functions of microglia and set forth. T cells. Cell Rep. 2015;11:592–604. 22. Tripathi P, et al. Reactive astrocytes promote ALS-like degeneration and Competing interests intracellular protein aggregation in human motor neurons by disrupting The authors declare that they have no competing interests. autophagy through TGF-β1. Stem cell reports. 2017;9:667–80. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 16 of 17 23. Phatnani, H.P., et al. Intricate interplay between astrocytes and motor and serum of patients with early Alzheimer's disease and normal controls. J neurons in ALS. Proc Natl Acad Sci U S A 110, E756–EE765 (2013). Alzheimers Dis. 2009;18:331–7. 24. Brambilla L, et al. Disruption of the astrocytic TNFR1-GDNF axis accelerates 50. Li G, et al. Cerebrospinal fluid concentration of brain-derived neurotrophic factor motor neuron degeneration and disease progression in amyotrophic lateral and cognitive function in non-demented subjects. PLoS One. 2009;4:e5424. sclerosis. Hum Mol Genet. 2016;25:3080–95. 51. Gurney ME, et al. Motor neuron degeneration in mice that express a human 25. Das MM, Svendsen CN. Astrocytes show reduced support of motor neurons Cu, Zn superoxide dismutase mutation. Science. 1994;264:1772–5. with aging that is accelerated in a rodent model of ALS. Neurobiol Aging. 52. Writing G, Edaravone ALSSG. Safety and efficacy of edaravone in well 2015;36:1130–9. defined patients with amyotrophic lateral sclerosis: a randomised, double- 26. Van Den Bosch L, et al. Effects of vascular endothelial growth factor (VEGF) blind, placebo-controlled trial. The Lancet Neurology. 2017;16:505–12. on motor neuron degeneration. Neurobiol Dis. 2004;17:21–8. 53. Bensimon G, Lacomblez L, Meininger V. A controlled trial of riluzole in 27. Oosthuyse B, et al. Deletion of the hypoxia-response element in the amyotrophic lateral sclerosis. ALS/Riluzole Study Group. N Engl J Med. 1994; vascular endothelial growth factor promoter causes motor neuron 330:585–91. degeneration. Nat Genet. 2001;28:131–8. 54. Lacomblez L, Bensimon G, Leigh PN, Guillet P, Meininger V. Dose-ranging 28. Bogaert E, et al. VEGF protects motor neurons against excitotoxicity by study of riluzole in amyotrophic lateral sclerosis. Amyotrophic Lateral upregulation of GluR2. Neurobiol Aging. 2010;31:2185–91. Sclerosis/Riluzole Study Group II. Lancet. 1996;347:1425–31. 29. Gros-Louis F, et al. Absence of mutations in the hypoxia response element 55. Miller RG, Mitchell JD, Moore DH. Riluzole for amyotrophic lateral sclerosis of VEGF in ALS. Muscle Nerve. 2003;28:774–5. (ALS)/motor neuron disease (MND). The Cochrane database of systematic reviews. 2012;14(3):CD001447. 30. Llado J, Tolosa L, Olmos G. Cellular and molecular mechanisms involved in the neuroprotective effects of VEGF on motoneurons. Front Cell Neurosci. 2013;7:181. 56. Petrov D, Mansfield C, Moussy A, Hermine O. ALS clinical trials review: 20 31. Mishra PS, et al. Astroglia acquires a toxic neuroinflammatory role in years of failure. are we any closer to registering a new treatment? Front response to the cerebrospinal fluid from amyotrophic lateral sclerosis Aging Neurosci. 2017;9:68. patients. J Neuroinflammation. 2016;13:212. 57. Izrael M, Slutsky S, Itskovitz-Eldor J. & Revel M. Astrocytes in Pathogenesis of 32. Re DB, et al. Necroptosis drives motor neuron death in models of both Neurodegenerative Diseases and Potential: Translation into Clinic. InTechOpen: sporadic and familial ALS. Neuron. 2014;81:1001–8. Astrocyte—Physiology and Pathology, ISBN 978–953–51-5760-1 (Mar 2018). 33. Hitomi J, et al. Identification of a molecular signaling network that regulates 58. Lee J, et al. Astrocytes and microglia as non-cell autonomous players in the a cellular necrotic cell death pathway. Cell. 2008;135:1311–23. pathogenesis of ALS. Experimental neurobiology. 2016;25:233–40. 59. Do-Ha D, Buskila Y, Ooi L. Impairments in motor neurons, interneurons and 34. Wiedemann FR, Manfredi G, Mawrin C, Beal MF, Schon EA. Mitochondrial astrocytes contribute to hyperexcitability in ALS: underlying mechanisms DNA and respiratory chain function in spinal cords of ALS patients. J Neurochem. 2002;80:616–25. and paths to therapy. Mol Neurobiol. 2017;55(2):1410–18. 35. Hirano A, et al. Fine structural study of neurofibrillary changes in a 60. Storkebaum E, Lambrechts D, Carmeliet P. VEGF: once regarded as a specific family with amyotrophic lateral sclerosis. J Neuropathol Exp Neurol. angiogenic factor, now implicated in neuroprotection. BioEssays. 2004;26:943–54. 1984;43:471–80. 61. Krakora D, et al. Synergistic effects of GDNF and VEGF on lifespan and 36. Sasaki S, Horie Y, Iwata M. Mitochondrial alterations in dorsal root ganglion cells disease progression in a familial ALS rat model. Molecular therapy : the in sporadic amyotrophic lateral sclerosis. Acta Neuropathol. 2007;114:633–9. journal of the American Society of Gene Therapy. 2013;21(8):1602–10. 37. Tan W, et al. Small peptides against the mutant SOD1/Bcl-2 toxic 62. Ramamohan PY, et al. Cerebrospinal fluid from amyotrophic lateral sclerosis mitochondrial complex restore mitochondrial function and cell viability in patients causes fragmentation of the Golgi apparatus in the neonatal rat spinal mutant SOD1-mediated ALS. The Journal of neuroscience : the official cord. Amyotrophic lateral sclerosis : official publication of the World Federation journal of the Society for Neuroscience. 2013;33:11588–98. of Neurology Research Group on Motor Neuron Diseases. 2007;8:79–82. 38. Nicaise C, Mitrecic D, Falnikar A, Lepore AC. Transplantation of stem cell- 63. Deepa P, et al. Down regulation of trophic factors in neonatal rat spinal derived astrocytes for the treatment of amyotrophic lateral sclerosis and cord after administration of cerebrospinal fluid from sporadic amyotrophic spinal cord injury. World J Stem Cells. 2015;7:380–98. lateral sclerosis patients. J Neural Transm. 2011;118:531–8. 39. Tannenbaum SE, et al. Derivation of xeno-free and GMP-grade human 64. Shruthi S, et al. Brain-derived neurotrophic factor facilitates functional embryonic stem cells–platforms for future clinical applications. PLoS One. recovery from ALS-cerebral spinal fluid-induced neurodegenerative changes 2012;7:e35325. in the NSC-34 motor neuron cell line. Neurodegener Dis. 2017;17:44–58. 40. Prathalingam N, et al. Production and validation of a good manufacturing 65. Rosito M, Deflorio C, Limatola C, Trettel F. CXCL16 orchestrates adenosine practice grade human fibroblast line for supporting human embryonic stem A3 receptor and MCP-1/CCL2 activity to protect neurons from excitotoxic cell derivation and culture. Stem Cell Res Ther. 2012;3:12. cell death in the CNS. The Journal of neuroscience : the official journal of 41. Izrael M, et al. Human oligodendrocytes derived from embryonic stem cells: the Society for Neuroscience. 2012;32:3154–63. effect of noggin on phenotypic differentiation in vitro and on myelination 66. Christopherson KS, et al. Thrombospondins are astrocyte-secreted proteins in vivo. Mol Cell Neurosci. 2007;34:310–23. that promote CNS synaptogenesis. Cell. 2005;120:421–33. 42. Scholze AR, Foo LC, Mulinyawe S, Barres BA. BMP signaling in astrocytes 67. Tyzack GE, et al. Astrocyte response to motor neuron injury promotes structural downregulates EGFR to modulate survival and maturation. PLoS One. 2014; synaptic plasticity via STAT3-regulated TSP-1 expression. Nat Commun. 2014;5:4294. 9:e110668. 68. Hafner A, et al. Neuroprotective role of gamma-enolase in microglia in a 43. Alfei L, et al. Hyaluronate receptor CD44 is expressed by astrocytes in the mouse model of Alzheimer's disease is regulated by cathepsin X. Aging Cell. 2013;12:604–14. adult chicken and in astrocyte cell precursors in early development of the chick spinal cord. Eur J Histochem. 1999;43:29–38. 69. Morisaki Y, et al. Selective expression of Osteopontin in ALS-resistant motor 44. Piao C, et al. Thrombin decreases expression of the glutamate transporter neurons is a critical determinant of late phase neurodegeneration mediated GLAST and inhibits glutamate uptake in primary cortical astrocytes via the by matrix metalloproteinase-9. Sci Rep. 2016;6:27354. Rho kinase pathway. Exp Neurol. 2015;273:288–300. 70. Kaplan A, et al. Neuronal matrix metalloproteinase-9 is a determinant of 45. Baba H, et al. GFAP gene expression during development of astrocyte. Dev selective neurodegeneration. Neuron. 2014;81:333–48. Neurosci. 1997;19:49–57. 71. Wright MC, et al. Novel roles for osteopontin and clusterin in peripheral 46. Hubbard JA, Hsu MS, Seldin MM, Binder DK. Expression of the astrocyte motor and sensory axon regeneration. The Journal of neuroscience : the water channel Aquaporin-4 in the mouse brain. ASN Neuro. 2015;7(5). official journal of the Society for Neuroscience. 2014;34:1689–700. 47. Gropp M, et al. Standardization of the teratoma assay for analysis of 72. Meller R, et al. Neuroprotection by osteopontin in stroke. Journal of cerebral blood flow and metabolism : official journal of the International Society of pluripotency of human ES cells and biosafety of their differentiated Cerebral Blood Flow and Metabolism. 2005;25:217–25. progeny. PLoS One. 2012;7:e45532. 48. Kawahara K, et al. Selective blockade of astrocytic glutamate transporter 73. Gardner J, Ghorpade A. Tissue inhibitor of metalloproteinase (TIMP)-1: the GLT-1 with dihydrokainate prevents neuronal death during ouabain TIMPed balance of matrix metalloproteinases in the central nervous system. treatment of astrocyte/neuron cocultures. Glia. 2002;40:337–49. J Neurosci Res. 2003;74:801–6. 49. Straten G, Eschweiler GW, Maetzler W, Laske C, Leyhe T. Glial cell-line 74. Israelson A, et al. Macrophage migration inhibitory factor as a chaperone derived neurotrophic factor (GDNF) concentrations in cerebrospinal fluid inhibiting accumulation of misfolded SOD1. Neuron. 2015;86:218–32. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 17 of 17 75. Cordero-Llana O, et al. Clusterin secreted by astrocytes enhances neuronal differentiation from human neural precursor cells. Cell Death Differ. 2011;18:907–13. 76. Winkler C, Yao S. The midkine family of growth factors: diverse roles in nervous system formation and maintenance. Br J Pharmacol. 2014;171:905–12. 77. Philips T, Rothstein JD. Rodent models of amyotrophic lateral sclerosis. Current protocols in pharmacology. 2015;69:5 67 61-65 67 21. 78. Collins MA, An J, Hood BL, Conrads TP, Bowser RP. Label-free LC-MS/MS proteomic analysis of cerebrospinal fluid identifies protein/pathway alterations and candidate biomarkers for amyotrophic lateral sclerosis. J Proteome Res. 2015;14:4486–501. 79. Tarasiuk J, Kulakowska A, Drozdowski W, Kornhuber J, Lewczuk P. CSF markers in amyotrophic lateral sclerosis. J Neural Transm. 2012;119:747–57. 80. Gao L, Zhou S, Cai H, Gong Z, Zang D. VEGF levels in CSF and serum in mild ALS patients. J Neurol Sci. 2014;346(1-2):216–20. 81. Varghese AM, et al. Chitotriosidase—a putative biomarker for sporadic amyotrophic lateral sclerosis. Clin Proteomics. 2013;10:19. 82. Przyborski SA. Differentiation of human embryonic stem cells after transplantation in immune-deficient mice. Stem Cells. 2005;23:1242–50. 83. Andrews PW, et al. Embryonic stem (ES) cells and embryonal carcinoma (EC) cells: opposite sides of the same coin. Biochem Soc Trans. 2005;33:1526–30. 84. Lee AS, et al. Effects of cell number on teratoma formation by human embryonic stem cells. Cell Cycle. 2009;8:2608–12. 85. Prokhorova TA, et al. Teratoma formation by human embryonic stem cells is site dependent and enhanced by the presence of Matrigel. Stem Cells Dev. 2009;18:47–54. 86. Ramot Y, et al. Compact MRI for the detection of teratoma development following intrathecal human embryonic stem cell injection in NOD-SCID mice. Neurotoxicology. 2017;59:27–32. 87. Priest CA, Manley NC, Denham J, Wirth ED 3rd, Lebkowski JS. Preclinical safety of human embryonic stem cell-derived oligodendrocyte progenitors supporting clinical trials in spinal cord injury. 2015;10(8):939–58. 88. Staff NP, et al. Safety of intrathecal autologous adipose-derived mesenchymal stromal cells in patients with ALS. Neurology. 2016;87:2230–4. 89. Oh KW, et al. Phase I trial of repeated intrathecal autologous bone marrow- derived mesenchymal stromal cells in amyotrophic lateral sclerosis. Stem Cells Transl Med. 2015;4:590–7. 90. Tian C, et al. Autologous bone marrow mesenchymal stem cell therapy in the subacute stage of traumatic brain injury by lumbar puncture. Experimental and clinical transplantation : official journal of the Middle East Society for Organ Transplantation. 2013;11:176–81. 91. Wang S, et al. Umbilical cord mesenchymal stem cell transplantation significantly improves neurological function in patients with sequelae of traumatic brain injury. Brain Res. 2013;1532:76–84. 92. Roybon L, et al. Human stem cell-derived spinal cord astrocytes with defined mature or reactive phenotypes. Cell Rep. 2013;4:1035–48. 93. Gupta K, Chandran S, Hardingham GE. Human stem cell-derived astrocytes and their application to studying Nrf2-mediated neuroprotective pathways and therapeutics in neurodegeneration. Br J Clin Pharmacol. 2013;75:907–18. 94. Petrou P, et al. Safety and clinical effects of mesenchymal stem cells secreting neurotrophic factor transplantation in patients with amyotrophic lateral sclerosis: results of phase 1/2 and 2a clinical trials. JAMA neurology. 2016;73:337–44. 95. Glass JD, et al. Lumbar intraspinal injection of neural stem cells in patients with amyotrophic lateral sclerosis: results of a phase I trial in 12 patients. Stem Cells. 2012;30:1144–51. 96. Mazzini L, et al. Human neural stem cell transplantation in ALS: initial results from a phase I trial. J Transl Med. 2015;13:17. 97. Kondo T, et al. Focal transplantation of human iPSC-derived glial-rich neural progenitors improves lifespan of ALS mice. Stem cell reports. 2014;3:242–9. 98. Nizzardo M, et al. iPSC-derived LewisX+CXCR4+β1-integrin+ neural stem cells improve the amyotrophic lateral sclerosis phenotype by preserving motor neurons and muscle innervation in human and rodent models. Hum Mol Genet. 2016;25:3152–63. 99. Scott S, et al. Design, power, and interpretation of studies in the standard murine model of ALS. Amyotrophic lateral sclerosis : official publication of the World Federation of Neurology Research Group on Motor Neuron Diseases. 2008;9:4–15. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Stem Cell Research & Therapy Springer Journals
Loading next page...
 
/lp/springer_journal/safety-and-efficacy-of-human-embryonic-stem-cell-derived-astrocytes-nM7OQx0R44
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Life Sciences; Stem Cells; Cell Biology
eISSN
1757-6512
D.O.I.
10.1186/s13287-018-0890-5
Publisher site
See Article on Publisher Site

Abstract

Background: Amyotrophic lateral sclerosis (ALS) is a motor neuron (MN) disease characterized by the loss of MNs in the central nervous system. As MNs die, patients progressively lose their ability to control voluntary movements, become paralyzed and eventually die from respiratory/deglutition failure. Despite the selective MN death in ALS, there is growing evidence that malfunctional astrocytes play a crucial role in disease progression. Thus, transplantation of healthy astrocytes may compensate for the diseased astrocytes. Methods: We developed a good manufacturing practice-grade protocol for generation of astrocytes from human embryonic stem cells (hESCs). The first stage of our protocol is derivation of astrocyte progenitor cells (APCs) from hESCs. These APCs can be expanded in large quantities and stored frozen as cell banks. Further differentiation of the APCs yields an enriched population of astrocytes with more than 90% GFAP expression (hES-AS). hES-AS were G93A injected intrathecally into hSOD1 transgenic mice and rats to evaluate their therapeutic potential. The safety and biodistribution of hES-AS were evaluated in a 9-month study conducted in immunodeficient NSG mice under good laboratory practice conditions. Results: In vitro, hES-AS possess the activities of functional healthy astrocytes, including glutamate uptake, promotion of axon outgrowth and protection of MNs from oxidative stress. A secretome analysis shows that these hES-AS also secrete several inhibitors of metalloproteases as well as a variety of neuroprotective factors (e.g. TIMP-1, G93A TIMP-2, OPN, MIF and Midkine). Intrathecal injections of the hES-AS into transgenic hSOD1 mice and rats significantly delayed disease onset and improved motor performance compared to sham-injected animals. A safety study in immunodeficient mice showed that intrathecal transplantation of hES-AS is safe. Transplanted hES-AS attached to the meninges along the neuroaxis and survived for the entire duration of the study without formation of tumors or teratomas. Cell-injected mice gained similar body weight to the sham-injected group and did not exhibit clinical signs that could be related to the treatment. No differences from the vehicle control were observed in hematological parameters or blood chemistry. Conclusion: Our findings demonstrate the safety and potential therapeutic benefits of intrathecal injection of hES-AS for the treatment of ALS. Keywords: Amyotrophic lateral sclerosis, Astrocytes, Human embryonic stem cells, Superoxide dismutase 1 * Correspondence: m.izrael@kadimastem.com Neurodegenerative Diseases Department at Kadimastem Ltd, Pinchas Sapir 7, Weizmann Science Park, Nes-Ziona, Israel Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 2 of 17 Background Inflammation-mediated neuronal injury is also recog- Amyotrophic lateral sclerosis (ALS) is an adult-onset nized as a major factor to promote ALS disease progres- disease characterized by the loss of both upper and sion and amplifies MN death-inducing processes. The lower motor neurons (MNs). Symptoms include pro- neuroimmune activation is not only a physiological reac- gressive paralysis of MN target muscles. The disease is tion to cell-autonomous death, but also an active compo- incurable, and fatal within 3–5 years of first symptoms, nent of non-autonomous cell death. Astrocytes participate due to respiratory failure when the diaphragm is affected in the cellular response to damage and danger signals by [1]. About 10–15% of cases of ALS are familial, and the releasing inflammation-related molecules like NO, IL-6, other cases are sporadic. Familial ALS includes muta- INF-γ, Prostaglandin D2, TGF-β and TNF-α that can 2+ 2+ tions in Cu /Zn superoxide dismutase-1 (SOD1) [2] induce the apoptosis of neurons observed in ALS disease and in RNA/DNA binding proteins FUS and TAR DNA [19–23]. In both physiological and pathological condi- binding protein-43 [3] However, the most frequent tions, astrocytes secrete a wide range of factors with genetic cause of ALS (40% of familial ALS) is an amplifi- multiple influences on their cellular neighbors. cation of a hexanucleotide in a noncoding region of the In addition, disruption of the astrocytic TNFR1– C9orf72 gene [4]. GDNF axis accelerates MN degeneration and disease The pathological mechanisms for ALS are still not well progression, as the levels of the protective agents for understood and the proposed mechanisms include MNs, glial-derived neurotrophic factor (GDNF), are inflammation, oxidative stress, glutamate cytotoxicity reduced [24]. Astrocytes in the ALS rat model acquire and protein aggregation. Although MNs are the main an accelerated senescent phenotype that shows reduced affected cells in the disease, a growing body of evidence support in MNs, that can be partially reversed by GDNF suggests the involvement of astrocytes in the pathology [25]. Another factor that plays a role in ALS pathology is of ALS in a non cell autonomous pathway. The contri- vascular endothelial growth factor (VEGF), originally bution of astrocytes to the pathology of ALS is probably described as a factor with a regulatory role in vascular a combination of loss of homeostatic functions and/or growth and development but it also directly affects neur- gain of toxic functions. Several mechanisms by which onal cells [26, 27]. Transgenic mice expressing reduced ALS patients’ astrocytes affect ALS pathology include levels of VEGF develop late-onset MN pathology, similar astrocyte toxicity; astrocytes that were isolated from to that of ALS [28, 29]. VEGF is secreted by astrocytes sporadic and familial postmortem ALS patients and and has been shown to protect MNs from excitotoxic astrocytes derived from iPSCs of ALS patients have been death, as occurs in ALS [30]. In line with these results, shown to be toxic to healthy (WT) MNs [5, 6]. Similar low levels of VEGF and GDNF were reported in the results were obtained by primary astrocytes isolated cerebrospinal fluid (CSF) of ALS patients [31]. Other G93A from the hSOD1 mouse model with both WT and mechanisms include activation of necroptosis [32] and MNs derived from ALS [7, 8]. The toxic effect of mitochondrial alterations [33–37]. astrocytes on MNs was also demonstrated by addition of These observations led to the rationale that ALS could astrocyte condition medium [9, 10]. This lead to the be treated by implantation of normal wild-type healthy notion that astrocytes of ALS patients secrete toxic/mu- astrocytes from an external source, to support or replace tated proteins that cause specific death of MNs. This dysfunctional ALS astrocytes [38]. In the present work, hypothesis is also supported by in-vivo studies in the we have used human embryonic stem cells (hESCs) as a G93A hSOD1 high copy number ALS models [11–14]. source for large-scale production of astrocyte progenitor Another proposed mechanism is the reduction of func- cells (APCs), which can be stored as frozen banks. These tional astrocytic glutamate uptake suggested to contrib- APCs can be further expanded and differentiated into an ute to glutamate excitotoxicity found in ALS patients enriched population of young committed astrocytes by [15]. GLT-1, a glutamate transporter (aka EAAT2), was removal of the growth factors for 7 days (hES-AS), found impaired in ALS patients [16, 17]. In-vivo studies which demonstrate functional properties of “healthy” as- have demonstrated that focal loss of GLT-1 in the ven- trocytes in vitro. These properties include: uptake of glu- tral horn of the spinal cord precedes disease onset in a tamate; production and secretion of a wide diversity of transgenic rat model for ALS overexpressing SOD1 [18]. neuroprotective factors, as seen by secretome analysis; Transplantation of SOD1(G93A) glial-restricted precur- promotion of axonal outgrowth; and protection of MNs sor cells–glial progenitors that are capable of differenti- from oxidative stress. In animal ALS models (high-copy G93A ating into astrocytes in the cervical spinal cord of WT number hSOD1 transgenic mice and rats), we show rats induced host MN ubiquitination and death, fore- that intrathecal injection of hES-AS into the CSF of G93A limb motor and respiratory dysfunction, and reactive hSOD1 mice and rats had significant effects on delay- astrocytosis and reduced GLT-1 transporter expression ing disease onset, maintaining motor performances and in WT animals [11]. delayed death. To obtain safety data that are relevant to Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 3 of 17 both hES-AS and to their proposed clinical use, we con- (7-day astrocytes, hES-AS), flow cytometry showed that ducted long-term safety and toxicology studies in NSG the percentages of GLAST, GFAP and AQP-4 astrocytic immune-deficient mice. These studies were designed to markers were increased compared to APCs (Fig. 1d). address key safety aspects associated with direct adminis- Upon differentiation of APCs toward committed young tration of hES-AS into the CSF by intrathecal injection, astrocytes there were no remaining undifferentiated cells, including toxicity, biodistribution, long-term engraftment as shown by the levels of TRA-1-60, SSEA-4 and EPCAM, and formation of tumors. which remained < 0.1% (Fig. 1e), indicating high purity and low risk of teratoma formation [47]. It is important to Results note that only few Ki-67-positive cells were observed in Direct differentiation of hESCs into astrocyte progenitor hES-AS cultures (Fig. 1f), indicating that most hES-AS cells and young astrocytes are post mitotic. Two hESC lines (HADC100 and NCL-14) were used to G93A produce astrocytes for engraftment in hSOD1 ALS Biological functionality of hES-AS animal models. Both hESC lines had a normal karyotype, Glutamate uptake capacity expressed pluripotency markers and were capable of The glutamate uptake capacity of hES-AS was tested by differentiating into all three embryonic germ layers incubating the cells in medium containing 0.5 mM glu- [39, 40]. We modified our previously reported protocol tamate and measuring the remaining concentration of [41] to generate an enriched population of APCs from the neurotransmitter at different times up to 120 min. hESCs, followed by further differentiation of the APCs Astrocytes from human spinal cord served as positive into functional astrocytes (Fig. 1a). The protocol was control and medium without cells as negative control. optimized to include good medical practice (GMP)-grade As shown in Fig. 2a, the hES-AS take up glutamate from materials and factors to be compatible for clinical use. In the medium occurred in a time-dependent manner simi- brief, hESC cultures having at least 70% of pluripotent lar to the control human spinal cord astrocytes. After stem cells expressing the SSEA4, TRA-1-60 and EPCAM 2 h, more than 85% of the glutamate was removed from markers were used as a starting material. The hESCs were the culture media. detached and cultured in suspension with stepwise To investigate whether GLT-1 (EAAT2) participates in changes in media composition (Fig. 1a, b). First, all-trans the glutamate uptake, the same experiment was done in retinoic acid and EGF were added for 7 days. This elicited the presence of either WAY-213,613 (1 μM) or dihydro- increased production of bone morphogenetic factors (i.e. kainic acid (DHK, 500 μM) [48]. With either of these BMP4, BMP6, BMP2, BMP7 and BMP11), which were GLT-1 inhibitors (Fig. 2b) the removal of glutamate in found to be essential for obtaining glial restricted cells, 60 min was inhibited by 60% (from 64.1% removal in the particularly astrocyte lineage cells [41, 42]. The suspension control to 25% with the inhibitors), demonstrating that a culture was continued with EGF resulting in the formation significant part of the glutamate uptake can be attributed of neurospheres, which were seeded in 2D culture on to GLT-1 activity in the hES-AS. laminin. The cells were expanded by successive passages in the presence of growth factors (bFGF and EGF) and human serum with the doubling time being 21 ± 2.6 h. This produced APCs that can be stored as frozen cell Neuroprotective effect against oxidative stress banks. The APC karyotype was tested at different passages Cultures of mouse spinal cord MNs were challenged (up to passage 12) and was found normal (Fig.1c). Flow with 150 μM hydrogen peroxide (H O ). The number of 2 2 cytometry analysis of APCs showed that the levels of apoptotic MNs was measured after staining for activated pluripotent markers, SSEA-4, EPCAM and Tra-1-60, were caspase-3 and the total number of MNs being measured <0.2% (Fig. 1e). Above 90% of APCs were positive for the by staining for tubulin-β3. Using high-content image astrocytic marker CD44 [43](Fig. 1d). The APCs had screening analysis, we calculated the percentage of apop- additional astrocytic markers such as the Glutamate totic MNs (seen as yellow cells, Fig. 3b, left panel). The Aspartate Transporter (GLAST, aka Excitatory Amino results (Fig. 3a) indicate a significant decrease (p < 0.05) Acid Transporter 1 (EAAT1)) [44], glial fibrillary acidic in MN death by adding conditioned medium from the protein (GFAP) [45] and Aquaporin-4 (AQP-4) [46], as hES-AS, as seen by the decrease in caspase-3-positive well as neuroepithelial stem cell markers Nestin, A2B5 cells (Fig. 3b, right panel). When the hES-AS were and CXCR-4 (Fig. 1d). The frozen/thawed APCs were added in coculture with the MNs, there was a greater further expanded for 2–3 weeks and then differentiated decrease in apoptosis resulting from oxidative stress toward committed astrocytes, by removing growth factors (Fig. 3a, p < 0.01) to levels similar to spontaneous apop- EGF and bFGF as well as human serum from the media tosis. These results demonstrate the neuroprotective and adding vitamin C. After 7 days without growth factors effects by hES-AS in vitro. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 4 of 17 d e Fig. 1 Differentiation of human embryonic stem cells into astrocyte progenitor cells and committed astrocytes. a Steps and timeline for differentiation of hESCs first into astrocyte progenitor cells (APCs) which can be stored frozen in APC banks. These APCs are further expanded with growth factors (bFGF, EGF and human serum), and then differentiated into astrocytes (hES-AS) by removal of growth factors for 7 days. b Representative images of different steps from hESCs to APCs (as in a, steps marked by asterisk). Arrows show selected neurospheres. c Representative spectral karyotyping analysis showing normal karyotype of APC cell bank at passage 12. d Flow cytometry analysis on nine batches of APC banks (grown with human serum, bFGF and EGF) versus 13 batches of astrocytes differentiated for 7 days showing expression of astrocytic markers (CD44, GLAST, GFAP, and Aquaporin-4) and neuroepithelial stem cell markers (Nestin, A2B5 and CXCR4). e Flow cytometry analysis of APCs and astrocytes differentiated for 7days(same batchesasin d) showing very low expression of pluripotent markers (below limit of detection, 0.1%). f Representative immunofluorescence images of astrocytes differentiated 7 days, showing expression of astrocyte markers (GFAP, GLAST, S100β and AQP-4) and very low proliferation marker (Ki-67, arrow). Scale bars = 100 μm. Error bars represent SD. hESC human embryonic stem cell, DAPI 4′,6-diamidino-2-phenylindole, GFAP Glial Fibrillary Acidic Protein, GLAST Glutamate Aspartate Transporter, RA Retinoic acid hES-AS stimulate axonal outgrowth in vitro The cultures were labeled by ICF with antibodies against We next assessed the ability of hES-AS to induce axonal axonal neurofilament-160 and GFAP markers. Represen- outgrowth in vitro. Rat primary cortical neurons derived tative images of the five conditions are shown in Fig. 4a. from day 18 embryos were precultured for 2 days in By high-content image screening analysis, the total area of Neurobasal medium (with B27) and then further cul- axons and neurites in the NF160-stained images was deter- tured for 4 more days in either medium alone or mined. A significant increase in axonal outgrowth was seen supplemented with 10 ng/ml Neurotrophin-3 (NT-3, as in the neurons cocultured with hES-AS (Fig. 4b, p < 0.01). positive control), or cocultured with hES-AS (1–2×10 Moreover, addition of the hES-AS conditioned medium cells), or cocultured with hES-AS conditioned medium was found to stimulate axonal outgrowth to a similar (collected from days 5 to 7 of astrocyte differentiation). extent as compared to the cocultures, indicating that Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 5 of 17 a b Fig. 2 hES-AS take up glutamate from medium. a Glutamate concentration measured in solutions with 500 μM glutamate that were incubated for indicated times either alone (black bars 1–2) or with hES-AS differentiated for 28 days (black bars 3–7). Kinetics of glutamate removal by hES-AS similar to that by astrocytes extracted from human spinal cord (gray bars). b Percentage of glutamate uptake after 60 min by hES-AS alone or in presence of inhibitors of glutamate transporter GLT-1, WAY-213,613 (1 μM) and DHK (500 μM). Error bars are SD of triplicates. *p < 0.05. hESC human embryonic stem cell, DHK dihydrokainic acid this neurogenic activity can be attributed to factors se- peptidases. This indicates that there is a complex set of creted by these astrocytes. As expected, GFAP-positive factors secreted by the hES-AS, beyond the classical neuro- cells were observed only in the cocultures, indicating trophic factors. Many of these factors may be responsible that the rat cortical neurons were not contaminated by for the neurogenic and neuroprotective activities observed rat astrocytes. earlier. Examples of the secreted factors with effects on neurons or with antiprotease activity are presented in Neurotrophic factor synthesis and secretion Table 1. Several of these factors may be relevant for poten- We first measured the levels of known neurotrophic tial therapeutic mechanism of action in ALS (e.g. Osteo- factors GDNF, BDNF, VEGF and IGF-I both in hES-AS pontin, tissue inhibitor of metalloproteinase (TIMP)-1 and culture supernatant media and in cell extracts (cell con- TIMP-2,Midkine,MIF;see Discussion). tent). VEGF was found to be secreted from hES-AS that G93A were differentiated without growth factors for 28 days Transplantation of hES-AS in SOD1 mouse and rat (Additional file 1: Figure S1). IGF-1 was also secreted, ALS models G93A whereas GDNF and BDNF were found inside the cells Both SOD1 mouse and rat models present a typical but less was secreted (Additional file 1: Figure S1). The pattern of ALS disease progression, in which onset of levels of these classical neurotrophic factors were in the the disease in hindlimbs precedes that in forelimbs, and range found in human CSF [49, 50]. in which the end stage results from compromised To have a more comprehensive view of the factors respiratory function [18, 51]. A dose of 2 × 10 hES-AS secreted by 7-day and 28-day differentiated hES-AS, we (differentiated for 7 days) were injected into the CSF of G93A carried out secretome analysis. The 48-h conditioned hSOD1 mice through the cisterna magna (CM), medium of replica cultures of hES-AS were analyzed using either once on day 67 ± 2 after birth or twice on days 67 the human Quantibody Kiloplex Array (RayBiotech), cap- ± 2 and 97 ± 2 (Additional file 3: Figure S2). Disease able of detecting 1000 proteins. A total of 220 protein fac- onset was determined by the loss of 3% of maximal body tors were found to be secreted at levels over the threshold weight. Results demonstrate that double transplantation in 7-day hES-AS, about 25% of which being more abun- of the hES-AS significantly delayed disease onset com- dant at 28 days (see Additional file 2:Table S1).Among pared to sham-injected controls (Additional file 3: Figure the highest 120, there were 25 proteins with activities in S2A; median 119 days vs 112 days; p = 0.0012, log-rank), neurogenesis, axon or neurite outgrowth or axon guidance. and was better than with a single injection. Motor per- Interestingly, there were 13 proteins with antiprotease formance, as measured by Rotarod test as well as by neuro- activity. In addition, there were extracellular matrix (ECM) logical scoring, was significantly improved in mice that were components, cell adhesion membrane proteins and a few injected twice with hES-AS, compared to sham-injected Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 6 of 17 Fig. 3 hES-AS protect MNs from oxidative stress. A Mouse motor neurons exposed in 96-well plates to 150 μMH O for6h (bar1)or left untreated 2 2 (bar 4). During H O treatment, neuron cultures supplemented with conditioned medium from hESC-derived astrocytes, differentiated for 28 days 2 2 (ACM, bar 2), or with 20,000 of the same hES-AS (bar 3). After fixation, cells double-stained with anti-tubulin β3 antibody (neuron marker, green) and anti-Caspase-3a (apoptotic marker, red). Percentage of apoptotic neurons (Caspase3a over tubulin β3-positive cells) counted using high-content image screening system (Arrayscan; Cellomics). Results represent average ± SD for 10 wells of 96 well-plate per treatment (for each well, 49 fields were analyzed). *p < 0.05; **p <0.01. b Left panel: representative image of neuron cultures with H O treatment. Apoptotic neuronal cell bodies yellow 2 2 (arrows, due to overlapping of red Caspase-3 staining with green tubulin β3). Right panel: with ACM, much less apoptotic yellow cells are seen. Scale bar: 100 μm. hESC human embryonic stem cell, H O hydrogen peroxide 2 2 mice (Additional file 3:FigureS2D,E; p <0.05). Two injec- curve (AUC) analysis). The disease onset was delayed very tions were better than a single dose. The survival of mice significantly by hES-AS treatment (Fig. 5b, p = 0.0001); injected twice with hES-AS was somewhat prolonged com- Kaplan–Meier analysis showed that 50% of treated rats pared to sham-injected mice (Additional file 3:FigureS2B; lost 3% of their body weight by day 175 compared to day median survival 130 days vs 126.5 days; but p=0.1, 157 in the sham-injected group. The hES-AS-treated rats log-rank). With the double injection there was also a trend maintained their body weight significantly longer (by for longer survival at late times, compared to one injection. about 30 days) than sham-injected rats (Fig. 5c, p =0.007). G93A We then shifted to the rat hSOD1 ALS model, A set of motor tests demonstrated the therapeutic which allows use of intrathecal injection by lumbar punc- effect of the hES-AS treatment. First the overall devel- ture (LP), a route of administration similar to what could opment of clinical symptoms, as evaluated by open field be applied in human patients. The rat model also allowed neurological scoring, was significantly delayed (Fig. 5d, administration of more cells. A total of 6 × 10 hES-AS p < 0.001). The decline of motor functionality, as mea- (differentiated for 7 days) was administered divided into sured by “time to fall” from a Rotarod, was markedly two injections, the first on day 50 ± 2 after birth and the slowed down by hES-AS treatment, the animals main- second on day 70 ± 2. A control group was sham-injected taining normal motor activity for more than 1 month with the vehicle solution. The LP injections were in the longer than the controls (Fig. 5e, p < 0.001). Likewise, subarachnoid space between L5 and L6 vertebra. The the loss of forelimb muscle strength, as measured by median survival of the hES-AS-treated rats was 216 days the grip strength test, was significantly slowed down, compared to 182 days in the sham-injected rats (Fig. 5a); just as the Rotarod performance (p <0.001; data not Kaplan–Meier analysis for the entire experiment showed shown). Other observations were that no tumors were an increased survival trend (p = 0.077 by area under the observed in the animals post mortem. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 7 of 17 a b Fig. 4 hES-AS and their conditioned medium stimulate axonal outgrowth in cortical neurons. a Mouse cortical neurons cocultured with hES-AS 4 4 (7-day differentiated APC) (2 × 10 and 4 × 10 cells), or with neurotrophin 3 (NT3) as positive control, or left untreated (negative control). Last row shows neurons cultured with conditioned medium from same hES-AS (taken after 48 h of culture). Representative images of cells stained with DAPI and by immunofluorescence for neurofilament-160 (NF160) and GFAP shown for each condition. Stimulation of axon and neurite outgrowth seen from NF160 stain and merge of NF160 (green) and GFAP (red). Scale bar = 100 μm b By high-content image screening analysis (Arrayscan; Cellomics), area covered by axon and neurite outgrowth quantified, using 49 fields for each of six replica wells from each experimental conditions. Error bars represent SD. *Student’s t test, p <0.05). DAPI 4′,6-diamidino-2-phenylindole, GFAP Glial Fibrillary Acidic Protein Assessment of safety, tumorigenicity and biodistribution attached to the pia mater. To assess the biodistribution of of hES-AS following a single injection to the cisterna hES-AS outside the CNS, the detection of human cells in magna of NSG mice mouse tissues was performed by quantitative real-time PCR The safety, tumorigenicity and biodistribution phases were (qPCR), targeting the specific sequence of the human Alu conducted in compliance with principles of good laboratory sequence. The detection was performed in nine organs practice (GLP) over a period of up to 9 months. hES-AS, including the spleen, kidney, testis/ovary, liver, heart, bone differentiated for 7 days, were injected intrathecally into the marrow of the femur, lungs, and cervical lymph nodes. The CSF of NSG mice through the CM with 0.4 × 10 cells/ qPCR method was validated prior to the study and both mouse, or with a vehicle. Mice were sacrificed 4, 17 and the limit of detection (LOD) and the limit of quantification 39 weeks post transplantation. No clinical signs were (LOQ) were set at one human cell (DNA equivalent) per attributable to treatment during the monitoring periods. 1 μg of mouse DNA. The PCR results showed no detection Cell-injected mice made similar body weight gain by 4, 17 of human DNA above the LOD in any of the tested organs and 39 weeks post dose to the vehicle control groups. In 4 and 17 weeks after transplantation. addition, there were no differences from the vehicle control We also examined the astrocytic identity of hES-AS in at the hematological and blood chemistry investigations at vivo 2 months after their transplantation in the CSF of 4, 17 and 39 weeks after dose administration (data not immunodeficient mice. Histological sections were shown). Histopathological evaluation of the brain and stained for the general human cytoplasmic specific spinal cord was performed to assess tumorigenicity. No marker Stem121 and for Stem123 (human-specific teratoma or other tumors that could be related to the treat- GFAP antibody) in order to ascertain the presence of ment were seen in the transplanted animals in any of tested human cells. All Stem121-positive cells were positive for time points. In order to evaluate the hES-AS distribution in human GFAP, demonstrating that the transplanted the CNS, the sections were stained using an in-situ hES-AS maintained their astrocytic identity in the CSF hybridization (ISH) technique with a human-specific Alu Y (Fig. 7). Further staining for the cell cycle marker sequence. Cells positive for Alu Y sequences were present Ki67 showed that 0.33 ± 0.15% of Stem121-positive at all levels of the CNS in similar incidences between the cells in the CSF were also positive for Ki67, indicating three study time points. The incidence for the various levels for the very low proliferative capacity of hES-AS in range between 17% (distal areas from injection site) and vivo (Fig. 7g). 80% (at vicinity of the injection site) after 4 weeks, between 13% and 97% after 17 weeks and between 21% and 96% Discussion after 39 weeks (Fig. 6 and Additional file 4:Table S2).The This work describes the derivation of young astrocytes cells were almost uniformly seen along the meninges, from human embryonic stem cells (hES-AS), which have Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 8 of 17 Table 1 hES-AS secrete a variety of factors with effects on multiplicity of mechanisms that underlie MN degener- neurons or with antiprotease activity ation in this disease. Thus, a potential therapy that acts 7-day astrocytes 28-day astrocytes through multiple mechanisms of action to treat the 6 6 (ng/ml/10 cells) (ng/ml/10 cells) broad pathological aspects of the disease is more likely Secreted factors with effects on neurons to be effective. An example for the complexity of the Osteopontin (OPN) 53.1 ± 29 56.8 ± 5.5 disease is the involvement of astrocytes in the degener- ation of MNs [5, 7, 8, 57]. Such noncell autonomous Dickkopf-3 (DKK-3) 43.1 ± 14.2 33.8 ± 1.6 death of MNs caused by ALS-type astrocytes supports Thrombospondin 22.7 ± 11.5 118.9 ± 36.8 the rationale that transplantation of healthy human as- (TSP-1) trocytes into the CNS of ALS patients may compensate Secreted Frizzled 20.8 ± 10.9 41.2 ± 23.0 Protein (sFRP3) for the malfunctional astrocytes and rescue dying MNs (review in [38]). Brevican proteoglycan 15.6 ± 4.9 12.6 ± 3.3 hES-AS exhibit multiple activities that were shown to Tripeptidyl peptidase 11.5 ± 4.2 20.1 ± 11.7 be impaired in ALS-type astrocytes. Astrocytes from (CLN2) ALS transgenic mice express more iNOS/NOS2, leading Clusterin 9.5 ± 3.2 6.5 ± 0.5 to increased release of NO, which exacerbates oxidative Midkine 8.4 ± 3.0 6.1 ± 3.5 stress leading to MN death [58]. We show in our study NSE 3.5 ± 1.8 0.9 ± 0.2 that hES-AS protect in-vitro spinal cord MNs from MIF chemokine 1.8 ± 0.6 0.4 ± 0.1 oxidative stress produced by H O In ALS patients, a 2 2. CXCL16 1.5 ± 0.8 2.1 ± 0.2 decrease of the astroglial GLT-1 glutamate transporter is observed [16], leading to decreased glutamate uptake in Thrombospondin-2 0.85 ± 0.4 2.3 ± 0.4 the synaptic clefts of the spinal cord. Accumulation of GRFα-1 0.45 ± 0.2 1.0 ± 0.6 excitatory glutamate makes MNs in ALS more suscep- VEGF 0.05 ± 0.02 0.23 ± 0.09 tible to excitotoxicity [59]. hES-AS express both glutam- Antiprotease activity ate transporters GLAST and GLT-1 and efficiently Fetuin A 1816.0 ± 677 1404.7 ±+ 129.4 uptake glutamate, which is in part due to their GLT-1 Tissue inhibitor of 16.6 ± 6.8 14.5 ± 0.8 expression, as shown by GLT-1 inhibitors. Another metalloprotease TIMP-2 mechanism by which the diseased astrocytes lead to MN PAI-1 Serpine 1 7.2 ± 6.2 54.9 ± 5.9 death is by a decrease in the secretion of neurotrophic protease inhibitor factors. hES-AS produce and secrete the neurotrophic Tissue inhibitor of 7.0 ± 3.8 6.5 ± 0.8 factors GDNF, BDNF, IGF-1 and VEGF in a comparable metalloprotease TIMP-1 amount to that of endogenous astrocytes. The neuro- Serpin A4 4.3 ± 2.5 4.0 ± 0.4 tropic property of hES-AS was demonstrated by cocul- Results shown as mean ± standard deviation for triplicates of hES-AS differentiated tures of hES-AS with neurons and by hES-AS conditioned for 7 days and duplicates of hES-AS differentiated for 28 days medium alone, indicating activity of soluble secreted Secretome analysis performed on 48-h conditioned media of hES-AS. Listed are factors with activities in neuroprotection, neurogenesis, axon growth or factors. Secreted VEGF is likely to play an important guidance, as well as antiproteases. For relevance to amyotrophic lateral sclerosis, role by protecting neurons in ALS, reducing excitotoxi- see Discussion. Complete secretome list presented in Additional file 2:Table S1 GRF GDNF family receptor, hES-AS human embryonic stem cell-derived city [28, 60], and its concentration is lower in the CSF astrocytes (differentiated from APCs for 7 days), MIF macrophage migration of ALS patients [31]. In addition, GDNF synergizes with inhibitory factor, NSE neuron specific enolase, PAI plasminogen activator VEGF to prolong survival in a murine ALS model [61]. inhibitor, VEGF vascular endothelial growth factor Intrathecal injection of CSF from sporadic ALS patients therapeutic activity in vivo following intrathecal injection to neonatal rats induces selective degeneration of MNs G93A into the CSF of transgenic SOD rats and mice. In [62] and downregulates the levels of both BDNF and addition, we describe the results of a preclinical safety IGF-1 in the spinal cord [63]. Supplementation of study in immunodeficient mice to assess the tumorigen- BDNF reverses the neurodegenerative changes induced icity potential and biodistribution of hES-AS in target by ALS-CSF in MN cultures [64]. and distal organs. The nature of the secreted factors was further investi- To date, two FDA-approved drugs, riluzole and Radi- gated by a secretome analysis, clearly illustrating the pleio- cava, were shown to modestly attenuate motor deterior- tropic activity of the cells. hES-AS secrete many factors ation in ALS patients [52–55]. Still, many late-phase having activities on neurons [65, 66–68] as well as several clinical trials failed to demonstrate a significant improve- antiproteases and factors which could remodel the ECM ment in slowing down disease progression when using (see Table 1). Among the more abundant factors found in single-target drugs [56]. ALS is a multifactorial disease the secretome analysis, several have been linked to ALS, and therapeutic approaches should take into account the thereby shedding new light on the possible mechanisms of Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 9 of 17 ab c e G93A Fig. 5 Effect of hES-AS transplantation on disease onset, motor activity and survival in hSOD1 rat ALS model. hES-AS (APCs differentiated for 7 days) injected intrathecally through lumbar puncture (L5–L6), in two doses of 3 × 10 cells each on days 50 and 70 after birth in hSOD1G93A rats. a Kaplan–Meir survival curves of rats treated with hES-AS (green) show prolongation of median survival compared to sham-injected group (vehicle, red). b Kaplan–Meir plot of disease onset (defined by 3% body weight loss) shows significant delay in hES-AS-treated ALS rats. c Body weight maintained significantly longer in hES-AS-treated ALS rats. d Neurological score. e Significant prolongation of motor performance on Rotarod in hES-AS-treated ALS rats. Same seen by grip strength measurement. c, d Values represent mean ± SEM action underlying the observed therapeutic effect in ALS capacity to save primary MNs from the degeneration models. One of the most abundant factors in the secre- caused by the ALS mutant SOD1 form, probably by acting tome is Osteopontin (OPN/SSP1), which in the mutant as a chaperone [74]. Also secreted is Clusterin, another SOD1 model of ALS is found to be associated with MNs chaperone, promoting axon regeneration, as observed on that are more resistant to degeneration early in the dis- peripheral sensory neurons [71], and increasing neuron ease, but low in the MNs more vulnerable to degeneration survival [75]. Midkine secreted by astrocytes is a known in ALS [69]. Conversely, the vulnerable MNs are high in neurotrophic factor promoting neurite outgrowth and high low matrix metalloproteinase MMP-9 (MMP9 /OPN ), neuron survival (review in [76]). The multiple nature of whereas MMP-9 is low and OPN is high in the the factors secreted by the hES-AS supports a mode of ALS-resistant MNs [69, 70]. Exogenous addition of OPN action much more diversified than merely through the has neurogenic effects, stimulating regeneration of motor classical neurotrophic factors. axons [71] and protecting neurons after ischemia in vitro The efficacy of hES-AS to delay disease onset and to and in vivo [72]. Although MMP9 was not detected in the ameliorate disease progression was evaluated in trans- G93A secretome of our astrocyte cultures, inhibitors of MMP9 genic high copy number SOD1 mouse and rat and other matrix metalloproteases were abundantly se- models, which recapitulate many of the clinical symp- creted, particularly the tissue inhibitors of metallopro- toms of the ALS disease in humans [18, 51, 77]. Intra- teases TIMP-1 and TIMP-2, which play a major role in thecal injection of hES-AS significantly delayed the preventing degradation of ECM components by MMPs or onset of the disease and slowed down the deterioration regulating ECM remodeling (review in [73]). Another of motor function. These effects were more pronounced chemokine found in the secretome is MIF, which has the when the cells were administered twice (3–4 weeks Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 10 of 17 b e Fig. 6 hES-AS distribute throughout CNS after intrathecal injection. hES-AS (400,000 cells) differentiated for 7 days transplanted intrathecally into NSG mice (into CSF through CM). a Illustration of brain and spinal cord sections performed: seven brain sections (L#1–L#7 as in [64]) and four of representative regions of spinal cord. b–d Graphical representation of AstroRx cell presence (as determined by Alu cell staining) and percent incidence of frequency scores ≥ 2 (one to three foci of 10–20 cells per foci) after 4-week (b), 17-week (c) and 39-week (d) follow up. AstroRx Cell presence calculated as incidence (%) from all samples (n) within each group. Frequency of score ≥ 2 calculated as incidence (%) of frequency scores ≥ 2 from only those sections in which AstroRx cells present. e–g Representative images of different sections demonstrating distribution of hES-AS throughout CNS using ISH with and Alu Y probe (human specific) of 17-week cohort. e Sacral region of spinal cord with numerous Alu cells (arrows) along surface of the spinal nerves (asterisks). f Brain, level 5. Arrows indicate cells along meningeal surface at many locations. g Brain, level 6. Arrows indicate Alu cells along meningeal surface along base of medulla at brain level 6. Cells attached to the pia mater (arrows). hES-AS human embryonic stem cell-derived astrocytes (differentiated from APCs for 7 days) apart) than with a single injection. Intrathecal injection A major safety concern associated with pluripotent stem into the CSF is in line with the proposed mode of cell-based therapies is the presence of residual undifferen- action, in which the healthy astrocytes would work at a tiated stem cells that might continue to divide without distance to modify the environment of brain and spinal control or develop teratoma after their transplantation in cord MNs. Indeed, the CSF composition shows several the body [82, 83]. We minimize the possibility of teratoma changes in the course of ALS [78, 79], including an formation by assuring a complete differentiation of hESCs increase in oxidative stress markers, an increase in into committed astrocytes with a normal diploid karyo- glutamate in at least 40% of patients and variations of type and minimal proliferation capacity. Teratoma forma- VEGF concentration correlating with the length of tion from undifferentiated hESCs depends on several survival [80], and other changes including OPN in- factors, among them the site of implantation and number crease [81]. Moreover, the fact that inoculation of CSF of transplanted cells. Several studies reported that undif- from ALS patients to animals is neurotoxic [63]dem- ferentiated hESCs develop teratomas within 6 weeks after onstrates that materials injected into the CSF can transplantation in immunodeficient mice [47, 82, 84, 85]. affect the parenchyma. We previously reported that injection of undifferentiated Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 11 of 17 b c Fig. 7 hES-AS are post mitotic and maintain their astrocytic identity in vivo. a–c High-content analysis of hES-AS cells in vitro displayed homogeneous + + + expression of human GFAP (Stem123). %Ki67 cells calculated as % Ki67 nuclei / total number of nuclei. Ki67 cells rarely found within hES-AS cell population (arrows). d–f Two million hES-AS injected intrathecally into the lumbar region twice, with interval of 21 days. Analysis of graft, 8 weeks post first cell injection, showed transplanted cells were located in subarachnoid space, attached to pia mater of lumbar spinal cord and nerve bundles. Cells + + maintained their astrocytic characters and homogeneously expressed human-origin GFAP. %Ki67 hES-AS cells calculated as % Ki67 nuclei / total + + number of nuclei of Stem123 cells. Ki67 staining very rare among hES-AS cells (arrows), indicating that cells are non-proliferative in vivo. hES-AS human embryonic stem cell-derived astrocytes (differentiated from APCs for 7 days), DAPI 4′,6-diamidino-2-phenylindole hESCs intrathecally into immunodeficient mice results in maintained stable over time, supporting that the cells re- teratoma formation within 5–7 weeks after injection [86]. main quiescent in the CSF. The effective biodistribution In our current study, we evaluated the formation of of hES-AS along the entire CSF supports the clinical G93A teratomas, or any other tumor, by hES-AS up to 39 weeks benefits we observed in SOD1 models. We found an after their intrathecal injection, long enough to allow attenuation in motor activity loss in both lower and development of teratomas. Histology evaluation showed upper limbs and the tail, indicating that the cells exert the cells survived in the CSF for the entire duration of the their activity on multiple regions of the CNS. The pos- study, attached to the pia mater along the neuroaxis, The sible migration of cells to distant organs was evaluated cells uniformly expressed astrocytic markers with very rare by qPCR for amplification of the Alu Y genomic se- coexpression of the cell cycle marker Ki67. Importantly, quence in nine organs. hES-AS were not found in any hES-AS did not develop teratoma or any other tumors in distant organ above the detection limit of the method (1 any of the treated mice. In line with these results, Priest et cell) at 4 and 17 weeks after their intrathecal injection. al. [87] also reported the absence of teratomas in the CNS This confined distribution of the cells to the CNS mini- following intraspinal injection of oligodendrocyte progeni- mizes any possible risk of presence of ectopic glial tissue tors derived from hESCs into the spinal cord of immuno- in nontarget organs outside the CNS. deficient rats. Large quantities of human astrocytes would be needed To access the CNS, we chose the CSF as the injection for the treatment of ALS patients worldwide. As shown site for hES-AS. The circulating CSF helps to distribute here, clinical-grade human ESCs provide a robust and the injected cells throughout the subarachnoid space. In controlled source of cells for mass production of glial addition, injection into the CSF by LP is a common progenitors that can give rise to functional astrocytes. To low-risk medical practice already demonstrated in comply with GMP standards, we adjusted our previous several clinical trials with cell-based therapies [88–91]. A protocol, originally aimed to produce both astrocytes and biodistribution evaluation of hES-AS in the CNS was oligodendrocytes [41], to include only GMP-grade mate- performed by in-situ hybridization of the Alu Y gene at rials. Under this protocol, large amounts of astrocyte pro- 4, 17 or 39 weeks following a single intrathecal injection genitor cells (APCs) are obtained, which can be frozen in of cells into immunodeficient mice. The analysis re- liquid nitrogen for long-term storage [41] as master and vealed the presence of hES-AS cells in the subarachnoid working cell banks for future expansion. Upon thawing of space throughout the entire CNS. Cell numbers were the APC vial, the differentiation into hES-AS is completed Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 12 of 17 within 7 days of culturing. In terms of yield, using our undifferentiated state of the hESCs was routinely protocol we can produce a total of 2 × 10 hES-AS from assessed by flow cytometry analysis of the surface a single batch of hESCs. Hence, the process described here markers SSEA-4, EpCAM and TRA-1-60, and by im- is suitable for mass production of clinical-grade hES-AS munofluorescence staining for the transcription factors per batch, which can potentially treat thousands of pa- NANOG and OCT4. Both lines were propagated in tients [92, 93]. undifferentiated state on a HFF feeder layer (25,000 In recent years, clinical trials of cell therapy in ALS cells/cm ) by passaging every 6–7 days using collage- have mainly used autologous transplantation of mesen- nase in order to detach the whole hESC colonies from chymal stem or stromal cells (MSCs) [89, 94], in which the feeder cell layers. The colonies were mechanically cells are taken from the patients and after in-vitro broken and seeded in a ratio of 1:3–6. The hESCs were culture are returned to the same patient. While giving grown in ES1 media composed of KO-DMEM, 14% (v/v) promising clinical efficacy, these autologous transplan- KO serum replacement, 2 mM glutamine, 1× MEM non- tations have limitations and it would be advantageous essential amino acids, 0.1 mM β-mercaptoethanol and to develop allogeneic cells as a shelf-product that would 25 U/ml penicillin, 25 μg/ml streptomycin (all from Life provide a treatment for all ALS patients. Given that Technologies) and 8 ng/ml bFGF (R&D). Important to intrathecal administration is effective (as seen with the note is that for generation of clinical-grade hESCs, the MSCs), it would be easier than injections in the spinal cells were adapted to feeder free conditions and the media cord anterior horn, which requires major surgery as composition was changed to Essential 8™ (E8) medium done in recent ALS clinical trials with neural stem cells (Thermo Fischer Scientific). taken from human organ donors [95, 96]. Future Formation of neurospheres (NS) was done in suspension clinical trials could use human pluripotent stem cell (3D) cultures. In brief, the harvested hESC colonies were cultures for mass production of neural cells, either transferred into 100-mm ultralow attachment culture plates from human iPSCs [97, 98] or from human ES cell lines (Corning) containing ITTSPP/B27 medium. ITTSPP/B27 is as described here. a mixture of DMEM/F12 containing 1% B27 supplement, 1% Glutamax, 1.5% Hepes at pH 7.4 (all from Thermo Sci- Conclusions entific), 1% penicillin/streptomycin/amphotericin solution Here we describe the derivation of a highly enriched (Biological Industries), 25 μg/ml human insulin (ActRapid; population of functional, clinical-grade, human astrocytes Novo Nordisk), 50 μg/ml human Apo-transferrin (Athens), (hES-AS) from embryonic stem cells. The hES-AS were 6.3 ng/ml progesterone, 10 μg/ml putrescine, 50 ng/ml shown to protect MNs by multiple mechanisms, similarly sodium selenite and 40 ng/ml triiodothyronine (T3) (all to normal astrocytes, including clearance of glutamate, from Sigma). ITTSPP/B27 was supplemented with 20 ng/ secretion of multiple NTFs, neutralization of ROS and ml r-human EGF (R&D Systems). After 2 days, the medium promotion of neural outgrowth. Intrathecal injection of was switched to ITTSPP/B27 supplemented with 20 ng/ml hES-AS to rodent models of ALS delays disease onset, EGF and 10 μM ATRA (Sigma). The culture was continued slows down disease progression and extends life expect- in suspension in the nonadherent plates for 7 days with ancy. A 9-month safety study conducted in an immunode- daily replacement of the medium (stage 2; Fig. 1). During ficient NSG animal model, under GLP conditions, showed the last step, which allows for NS ripening, the culture was that intrathecal transplantation of hES-AS cells to the continued in ITTSPP/B27 medium supplemented with cerebrospinal fluid (CSF) is safe. Thus, these findings 20 ng/ml EGF for 18 days. Medium was replaced every demonstrate the feasibility, safety and potential efficacy of other day (stage 3; Fig. 1). For APC expansion, round yel- intrathecal injections of hES-AS for the treatment of ALS. low NS were manually selected using a stereoscopic micro- The safety and efficacy of hES-AS treatment in ALS scope and transferred into six-well plates coated with patients will be tested in a phase I/IIa clinical trial (Clini- Matrigel or GMP-compliant laminin 521 (from Biolamina) calTrials.gov identifier: NCT03482050). in ITTSPP/B27 supplemented with 20 ng/ml EGF. Medium was replaced every other day for 7–10 days (passage 0). In Methods order to produce a monolayer, the spheres were dissociated Derivation of astrocyte progenitor cells and committed with TryplE (Thermo Scientific) and reseeded on ECM astrocytes from hESCs (passage 1) in N2/B27 medium consisting of DMEM/F12 Two clinical-grade hESC lines, were used: NCL14, licensed with 0.5% (v/v) N supplement, 1% (v/v) B27 supplement, from the University of Newcastle; and HADC100, 1% Glutamax and 1.5% Hepes at pH 7.4 (all from Thermo obtained from the Hadassah Medical Organization Scientific). The growth factors EGF and bFGF (R&D (HMO), Jerusalem (Prof. Benjamin Reubinoff). Master Systems) were added at 10 ng/ml each. The monolayer cells cell banks (MCB) and working cell banks (WCB) of were further passaged weekly until a sufficient number of these hESCs were created at Kadimastem Ltd. The cells was generated. Cells were then frozen in liquid Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 13 of 17 nitrogen andstoredas banks of APCs.ThawedAPCswere Department of Life Sciences Core Facilities, Weizmann further expanded as described earlier for 2–3 weeks. In Institute of Science. order to differentiate the APCs toward astrocytes, EGF and bFGF were removed from the media, 50 μg/ml ascorbic Flow cytometry acid (Sigma) was added and the culture was continued for Cells were analyzed by flow cytometry for identity and pur- 7or28days. ity markers using the following antibodies: anti-A2B5 (1:20; Miltenibiotec), anti-GLAST (1:20; Miltenibiotec), anti-CD44 (1:20; BD Pharmingen), anti-CXCR4 (1:20; Immunocytofluorescence assays Biolegend), anti-TRA-1-60 (1:50; Biolegend), anti-EPCAM Cells were fixed with 4% paraformaldehyde (PFA), washed (1:50; Biolegend), anti-SSEA4 (1:50; Biolegend), anti-GFAP with PBS and kept at 4 °C before staining. Permeabilization (1:2000; Sigma), Nestin (1:500; BD Pharmingen) and was done by 0.5% Triton X-100 in Blocking solution (5% anti-AQP-4 (1:2000; Abcam). The Flow Cytometer FACS BSA; Sigma) and 3% horse serum (w/v in PBS; Biological Canto II (BD) was operated with FACSDIVA software Industries). Incubation in the same blocking solution was (BD). At least 10,000 events were collected per sample. done for 1 h at RT. Primary antibodies, diluted in blocking solution, were as follows: anti-Nanog, anti-Nestin (1:500; Glutamate uptake assay BD Pharmingen), anti-GFAP-cy3 (mouse monoclonal anti- Glutamate uptake capability of the cells was measured in body (Mc), 1:500; Sigma), anti-GLAST (rabbit Mc, 1:100; 28-day differentiated hESC-derived astrocytes. Glutamic Miltenibiotec), anti-S100 (rabbit polyclonal antibody, 1:100; acid (0.5 mM; Sigma) in Hanks’ Balanced Salt Solution DAKO), anti-AQP-4 (rabbit, 1:2000; Mc Abcam) and (Gibco) was added to 1 × 10 cells/ml. After 0, 10, 30, 60 anti-Ki67 (rabbit, 1:50; Mc Cell Marque). After overnight and 120 min, the solution was aspirated and kept at 4 °C incubation at 4 °C, secondary antibody (1:200; Jackson until further testing. Human astrocytes derived from the Immuno Research) was added for 1 h at RT, followed spinal cord (from Thermo Scientific) served as positive by the nuclear fluorescent dye DAPI (0.05 μg/ml; control, while 0.5 mM glutamic acid kept at 37 °C for Sigma). Pictures were taken using Arrayscan VTI 120 min served as negative control. In addition, 0.5 mM (Thermo Scientific, Cellomics). glutamic acid kept at 4 °C for 120 min served as time 0 concentration control. The EnzyChrom™ Glutamate Assay Immunohistochemical staining Kit (BioAssay Systems) was used to measure the concen- Brain and spinal cord tissues were trimmed, decalcified tration of glutamate in the collected samples according to and embedded in paraffin, sectioned at approximately the manufacturer’s protocol and recommendations. The 5 μm thickness and stained with hematoxylin and eosin optical density was read at 565 nm using the iMark (H&E). For immune-cytofluorescence assays, tissues were Microplate reader (Bio Rad). Dihydrokainic acid (DHK, deparaffinized using the following washes: xylene (Sigma), 500 μM; Sigma) or 1 μM WAY-213,613 (Sigma) were used two washes × 5 min; 100% ethanol, two washes × 5 min; as inhibitors of GLT-1. 95% ethanol, one wash × 5 min; 70% ethanol, one wash × 5 min; and cold tap water, two washes × 5 min. Secretome analysis Heat-induced epitope retrieval was performed by boiling In order to promote astrocyte differentiation, APCs the sections in a domestic microwave, twice for 10 min, were deprived from growth factors (bFGF and EGF) using 100× H-3300 citrate-based solution (Vector Labora- and vitamin C was added for 7 days and 28 days. tories). Permeabilization was done by 0.5% Triton X-100 Conditioned media were collected after 48 h from each in blocking solution as described earlier, and incubation experimental well. The number of cells for each well continued in the same blocking solution for 1 h at RT. was counted (at least two replicas per each cell type) Primary mouse Mc antibody Stem123 or Stem121 (1:500; and secretome analysis was performed by multiplex Stem Cells) were added overnight and kept at 4 °C. ELISA using the human quantibody kiloplex Array Secondary antibody goat anti mouse Cy2 or Cy3 (1:200; (Raybiotech). The values obtained in secretome analysis Jackson Immuno Research) were added for 1 h at RT, 6 were normalized to 1 × 10 cells/ml. followed by the nuclear fluorescent dye DAPI (0.05 μg/ml; Sigma). G93A Transplantation of hES-AS in the hSOD1 animal model G93A Karyotype Transgenic hSOD1 mice aged 8–9 weeks of mixed The test was performed using spectral karyotyping gender (B6SJL-Tg(SOD1*G93A)1Gur/J) were purchased analysis (SKY) on cells from two APC banks (passages from The Jackson Laboratory (Bar Harbor, ME, USA; G93A 11 and 12). The analysis was performed by the Stem https://www.jax.org/). Transgenic hSOD1 rats aged G93A Cell Core and Advanced Cell Technologies Unit, 5–6 weeks of mixed gender (NTac:SD-Tg(SOD1 )L26H) Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 14 of 17 were purchased from Taconic Biosciences Inc. (Hudson, and lasting all throughout the duration of the experiment. NY, USA; http://www.taconic.com). Allanimalcareand CellCept was administered orally twice a day at a dose of surgical procedures described here were carried out accord- 15 mg/kg (total daily dose was 30 mg/kg). Dosing started ing to protocols approved by the Israeli National Commit- 3 days prior to the treatment and lasted for a total of 10 tee for Animal Care. The animals were kept in a certified consecutive days. Cohort 3, which was given the treatment animal facility in IVC cages with a light cycle of 12 h and at twice, started receiving CellCept 3 days prior to each treat- temperature of 22 ± 2 °C. Rodent diet and drinking water ment injection for 10 consecutive days. were provided ad libitum. Measurements Intrathecal injection through the cisterna magna Measurement of body weight and all motor tests took Mice were anesthetized with an i.p. injection of keta- place 7–10 days prior to cell transplantation and rou- mine/xylazine (K4138; Sigma) and then mounted on a tinely afterward. Motor function was tested using an stereotaxic frame. The head was then bent, resulting in acceleration Rotarod device (Rotarod 7650; Ugo Basile, nape distention. A midline skin incision was made at the Comerio, Italy) for the duration of 180 s. The time it nape area to expose the sagittal suture of the cranium took each mouse to fall from the rod was recorded. and midline of the nape. Under a dissection microscope, Animals were trained for 1 week prior to conducting the the subcutaneous tissue and muscles were separated by test. Forelimb muscle grip strength was determined blunt dissection with forceps to expose the cleft between using a Grip Strength Meter 47,200 (UGO Basile). Grip the occipital bone and the atlas vertebra. The muscles strength testing was performed by allowing the animals were held apart to expose the dura mater which was to grasp a thin bar attached to the force gauge. This is carefully penetrated using a 29G-gauge 45° beveled done by pulling the animal away from the gauge until needle (Hamilton, Reno, NV, USA) connected to a 10-μl the mice forelimbs released the bar. The procedure Hamilton syringe preloaded with 10 μl of cell suspension provides a value of the force of maximal grip strength. or vehicle (DMEM/F12 medium). Then 2 × 10 hES-AS The force measurements were recorded in three separate (APCs differentiated for 7 days) were injected once on trials, and the averages were used in the statistical day 67 ± 2 (CellsX1 group, n = 14 mice) or twice on day analysis. Neurological scoring was done according to 67 ± 2 and on day 97 ± 2 at interval of 30 days (CellsX2 neurological score on a scale from 0 to 5 [99]. group, n = 13), or injected with DMEM F12 (Sham group, n = 10) into the CSF through the CM. The Statistical analysis G93A syringe was held in position for 3 min before being grad- Kaplan–Meier analysis of the SOD1 mice and rats ually pulled away to avoid liquid outflow along the was conducted using the statistical software Sigmastat needle tract. The skin cut was secured with stainless (SAS Software) to analyze survival, disease onset and dur- steel surgical clips and wiped with 70% ethanol. ation data. Weight, time to fall from the Rotarod, neuro- logical score and grip strength results were analyzed via Injection of the cells by lumbar puncture repeated-measures ANOVA. All data are presented as The rats were anesthetized with ketamine/xylazine. The mean ± SEM, and significance level was set at p ≤ 0.05. lumbar region was shaved, sterilized with iodine and the Statistical analysis was performed by MediStat Ltd, Israel. intervertebral spaces widened by placing the animal on a 15-ml conical plastic tube. The injections were per- Transplantation of hES-AS in NSG mice formed by inserting a 29-gauge 45° beveled needle The mouse was mounted on a stereotaxic frame. A mid- (Hamilton) connected to a 10-μl Hamilton syringe into line skin incision was made at the nape area to expose the the tissues between the dorsal aspects of L5 and L6. sagittal suture of the cranium and midline of the nape. Correct subarachnoid positioning of the tip of the needle The head was then bent, resulting in nape distention. was verified by a tail flick test. A volume of 10 μl Under a dissection microscope, the subcutaneous tissue containing 3 × 10 APCs was injected twice on day 50 ± and muscles were separated by blunt dissection with for- 2 and on day 70 ± 2 (n = 7), or vehicle (DMEM/12 ceps to expose the cleft between the occipital bone and medium, n = 7) was injected. The syringe was held in the atlas vertebra. The muscles were held apart to expose position for 30 s before being progressively pulled away. the dura mater which was penetrated using a 29G needle connected to a Hamilton syringe, preloaded with 10 μlof Immunosuppression 0.4 × 10 hES-AS. The cells were injected within 30 s into Immunosuppression was used only in the transplantation the CSF space. The needle was held for about 30 s G93A experiment in SOD1 mice. In this experiment, Cyclo- after injection and then withdrawn. The skin cut was sporin A was given daily by intraperitoneal injection, at a secured with stainless steel surgical clips and wiped dose of 10 mg/kg, starting 3 days prior to the treatment with polydine solution. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 15 of 17 Additional files Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Additional file 1: Figure S1. hES-AS produce and secrete neurotrophic factors. Conditioned media of 24 h from cultures of hES-AS differentiated Author details for 28 days as well as cell extracts used to measure level of neurotrophic Neurodegenerative Diseases Department at Kadimastem Ltd, Pinchas Sapir factors GDNF, BDNF, VEGF and IGF-1. For each factor, bars show cell content, 7, Weizmann Science Park, Nes-Ziona, Israel. Department of Molecular amount secreted and negative control (medium only), expressed in pg/10 Genetics, Weizmann Institute of Science, 76100 Rehovot, Israel. cells (triplicates ± SD) (PDF 91 kb) Additional file 2: Table S1. Secretome analysis of hES-AS, differentiated for Received: 5 April 2018 Revised: 30 April 2018 7 days or 28 days. The 220 most secreted factors by the 7-day differentiated Accepted: 1 May 2018 hES-AS sorted by mean ng/ml/10 cells ± SD (PDF 140 kb) Additional file 3: Figure S2. Effect of hES-AS transplantation on disease References G93A onset, progression and survival in hSOD1 mice. hES-AS, differentiated 1. Rowland LP, Shneider NA. Amyotrophic lateral sclerosis. N Engl J Med. 2001; G93A for 7 days, transplanted intrathecally through CM of hSOD1 mice. A 344:1688–700. Three experimental groups tested, single injection of 2 × 10 hES-AS on 2. Rosen DR. Mutations in Cu/Zn superoxide dismutase gene are associated day 67 of life (Cellsx1), two injections of 2 × 10 hES-AS each on days 67 with familial amyotrophic lateral sclerosis. Nature. 1993;364:362. and 97 (Cellsx2) and once sham-injected mice (vehicle). Kaplan–Meir plot 3. Lagier-Tourenne C, Cleveland DW. Rethinking ALS: the FUS about TDP-43. of disease onset (measured by 3% body weight loss from maximal weight) Cell. 2009;136:1001–4. showing more delay in twice-injected group. B Kaplan–Meier survival curves 4. Renton AE, Chio A, Traynor BJ. State of play in amyotrophic lateral sclerosis with similar trends. C Body weight maintained longer in hES-AS-treated mice. genetics. Nat Neurosci. 2014;17:17–23. Note that a few days after second injection, day 97, weight loss occurred 5. Haidet-Phillips AM, et al. Astrocytes from familial and sporadic ALS patients related to injection. D Neurological score. E Significant improvement in motor are toxic to motor neurons. Nat Biotechnol. 2011;29:824–8. performance (Rotarod test) for hSOD1 mice transplanted twice with hES-AS. 6. Meyer K, et al. Direct conversion of patient fibroblasts demonstrates non- C, D Values are mean ± SEM (PDF 262 kb) cell autonomous toxicity of astrocytes to motor neurons in familial and Additional file 4: Table S2. Percent of cell presence and percent of sporadic ALS. Proc Natl Acad Sci U S A. 2014;111:829–32. frequency scores greater than, or equal to ‘2’ (one to three foci of 10-20 7. Di Giorgio FP, Carrasco MA, Siao MC, Maniatis T, Eggan K. Non-cell cells per foci) for each follow up time (4, 17 and 39 weeks after hES-AS autonomous effect of glia on motor neurons in an embryonic stem cell- transplantation). Supplementary materials and methods. (ZIP 150 kb) based ALS model. Nat Neurosci. 2007;10:608–14. 8. Nagai M, et al. Astrocytes expressing ALS-linked mutated SOD1 release factors selectively toxic to motor neurons. Nat Neurosci. 2007;10:615–22. Abbreviations 9. Di Giorgio FP, Boulting GL, Bobrowicz S, Eggan KC. Human embryonic stem ALS: Amyotrophic lateral sclerosis; APC: Astrocyte progenitor cell; AQP- cell-derived motor neurons are sensitive to the toxic effect of glial cells 4: Aquaporin-4; CM: Cisterna magna; CNS: Central nervous system; carrying an ALS-causing mutation. Cell Stem Cell. 2008;3:637–48. CSF: Cerebrospinal fluid; G93A mutation: Glycine 93 changed to alanine; 10. Marchetto MC, et al. Non-cell-autonomous effect of human SOD1 G37R GFAP: Glial Fibrillary Acidic Protein; GLP: Good laboratory practice; GLT- astrocytes on motor neurons derived from human embryonic stem cells. 1: Glutamate transporter 1; GMP: Good manufacturing practice; hES- Cell Stem Cell. 2008;3:649–57. AS: Human embryonic stem cell-derived astrocytes (differentiated from APCs 11. Papadeas ST, Kraig SE, O'Banion C, Lepore AC, Maragakis NJ. Astrocytes for 7 days); hESC: Human embryonic stem cell; hSOD1: Human superoxide carrying the superoxide dismutase 1 (SOD1G93A) mutation induce wild- dismutase 1; LOD: Limit of detection; LP: Lumbar puncture; MN: Motor type motor neuron degeneration in vivo. Proc Natl Acad Sci U S A. 2011; neuron; NTF: Neurotrophic factor; SOD1: Superoxide dismutase 1; 108:17803–8. TIMP: Tissue inhibitor of metalloproteinase; VEGF: Vascular endothelial 12. Yamanaka K, et al. Mutant SOD1 in cell types other than motor neurons and growth factor oligodendrocytes accelerates onset of disease in ALS mice. Proc Natl Acad Sci U S A. 2008;105:7594–9. Acknowledgements 13. Tong J, et al. Expression of ALS-linked TDP-43 mutant in astrocytes causes The authors would like to thank Science in Action and Envigo-IL animal non-cell-autonomous motor neuron death in rats. EMBO J. 2013;32:1917–26. facilities for conducting animal studies. 14. Rossi S, Cozzolino M, Carri MT. Old versus new mechanisms in the pathogenesis of ALS. Brain Pathol. 2016;26:276–86. Funding 15. Foran E, Trotti D. Glutamate transporters and the excitotoxic path to motor This work was supported by the Israel Innovation Authority National Natural neuron degeneration in amyotrophic lateral sclerosis. Antioxid Redox Signal. (grant No. 59652). 2009;11:1587–602. 16. Rothstein JD, Van Kammen M, Levey AI, Martin LJ, Kuncl RW. Selective loss Availability of data and materials of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann The data that support the findings of this study are available from the Neurol. 1995;38:73–84. corresponding author upon reasonable request. 17. Lin CL, et al. Aberrant RNA processing in a neurodegenerative disease: the cause for absent EAAT2, a glutamate transporter, in amyotrophic lateral Authors’ contributions sclerosis. Neuron. 1998;20:589–602. RM, IM, SSG, CJ and HA, JI-E conceived and designed the studies. TA, LC, GA, 18. Howland DS, et al. Focal loss of the glutamate transporter EAAT2 in a KPL, LN, SYI, SJL, ZR, ZA and EV performed the experiments. KG was respon- transgenic rat model of SOD1 mutant-mediated amyotrophic lateral sible for quality assurance. IM and SSG analyzed the data. IM, SSG, RM and CJ sclerosis (ALS). Proc Natl Acad Sci U S A. 2002;99:1604–9. wrote the manuscript. SSG and IM interpreted the data and reviewed the 19. Raoul C, et al. Motoneuron death triggered by a specific pathway manuscript. All authors read and approved the final manuscript. downstream of Fas. potentiation by ALS-linked SOD1 mutations. Neuron. 2002;35:1067–83. Ethics approval and consent to participate 20. Barbeito LH, et al. A role for astrocytes in motor neuron loss in amyotrophic This study was performed following an application-form review by the lateral sclerosis. Brain Res Brain Res Rev. 2004;47:263–74. National Council for Animal Experimentation and after receiving approval 21. Endo F, et al. Astrocyte-derived TGF-β1 accelerates disease progression in (No. IL-16-09-279) that the study complies with the rules and regulations ALS mice by interfering with the neuroprotective functions of microglia and set forth. T cells. Cell Rep. 2015;11:592–604. 22. Tripathi P, et al. Reactive astrocytes promote ALS-like degeneration and Competing interests intracellular protein aggregation in human motor neurons by disrupting The authors declare that they have no competing interests. autophagy through TGF-β1. Stem cell reports. 2017;9:667–80. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 16 of 17 23. Phatnani, H.P., et al. Intricate interplay between astrocytes and motor and serum of patients with early Alzheimer's disease and normal controls. J neurons in ALS. Proc Natl Acad Sci U S A 110, E756–EE765 (2013). Alzheimers Dis. 2009;18:331–7. 24. Brambilla L, et al. Disruption of the astrocytic TNFR1-GDNF axis accelerates 50. Li G, et al. Cerebrospinal fluid concentration of brain-derived neurotrophic factor motor neuron degeneration and disease progression in amyotrophic lateral and cognitive function in non-demented subjects. PLoS One. 2009;4:e5424. sclerosis. Hum Mol Genet. 2016;25:3080–95. 51. Gurney ME, et al. Motor neuron degeneration in mice that express a human 25. Das MM, Svendsen CN. Astrocytes show reduced support of motor neurons Cu, Zn superoxide dismutase mutation. Science. 1994;264:1772–5. with aging that is accelerated in a rodent model of ALS. Neurobiol Aging. 52. Writing G, Edaravone ALSSG. Safety and efficacy of edaravone in well 2015;36:1130–9. defined patients with amyotrophic lateral sclerosis: a randomised, double- 26. Van Den Bosch L, et al. Effects of vascular endothelial growth factor (VEGF) blind, placebo-controlled trial. The Lancet Neurology. 2017;16:505–12. on motor neuron degeneration. Neurobiol Dis. 2004;17:21–8. 53. Bensimon G, Lacomblez L, Meininger V. A controlled trial of riluzole in 27. Oosthuyse B, et al. Deletion of the hypoxia-response element in the amyotrophic lateral sclerosis. ALS/Riluzole Study Group. N Engl J Med. 1994; vascular endothelial growth factor promoter causes motor neuron 330:585–91. degeneration. Nat Genet. 2001;28:131–8. 54. Lacomblez L, Bensimon G, Leigh PN, Guillet P, Meininger V. Dose-ranging 28. Bogaert E, et al. VEGF protects motor neurons against excitotoxicity by study of riluzole in amyotrophic lateral sclerosis. Amyotrophic Lateral upregulation of GluR2. Neurobiol Aging. 2010;31:2185–91. Sclerosis/Riluzole Study Group II. Lancet. 1996;347:1425–31. 29. Gros-Louis F, et al. Absence of mutations in the hypoxia response element 55. Miller RG, Mitchell JD, Moore DH. Riluzole for amyotrophic lateral sclerosis of VEGF in ALS. Muscle Nerve. 2003;28:774–5. (ALS)/motor neuron disease (MND). The Cochrane database of systematic reviews. 2012;14(3):CD001447. 30. Llado J, Tolosa L, Olmos G. Cellular and molecular mechanisms involved in the neuroprotective effects of VEGF on motoneurons. Front Cell Neurosci. 2013;7:181. 56. Petrov D, Mansfield C, Moussy A, Hermine O. ALS clinical trials review: 20 31. Mishra PS, et al. Astroglia acquires a toxic neuroinflammatory role in years of failure. are we any closer to registering a new treatment? Front response to the cerebrospinal fluid from amyotrophic lateral sclerosis Aging Neurosci. 2017;9:68. patients. J Neuroinflammation. 2016;13:212. 57. Izrael M, Slutsky S, Itskovitz-Eldor J. & Revel M. Astrocytes in Pathogenesis of 32. Re DB, et al. Necroptosis drives motor neuron death in models of both Neurodegenerative Diseases and Potential: Translation into Clinic. InTechOpen: sporadic and familial ALS. Neuron. 2014;81:1001–8. Astrocyte—Physiology and Pathology, ISBN 978–953–51-5760-1 (Mar 2018). 33. Hitomi J, et al. Identification of a molecular signaling network that regulates 58. Lee J, et al. Astrocytes and microglia as non-cell autonomous players in the a cellular necrotic cell death pathway. Cell. 2008;135:1311–23. pathogenesis of ALS. Experimental neurobiology. 2016;25:233–40. 59. Do-Ha D, Buskila Y, Ooi L. Impairments in motor neurons, interneurons and 34. Wiedemann FR, Manfredi G, Mawrin C, Beal MF, Schon EA. Mitochondrial astrocytes contribute to hyperexcitability in ALS: underlying mechanisms DNA and respiratory chain function in spinal cords of ALS patients. J Neurochem. 2002;80:616–25. and paths to therapy. Mol Neurobiol. 2017;55(2):1410–18. 35. Hirano A, et al. Fine structural study of neurofibrillary changes in a 60. Storkebaum E, Lambrechts D, Carmeliet P. VEGF: once regarded as a specific family with amyotrophic lateral sclerosis. J Neuropathol Exp Neurol. angiogenic factor, now implicated in neuroprotection. BioEssays. 2004;26:943–54. 1984;43:471–80. 61. Krakora D, et al. Synergistic effects of GDNF and VEGF on lifespan and 36. Sasaki S, Horie Y, Iwata M. Mitochondrial alterations in dorsal root ganglion cells disease progression in a familial ALS rat model. Molecular therapy : the in sporadic amyotrophic lateral sclerosis. Acta Neuropathol. 2007;114:633–9. journal of the American Society of Gene Therapy. 2013;21(8):1602–10. 37. Tan W, et al. Small peptides against the mutant SOD1/Bcl-2 toxic 62. Ramamohan PY, et al. Cerebrospinal fluid from amyotrophic lateral sclerosis mitochondrial complex restore mitochondrial function and cell viability in patients causes fragmentation of the Golgi apparatus in the neonatal rat spinal mutant SOD1-mediated ALS. The Journal of neuroscience : the official cord. Amyotrophic lateral sclerosis : official publication of the World Federation journal of the Society for Neuroscience. 2013;33:11588–98. of Neurology Research Group on Motor Neuron Diseases. 2007;8:79–82. 38. Nicaise C, Mitrecic D, Falnikar A, Lepore AC. Transplantation of stem cell- 63. Deepa P, et al. Down regulation of trophic factors in neonatal rat spinal derived astrocytes for the treatment of amyotrophic lateral sclerosis and cord after administration of cerebrospinal fluid from sporadic amyotrophic spinal cord injury. World J Stem Cells. 2015;7:380–98. lateral sclerosis patients. J Neural Transm. 2011;118:531–8. 39. Tannenbaum SE, et al. Derivation of xeno-free and GMP-grade human 64. Shruthi S, et al. Brain-derived neurotrophic factor facilitates functional embryonic stem cells–platforms for future clinical applications. PLoS One. recovery from ALS-cerebral spinal fluid-induced neurodegenerative changes 2012;7:e35325. in the NSC-34 motor neuron cell line. Neurodegener Dis. 2017;17:44–58. 40. Prathalingam N, et al. Production and validation of a good manufacturing 65. Rosito M, Deflorio C, Limatola C, Trettel F. CXCL16 orchestrates adenosine practice grade human fibroblast line for supporting human embryonic stem A3 receptor and MCP-1/CCL2 activity to protect neurons from excitotoxic cell derivation and culture. Stem Cell Res Ther. 2012;3:12. cell death in the CNS. The Journal of neuroscience : the official journal of 41. Izrael M, et al. Human oligodendrocytes derived from embryonic stem cells: the Society for Neuroscience. 2012;32:3154–63. effect of noggin on phenotypic differentiation in vitro and on myelination 66. Christopherson KS, et al. Thrombospondins are astrocyte-secreted proteins in vivo. Mol Cell Neurosci. 2007;34:310–23. that promote CNS synaptogenesis. Cell. 2005;120:421–33. 42. Scholze AR, Foo LC, Mulinyawe S, Barres BA. BMP signaling in astrocytes 67. Tyzack GE, et al. Astrocyte response to motor neuron injury promotes structural downregulates EGFR to modulate survival and maturation. PLoS One. 2014; synaptic plasticity via STAT3-regulated TSP-1 expression. Nat Commun. 2014;5:4294. 9:e110668. 68. Hafner A, et al. Neuroprotective role of gamma-enolase in microglia in a 43. Alfei L, et al. Hyaluronate receptor CD44 is expressed by astrocytes in the mouse model of Alzheimer's disease is regulated by cathepsin X. Aging Cell. 2013;12:604–14. adult chicken and in astrocyte cell precursors in early development of the chick spinal cord. Eur J Histochem. 1999;43:29–38. 69. Morisaki Y, et al. Selective expression of Osteopontin in ALS-resistant motor 44. Piao C, et al. Thrombin decreases expression of the glutamate transporter neurons is a critical determinant of late phase neurodegeneration mediated GLAST and inhibits glutamate uptake in primary cortical astrocytes via the by matrix metalloproteinase-9. Sci Rep. 2016;6:27354. Rho kinase pathway. Exp Neurol. 2015;273:288–300. 70. Kaplan A, et al. Neuronal matrix metalloproteinase-9 is a determinant of 45. Baba H, et al. GFAP gene expression during development of astrocyte. Dev selective neurodegeneration. Neuron. 2014;81:333–48. Neurosci. 1997;19:49–57. 71. Wright MC, et al. Novel roles for osteopontin and clusterin in peripheral 46. Hubbard JA, Hsu MS, Seldin MM, Binder DK. Expression of the astrocyte motor and sensory axon regeneration. The Journal of neuroscience : the water channel Aquaporin-4 in the mouse brain. ASN Neuro. 2015;7(5). official journal of the Society for Neuroscience. 2014;34:1689–700. 47. Gropp M, et al. Standardization of the teratoma assay for analysis of 72. Meller R, et al. Neuroprotection by osteopontin in stroke. Journal of cerebral blood flow and metabolism : official journal of the International Society of pluripotency of human ES cells and biosafety of their differentiated Cerebral Blood Flow and Metabolism. 2005;25:217–25. progeny. PLoS One. 2012;7:e45532. 48. Kawahara K, et al. Selective blockade of astrocytic glutamate transporter 73. Gardner J, Ghorpade A. Tissue inhibitor of metalloproteinase (TIMP)-1: the GLT-1 with dihydrokainate prevents neuronal death during ouabain TIMPed balance of matrix metalloproteinases in the central nervous system. treatment of astrocyte/neuron cocultures. Glia. 2002;40:337–49. J Neurosci Res. 2003;74:801–6. 49. Straten G, Eschweiler GW, Maetzler W, Laske C, Leyhe T. Glial cell-line 74. Israelson A, et al. Macrophage migration inhibitory factor as a chaperone derived neurotrophic factor (GDNF) concentrations in cerebrospinal fluid inhibiting accumulation of misfolded SOD1. Neuron. 2015;86:218–32. Izrael et al. Stem Cell Research & Therapy (2018) 9:152 Page 17 of 17 75. Cordero-Llana O, et al. Clusterin secreted by astrocytes enhances neuronal differentiation from human neural precursor cells. Cell Death Differ. 2011;18:907–13. 76. Winkler C, Yao S. The midkine family of growth factors: diverse roles in nervous system formation and maintenance. Br J Pharmacol. 2014;171:905–12. 77. Philips T, Rothstein JD. Rodent models of amyotrophic lateral sclerosis. Current protocols in pharmacology. 2015;69:5 67 61-65 67 21. 78. Collins MA, An J, Hood BL, Conrads TP, Bowser RP. Label-free LC-MS/MS proteomic analysis of cerebrospinal fluid identifies protein/pathway alterations and candidate biomarkers for amyotrophic lateral sclerosis. J Proteome Res. 2015;14:4486–501. 79. Tarasiuk J, Kulakowska A, Drozdowski W, Kornhuber J, Lewczuk P. CSF markers in amyotrophic lateral sclerosis. J Neural Transm. 2012;119:747–57. 80. Gao L, Zhou S, Cai H, Gong Z, Zang D. VEGF levels in CSF and serum in mild ALS patients. J Neurol Sci. 2014;346(1-2):216–20. 81. Varghese AM, et al. Chitotriosidase—a putative biomarker for sporadic amyotrophic lateral sclerosis. Clin Proteomics. 2013;10:19. 82. Przyborski SA. Differentiation of human embryonic stem cells after transplantation in immune-deficient mice. Stem Cells. 2005;23:1242–50. 83. Andrews PW, et al. Embryonic stem (ES) cells and embryonal carcinoma (EC) cells: opposite sides of the same coin. Biochem Soc Trans. 2005;33:1526–30. 84. Lee AS, et al. Effects of cell number on teratoma formation by human embryonic stem cells. Cell Cycle. 2009;8:2608–12. 85. Prokhorova TA, et al. Teratoma formation by human embryonic stem cells is site dependent and enhanced by the presence of Matrigel. Stem Cells Dev. 2009;18:47–54. 86. Ramot Y, et al. Compact MRI for the detection of teratoma development following intrathecal human embryonic stem cell injection in NOD-SCID mice. Neurotoxicology. 2017;59:27–32. 87. Priest CA, Manley NC, Denham J, Wirth ED 3rd, Lebkowski JS. Preclinical safety of human embryonic stem cell-derived oligodendrocyte progenitors supporting clinical trials in spinal cord injury. 2015;10(8):939–58. 88. Staff NP, et al. Safety of intrathecal autologous adipose-derived mesenchymal stromal cells in patients with ALS. Neurology. 2016;87:2230–4. 89. Oh KW, et al. Phase I trial of repeated intrathecal autologous bone marrow- derived mesenchymal stromal cells in amyotrophic lateral sclerosis. Stem Cells Transl Med. 2015;4:590–7. 90. Tian C, et al. Autologous bone marrow mesenchymal stem cell therapy in the subacute stage of traumatic brain injury by lumbar puncture. Experimental and clinical transplantation : official journal of the Middle East Society for Organ Transplantation. 2013;11:176–81. 91. Wang S, et al. Umbilical cord mesenchymal stem cell transplantation significantly improves neurological function in patients with sequelae of traumatic brain injury. Brain Res. 2013;1532:76–84. 92. Roybon L, et al. Human stem cell-derived spinal cord astrocytes with defined mature or reactive phenotypes. Cell Rep. 2013;4:1035–48. 93. Gupta K, Chandran S, Hardingham GE. Human stem cell-derived astrocytes and their application to studying Nrf2-mediated neuroprotective pathways and therapeutics in neurodegeneration. Br J Clin Pharmacol. 2013;75:907–18. 94. Petrou P, et al. Safety and clinical effects of mesenchymal stem cells secreting neurotrophic factor transplantation in patients with amyotrophic lateral sclerosis: results of phase 1/2 and 2a clinical trials. JAMA neurology. 2016;73:337–44. 95. Glass JD, et al. Lumbar intraspinal injection of neural stem cells in patients with amyotrophic lateral sclerosis: results of a phase I trial in 12 patients. Stem Cells. 2012;30:1144–51. 96. Mazzini L, et al. Human neural stem cell transplantation in ALS: initial results from a phase I trial. J Transl Med. 2015;13:17. 97. Kondo T, et al. Focal transplantation of human iPSC-derived glial-rich neural progenitors improves lifespan of ALS mice. Stem cell reports. 2014;3:242–9. 98. Nizzardo M, et al. iPSC-derived LewisX+CXCR4+β1-integrin+ neural stem cells improve the amyotrophic lateral sclerosis phenotype by preserving motor neurons and muscle innervation in human and rodent models. Hum Mol Genet. 2016;25:3152–63. 99. Scott S, et al. Design, power, and interpretation of studies in the standard murine model of ALS. Amyotrophic lateral sclerosis : official publication of the World Federation of Neurology Research Group on Motor Neuron Diseases. 2008;9:4–15.

Journal

Stem Cell Research & TherapySpringer Journals

Published: Jun 6, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off