Human Motor Neurons Generated from Neural Stem Cells Delay Clinical Onset and Prolong Life in ALS Mouse Model

Amyotrophic lateral sclerosis (ALS) is the most common adult onset motor neuron disease. The etiology and pathogenic mechanisms of the disease remain unknown, and there is no effective treatment. Here we show that intrathecal transplantation of human motor neurons derived from neural stem cells (NSCs) in spinal cord of the SOD1G93A mouse ALS model delayed disease onset and extended life span of the animals. When HB1.F3.Olig2 (F3.Olig2) cells, stable immortalized human NSCs encoding the human Olig2 gene, were treated with sonic hedgehog (Shh) protein for 5–7 days, the cells expressed motor neuron cell type-specific phenotypes Hb9, Isl-1 and choline acetyltransferase (ChAT). These F3.Olig2-Shh human motor neurons were transplanted intrathecally in L5–L6 spinal cord of SOD1G93A mice, and at 4 weeks post-transplantation, transplanted F3.Olig2-Shh motor neurons expressing the neuronal phenotype markers NF, MAP2, Hb9, and ChAT were found in the ventral horn of the spinal cord. Onset of clinical signs in ALS mice with F3.Olig2-Shh motor neuron implants was delayed for 7 days and life span of animals was significantly extended by 20 days. Our results indicate that this treatment modality of intrathecal transplantation of human motor neurons derived from NSCs might be of value in the treatment of ALS patients without significant adverse effects.


Introduction
Amyotrophic lateral sclerosis (ALS) is a relentlessly progressive, adult onset neurodegenerative disease characterized by degeneration and loss of motor neurons in the cerebral cortex, brain stem and spinal cord, leading to muscle wasting and weakness, and eventually to death within five years after clinical onset [1]. The proposed pathogenetic mechanisms of ALS, albeit not fully elucidated, include oxidative stress, protein aggregation, mitochondrial dysfunction, impaired axonal transport, glutamatemediated excitotoxicity, and insufficient supply of neurotrophic factors [2]. To date there is no effective treatment.
Stem cell-based cell therapy is one of the most promising approaches for the treatment of neurological diseases including ALS [3][4][5][6]. Recent studies have indicated that it is possible to generate motor neurons in culture from several types of stem cells, including embryonic stem cells (ESCs), induced pluripotent stem cells (iPSCs) and neural stem cells (NSCs) [7][8][9][10]. Mouse ESCderived motor neurons transplanted into motor neuron-injured rat spinal cord survived and extended axons into ventral root [8 9], and human ESCs transplanted into cerebrospinal fluid of rats with motor neuron injury migrated into the spinal cord and led to improved motor function [11].
Previous studies have demonstrated that delivery of vascular endothelial cell growth factor (VEGF) significantly delayed disease onset and prolonged the survival of ALS animal models [12,13], and we have previously demonstrated that human NSCs overexpressing VEGF transplanted in spinal cord of transgenic SOD1G93A mice induced functional improvement, delayed disease onset for 7 days and extended survival of animals for 15 days [14].
In the present study, we wish to establish proof of prnciple that transplantation of human motor neurons generated from NSCs into spinal cords of SOD1G93A mice can lead to clinical improvement and extend life in this mouse model of ALS.

Ethics Statement
Use of fetal brain tissue collected for research purpose was approved by the Clinical Research Screening Committee and the Internal Review Board of the University of British Columbia (For preparation of immortalized human NSC line used in the present study). Pregnant woman gave a written informed consent for clinical procedure and research use of the embryonic tissue in accordance with the declaration of Helsinki. Use of laboratory animals for the study was approved by the Chung-Ang University Animal Care Committee and was accordance with the Guide for the care and use of laboratory animals as published by the US National Institute of Health.

Establishment of F3 Human NSCs Encoding Olig2 Transcription Factor
Primary cultures of dissociated human fetal telencephalon (15 weeks gestation) were prepared as reported previously [15,16]. The brain cells were transfected with a retroviral vector encoding v-myc and selected by neomycin resistance. One of the isolated clones, HB1.F3 (F3) human NSC line, which was expanded for the present study expresses NSC-specific markers, ABCG2, nestin and Musashi-1 [15,16].

Formation of Neuromuscular Junctions
Thigh muscle isolated from neonatal ICR mice was incubated in PBS containing 0.25% trypsin for 20 min at 37uC, washed in PBS, and dissociated into single cells by repeated pipetting. Dissociated muscle cells were suspended in DMEM with high glucose containing 10% FBS, 2 mM L-glutamine and 20 mg/mL gentamicin, and plated on gelatin-coated Aclar cover slips (12 mm round) at low cell density. Two days later, the medium was supplemented with cytosine arabinoside (Ara-C; 10 mM, Sigma) and cultured for another 2 days to eliminate dividing fibroblasts. Two days later F3.Olig2 NSCs treated with 100 ng/mL sonic hedgehog (Shh) protein for 7 days were seeded on top of the muscle cell culture to induce neuromuscular junction formation between F3.Olg2-Shh human NSCs and muscle cells.
In brief, spinal cord sections were incubated with primary antibodies overnight at 4uC, followed by secondary antibodies conjugated with Alexa Fluor-488 or -594 (1:400, Molecular Probes) for 2 hrs at RT.

Electrophysiological Recordings
Electrophysiological measurements of differentiated F3.Olig2-Shh cells were made by the tight-seal voltage clamp method, and whole-cell currents were recorded with an Axopatch 200B patch  clamp amplifier (Axon Instruments, Union City, CA). pCLAMP 9.0 software (Axon Instruments) was used for data acquisition and analysis.

Behavioral Tests
Motor strength and motor coordination were evaluated on a Rotarod (Daejong Instrument, Seoul, Korea) after a 1 week training period. A mouse running time of 300 seconds was selected as the arbitrary cut-off time, measured on the Rotarod rotating at a constant speed of 15 rpm. The animals performed the test twice per week until they no longer could perform the task. Paw grip endurance (PaGE) was used as the index of a mouse's grip strength. The wire-lid was gently shaken to prompt the mouse to hold onto the grid before the lid was swiftly turned upside down. The latency time for the mouse to let go off the grid with at least both hind limbs was measured. Each mouse was allowed up to three attempts to hold onto the inverted lid for an arbitrary maximum of 200 seconds, and the longest latency was recorded. An extension reflex was evaluated by the following 4-point scoring system: 4 for normal extension reflex of all hind limbs with balance, 3 for imbalance of extension, 2 for the extension reflex of only one hind limb, 1 for the absence of any hind limb extension and 0 for the total paralysis. The time of death was defined as the date on which the mouse could no longer roll over within 30 seconds after being placed on its side [14].

Quantitative Cell Counts
Total number of anti-human mitochondria antibody (hMit)positive F3.Olig2 and F3.Olig2-Shh cells in spinal cord sections was determined by stereological estimation. Counting was performed in whole spinal cord section areas divided into gray matter and ventral horn. Totally 30 sections taken from the serial section sets at an equal distance (1 mm) from cervical to lumber levels of spinal cord were counted. The estimate of the total number of hMit-positive cells were calculated using the optical fractionator formula [20].

Statistical Analysis
All statistical analyses were performed using the statistics software package SPSS (version 12, SPSS, Chicago, IL). All values are expressed as means 6 SEM. Differences among experimental groups were evaluated by a one-way ANOVA followed by a Tukey's posthoc analysis and Kaplan-Meier survival curves and survival times of groups of mice were compared using a log-rank test. Significance for all statistical analysis was set at p,0.05.

Results
Characterization of F3.Olig2 and F3.Olig2-Shh Human NSCs F3.Olig2 and F3.Olig2-Shh human NSCs exhibited multiple thin-branched cytoplasmic processes in a phase-contrast image as in F3 parental NSCs (Fig. 1A), RT-PCR analysis confirmed expression of Olig2 mRNA in F3.Olig2 and F3.Olig2-Shh cells (Fig. 1B). Pax6, a transcription factor which directly regulates the differentiation and maturation of oligodendrocytes, was not expressed in parental F3 NSCs but was newly expressed in F3.Olig2 cells and F3.Olig2-Shh cells after introduction of the Olig2 gene (Fig. 1B). Expression of Nkx-6.1, another transcription factor important in differentiation of motor neurons, was elevated in F3.Olig2-Shh cells (Fig. 1B). Expression of nestin, a cell typespecific marker for NSCs, was found in both F3 cells and F3.Olig2 cells but not in F3.Olig2-Shh cells. CNPase, a cell type-specific marker for oligodendrocytes, was demonstrated in F3.Olig2 cells but not in F3.Olig2-Shh cells indicating that the Shh treatment of F3.Olig2 cells annuls expression of a cell type specific marker of oligodendrocyte but newly induces motor neuron phenotype markers in F3.Olig2-Shh cells (Fig. 1B). Specifically, expression of three cell type-specific markers for motor neurons, Hb9, Isl-1 and ChAT, was demonstrated in F3.Olig2-Shh cells following 7 days of Shh treatment but not in F3 or F3.Olig2 cells (Fig. 1B). These RT-PCR results are also confirmed by immunocytochemical staining of ChAT and Hb9 immunorection in F3.Olig2-Shh cells. These results indicate that Shh induced gene expression in the pattern of normal spinal cord development, generating motor neuronal phenotypes from the NSCs expressing Olig2.
Previous studies have demonstrated that delivery of trophic factors such as VEGF significantly delayed disease onset and prolonged the survival of ALS animals [12,13].
We determined in vitro expression profiles of neurotrophic factors including VEGF in F3.Olig2-Shh cells, parental F3 and F3.Olig2 cells. F3.Olig2-Shh cells are shown to express high levels of GDNF and HGF, while F3 parental cells express both factors very low (Fig. 1C). F3 parental cells expressed high levels of NGF and VEGF, but F3.Olig2-Shh cells do not express these important neurotrophic factors. It is worthy to note that F3.Olig2-Shh cells express considerably high level of BDNF (Fig. 1C) which is known for it's significant neuroprotective function in neurodegenerative diseases, ischemia and brain injury.

F3.Olig2-Shh NSCs Migrate to the ALS Mouse Spinal Cord Ventral Horn
We performed immunohistochemical staining with anti-human moitochondria (hMit)-antibody to localize the transplanted F3.Olig2-Shh human cells and co-stained them with anti-MAP2 or anti-NF antibodies to identify the migration of F3.Olig2-Shh

Discussion
The present study shows that human NSCs transduced with Olig2 gene (F3.Olig2) express motor neuron-specific phenotypes that include Hb9, Isl-1 and ChAT following treatment with Shh protein. Intrathecal transplantation of F3.Olig2-Shh motor neurons into the L5-L6 lumbar spinal cord of SOD1G93A mice migrated into the ventral horn area and improved ambulatory functions, including extension reflex, paw grip endurance (PaGE) and rotarod test. Our findings are in good agreement with our earlier results that the clinical onset was delayed and life span of animals extended for 17 days in SOD1G93A mice after intrathecal grafting of the same human NSCs genetically modified to produce vascular endothelial growth factor (F3.VEGF) [14].
During the past years, stem cell replacement therapy has shown great promise as a prospective therapeutic strategy for ALS. Motor neurons derived from mouse ES cells that were transplanted into motor neuron-injured rats survived and extended axons into ventral roots [8]. Human ES cells also differentiated into motor neurons upon treatment with Shh and retinoic acid and transplanted into spinal cord of chick embryos and mature rats [21]. Mesenchymal stem cells (MSCs) derived from bone marrow or cord blood were proposed as candidates for cell replacement therapy in ALS because they could be induced to form neurons and to serve for autologous transplantation [22][23][24][25]. However, rather than differentiating into motor neurons and replacing comparable degenerated neurons of the host, it appears that the main role of the injected MSCs was to rescue many host neurons from impending cell death by provision of neurotrophic and survival factors [26][27][28]. It is worthy to note that F3.Olig2-Shh motor neurons in the present study are also capable of inducing neuroprotection of host neurons in SOD1G93A mutant mouse spinal cord by secreting neurotrophic factors such as BDNF, GDNF and HGF (Fig. 1C). Motor neurons were also developed from iPSCs isolated from an ALS patient [10]. Neurons and glia induced from patient-derived iPSCs are autologous, readily accessible, and not subject to immune rejection or ethical problems. Such patient-derived neurons could become an ideal cellular source for the screening of new drug candidates. A recent U.S. phase I trial of intraspinal injections of the same human NSCs as the rat ALS study was carried out in 12 ALS patients. In each patient 10 injections (100,000 cells per injection) were made into the lumbar spinal cord, with clinical assessments ranging from 6 to 18 months after transplantation, and no acceleration of disease progression was detected [29]. A recent metaanalysis study has reported that transplantation of NSCs, both mouse and human, in SOD1G93A transgenic mice delayed onset and progression of clinical signs and prolonged life span of animals. The beneficial effects of NSC transplantation appear to be mediated by NSCs' ability to produce neurotrophic factors, preserve neuromuscular function, and reduce astrogliosis and inflammation [30]. It is interesting to note that less than 5-10% of transplanted NSCs were found to differentiate into neurons or glia and 90 to 95% of grafted cells remain as quiescent nestin-positive neural progenitor cells [30]. In contrast to the previous study, our results indicate that better than 90% of hMit-positive human F3.Olig2-Shh cells in the SOD1G93A mouse spinal cord sections are doubly positive with NF or MAP2, both cell type-specific markers for neurons. For this reason, it is beneficial to transplant motor neurons terminally differentiated from NSCs in spinal cord of ALS animals as reported in the present study as contrasted with NSC transplantation reported by others [30].
In the developing ventral neural tube, Shh secreted by axial midline cells of the notochord and floor plate become distributed along a gradient that provides positional information that induces the apperance of motor neurons at defined positions in the ventral spinal cord [31,32]. Shh regulates homeodomain proteins that identify five domains of progenitors. One progenitor domain, pMN, produces somatic motor neurons identified by expression of two homeodomain proteins, Pax 6 and Nkx 6.1. The expression of Nkx 6.1 is up-regulated by high concentration of Shh at pMN, V2, V1 and V0 domains in the ventral tube [33]. Our observations of gene expression profiles in F3.Olig2-Shh cells and the capability of F3.Olig2 cells to differentiate into motor neurons following treatment with Shh treatment are in good agreement with the previously reported phenotypic expression of pMN motor neurons [31][32][33].
Since ALS is a multifocal disease affecting various parts of the CNS including spinal cord and motor cortex, it is important to determine the most appropriate route for stem cell transplantation. Intrathecal injection is less invasive than intraspinal injection, and delivery via cerebrospinal fluid (CSF) is more extensive than by direct intraspinal injection. Intravenous transplantation could give broader delivery via the vascular system, but the injected cells come to reside in unexpected tissues such as spleen, liver, lung and kidney [34]. Thus, we transplanted F3.Olig2-Shh cells intrathecailly, and verified the presence of injected cells within the spinal cord. In a previous study, we transplanted intrathecally F3.VEGF human NSCs (same parental origin as F3.Olig2, but expressing the VEGF gene) in SOD1G93A ALS mice and found clinical onset delay and extended life span in the transplanted animals [14]. VEGF is an angiogenic growth factor acting as both a mitogen and a chemoattractive factor for endothelial cells, and is known also to have neuroprotective effects against brain injury. Of the transplanted F3.VEGF NSCs, 12.3% were found within spinal cord gray matter, compare with only 1.1% of the parental F3 NSC controls. In the present study, 17.1% of injected F3.Olig2-Shh cells were found within gray matter, most of them localized in the ventral horn (Fig. 7).
In conclusion, F3.Olig2-Shh human motor neuron -like cells intrathecally transplanted into spinal cords of ALS model mice were capable of replacing lost motor neurons, and led to behavioral improvement and prolonged survival of transplanted ALS mice. Consequently, combination of Olig2 gene transduction and Shh treatment can induce motor neuronal differentiation of human NSCs, and these NSC-derived motor neurons might serve as a source of cell therapy for ALS patients.