Introduction

Spinal muscular atrophies (SMAs) include a series of neuromuscular genetic diseases in which spinal motor neurons degenerate leading to progressive paralysis with proximal muscular atrophy. The most frequent condition is caused by mutations of the survival motor neuron 1 ( SMN1 ) gene on chromosome 5q13 . The human genome also contains a highly homologous gene on the same chromosome, called SMN2. It differs from SMN1 by the presence of a C→T nucleotidic replacement in a critical position on exon 7. This results in an altered splicing of the transcript leading to the production of a very small amount of SMN full-length functional protein.

The 5q form of SMA represents the most common genetic cause of death during childhood affecting about 1 in 10,000 newborns. SMA clinical manifestations include muscular weakness and atrophy especially affecting lower limbs and proximal segments. Muscular wasting leads to severe difficulties in breathing and swallowing; death usually occurs due to cardiorespiratory failure. The number of copies of the paralogous gene, SMN2 , which can contribute only a small amount of SMN full-length protein, appears to determine the severity of the clinical picture. The SMA type 1 form represents the most frequent and severe phenotype; affected children manifest the disorder within 5–6 months after birth, and never reach motor-developmental milestones. They cannot sit or walk autonomously, and require early ventilator support. SMA type 2 displays a milder clinical phenotype; patients become symptomatic between 6 and 18 months from birth and are able to sit independently. SMAs 3 and 4 become overt later (the latter during adulthood), and while patients display symptoms related to muscular weakness and altered motor performance, their life expectancy is comparable to the general population. There is no effective treatment clinically available for SMA patients, who require a 24-h assistance with specific sanitary devices. A multidisciplinary environment in which clinicians, physiotherapists, psychologists, and speech therapists work collectively currently represents the most proper supportive approach. Due to its high frequency, monogenic etiology, and severe clinical and social burden, SMA represents a valuable target for experimental therapeutic approaches.

Potential Therapeutic Effects of Stem Cells on Spinal Muscular Atrophy Disease Mechanisms

Although the genetic origin of SMA has been revealed during the last decade, many questions still remain about downstream SMN1 mutation molecular pathogenic mechanisms. The precise physiological role of SMN protein is the focus of several ongoing researches. SMN appears to complex with Gemin proteins forming a structure that cooperates into the biogenesis of uridine-rich small nuclear ribonuclear proteins. They are a crucial contributor to spliceosome formation and splicing process. Other functions have been attributed to SMN including a role in axon maturation and myelination. The rather selective vulnerability of motor neurons to SMN reduction remains a major issue to be addressed.

Several research groups have exploited the murine model to investigate pathological mechanisms involved in SMA development. One model has been obtained by introducing the human SMN2 transgene on a Smn-null background. The so-called SMNΔ7 mice mimic SMA1 human phenotype in terms of neuromuscular waste, with a rather selective involvement of spinal motor neurons. However, SMNΔ7 animals display a systemic multiorgan involvement with distal tissue necrosis, which is more severe than it appears to be in humans, and it significantly affects mouse life span. This and other differences are rather species-specific and need to be taken into account during the analyses of the experimental data. The discovery of human pluripotent stem cells and their use in disease-modeling-studies has led to the precious possibility of investigating pathogenic mechanisms peculiar to human pathology.

Several studies on murine models and human pluripotent stem cells have highlighted widespread axonal and synaptic alterations in SMA motor neurons, including scarcity of dendrites and spines, impairment of calcium metabolism and dysfunctional remodeling ability. SMA human pluripotent stem cells are able to differentiate into motor neurons but derived cells are characterized by a reduced life span; a crucial role of apoptosis is noted in the disease onset, with increased caspase-8 and-3 activation.

Moreover, other cell types beside motor neurons have been recently implied in SMA onset. Interneurons and sensory neurons have been shown to be affected in the disease course thus suggesting the SMN loss could lead to a widespread perturbation of the neural network.

Non-cell-autonomous mechanisms due to a toxic activation of glia cells are emerging as a central contributor in motor neuron death. Astrogliosis has been known to play a pathogenetic role in amyotrophic lateral sclerosis (ALS), and recent discoveries have pointed out that SMA astrocytes could trigger motor neuron degeneration, both by losing their trophic function and acquiring inflammatory features. Overall, SMA pathogenesis appears to involve complex molecular mechanisms in multiple cell types highly interactive in creating a pathological microenvironment.

In this context, stem-cell-based therapeutic approaches could counteract several disease mechanisms at the same time. Pluripotent stem cells are able to give rise to mature motor neurons reacting to exogenous signals and thus replacing lost cells after engraftment, similar to what happens during physiological neurogenesis after injury. Stem cells are also able to differentiate in glia lineage and substitute toxic astrocytes with healthy ones, which are able to provide neuroprotection to endogenous motor neurons. Indeed, transplanting neural stem cells (NSCs) could be more effective than transplanting more mature cell subtypes. NSCs are more robust and intrinsically plastic; they can give rise to different subpopulations of healthy cells to replace and sustain endogenous ones. They could contribute to alleviate the neural network impairment by building alternative circuitry and stimulating the formation of new synapses. Furthermore, the role of oligodendrocytes and myelination dysfunction in SMA pathogenesis as potential therapeutic targets has to be further investigated. As a consequence, stem cell-based therapeutic approaches could be clinically effective through a multifactorial action consisting of modulation of glial impairment, trophic sustainment of the endogenous cells, and enrichment of the unhealthy microenvironment.

Indeed, several methods exploit stem cells in order to give rise to de novo production and exogenous import of neurotrophic molecules to the diseased spinal cord. Human NSCs are able to express several trophic factors and can be manipulated to produce specific substances. Insulin-like growth factor-I (IGF-I), brain-derived neurotrophic factor (BDNF), glial-derived neurotrophic factor (GDNF), and vascular endothelial growth factor (VEGF) have been demonstrated to alleviate neurodegeneration and provide environmental support in a series of models of neurodegenerative diseases. Moreover, the SMA nervous system showed an impairment in IGF1 signaling, which could represent a valid target for this approach in stem cell-mediated therapy.

Overall, these considerations support the hypothesis that cell-based therapies may function in supporting SMA motor neurons, especially by providing a trophic and protective environment within the spinal cord, and thus counteracting the multifactorial pathogenetic mechanisms of neuronal death. Stem cell transplants could represent a complementary approach to SMA molecular and gene therapies. The latter are able to precisely target the loss of SMN restoring its function, but they could have a minor impact on later-diagnosed patients, while cell therapy could be useful in managing symptomatic phases of the disease. However, it is still worth considering that mechanisms downstream of stem cell activity are poorly understood, as are the molecular pathways underlying neural repair. Elucidating these aspects will provide invaluable insight into SMA pathogenesis and possibilities for therapy, which could be suitable also for other motor neuron diseases.

The Selection of Cell Types to be Transplanted

During the last decades, a few relevant preclinical studies have been conducted on the use of different cell subtypes for SMA therapy. Cells have been screened for their ability to target the injured area, survive, and correctly engraft exerting a therapeutic action.

Many of these studies employed NSCs or derived motor neurons. A series of relevant preclinical studies have been reported (see Table 11.1 ). Hopefully, these results could open the way to clinical translation, even though more experiments will be necessary to accurately assess safety and effectiveness of stem cell based approaches.

Embryonic Stem Cell Derived Motor Neuron Precursors for Spinal Muscular Atrophy

A novel technique to obtain highly enriched human motor neuron precursor (hMNP) cultures was established by California Stem Cell Inc. HMNPs were derived from human embryonic stem cell (ESC) lines that were expanded on Matrigel for the first 3 weeks in a media supplemented with basic fibroblast growth factor (bFGF) . Cells were then transferred to ultralow binding dishes and suspended in MN differentiation media, including Glutamax, B27, insulin, sodium selenite, transferrin, MgSO4, and bFGF. Cells were exposed to the differentiation media for 5 days, supplemented with retinoic acid (RA), and then fully characterized. The transplantation of derived cells in the spinal cord of SMA models resulted in a proper engraftment and survival. Moreover, transplanted hMNP were able to produce trophic factors in situ (i.e., neurotrophin-3 (NT-3) and nerve growth factor). The number of endogenous motor neurons was significantly increased. Given these results, the research group designed a clinical trial, but until now, the FDA has not approved the study.

Induced Pluripotent Stem Cell Derived Motor Neurons as a Cell Source for Transplantation

Spinal motor neurons appear to degenerate rather selectively during the course of SMA and this background led to the experimental trial to transplant healthy motor neurons in SMA murine models.

Human motor neurons can be obtained by differentiating human embryonic or induced pluripotent stem cells (iPSCs). ESCs can be isolated from the blastocyst and are characterized by the ability to give rise to all the three germ layers. iPSCs derive from patients’ somatic cells ( Fig. 11.1 ), which have been reprogrammed to an embryonic stage. Both ESCs and iPSCs can be differentiated toward a motor neuron lineage with established protocols including the use of RA and Sonic hedgehog, which play a fundamental role in physiological motor neuron development. Derived motor neurons express specific markers (i.e., choline acetyltransferase (ChAT) and HB9) and spread axons to neuromuscular junctions after in vivo transplantation.

Overview on the potential use of human pluripotent stem cells for spinal muscular atrophy (SMA) modeling and therapy. SMA patients’ fibroblasts can be reprogrammed into induced pluripotent stem cells and genetically corrected to be used for autologous cell therapy. Induced pluripotent stem cells (iPSC)-derived motor neurons (MNs) represent an invaluable source for disease-modeling-studies. CSF , cerebrospinal fluid; IV , intravenous; NSC , neural stem cells; SMN , survival motor neuron.

We firstly transplanted a purified motor neuron population differentiated from LeX+stem cells derived from embryonic spinal cords into an SMA respiratory distress type 1 (SMARD1) transgenic model. LeX is a developmentally regulated tetrasaccharide carbohydrate, which is expressed in embryonic and adult NSCs. It can be exploited to select primitive stem cell fractions. A pharmacological treatment aiming to promote axonal growth directed at the neuromuscular junction was administered in combination. More specifically, isolated motoneurons were treated with a solution containing dbcAMP to enhance cell survival and axonal growth. Animals were also systemically injected with rolipram, a pharmacological compound modulating phosphodiesterase type 4 activity with the aim of counteracting the inhibitory effects of myelin proteins on axonal outgrowth. Moreover, GDNF was administered to the animals with intramuscular injections in order to direct donor axons toward their targets. Analyses at the end stage of the disease revealed that transplanted cells were integrated into the host anterior horn and their axons were directed toward the muscular target. The disease phenotype of SMARD1 mice was improved with better motor performance and increased survival. Moreover, neuroinflammation within the host spinal cord was significantly reduced after treatment.

Our research group exploited iPSC-derived motor neurons as a tool for study SMA pathology and as an experimental therapeutic approach after genetic manipulation ( Fig. 11.1 ). We derived human iPSCs from SMA patients’ fibroblasts, reprogramming them with the use of nonintegrating and nonviral episomal vectors. This reprogramming strategy is slightly less efficient than the viral reprogramming method but could ensure a more rapid translation to clinical trials. Generated cells were treated with single-stranded DNA oligonucleotides designed to genetically edit SMN2 inducing a substitution of a single nucleotide within exon 7, thus ensuring the inclusion of the exon within the transcript and as a consequence the production of a SMN1 -like functional protein.

Engineered iPSCs were then differentiated toward motor neurons employing a multistage differentiation protocol already validated for human ESCs. After 1 month, cells in culture expressed motor neuron markers (i.e., Hb9, ISLET1), pan-neuronal markers (TuJ1, Neurofilament, and MAP2), or spinal cord progenitor markers (OLIG2). Hb9- and ISLET1-positive cells could also be stained for ChAT and SMI32, thus showing a motor neuronal lineage. Uncorrected and modified SMA iPSC-derived motor neurons displayed significant differences in terms of morphology and survival. SMA cell phenotype was characterized by reduced axonal length and neuromuscular junction formation with a decreased survival in culture. These features were substantially rescued in motor neurons derived from corrected SMA iPSCs. An in-depth analysis of the transcriptome of both cell populations (i.e., treated and untreated ones with oligonucleotides) revealed a difference in a specific group of genes responsible for RNA metabolism, motor neuronal and axonal development. SMA iPSC-derived motor neurons were then transplanted into the spinal cord of SMA transgenic model and they were able to survive and properly engraft into the host microenvironment without modifications of the motor neuronal phenotype. An amelioration of SMA mice phenotype could be observed after transplant; treated SMA mice presented an increased survival (about 50%) in comparison to vehicle-treated mice. The beneficial effect was more evident with the transplant of corrected iPSC-derived motor neurons. Moreover, transplanted SMA iPSC motor neurons showed a decreased life span with a reduced ability to spread their axons when compared to motor neurons generated from oligonucleotide-corrected stem cells.

Overall, these data showed a way to exploit transplanted SMA-iPSC motor neurons to create an in vivo model of the disease. Moreover, the possibility of effectively correcting the SMA motor neuron phenotype by using molecular strategies could pave the way for experimental therapeutic approaches with autologous cells. It is worth considering that the proper engraftment and integration within the host microenvironment of motor neurons differentiated from oligonucleotide-treated SMA iPSCs appeared to be modulated by the time of correction. A very early treatment with oligonucleotides was more effective in ensuring beneficial effects, thus reinforcing the hypothesis of a “window of opportunity” for molecular therapeutic approaches.

In our study, transplanted human iPSC-derived motor neurons were able to provide a protective environment for endogenous cells by secreting neurotrophins and growth factors. This beneficial action appeared to be related to a successful engraftment in a proportional way. Indeed, motor neurons obtained from genetically manipulated SMA iPSCs displayed a major therapeutic effect on endogenous motor neurons, which presented an increase in cell size with enhanced axonal length and a prolonged life span.

Taken together, the results from our study provided data on the possibility of deriving SMA patients’ iPSCs, which could be genetically manipulated and used to derive precious insight into SMA pathogenesis, and experimental stem cell based therapeutic approaches.

Neural Stem Cells

NSCs represent the progenitors of all the three neuroectodermal progenies within the central nervous system (CNS). NSCs can be directly derived from embryonic or mature neural tissue, or can be obtained through differentiation of ESCs or iPSCs using well-established protocols. When transplanted, NSCs are able to improve the phenotype in different transgenic models of motor neuron disease. Our research group investigated, for the first time, the possibility of transplanting NSCs to ameliorate the disease phenotype in SMA and SMARD1 animal models.

Transplantation of Specific Neural Stem Cell Subpopulations

NSCs represent a group of heterogeneous cells, which can be selected for their biological potential to survive within the host environment after transplant, migrate to the injury site and properly engraft, and finally differentiate into the relevant phenotypes. Our research group has shown that choosing specific subpopulations of NSCs could be advantageous for therapeutic purposes and for clinical application in motor neuron disorders such as SMA. In our first work, we studied the therapeutic action of a primary NSC population selected for the levels of aldehyde dehydrogenase activity and low side scatter (ALDHhiSSClo cells), and their ability to modulate the disease course of the nmd mouse, a SMARD1 transgenic model. ALDH activity has been used as a parameter to isolate hematopoietic stem cells, and multiple research studies have described that high ALDH activity is related to the stem cell state.

Intrathecal injection of derived stem cells in nmd mice resulted in delayed disease progression, protection of motor neurons, and increased survival. Transplanted motor neuron precursors correctly migrated to the ventral horns of the host spinal cord. In mutated and treated nmd spinal cord, we showed with microarray and RT-PCR analyses a reduction of genes involved in glutamate damage and oxidative toxicity, and an overexpression of genes linked to the chromatin structure. These genes might be involved in SMARD1 development. Spinal cord of nmd- treated animals expressed high transcript levels of genes involved in neurogenesis (i.e., LIS1, doublecortin (DCX) and drebrin). The detection of DCX-positive cells within adult nmd spinal cord led to the speculation that both exogenous and endogenous neurogenesis could contribute to the nmd mice phenotypic improvement. We then tested the same selected subpopulation derived from spinal cord murine NSCs on SMA mice. ALDHhiSSClo cells were derived from fetal murine spinal cord neurospheres and cultured with specific priming and motor neuron differentiation protocols. GFP+ALDHhiSSClo-derived cells exhibited morphological and histochemical characteristics of fully differentiated motor neuron precursors in vitro. After ALDHhiSSClo cell intrathecal injection, 1-day-old SMA mice showed ameliorated functional (neuromuscular function and life span) and histological features (motor neuron cell number and dimension and neuromuscular junction size) compared to untreated SMA mice. . SMA mice treated with primed ALDHhiSSClo NSCs survived significantly longer also compared to SMA mice injected with undifferentiated ALDHhiSSClo cells, ALDHhiSSClo-derived astrocytes, and murine primary fibroblasts. Transplanted cells (HB9-GFP+) resembled true motor neurons morphologically, migrated, and engrafted along the recipient spinal cords. Full gene expression analyses of laser-capture-microdissected motor neurons derived from transplanted animals through microarray analysis showed an amelioration of SMA-diseased features in the direction of the wild-type pattern (for primary data refer to the GEO website: http://www-ncbi-nlm-nih-gov.easyaccess1.lib.cuhk.edu.hk/geo/ ; GEO accession number: GSE10224). Moreover, with Luminex multianalyte profiling technology, we detected different factors secreted by NSCs including VEGF, granulocyte colony-stimulating factor (G-CSF), and several neurotrophins (BDNF, GDNF, NT-3, and transforming growth factor alpha (TGF-α)), which may exert neuroprotective effects on motor neurons after ALDHhiSSClo cell transplantation. These data demonstrated the feasibility and therapeutic potential of NSC transplantation in the SMA mouse model.

In a second study, we investigated the hypothesis that NSCs obtained from murine pluripotent stem cells could improve the SMA mouse model phenotype. We used NSCs derived from wild-type murine ESCs and drug-selectable (with ganciclovir and G418) ESC lines, to promote neuronal differentiation and increase cell safety. Pluripotent-derived NSCs migrated into the spinal cord of SMA mice after intrathecal transplantation and improved phenotype and survival of SMA mice through replacement of motor neuron cells and conferring neuroprotection.

To differentiate drug-selectable mouse ESCs, the differentiation medium was supplemented with G418 and ganciclovir to select for neuroepithelial cells and against undifferentiated ESCs. We obtained motor neurons with classical phenotype and expressing correct neuroectodermal factors and cholinergic molecules using various signaling molecules in culture. The drug-selected NSCs promoted longer survival than the wild-type ESCs and the primary NSCs of our previous report. Moreover, we demonstrated that the differentiation toward motor neurons could be promoted once NSC phenotype had been established and reinforced. Combining cell transplants with pharmacological or gene therapy could enhance the therapeutic effectiveness of this strategy to a clinically relevant threshold.

In a more recent study, we transplanted murine LewisX+; CXCR4+NSCs directed toward a motor neuron fate into the spinal cord of SMARD1 transgenic mice. Treated animals presented a prolonged survival with increased motor neuron numbers within the spinal cord. About 18% of transplanted cells expressed motor neuronal markers at the end stage of the disease and some of them have spread their axons toward the anterior roots.

Glial Cells

Astrocytes play a key role in maintaining a healthy and trophic environment for motor neurons. SMA astrocytes seem to lose their protective action and undergo phenotypical and functional modifications. McGivern’s group showed that SMA astrocytes displayed morphological changes suggestive of overactivation even before motor neuron degeneration. Activated cells can secrete inflammatory molecules provoking neuronal death through the apoptotic signaling. Moreover, astrocytes derived from SMA patients’ iPSCs produced less GDNF (a key growth factor) in culture. Impaired astrocyte function can trigger motor neuron degeneration and, complementary, transplantation of Glial-Restricted Progenitors has been demonstrated to give rise to enriched healthy astrocytes within an ALS animal model (SOD1G93A rats). This resulted in delay of the disease progression. The beneficial effect was due in part to the restoration of astrocyte GLT1, which is essential for glutamate balance in the extracellular fluid. The site of transplantation was chosen around cervical respiratory motor neurons, which are responsible for respiratory failure in patients affected by motor neuron diseases. These data could open the path to the employment of astrocytes for SMA therapy, addressing neuroinflammation within the spinal cord, which has not been performed up to now. Complementarily, the role of oligodendrocytes in contributing for SMA pathogenesis has been considered; alterations in myelination have been reported in SMA murine models. More studies will be needed to clearly define the role of these modifications and the potential therapeutic effects of stem cell derived oligodendrocyte transplant.