Analyses of spinal muscular atrophy (SMA) modifiers and drug-dependent responses using motoneurons (MNs) derived from induced pluripotent stem cells (iPSCs)

Spinal muscular atrophy (SMA) is an autosomal-recessive neurological heredopathia caused by homozygous loss of the survival of motor neuron 1 (SMN1) gene. Although SMN protein is ubiquitously expressed as core part of the nuclear spliceosomal machinery, SMN depletion exerts its deleterious effects mainly in lower α-motoneurons of the spinal cord. The subsequent motoneuronal death and disruption of neuromuscular connectivity evokes de-nervation and atrophy of skeletal muscles in proximal limbs and trunk causing a high morbidity in affected infants. In humans there is naturally a second SMN gene copy (SMN2)present almost identical to SMN1. Notably, a C>T transition in exon 7 interrupts an important exonic splicing enhancer site in SMN2 nucleotide sequence. Consequently, diminished intron/exon border recognition frequently causes exon 7 skipping so that approximately 90% of total transcripts are shortened (Δ7-SMN isoform). Still, about 10% full-length SMN transcripts (FL-SMN) are produced per SMN2 gene copy ensuring a basal expression of fully functional FL-SMN protein even in SMN1-deleted SMA patients, after all. Variable numbers of SMN2 (2-6 copies) are present within the population and SMA patients due to high dynamics in this genomic region. Hence, the number of SMN2 copies inversely correlates with SMA severity. SMN2 was regarded as the sole genetic modifier in SMA until the discovery of actin bundling protein plastin 3 (PLS3) (Oprea et al. 2008). In rare “discordant” families, some individuals over-expressing PLS3 remain phenotypically asymptomatic despite sharing the same genotype as SMA affected siblings. Thus, PLS3 obviously acts as protective modifier in SMA. Findings in cell culture models as well as in zebrafish and mouse models proved PLS3-induced rescue of SMA-mediated deficits in respect of axonal outgrowth and NMJ maintenance. Yet, the ameliorating effects of PLS3 could not be directly investigated in motoneurons of discordant family members. The same problem occurred in examining the effects of SMA drugs such as histone deacetylase (HDAC) inhibitor valproic acid (VPA) (Brichta et al. 2006; Swoboda et al. 2011). However, not all SMA patients exhibited the desired SMN2 activation and elevation of blood FL-SMN levels pointing to a different responsiveness to VPA administration (Brichta et al. 2006; Piepers et al. 2011). The underlying molecular mechanisms remained elusive so far. The general infeasibility to obtain living neuronal tissue inheres in any neurological illness in humans. Therefore, introduction of induced pluripotent stem cell (iPSC) technology (Takahashi et al. 2006) paved the way for the generation of individualised patient-derived in vitro cell culture models. Via ectopic over-expression of pluripotency-related transcription factors (OCT4, SOX2, KLF4, c-MYC) human somatic target cells are reprogrammed into a pluripotent state displaying typical characteristics of naturally pluripotent embryonic stem cells (ESCs) (Takahashi et al. 2007). Directed re-differentiation of these iPSCs towards spinal motoneurons enables the generation of a patient-derived in vitro SMA model (Ebert et al. 2009). The goal of this doctoral thesis was to establish such a personalised iPSC-based in vitro cell culture model of PSL3-discordant family members as well as individuals with different responsiveness towards VPA.Fibroblasts of VPA responder/non-responder SMA patients as well as two PLS3 discordant SMA families (SMA III affected and their corresponding asymptomatic SMN1-deleted siblings)were successfully reprogrammed by classical retroviral 4F transduction or by application of state-of-the-art non-integrative Sendai virus. Subsequently, twelve iPSC lines were clonally expanded and subjected to standardised validation procedures affirming bona fide pluripotency in all iPSC lines, indeed. In addition, an optimised iPSC-derived embryoid body (EB)-based motoneuron differentiation protocol was set up with which patient-derived iPSCs were differentiated into human motoneuron (MNs) cultures lacking sufficient MN numbers for meaningful studies, however. Instead, generation of stable small molecule neural precursor cell (smNPC) lines provided a valuable tool in SMA studies representing a homogeneous cell population with high MN yield upon differentiation. Stringent examination of four phenotypic groups (i.e. healthy controls, SMA I and PLS3 discordant SMA III and asymptomatic siblings) in several different cell populations representing the iPSC-derived in vitro MN development (i.e. fibroblasts, iPSCs, smNPCs and MN cultures) enabled monitoring of cell-specific features during developmental course. Subsequent MN differentiation of correctly patterned smNPCs resulted in considerable yield of bona fide MN cultures stunningly displaying significant differences in MN survival rate with SMA I exhibiting a massive MN decline while SMA III and asymptomatic showed an intermediate MN number relative to control upon maturation. These findings perfectly matched patients’ respective phenotypes. Furthermore, SMN expression RNA and protein levels and concomitant gem numbers broadly mirrored familiar SMN1/SMN2 copy number distribution amongst phenotype classes with regard to principally known decreasing SMN expression in adult vs. foetal cells or in proliferating cells vs. post-mitotic cells: SMA I cultures indicated lowest SMN expression rates as well as fewest gem numbers in comparison to controls. In contrast, only a moderate reduction in gem numbers and SMN expression levels was found in SMA III and asymptomatic siblings relative to controls. In addition, PLS3 expression on RNA and protein levels depicted strong over-expression in neural lineages of asymptomatic siblings highlighting the importance of tissue-specific PLS3 over-expression for exertion of the protective effect in mild SMA III. Different neurite length was not observed among phenotype classes in early MN cultures. PLS3/actin co-localisation occured in neuronal growth cones of MN cultures. Co-culturing iPSC-derived motoneurons with human myotubes successfully established the initial prerequisites for studying deficits in neuromuscular synapse formation, a key feature of SMA pathology. Application of iPSCs successfully presented VPA response in GABAergic neuronal cultures of VPA (non)responders. Altered GABA release ultimately corroborated data disclosing membrane fatty acid transporter C36 as causative element in VPA responsiveness (Garbes et al. 2013). For the first time, this study profoundly and reliably recapitulated SMN deficiency-conditioned deficits in SMA diseased cell lines of different severity grades and healthy control at decisive steps of MN differentiation directly in the target tissue. Moreover, PLS3 expression course during MN development gave new insight into putative PLS3-mediated protection specifically in MN cultures of asymptomatic asymptomatic siblings. Additionally, similar VPA responsiveness as in fibroblasts and blood was verified in neuronal cultures