Profiling the inherent vulnerability of motor neuron subtypes

Abstract

Motor neuron diseases (MNDs) are a heterogeneous group of disorders that result in selective degeneration of upper and/or lower motor neurons (MNs). The precise causes of most MNDs remain poorly resolved and treatment options are currently mostly limited to the easing of symptoms and intensive care efforts. Remarkably, distinct MN subtypes display dramatic differences regarding their vulnerability towards degeneration in MNDs. The mechanistic basis for these inherent differences of the MN subtypes remains unknown. In order to understand what renders different motor neuron subtypes vulnerable, or resistant the thesis project first aimed to identify genetic factors linked to the different MN subtype. To achieve this, a transcriptome-wide screen of vulnerable versus resistant MN subtypes was performed. Firstly, this comprised identifying the transcriptome of MNs innervating the M. Rectus femoris (RF), a predominantly fast muscle innervated by fast MNs that are highly susceptible to MNDs. Secondly, the RF transcriptome was compared to that of the MNs innervating the M. Soleus (S), a slow muscle innervated mainly by slow MNs that are relatively resistant towards MNDs. Thirdly, as an example for resistant MNs, the gene signatures of both the RF and S MNs were compared to the transcriptome of the MNs innervating the M. Bulbocavernosus (BCV) muscle. The results showed a highly differential expression profile between the resistant BCV and S MNs on the one hand and the vulnerable (RF) MNs on the other. Promising candidate genes were defined by in silico data analysis and verified via qPCR and in situ hybridization. For instance, the identified genes Cartpt and Uts2/Uts2d provided the first confirmed molecular markers for resistant and relatively resistant MNs. Moreover, genes including Cart, Lxn, Lifr, Ubxn4, Calb2 and Pvalb that could be shown to be selectively expressed by resistant MNs have previously been linked to mediating neuroprotective roles. In parallel to the in silico analysis, all 200 identified candidate genes that showed an association with a resistant MN profile (high expressed in BCV plus S, but low expression in RF) were systematically screened to identify potential modifiers of neurotoxicity in four separate neurodegeneration models in the fruitfly Drosophila melanogaster. This screen for instance identified 36 modifiers of TPD-43-mediated retinal neuron loss as a model for MND-linked neurodegeneration. Interestingly, a substantial portion of the pool of the MN resistance-associated genes that showed a modifier activity in this screen, were linked to the ubiquiting/proteasome pathway. One of these genes encoded the ubiquitin ligase Ubxn4, a key co-factor of the central component of the ER-associated protein degradation pathway via VCP. Mutations in VCP in turn have recently been linked to familial MNDs and were shown to trigger aggregation of MND-linked TDP43. To allow confirming potential modifiers of MND-linked neurodengeneration a novel vertebrate model system for studying the MN loss in vivo was developed. This allowed stable neuron subtype-specific transgene expression of human TDP-43 in chick, which leads to progressive MN loss, in part mediated by caspase3-dependent apoptosis. This further showed for the first time that neurotoxicity mediated by the MND-linked protein TDP-43 requires its endogenous RNA binding activity. This provides the basis for systematically testing the activities of selected candidate genes on MND-like neurodegeneration of MNs in vivo. Taken together, this work revealed the first gene signatures and molecular markers associated with MNs that show resistance towards neurodegeneration in MNDs. In addition, the thesis project further provided the basis to test a defined set of these genes for a role in mediating MN resistance in MND models in vivo.