RNA interference: a new tool to study gene functions in adult mammalian muscle "in vivo"

RNA interference (RNAi) is a powerful method for sequence-specific posttranscriptional
gene silencing (PTGS), which allows rapid survey of gene functions using
double-stranded RNA (dsRNA). At the time when we started this work, RNAi was a
recently developed tool that had been successfully applied to many organisms, in
particular C. elegans and Drosophila, but not to any mammalian system. It was generally
doubted that RNAi would also work in mammals in vivo, because the introduction of
dsRNA can induce general shutdown of translation and apoptosis in several mammalian
cell types. One excellent model system for investigating this open question is the nervemuscle
synapse known as the neuromuscular junction (NMJ).
Characteristic for the NMJ is the precise apposition of the neurotransmitter
release machinery on the nerve terminal side and the neurotransmitter receptors on the
muscle fiber membrane. At least two mechanisms underlie the formation and
maintenance of a postsynaptic apparatus on the muscle fiber membrane. Both
mechanisms are triggered by the heparan sulfate proteoglycan agrin, which is released
by the motor neuron. First, neural agrin activates all the cellular mechanisms necessary
to assemble a fully functional postsynaptic structure including aggregates of acetylcholine
receptors (AChRs). Besides this redistribution of preexisting molecules, agrin signaling
restricts the transcription of postsynaptic proteins to myonuclei located in the NMJ. Still
little is known about the agrin signaling cascade. Therefore, once RNAi could be
developed for mammals system, it will in turn provide a unique tool to address the role of
newly identified genes in the postsynaptic differentiation, since there are no tools
available for the fast and reliable perturbation of gene function in vivo.
In the first part of this work, we investigated the potential of RNAi in perturbing
the formation and stability of postsynaptic structures in adult muscle in vivo (chapter 2
and 3). First, we used the experimental paradigm where neural agrin expressed in nonjunctional
regions of rat soleus muscle induces formation of ectopic AChR aggregates.
Knockout experiments have shown that this agrin activity requires the receptor tyrosine kinase MuSK and the AChR-associated scaffolding molecule rapsyn, but not the
cytoskeletal proteins sarcoglycan α (SGCA) and utrophin. In our experiments, we show
that co-injection of dsRNAs derived from MuSK or rapsyn perturbed agrin-induced
formation of ectopic AChR aggregates, while dsRNAs derived from SGCA or utrophin had
no significant effect. In a further step, we used RNAi to study the role of MuSK at adult
NMJs. Here, the electroporation of plasmids encoding short hairpin-based 21-bp small
interfering RNAs (siRNAs) or long hairpin dsRNAs, which allow global and sustained
perturbation of MuSK expression, leads to the disassembly of NMJs in adult mice. These
results are consistent with the finding that auto-antibodies to MuSK, which also lower the
amount of MuSK protein, cause severe forms of myasthenia gravis. In summary, these
results demonstrate for the first time the effectiveness of long dsRNA as well as siRNA in
silencing endogenous genes in adult mammalian muscle in vivo and they provide strong
evidence that continuous expression MuSK is required to maintain the NMJ.
The second part of this work aimed to establishing RNAi in adult muscle to study
the role of newly identified genes in the development of the NMJ and in the growth of
muscle fibers (chapter 4 and appendix). First, we used RNAi to perturb neural agrininduced
formation of ectopic AChR aggregates on mouse soleus muscle. We show that
electroporation of plasmids encoding short hairpin- derived siRNA for MuSK leads to the
perturbation of ectopic AChR aggregation, regardless whether agrin expression vector or
recombinant protein was applied to innervated or denervated muscles. These results
clearly show the reliability of RNAi in adult muscle in vivo and therefore set the stage for
experiments aimed to study the function of genes, whose expression is altered during the
formation postsynaptic structures. A protocol was established to identify functional siRNA
target sites in several genes. Plasmids were designed that encoded short hairpin RNAs
(shRNAs) derived from different putative effectors of the mammalian target of rapamycin
(mTOR) signaling pathway. For some candidate effectors, electroporation of the
corresponding plasmids into mouse soleus muscle leads to altered muscle fiber size.
These preliminary results are consistent with several reported findings, which indicate
that the mTOR signaling pathway is a central controller of muscle fiber atrophy and
hypertrophy. The efficiently induced RNAi in those experiments demonstrates that our protocol is useful for identifying siRNA targets. In summary, these results demonstrate
that we have successfully established RNAi as a fast and reliable gene knockdown
method in muscle fibers of mammals. This method will be important for future
investigation of gene functions in adult mammalian muscle in vivo.