While it is of utmost interest to find a cure for ALS soon, we should take the time to research and evaluate the cellular processes that lead to disease and take this knowledge to develop highly effective and target-specific therapeutic strategies. Session 6A aimed to do precisely that: to look more closely at the cell biology and pathology involved in ALS.
The session was opened by Francisco Baralle, who talked about the characterisation of TDP43 and its interactions. TDP43 first became famous for regulating exon 9 splicing in the cystic fibrosis transmembrane conductance regulator (CFTR) transcript, but has since been shown to be involved in numerous aspects of RNA metabolism. TDP43 binds to GU-rich RNA sequences and fellow heterogeneous nuclear ribonucleoproteins (hnRNPs). The majority of protein-protein-interactions are brought about by the protein’s Q/N-rich C-terminus, spanning amino acids 321-366.
TDP43 controls its cellular levels by a self-regulatory mechanism: binding to the 3’UTR of its own transcript leads to splicing disruption and subsequent degradation of the pre-mRNA in the nucleus. Cytoplasmic aggregation of TDP43, which occurs during disease, may break the negative feedback loop and even stimulate TDP43 production. More generally, mislocalisation of the protein in the cytoplasm and resulting toxic gain of function have been proposed to be pivotal disease mechanisms. In line with this, mutant TDP43 has been associated with downregulation of Syntaxin levels – caused by aberrant binding to syntaxin mRNA in the cytoplasm – and consequently impairment of synaptic transmission.
The TDP43 fly homologue TBPH has a short half-life in vivo and its synthesis is permanently required for larval motility and neuromuscular junction (NMJ) assembly. However, overexpression of TBPH leads to increased cell death, as seen in flies expressing TBPH under the control of the eye-specific GMR-GAL4 driver. Prof Baralle and co-workers could rescue this phenotype by co-expressing an aggregation-prone TDP43 construct, consisting of the complete human TDP43 sequence and a tag of twelve consecutive Q/N-rich regions (12xQ/N). Formation of non-toxic, insoluble aggregates resulted in the sequestration of soluble TBPH and thereby abolishment of cytotoxicity. It is important to note that sequences outside the Q/N-rich region can influence TDP43 aggregation and toxicity: the nuclear localisation sequence (NLS)-containing N-terminus for example is a requirement for the aggregation of the protein, while intermediate regions mitigate toxicity.
In the second talk Daniela Zarnescu described her group’s search for in vivo translational targets of TDP43. While TDP43 had already been shown to be involved in miRNA processing, RNA transport and RNA stress granule formation, its role in protein synthesis had remained unclear. Computational analysis identified futsch/MAP1B mRNA as a candidate translational target.
Futsch is a microtubule stabilising protein, which regulates axonal and dendritic development and confers synaptic stability. TDP43 interacts with futsch mRNA at the NMJ, thereby controlling its localisation, expression and translation. Dr Zarnescu showed that overexpression of both wild type and mutant TDP43 resulted in a negative regulation of futsch expression through a shift of the futsch mRNA from actively translating polysomes to non-translating RNPs. Interestingly, Futsch overexpression was neuroprotective in TDP43 fly models: it was shown to rescue motor deficits, mitigate architectural defects at the NMJ, significantly reduce TDP43 aggregation and extend life span.
Next, Janice Robertson reported on her team’s efforts to produce isoform-specific antibodies against C9ORF72 in order to characterise the biochemical profile and expression patterns of C9ORF72 long and short isoforms (aa 1-481 and aa 1-222, respectively) in brains of C9ORF72 and sporadic ALS patients. Polyclonal antibodies against the two isoforms were successfully generated and antibody specificity was confirmed using recombinant proteins. Staining of post mortem tissue and Western blot analysis of protein lysates showed that the two isoforms were indeed present in human patient brain samples. Zooming in on different areas of the brain, Dr Robertson’s group found that levels of the long C9ORF72 isoform were reduced in the frontal cortex of C9ORF72 ALS patients in comparison to sporadic cases. In addition, the two isoforms showed distinct subcellular localisations in Purkinje cells: C9ORF72 short for example associated with the nuclear membrane through interaction with nucleocytoplasmic transport proteins. It has to be noted that sample sizes were fairly small and that more work needs to be done to fully understand differential expression and splicing of C9ORF72 under normal and pathological conditions.
In the fourth talk Steven Boeynaems, as several others had before him, pointed out that toxicity of C9ORF72 expansions is multifactorial or multi-layered. Haploinsufficiency, RNA toxicity and repeat-associated non-AUG initiated (RAN) translation – yielding dipeptide repeat (DPR) protein aggregates – have been proposed as potential mechanisms of toxicity. Steven and his team investigated the involvement of all three mechanisms in a zebrafish model of C9ORF72 ALS/FTD. In order to test for a loss of C9ORF72 function, they performed a morpholino-mediated knockdown of the endogenous C9ORF72 homologue. Reduced levels of C9ORF72 have been described in patients and secondary structures within the GGGGCC repeat expansion may play an important role in this process. Unsurprisingly, knockdown of C9ORF72 led to induction of axonopathy in fish, which could be rescued by overexpression of the human protein.
The group went on to investigate RNA toxicity and found that C9ORF72 sense RNA was neurotoxic, while antisense RNA – at least in their hands (expression constructs contained 40-60 repeats) – did not induce axonopathy. Examining secondary RNA structures by atomic force microscopy, they learned that sense RNA forms G-quadruplexes, structures that are predicted to increase stability of the RNA and to sequester RNA-binding proteins like Pur-alpha. Overexpression of human Pur-alpha in the C9ORF72 context prevented RNA toxicity, most likely through saturation of binding sites and consequently inhibition of aberrant protein interactions. RAN translation did not appear to be of importance in zebrafish, however exogenous arginine-rich DPR proteins were found to induce neurotoxicity.
Afterwards, Vladimir Buchman convincingly showed that aggregation of FUS in the cytoplasm is a multistep process, which involves both RNA-dependent and RNA-independent mechanisms. RNA-dependent aggregation results in the formation of so-called FUS granules (FGs), physiological structures, which are structurally similar, but not identical to RNA transport granules. This process is reversible and requires the N-terminal prion-like domain as well as specific RNAs. Clustering of FGs and recruitment of further RNAs and proteins produces larger FUS aggregates (FAs), which are similar to stress granules. Upon inhibition of transcription, FAs dissociate into RNA-free complexes and undergo transformation to form RNase-resistant cytoplasmic aggregates associated with ALS. This process is irreversible and RNA-independent.
In the last talk of the session André Bento-Abreu introduced Elongator protein 3 (Elp3), the catalytic subunit of the Elongator complex, as a disease modifier of ALS. Elongator associates with RNA polymerase II to regulate transcription elongation, but has also been shown to increase translation efficiency through tRNA wobble modifications. Previous studies have linked a polymorphism in the ELP3 gene to ALS and a reduction in Elp3 protein levels to axonal / synaptic defects in flies and zebrafish. Dr Bento-Abreu’s team therefore set out to examine the involvement of Elp3 in ALS.
The researchers demonstrated that AAV9-mediated overexpression of ELP3 delayed disease onset and prolonged survival in SOD1 G93A mice. Strikingly, constitutive knockout of ELP3 was embryonically lethal and conditional knockdown in adult mice resulted in death within 40 days. Overexpression of ELP3 was furthermore found to rescue axonopathy in SOD1 A4V zebrafish: while motor neuron counts remained unchanged, the number of NMJs increased, indicating a re-innervation of the muscle. Interestingly, this neuroprotective effect was abrogated by mutations in the SAM domain (mediating methylation reactions), but not the HAT domain (mediating acetylation reactions).
Naturally, this session only covered a minute fraction of the cellular processes implicated in ALS pathology. However, knowledge of the presented pathways and mechanisms may aid the development of a therapeutic strategy against ALS.