October's ALS Research Roundup

CRISPR/Cas9-mediated excision of ALS/FTD-causing hexanucleotide repeat expansion in C9ORF72 rescues major disease mechanisms in vivo and in vitro

(Meijboom et al., 2022)

Gene therapy is an emerging technology with the potential to reshape the genome, addressing diseases on the genetic level. The most well-known and well-studied gene therapy in recent years is CRISPR/Cas9 gene-editing, in which DNA is ‘cut’ at a specific site and a new segment ‘pasted’ in the break. It is currently being tested for the treatment of a number of human diseases including sickle cell disease and transthyretin amyloidosis. This study addresses the use of CRISPR/Cas9 gene editing in C9orf72-mediated ALS (C9-ALS), the most common genetic ALS variant which has been linked to 40% of familial and 5-6% of sporadic cases. C9-ALS pathology involves a repeat expansion, in which a 6-nucleotide segment in the gene C9orf72 goes from being repeated several dozen times in healthy people to hundreds in patients with ALS. This leads into ALS through multiple pathways, including the loss of C9orf72’s healthy function (it is involved in immune activity and the removal of pathological proteins), formation of foci which sequester other functional RNA, and the production of toxic proteins from the repeat expansion.

While not the first to attempt removing the C9orf72 repeat expansion in ALS, this study more comprehensively addressed the disease mechanisms. They used an adeno-associated viral vector (AAV; small controlled viruses) to deliver the CRISPR/Cas9 rather than transfecting them as DNA, RNA or RNA-protein complexes (RNPs). AAV is likely to be superior as a therapeutic, being the safest and most effective gene therapy vector for the CNS, and can produce long-lasting effects from a single treatment. Additionally, simply removing the repeat expansion doesn’t address the loss of function pathological pathway, so this study combined the excision with guide RNAs (gRNAs) which restored C9orf72 to normal levels. This is a point of superiority over many therapeutic approaches for C9-ALS such as antibody therapies, which are only able to stop toxic activity and not recover the healthy state.

These methods proved to be successful in treating multiple models of C9-ALS pathology including induced motor neurons and brain organoids grown from patient-derived stem cells (iPSCs), neurons from transgenic mouse models with the repeat expansion and living mice. A single treatment in patient-derived cells was able to increase production of healthy C9orf72 proteins by 50%. In the mouse models (cells and whole animals) it showed the ability to reduce levels of the RNA-binding foci. Some of the toxic proteins produced by C9orf72 repeat expansions were shown to be suppressed in the different models, however only the mouse models suppressed poly-GR, one of the known disease-driving proteins.

While it is easy to focus on simply the removal of the offending component when addressing this disease, ALS is far too complex for that to form a comprehensive treatment. This study presents a multi-pronged approach to gene therapy against C9orf72-mediated ALS, both counteracting toxic function and recovering a healthy state. These processes are believed to both contribute to ALS independently and to compound each other, so a holistic approach is valuable. However, the authors did note that the AAV-mediated CRISPR used here is more of a proof-of-mechanism rather than a viable human treatment. Long-term gene editing can have consequences such as destabilising the genome with repeated ‘cuts’, off-target effects and immune responses to the Cas9 protein. Other methods of CRISPR delivery without these effects are in the works, but not yet available. However, once they are developed this combined approach may serve as the core for a valuable gene therapy which can address the source of the most common form of ALS, helping to stop the progression of the disease in patients and possibly even preventing familial cases before pathology can properly emerge.

TAGS: therapeutic, gene_therapy, C9orf72, iPSC, mouse

Two FTD-ALS genes converge on the endosomal pathway to induce TDP-43 pathology and degeneration

(Shao et al., 2022)

ALS is unusual in the sheer quantity of different mutations and genetic variants which have been associated with it. Mutations of dozens of genes have been linked to ALS as possible causative elements and/or disease modifiers, some in only one or a few cases, and others in large percentages of the ALS population. Some cases are not limited to one, either, with multiple mutations and pathological proteins co-occurring. This paper investigated the interplay between three different elements of ALS pathogenesis. C9orf72 is the most common genetic factor linked to ALS, appearing in 40% of familial and 5-6% of sporadic cases. Loss-of-function mutations in TBK1 are considered risk factors, and can increase the severity of ALS cases. TDP-43 is a protein which becomes misfolded and aggregates in the vast majority of cases, even without direct mutations in the gene which codes for it. Here, several genetically modified mouse models of ALS were used to confirm how these elements interact, and the implications for ALS pathology.

TBK1 levels were found not to differ between mice with and without the C9orf72 repeat expansion mutation. However, there were differences in the amount of TBK1 which was phosphorylated (pTBK1), a post-translational modification which can alter the protein’s activity. pTBK1 was found in proximity with poly-GA, a toxic protein produced from the C9orf72 repeat expansion. These combinations of poly-GA and pTBK1 were found in the frontal cortex and hippocampus of C9-ALS patients, but not healthy controls. Even more specifically, pTBK1 was linked to poly-GA which took on a compact aggregated form known as an oligomer, and tests suggested that clustering of TBK1 into these aggregates is required for its unusual phosphorylation. Oligomeric forms of other disease-associated proteins including TDP-43 have also been associated with ALS, and are closer linked to toxicity than larger aggregates. The authors proposed that these poly-GA oligomers imprison pTBK1 in affected neurons, preventing its regular function. As TBK1 regulates the clearance of protein aggregates through autophagy, this contributes to a self-reinforcing cycle in which aggregates produce a better environment for further aggregation, of both TBK1 and TDP-43. This may explain why TBK1 mutations accelerate ALS pathology, weakening the initial protective autophagy systems and so reducing the barriers to initial aggregation. Double-mutant mice were created to model this interaction, with TBK1 dysfunction enhancing various motor dysfunctions in mice expressing poly-GA. ?

Understanding the interplay between these elements of pathology can help us to therapeutically address them. The study showed that overexpression of healthy TBK1 in mice, even in the presence of poly-GA, increased the clearance of aggregates and so weakened the pathology. Gene therapies which bolster TBK1 levels, or prevent its phosphorylation and sequestration, could help to slow ALS pathology, or even prevent it from developing into a self-propagating cascade if caught early. Alternatively, bolstering the protein clearance pathways downstream of TBK1 may have a similar effect, and one such pathway involving the autophagy adaptor p62 is already of interest in ALS research. While this pathway is not present in all ALS cases, understanding how it functions as a risk factor may help us to identify others with similar activity, and exactly what tips the scales from possible ALS to full-blown pathology.

TAGS: mechanistic, autophagy, TBK1, C9orf72, TDP-43, mouse

MND Phenotypes Differentiation: The Role of Multimodal Characterization at the Time of Diagnosis

(Meo et al., 2022)

Motor neuron disease (MND) refers to a small family of ALS-like diseases, all characterised by loss of muscle function due to degeneration of motor neurons. Distinguishing between these is often difficult, as they can vary in both large and subtle ways. One method is based on the site of onset, dividing cases into classical ALS or MND with purely/predominantly upper (pUMN) or lower motor neuron (pLMN) symptoms. While this distinction seems simple, these divisions differ not just in their clinical symptoms but in prognosis, with pUMN and pLMN cases progressing much more rapidly than classical ALS. This can have significant implications for how treatments are applied, as well as lifestyle changes and preparations by patients themselves. However, separating these subsets, especially early stages, is difficult, and made more so by some diagnoses requiring measurement over years. The present study aimed to use a variety of evidence available at the time of onset to differentiate them. Data used included clinical, cognitive/behavioural, genetic and neurophysiological measures.

Disease subtypes were not significantly different in terms of demographics or disease onset. Cognitive and behavioural symptoms were also the same between pLMN and pUMN. Functional changes and neurophysiological features, such as patterns of motor neuron degeneration and electrical activity in neurons, remained similar between pLMN and classical ALS. pLMN was distinguished by greater weakness in the right leg, in what is known as the ‘split leg phenomenon’. This is not totally understood, but has been hypothesised to be largely influenced by factors in the periphery, rather than the central nervous system. However, the largest predictor was the rate of degeneration in upper motor neurons, allowing identification of pLMN with over 80% accuracy. pUMN patients were more likely to have onset in the lower limbs, due to a higher rate of involvement of the spine. They also had the lowest functional loss, both at diagnosis and over time, which proved the best tool for distinguishing pUMN disease at 86% specificity. Upper motor neuron degeneration was more variable than in pLMN, ranging between 60% and over 90% specificity. ?

This study was able to distinguish both pLMN and pUMN from classical ALS and each other with at least 80% accuracy, with measures that can be taken at or soon after diagnosis. While the main value is in prognosis, it may also help to understand what differs between these disease variants on a physiological level. It is currently unknown why certain neurons, both motor and non-motor, are targeted or spared. If this was better understood it could help dramatically in treatment, as well as predicting the development of both ALS and other MNDs.

TAGS: predictive, stratification, human_clinical

Mutation-specific metabolic profiles in presymptomatic amyotrophic lateral sclerosis

(Xia et al., 2022)

While it is easy to consider ALS as a disorder that just affects the muscles and motor neurons, in reality it has wide-reaching systemic effects. These are far from consistent, varying in large and small ways between individual cases and subsets of disease. The genetic subset was the focus here, with different ALS-associated mutations assessed for associated changes in patients’ metabolisms. Patients’ metabolism has been tentatively linked to both the onset and symptoms of ALS, with a common feature being disturbance of energy generation systems within the body. Higher resting energy expenditure and weight loss are also commonly seen, sometimes even before the onset of clinical symptoms. Metabolic factors studied here included body mass index (BMI), blood pressure and serum levels of blood glucose, total cholesterol, triglycerides, high-density lipoprotein (HDL) and low-density lipoprotein. These were compared in patients carrying mutations in a range of ALS-associated genes including C9orf72, SOD1, FUS, KIF5A, NEK1, SETX, TBK1 and TDP-43.

In a combined dataset containing all of these mutations, no significant differences were detected between presymptomatic ALS gene carriers and healthy controls. However, this was not the case for carriers of C9orf72 and SOD1 mutations. C9orf72 patients had lower BMI, systolic blood pressure, fasting serum glucose and higher HDL, contributing to a low cardiovascular risk profile. High cardiovascular risk profiles have previously been found to predict less aggressive disease, so these measures appear to either contribute to or be indicative of worse prognosis. Interestingly, SOD1 patients had higher cardiovascular risk profiles including higher BMI, higher fasting serum glucose and lower HDL. However, these were most prominent in patients carrying SOD1 mutations associated with slow disease progression. This protective metabolic profile could be the reason for this less aggressive disease profile.

There is a large and ever-growing body of research connecting metabolic changes to ALS disease progression, although the direction of association is unclear. It may be that metabolic changes induce the production of toxic disease-drivers such as reactive oxygen species, or that the metabolism is downstream of other pathological effects. Maybe both are true, and form a self-reinforcing cycle. However, we do not need to know exactly how they interact to use them as a diagnostic or prognostic tool. Presymptomatic biomarkers are some of the most valuable, as identifying cases early dramatically improves treatment outcomes. These markers are also likely to be at or near the start of the pathological cascade, and so give us the best understanding of the true causes of ALS. Given that C9orf72 is the most common gene associated with ALS pathology, being able to predict cases early may contribute greatly to improving quality of life and treatment for patients. While the slow-progressing SOD1 variant is not as common, being able to understand what makes this variant less aggressive may help with broader prognostic assessment, or even treatments to help mimic that disease environment in others.

TAGS: mechanistic, metabolism, C9orf72, SOD1, human_clinical??

Multi-omic single-cell velocity models epigenome–transcriptome interactions and improves cell fate prediction

(Li et al., 2022)

In the modern medical world, diseases can be studied on many levels, with genomics, proteomics, epigenomics, and many others. The incorporation of several of these simultaneously is referred to as ‘multi-omics’. Multi-omics has become particularly relevant following the advent of machine learning technology, which is capable of integrating a far larger amount of multidimensional information than the human mind working alone. As experimental measurement of cells involves destroying the cells in question, studying cells can only give ‘snapshots’ of moments in time. Computational models can use data from multiple cells to infer changes over time, developing cell ‘trajectories’. Multi-omic models have extended this technology, providing a more complete digital model which more accurately mimics the biological environment.

In this study, multi-omic work was used to create a digital model known as MultiVelo, which maps and predicts the degeneration fates of cells using data from single cells from the brain, skin or blood. This model was developed using transcriptomics, the study of RNA molecules in a cell, and epigenomics which studies modifications of the genome, particularly in response to lifestyle or environmental factors. Epigenetic changes can regulate the transcription of genes into RNA, so the overlap of the two was proposed as a key factor for accurate cell modelling. MultiVelo was able to accurately model a variety of factors in single-cell datasets from humans and mice including their gene expression levels and chromatin accessibility, which alters transcription rates. According to the authors, the most exciting new direction opened by MultiVelo is in relating epigenomic and transcriptional changes during cell differentiation. For example, they were able to identify two classes of genes which are associated with chromatin closing and repression of RNA transcription, and validated these across various tissues. They also highlighted four major cell states, two in which the epigenome and transcriptome were coupled and two where they were not. This is of particular importance in diseases such as ALS, in which there is extensive evidence of post-translational epigenetic changes to proteins.

While not directly associated with ALS, this study demonstrated the potential of multi-omic data integration platforms and models. Together, they can provide more holistic models of biological interactions, and highlight states in which they are and are not coupled. Given the complex physiological interactions observed in ALS, these could greatly contribute to our understanding of what factors truly contribute to ALS pathogenesis, and under what conditions. It may be that inducing decoupled cell states could help to inhibit disease processes, or vice versa. The authors also mention coupling of their model with others in order to fully model global protein dynamics, although this seems ambitious. However, if successful this could greatly help with understanding and predicting disease paradigms and developing appropriate treatments and therapies.

TAGS: model, multi-omics, machine_learning, non-ALS

Discovery of Mitophagy Inhibitors with Therapeutic Potential in Different Familial Amyotrophic Lateral Sclerosis Mutations

(Maestro et al., 2022)

Mitophagy is a process by which defective or otherwise damaged mitochondria are broken down, through a broader process known as autophagy. Dysfunctional mitochondria are a well-established factor in ALS pathology, and may contribute to both the spread of pathology-driving proteins and the production of toxic reactive oxygen species. Data on mitophagy itself is inconsistent, with some studies demonstrating blockage of autophagy in ALS while others found it to be over-activated. This may be due to different responses across ALS variants, or due to complex interactions such as attempted compensation for lost function at different stages of the autophagy pathway. Properly studying this, and attempting to recover healthy function, requires access to drugs and/or other molecules which can modulate this process. In this study, a cell line which expresses a mitophagy reporter was used to screen for compounds which can modulate both autophagy and mitophagy.

48 compounds were pre-selected from a library of candidates, and tested against the cell line. Of these only one, IGS2.7, was found to have an appreciable inhibitory effect on mitophagy. Notably, this activity required relatively high mitophagy levels to activate, suggesting potential in suppressing overactive mitophagy. It was found to likely act through targeting and inhibiting ULK1, an autophagy-related kinase, and this inhibition leads to accumulation of mitochondrial proteins. While not the primary driver of its mitophagy inhibition, IGS2.7 was also linked to CK1, a protein kinase with a likely function in autophagy which was previously linked to ALS via the probable disease-driving protein TDP-43. Treatment with IGS2.7 was able to reduce pathological phospho-TDP-43 in lymphoblasts from patients but also penetrate the blood-brain barrier to modulate the disease in TDP-43 mice.

The authors highlight that the value of IGS2.7 as a therapeutic model is that despite being identified for its association with mitophagy, it has multiple mechanisms of therapeutic action. Targeting both mitophagy, probably a downstream toxic effector, and TDP-43, whose exact role is unclear but likely contributes significantly to motor neuron death, is likely to address the broader ALS pathology more strongly than a more traditional ‘single-pathway’ treatment. This molecule was shown to be effective in treating both TDP-43 and SOD1-based ALS models, supporting its more wide-ranging therapeutic potential. There is a significant body of evidence suggesting that the loss of cellular quality control mechanisms is a major driver of disease, so it is plausible that modulating autophagy could have a notable effect. The testing in cell and mouse models was fairly broad, but many treatments have failed to make the jump across to human treatment. Hopefully, this will not be one of them.

TAGS: therapeutics, autophagy, TDP-43, SOD1, mouse, human_cell ?

Effects of Global Warming on Patients with Dementia, Motor Neuron or Parkinson’s Diseases: A Comparison among Cortical and Subcortical Disorders

(Bongioanni et al., 2022)

It is easy to think about non-infectious diseases as being contained within the body, the result of dysfunctions which are inevitable consequences of particular genetics. However, genetics are surprisingly fluid, and can be influenced dramatically by environmental and lifestyle factors. As such, it can be interesting to study just how much the environment in which we life can influence diseases. This study looked particularly at the effects of global warming on the incidence of various neurodegenerative diseases, including ALS, Alzheimer’s disease (AD) and Parkinson’s disease (PD). They studied data from 184 different countries in 1990 and 2016.

No association was found between warming and incidence of AD or ALS. However, annual average temperature did have an effect in ALS, with warmer temperatures linked to lower numbers of cases. This mirrors a previous finding that sunlight exposure reduced ALS incidence in a small population. These observations undermine a hypothesis made regarding a prior study into PD which linked global warming to heat-related degeneration effects, which can lead to neurotoxic oxidative stress, neuroinflammation and neurodegeneration. This may be due to the substantia nigra cells which are lost in PD being more vulnerable to heat-related degradation than other neurons.

This study was limited by the relatively small period of time assessed on a geological time scale, with a global average temperature change of ~0.5oC over the 27 year interval. It is interesting that ALS incidence was lowered with temperature. Further exploration of the mechanisms underlying this would be interesting, and could contribute to lifestyle changes for ALS patients.

TAGS: demographics, human

References

Bongioanni, P., Del Carratore, R., Dolciotti, C., Diana, A., and Buizza, R. (2022). Effects of Global Warming on Patients with Dementia, Motor Neuron or Parkinson’s Diseases: A Comparison among Cortical and Subcortical Disorders. International Journal of Environmental Research and Public Health 19, 13429.

Li, C., Virgilio, M.C., Collins, K.L., and Welch, J.D. (2022). Multi-omic single-cell velocity models epigenome–transcriptome interactions and improves cell fate prediction. Nature Biotechnology.

Maestro, I., De La Ballina, L.R., Porras, G., Corrochano, S., De Lago, E., Simonsen, A., Boya, P., and Martinez, A. 2022. Discovery of Mitophagy Inhibitors with Therapeutic Potential in Different Familial Amyotrophic Lateral Sclerosis Mutations. International Journal of Molecular Sciences [Online], 23.

Meijboom, K.E., Abdallah, A., Fordham, N.P., Nagase, H., Rodriguez, T., Kraus, C., Gendron, T.F., Krishnan, G., Esanov, R., Andrade, N.S., Rybin, M.J., Ramic, M., Stephens, Z.D., Edraki, A., Blackwood, M.T., Kahriman, A., Henninger, N., Kocher, J.-P.A., Benatar, M., Brodsky, M.H., Petrucelli, L., Gao, F.-B., Sontheimer, E.J., Brown, R.H., Zeier, Z., and Mueller, C. (2022). CRISPR/Cas9-mediated excision of ALS/FTD-causing hexanucleotide repeat expansion in C9ORF72 rescues major disease mechanisms in vivo and in vitro. Nature Communications 13, 6286.

Meo, G., Ferraro, P.M., Cillerai, M., Gemelli, C., Cabona, C., Zaottini, F., Roccatagliata, L., Villani, F., Schenone, A., and Caponnetto, C. 2022. MND Phenotypes Differentiation: The Role of Multimodal Characterization at the Time of Diagnosis. Life [Online], 12.

Shao, W., Todd, T.W., Wu, Y., Jones, C.Y., Tong, J., Jansen-West, K., Daughrity, L.M., Park, J., Koike, Y., Kurti, A., Yue, M., Castanedes-Casey, M., Del Rosso, G., Dunmore, J.A., Zanetti Alepuz, D., Oskarsson, B., Dickson, D.W., Cook, C.N., Prudencio, M., Gendron, T.F., Fryer, J.D., Zhang, Y.-J., and Petrucelli, L. (2022). Two FTD-ALS genes converge on the endosomal pathway to induce TDP-43 pathology and degeneration. Science 378, 94-99.

Xia, K., Witzel, S., Witzel, C., Klose, V., Fan, D., Ludolph, A.C., and Dorst, J. (2022). Mutation-specific metabolic profiles in presymptomatic amyotrophic lateral sclerosis. Eur J Neurol.