Investigating a neuronal subcellular transcriptome by the novel technique of RNA TU-tagging, in a normal and ALS-related mouse model.

Amyotrophic Lateral Sclerosis (ALS, 'motor neuron disease') is a devastating neurodegenerative disorder which causes progressive loss of muscle function and paralysis. ALS leads to death, usually caused by the inability to breathe, on average only 3 years after diagnosis, with a lifetime risk of ~1 in 250 by 85 years old.

The principal cells affected in this disease are nerve cells called motor neurons (MNs). MNs connect the brain to the muscles therefore making movement possible. MNs progressively die during the course of ALS. MNs are amongst the largest cells of the body. Their main body lies in the spinal cord and contains numerous thin branching processes called dendrites. They also have one thin process, named the axon, which extends from the spinal cord out to each of our muscles. A single axon can measure over a meter, running from the spinal cord to ends of our fingers or toes. The connection between the MN and muscles is called the neuromuscular junction (NMJ).

All cells in an individual's body, although very diverse from each other, contain the same DNA, the genetic material that gives instructions to each cell. So the identity of each cell type (whether the cell is a nerve cell or a heart cell, for example) is the result of which regions of DNA are active and produce another type of chemical called RNA. RNA carries all the necessary information for the cell to function. The sum of all the RNA in a cell, named the transcriptome, is the signature that characterizes each cell type.

Knowing one cell transcriptome provides insights into its biology and helps determine the causes of disease. This is particularly relevant with MNs in ALS since there is good evidence showing that the biological processes linked to RNA 'metabolism' are primarily affected in ALS.

Importantly, we now know that RNA is transported and functions in different regions within an individual cell; in MNs it is transported in axons and to NMJs for specific roles. NMJs and axons are thought to be the first parts of the MNs to be affected in ALS.

Therefore it is important to know which RNAs are present in cell bodies and dendrites, and in axons, and at neuromuscular junctions of MNs, to understand how they function normally and what goes wrong in ALS.

It is possible to isolate and identify RNA from neurons and their axons when these are artificially grown in a culture dish or when cell bodies are dissected out under a microscope from fixed tissues. These findings have shown that thousands of different RNA species are actively transported to the axons, but it is difficult from these experiments to extract information that is relevant to mature neurons in their natural context in vivo and therefore to disease.

We will work with a new technique, 'TU-tagging', which has already been successfully used in mouse and which allows us to 'tag' RNA in specific cells in the context of a living animal. The tagged RNA can then be isolated and identified through high-throughput sequencing. We will apply this technique to MNs so for the first time we can isolate and identify RNA from MNs cell bodies and dendrites, and from their axons and NMJs in the living adult mouse.

We will work with normal mice, and with a new mouse model that we have developed that has a defect in a gene, Tardbp (also known as Tdp-43) that causes ALS. From our current work we know defects in this gene give an aberrant RNA profile in the MN cell bodies of this mouse. Currently no Tdp-43 mutant mice exactly model human ALS, but they teach us a great amount about how Tdp-43 functions in the normal and abnormal state.

These results will be extremely helpful in furthering our understanding of MN biology and of what causes these cells to be so specifically vulnerable in ALS.

This project will help research in the many other diseases in which RNA metabolism in different cell regions is important, and in response to nerve injury where again it plays a key role.

Coding and non-coding RNAs are transported to different subcellular localisations for local translation and regulation. Localised 'RNA metabolism' is important in polarised cells such as neurons, but techniques to study RNA localisation are not able to isolate subcellular transcriptomes in vivo.

Motor neurons (MNs) have cell bodies in the spinal cord, numerous arborizing dendrites and they connect to muscle fibres at the neuromuscular junction (NMJ) through an axon that can exceed 1 meter in length.

We are interested in MNs because of our research into Amyotrophic Lateral Sclerosis (ALS), a fatal neurodegenerative disorder that is characterized by the progressive loss of motor neurons that 'die back' from NMJs. ALS-causative mutations in RNA-binding and -transport genes (TARDBP 'TDP-43', FUS), and in C9ORF72 that induce RNA foci - have highlighted aberrant subcellular RNA metabolism in ALS pathogenesis.

Thus there is a clear need to identify the neuronal subcellular transcriptome in vivo to understand normal function, and dysfunction in our paradigm disease, ALS.

We will use a new technique, 'TU-tagging' (already successful in mouse) to identify the subcellular transcriptome of three compartments of adult motor neurons in normal mouse and in a mouse with a mutation in Tdp-43. Our preliminary data from a Tdp-43 mutant we are characterising shows disruption of many downstream genes including b-synuclein and tau.

TU-tagging works through the cell-specific expression of an enzyme (UPRT) and treatment with a drug (4TU), allowing incorporation of thio-uridine into RNA in specific cell types. Thio-uridine RNA is biotinylated and then pulled-down. We will use TU-tagging to identify RNA from (1) MN cell bodies and dendrites, (2) MN axons, (3) NMJs.

This will allow us insight in the biology of MNs, axons and NMJs and their susceptibility in disease.

TU-tagging is applicable to all models of normal function and disease, and of neuronal injury.

The impact of this research lies in two areas (1) understanding the biology of the transcriptome in subcellular locations, (2) amyotrophic lateral sclerosis and the TDP-43 proteinopathies. These two impacts are quite different in that (1) provides a new technique (albeit previously used in mouse) that we believe will become a standard approach to isolating the subcellular transcriptome and so with very broad impact, whereas (2) tackles a biomedical issue, well-known neurodegenerative diseases for which no cure exists.


(1) TU-tagging is a technique that is relevant to every study with a need to catalogue and quantify the localised RNA transcriptome within a cell type. This has obvious application to early development of the Drosophila and mammalian embryos, but also to ALL other studies in which subcellular RNA pools are important.

Therefore we envisage this becoming a standardly used technique, only limited currently by the availability of suitable mouse models. We note that with the number of freely available mice with individual gene mutations and Cre transgenics with precisely defined patterns of expression set to rise dramatically over the next few years, TU-tagging will be very widely applicable to both academic and industrial applications. Knowing the subcellular transcriptome will have impact on academic studies of all areas of biology and also on industry studies for example in highlighting new mechanisms and targets, notably in non-coding RNAs. One immediate application for both academic and commercial interests is in the neuronal response to injury which involves local translation to form a signalling complex that is carried to the cell body.


(2) ALS is a devastating incurable mid-life disorder for which no cure or effective treatment exists. It is clear that localised RNA biology is involved within neurons and so our research will shed light on pathways and networks that are important for motor neuron health and disease. This will have impact on the ALS community.

As our paradigm, we have chosen to work with a Tdp-43 model related to ALS - this gene/protein is also dysfunctional in many common neurodegenerative disorders, including Alzheimer disease and Parkinson disease, and so this project may well also have a wider impact as more commonalities are found in the 'TDP-43 proteinopathy' neurodegenerative diseases, particularly those clearly involving disruption to RNA metabolism disruption.

Thus our findings may benefit biotech/pharma by providing new insights and new targets for the development of therapeutics for ALS, and possibly other RNA disorders. It is intriguing also that several of the new ALS genes also play roles in specific cancers when translocated and so it is possible this research may have an impact on cancer studies also.



We note also that the sub-cellularly localised RNA transcriptome also plays a role in the response to nerve injury and thus our approach should be of interest in studying the impact of neuronal damage.

Finally, increasingly sophisticated techniques are being used for making RNA a therapeutic molecule, and it is likely that in the future knowledge of the localised transcriptome may further refine these efforts.