Early detection of progression in multiple sclerosis using OPM-MEG

Multiple sclerosis (MS) affects 130,000 people in the UK, with 130 more people diagnosed each week. Over time, MS gradually worsens, and treatments for this phase of disease are extremely limited. There are no ways to detect early stages of progression, making it difficult to develop treatments.

This study will apply a new scanning technology, OPM-MEG, to detect progression in MS. People with progressive MS, stable MS and healthy controls will be recruited. They will receive three OPM-MEG scans: an initial baseline scan, after six months and after one year. They will receive clinical assessments on their first and final visits. I will analyse how brain networks change over time using OPM-MEG markers. I will investigate if these markers worsen in those with progressive MS, and remain the same in those with stable MS.

I aim to distinguish who is silently progressing from who is stable, by using two OPM-MEG scans within a year. The long-term vision is that those who have progressive disease could then join clinical trials, and stable patients could avoid intensive treatment. The new marker will be objective, providing greater certainty for patients and neurologists. If we link decline in brain networks to progression, this will change how we think about disability in MS. This may alter how we perform clinical trials in the future. Currently there are no tools to detect early progression in MS, so this research is desperately needed.

Exploring PDE10A levels and their relation to disease markers in far-from-onset Huntington’s disease

Huntington's Disease (HD) is a serious brain disorder caused by a mutation in the huntington gene. The gene contains too many building blocks called CAG repeats, making it abnormally long. This causes certain brain cells to break down, leading to problems with movement, thinking and mood. New research suggests that in some brain cells, the CAG repeat keeps getting longer throughout life, in a process called somatic expansion. When it reaches a certain length, it sets off a chain reaction that causes the brain cell to stop functioning normally and die. During this process, the production of a wide range of proteins changes, but one of the largest is reduction in the levels of an enzyme called PDE10A.

To prevent the disease we need to give treatments early, before the chain reaction has started. However, it's difficult to know if a treatment works when people don’t have symptoms. To do this, we need special tests to detect the first signs of disease within the brain. The HD Young Adult Study (HD-YAS) is a research study on young people with the HD gene expansion who are 20-25 years from showing symptoms.

Positron Emission Tomography (PET) scans are able to measure PDE10A levels in living people. This study will perform PDE10A-PET scans in participants with and without the gene mutation, from HD-YAS. Each person will have two scans, one year apart. This will show how levels change in HD and correspond with markers of brain damage and symptoms. This will help us understand how the disease worsens over time and give us important starting information for future clinical trials.

Identification of biomarkers to predict early diagnosis and progression in patients with multiple system atrophy

Multiple system atrophy (MSA) is a relentless neurological disease that causes significant disability and is ultimately life-limiting. People present with a combination of problems, including difficulties with walking, balance, tremor, blood pressure, bowel and bladder dysfunction. Doctors use the signs and symptoms to guide them in diagnosing MSA and can only make a definite diagnosis of MSA after the patient has died. There is no test that a clinician can use to help predict how fast the disease will advance and how long the person has left to live, some of the most important questions that patients and their families have.

The disease is characterised by abnormal deposition of the protein alpha-synuclein within specific cells within the brain. We will use clinical, imaging, blood and spinal fluid data already obtained from MSA patients withinthe PROSPECT-M-UK study to look for a test that we could use to predict the progression of patients. We will also continue to recruit further patients for this study.

I will also focus on the use of new techniques to assess blood and spinal fluid samples to detect abnormal alpha-synuclein as an early test to diagnose MSA. Being able to make an accurate and early diagnosis is of increasing importance given the advent of trials of new treatments for MSA that are aimed at preventing or slowing down disease progression.

The knowledge gained will also improve understanding of the underlying disease biology and enable better monitoring and discussion about disease progression.

Predicting dementia using quantitative electroencephalography: development of an EEG Dementia Risk Signature

I will develop a first-of-its-kind screening test using electroencephalography (EEG) – a brainwave measurement device - that can accurately identify healthy people in the general population who have a high chance of developing dementia several years or decades later.

Currently, there is no single test that can accurately identify who in the general population is going to develop dementia before they experience memory and thinking problems. Therefore, it is unclear who would benefit most from disease slowing drugs and other early prevention strategies. There is an urgent need for a new test that can be used alone or alongside others (such as blood tests for Alzheimer's disease) to enable early and accurate prediction of long-term dementia risk.

My work during the past year showed that a single EEG characteristic (which reflects slowing of brainwaves) was linked to an increased chance of developing dementia 12 years later among 3,000 healthy older adults. I will now investigate whether other EEG characteristics (reflecting diverse aspects of brainwave function) may also predict future dementia. Then, using artificial intelligence, I will examine whether combining multiple EEG characteristics into a single measure (ERS-dementia) would improve the ability to predict future dementia among >4,000 middle-aged and older adults.

If successful, my work will introduce a new screening test (ERS-dementia) that will contribute to dementia prevention around the world. Furthermore, I will work with engineers to develop a cheap portable device that will enable ERS-dementia to be used by anyone in their own homes within the next five years.

Discovering intergenerational genes in Huntington’s Disease

Huntington’s disease (HD) is a genetic movement disorder with dementia and psychiatric symptoms. This research is a sperm study aiming to better understand HD inheritance to improve genetic counselling of HD gene carriers. HD therapies under development may reach clinical trial stage within the next 5 years. This study will help determine whether sperm are affected by these drugs and therefore has important implications for drug development.

DNA is made up of four molecules; A(adenine), T(thymine), C(cytosine) and G(guanine) and contains ‘repeat regions’; repetitive combinations of these molecules. The huntingtin (HTT) gene is a repeat region with repeated CAG units. Normally there are up to 26 CAGs in HTT but those carrying the HD mutation have 36 or more.

Over 50 neurological human diseases exist, called repeat expansion disorders (REDs), are caused by abnormal repeat numbers in different DNA regions.

We know in HD, and many REDs, that when the mutation is inherited the gene length may change in sperm and sometimes children inherit a different sized gene compared to their parent. Expansions can lead to the child developing disease symptoms earlier than their parent, termed ‘genetic anticipation’. Although genetic anticipation is widely recognised, the reason it happens in some families and not others is unclear.

This project involves donation of blood and semen from men with HD for genetic analysis of repeat region changes in sperm and to identify potential causative genetic factors. This will improve understanding of HTT mutation inheritance risk and may reveal biological pathways for further investigation which may be applicable to other REDs.

Investigating the consequences of HnRNPM mislocalisation in amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) results in degeneration and death of motor nerve cells (motor neurons,‘MNs’) which control walking, speaking, swallowing and breathing. It is fatal and untreatable. To develop treatments, we must understand precisely what causes MN degeneration.

Important proteins which bind and regulate RNA messages are dysfunctional in MNs in ALS. I have discovered that another protein in this class, called HnRNPM, is abnormally ‘mislocalised’ from nucleus to cytoplasm in MNs in ALS. This is especially interesting since HnRNPM binds a specific abnormal RNA message in ALS and could be a new target for treatments.

To study ALS, we transform skin cells from ALS patients and healthy volunteers into stem cells then into MNs in the lab, and compare these. I hypothesise a loss of HnRNPM’s normal functions due to its mislocalisation in ALS. To model this, I will deplete HnRNPM in healthy MNs, then look for typical MN features of ALS, and for differences in RNA messages due to loss of HnRNPM's functions. I will also examine all interactions of HnRNPM with RNA messages in healthy and in ALS MNs. I will assess whether increasing HnRNPM levels can resolve features of ALS.

This work will help us understand the contribution of HnRNPM to MN degeneration, and whether manipulating HnRNPM could treat ALS. If HnRNPM manipulation shows promise as an ALS treatment, development could progress to animal studies and subsequently clinical trials. From initial studies to approval of a treatment for patients could take approximately 10-15 years.

Using machine learning to predict prognosis and treatment response following silent MRI activity in multiple sclerosis

Personalised treatment is needed to reduce long-term disability in multiple sclerosis (MS), while limiting the risks of treatment.

MS is one of the most common causes of disability in young adults. Many MS treatments can reduce the development of disability, but the strongest treatments can have more serious side effects. I aim to solve three common problems to improve MS management.

Firstly, even when someone with MS is on a treatment and has no new symptoms, MRI scans often show new “silent lesions”. This indicates that the disease is still active. We do not know the long-term prognosis following silent lesions, or whether changing to stronger treatments improves outcomes. I aim to solve this problem by analysing the largest international database of people with MS.

The second problem is that lesions are often missed when doctors report MRI scans. Recent studies have shown that artificial intelligence (AI) can detect most of these missed MS lesions. However, since these lesions are harder to detect, they may also be less important. I will analyse whether this is the case. This will be crucial if we are to use AI to help report MRI scans in MS clinics.

Finally, the severity of MS varies widely between affected people. We are currently unable to accurately predict which people could benefit most from early use of the strongest treatments. To address this, I will use existing clinical and MRI data from thousands of people with MS to train cutting-edge AI tools. This will be the first step towards developing a software tool to assist with personalised treatment decisions in MS clinics.

Interoception as a diagnostic, mechanistic and therapeutic prism for functional seizures

Functional seizures (FS) are as common as multiple sclerosis and associated with elevated morbidity/mortality rates. We have shown that interoception, the brain-body processing of visceral signals, is central to FS. We aim to demonstrate the therapeutic, mechanistic and diagnostic importance of interoception via three experiments.

Experiment 1:

Resonance breathing (respiration rate=0.1 Hz) increases heartbeat discrimination task (HDT) interoceptive performance and reduces dissociation, via improvements in heartrate variability (HRV). We have shown that HDT performance corresponds with FS frequency and dissociation. We will assess the feasibility of resonance breathing training to improve HRV, and HDT performance, and reduce FS frequency and dissociation, thereby providing data for a pilot treatment study.

Experiment 2:

We have shown that reduced HDT interoceptive performance correlates with reduced sense of agency (SoA) via the Libet clock paradigm in healthy individuals. We will extend this to FS patients. We hypothesise that in hyperkinetic FS, reduced interoceptive accuracy is associated with reduced SoA. This will aid mechanistic understanding of the involuntariness of these seizures.

Experiment 3:

We have shown that the EEG correlate of interoception, the heartbeat evoked potential/HEP, changes pre-ictally in FS but not epileptic seizures (ES). Using scalp EEG archives, we will focus on cases where qualitative ictal EEG analysis struggles to distinguish ES/FS, and in patients where ES+FS coexist. We hypothesise the absence/presence of quantitative pre-ictal HEP changes in semiologically similar ES/FS respectively. We will derive a predictive model and test its sensitivity/specificity in an independent home scalp EEG dataset. This study has diagnostic implications.

The immunobiology of herpes simplex virus encephalitis and post-infectious antibody-mediated disease

Encephalitis is a serious condition whereby the brain becomes swollen and inflamed. Herpes simplex virus is its leading cause with an untreated death rate of 70% reducing to around one-quarter with early treatment. In one-quarter of patients the illness can return leading to further severe health problems.

In recent times, the role of the immune system following herpes simplex virus encephalitis (HSE) has become increasingly recognised as a possible cause for such disability. The immune system may become confused and mistakenly attack the brain further whilst trying to control the virus. Therefore, controlling the immune system’s inflammatory reaction may help limit additional brain swelling. This is a valuable but underexplored area of research.

My work, in the world-renowned Oxford Autoimmune Neurology Group, will apply emerging knowledge of unwanted inflammation by the immune system, to its role in HSE. We will study different compartments of the body where the immune system may first be activated to explore direct mechanisms behind its disruption.

Our findings will explore whether tailored treatment targeting the immune system in HSE may help patients with this condition in the future. We will complete this work within three years and communicate our findings to researchers and the public through publications in journals, conference presentations and patient-public engagement days.

Investigating the role of non-canonical PIKfyve/PI(5)P autophagy pathway on progression of neurodegenerative disease

Maintenance of protein homeostasis is critical for cellular viability and involves strict control of protein synthesis, folding and degradation over organismal lifespan. Conversely, excessive accumulation of intracytoplasmic, aggregate-prone proteins is a hallmark of many age-related neurodegenerative disorders. Such misfolded proteins exert their effects via toxic gain-of-function mechanisms, thus processes involved in removal of the aggregate-prone proteins could represent a rational therapeutic strategy in a range of neurodegenerative disorders.

Autophagy, which is often likened to a cellular waste disposal system, is a highly conserved pathway where cells engulf various cytoplasmic contents into double-membraned autophagosomes and traffic their contents to lysosomes for degradation and recycling. Rubinsztein lab has previously demonstrated that intracellular aggregate-prone proteins are autophagy substrates and that autophagy plays a key role in clearance of a number of proteins associated with neurodegenerative diseases.

The aim of my project is to explore a novel, recently described autophagy pathway as a potential therapeutic target in neurodegeneration. I will perform overexpression experiments in vitro and in vivo using zebrafish models of neurodegenerative disease, including tauopathy and Huntington’s disease models. I will then assess its effects on autophagic flux, clearance of intracytoplasmic protein aggregates as well as disease progression. In the second part of my project, I will aim to identify pharmacological inducers of this pathway and most promising compounds in induced pluripotent stem cell-derived cortical neurons and further animal models.

Ultimately, our goal is to identify pharmacological autophagy stimulators that could have therapeutic potential in treatment of a range of neurodegenerative conditions.

The Pathophysiology of Myelin Oligodendrocyte Glycoprotein Antibody Disease

Myelin oligodendrocyte glycoprotein (MOG) antibody disease is an important disease to understand, potentially causing devastating disability in children, including infants, as well as adults by attacking the eyes, brain and spinal cord. It can leave people blind, immobile and with impaired cognition. It has similarities with multiple sclerosis, with both diseases targeting “myelin”, the insulation covering nerve fibres, causing “demyelination”. MOG antibody disease (MOGAD) is the commonest cause of acute demyelination in children.

MOGAD is diagnosed when people develop neurological disease with MOG antibodies in their blood. Antibodies are produced in healthy people to fight infections. In MOGAD, antibodies are erroneously produced against MOG, a protein found on myelin. It is unclear whether these antibodies cause demyelination in MOGAD. The function of MOG is currently unknown, but there is some evidence suggesting that it may play an important role in autoimmune diseases in the brain, where the body’s own defence system malfunctions and attacks the brain.

My research will focus on answering two questions: (1) Does MOG interact with an immune cell in the brain, called “microglia” (2) Do antibodies found against MOG in patients with MOGAD cause disease? Understanding this may help us to find better treatments for MOGAD and other similar diseases like multiple sclerosis.

Using digital biomarkers to assess cognitive, behavioural, and motor features of genetic frontotemporal dementia

Can we detect clinical changes in frontotemporal dementia (FTD) before the onset of symptoms using new digital technologies?

FTD is a common cause of dementia in those aged under 65, and around a third of cases are caused by genetic mutations. Studying this group of people allows us to characterise the disease in its earliest stages before symptoms begin by studying ‘at-risk’ family members. Understanding this phase is important for the design of clinical trials because it is likely that treatments will have their greatest effect in this early period, once the disease process has begun but before irreversible damage has occurred.

My work will take place within the UK arm of the GENetic FTD Initiative (GENFI) study, which has enrolled approximately 200 people at risk of FTD because of a genetic mutation within the family. So far GENFI has not investigated the ability of wearable devices and computerised tests to detect early changes of FTD from a person’s own home. Advantages of this approach over traditional evaluation include the reduced need to travel to healthcare institutions, particularly important given the COVID-19 pandemic, as well as the real-world setting of the testing.

My study will use computer-based tests and wearable devices to assess four domains: cognition, activity levels, upper limb function, and gait.

My aim is to identify early changes that might help guide the timing of enrolment of participants into clinical trials, and that can act as outcome measures when assessing the effect of a potential treatment for FTD.

Transcutaneous auricular vagus nerve stimulation (tVNS) for upper limb motor recovery in chronic stroke.

Between 30-60% of stroke survivors have lasting arm weakness. A recent ground-breaking clinical trial found that electrically stimulating a nerve called the vagus nerve, combined with physiotherapy, led to better recovery of arm strength than doing physiotherapy alone. However, the participants needed to have the stimulator inserted surgically and the electrical stimulations needed to be done in hospital.

The vagus nerve can also be stimulated through the skin using a simple earpiece– this is called transcutaneous vagus nerve stimulation (tVNS). This does not require an operation and treatment can take place at home. In a large upcoming study, participants with arm weakness from a previous stroke will receive tVNS alongside regular home physiotherapy sessions for 3 months to find out whether tVNS improves their arm strength.

How tVNS improves brain function and which patients benefit is not fully known. This fellowship will recruit participants from this upcoming trial to study how tVNS works. We will collect blood samples and perform special brain scans before and after treatment to find out whether tVNS reduces inflammation in the blood or changes the structure and function of the brain. For the millions of stroke survivors worldwide, tVNS could potentially be a cost-effective treatment to improve arm strength within the next 5-10 years. The results of this fellowship could help scientists and doctors discover how tVNS works, monitor its effects and determine which patients benefit from it.

Characterising B Cell Tolerance in Autoimmune Encephalitis

Encephalitis is a swelling of the brain which leads to confusion, mood changes and seizures. In the last decade, we have learned that this swelling is most commonly caused by an overactive immune system.

For reasons that we don’t yet understand, the immune system’s ‘B’ cells recognise healthy brain tissue in error. The B cells then produce proteins called antibodies which target the brain tissue and cause the swelling and damage.

The overactivity of B cells is likely to be caused by the failure of the normal checks and controls in the immune system. Knowing where these breakdowns occur would allow us to:

1. identify a fundamental disturbance which underlies the disease;
2. explain disease causation to patients; and
3. characterise the B cells which could be targeted with future treatments to reduce production of the causative antibodies and improve patient outcomes.

Once we understand where the immune system goes wrong, we can focus treatments that already exist on these targets. By exploiting this available knowledge, we can potentially rapidly bring treatments to patients, perhaps even within 2-3 years. In addition, the discovery of new targets in the faulty immune system will also allow us to develop new treatments.

As these checks and controls operate similarly across several conditions of the immune system, our findings will have wide ranging implications, across many neurological and non-neurological diseases.

Towards an optimal mouse model of CJD for testing treatments with fluid biomarker and automated behavioural readouts

Prion diseases are transmissible, fatal disorders of humans and other mammalian species. The most common human prion disease, sporadic Creutzfeldt-Jakob disease, manifests as a rapidly progressive dementia and has no effective treatment. Unlike other neurodegenerative disorders, laboratory mice are naturally susceptible to prion disease by inoculation and recapitulate the clinical features of a fatal neurodegenerative disease and the full spectrum of pathological changes seen in patients. Several therapeutic interventions are effective in mice when treatment is started pre-symptomatically. However, this paradigm is not a representative model for sporadic CJD, which is only diagnosed after symptom onset.

My project seeks to bridge the gap in mouse modelling by developing a profile equivalent to changes in clinical features and biomarkers in patients. I will characterise the field-standard RML prion-inoculated FVB mouse model of prion disease with fluid biomarkers and automated behavioural and gait analysis and develop a streamlined profile of parameters to measure progression of disease. Automated testing using the IntelliCage and MotoRater systems will measure activity, learning, cognitive flexibility and gait. These systems offer superior reproducibility and efficiency compared to traditional behavioural testing and reduce animal stress due to handling. Ultra-sensitive immunoassays will be used to measure neurofilament light chain and tau in blood, a convenient analyte for serial measurement to assess treatment response. These techniques will be used to test candidate therapeutics at the clinically relevant symptomatic stage, to assess the impact of therapies on behaviour, gait and biomarkers of neurodegeneration. I also plan to utilise specialised assays of infectious (but not directly neurotoxic) prions and prion-associated neurotoxicty to gain insights into the mechanisms of treatment success or failure.

My aim is to use these techniques to identify therapies, and potentially synergistic combination therapies, effective at the clinically relevant symptomatic timepoint, for progression to human study. This approach may have broader applications in preclinical selection of candidate therapeutics for neurodegeneration.

High-Throughput Single Cell Multi-omic Analysis of Multiple Sclerosis Patient Blood to Identify Causal Immune Cell Subsets

Multiple sclerosis (MS) is a common disease in which the immune system mistakenly attacks the brain and spinal cord resulting in neurological disability. The immune system is made up of cells that play an important role in fighting infection and cancer. Exactly how a healthy immune system goes wrong and causes MS is not known.

Currently, the most effective treatments for MS work by broadly suppressing immune cells to prevent this attack. However, these treatments can lead to potentially serious side effects from a weakened immune system and are also unable to completely cure the disease. Most patients will therefore eventually become permanently disabled.

This innovative study will examine the immune system of MS patients at a high level of detail with new cutting-edge technologies that can analyse the diverse array of immune cells on an individual level. Our aim is to identify and disentangle those cells that are causing the disease from those that are needed to maintain a healthy immune system and may also be protective against the disease.

Our findings aim to promote and guide drug development to specifically target these causal cells while sparing the healthy cells, which should be more effective and have fewer side effects. This would allow the most effective treatments to be safely given early after diagnosis in all patients, before they develop irreversible disability.

We hope to provide key insights into immune dysfunction in MS to greatly speed up the discovery of improved treatments for patients within the next decade.

Brain Waves, Bandwidth, and Behavioural Flexibility in Health and Disease

Mammalian brains exhibit a remarkable ability to reconfigure the way that different areas communicate. For instance, the same language centre can either process visual or auditory information. It has been proposed that our ability to switch between different information streams involves brain waves. Such waves consist of waxing and waning synchronization of electrical signals from groups of neurons (brain cells). Importantly, abnormal brain waves are seen in diseases such as epilepsy and encephalitis, and in common disordered brain states such as delirium and psychosis.

All communication systems require energetically expensive infrastructure with limited bandwidth. To provide flexible selective communication at scale and rapidly, these systems must use ‘multiplexing’ - the process of combining multiple signals into a single signal over a shared medium. Signals can be separated in time (e.g. Northern Line from Kings Cross to either Edgware every 3 minutes or High Barnet every 5 minutes), (2) by frequency (e.g. different AM radio stations) or (3) in space (e.g. different lanes on the highway).

Although correlational and computational evidence supports the idea that brain waves can implement multiplexing by time and frequency, definitive evidence is lacking. This research proposal aims to fill this gap, by manipulating the frequency, amplitude and phase of on-going brain waves in visual perception tasks in mouse models of health and neurological disorders, where gamma-frequency (25-140Hz) waves affects perception.

Cardiovascular and cerebrovascular health in patients with late onset unprovoked seizures and epilepsy- a target for secondary prevention? (CHeLOSE)

The primary aim of this project is to assess heart and brain blood vessel (cardio-/cerebro- vascular) health of patients with late onset seizures and epilepsy (LOUSE).

The number of patients with LOUSE is increasing due to an ageing population. Patients with unprovoked seizures in adulthood are at increased risk of stroke. It is not known whether this increased risk of stroke is due to untreated risk factors such as diabetes or hypertension; or whether there is impairment of brain vessel function irrespective of these. Presenting with a seizure in adulthood could be an important point in the patient’s journey to initiate vascular health assessment and treatment to reduce the future risk of adverse cerebrovascular (stroke) or cardiovascular (heart attack) events.

In this project I will undertake the following:

1. Assess vascular health in patients with LOUSE at diagnosis compared to healthy people and people with high blood pressure (a common risk factor for poor vascular health)
2. Assess vascular health over time, at 6 and 12 months, in patients with LOUSE and compare to patients with high blood pressure who have had their vascular risk factors managed
3. Establish a registry of patients with LOUSE to evaluate markers of poor vascular health, MRI brain imaging data, and review the rate of stroke and heart disease at 12 months.

This will provide insight into vascular health in patients with LOUSE and inform understanding of the relationship between cardiovascular risk factors and LOUSE. Results will inform clinical practice and future trials.