Delayed, subcutaneous neurotrophin-3 infusion improves recovery after stroke: mechanisms of recovery, automation of testing and markerless kinematics

Background: Stroke occurs in the brain when a clot restricts blood flow or when a blood vessel breaks. Without oxygen and energy, brain cells die rapidly and the opposite side of the body is often disabled, with a hand often weak or impaired. There are no fully restorative therapies for stroke. Millions of people are affected.

Neurotrophin-3 (NT3) is a growth factor normally produced by muscles during infancy in humans and other mammals. It is essential for wiring up the nervous system from muscles to the spinal cord and these circuits are important for normal movement. Levels of NT3 decline with age. We set out to determine whether supplementary NT3 protein can improve outcome after stroke when it is administered in a clinically-feasible time frame.

Innovation: Our work is novel because we have discovered that infusion of NT3 subcutaneously (under the skin) normalises a spinal circuit that controls a hand muscle, and improves dexterity and walking in elderly rats after stroke.

Clinical relevance: These discoveries are exciting for three reasons:

1) We start treatment 24 hours after CNS injury. Most new stroke survivors could be treated within this time frame (unlike the existing clot-busting therapies for stroke which need to be given within a few hours).

2) We deliver the treatment by a clinically-straightforward route that does not require invasive neurosurgery and could, in principle, be delivered during rehabilitation using existing small, wearable infusion pumps.

3) Clinical trials have shown that repeated, subcutaneous, high doses of NT3 are well-tolerated and safe in more than 200 humans. This paves the way for NT3 as a therapy for stroke.

Gaps in knowledge: It is not yet known how Neurotrophin-3 promotes recovery after stroke.

Perhaps surprisingly, our evidence indicates that NT3 promotes recovery by travelling in the bloodstream and by binding to "receptors" (molecular targets) that are located outside the central nervous system on peripheral sensory neurons that convey information about body position from muscle to the spinal cord. This grant seeks to understand how subcutaneously infused NT3 promotes recovery because this information could help a future clinical trial of NT3 in stroke survivors.

Aim 1: We wish to assess the mechanisms whereby NT3 improves arm and hand function in mice. NT3 will be infused subcutaneously 24 hours after stroke for one month. Dexterity and mobility will be measured weekly for 12 weeks. Our "MouseBots" are novel in-cage automated devices which automatically assess grasping in group-housed mice 24/7 (freeing us from repetitive manual testing). Neural stimulation and recording will be performed fortnightly to assess recovery of spinal circuits involved in movement. To confirm the identity of the critically-important neural pathways that restore dexterity after NT3 treatment, we will silence pathways from the brain using inducible genetic and pharmacological methods.

Aim 2: We will discover where NT3 acts. We predict that deletion of the TrkC receptors in peripheral sensory neurons (and not neurons in motor cortex) of transgenic mice after stroke will prevent recovery induced by NT3. If true, this would prove that this therapy does not need to enter the brain or spinal cord for it to promote recovery, and would facilitate clinical trials.

Aim 3: We will automate all our behavioural analyses by developing software which uses artificial neural networks to track digits, snout and pellets (during reaching and grasping) and limb movements (during walking) without the need to attach reflective markers to joints. This will enable us and other users worldwide to accelerate research into stroke and other diseases affecting limb movements.

These experiments will help us take this potential therapy one step closer to clinical trials.

After stroke, sensorimotor reflexes become abnormal; grasping and walking are commonly impaired. Neurotrophin-3 (NT3) is a growth factor that is made at high levels in infant skeletal muscles and which is required for development of proprioceptive circuits from muscle to the spinal cord. However, the levels of NT3 are very low in adult and elderly humans. We have shown that elevating the levels of NT3 normalised a proprioceptive reflex to a hand muscle and promoted sensorimotor recovery in adult and elderly rats when infused subcutaneously for one month, starting 24 hours after stroke. Excitingly, five trials show that NT3 is well tolerated when given subcutaneously at high doses to people with other conditions.

We will test the hypothesis that:

Neurotrophin-3 normalises proprioceptive spinal circuits and promotes sensorimotor recovery after stroke by binding to TrkC receptors on sensory neurons in dorsal root ganglia when subcutaneous infusion is initiated after 24 hours.

Aim 1: We will assess the mechanism of recovery of arm, hand and leg movements in wildtype mice after large cortical strokes and 24h-delayed, one-month subcutaneous infusion of NT3. We will assess dexterity using novel automated in-cage MouseBots and we will assess mechanisms of recovery using viral tracing and chemogenetic silencing, neurophysiology and biochemistry.

Aim 2: We will determine whether deletion of TrkC neurotrophin receptors in Dorsal Root Ganglia or neurons in motor cortex 24h after stroke prevents recovery after subcutaneous NT3 infusion. We will use inducible, conditional genetic strategies to delete TrkC neurotrophin receptors in peripheral sensory neurons or neurons in motor cortex after stroke.

Aim 3: We will use deep learning algorithms to track limbs and digits without the need to attach reflective markers to the joints, allowing automated kinematics and quantification of recovery.

These advances will help move this therapy towards the clinic.

Aim 1 and Aim 2:

Our ultimate goal is to develop a safe and effective therapy which improves hand, arm and leg function after stroke in humans. The translational potential for NT3 is described in our Case for Support. If the MRC funds this intermediate step and it is successful, then Phase I/IIa human clinical trials might be warranted as NT3 has already been shown to be safe in humans when repeated high doses are given subcutaneously.

Potential beneficiaries and benefits might be as follows:

1. Stroke survivors. Worldwide, there are 31 million stroke survivors, with another 9 million new strokes annually. Some ischemic strokes can be treated with clot-busting therapies but treatment after a few hours is very limited (e.g., rehabilitation) and is rarely fully restorative. There are no neurorestorative drugs approved for use after stroke. Success would benefit elderly people as well as young adults: >91% of EU stroke victims are >65 years old. Moreover, NT3 might be suitable for treating both ischemic and haemorrhagic stroke because its proposed mechanism of action is common to both.

2. Wealth: Each year, stroke costs the EU 38 billion Euros and the USA $53.6 billion. In the UK, the economic cost of stroke exceeds £8.9 billion per annum; the direct care costs constitute 5% of the National Health Services' budget while the indirect costs (costs to social services and carers, loss of earnings) are enormous (Saka et al 2009). Yet stroke research is enormously underfunded relative to diseases of comparable burden (Pendlebury, 2007); increased funding of pre-clinical and clinical stroke research is required.

3. Health and quality of life: A stroke therapy that conferred even a modest amount of leg, arm or hand recovery could lead to significant benefits in terms of independent living and in terms of reduced, costly dependence on carers, hospitals and social services.

4. Spinal cord injury survivors and their carers. Spinal cord injury affects more than 2.5 million people worldwide, and more than 130,000 new cases are reported each year. The estimated annual cost in the UK is over £500 million pounds and in the USA is $7.7 billion.

5. Economic competitiveness: Historic patents covering NT3 (held largely by US companies) have expired or expire soon. This increases the UK's 'freedom to operate' and prospects of for-profit or not-for-profit production of NT3 for therapy are increasing.


Aim 3: An automated method for kinematic analysis would enable more labs to adopt this sensitive measure of impairment. Automation enables collection of more data from more animals (and more data from each animal) which will improve experimental power and estimates of precision, respectively. A measurement of impairment (e.g., kinematics) may be better suited than measurements of function/activity for evaluating neurorestorative therapies (Krakauer and Carmichael, 2017). Using similar kinematic outcome measures in both animal trials and human clinical trials may help close the "translational disconnect". Moreover, kinematics can help researchers to distinguish true recovery from compensatory movements; this could help inform rehabilitative co-therapy.

Additional Impact:

A) MRC Strategic Skills Priorities: This project addresses three of MRC's Strategic Skills Priorities: 1) Advanced in vivo sciences, 2) Regenerative Medicine and 3) Interdisciplinary Strategic skills. Sotiris Kakanos will be trained in these additional methods to amplify benefit.

B) MouseBots enable scaling-up of experiments. This could enable other researchers in academia and industry to assess therapies for stroke more quickly, more cost-effectively, and with less bias (because the MouseBots assess each mouse automatically without knowledge of group).

Acceleration of stroke research is vital because in England alone, a person suffers stroke every five minutes.

Pendlebury (2007) Int. J. Stroke 2:80-87
Saka et al (2009) Age and Ageing 38:27-32