A high-throughput spheroid fusion platform for the templated-assembly of 3D neuromuscular junctions

Lead Research Organisation: UNIVERSITY OF EXETER
Department Name: Physics

Abstract

Neurons help control movement by sending conveying electrical impulses from the brain and spinal cord to the muscles. The connection between the neurons and muscles happens in the human body at specific places called neuromuscular junctions (NMJ). Understanding the interactions between neurons and muscle cells is crucial to understand neuromuscular development and dysfunction. At the moment, there are no available human three-dimensional models that can be used to study this junction, hence animal models are routinely used instead. However, animal NMJ differs significantly from the human ones. Hence, it is critical to develop accurate human models of the NMJ. In the past few years, our team has developed novel technologies that permit the formation of lab-grown tissue that can be used to recapitulate human physiology.

In this project, we will create a novel in-vitro model neuromuscular junction to alleviate the need for animal models. We will grow both neurons and muscle cells into three dimensional cultures and force them to assemble using networks of microtraps. We will generate for the first time truly 3D organoids with hundreds of such junctions per experiment, a number at least one order of magnitude higher than previous attempts. The junctions will be validated by measuring muscle contraction upon chemical stimulation of the neurons. This will ensure that the models we create have the highest relevance to human physiology and can be used as a platform for future biomedical research.

Technical Summary

We will controllably fuse pre-formed 3D cell cultures (diameter 150-200 microns) derived from iPSC-derived spinal motor neurons or skeletal muscle cells in a single microfluidic device. To achieve high reproducibility, we will include a spheroid sorting step based on real-time brightfield imaging. The PI has pioneered this approach with established sorting rates of up to 40 spheroids/second with a single stage deep learning object detector (YOLOv4). In this project we will train a novel model to recognise neuron and muscle spheroids based on either size or morphology. We will program sorting of both types of pre-formed 3D cell cultures in alternation in flow. The sorted spheroids will be directed towards a microfluidic trapping and pairing array in which spheroids of either type will be trapped in alternation. The array is expected to hold up to 400 spheroid pairs. Long-term perfusion under controlled environmental conditions will permit subsequent formation of neurites and establishment of self-organised spinal organoids.

As a second step, we will quantify spheroid fusogenicity, by using two types of measurements:
1. A 3D morphological assay, in which we will track neurites, and key morphological parameters of the doublets as a function of time using confocal fluorescence microscopy.

2. Map the presence of acetylcholine receptors, indicating the presence of neuromuscular junctions. We will extract selected spheroids, slice them into micron-sized sheets and image the sheets using fluorescently-labelled alpha-Bungarotoxin.

Finally, we will assess functionality of the NMJs using glutamate as excitatory neurotransmitter. Glutamate will be added to the perfusion media to induce rapid muscle contraction which will be assessed via bright-field and fluorescence imaging of the muscle cells, comparing the volume of the spheroids before and after contraction.

Publications

10 25 50