Development of a high-throughput drug screening platform for oligodendrocyte myelination (for progressive multiple sclerosis)

Abstract

Aim As part of the broad strategy in Edinburgh and beyond to discover new treatments for progressive Multiple Sclerosis (MS), the aims for my PhD project were: (1) to address the lack of an in vitro phenotypic drug screening platform that is able to fully recapitulate myelin sheath formation, with the long-term goal being to enhance the discovery of pro remyelination therapies in progressive MS, and target the poor drug discovery rates in brain disorders which is partially due to poor disease modelling, and (2) the exploration of non linear optical imaging microscopy techniques that targets both the resolution and speed to study myelination.

Background In multiple sclerosis, an inflammatory autoimmune process destroys the oligodendrocytes that provide neuronal support by forming multi-layered compact myelin sheaths around axons, leading to neurodegeneration. Although there are drugs available to suppress the inflammatory attack to limit the formation of demyelinated lesions, no treatments currently exist to promote the regeneration of these myelin sheaths once the damage has occurred. Cell based screens have formed an important part of the strategy used to discover such regenerative drugs. The majority of published cell-based phenotypic drug screens to target repair have focused on the differentiation of oligodendrocyte precursor cells into oligodendrocytes rather than their ability to form mature protective myelin sheaths. However, in many MS lesions, pre-myelinating oligodendrocytes are present, and studies on oligodendrocyte biology show that differentiation and myelination are regulated by distinct mechanisms. There is therefore a need for novel drug screens that target the later myelinating stages of oligodendrocyte development.

Results Using a 3D microfibre system for in vitro myelin sheath formation (described by Bechler, Byrne and ffrench-Constant (2015)), I first asked whether compounds that had been v

identified as increasing differentiation in conventional 2D culture systems enhanced myelination in the 3D cultures. While Benztropine and Clemastine showed an increase in the number of MBP+ (differentiated) oligodendrocytes in the 2D system, consistent with previous publications, no increase in myelin sheath formation was seen with any of the drugs in the 3D system, highlighting the potentially important differences between differentiation and myelin sheath formation for drug discovery. Next I developed a multiwell plate based assay to allow 3D myelination assays to be used for drug screening. Using electrospinning to produce PLA microfibres, I was able to develop and optimise a technique to insert and suspend the fibres across the bottom of a 96-well plate that can be incorporated into an automated pipeline for high-throughput drug screening. The 3D myelin sheaths could be imaged using the Leica SP8 confocal system with the MatrixScreener extension and the Opera Phenix high-content screening system. Finally, I addressed the problem of imaging myelin in such screens. CARS was shown to be able to preferentially detect myelin sheaths in fixed and live slices in regions and time-points of varying myelin densities. As the development of new myelin sheaths requires the formation of lipid and therefore the incorporation of hydrogen, the consumption of D2O (heavy water) with a 2H atom allowed the non-invasive labelling and detection of myelin sheaths using SRS. Future experiments will allow us to confirm whether this deuterium detection is preferential and/or specific to new myelin sheaths.

Significance With only about 10% of drugs that enter Phase I trials successfully launching into clinics, there is an important need for more effective drug screens that better model disease-relevant processes and so reduce late-stage failures. The high-throughput-compatible 96-well plate with suspended PLA microfibres that combines the recent progresses in 3D cellular model systems with bioengineering and the recent advances in high content imaging systems may give us the opportunity to more accurately model these disease-relevant structures in vitro, and therefore improve drug discovery for regenerative therapies in multiple sclerosis and other myelin diseases. The research on Raman-based label-free imaging of myelin sheaths is not only applicable for imaging myelin sheaths in the context of drug validation, but could be important for live imaging of brain slices and detection of newly-formed myelin sheaths without the need for complex and expensive transgenic animals.