Degeneration or breakdown of the myelin sheath that wraps and insulates axons of the central nervous system,
as well as peripheral nerves, is a factor in a large array of neurological disorders. These include multiple
sclerosis, stroke, age-related neurodegenerative diseases like Alzheimer’s, and in non-neurodegenerative age-
related cognitive impairment. To date, myelin breakdown has been assessed indirectly with neurophysiological
assays of conduction, post-mortem myelin staining, or in-vivo diffusion MRI. In the normal aging rhesus monkey,
post-mortem electron microscopy (EM) has confirmed myelin breakdown at the ultrastructural level; however,
EM studies cannot be scaled up to evaluate the entire 100-cc monkey brain or the larger human brain. Myelin
stains suffer from various technical and practical limitations that impede quantification. Other methods, like label-
free spectral confocal reflectance imaging (SCoRe), and coherent anti-Stokes Raman spectroscopy (CARS), are
promising, but like EM, are expensive, slow and difficult to apply to large brain sections. While SCoRe and CARS
microscopy work in reflectance, opening the possibility of in-vivo imaging through a cranial window in the small-
animal brain, a complementary method is needed that is validated, quantitative and high-resolution, to survey
the status of myelin in whole-brain sections of monkey or human brain. Here we propose to optimize and
validate quantitative birefringence microscopy (qBRM) for high-resolution imaging of normal and
abnormal myelin. The qBRM image provides, for every pixel, two quantitative measures: a) the relative phase
retardance, which is linearly proportional to the density of the birefringent medium and the organization of its
anisotropy; and b) the local orientation of the optic axis of the myelin, which corresponds to the direction of the
structural anisotropy. Archived brain tissues from two rhesus monkey models of myelin damage are available for
this project at the Boston University Medical Campus (BUMC). In Aim 1 we will automate the function of our
birefringence microscope with computer-controlled, motorized polarizer components and stage-positioning, to
facilitate fast, automated image acquisition (montaging) of large tissue sections (up to 40x40 mm). We will also
employ massively-parallel computing resources at BU to rapidly render the multi-megapixel quantitative images.
In Aim 2 we will validate our qBRM image measures by comparing them, in the same sections, with myelin
identified by myelin stains, such as FluoroMyelin Red, and with label-free spectral confocal reflectance (SCoRe)
microscopy. In Aim 3, We will demonstrate the untility of qBRM for quantifying myelin by imaging tissue sections
from three monkey models of myelin pathology: a) behaviorally characterized young and old monkeys with
cognitive assessment; b) behaviorally characterized old monkeys with impaired hand movement produced by
motor-cortex injury and treated with mesenchymal stem-cell derived extracellular vesicles (EVs); c) behaviorally
characterized old monkeys with demonstrated cognitive impairments treated with EVs. Together this will
establish the validity and utility of qBRM for assessing myelin in the central nervous system in health and disease.
