Intravital assessment of myelin molecular order with polarimetric multiphoton microscopy

Abstract

Myelin plays an essential role in the nervous system and its disruption in diseases such as multiple sclerosis may lead to neuronal death, thus causing irreversible functional impairments. Understanding myelin biology is therefore of fundamental and clinical importance, but no tools currently exist to describe the fine spatial organization of myelin sheaths in vivo. Here we demonstrate intravital quantification of the myelin molecular structure using a microscopy method based on polarization-resolved coherent Raman scattering. Developmental myelination was imaged noninvasively in live zebrafish. Longitudinal imaging of individual axons revealed changes in myelin organization beyond the diffraction limit. Applied to promyelination drug screening, the method uniquely enabled the identification of focal myelin regions with differential architectures. These observations indicate that the study of myelin biology and the identification of therapeutic compounds will largely benefit from a method to quantify the myelin molecular organization in vivo.

Figure 1. P-CARS zebrafish spinal cord model.

(a) The same spinal cord segment can be identified on different imaging sessions using morphological references. (b) The ventral spinal cord contains two Mauthner axons, one on each lateral side (image width: 80 μm). (c) The Mauthner axon can be easily identified with CARS microscopy (image width: 112.5 μm) based on the visualization of (d) its myelin sheaths (image width: 7 μm). (e) Imaging of the same location as in (c) with an orthogonal polarization orientation showing the modulation in CARS intensity. The arrows indicate the orientation of the linear polarization.

Figure 2. Evaluation of promyelination chemical treatments.

(a) The myelin thickness, (b) the MM, and (c) the focal myelination index (FMI) as a function of chemical treatments (Mann-Whitney test, n = 4). (d) CARS image of a Mauthner axon in a GC1-treated zebrafish. (d) Mapping of the MM from P-CARS reveals a discontinuity on the left side. Blue ROIs indicate where the polarization dependence is plot in (f). The black curve is from the focal myelin discontinuity (e-left, MM = 0.17 ± 0.04) and the red curve from the normal region (e-right, MM = 0.34 ± 0.03).

Figure 3. Developmental myelination of the Mauthner axon in live nacre zebrafish.

(a) CARS images of the Mauthner axon on different days (image width: 8.4 μm). Representative images were selected from longitudinal imaging experiments (repeated imaging over time of same fish). (b) The myelin thickness and (c) the MM are plotted as a function of larval age. A steady increase in MM, but not in myelin thickness, was measured between 3 and 5 dpf. R² for linear regression and p-value for non-zero slope test are given in each graph.

Figure 4. Electron microscopy confirms that P-CARS informs on the nanostructure.

(a) Electron microscopy images at 3 and 5 dpf show the difference in myelin lamellae organization. (b) The entropy decreases significantly between 3 and 5 dpf (Mann-Whitney test, n = 4).