Diffusion tensor imaging measures of white matter compared to myelin basic protein immunofluorescence in tissue cleared intact brains

We provide datasets from combined ex vivo diffusion tensor imaging (DTI) and Clear Lipid-exchanged, Anatomically Rigid, Imaging/immunostaining compatible, Tissue hYdrogel (CLARITY) performed on intact mouse brains. DTI-derived measures of fractional anisotropy (FA), radial diffusivity (RD), and axial diffusivity (AD) were compared to antibody-based labeling of myelin basic protein (MBP), as measured by fluorescence microscopy. We used a customized CLARITY hydrogel solution to facilitate whole brain tissue clearing and subsequent immunolabeling. We describe how CLARITY was made compatible with magnetic resonance imaging with the intention of facilitating future multimodal imaging studies that may combine noninvasive imaging with 3D immunohistochemistry. These data and methods are related to the accompanying research article entitled, ‘The role of myelination in measures of white matter integrity: Combination of diffusion tensor imaging and two-photon microscopy of CLARITY intact brains’ (E.H. Chang, M. Argyelan, M. Aggarwal, T-S. Chandon, K.H. Karlsgodt, S. Mori, A.K. Malhotra, 2016) [1].


Myelination
Radial diffusivity with the intention of facilitating future multimodal imaging studies that may combine noninvasive imaging with 3D immunohistochemistry. These data and methods are related to the accompanying research article entitled, 'The role of myelination in measures of white matter integrity: Combination of diffusion tensor imaging and two-photon microscopy of CLARITY intact brains' (E.H. Chang

Subject area
Neuroscience, Radiology More specific subject area

Diffusion tensor imaging and immunohistochemistry
Type of data Mice were transcardially perfused with a CLARITY liquid hydrogel, ex vivo brains were then immersed in PBS and 0.1 mM gadopentetate dimeglumine prior to DTI. During scanning, brains were immersed in Fomblin for susceptibility matching and to limit tissue dehydration.

Experimental features
TrackVis was used to create white matter regions-of-interest (ROIs) from DTIderived FA maps. CLARITY images were analyzed using Imaris 8.0 with the Surfaces module to trace individual MBP-positive ROIs.

Data source location
Manhasset, New York, USA Data accessibility Data is provided in this article and is related to the research article [1].

Value of the data
This is the first dataset of CLARITY whole-brain MBP immunolabeling and the first study, to our knowledge, combining DTI with CLARITY in the same brain samples.
CLARITY immunolabeling can be used in future studies to examine other protein targets found in brain structures with complex 3D architectures and distributions. This enables molecular analysis of brain targets across intact macroscopic circuits, while maintaining microscopic resolution.
Our dataset demonstrates that within-subject multimodal imaging studies can provide useful insight into the biological basis of noninvasive imaging methods.

Data
The dataset includes images, videos, and plots from experiments combining ex vivo DTI and twophoton microscopy of CLARITY mouse brains. Figs. 1-3 show both raw and analyzed data from these experiments indicating the relationship between various diffusivity measures and MBP immunofluorescence. Table 1 shows DTI-derived measures and MBP immunofluorescence values for major myelinated white matter tracts of the mouse brain.

Experimental design and animal subjects
For all experiments, we used C57BL/6J mice from Jackson Laboratories (Bar Harbor, ME). All animal procedures were approved by the Feinstein Institute Medical Research Institutional Animal Care and Use Committee and maintained according to National Institutes of Health guidelines. We first piloted our combined DTI and CLARITY [2] approach using Thy1-eYFP-H mice (n ¼3) that endogenously produce fluorescence signal without immunolabeling (Fig. 1C). Once we finalized a working protocol Transmittance values were normalized to H 2 0. Images on the right show an example whole brain prior to clearing and a CLARITY-cleared brain. (C) Example images of a CLARITY Thy1-eYFP-H intact brain demonstrating the potential for examination of fluorescence on multiple biological scales. that enabled DTI followed by CLARITY, we used four C57BL/6J mice for the main experiments. The mice were first perfused with a CLARITY liquid hydrogel solution, then brains were removed and underwent ex vivo DTI scanning. Subsequently the CLARITY hydrogels were polymerized, tissue cleared, MBP immunolabeled, and finally imaged intact using a two-photon laser scanning microscope. Details of each step can be found below and also in the research article [1].

ex vivo DTI
Mouse brains were scanned in a CLARITY liquid hydrogel containing 4% paraformaldehyde and 0.1 mM gadopentetate dimeglumine (Gd-DTPA). The Gd-DTPA was used as a T1 shortening contrast agent to achieve shorter repetition times (TR), thereby allowed faster DTI acquisitions while maintaining good signal-to-noise [3]. Diffusion-weighted magnetic resonance (MR) images were acquired on an 11.7 T NMR spectrometer (Bruker BioSpin, Billerica, MA) using a three-dimensional gradientand-spin-echo (3D DW-GRASE) sequence with twin navigator-echo phase correction [3], along 15 independent directions. A b-value of 1500 s/mm 2 was used and scanning time was $ 16.5 h for each mouse brain.

CLARITY clearing and MBP immunolabeling
CLARITY tissue clearing and immunolabeling were performed as previously described [2] with important modifications made to the hydrogel solution in order to obtain sufficient whole brain immunolabeling, while also maintaining tissue rigidity. We experimented with different concentrations of paraformaldehyde (PFA) and bis-acrylamide before arriving at an optimal hydrogel mixture containing 4% paraformaldehyde, 1.75% acrylamide, 0.01875% bis-acrylamide, 0.25% VA-044 initiator (Wako Chemicals USA) and 1X PBS. This resulted in a hydrogel-tissue matrix that facilitated lipid clearance and allowed for deep immunolabeling of whole mouse brains. Instead of electrophoretic tissue clearing (ETC) at high temperature, we used passive clearing at 37°C (without agitation) in order to minimize changes in overall brain volume that commonly occur with ETC. While this passive technique is markedly slower, we found that it achieved excellent tissue transparency and structural preservation without tissue inflation ( Fig. 1A and B). Following clearing, intact brains were immunlabeled with anti-MBP (1:50, EMD Millipore), followed by secondary antibody labeling with   [4]) and imaged using an Olympus FV1000-MPE with Mai Tai DeepSee Ti:Sapphire laser (SpectraPhysics).

DTI-CLARITY analysis
From the diffusion-weighted MR images, maps of FA, AD, RD, and MD (Table 1; Fig. 3) were calculated using FSL (http://www.fmrib.ox.ac.uk/fsl) software. TrackVis (http://www.trackvis.org) was used for post hoc calculations and visualization. White matter regions-of-interest (ROIs) were drawn in TrackVis from the FA map. For visualization and analysis of two-photon image stacks, we used Imaris 8.0 software (Bitplane). ROI-based analyses (Fig. 2) were conducted in the DTI space and CLARITY space with statistical comparisons using Spearman's rank correlations or mixed effect models performed in R software (version 3.0). Statistical significance for all tests was defined as P o0.05.