Chemical clearing and dehydration of GFP expressing mouse brains

Abstract

Generally, chemical tissue clearing is performed by a solution consisting of two parts benzyl benzoate and one part benzyl alcohol. However, prolonged exposure to this mixture markedly reduces the fluorescence of GFP expressing specimens, so that one has to compromise between clearing quality and fluorescence preservation. This can be a severe drawback when working with specimens exhibiting low GFP expression rates. Thus, we screened for a substitute and found that dibenzyl ether (phenylmethoxymethylbenzene, CAS 103-50-4) can be applied as a more GFP-friendly clearing medium. Clearing with dibenzyl ether provides improved tissue transparency and strikingly improved fluorescence intensity in GFP expressing mouse brains and other samples as mouse spinal cords, or embryos. Chemical clearing, staining, and embedding of biological samples mostly requires careful foregoing tissue dehydration. The commonly applied tissue dehydration medium is ethanol, which also can markedly impair GFP fluorescence. Screening for a substitute also for ethanol we found that tetrahydrofuran (CAS 109-99-9) is a more GFP-friendly dehydration medium than ethanol, providing better tissue transparency obtained by successive clearing. Combined, tetrahydrofuran and dibenzyl ether allow dehydration and chemical clearing of even delicate samples for UM, confocal microscopy, and other microscopy techniques.

Figure 1. Chemical structures of tetrahydrofuran (THF) and dibenzyl ether (DBE).

A: THF. B: DBE. Interestingly, both compounds are ethers without further functional groups.

Figure 2. Comparison of THF/DBE clearing with ethanol/BABB.

A–B: Left and right brain hemispheres from two mice (postnatal day 101). The left hemispheres of each animal were dehydrated with ethanol, or THF. The corresponding right hemispheres were dehydrated with ethanol as a control. Afterwards, the hemispheres were rendered transparent with BABB (A1 and B1. C: Left and right brain hemisphere of an exemplary mouse (postnatal day 101). The left hemisphere was dehydrated with THF and cleared with DBE. The right hemisphere was dehydrated with ethanol and then cleared with BABB as a control (C1). From all hemispheres 3D reconstruction were performed by UM (A2–C2, Olympus objective XLFluar 4×, N.A. 0.28). As evident from A–C, THF-dehydration provides an improved fluorescence signal (B2), and considerably reduced background fluorescence (B2), while DBE-clearing provides considerably better tissue transparency (C1) and strikingly better fluorescence preservation (C2). The black bars below A2–C2 represent intensity histograms of the respective image stacks used for 3D reconstruction. Length of scale bars in Figures A1–C1 is 2 mm. Scales are identical for all histograms A2–C2 and thus are only shown in the upper two histograms.

Figure 3. UM-reconstructions of isolated hippocampi of a thy-1 EGFP-M mouse.

DBE clearing provides improved fluorescence and visibility of details. A: left hippocampus dehydrated with THF and cleared with BABB. B: corresponding right hippocampus dehydrated with THF and cleared with DBE. A1–B1: Olympus objective XLFluar 4×, N.A. 0.28, scale bar 100 µm. (A2–B2) 4× Olympus objective XLFluar 4×, N.A. 0.28, scale bar 20 µm. A3–B3: 4× Olympus objective XLFluar 4×, N.A. 0.28, scale bar 10 µm. CA1: cornu ammonis region one. gl: granular cell layer py: pyramidal cells.

Figure 4. Dehydration and clearing chemicals tested.

Most of the substances found to be GFP-incompatible possess hydroxyl, or carbonyl groups.

Figure 5. Removal of peroxides from THF and DBE.

A: Apparatus for peroxide cleaning of THF 1: Dropping funnel with pressure compensation. 2: Chromatography column filled with basic activated aluminum oxide activity grade Brockman 1 3: Two necked round bottom flask 4: Drying tube filled with calcium chloride. B: Apparatus for peroxide removal in DBE and BABB 1: Filter unit with filter plate (16–40 µm pore size) 2: Vacuum tight filtering flask.