Brain tissue preparation for EM

31

After wet brain sectioning and LM imaging, several wet brain sections containing the center portion (in the medial-lateral axis) of HVC of each experimental bird were prepared for EM study. The following procedure was adapted from (Knott et al.

2008), to increase the cell membrane contrast and reduce the background staining of the intracellular space in the EM.

Figure 2. 3: Localization of HVC with LM and EM. (a) A wide-field LM image of a wet brain section. The red dashed oval outlines the HVC region, as also seen in (b), (c) and (d). (b) Wide-field LM image of the same brain section after EM staining, embedding and ultramicrotomy. (c) An electron micrograph of serial ultrathin sections that were cut from the surface of the brain section in (b). (d) An electron micrograph of the brain tissue block taken with FIB-SEM after an imaging session. The red arrow indicates the region where the imaging and milling occurred, which resulted in a trench in the tissue.

Each selected wet brain section was flattened on the bottom of a glass vial. The EM staining solutions were then added to and removed from the vials with either a glass pipette (during osmification) or plastic pipette (during dehydration). The sections were first washed in cacodylate buffer (0.1 M, pH 7.4) and then incubated in 1.5%

potassium ferrocyanide and 1% OsO4 in cacodylate buffer for 40 minutes, following another 40-minute incubation in only 1% OsO4. Afterwards, the sections were incubated in 1% uranyl acetate in double distilled water (ddH2O) for another 40 minutes. After these heavy metal stainings, the sections were dehydrated by 10-minute incubations each in a gradient of ethanol solutions with increasing concentrations (50%, 70%, 90%, and 95%). The sections were further dehydrated

32

with twice incubations in 100% ethanol following twice incubations in propylene oxide, each for 15 minutes. During the dehydration, small amounts of liquid from the preceding steps remained inside the vial to prevent the tissue from drying and cracking. After dehydration, the sections were immersed into a freshly made epoxy resin for at least 12 hours to reach complete tissue infiltration. The epoxy resin (Durcupan ACM, Sigma-Aldrich Fine Chemicals, Buchs, Switzerland) consisted of:

10 g : 10 g : 0.3 g : 0.2 g of component A/M, B, C, and D, respectively. Component C was stored at 4°C, while the other three components were stored at room temperature. After incubation, the sections were placed between two Aclar films (ACLAR^® sheets, Agar Scientific Ltd., Stansted, UK) that were sandwiched between two glass slides. A small weight (~40 g) was placed on top of the upper glass slide. The sections were then cured for 48 hours at 52°C. After curing, the sections should be sufficiently hardened for subsequent ultramicrotomy.

Before proceeding to ultramicrotomy, first, the ROI in HVC was located by comparing the LM images of the same section before and after embedding (Figure 2.

3, a and b). Once HVC was identified, the location was drawn with a red-color marker onto the Aclar film. The two Aclar films surrounding the sample were separated with the embedded tissue usually sticks to one of the two films. With a scalpel blade (cat. # 10050-00; Fine Science Tools GmbH., Heidelberg, Germany), a small piece of tissue containing HVC was dissected from the marked region. The small tissue piece was then glued on top of a resin block with the brain tissue side facing down (remaining Aclar film facing up).

The resin block was then mounted onto an ultramicrotome (Leica FC6; located at The Scientific Center for Optical and Electron Microscopy (ScopeM), ETH Zürich, Zürich, Switzerland). Referred to the LM images of the same tissue section, HVC was located and trimmed in a trapezoidal shape with a diamond trimming knife (trimtool 20, DiATOME Ltd., Nidau, Switzerland) (Figure 2. 3, b). The Aclar film on top was trimmed away, and the surface of the brain tissue block was polished.

After trimming, another diamond knife (Histo Jumbo, DiATOME Ltd) was used with a bath filled with pure water. Serial 70-nm-thick ultrathin sections were then cut. The sections formed a ribbon of tissue that were floating on the surface of the water bath. During the cutting, the thickness of the ultrathin sections can be monitored with their reflection index (golden: >100 nm; silver-gray: 50 – 100 nm).

Sections that were too thick were discarded from the collection. When the desired ribbon of serial ultrathin sections was produced, it was detached from the diamond blade and moved on the surface of the water bath with a human eyelash that was fixed to a toothpick. A custom-made silicon wafer (Si-Mat Silicon Materials, Kaufering, Germany) was first deionized with a charging generator (EN SL, Haug Biel AG, Biel, Switzerland) and then slowly dipped into the water bath. The floating ribbon of ultra-thin sections was flattened and moved to and attached onto the silicon wafer, and both were withdrawn from the water bath. These procedures enabled the collection of the intact ribbon of ultrathin sections onto the silicon wafer, and both were dried and stored. Prior to the ssSEM imaging of the ultrathin sections,

33

the silicon wafer was fixed with clips onto a SEM sample holder (cat. # 16112-20, Plano GmbH, Wetzlar, Germany).

In contrast, after the ultramicrotomy, the brain sample block with the polished surface was then fixed onto a SEM sample stub holder (cat. # G301F, Plano GmbH).

The fixed sample block were then carbon-coated (20 nm, Bal-tec MED 020; located at The Center for Microscopy and Image Analysis (ZMB), University of Zürich) to improve surface conductivity. The carbon-coated tissue block was either stored or immediately processed for FIB-SEM imaging.