Cryo-Confocal Imaging
Figure 2: LM imaging of frozen tissue and identification of the target spine imaged with 2-photon LM. A) Low-magnifi- cation view of frozen tissue containing hippocampal pyramidal neurons expressing GFP. B) Higher magnification of the same region (box in A). The indexed grid provides both a substrate for the tissue and a system for identifying the location of the target cell. Arrow shows a clear fiducial that can be used to estimate the position of other alpha-numeric fiducials (overlaid onto the image). Circle shows location of target cell at C3. C) View of the sample following EM sample prepara- tion and the location of the cell (circle). D) Epifluorescence overview of the target cell. E) Higher-magnification view of the same cell using CLSM. Outlined arrowhead indicates the base of the dendritic branch having the target spine. F) Two- photon GFP imaging done prior to confocal imaging shows the same structure as seen in E. Outlined arrowhead indicates the base of the same dendritic branch in E. G) High-magnification view of the same region (box in F). Arrowhead marks the target spine that received glutamate uncaging. Scale bars: A–C, 100 μm; D, 50 μm; E–F, 10 μm; G, 5 μm.
Tissue was then post-fixed for 10 min in 2% GA buffered in PBS, rinsed with PBS followed by a water rinse, and placed in a solution of Silver Enhancement Kits (HQ Silver, Nanoprobe) for 9 min. Tissue was rinsed in water and then treated with 0.5% OsO4
for 20 min and 1% aqueous UA for 35 min. Tissue
was dehydrated in a graded series of ethanol and acetone, infil- trated into Durcupan resin (Sigma), then flat embedded and polymerized for 2 days at 60°C. Sectioning and ssSEM image acquisition. A piece of containing the neuron of interest was
approximately 1 mm2
trimmed from the tissue using the CLSM image as a map. Serial sections were cut at 50 nm and collected onto Kapton tape using the ATUMtome. Te tape was arrayed onto 10 cm silicon wafers using double-sided carbon tape and imaged using the ZEISS Gemini 300 and ATLAS 5 AT soſtware. Mid- magnification images were captured at 5,600× magnification, 5 kV accelerating voltage, 1.3 nA beam current, and 20 nm/ pixel resolution using the Gatan OnPoint™ detector. High- magnification images were captured at 28,000× magnification and 4 nm/pixel. All manual image correlation was performed with Adobe Photoshop CS6, and image stacks were aligned using the TrakEM2 plug-in of Fiji [13].
Results Te vitrified brain slice was kept at liquid nitrogen tempera- ture within the Linkam cryo-stage during imaging. First, we used
2021 September •
www.microscopy-today.com
epifluorescence to visualize the entire shape of the frozen slice. Ten we used CLSM to locate and identify the GFP-expressing neurons in the >150 μm thick- ness of the tissue (Figure 2). Te fluorescence signal was signifi- cantly impacted by the inhomo- geneous thickness of the frozen sample, caused by the uneven thickness of the slice aſter cul- turing for days, the amount of the media filled in the carrier, and frost accumulation on the sample during imaging. To assist correlation between the LM and EM images, we first captured an overview image of the entire slice with the index grid at low mag- nification using epifluorescence (Figure 2A). Because only a handful of neurons per cultured slice were labeled with GFP via biolistic transfection, we were able to identify the target cell by matching the general shape of the neuron from the 2-photon images to the epifluorescence image. Although thicker areas of the tissue tended to obscure any of the underlying fiducial markers of the index grid, we always used the areas with vis-
ible letters of the grid as landmarks to estimate the location of the target cell (Figure 2B). Tis greatly narrowed our examination area for the neuron in the embedded tissue and let us trim the slice to a smaller size (Figure 2C). Once we identified the target neuron in epifluorescence (Figure 2D), we switched to CLSM to see more details (Figure 2E). Using the original 2-photon images as guides (Figure 2F, 2G), we could trace the dendrite in the confocal image outward from the cell body until we found the specific dendritic segment containing the targeted spine (Figure 2E). We then searched for the target spine by capturing z-stack
images of the target dendritic segment, using both 10× (C Epi- plan-Apochromat, NA 0.4, ZEISS) and 20× (Plan-Apo, NA 0.8, ZEISS) objectives. Ultimately, we were limited to using the 10× objective, as the 20× objective did not provide a sufficient SNR to depict the target spine. A buildup of ice and frost on the sur- face of the tissue further diminished the image quality, caus- ing small structures such as spines to be almost impossible to see using standard confocal detection (Figure 3A, arrowhead). By switching to the more sensitive Airyscan detection on the LSM 980, we were able to increase the SNR enough to visual- ize and identify the spine of interest (Figure 3B, arrowhead). Troughout the imaging process we observed negligible pho- tobleaching, which enabled us to repeatedly image the neuron even with extended beam dwell times. Aſter the cryo-CLSM, the sample went through the freeze- substitution process, and we performed an additional step of
37
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80
