1. Neuroscience
  2. Structural Biology and Molecular Biophysics
Download icon

High-quality ultrastructural preservation using cryofixation for 3D electron microscopy of genetically labeled tissues

Tools and Resources
Cite this article as: eLife 2018;7:e35524 doi: 10.7554/eLife.35524
5 figures, 2 tables, 5 data sets and 1 additional file

Figures

Flowchart of the CryoChem method.

After cryofixation by high-pressure freezing and freeze-substitution, cryofixed samples are rehydrated gradually. Rehydrated samples can then be imaged for fluorescence, subjected to DAB labeling reaction or en bloc stained with a substantial amount of heavy metals. The protocol is modular; the first three processes are the core steps of CCM and the starred steps are optional depending on the experimental design. The samples are then dehydrated for resin infiltration and embedding, followed by imaging with any EM technique of choice. Blue and grey denote hydrated and dehydrated states of the sample, respectively.

https://doi.org/10.7554/eLife.35524.003
Figure 2 with 2 supplements
CryoChem method offers high-quality ultrastructural preservation and sufficient en bloc staining for SBEM.

EM images were acquired to assess the morphology of CCM-processed tissues. (A-B) The quality of preservation was markedly improved in the CCM-processed Drosophila antenna compared to the chemically fixed counterpart. Pixel resolution of SBEM images (x,y): 6.5 nm. (A1 and B1) Unlike the CCM-processed antenna, the chemically fixed antenna showed signs of extraction (arrow) and disorganized membranes. ORN: olfactory receptor neuron; ML: microlamella. Scale bars: 1 µm. (A2 and B2) The microlamellae were well-preserved in the CCM-processed antenna, compared to the chemically fixed samples. Scale bars: 1 µm. (A3 and B3) In the enlarged views of the boxed regions, the microlamellae in the CCM-processed antenna appeared uniform in size and shape, unlike the chemically fixed ones which were distorted. Scale bars: 200 nm. (C-D) CCM enhanced the morphological preservation of aldehyde-perfused mouse brain. The initial dehydration in standard EM preparation took place on ice for 1 hr, but it occurred during freeze-substitution at −90 °C to −30 °C for over 5 days in CCM processing. (C1 and D1) The smoothness of membranes was improved by CCM processing. ST: synaptic terminal. Scale bars: 200 nm. Pixel resolution of TEM images (x,y): 1.92 nm. (C2 and D2) The preservation of nuclear envelope was improved by CCM processing. N: nucleus. Scale bars: 500 nm. Pixel resolution (x,y): 2.88 nm. (C3 and D3) In the enlarged views of the boxed regions, the nuclear envelope (NE; arrows) appeared smoother and the cytoplasmic density (asterisk) was increased with CCM processing. We note that the chromatin was more heavily stained in the CCM-processed specimen, likely due to the additional exposure to uranyl acetate during freeze-substitution. Scale bars: 100 nm. Pixel resolution (x,y): 1.14 nm.

https://doi.org/10.7554/eLife.35524.004
Figure 2—figure supplement 1
TEM images showed well-preserved ultrastructures in the CCM-processed HEK 293T cells.

In the enlarged view of the boxed region (bottom panel), smooth nuclear membranes were observed. Scale bars: 1 µm for the top panel, 500 nm for the bottom panel. Pixel resolution (x,y): 6.01 nm.

https://doi.org/10.7554/eLife.35524.005
Figure 2—video 1
A SBEM volume from a CryoChem-processed Drosophila antenna.

Scale bar: 500 nm. SBEM imaging parameters (only a small region of the SBEM volume acquired is shown in the video): Z step: 50 nm; Z dimension: 1200 sections; raster size: 12k × 9k; pixel size: 3.8 nm.

https://doi.org/10.7554/eLife.35524.006
Figure 3 with 1 supplement
CryoChem method enables DAB labeling by APEX2 in cryofixed tissues.

In the CCM-processed cultured cells and Drosophila antennae, DAB labeling was observed in cells expressing APEX2. (A) Flowchart for DAB labeling of target structures expressing APEX2 in CCM-processed samples. In our experiments, the cultured cells were imaged with TEM and the Drosophila antennae were imaged with X-ray microscopy, followed by SBEM. (B) Mitochondria in the HEK 293T cell transfected with Mito-APEX2 were labeled with DAB (bottom panel), in contrast to the untransfected control cell (top panel). Scale bars: 200 nm. Pixel resolution (x,y): 3.97 nm (top panel); 2.88 nm (bottom panel). (C) An APEX2-expressing olfactory receptor neuron (ORN) was labeled with DAB (arrow) in the Drosophila antenna (10XUAS-myc-APEX2-Orco; Or47b-GAL4). Asterisks denote ORNs without APEX2 expression. Scale bar: 500 nm. Pixel resolution of SBEM images (x,y): 6.5 nm. (D) A series of SBEM images showing the same DAB labeled Drosophila ORN (arrow) in different planes of section. Asterisks denote ORNs without APEX2 expression. The images were acquired using standard imaging methods without charge compensation by nitrogen gas injection (Deerinck et al., 2018). These images, together with the rest of the EM volume acquired using SBEM, enabled semi-automatic segmentation and 3D reconstruction of the labeled ORN (right panel). Scale bars: 500 nm for SBEM images, 2 µm for the 3D model of ORN. SBEM imaging parameters: Z step: 50 nm; Z dimension: 1200 sections; raster size: 12k × 9k; pixel size: 3.8 nm.

https://doi.org/10.7554/eLife.35524.007
Figure 3—video 1
An X-ray micro-computed tomography volume from a CCM-processed Drosophila antenna showing DAB labeling in subsets of ORNs expressing APEX2 (10XUAS-mCD8GFP-APEX2; Or22a-GAL4).

The damaged region on the opposite side of the labeled cells indicates the hole poked in the antenna to facilitate solution exchange. Scale bar: 30 μm.

https://doi.org/10.7554/eLife.35524.008
Figure 4 with 1 supplement
GFP fluorescence is well-preserved in CryoChem-processed samples.

Confocal images were taken to quantify the level of GFP fluorescence in Drosophila ORNs. Antennae were collected from transgenic flies expressing GFP in a subset of ORNs. (A) GFP fluorescence intensity distributions of the ORN soma in the freshly-dissected, unfixed antennae (left panel) and the CCM-processed antennae (right panel) are not significantly different. p=0.810, Kolmogorov-Smirnov test. Insets show representative images, with ORN soma outlined. Scale bar: 2 µm. (B) Comparison of the average fluorescence intensities. GFP intensities are virtually identical between neurons in the unfixed antennae and the frozen-rehydrated antennae. n = 3 antennae, Error bars denote SEM, p=0.950, Mann-Whitney U Test.

https://doi.org/10.7554/eLife.35524.009
Figure 4—source data 1

GPF fluorescence intensities in unfixed and CryoChem-processed Drosophila ORN soma.

https://doi.org/10.7554/eLife.35524.010
Figure 4—figure supplement 1
GFP and tdTomato fluorescence in a cryofixed-rehydrated mouse brain.

(A) GFP- and (B) tdTomato-positive neurons (C) Co-expression of GFP and tdTomato fluorescence was detected in the cryofixed-rehydrated mouse brain. Scale bar: 50 µm.

https://doi.org/10.7554/eLife.35524.011
Figure 5 with 2 supplements
3D correlated light and electron microscopy (CLEM) in CCM-processed mouse brain.

Mouse brain slices with fluorescently labeled neurons were processed with CCM, imaged with confocal microscopy, X-ray microscopy and SBEM for 3D CLEM. (A) Flowchart for performing 3D CLEM with CCM-processed samples. Similar to a typical CCM protocol, cryofixed samples are first freeze-substituted and rehydrated. The frozen-rehydrated sample is then stained with DRAQ5 to label DNA in the nuclei. Next, the region of interest (ROI) is identified using confocal microscopy based on fluorescent signals, while the DRAQ5 signals are also acquired to serve as fiducial markers. Subsequently, the sample is stained, dehydrated and embedded for X-ray microscopy and SBEM. Using the DRAQ5 signals as fiducial markers, the confocal volumes can be registered to the X-ray volume such that the ROI for SBEM can be identified. Once the SBEM volume is acquired, it can be registered to the confocal volumes based on the positions of the nuclei for 3D CLEM. (B) An example of the DRAQ5 fluorescence signals (left), the corresponding ROI in X-ray volume (middle) and the overlay (right). This image registration process facilitates ROI identification in SBEM. Scale bar: 20 µm. (C) DRAQ5 fluorescence labeling served as fiducial points for registering the confocal volume to the SBEM volume. Scale bar: 5 µm. (D) The cell body of a tdTomato-expressing neuron (left) was identified in the SBEM volume (middle) through CLEM (right). (E) Neuronal processes expressing tdTomato (left) were also identified in the SBEM volume (middle) through CLEM (right). Scale bars: 2 µm, for both (D) and (E). SBEM imaging parameters: Z step: 70 nm; Z dimension: 695 sections; raster size: 10k × 15k; pixel size: 6.8 nm.

https://doi.org/10.7554/eLife.35524.012
Figure 5—figure supplement 1
A correlation of subcellular structure can be achieved using 3D CLEM performed on the CCM-processed mouse brain.

A heterochromatin structure (arrow) labeled by DRAQ5 in fluorescence microscopy was identified in the SBEM volume through CLEM. This heterochromatin structure was not used as a fiducial marker for image registration. Scale bar: 1 µm.

https://doi.org/10.7554/eLife.35524.013
Figure 5—video 1
A volume showing 3D CLEM in the CCM-processed mouse brain.

tdTomato-expressing neurons were clearly identified in the SBEM volume. Scale bar: 10 µm.

https://doi.org/10.7554/eLife.35524.014

Tables

Table 1
Comparison of the advantages and limitations of different sample preparation methods for electron microscopy.

The CryoChem Method (CCM) combines the advantages of chemical fixation and cryofixation. With CCM, samples are fixed with high-pressure freezing and freeze-substitution to achieve quality ultrastructural preservation. This approach allows preservation of tissues with cuticle or cell wall and captures biological events with high temporal resolution. A rehydration step is introduced to enable fluorescence imaging, DAB labeling by genetically encoded EM tags and high-contrast en bloc heavy metal staining of cryofixed samples. The high-contrast en bloc heavy metal staining permitted by CCM reduces the need for post-staining on sections, and makes CCM compatible with serial block-face scanning electron microscopy (SBEM). Common limitations of chemical fixation and cryofixation are denoted in red.

https://doi.org/10.7554/eLife.35524.002
Chemical fixationCryofixationCryoChem
FixationAldehyde fixatives
(4 °C)
(1) High-pressure freezing
(−196 °C,~2100 bar)
(2) Freeze-substitution in organic solvents
(1) High-pressure freezing
(−196 °C,~2100 bar)
(2) Freeze-substitution in organic solvents
Ultrastructural preservationFairExcellentExcellent
Tissues with cuticle or cell wallIncompatibleCompatibleCompatible
Temporal resolution of events capturedLowHighHigh
Hydration state of the sampleHydratedDehydrated after freeze-substitutionHydrated after rehydration
Fluorescence imaging after fixationCompatibleGenerally incompatibleCompatible
DAB labeling by genetic EM tagsCompatibleIncompatible due to dehydrationCompatible due to rehydration
High-contrast en bloc heavy metal stainingCompatibleLimitedCompatible due to rehydration
Post-staining on sectionsOptionalOften requiredOptional
SBEM compatibilityCompatibleIncompatibleCompatible
Key resources table
Reagent type (species)
or resource
DesignationSource or referenceIdentifiersAdditional
information
genetic reagent
(Drosophila melanogaster)
Or47b-GAL4(Fishilevich and Vosshall, 2005)RRID:BDSC_9984
genetic reagent
(D. melanogaster)
Or22a-GAL4(Dobritsa et al., 2003)RRID:BDSC_9951
genetic reagent
(D. melanogaster)
10XUAS-myc-APEX2-Orcothis studysee Materials and methods
genetic reagent
(D. melanogaster)
10XUAS-mCD8GFP-APEX2this studysee Materials and methods
genetic reagent
(Mus musculus)
B6.Cg-Crhtm1(cre)Zjh/JJackson Laboratory031559
RRID:IMSR_JAX:031559
genetic reagent
(M. musculus)
B6.Cg-Gt.ROSA.26
Sortm14(CAG-tdTomato)Hze/J
Jackson Laboratory007914
RRID:IMSR_JAX:007914
genetic reagent
(M. musculus)
TH-GFP(Kessler et al., 2003)
cell lineHEK 293TATCCCRL-3216
RRID:CVCL_0063
antibodyDRAQ5Cell Signaling Technology40841:1000
recombinant DNA
reagent (plasmids)
pcDNA3-Mito-V5-APEX2(Lam et al., 2015)Addgene_72480
recombinant DNA
reagent (plasmids)
APEX2 DNA(Lam et al., 2015)Addgene_49386
recombinant DNA
reagent (plasmids)
APEX2-Orcothis studysee Materials and methods
recombinant DNA
reagent (plasmids)
mCD8GFP-APEX2this studysee Materials and methods
chemical compound, drugparaformaldehydeFisher Scientific04042–500
chemical compound, drugglutaraldehydeTed Pella18426
chemical compound, drugsodium cacodylateTed Pella18851
chemical compound, drugCaCl2Sigma-Aldrich223506
chemical compound, drugglycineBio-Rad Laboratories161–0718
chemical compound, drugBSASigma-Aldrich9048-46-8
chemical compound, drug1-hexadeceneSigma-AldrichH2131
chemical compound, druguranyl acetateElectron Microscopy Sciences22400
chemical compound, drugmethanolFisher ScientificA412-4
chemical compound, drugacetoneACROS OrganicsAC326800010
chemical compound, drugHEPESGibco15-630-080
chemical compound, drugdiaminobenzidine; DABSigma-AldrichD5637
chemical compound, drugH2O2Fisher ScientificH325-100
chemical compound, drugosmium tetroxide; OsO4Electron Microscopy Sciences19190
chemical compound, drugpotassium ferrocyanideMallinckrodt6932
chemical compound, drugthiocarbohydrazideElectron Microscopy Sciences21900
chemical compound, drugDurcupan ACM resin component ASigma-Aldrich44611
chemical compound, drugDurcupan ACM resin component BSigma-Aldrich44612
chemical compound, drugDurcupan ACM resin component CSigma-Aldrich44613
chemical compound, drugDurcupan ACM resin component DSigma-Aldrich44614
chemical compound, drugconductive silver epoxyTed Pella16043
software, algorithmIMOD(Kremer et al., 1996)RRID:SCR_003297http://bio3d.colorado.edu/imod/
software, algorithmImageJNIHRRID:SCR_003070https://imagej.nih.gov/ij/
software, algorithmAmira 6.3ThermoFisherRRID:SCR_014305
otherAclarElectron Microscopy Sciences50426
otherFocusClearCedarlane LabsFC-101

Data availability

A source data file has been provided for Figure 4 (Figure 4-source data 1). The SBEM volume of a Drosophila antenna presented in this study has been deposited to the Cell Image Library. The SBEM volume, the tdTomato confocal volume and the DRAQ5 confocal volume used for 3D CLEM in a mouse brain (corresponding to Figure 5) have also been deposited to the Cell Image Library. The video of 3D CLEM in a mouse brain expressing tdTomato that corresponds to Figure 5-video supplement 1 has been deposited to the Cell Image Library (http://cellimagelibrary.org/groups/50451).

The following data sets were generated
  1. 1
    SBEM of Drosophila antenna (10XUAS-myc-APEX2-Orco; Or47b-GAL4)
    1. Tsang TK
    2. Bushong EA
    3. Boassa D
    4. Hu J
    5. Romoli B
    6. Phan S
    7. Dulcis D
    8. Su C-Y
    9. Ellisman MH
    (2018)
    Publicly available at the Cell Image Library under accession number CIL:50452.
  2. 2
    SBEM volume used for 3D CLEM in mouse brain
    1. Tsang TK
    2. Bushong EA
    3. Boassa D
    4. Hu J
    5. Romoli B
    6. Phan S
    7. Dulcis D
    8. Su C-Y
    9. Ellisman MH
    (2018)
    Publicly available at the Cell Image Library under accession number CIL:50451.
  3. 3
    tdTomato confocal volume used for 3D CLEM in a mouse brain
    1. Tsang TK
    2. Bushong EA
    3. Boassa D
    4. Hu J
    5. Romoli B
    6. Phan S
    7. Dulcis D
    8. Su C-Y
    9. Ellisman MH
    (2018)
    Publicly available at the Cell Image Library under accession number CIL:50453.
  4. 4
    DRAQ5 confocal volume used for 3D CLEM in a mouse brain
    1. Tsang TK
    2. Bushong EA
    3. Boassa D
    4. Hu J
    5. Romoli B
    6. Phan S
    7. Dulcis D
    8. Su C-Y
    9. Ellisman MH
    (2018)
    Publicly available at the Cell Image Library under accession number CIL:50454.
  5. 5
    Video of 3D CLEM in a mouse brain expressing tdTomato
    1. Tsang TK
    2. Bushong EA
    3. Boassa D
    4. Hu J
    5. Romoli B
    6. Phan S
    7. Dulcis D
    8. Su C-Y
    9. Ellisman MH
    (2018)
    Publicly available at the Cell Image Library under accession number CIL:50401.

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)