Serial block-face scanning electron microscopy provides an unprecedented view of three-dimensional tissue ultra structures, and can answer a host of questions previously impossible or exceedingly difficult to pursue. This technique allows the rapid and reproducible production of three-dimensional datasets with minimal tissue charging and artifacts. And importantly, it allows the automated capture of thousands of images daily.
Standard transmission electron microscopy provides excellent cellular ultra structure. However, these images lack three-dimensional context, and can be difficult to interpret. Serial block-face scanning electron microscopy provides that three-dimensional context, allowing the interpretation of complex processes.
While this protocol is reliable and reproducible, certain steps require precision and are critically important for the success of this methodology. This video will demonstrate these critical steps in detail. After fixing samples in 0.1 molar sodium cacodylate buffer, containing 2.5%glutaraldehyde, and two millimolar calcium chloride, samples should be cut into blocks no larger than two millimeters cubed, and washed with 0.1 molar sodium cacodylate buffer containing two millimolar calcium chloride.
The tissue is then sequentially stained with osmium ferrocyanide, thiocarbohydrazide, and osmium tetroxide. After osmium tetroxide incubation, treat the tissue with 1%aqueous uranyl acetate at four degrees Celsius overnight. The next morning, dissolve 0.066 grams of lead nitrate in 10 milliliters of 0.03 molar aspartic acid solution, and use one normal potassium hydroxide to adjust the pH of the Walton's lead aspartate solution to 5.5.
After warming the solution in the oven, wash the tissue five times for three minutes per wash in room temperature, double-distilled water, before transferring the tissue into a 60 degree Celsius warmed Walton's lead aspartate solution for 30 minutes at 60 degrees Celsius. At the end of the incubation, wash the tissue and dehydrate it in an ascending, ice-cold acetone series, followed by a 10-minute incubation in room temperature acetone. Then place the tissue in a 1:3 ratio of hard-mixed resin in acetone for four hours of infiltration on a rotating platform, followed by an eight-hour or overnight infiltration in a 1:1 ratio of resin to acetone solution on a rotating platform, then, an overnight infiltration in a 3:1 ratio of resin to acetone on the rotating platform.
The next morning, infiltrate the tissue in fresh 100%resin for one four-to eight-hour, one overnight, and one four-hour infiltration at room temperature on a rotating platform. The next morning, use a wooden stick to mix a small amount of resin with carbon black powder until the resin is saturated with the powder, but is still fluid and does not become grainy. The residence should resemble thick ink, and be able to slowly drip without visible clumps Place the tissue sample in a silicone rubber mold, and capture an image for later reference of the sample orientation within the resin block.
When the sample is in place, cover the tissue in the carbon black saturated resin at the tip of the silicone mold, with the label indicating the experimental and tissue details in the mold at the opposite end of the resin. Then, place the mold into a 65 degree Celsius oven at an incline for approximately one hour. When the carbon black infused resin has sufficiently cured, remove the mold from the oven, and beginning with an empty well, fill the remainder of the mold with clear resin, taking care that the label remains visible.
When the sample has been covered with clear resin, place the silicone mold back into the 65 degree Celsius oven, not on an incline, for 48 hours, then remove the mold. To prepare the block for imaging, place the resin-embedded specimen onto a microtome chuck with the tapered end sticking approximately five to six millimeters out of the chuck. Lock the specimen in place with the set screw, and place the specimen under a heat lamp.
After several minutes, place the chuck into a stereo microscope holder, and use a new, double-edged razor blade to make thin sections parallel to the block face until the tissue is visible, which will be less reflective and granular compared to the portions of the resin that are devoid of tissue. Secure an aluminum specimen pin in the trimming pin holder, and make several deep, crisscrossing scratches in the face of the pin to provide a larger surface area for the glue used to hold the specimen in place. Place the specimen under a heat lamp, and push the razor one to two millimeters straight down into the resin block before making a second, perpendicular cut of equal depth into the block, to trim away the excess resin from the tissue sample so that the block is approximately three millimeters in diameter by two to three millimeters in height.
After this initial trimming, warm the block under the heat lamp again, and use a new, double-edged razor blade to cut off the top of the resin block roughly one millimeter below the trimmed portion in a single, smooth cut. Place the trimming pin holder, containing the cut aluminum pin, in the stereomicroscope receptacle, and apply a thin layer of cyanoacrylate glue to the pin face such that it completely covers the pin without forming a visible meniscus. Use forceps to press the trimmed piece of tissue block onto the center of the specimen pin face for several seconds, and allow the glue to set for several minutes.
When the glue has thoroughly dried, locate the tissue on the raised portion of the resin block, and use a double-edged razor to trim the raised portion of the resin containing the tissue sample to an area no larger than one square millimeter. Slowly and carefully, remove as much excess resin as possible, leaving the block slightly longer in one dimension, and use a final metal file to angle the excess resin in the area outside of the raised portion containing the tissue sample down toward the edge of the pin. Remove resin particles and dust from the prepared sample before applying a thin coat of silver paint and gold sputtering to the entire sample block surface.
After coating, trim any excess silver paint from the block face surfaces, and place the mounted and trimmed block in the bulbous end of a custom-made transfer pipette tube with the appropriate label attached. Using SBF-SEM imaging, a network of elastin-free microfibril bundles have been identified within the adult mouse cornea. This network is organized in distinct layers, with the fibers closely associated with keratocytes, even lying within shallow invaginations on the keratocyte surface.
Application of this protocol has led to the discovery of a previously unknown population of central corneal nerves that fuse with basal epithelial cells at the stromal-epithelial border. SBF-SEM imaging of the central cornea reveals the presence of limbal vasculature, nerve bundles, and associated cells that can be manually segmented for 3D reconstruction. Some pitfalls of this imaging procedure include using a too-long pixel dwell time, which can cause subsequent images to be wavy and distorted, small discharges of electrons from the block face, leading to rapid contrast changes in lines, knife scratches on the block face due to a damaged knife or debris accumulation on the edge of the knife, artifacts from an extended electron beam focus on the block face while the sample is still in the imaging chamber, improper tissue fixation, leading to the separation of cellular structures and connective tissue, and image-skipping as a result of a large amount of charging within the tissue or resin block.
When cutting the tissue block, it is important to have a good understanding of where the tissue lies within the block, So as to not accidentally trim away any important sample regions. If higher resolution images of specific cellular structures or interactions are desired, the imaging can be halted, the block removed from the microscope, and sections cut on an ultramicrotome where they can be viewed in a transmission electron microscope Serial block-face scanning electron microscopy allows a wholly unique view of biological tissues. Using this protocol, we have been able to rapidly acquire quality datasets, and to expand the range of tissues we can explore.
