Robotic Optimization of Specimen Preparation Protocol for Astrocytes Seeded on Coverslips for Imaging by Transmission Electron Microscopy (TEM)

Scientists at the UW-Madison routinely differentiate human pluripotent stem cells (hPSCs) into astrocytes to study their physiology. These cells are often seeded onto glass coverslips in 24-well polystyrene culture plates [1]. To optimize visualization of ultrastructural details of these cells when chemically fixed for imaging by Transmission Electron Microscopy (TEM), we asked two questions: Are there detectable differences between cacodylate or phosphate buffer used during fixation, and does post-fixation in osmium tetroxide or osmium tetroxide reduced with potassium ferrocyanide affect contrast in terms of membrane preservation? Reliable and consistent fixation and specimen preparation of astrocytes cultured on glass coverslips is vital for generating reproducible and timely experimental results. Traditional methods often involve tediously moving liquid reagents to and from cumbersome containers to perform numerous manual liquid reagent exchanges. Glass coverslips can be damaged with excessive handling. Improved sample preparation protocols are universally desired however time and resource limitations often force TEM laboratory technicians to use inherited methods rather than perform head-to-head comparisons of potentially better performing reagents. The ability to simply and easily prepare multiple identical specimens to test various fixation and staining protocols is now possible using robotic processing, whereas it would be very difficult to compare these variables using the traditional manual technique. A laboratory robot was used to compare various conditions used to prepare cells cultured on glass coverslips in the same polystyrene plate in which they were cultured. After initial fixation, the 24 well polystyrene plate was placed on the robot’s deck. The remainder of the protocol (see Table 1 below) was fully automated using precise robotic pipetting until resin embedding. Fully automated robotic pipetting ensured that all reagents were delivered to their respective sample wells which, in turn, ensured highly reliable and reproducible results. After robotic rinsing in their respective buffers, the robot delivered aliquots of osmium tetroxide for post-fixation; two samples for each of the two buffer groups were treated with 1% osmium tetroxide in its respective buffer (PB or cacodylate according to the initial fixation schedule) reduced with 1% potassium ferrocyanide and other two samples from each buffer group were treated with 1% osmium tetroxide without potassium ferrocyanide reduction (See Table 2 below). All sample groups were postfixed for 30 minutes. No differences in overall image quality were detected between cacodylate buffer and phosphate buffer when combined with either experimental osmium conditions. Clear differences were observed, especially in membrane preservation, overall contrast, and the ability to discern cytoskeletal elements, (i.e., microtubules and neurofilaments) in the samples post-fixed with potassium ferrocyanide reduced osmium as compared with unreduced osmium. The use of phosphate buffer vs. cacodylate buffer did not appear to have any significant effect on image quality, membrane preservation, contrast, or ultrastructural details. The laboratory robot enabled us to easily test 4 variables in one simple experiment using computer-controlled precise automated liquid handling with zero possibility of human error induced from trying to manually distribute a complicated array of reagents to each individual sample properly. This revolutionary increase in performance is due to the benefit of a programmable lab robot having zero possibility of decision-making errors, plus liquid handling accuracy and precision unattainable by a human technician. Robotic protocol Deck of the Prepmaster 5100 specimen preparation robot. Note the darker color of the reduced osmium in the reservoir plate in the lower left corner. Location of reagents in processing plate Phosphate buffer/unreduced osmium tetroxide produces identifiable Golgi and centrosome. Phosphate buffer/reduced osmium tetroxide gives low cytoplasmic background & well-defined centrosome. Cacodylate buffer/unreduced osmium tetroxide shows lack of clear and easily discernable membranes. Cacodylate buffer/reduced osmium tetroxide shows beautifully preserved Golgi apparatus membranes, low cytoplasmic background, and easily visible microtubules.