Computational protocol: Acoustic transfer of protein crystals from agarose pedestals to micromeshes for high-throughput screening

Similar protocols

Protocol publication

[…] A 2% agarose solution was heated (100°C for ∼10 min) in a water bath until it reached a random-coil state (a polymer conformation where monomer subunits are randomly oriented but are still bound to adjacent subunits). The agarose solution was then cooled to 70°C and mixed in a 1:1 ratio with the following mother-liquor solutions: 0.2 M sodium acetate, 8% NaCl for lysozyme, 0.05 M NaOH, 15% ammonium sulfate for thermolysin, 10% glycerol, 10% PEG 3350, 25 mM hexammine cobalt chloride, 100 mM HEPES pH 7.0 for stachydrine demethylase and 40% PEG 5000 for photosystem II.In order to achieve a concave basin, the wells must be over-filled with tacky agarose (when cooled to ∼70°C, agarose becomes somewhat adhesive) and mother-liquor solution so that the agarose adheres to the walls of the source well, resulting in a bowl-shaped surface when excess agarose is removed from the center of each well. The wells of a 384-well polypropylene source microplate were overfilled with 70 µl of the agarose and mother-liquor mixture using a pipette. After allowing 3 s for the agarose to adhere to the sides of the well, 40 µl were aspirated out of the well from the center. This created a concave basin in the agarose gel (Fig. 1). A custom-made positioning tool secured the pipette tip in the center of each well to ensure a symmetric bowl shape.Crystals of lysozyme (50 mg ml−1), thermolysin (50 mg ml−1) and stachydrine demethylase (20 mg ml−1) were grown by standard hanging-drop protocols (4 µl of protein solution combined in a 1:1 ratio with mother-liquor solution over a 500 µl reservoir). The photosystem II crystals were donated. Crystals were manually pipetted from each hanging drop onto the agarose pedestal, where gravity led them to accumulate in the center. The plate was sealed with adhesive plastic. Using the Echo 550, the supernatant above the crystals was removed in 1 µl increments (by serial ejection onto the plastic adhesive that sealed the source plate; no pin platform box was present) until crystals were observed in the ejecta using a light microscope (supernatant removal). A 1 µl volume was chosen because the emergence of crystals from the CAP was observed to be gradual, so the number of crystals lost in the 1 µl supernatant-removal procedure was small compared with the total number in the well. The adhesive plastic was peeled off after the supernatant was removed. 50 nl of crystal suspension was then acoustically transferred from the CAP to each micromesh (Fig. 2). In cases where the crystal concentration was high (lysozyme and thermolysin), each micromesh contained an average of approximately five crystals. In cases where the crystal concentration was low (stachydrine demethylase and photosystem II), only mother liquor was ejected onto some of the micromeshes. If crystals were not observed on each micromesh (using a Leica microscope) then additional transfers were made.Each micromesh that contained crystals was cryocooled. When cryocooling many crystals on pin-mounted micromeshes, the entire pin platform was manually dropped into liquid nitrogen (see §3.3). When cryocooling only a few crystals on pin-mounted micromeshes, each crystal was individually cooled by hand. Diffraction data were collected on beamlines X12C and X29 at the National Synchrotron Light Source (NSLS). Data sets were processed with HKL-2000 (Otwinowski & Minor, 2001) and further processed using CTRUNCATE in the CCP4i suite (Winn et al., 2011). Structures were obtained by molecular substitution from published models and were refined using REFMAC (Winn et al., 2003) and ARP/wARP (Perrakis et al., 2001) (starting models: lysozyme, PDB entry 1lyz; thermolysin, 4tln; stachydrine demethylase, 3vca; photosystem II, 1fe1; Diamond, 1974; Holmes & Matthews, 1981; Daughtry et al., 2012; Zouni et al., 2001). Each atomic model was further screened for binding to agarose (ZINC database 87496095) using AutoDock Vina (Trott & Olson, 2010), confirming that the tightest predicted binding pose for agarose monomers has zero electron density (we could not find any electron density for sugar molecules that might have originated from the agarose gel). […]

Pipeline specifications

Software tools HKL-2000, ARP/wARP
Application Protein structure analysis