Finding Perfect Protein Crystals
Protein crystallography is an important tool for investigation of biological molecules. Protein molecules can crystallize under certain conditions, forming regular lattices composed of multiple copies of the same molecule. When such a crystal is irradiated by a photon beam (x-ray), the photons scatter from the atoms and concentrate in sharp intense spots.
The molecular structure can be determined by analysis of the intensities and positions of the diffraction spots. Co-crystallization of interacting biomolecules or a biological molecule and a chemical drug helps elucidate the protein’s detailed function or find the inhibitor of this function.
Because of the complexity of such experiments, many structural studies require screening hundreds of samples until the crystal with the best resolution is identified.
Traditionally, the rate-limiting step of crystallography studies was a manual sample-loading step. Precise manipulations are required to extract the protein crystal from a liquid nitrogen storage container with special forceps and to place it in the exact position where the crystal can interact with the x-ray beam.
“Automated mounting was a key step in developing fully automated sample screening,” says Clyde Smith, Ph.D., senior staff scientist, Stanford Radiation Laboratory. “Without this manual step, remote data collection also became possible. Now our collaborators all over the world are able to visualize and collect the data stream via a computer interface.”
Stanford Automated Mounter is a robotic arm capable of locating the samples under liquid nitrogen, extracting the sample from storage cassette, and precisely transferring the crystal to the goniometer, a platform that rotates the crystal exposing it to the x-ray beam from various angles.
The robot uses a set of coordinates to find the sample location within the liquid nitrogen dewar. It is also equipped with fast response sensors, preventing it from bumping into the hardware. Periodic calibration ensures precision operation with no sample loss.
“Structure determination of RNA polymerase, a large and complex enzyme, was made possible only by automation of the crystallography process,” says Dr. Smith.
“The team had to screen several hundred samples to find a single crystal with required resolution. Manual screening on this scale is simply not practical.” To date, over 300,000 samples have been screened using Stanford Radiation Laboratory’s automated process.
Faster Stem Cell Therapy Development
Discovery of compounds affecting growth and differentiation of pluripotent stem cell is an active area of research. But the process of adaptation of stem cell cultures to the high-throughput format required for drug discovery has been slow.
Stem cells are difficult to grow, and the cultures have a tendency to spontaneously differentiate, or they do not differentiate consistently. To induce differentiation, embryonic stem cells are typically grown in hanging drops suspended from the lids of Petri dishes. In these drops, ES cells form embryoid bodies that need to be carefully collected and transferred into tissue culture plates for adherence.
“We started with a clear unmet need: to translate this manual process into an HTS format,” comments Michael Kowalski, Ph.D., senior applications scientist at Beckman Coulter.
“Moreover, we wanted to automate the optimization of the differentiation process itself. This means testing numerous combinations of differentiation factors at different concentrations.”
The BioRAPTR FRD noncontact dispenser became the key component of the solution. The device is able to dispense specific volumes of reagents in each well of a 384-well plate individually. Polypropylene plates were chosen because of low cell adherence, resulting in over 99% of wells forming a single embryoid body.
A Biomek Laboratory Automation Workstation transferred the embryoid bodies into the gelatin-coated plates. Again using the BioRAPTR, the team was able to test growth factors in hundreds of conditions on the same plate with the goal of identifying the set optimal for cardiomyocyte differentiation.
The “procardiomyocytic” compounds were selected from the literature. The statistical software created combinations of these factors resulting in an easily visualizable matrix. Automated Assay Optimization for BioRAPTR software converted the matrix into the dispense volumes. The BioRAPTR dispenser completed 914 individual pipetting steps in 7 minutes, Dr. Kowalski reports.
“We started with 6 percent differentiated cardiomyocytes,” he continues. “After completing the optimization schema, we found a set of conditions that can reproducibly generate over 40 percent cardiomyocytes on the plate. Automation provided a remarkable consistency for this highly complex biological process.” The team is currently working on optimizing differentiation of pluripotent human cells.
Post-translational modifications of chromatin facilitate or suppress gene expression. Changes in gene transcription due to dysregulations of these modifications have been linked to certain disease pathologies, primarily to various types of cancers. Consequently, enzymes that promote post-translational modifications have become important targets of drug discovery efforts.