February 1, 2007 (Vol. 27, No. 3)
Joan Zimmerman
Increasing Data Integrity and Reducing Volume and Raw Material Usage Are Critical
Robotics, automation, and fine-tuning of microplate assays have generally smoothened the way for high-throughput screening (HTS). Now the pressure points are building in the validation and assay design areas as well as in secondary screening, squeezing the upstream and downstream process. Some clever tweaking has become necessary to reduce false hits and shrink the error bars that can expand so rapidly when screening large numbers of compounds.
Accounting for growing knowledge of genome variability, delving deeper into cell cycle and behavior, separating screening components of individual assays, and simplifying nanoliter volume assay assembly and detection are helping to overcome these obstacles. The aim overall is to increase data integrity and reduce volume and raw material usage, thus increasing the percentage of true hits.
Mike Biros, company spokesperson for SpinX Technologies (www.spinx-technologies.com), reports that his company’s nanoliter assay platform assists in “compound profiling, downstream of HTS.” SpinX’ approach offers “homogenous assays in nanoliter volumes, performing dose-response with end-point or time-course readouts.”
Generally, companies have reduced the volume of HTS assays to a sweet spot that balances cost savings with technological complexity. After HTS, however, performing assays that vary component concentrations or read at multiple time-points is still a challenge at low volumes. The SpinX technology addresses this challenge by moving and combining nanoliter-volumes of assay components on gCard™, a plastic microfluidic device, using centrifugation and a unique valve mechanism, called VLV™.
Its advantage, in comparison to available platforms, continues Biros, lies in the fact that “SpinX technology allows robust assays in 200-nanoliter volumes without characterizing the physical properties of the assay components. For instance, there is no need to define liquid classes or calibrate based on viscosity. Furthermore, classic detection modalities can be used with this technology.” In addition, “because the gCards are being spun past the detector, kinetic data is inherently collected from hundreds of assay chambers simultaneously.”
The user interacts with a gStack™, “which looks a bit like a deep-well plate. Each gStack has 12 gCards with reservoirs that can be filled with different buffers and enzymes in stock form—no dilutions are necessary. The individual cards are automatically loaded onto a rotor, where the liquids are forced through microfluidic channels into chambers. The user can specify the amount of liquid that can go into the chambers and when that liquid is moved, enabling concentration and time-dependent assays,” says Biros.
“As the rotor spins, a laser perforates a hole between two adjacent microfluidic components, either channels or chambers. Centrifugal force moves the liquid to a new chamber. It’s a very robust valve mechanism with no moving parts or external mechanical or electrical contacts.” As an example of the system’s ability, 96 10-point IC50 curves can be run after loading a single 384-well gStack in a couple of hours, Biros claims.
Assisting Assay Developers
Nanostream’s (www.nanostream.com) LD System, a miniaturized parallel high-performance liquid chromatography system, is touted by CEO Steve O’Connor, Ph.D., as one means of overcoming assay development bottlenecks. Typically, says Dr. O’Connor, “in screening, one buys or makes an enzyme and develops the assay to fit a platform, often spending a lot of time in assay development. It takes two to six months to develop an assay for an HTS platform, and then the data is acquired in just a single read per well.
“The Nanostream LD System assists assay developers at the beginning of the process by allowing them to separate the components of the assay before it is read. The system directly measures product and substrate separately and then takes a ratio to determine the conversion of the enzyme. It lets you see rates of product formation as low as one percent. In addition, the LD System fits into the existing workflow,” helping to determine assay conditions quickly.
For secondary screening, the Nanostream LD System is useful for accurately and specifically identifying false positives. It is currently being used in the NIH Molecular Libraries Screening Centers Network. For focused library screening, “unlike a lot of other platforms that require a charge change, the LD System is not limited to ions. It can do lipid-based assays; deacetylase assays, for example. It is a complete solution,” points out Dr. O’Connor.
Cyanine Fluorescent Dye
At the GE Global Research Center (www.ge.com), researchers have developed phagocytosis and endocytosis assays around the use of a pH-sensitive cyanine fluorescent dye with built-in properties that facilitate new assay development and improvement. Anton Beletskii, Ph.D., cell biologist, says that the dye, CypHer5E, was developed to assay “uptake of nanoparticles, large particles, and bacterial cells by professional phagocyte cells, usually macrophages, that serve as a front line against infection.
The CypHer5E dye “is exactly matched to conditions inside the uptake vesicles, where the pH is close to 5.5, and is designed to reach its fluorescent peak at that pH. In traditional uptake assays, there are three fates for the particles—they remain in the media, they attach to the surface of the cell, or they are ingested,” Dr. Beletskii says. These conditions, when reported with traditional fluorescent dyes, are hard to distinguish. However, CypHer5E will fluoresce only when properly ingested, indicating that the molecule under investigation is indeed taken up into the cell.
Dr. Beletskii explains that “in some cases researchers have requested CypHer5E dye and protocols to determine how drugs affect phagocytosis during the screening and discovery phases. In addition, bacteria, such as those that cause tularemia and tuberculosis, often evade complete phagocytosis by preventing acidification of the uptake vesicle, and CypHer5E can be used to detect that.”
In the endocytosis applications, CypHer5E can be used to elucidate the proteins involved in cell signaling, to track the fate of molecules as they travel from surface receptor into the cell. As receptors are taken into endosomes, “most of the signaling goes through acid compartments, which can then be measured with CypHer5E. No matter what you label, once it participates in the endocytic pathway, it lights up.” Dr. Beletskii adds that because endocytic uptake occurs in modified low-density lipoprotein (LDL) processing, GE scientists have been able to use CypHer5E to “very efficiently study macrophage-mediated mechanisms of LDL uptake to model atherosclerotic plaque formation in a high-throughput setting before the use of animal models.”
Overall, CypHer5E labeling greatly reduces background fluorescence from particles or molecules that have not been internalized, in addition the dye is resistant to fluorescent quenching that can occur in acidic conditions of the uptake vesicles. The assays have been successfully demonstrated in human monocytes, dendritic, and endothelial cells, as well as mouse macrophages.
Genome-scale SNP Genotyping
Roche (www.roche.com) is employing human genetics approaches, such as SNP genotyping techniques, to address data before it enters the HTS pipeline. Mitchell Martin, Ph.D., research leader, research informatics, genetics and genomics, explains that addressing the growing insight into the variability of the human genome will improve clinical outcomes.
“Currently, only about 25% of compounds that have undergone Phase II testing are able to enter Phase III trials. Most of that attrition is due to lack of efficacy. The compound binds to the target but doesn’t work as intended. The key decision is which target to put into HTS. SNP genotyping can be used at the outset to actually measure (the contribution of a gene) to disease at the clinical level. Also, by genotyping patient populations from the beginning (of the HTS process), we can pick the most representative form of the gene for cloning and expression prior to HTS.”
Roche has been supporting human genetics in discovery and development for at least five years, says Dr. Martin. He adds that the methodology is “better at narrowing targets than expected. For example, we now have a series of compounds in the glucokinase activator family that represent a real breakthrough for diabetes treatment. This gene family was first identified in genetic studies of Maturity-Onset Diabetes of the Young and represents a promising area for successful intervention.”
In addition, genome-scale SNP genotyping, as practiced by Roche, can address another aspect of genomic data that can inform HTS. “In cell-based assays, many of the familiar cell lines can be monitored over time for chromosomal aberrations, deletions and duplications, for example, thus assuring stability of cell lines to ensure the integrity and comparability of screening results.
“We know that the notion of a static genome is a gross oversimplification. Cell lines that have been maintained for decades, such as HeLa and CHO cells, may not contain a faithful version of a gene represented in patients. This new technology can give us a window into how these cell lines behave.”
Dr. Martin believes that within the next few years, as data accumulate from basic studies in genomic variability, SNP genotyping will prove to be a more widely applied tool for candidate selection.
GPCR Signaling
Two additional techniques of fairly recent vintage are DiscoveRx’ (www.discoverx.com) pathway profiling and analysis of cell signaling and Euroscreen’s (www.euroscreen.com) culture-free functional assays on G-protein coupled receptor (GPCRs). DiscoveRx pioneered the use of enzyme-fragment complementation assays and recently added to its portfolio a “novel protein-protein interaction approach for generic analysis of GPCR signaling based on beta-arrestin recruitment to activated receptors.”
Euroscreen offers a wide variety of off-the-shelf and customizable receptor-based assays, and was issued a U.S. patent early in 2006 that “covers the use of an important human neurology drug target, a GPCR known as the human nociceptin (ORL1) receptor.”
