How to Choose Your HCS Reader?
Resolution, Throughput and Ease-of-use
High-content screening (HCS) is celebrating 10 years of existence. The technology has proven its value for the drug discovery process and is continuously increasing its role. In parallel, academia has also recently discovered the potential of HCS. The entry into the technology is associated with a high investment in terms of cost, knowledge and human resources. This article aims at highlighting some points that should be considered when purchasing a HCS reader.
HCS: A Growing Market
HCS is an image-based screening methodology applying automated microscopes and automated image analysis. Analysis of digital fluorescence images quantifies intensity, shape and spatial relationships of objects. HCS was first adopted by the drug development industry as it allows assaying drugs in the physiologically relevant environment of a cell using multiple parameters. Multivariate analysis of individual cells in a population has been shown to be very powerful and more efficient compared to univariate assays [1, 2]. As the technology is image-based, events such as translocation or morphology can be assayed, allowing novel types of assays to be screened . Since the creation of various academic drug screening programs and the advent of systems biology, high-content screening is also applied in academia [4, 5]. Thus the market of HCS is continuously growing.
HCS is multi-disciplinary, requiring expertise in cell biology, microscopy, data management, computer vision, statistics and automation. All these fields require investment in hardware, software and human resources and the initial cost for a minimal HCS setup can run up to € 500.000 with yearly running costs of € 150.000. In the following, I will focus on guidelines for purchasing HCS readers, with particular mention of three platforms currently in use in our laboratory: Thermofischer Arrayscan VTI, BDPathway 855 and Perkin Elmer Opera.
Points to Consider When Acquiring an HCS Reader
The main points to consider when purchasing an HCS reader are: resolution, throughput and ease-of-use. It is paramount to test several readers and to talk to experienced users.
The test period should be carefully prepared and a series of conclusive experiments planned ahead. The readers should be tested over a period of several weeks to get a good feel for the reader.
Abbe's law describes the optical resolution of a microscope (table 1). The numerical aperture of the objective is crucial and the highest possible numerical aperture should be chosen to obtain the best resolution. Confocality does not increase the resolution much, but mainly improves contrast compared to widefield systems. For assays measuring morphological changes of cells, translocation events or reporter gene induction, widefield microscopes are adequate. For assays measuring sub-cellular objects, colocalization or objects with high background intensities, confocal systems are better suited than wide field systems due to their superior contrast (fig. 1).
Many factors influence the throughput of an HCS reader, but the two most crucial factors are parallel acquisition and auto-focus mode. Readers with several cameras allow acquisition of multiple channels simultaneously, speeding up the acquisition process dramatically. Laser-based auto-focus is fast and is preferable to the much slower image-based auto-focus. To determine the speed of the HCS reader, several plates need to be imaged using different magnifications and acquiring several images in several channels in each well.
Every HCS reader is equipped with software to control the microscope and perform image analysis. The most crucial and difficult part of the software deals with image analysis and it should be carefully evaluated. Some HCS image analysis software are designed as turnkey solutions requiring little training. These systems are designed to analyze typical HCS assays such as target activation, neurite outgrowth, translocation events etc. Parameters for the various assays are adjusted on a few test images and the software is then ready to analyze a screen. Other software do not have ready-made assays, but provide image analysis algorithms to develop pipelines to analyze assays. Typically, these software perform better but require more knowledge and training.
The Thermofischer ArrayscanVTI is a very successful widefield HCS reader of medium throughput. The resolution and sensitivity are adequate for most simple cellular assays. The software (BioApplication) is user-friendly and provides many solutions for standard assays. The image analysis solutions can be adapted, when performing a screen for which no turnkey solution is provided. The ArrayscanVTI can be equipped with a live cell chamber and injection arm. In summary, the ArrayscanVTI is an instrument requiring little experience that can enable a laboratory to quickly perform its first HCS screen.
The BDPathway 855 is both a widefield and confocal system. A Nipkow disk can be inserted into the light path to obtain confocality. The BDPathway 855 has many illumination settings with two mercury lamps and two filter wheels providing 16 color combinations. The objectives are air objectives of excellent quality and allow high-resolution imaging. Both laser-based and image-based auto-focus are available. The software (Attovision) controlling the reader and performing image analysis has some unfortunate features that strongly limit its usefulness. There are only very few turnkey image analysis solutions and generally a pipeline has to be built. The major problem of the software is that only one image per well can be analyzed which is insufficient to obtain reliable statistics and completely unusable for live cells. If choosing the BDPathway to profit from its flexibility and excellent optical properties, we recommend acquiring third party image analysis software. The BDPathway 855 is equipped with an incubation chamber and an injection system for live cell imaging.
The confocal Perkin Elmer Opera is the most sophisticated instrument on the market currently. It is equipped with a Yokogawa spinning disk, ensuring minimal loss of light. The resolution is very high due to excellent water immersion objectives. The cameras are very sensitive of which four can be mounted, allowing the simultaneous acquisition of multiple channels dramatically increasing the throughput. The Opera is equipped with laser auto-focus ensuring fast acquisition. The Acapella controlling and analysis software is complex and requires some training. Often a dedicated person to script analysis programs is necessary.
 Young D.W. et al.: Nat Chem Biol 4, 59-68 (2008)
 Low J. et al.: Mol Cancer Ther 7, 2455-2463 (2008)
 Taylor D.L. and Giuliano K.A.: Drug Discov Today: Technologies 2, 149-154 (2005)
 Abraham V.C. et al.:Trends Biotechnol 22, 15 (2004)
 Simpson J.C. et al.: Journal of Biotechnology 129, 352 (2007)