High-Content Analysis Comes Full Circle in Only Eight Years

In eight short years, high-content analysis (HCA) has circled back to its roots, where many practitioners are again using personal instruments. Sure, the large core facilities were created and became dominant during the preceding seven years, but now they are busy with million-well screens and spew forth gigantic files of data. The high growth part of the hundred-million-dollar global market for HCA systems and components is now in small laboratories working with smaller, more focused screens. This is precisely where HCA started. Many firms found that individual investigators, often several in a large facility, were independently creating screening systems. Most were spending too much time solving the same problem. A typical laboratory might have had a liquid handler coupled via a robot to a plate reader. Computers for instrument control and data analysis were tucked in the janitor’s closet alongside the mops and pails. Data storage and retrieval were not yet much of a problem, since just running the experiment was all-consuming and generally not that comparable to others. Today, individual investigators can buy highly engineered components that work together reliably when assembled into a system.

In a typical experiment, the liquid handlers prepare the sample, including cells, nutrients, and test compounds (including stains), in multiple-well plates (96, 384, or 1536 wells). Robots transfer the plates to an imager, which usually records fluorescence images. These images are then characterized by as many as 1000 measurements, each of which is tied to some facet of cell function. For example, cell volume, number and size of nuclei, cell morphology, etc., are all measured and archived. The data are then ready for use in the study, which often is related to dose and corresponding response plots.

The eighth annual High-Content Analysis Meeting (HCA 2011) was held at the historic Fairmont Hotel in San Francisco, CA, January 12–15, 2011. Four topics dominated: assays, imaging agents, personal imagers, and information technology. Proficiency in all four is required for success in HCA, even on the individual scale.

Personal imagers

Major pharmaceutical firms have begun to recognize the promise of the technology, but they also wanted to give it sufficient funding so that HCA could be successful, hence the creation of enterprise-wide core facilities. These facilities managed to create very sophisticated systems, including software, and implement HCA runs. The most commonly mentioned failing is cost per well, which is still often much too high ($0.25/well or more).

At HCA 2011, at least four vendors (Idea Bio-Medical [Rehovot, Israel], Thermo Fisher Scientific [Pittsburgh, PA], Vala Sciences [San Diego, CA], and Yokogawa Electric Corp. [Ishikawa, Japan]) introduced new instruments designed around the moderate-throughput needs of individual laboratories. Representatives of several companies confirmed that the focus today is on individual laboratories. The core laboratories made their major buy about four years ago.

Cellnsight, the new personal cell imager from Thermo Fisher Scientific, provides assay results similar to the company’s ArrayScan VTI HCS reader. Both use the same software and have similar optical performance. The Cellnsight can image and analyze a 96-well plate in less than 4 min. Four-color solid-state light sources provide vivid images with fine details. In contrast to some others, the Cellnsight is compatible with plate handling robots, SBS standard microplates, and slides.

Vala Sciences introduced the IC200 high-content screening imager, which uses novel chromatic aberration to correct for differences in focal plane for colors. This produces a rapid focus on objects in the Z direction and greatly reduces artifacts arising from variations in the bottom of plastic multiple-well plates. One study showed that this can affect the Z focus by as much as 3 μm. This adversely affects the coefficient of variation for the cytometric features. The IC200 can be fitted with an electrode assembly that adds kinetic imaging produced with electrical stimulation of cells such as cardiomyocytes. This is called the Kinetic Image Cytometer (KIC). Electrical stimulation modulates the calcium flux across the cell membrane. With the KIC, the flux is measured for individual cells from movies recorded at 30 Hz. Special software identifies and tracks the calcium transients for each cell in the view field. Vala expects that the KIC will be the instrument of choice for cardiotoxicity testing.

Three-dimensional imaging is new and potentially much more important in cell biology than in entertainment. Conventional cell imaging typically looked at a layer of cells, but cells often grow in three dimensions. Bellbrook Labs (Madison, WI) introduced the IUVO 3-D cell-based assay service, in which microchannels are filled with cells embedded in collagen. An aliquot of test reagent is added to the sample port, and a cocktail of antibodies and detection reagents is added to the input port. The microchannels are incubated for a predetermined time, usually 16 hr. A quick scan of the plate using a conventional whole-well plate imager is summarized in a conventional dose response plot. The company claims that when cells are given a proper 3-D scaffold, they undergo polarization and differentiation resembling in vivo structures. Differences have already been shown in cell migration and therapeutic efficacy between flat plastic and 3-D environments.

Kinetics is another topic that is gaining interest, as well it should. Many of the HCA experiments use a single time point, i.e., measure after 16 hr of incubation. This may be convenient, but others report that a good scan needs to include the kinetics as well, especially when comparing therapeutic activity to toxicity. The mechanisms are often different, with toxicity being much quicker than therapeutic response. However, making screens at several time points increases the run time and data file proportionally, and the data files are already large. In spite of this, interest in imaging live cells appears to be growing.