Cells transfected with baculovirus-expressed fluorescent markers allow researchers to track microtubules and DNA during mitosis. (All images: Life Technologies)Technology advances allow researchers to gain more information from live cells.

The old pharmaceutical mantra to “fail early” hasn’t been working out very well recently. Bristol Myers’s BMS-094 hepatitis C drug took a $1.7 billion Phase 3 dive due to toxicity, while monoclonal antibodies against Alzheimer’s disease plaques from Pfizer, Johnson & Johnson, Eli Lilly, and Medivation have all met with costly late-stage failures of their own. And that’s just the most recent news.
Besides re-evaluating the design of some of their clinical trials, companies are taking a sharp eye to every part of the developmental pipeline, including cellular assays for the earliest phases of drug discovery. Traditionally, early-stage screening meant measuring a few obvious endpoints, such as proliferation and apoptosis, using biochemical tests or fixed cells. Aided by advances in basic science, such as induced pluripotent stem cells, new microscopy techniques, and sophisticated data analysis software, researchers can now glean vast amounts of information from live cells. As more companies start applying these technologies to their early-stage screening programs and more vendors build equipment for it, they’re discovering firsthand how powerful—and challenging—live-cell assays can be.
Game on
Joel Schwartz, principal scientist for biology at Pfizer (New York), likens the evolution of drug screening assays to following a football game: biochemical assays give only the final score, while fixed-cell experiments are more like a box score that also tells how individual players performed. “Live-cell imaging is this continued evolution into the idea that you watch the game,” says Schwartz.
A microscope remains the central piece of equipment for this spectator sport, while the critical advances in live-cell imaging involve ancillary tools. “The actual insides of a microscope, the inner guts, have not really changed in about a hundred years,” says Schwartz, adding that “the big jumps have come in the use of additional technologies adjacent to the microscope.”
Because animal cells are largely transparent, advances such as confocal microscopy, phase contrast, and fluorescent markers are essential for live-cell assays. More recently, super-resolution microscopy has also allowed researchers to probe smaller structures, while endomicroscopy lets them view tissues inside living animals in unprecedented detail.
Each of these new techniques carries a price measured in both dollars and time. In general, greater resolution comes at the expense of slower throughput, so Schwartz argues that live-cell imaging will augment the relatively fast and cheap fixed-cell assays rather than replace them completely. 
In addition to new imaging methods, drug developers are exploiting new types of cultures, such as induced pluripotent stem cells that can form interconnected neurons and other tissues. That could be especially useful for developing neurological drugs, where existing preclinical assays are spotty. “Neurological problems are inherently a problem of connectivity. If you cannot measure the function of the circuit in its entirety, it’s really difficult to extrapolate useful biological information,” says Schwartz. 
Living on the stage
Watching live, networked neurons and other types of cells respond to your drug can be immensely useful, of course, but it also raises challenges that range from esoteric to mundane. One fundamental problem is keeping the cells alive as they spend several hours on a microscope stage.
Standard cell culture medium isn’t compatible with many types of microscopy, as the phenol red dye in it interferes with imaging. “Switching to a physiological salt solution is often ideal for that, however there are various issues of pH control with those buffers,” explains Nick Dolman, senior staff scientist at Life Technologies in Carlsbad, Calif. To address that problem, Life Technologies recently introduced a new buffer specially formulated for live-cell imaging, allowing researchers to maintain a constant pH for their cells during microscopic monitoring.
For long experiments cells also need to be kept warm, so equipment makers are now combining microscopes with cell culture incubators. Nikon, for example, offers the Biostation IM, which incorporates a microscope and incubator as well as a robotic sample handler and a built-in digital camera. Besides keeping the cells alive, this type of system can speed up some experiments. “It takes significantly less time for the software of the Biostation to identify stem cell colonies, so ... you can select them and then plate those out to grow stem cell colonies within the Biostation about twice as fast as you can by the human eye,” says Steve Ross, general manager of the product and marketing department at Nikon USA in Melville, N.Y. That type of acceleration is especially important for drug screening, where investigators need large numbers of stem cell colonies to test multiple compounds.
Observing without interferingThe Premo FUCCI cell cycle sensor assay follows a live cell’s progress through the cell cycle.

As more researchers embrace live-cell imaging, equipment makers are adding more features to their workstations. Systems such as PerkinElmer’s Ultraview 3-D are representative of this trend. “Every single feature on that system has been carefully designed to provide the best imaging capability for live-cell imaging. We can capture the cell as it develops and then follow the fate of that cell in 4, 5, 6 dimensions if you want to,” says Jacob Tesdorpf, director of high-content instruments and applications at PerkinElmer. In addition to three spatial dimensions, the Ultraview can track multiple channels of fluorescent markers to follow different cellular processes.
Though the Ultraview is designed for basic researchers, Tesdorpf says several pharmaceutical companies have purchased it as well, underscoring the industry’s growing interest in using live-cell imaging to discover new drug targets and fundamental disease mechanisms. 
The ability to track multiple fluorescent markers in three dimensions is powerful, but also problematic. Traditional fluorescence microscopy light sources and dyes can interfere with a cell’s normal biology over the course of a long experiment. In response, many companies have introduced more sensitive imaging systems that allow researchers to use lower light levels, as well as less disruptive dyes. 
Longer experiments also call for automating procedures such as focusing. “That’s something that has gotten much better in the past couple of years, with dynamic focusing systems like our Perfect Focus system, which actually tracks the coverslip using a separate optical path to hold focus over long periods of time,” says Nikon’s Ross.
The meaning of life
Once investigators have addressed all of the technical challenges of a live-cell imaging experiment, one final hurdle looms large: analyzing the data. Time-lapse and other live-imaging techniques often leave researchers buried in information but starving for insight. Equipment vendors are keenly aware of the problem, and most provide data analysis software along with their live-cell imaging systems. PerkinElmer, for example, offers the Columbus software package, which stores a research program’s entire trove of imaging information on a central server. Scientists can then access it remotely and use the server’s computing power to perform sophisticated analyses.
Even with central data storage, though, the analysis itself can be daunting. Some experts in the field hope new machine-learning algorithms will help. “Automated image acquisition and analysis is probably going to be the biggest thing, because then we can send datasets into queues to be analyzed overnight,” says Dolman, adding “whilst it’ll still take the same amount of time, at least it’s hands-off time.”
Despite the challenges, drug developers are keen to do more live-cell imaging, and many hope that it will help improve companies’ success rates in the clinic. As Regis Doyonnas, senior principal scientist at Pfizer, explains, “We really see that as something that’s going to be very important for even the screening part, and adding a lot of impact all along the pipeline for compound development.”
About the author
Originally trained as a virologist, Alan Dove is now a science journalist whose work appears regularly in a variety of trade and scientific journals and online publications. He also co-hosts the popular podcast “This Week in Virology.”