Courtesy of The Automation Partnership

Automated tissue culture systems, like the SelecT automated mammalian cell culture system shown here, provide a level of plate-to-plate uniformity that can be difficult to achieve manually. And because the systems work 24/7, they can have assay-ready plates available early on a Monday morning.

Much has been written of how robots have been used to streamline drug development efforts. Robots never vary their routines, never tire, and never make mistakes. But they also require a steady stream of input material to work with, and when it comes to cultured cells, that can be a problem. Although a single assay may require only a few hundred or thousand cells, high-throughput screening (HTS) programs often involve hundreds of thousands or even millions of tests.

That volume of work strains resources to the point that even giant drug companies like AstraZeneca have trouble keeping...


One of the groups championing automated cell culture is The Automation Partnership (TAP), of Cambridge, UK. TAP led a design consortium of six top pharmaceutical companies to produce a fully automated cell culture system called SelecT, which "is focused on managing lots of different cell lines in parallel and can prepare plates of assay-ready cells," says Tim Ward, TAP's cell culture business unit director. The system can continuously maintain up to 182 T-flask cultures, either a single line or 182 different lines.

Based on a barcode-linked protocol, SelecT can grow, feed, harvest, count, and pass cells into new flask, readying up to 400 plates as needed. "Cells will be ready whenever you want, rather than according to the availability of the cell culture technicians," says Ward.

And those plates are sure to be of consistent quality. "When the robot does it [cell culture work], it takes time, but every movement is identical," says Siqi Lin, group leader for Cell-Based HTS at GlaxoSmithKline, King of Prussia, Pa. "When technicians do it, the first 10 flasks are fast, the next 10 are okay, but then because of fatigue, we definitely see a difference in the quality of the data we get from these cells."

But automation also provides a cost benefit, by reducing the number of necessary cell culture technicians. "Without SelecT, I needed three full-time people dedicated to cell culture," says Lin. "Now I need one."

Yet machines like SelecT are not quite ready to replace scientists altogether. There are some tasks the machines cannot do, like determining if cells are confluent. "That's one area we would like to see automated," says Lin. People are required to "feed" the machine with fresh media, flasks, and plates and to communicate needs to it. "You have to understand about growing those cells, when they need feeding and splitting," says Ward. "The system can't do that for you."


Automation is playing a role outside of Big Pharma, too. For instance, RTS has codeveloped a system to simplify the generation of immortalized cell lines – to provide a never-ending supply of DNA for genetic studies – for the Avon Longitudinal Study of Parents and Children (ALSPAC).

"The intention was to create a genetic resource for research on any condition which could be described as a common disease of complex causation, like heart disease or diabetes," says Richard Jones, head of biological samples for ALSPAC. With help from the Wellcome Foundation, ALSPAC, based at the University of Bristol, UK, will be able to create cell lines from all the families, eliminating the need to go back to 14,000 potential cohort members for more DNA.

The process of cell immortalization involves multiple steps, not all of which are amenable to automation. Researchers isolate peripheral blood lymphocytes and coculture them with Epstein-Barr virus in one well of a 24-well plate. Once established, the transforming cells are placed into the automated system, which controls feeding and expansion. The centrifugation and washing steps are not amenable to automation, says Jones, but "after that, the whole process of tending the cells as they transform and grow up was the ideal type of repetitive activity that could be automated."

Each plate is barcoded, associating that dish with a set of protocols and tasks. Once the plates are loaded into automated incubators, the system reads the barcodes and decides which protocol to run. A lift mechanism works its way up and down each stack of plates, shuttling them into the airflow cabinet containing a liquid-handling robot. The robot then removes and disposes of used media and delivers fresh media. "The whole system is really about delivering plates to a work station and removing them again in a reliable way," says Jones.



Courtesy of Chris Rockhold, Payload Systems

A 3D CAD model of the carousel assembly which supports the Cell Specimen Chambers (CSC) inside the MARS unit's cell culture unit. It is shown here configured for nine CSCs.

One field in which automated tissue culture work makes sense is astrobiology. Since putting a person into orbit is a tremendous expense, having a robot that can do the job would be a tremendous benefit. Now the Miniature Automated bioReactor System (MARS) unit, the terrestrial version of a cell culture unit (CCU) slated to fly on the International Space Station in 2007, may soon be available to the earth-bound scientific community, courtesy of Payload Systems, Cambridge, Mass.

Explaining the decision to develop MARS, Joe Parrish, president of Payload Systems, says, "I don't know any investigator who wouldn't rather run 10 experiments using one lab technician or graduate student than one. So the notion of providing a highly automated system is just as relevant for the terrestrial environment as it is for space." The unique contributions MARS makes, says Parrish, are "the combination of high levels of automation, monitoring, and control of the cell's environment and the modularity of the system."

The CCU can be used to grow a variety of cell types, including mammalian cells, yeast, tobacco cells, and Euglena. Although operating on a smaller scale than the vast pharma-based automation systems, the CCU accommodates specimens in up to 24 individual, environmentally controlled culture chambers, each with its own gas exchanger, medium supply, additives, sensors, and online sampling. The system can expand cell lines by extracting cells from one chamber and distributing them into others.

These controlled environments are especially useful for tissue engineers, for whom quality control is a critical issue. "When you engineer cardiac muscle, it just doesn't work in a dish," says Gordana Vunjak-Novakovic, a tissue engineer scientist at Harvard-Massachusetts Institute of Technology (MIT) Division of Health Sciences and Technology at MIT, Cambridge, Mass. High-level automation is required to engineer a piece of compact, functional, clinically thick cardiac tissue. Vunjak-Novakovic adds, "The level of sophistication increases, but your output is much greater than you can achieve without all the advances."

Spurred by interest from the scientific community, MARS units are ready to be tested by independent research institutions.


Although researchers have shown interest in automated systems, many remain skeptical. "Tissue culture is regarded as a rather magical process, especially cell-line transformation," says ALSPAC's Jones. "So those who have been doing this for many years have a strong conviction you can only do this sort of thing by hand." He adds, "We've been funded, but referees comments are cautious."

And despite the advances, there remain many aspects of cell culturing that are yet to go robotic. Examples include cryopreservation, banking and retrieval of frozen cell lines, and clonal expansion. Currently, cells expressing surface receptors or specific genes are sorted and grown up in 96-well microtiter plates, one cell per well. Automation in the arena of identifying, expanding, and monitoring clone expression would significantly speed an otherwise slow process.

Yet Jones suggests that enhanced quality control is one reason to consider automation, especially when working with large groups of study participants. People, he says, are not as good as robots at repetitive, tedious tasks. "With our system, people spend their time looking at cultures, making decisions, and doing quality control," says Jones. "The robots do what they're good at, and the people do what they're good at."

Nevertheless, the more industrialized approach toward research is not an easy sell to academicians who typically have strict protocols based on years of experience – not to mention tight budgets. But, Ward concedes, "we've been surprised by the level of interest in our system from academicians." He adds, "It could have value in an environment where enough people could pool their work who consistently use cells."

In the end, cell culture automation has the same value in a pharmaceutical company as it does in a research lab. It speeds up research. There isn't a researcher out there who would scoff at that.

Linda Schultz LBSchultz@sciwriting.com is a freelance science writer and author in Georgia.


The Automation Partnership SelecT, Cello, and Cellmate http://www.automationpartnership.com

Tecan Cellerity http://www.tecan.com

BioCrystal Robotic Cell Culture System http://www.biocrystal.com

Thermo Electron

CRS Automated Cell Culture System http://www.thermo.com

RTS Life Science acCellerator http://www.rts-group.com

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?