SAN FRANCISCO — Fujitsu Computer Systems' Biosciences group last week said that the Whitehead Institute for Biomedical Research will evaluate its CelInjector platform for automated cellular microinjection for use in high-throughput cell-based research.
The placement marks the first foray for Fujitsu's Biosciences group into the life science instrument market in the US since the tool debuted three months ago, Michael McManus, vice president of Fujitsu's biosciences group, told CBA News at the American Society for Cell Biology meeting, held here last week.
The company had previously only offered computational and informatics offerings for life sciences research in the US, although it had placed several CellInjector systems in Japan.
"Fujitsu has, like a lot of large companies, its own research labs that do a lot of advanced research," McManus said. "Often times in these companies, and also at Fujitsu, this research goes uncommercialized, but every so often something comes along that has some market potential. CellInjector was such a piece of research. Our group here in Boston and our corresponding group in Tokyo were asked to try and commercialize this for the US, European, and Japanese markets."
"There is no reliable way of transfecting any given cell. You tend to work with cells that are amenable to transfection."
Whitehead Institute researchers will use CellInjector for a variety of research projects, including peptide expression and trafficking in stem cells, assessment of gene expression controls, and the study of cell-signaling cascades, Fujitsu said in a statement. The platform will be housed at the Whitehead-MIT Bioimaging Center.
Last week, Paul Matsudaira, director of the Whitehead-MIT Bioimaging Center, provided more details about how his group would be using the CellInjector platform.
"We are trying to express 100 different genes involved in cell adhesion and cell motility involved in macrophage cells and other kinds of cells," Matsudaira said. Other researchers at Whitehead are testing the microinjector to inject other genes in other cells, he said, but declined to provide further details due to confidentiality.
The current version of CellInjector, which is set to be officially launched sometime next year to coincide with the beginning of Fujitsu's fiscal year, is capable of transfecting "floating" cells only, McManus said, although another version is in the works that will be able to microinject adherent cells.
CellInjector comprises a silicon chip on a perforated dish that sits on a moveable platform. According to McManus, suspended cells are placed in the dish, and a suction is applied. "There are 1,043 holes on that little silicon wafer, and 1,043 seats are therefore available for cells to occupy," McManus explained.
The cells fall into the holes, the excess cells are diluted away, and an automated injection arm — which consists of a CCD camera-enabled microscope — "goes through and rather rhythmically injects all of these cells in the holes on the silicon chip," McManus said. According to McManus, it takes about 15 minutes to inject all of the cells.
Depending on the cell type, transfection efficiency is on the order of 60 to 70 percent, McManus said, although researchers actually should take three numbers into account, he said: Capture efficiency, injection efficiency, and expression efficiency. Fujitsu presented a poster at the ASCB meeting that reported, for example, 75-percent capture efficiency, 68-percent capture efficiency, and 98-percent expression efficiency for K562 cells; and 67, 75, and 76 percent, respectively, for murine macrophages.
According to Matsudaira, the efficiency for macrophages — and other difficult-to-transfect cell types like stem cells — is particularly important.
"This issue of hard-to-transfect cells is really a stumbling block, and in our case, macrophages are very difficult to transfect," he said. "The efficiencies are typically in fractions of a percent, or a very low percent. There is no reliable way of transfecting any given cell. You tend to work with cells that are amenable to transfection."
A number of other methods exist for transfecting cells, with lipofection, viral vectors, electroporation, and laser-based manipulation being the most well-known. In fact, several companies have developed automated transfection platforms based on some of these techniques, such as Cellectricon with its newly released CellAxess platform; Cyntellect with its LEAP system; and Amaxa's Nucleofector.
However, each of these methods has its own problems — such as inefficient genetic transfer and possible change of cellular phenotype — which often makes manual injection the method of choice.
"The problem with lipofection is that it's just very hard to control how much material gets into a cell," Matsudaira said. "With electroporation it's the same sort of thing, whereas with direct microinjection you can control roughly the amounts that go in.
"The trade-off is that usually with microinjection, you cannot inject large numbers of cells," he added. "We have a manual microinjector — it's a technique that lots of labs use. But the Fujitsu technology appealed to us because it can do thousands of cells, and that gets us into the level of cells we need to manipulate in order to measure the effects of perturbation in a statistically reliable way."
Matsudaira's comments seem to indicate that the Whitehead is sold on the platform's capabilities, but it isn't officially sold yet. McManus characterized the placement as an "evaluation," and many of the planned projects will begin next year. The main value to Fujitsu now is the possibility of presentations and publications validating the efficacy of the platform as a high-throughput cellular transfection tool.
Although the platform is not yet officially for sale, McManus said it would sell one to someone if they were interested at this point. He said that Fujitsu has several users in Japan, and is negotiating another evaluation agreement with a West-coast US-based academic institution.
McManus declined to comment on the price range of CellInjector, but said it would "cost less than $500,000."
— Ben Butkus ([email protected])