By Julia Boguslavsky
July 15, 2003 | Comprehensive proteomic analysis using conventional 2-D gel electrophoresis typically requires microgram quantities of protein — amounts that can prove hard to come by. After spending the past two decades developing ultrasensitive tools to study biological molecules, Norman J. Dovichi, professor of analytical chemistry at the University of Washington, has finally put single-cell proteomics within reach by successfully studying the expression of relatively abundant proteins in single cells.
"Our technology offers a dramatic improvement in sensitivity, which allows us to study protein expression in single cells," Dovichi says. "This sensitivity provides many exciting research opportunities in cell biology." Similar technology, based on coupling 2-D capillary electrophoresis separation with laser-induced fluorescence detection, has been applied to DNA sequencing.
To fully appreciate single-cell proteomics, consider that a typical mammalian cell is a mere 10 µm in diameter, has a volume of about 500 fL, and contains only about 50 pg (2 femtomoles) of total protein. "Our system can detect proteins present at a few thousand copies per cell," Dovichi says.
To study protein expression in single cells, Dovichi's team had to master the handling of individual cells and develop tools to label, separate, and detect the minute amounts of protein contained therein.
After analyzing the cellular suspension through standard microscopy techniques, the cell of interest is aspirated into a 50µm-diameter capillary by a brief pulse of vacuum. The cell is then lysed inside the capillary to release its complement of protein. This is where things get tricky. The labeled proteins are then separated inside the same capillary by their molecular weight through capillary sieving electrophoresis (CSE), which is the capillary version of conventional polyacrylamide gel electrophoresis. To separate the proteins further, successive fractions are transferred to a second capillary. Applying an electric field across this capillary separates proteins based on their interaction with the surfactant in a micellar electrophoretic separation. Repeating this process about 100 times under computer control comprehensively separates the protein content of the single cell.
The minute amounts of protein that result are detected by ultrasensitive laser-induced fluorescence. Software identifies the proteins differentially expressed between cells.
"This is an exciting technology that has great potential for the future," says Ruedi Aebersold, co-founder of the Institute for Systems Biology, who recently helped Dovichi's team identify some proteins. "The strengths include very high detection sensitivity of proteins and the demonstrated ability to manipulate a single cell and to extract, label, separate, and detect proteins from that single cell."
"The potential," Aebersold explains, "is the detection of changes in the proteome at the single cell level. This is significant because it will allow us to dissect responses in single cells and then statistically analyze the variation of cell responses, as opposed to measuring the average response of a cell population as is currently practiced."
Understanding Cellular Variation
Obtaining the capillary electrophoresis "fingerprint" of a single-cell proteome enables researchers to study variation in protein expression among individual cells. In a recent study, for example, Dovichi's team explored the changes in protein expression as cells progress through the cell cycle.
Eventually, Dovichi's team aims to quantitate the inherent "noise" in gene expression and provide a "baseline" of cell-to-cell variability in normal tissue by studying variation in protein expression between nearly identical single cells. The variability caused by a different cellular environment or phase in the cell cycle may be minimized by comparing two daughter cells immediately following division.
"These daughter cells should have nearly identical composition, and differences should be due to only the variability in the packaging of proteins during mitosis," Dovichi says. "Repeating the experiment as a function of time after mitosis will determine the noise in protein expression."
Dovichi predicts that this technology may be applied to cataloging the cellular heterogeneity in the central nervous system; to monitor embryo development or cellular differentiation; and to observe the pharmacokinetics in single rare cells, such as chemotherapy-resistant cells or adult stem cells.
Nor would it be surprising to find this technology ending up in the clinic. "We are interested in cancer prognosis applications," Dovichi says. "Cell-to-cell variability in a tumor increases as the disease progresses. By obtaining protein maps from a hundred tumor cells, we could generate a molecular index that provides prognostic information to guide patient therapy."
Currently, it takes about five hours to generate a protein map from a single cell. But the next-generation instrument, based on multiple capillary sheath flow cuvettes used in large-scale DNA sequencing, could process 500 cells a day.
"We are interested in commercialization of this technology," Dovichi says. "Unfortunately, the investment community has been hesitant to invest in new technology development. We anxiously await a return to a more active investment climate."
Julia Boguslavsky is the conference director for Cambridge Healthtech Institute. She can be reached at firstname.lastname@example.org.