Science News - October 8, 2011

the best treatment, Altschuler says this new approach differs because, in principle, it would allow a doctor to look at every cell in a tumor biopsy, identifying clues that would get lost in a “population averaging” approach.

Multiplying traits

But fluorescent microscopy has its own limitations. Scientists can look for only a handful of variations, or biomarkers, in a cell at one time. Though cells carrying biomarker A could be distinguished from those carrying B or C, for example, scientists can’t simultaneously scan the cells for a second characteristic that might be important in understanding the first.

Altschuler and Wu’s team has found ways to maximize the use of fluorescent probes to get more details on cell-to-cell differences. Using three out of four available probes to home in on subpopulations of interest, a fourth probe can be free to search for a second variation within each of those groups. By running a series of experiments, rotating the fourth marker’s target each time, scientists can get more detailed information on members in each group.

Say you’re trying to understand peo-
ple at a baseball stadium, instead of
cells, and you want to figure out differ-
ences in the way the players, vendors
and fans dress. Three out of four fluores-
cent probes would be used to separate
people into each of these groups. The
fourth probe could then search for a spe-
cific item, such as sneakers, which may
be worn by members in each group. In
subsequent experiments, the first three
probes remain set to identify players,
vendors and fans, while the fourth spots
people sporting home-team jerseys, then
baseball caps and so on. A computer pro-
gram stitches together the information
gleaned from the serial studies to show
what people in each group wear.

Live on and inform

Such clues to what’s going on inside single cells often come at a price: Cells are damaged or killed in the process, providing only a single snapshot in time. This makes it hard to predict how a single cell will respond to a given stimulus.

J. Christopher Love, a chemical engi-neer-turned immunologist at MIT, combines the idea of separating cells and the helpfulness of fluorescent dyes to follow biochemical processes over time in living cells. His lab takes an array of sub-nanoliter-sized wells and fills each with either a single immune cell or a small group of them. With tens of thousands of wells, the unit looks like an ice cube tray the size of a microscope slide. Fluorescent markers identify the various proteins in or on each cell, distinguishing one cell type from another.

Over the course of several hours or days, Love studies the proteins and chemicals secreted by the cells to see how they change. Borrowing a technique from the printing of U.S. dollar bills, he presses the microarray against a glass plate to get a printout (so to speak) of the proteins and chemicals secreted by the cells. Unlike Allbritton’s work, the

Trouble with averages for a long time scientists had no choice but to study cells by looking at the average behavior of an entire population. today, many researchers are developing analytical techniques that get around the inherent problems (some below) of this approach.