Membrane Close-Up

A team of researchers has shown that they can use imaging mass spectrometry to obtain quantitative, chemically specific pictures of domains in lipid bilayers with spatial resolution of less than 100 nm. Such resolution could turn imaging mass spectrometry into a tool to answer long-standing questions about the composition of similar domains in biological membranes.

The team's images of a membrane model system are "the precursor to the dream that one might be able to actually image the lateral organization of real biological membranes on a length scale and with chemical specificity at a level where you might be able to say something as opposed to infer something," says team leader Steven G. Boxer, a professor of chemistry at Stanford University.

Boxer, postdoc Mary L. Kraft, and their collaborators used nanoscale secondary ion mass spectrometry (nanoSIMS) to image lipid bilayers made of two different lipids, ¹⁵N-labeled 1,2-dilauroylphosphatidylcholine (¹⁵N-DLPC) and ¹³C-labeled 1,2-distearoylphosphatidylcholine (¹³C-DSPC) (Science 2006, 313, 1948).

In the nanoSIMS method, the lipid bilayer is bombarded by a tightly focused beam of cesium ions, which continuously ablate the surface to generate secondary ions that can be identified and measured by mass spectrometry. The cesium beam is moved over the surface and spectra are collected at each location. The spectra are used to generate images of the distribution of the lipids. The lipids' isotope labels let the researchers quantify the lipids. They find that the model membrane consists of distinct microdomains of ¹³C-DSPC within a matrix of ¹⁵N-DLPC.

"The quality of the images is spectacular in my view, and the fact that Boxer can get quantitative information is icing on the cake," says Nicholas Winograd, a chemistry professor at Pennsylvania State University and a pioneer in imaging mass spectrometry. He adds that even though fluorescence labeling of lipids is ultimately a more sensitive method for probing bilayer composition, "the mass spec approach avoids the possibility of disturbing the energetics of fragile systems like lipids and their domains."

The researchers characterized the membranes by using nanoSIMS in parallel with atomic force microscopy, and their results demonstrate the power of combining these two analytical methods. The nanoSIMS helped them interpret unusual AFM images that showed depressions in the middle of lipid domains. These depressions had none of the typical secondary ions from lipids and turned out to be debris rather than lipids. "Those are the most expensive images of dirt you're ever going to see," Boxer says. "I don't think the mass spec community would believe that we can get such good spatial resolution and information content if it weren't for the AFM image."

Methods for characterizing biological membranes generally work at lengths less than 10 nm or longer than 300 nm. "The need for imaging biomembrane organization at length scales of 10 to 300 nm thus remains largely unmet," writes Jay T. Groves, a chemistry professor at the University of California, Berkeley, in an accompanying commentary (Science 2006, 313, 1901). "Although imaging mass spectrometry is still in its infancy, it is emerging as a powerful technique that uniquely accesses a strategic gap in our knowledge of cell membrane structure."