Published online 12 January 2009 | Nature | doi:10.1038/news.2009.15


MRI goes to the nanoscale

Picture of virus points way to kinder, gentler molecular imaging.

Virus particles on the tip of a silicon lever.IBM

By taking a snapshot of a virus, researchers have created the highest-resolution image ever made by magnetic resonance imaging (MRI). Such nanoscale MRI could one day offer a non-invasive way to reveal the three-dimensional details of biological structures such as proteins.

The image was made by a team based at IBM's Almaden Research Center in San Jose, California. The new technique can resolve a volume 100 million times smaller than the best achieved using conventional MRI.

"This opens a door to a huge vista," says Scott Fraser, director of the Biological Imaging Center at the California Institute of Technology's Beckman Institute in Pasadena. "This is a very, very important step toward the goal of non-destructive, non-invasive imaging."

Conventional MRI, of the type used in hospitals, uses electrical coils to pick up magnetic signals from hydrogen nuclei, and processes the signals to make a three-dimensional image. But the magnetic forces are so weak that typical MRI instruments can't detect anything smaller than a micrometre across, about the size of a bacterium.

Another dimension

In pursuit of finer resolution, physicist Daniel Rugar and his colleagues attached tobacco mosaic virus particles to a tiny silicon lever. They then exposed the lever to a fluctuating magnetic field, which caused the hydrogen nuclei in the viruses to emit their own magnetic signals. The nuclei alternately attracted and repelled a nearby magnetic tip, creating ever-so-slight vibrations in the lever that the researchers could use to reconstruct an image of the virus.

"I am just stunned by these images," says Chris Hammel, a physicist at Ohio State University in Columbus. "I think it's really remarkable."

Rugar and others first demonstrated the basic technique, called magnetic resonance force microscopy (MRFM), more than 15 years ago1. But by improving the instrument's sensitivity and developing new signal-processing algorithms, the IBM team has now made the first three-dimensional nanoscale MRFM image of a biological specimen.

The image, which has a resolution of less than 10 nanometres — comparable to a picture from a scanning electron microscope — shows individual virus particles. The findings are published in the Proceedings of the National Academy of Sciences2.

MRFM might eventually reveal molecular structures invisible to two-dimensional imaging techniques such as scanning probe microscopy, which "get excellent resolution, but only on the surface", says Rugar. "What these people do in two dimensions, we want to do in three dimensions."

Destroy after reading

There are other three-dimensional imaging techniques, but they have their own limitations. X-ray analysis requires the sample to be crystallized, which isn't possible for all proteins, and nuclear magnetic resonance spectroscopy doesn't work well for larger molecules. Cryo-electron microscopy destroys the sample, and scans of many copies of the same molecule are needed to achieve fine resolution.

In contrast, MRFM works on a single specimen and does not use damaging radiation. With some tweaking of the frequency of the magnetic field and additional labelling techniques, it could also image individual chemical elements in a sample.

Hammel notes that the experiment was performed at cryogenic temperatures and that the delicate handling needed to coat the sample on the lever is a specialist skill. Still, he says, the technique might be applicable to almost any molecule. Fraser points out that the virus particles were dried, which could alter their structure, and says the method will need to work with samples preserved in ice if it is to become a general tool. 

  • References

    1. Rugar, D., Yannoni, C. S. & Sidles, J. A. Nature 360, 563–566 (1992). | Article | ISI |
    2. Degen, C. L., Poggio, M., Mamin, H. J., Rettner, C. T. & Rugar, D. Proc. Natl Acad. Sci USA Advance online publication doi:10.1073/pnas.0812068106 (2009).
Commenting is now closed.