One researcher's DNA is another's unicorn.

One researcher's DNA is another's unicorn

It looks like a DNA molecule. It meanders like a DNA molecule. It even appears to twist like a DNA molecule. But that doesn't mean it is one, warn two chemists at the University of Utah in Salt Lake City.

In the Feb. 8 SCIENCE, Thomas P. Beebe Jr. and Carol R. Clemmer caution that researchers using scanning tunneling microscopes (STMs) to image biological molecules deposited onto a commonly used graphite substrate often mistake biological-looking features of the graphite surface for the biomolecules they aim to study.

Since their introduction in the early 1980s, STMs and related "scanning probe" instruments have provided researchers with observational revelations and unprecedented manipulative powers, even at the atomic level. Scientific journals now teem with striking STM images. "Awash with enthusiasm, the first pretty picture [that seems to show a biomolecule! gets published," says biophysicist and STM researcher Stuart Lindsay of Arizona State University in Tempe.

But pitfalls lurk for the unwary. Last August, several speakers at an STM conference in Baltimore remarked that features of a substrate called highly ordered pyrolytic graphite (HOPG) can muddle interpretations of biomolecular studies. "People have been talking about [this problem! for a while, but until now no one had actually addressed it in a paper," Clemmer says.

She and Beebe used their STM to examine hundreds of blank HOPG surfaces. They found numerous features resembling what they would expect to see if they had deposited DNA or other biomolecules onto the blanks. Beebe admits he has fallen prey to these ambiguities in the past.

Looking at HOPG under an STM is "like looking at a marble floor in a bathroom: You can see anything you want to in it," Lindsay notes.

On HOPG surfaces, the Utah scientists and others have observed linear, periodic features easily mistaken for a chain-like DNA or protein molecule. Sometimes these features "meander" over atomic steps on the substrate, giving the false impression that they are independent of the surface. Some even appear to have a left-handed twist remarkably similar to the helical pitch of many biomolecules.

Beebe acknowledges a "very small probability" that his team's observations resulted from unexpected DNA contamination on the supposedly naked graphite. However, he says, "I strongly feel that that's not the case."

Graphite's propensity for harboring long, twisting, polymer-like features that match textbook representations of molecules like DNA can create a dangerous decoy for scientists who hunt for such molecules. "There are many reports due to come out which we think could be vulnerable to these problems," Beebe says.

He and Clemmer urge biomolecular researchers to abandon graphite for substrates less likely to mimic biological molecules. Some STM users, they note, are turning to gold, often deposited onto a mica surface, as a preferred substrate for such studies.