When, in 1996, a group of NASA researchers presented several lines of evidence for fossil bacteria in a Martian meteorite, a wave of excitement passed through the public and the scientific community alike. Of course, that wave was followed by a storm of controversy.

Five years of scrutiny and debate over the NASA group's claims have since brought all but one of their arguments unceremoniously back to Earth. Non-biological processes and contamination could explain the "bacterium-shaped objects" and organic chemicals found in the meteorite, other scientists have argued.

Only one line of evidence for bacterial life in the meteorite still stands: Microscopic crystals of a mineral called magnetite. According to the NASA scientists, the magnetite crystals found in the meteorite are so structurally perfect, chemically pure, and have such unique, distinctive three-dimensional shapes that only bacteria could have produced them, not any inorganic process. This claim, too, is now being assailed by new data and criticisms from an Arizona State University research team and their collaborators.

Top: Modern magnetotactic bacteria, one showing a chain of magnetite crystals, as seen in the backscattered scanning electron microscope. Bottom: Magnetite crystals and chains of magnetite crystals in the Martian meteorite ALH84001 in the backscattered scanning electron microscope. One conspicuous chain indicated by arrows. The diameter of a single crystal is approximately one-millionth of an inch. Photo: NASA

Peter Buseck, Regent's Professor of geological sciences and professor of chemistry and biochemistry at ASU, and Martha McCartney, a research scientist at the ASU Center for Solid State Science, argue that the match between the meteoritic crystals and those in bacteria is at best ambiguous. At worst, they say, the data used in the NASA group's analysis is mistaken.

In their paper, "Magnetite Morphology and Life on Mars," published November 20, 2001, in the Proceedings of the National Academy of Sciences, Buseck and his co-authors assert that the evidence for bacterial magnetite crystals on the Martian meteorite is inadequate. In doing so, they may have cut the Martian meteorite's last tenuous hold on life.

The magnetite crystals in the meteorite are tiny, even by an electron microscopist's standards, at only 40 to 100 billionths of a meter wide. And there's the rub. The technology necessary to accurately describe the three-dimensional shape of such small crystals has become available only in the last few years, and has not yet been used to study the magnetite grains in the meteorite. Therefore, says Buseck, it is too early to say for sure what the exact shapes of the meteoritic crystals are, let alone whether they provide identical matches to those in bacteria.

The only kind of microscope powerful enough to produce clear images of such small crystals is a transmission electron microscope, or TEM. By using a beam of electrons rather than a beam of light to view the sample, the TEM allows researchers to see objects smaller than one billionth of a meter wide. But a TEM sees only in two dimensions. It generates a spectacular silhouette image of the sample, but conveys little about its thickness.

An accurate description of the crystals' complex three-dimensional shapes requires that they be examined from a variety of perspectives. Discriminating between their flat facets and tapered edges is a particular challenge - when viewed in profile, the two are indistinguishable straight edges. Only by tilting each crystal at dozens of angles can scientists unequivocally identify their three-dimensional shapes, says Buseck.

Multiple views of a magnetite (Fe3O4) nanocrystal from within a magnetotactic bacterium. This tomographic reconstruction shows the three-dimensional shape of the crystal viewed from several directions. The images were obtained in a transmission electron microscope with a field-emission gun. A chain of such crystals from within a single bacterial cell runs from the upper left to the lower right. Photo: ASU

At the time of the NASA group's study, the tilting experiments could be done only by hand, with great technical difficulty. "It's a lot of work and it's not very precise," says McCartney. The NASA group used this approach to create images of the magnetite crystals from both the meteorite and from one strain of bacteria.

Since then, scientists studying the three-dimensional shapes of crystals have upgraded TEM technology and merged it with computer technology. "The microscope stages and beam shifts and focuses have come under computer control, which makes the experiments much more doable" and more precise, says McCartney.

Only two laboratories, Buseck and McCartney's and that of their co-authors in Cambridge, have applied the new technology to study magnetite crystal shapes. Using these new developments, they have reexamined the evidence described in the NASA team's study.

"The shape [the NASA group] came up with disagreed with what we thought the shape was," says McCartney. This difference calls into question whether the shapes of the meteoritic crystals are accurately known and whether the claim of an exact match - the only remaining evidence for bacterial life on the meteorite - is accurate.

Buseck's team also criticizes several other underpinnings of the Martian life claim. The NASA group selected only 27 percent of all the magnetite crystals present in the Martian meteorite for comparison with bacterial crystals. The Buseck group implicitly questions both the objectivity of their selection and the effect of such a limited comparison on their conclusions.

Further, Buseck and McCartney's team demonstrates that the shapes of bacterial magnetite grains vary more than scientists had previously thought. The shapes and sizes differ among bacterial strains and even within individual bacteria. That expanded variety makes it more likely that bacterial and meteoritic magnetite grains could appear to match by simple chance. Lacking sufficiently precise data and resting on a restricted analysis, the NASA team's claims must be considered best guesses, Buseck and his co-authors argue.

However, they have not eliminated the possibility that the Martian crystals could have a biological origin. With more advanced technology now at their disposal, Buseck and his collaborators plan more conclusive studies of the magnetite crystals from both the meteorite and several strains of terrestrial bacteria.

"We will look at them in far greater detail than others have been able to do before," says Buseck.

Buseck and McCartney's co-authors on the paper are Rafal Dunin-Borkowski, Paul Midgley, Matthew Weyland (all of Cambridge University, England), Bertrand Devouard (of Blaise Pascal University, France), Richard Frankel (of California Polytechnic State University), and Mihály Pósfai (of the University of Veszprem, Hungary).