Researchers are puzzling out a central mechanism for how some proteins navigate inside bacterial cells: Rather than using biochemical cues, they appear to rely on the cells' geometry, sensing the membrane's curvature, two recent studies suggest.

Gram-stained Bacillus subtilis Image: Wikipedia

"This is an important and fundamental observation," said Lucy Shapiro at Stanford University, who did not participate in the research.

Because bacterial cells are so small, researchers for years thought they were essentially bags of enzymes, with no mechanisms for protein localization; proteins could just diffuse wherever they needed to go in milliseconds, she explained. In the last decade, however, that idea has been overturned, with her own lab and others identifying a high level of organization in the bacterial cell.

Proteins aggregate at spots where they are needed by a mechanism called diffusion capture -- after being made at the ribosome, a protein travels through the cell by Brownian motion and is "captured" by some mechanism at a specific position. Its presence in that location acts as a cue to attract more such proteins. "So the ultimate question is, well how does the first one get there?" Shapiro said.

In a study published in Science in March, Richard Losick's group at Harvard showed that a protein involved in spore formation in Bacillus subtilis identified positive -- that is, convex -- curvature in the cell membrane. Because bacteria contain no organelles, they "don't usually have convex surfaces," said Kumaran Ramamurthi, a postdoc in Losick's lab and first author of that paper, and spore formation is an exception to that rule.

In the Science paper, Ramamurthi and his colleagues purified the protein, called SpoVM, and incubated it with different-sized vesicles in vitro, finding that it preferentially located to the smallest (and thus the most convex) vesicles. They then put the protein into E. coli cells and yeast cells mutated to form vesicles similar to those formed during sporulation. Again, the protein preferentially localized to the convex membranes of those vesicles -- suggesting it recognizes geometric cues rather than cell-specific ones, which are likely to be chemical.

Now, in another study published in PNAS this week, the same group showed that a second protein called DivIVA, which helps regulate cell division in Bacillus subtilis, identifies negative, or concave, curves in the membrane. DivIVA normally localizes to the most concave part of the cell during cell division -- the ring at the emerging septum of the cell. It is also enriched at the hemispheric poles of the cell, also a highly concave environment. When the researchers prevented cytokinesis, the ring didn't form, and the protein redistributed to the poles. When they used an enzyme to transform rod-shaped Bacillus cells into spheres, the protein redistributed evenly along the membrane, showing its preference for concavity.

"Now we have two proteins that seem to recognize two types of membrane curvature," said Ramamurthi, also the first author of the PNAS study. "So now, we're wondering aloud whether recognition of cell geometry could be a general mechanism by which proteins are deployed in bacterial cells."

For an individual protein, the curvature is slight and almost imperceptible, Ramamurthi explained. "It would be like us standing on the surface of the earth and sensing that it is curved." In their Science paper, the group showed that although a single protein could not sense such a small curvature, a cluster of proteins functioning cooperatively could.

Their evidence for such a mechanism was "just beautiful, really rock solid," said Christine Jacobs-Wagner of Yale University, who did not participate in the research. The next question, she said, is how exactly such sensing works. A slew of recent work in the field has identified an increasing number of proteins that localize to the hemispheric poles of the cell, she noted, and the two studies together "suggest that the poles are particularly special places."

Shapiro noted, though, that there are proteins that don't use this mechanism to localize, and future studies will surely identify them. "I should stress that this is just one mechanism," she said. "I predict there will be multiple, multiple mechanisms."


Related stories:

LOV story
[May 2009]

How bacteria talk
[June 2006]

Bacterial proteins on the move
[1st July 2005]