Harvard researchers examining the activity of a common type of soil bacteria have taken another step in understanding the inner workings of cells, showing that proteins can arrange themselves according to a cell’s inner geometry.

The work was a collaboration between Harvard engineers and molecular biologists, led by Richard Losick, the Maria Moors Cabot Professor of Biology, and Howard Stone, Vicky Joseph Professor of Engineering and Applied Mathematics. The research, published in the March 6 issue of the journal Science, was largely conducted by Kumaran Ramamurthi, a research associate in Losick’s lab, and Sigolene Lecuyer, a postdoctoral fellow in Stone’s lab.

The work focused on Bacillus subtilis, a type of bacteria that has long been a subject of Losick’s research and is a model scientific organism for those seeking to better understand how cells grow, divide, and specialize. In times of environmental stress, these bacteria form tough, heat-resistant spores that can persist in soil for a long time.

During spore formation, the bacteria undergo a modified process of cell division. Instead of dividing into two identical cells, they divide asymmetrically, creating a smaller cell that the larger cell then engulfs and surrounds with a tough, resistant coat. The larger cell then withers away, leaving the spore to wait for better conditions.

Fifteen years ago, Losick began wondering how the proteins generated by the bacterium know to go to the spore’s membrane rather than to the inner side of the bacterium’s own outer membrane.

Over the years, Losick and colleagues examined the 70 proteins that make up the spore’s outer coat. They found that each one moved to the spore by following the protein that was ahead of it. Researchers eventually ran out of prior proteins, however, and were left figuring out how the first protein, called SpoVM, knew to adhere to the spore’s wall.

“For the life of us, we couldn’t figure it out. We examined other molecules on the spore. We were really mystified. Then one day it occurred to us that maybe the answer was simple and right under our noses,” Losick said.

About two years ago, Losick and Ramamurthi began to suspect that perhaps it had something to do with the spore’s geometry, since the spore’s surface is the only convex surface inside the cell. The only other membrane is the one that forms the bacteria’s cell wall, which, looked at from inside the cell, is concave.

To test the idea, scientists labeled the protein with a fluorescent molecule so that they could track it and then observed the process of spore formation. Then they tried the same experiment with yeast and another type of bacteria, neither of which formed spores but both of which contained spherical structures similar in shape to a spore but chemically different. They found that the protein adhered to the convex shapes inside the other cells as well.

“We kept trying to disprove the hypothesis. But everything we did was consistent with the hypothesis,” Ramamurthi said.

The final test came with the help of Stone and Lecuyer’s engineering skills. The researchers created artificial lipid spheres that mimicked the shape of a spore, only without the vast array of molecules and compounds that are normally present in the bacterium.

Lecuyer said it was a bit tricky to create vesicles of the right size and then to figure out how to image them, but all those problems were eventually worked out.

To the mix of different-sized vesicles, they added a purified solution of the fluorescently labeled SpoVM protein. They found that the protein adhered to the artificial spheres just as it had to the natural spore. Further, they also found that the closer the vesicle’s size was to that of an actual spore, the more protein adhered, resulting in brighter fluorescence.

“Working with Rich and Kumaran was a great experience for Sigolene and me,” Stone said. “We learned about the microbiology of a new system, and we were able to test — using in vitro experiments with phospholipid vesicles — an original hypothesis from Rich and Kumaran that small molecules can localize to only one of the membranes — distinguished by the sign and magnitude of the curvature — in the bacterial cell. It was very exciting and more work remains.”

Losick said these findings could potentially lead to the ability to create artificial spores for drug delivery and other uses. The findings may also open the way to a new understanding of protein movement in bacteria, including that of proteins to “poles” in the bacteria’s narrow ends. If determined by the cell’s geometry, Losick speculated that movement might be driven by the extreme concavity of each end of the cell.

The team is also continuing the work, seeking to understand the mechanism by which the proteins move and adhere to the spores. Losick said he’s still troubled by what he terms the “Christopher Columbus” problem. Like Columbus looking out his window in Spain and lacking the perspective to see a curved Earth, the sphere itself is vastly larger than the protein, raising the question of how the protein detects what to it must appear as an extremely slight curvature.

The answer, he said, might lie in the assembly of rafts of proteins providing a larger patch over which to gain perspective, though the answer awaits future exploration.