Inhibition of transcription (left) but not translation (right) slows DNA segregation.


Bacterial genetics was meant to be easy. But, genetics or no genetics, the identity of the motor for segregating bacterial DNA remained a mystery. “People should have found it,” says Jonathan Dworkin. “It's something that is relatively easy to score for.”

Dworkin and Richard Losick (Harvard University, Cambridge, MA) now claim that the motor has never been found before because it is RNA polymerase. “You use a common mechanism to drive [segregation]—we like the parsimony of that,” says Dworkin. “But one of the things that is appealing about the model also makes it difficult to test”—and impossible to discover by genetics.

Dworkin and Losick have tested the model by direct chemical inhibition of RNA polymerase. They found that the segregation of bacterial chromosomes slows drastically, even though a translation inhibitor has no such effect. This transcription dependence holds true even if replication restarts from a starvation-induced pause.

The researchers imagine that back-to-back DNA polymerase machines at the origin of replication get the two DNA molecules headed off in opposite directions. But this motor is unlikely to sustain lengthy linear movement, as DNA is too flexible to be pushed successfully from one end.

This is where RNA polymerase comes in. The enzyme, which is known to exert greater than five times more force than myosin, is thought to be tethered by its sheer size or its attachment to ribosomes. The directionality of force comes from genome organization. Two-thirds of genes are transcribed in the direction of replication (away from the origin of replication), and up to 75% of transcription involves rRNA genes, all of which point away from the origin. Thus, successive RNA polymerase complexes should spool DNA away from the center of the cell.

“It's difficult to do experiments that are clean,” says Dworkin. “We do feel cautious about [the model].” But he notes a pattern that supports the model. Nearly all surviving large DNA rearrangements are almost symmetrical about the origin of replication, thus retaining the directional bias of transcription. The one known exception has a severe defect in chromosome segregation. Dworkin now plans to test whether this is a fluke or a general rule by engineering multiple origin-asymmetric rearrangements using recombinases. If the theory stands, this may be the original and most direct way of coupling growth (transcription levels) with segregation, and thus division. ▪


Dworkin, J., and R. Losick. 2002. Proc. Natl. Acad. Sci. USA. 10.1073/pnas.182539899.