ABSTRACT
There has been recent debate as to the source of constriction force during cell division. FtsZ can generate a constriction force on tubular membranes in vitro, suggesting it may generate the constriction force in vivo. However, another study showed that mutants of FtsZ did not affect the rate of constriction, whereas mutants of the PG assembly did, suggesting that PG assembly may push the constriction from the outside. Supporting this model, two groups found that cells that have initiated constriction can complete septation while the Z ring is poisoned with the FtsZ targeting antibiotic PC190723. PC19 arrests treadmilling but leaves FtsZ in place. We sought to determine if a fully assembled Z ring is necessary during constriction. To do this, we used a temperature-sensitive FtsZ mutant, FtsZ84. FtsZ84 supports cell division at 30 °C, but it disassembles from the Z ring within 1 min upon a temperature jump to 42 °C. Following the temperature jump we found that cells in early constriction stop constricting. Cells that had progressed to the later stage of division finished constriction without a Z ring. These results show that in Escherichia coli, an assembled Z ring is essential for constriction except in the final stage, contradicting the simplest interpretation of previous studies using PC19.