Challenging a decades-old paradigm: ProB and ProA do not channel the unstable intermediate in proline synthesis after all

The pathway for synthesis of proline in most forms of life produces a highly unstable intermediate, γ-L-glutamyl 5-phosphate (GP). For nearly 70 y, channeling of this intermediate from the active site of glutamate 5-kinase to the active site of GP reductase has been believed to protect GP from cyclization to a dead-end product. However, the evidence presented in support of this idea is not conclusive. We show that changes in the structures of the kinase or reductase that should preclude a protein-protein interaction do not compromise proline synthesis in <i>Escherichia coli</i>, demonstrating that channeling does not occur. We calculate that the half-life of GP is 320 ms. Although GP is indeed unstable, it should diffuse the length of an <i>E. coli</i> cell in less than 3 ms. Thus, most GP produced by glutamate 5-kinase should encounter the active site of GP reductase before cyclization occurs.