The structure-specific nuclease human flap endonuclease-1 (hFEN1) plays a key role in DNA replication and repair and may be of interest as an oncology target. We present the first crystal structure of inhibitor-bound hFEN1 and show a cyclic N-hydroxyurea bound in the active site coordinated to two magnesium ions. Three such compounds had similar IC50 values but differed subtly in mode of action. One had comparable affinity for protein and protein–substrate complex and prevented reaction by binding to active site catalytic metal ions, blocking the unpairing of substrate DNA necessary for reaction. Other compounds were more competitive with substrate. Cellular thermal shift data showed engagement of both inhibitor types with hFEN1 in cells with activation of the DNA damage response evident upon treatment. However, cellular EC50s were significantly higher than in vitro inhibition constants and the implications of this for exploitation of hFEN1 as a drug target are discussed.

The ammonium‐dependent posttranslational regulation of nitrogenase activity in Azospirillum brasilense requires dinitrogenase reductase ADP‐ribosyl transferase (DraT) and dinitrogenase reductase ADP‐glycohydrolase (DraG). These enzymes are reciprocally regulated by interaction with the PII proteins, GlnB and GlnZ. In this study, purified ADP‐ribosylated Fe‐protein was used as substrate to study the mechanism involved in the regulation of A. brasilense DraG in vitro. The data show that DraG is partially inhibited by GlnZ and that DraG inhibition is further enhanced by the simultaneous presence of GlnZ and AmtB. These results are the first to demonstrate experimentally that DraG inactivation requires the formation of a ternary DraG‐GlnZ‐AmtB complex in vitro. Previous structural data have revealed that when the DraG‐GlnZ complex associates with AmtB, the flexible T‐loops of the trimeric GlnZ bind to AmtB and become rigid; these molecular events stabilize the DraG‐GlnZ complex, resulting in DraG inactivation. To determine whether restraining the flexibility of the GlnZ T‐loops is a limiting factor in DraG inhibition, we used a GlnZ variant that carries a partial deletion of the T‐loop (GlnZΔ42‐54). However, although the GlnZΔ42‐54 variant was more effective in inhibiting DraG in vitro, it bound to DraG with a slightly lower affinity than does wild‐type GlnZ and was not competent to completely inhibit DraG activity either in vitro or in vivo. We, therefore, conclude that the formation of a ternary complex between DraG‐GlnZ‐AmtB is necessary for the inactivation of DraG.