Finally, protein–protein interactions were identified by SDS-PAGE. As shown in Fig. 2a, 16.7 kDa His6–WhcA and 62.2 kDa GST–SpiA coeluted together,

indicating specific binding. Nonspecific binding of the GST–SpiA protein to the beads was not observed (data not shown). Purified maltose-binding protein, which was used to assess nonspecific interactions, did not bind to the bait His6–WhcA (data not shown). However, the band intensity of the GST–SpiA protein was lighter than expected (Fig. 2a), suggesting a weak protein Thiazovivin manufacturer interaction. If protein–protein interactions occurred at a 1 : 1 molar ratio, the band intensity of the GST–SpiA protein, which was three times larger than the His6–WhcA in size, should be approximately three times stronger than that of the His6–WhcA band. This discrepancy could be due

simply to inefficient refolding, leaving only a fraction of Sunitinib manufacturer the bead-bound His6–WhcA in the correct conformation. Alternatively, fractions of the refolded His6–WhcA could have lost their Fe–S cluster during the denaturation–refolding process, thus remaining in an alternative conformation that does not interact with GST–SpiA (see Discussion). Nevertheless, the pull-down assay indicated that WhcA can specifically bind the SpiA protein. So far, we were able to show that WhcA interacts with SpiA via in vivo and in vitro assays. As the WhcA protein was found to play a negative role in the oxidative stress response pathway, we postulated that the protein–protein interaction could be affected by external factors, such as external redox environments. When oxidant diamide was applied to growing HL1387 cells, the interaction between WhcA and SpiA was significantly reduced to 34% relative to those of positive and negative control strains (Fig. 3a). The effect of oxidant menadione

was observable but rather marginal (Fig. 3b), whereas reductant dithiothreitol was not effective at all in disrupting the protein–protein interaction (data not shown). Whereas the thiol-specific oxidant diamide specifically oxidizes sulfhydryl groups (Kosower & Kosower, 1995), the redox-cycling compound menadione exerts its toxic effects via stimulating intracellular production of superoxide radicals and hydrogen peroxide (Hassan & Fridovich, 1979). Phospholipase D1 However, the redox-cycling compound is also known to drain electrons from the reductive pathways, including the thioredoxin system (Holmgren, 1979), thus inducing disulfide bond formation in cells. The differential response of the protein to diamide and menadione may suggest that the cysteine residues of the WhcA protein are involved in disulfide bond formation. To study the effect of diamide on in vitro protein–protein interactions, the pull-down assay was performed in the presence of oxidant diamide, as described in Materials and methods.