The role of rigidity in DNA looping-unlooping by AraC

Data obtained from both in vivo and in vitro experiments have shown that, in the absence of arabinose, the dimeric AraC protein prefers binding to the well separated I1 and O2 half-sites and forming a DNA loop (1–4), Fig. ​Fig.1.1. On the addition of arabinose, the protein's affinity for the I1 and I2 half-sites increases by about 50-fold, leading the protein to prefer to bind to these adjacent half-sites and induce the pBAD promoter rather than loop and repress pBAD (3, 5–8). The basis for the change in the DNA binding properties appears to result from an arabinose-induced shift of the N-terminal arms from associating with the DNA-binding domains of AraC to associating with the dimerization domains (refs. 4, 9, and 10; Fig. ​Fig.1).1). The process involving this shift is called the light switch mechanism because the arms switch expression of the system on and off. Figure 1 The regulatory region of the araBAD operon showing the binding sites for AraC, the two promoters (pBAD and pC), and the light switch mechanism for AraC action. AraC bound to the O1 site in the presence of arabinose is shown in gray because occupancy ... In the absence of arabinose, the arms from the dimerization domains are thought to bind to the DNA-binding domains and rather rigidly hold them apart from one another and in an orientation that favors DNA looping. Whether the arms also affect the DNA binding properties of the individual DNA binding domains has not been addressed before this work. A rigid connection between the domains mediated by the arms is consistent with the fact that reversing the head-to-tail orientation of the asymmetric O2 half-site while retaining its AraC-contacting area on the same face of the DNA eliminates DNA looping (4). Another property that can be attributed to rigidity can be seen in the binding of AraC in the absence of arabinose to two adjacent I1 and I2 half-sites. Whereas such binding normally does not occur in vivo, it can be observed in vitro (3, 6). According to the current view, to bind to adjacent half-sites, part of the DNA binding energy must be used to distort AraC so that the DNA binding domains are correctly positioned adjacent to one another. In the presence of arabinose, however, the DNA-binding domains of AraC are freed from the arms and can easily adopt an orientation compatible with the adjacent DNA half-sites. Thus, the presence of arabinose increases the affinity of AraC for adjacent half-sites, whether they are in direct repeat or inverted repeat orientation (6, 11). One important input to the formulation of the light switch mechanism was the observation that deleting the N-terminal arms of AraC shifts the protein's preference in the absence of arabinose to that of preferring to bind adjacent half-sites rather than looping (9). Another important input was genetic information indicating an interaction between the N-terminal arm of AraC and the DNA binding domain of AraC (9). Whereas the light switch mechanism seems to be the simplest model compatible with the existing genetic and physiological data, and is consistent with the structure of the dimerization domain in the presence and absence of arabinose (12), the model has not been subjected to extensive biophysical testing. Here, we describe experiments designed to explore this latter area, in particular, the role of rigidity and flexibility of AraC and the DNA in the behavior of AraC in response to arabinose. The experiments also explore in a general way the signals that are sent from the dimerization domains to the DNA binding domains and whether the intrinsic affinity of the DNA binding domains for DNA is modulated. The experimental approaches we used should be applicable to the study of many ligand-regulated DNA binding proteins. In the case of AraC, the results rule out entire classes of general models, but are entirely consistent with the light switch mechanism, and in fact, refine the mechanism.