The catabolite activator protein (CAP) activates the expression of more than 100 genes involved in secondary sugar metabolism in E.coli. Apparently, it always binds in sites that are away from -10 and -35 elements of the promoter by multiples of 10 base pairs (-70 or -60, for example). How could this be related to its mechanism of action?
I've searched about this particular activator and I found that it acts mainly by recruiting RNAP to weak promoters, although it does alter the architecture of the DNA, introducing an abrupt bend (for some CAP-dependent genes, this favours conformational changes in RNAP involved in the initialization of transcription): here's the crystallographic data.
However, I did not find anything conclusive about the pattern of binding to distances which are usually multiple of 10 base pairs (which is close to the number of base pair in a complete turn of the double helix).