The cyclic AMP (cAMP) receptor protein (CRP) is a DNA binding protein from Escherichia coli that can activate transcription when bound to specific sites located near promoters. CRP can serve as a model system for the study of gene regulation. We have been studying several steps in the pathway to CRP activation of transcription. (i) In order for CRP to become activated as a DNA binding protein, cAMP must bind to the protein and cause an allosteric change. We have isolated CRP mutants that are defective in the cAMP-induced allosteric change. In collaboration with Richard Brennan at the Oregon Health Sciences University, we have begun the X-ray crystallographic analysis of one of the CRP mutants. To further assign residues responsible for the cyclic AMP-induced change in CRP, we have selected intragenic suppressor mutations for the allosteric defect. The positions of these suppressor mutations confirm our earlier hypothesis as to how the conformational change occurs in CRP. We have also addressed the issue of how many cAMP molecules must bind to CRP to activate it. We have made CRP heterodimers that contain one wild type subunit and one subunit that cannot bind cAMP. It appears that CRP can function with only one cAMP bound, although the subunit with the cAMP bound binds more easily to DNA. (ii) We are addressing how CRP activates transcription in vivo, since there appears to be far more CRP activation of lactose operon transcription in vivo than in vitro. We have isolated E. coli mutants that are defective in CRP activation of the lac operon; one of the mutations is in the gene encoding isocitrate dehydrogenase. A likely explanation for this is that accumulation of citrate or isocitrate prevents CRP activation of transcription. We are currently searching for the target of the citrate/isocitrate toxicity, and are in the process of mapping E. coli mutations that suppress the citrate/isocitrate toxicity. We have also investigated the effects of citrate and isocitrate on in vitro transcription. Citrate and isocitrate inhibit most transcription, and this can be compensated for by adding back excess Mg++. What is surprising is that CRP-activated transcription can survive the Mg++-depletion. This result is the opposite to what we would have expected based on the in vivo results, since in vivo it appeared that excess citrate inhibited CRP activation of transcription. We are continuing to investigate this in vitro/in vivo discrepancy. We are also continuing our study of our previous finding that the small DNA binding protein HU decreases the amount of lac transcription in the absence of CRP. Therefore, HU accounts for some of the discrepancy between and in vitro and in vivo results.
