Interactions between activating region 3 of the Escherichia coli cyclic AMP receptor protein and region 4 of the RNA polymerase sigma(70) subunit: Application of suppression genetics

The Escherichia coli cyclic AMP receptor protein, CRP, induces transcriptio n at Class II CRP-dependent promoters by making three different activatory contacts with different surfaces of hole RNA polymerase. One contact surfac e of CRP, known as Activating Region 3 (AR3), is functional in the downstre am subunit of the CRP dimer and is predicted to interact with region 4 of t he RNAP sigma(70) subunit. We have previously shown that a mutant CRP deriv ative that activates transcription primarily via AR3, CRP HL159 KE101 KN52, requires the positively charged :residues K593, K597 and R599 in sigma(70) for activation. Here, we have used the positive control substitution, EK58 , to disrupt ARS-dependent activation by CRP HL159 KE101 KN52. We then scre ened random mutant libraries and an alanine scan library of sigma(70) for c andidates that restore activation by CRP HL159 KE101 KN52 EK58. We found th at changes at R596 and R599 in sigma(70) can restore activation by CRP HL15 9 KE101 KN52 EK58. This suggests that the side-chains of both R596 and R599 in sigma(70) clash with K58 in CRP. Maximal activation by CRP HL159 KE101 KN52 EK58 is achieved with the substitutions RE596 or RD596 in sigma(70). W e propose that there are specific charge-charge interactions between E596 o r D596 in sigma(70) and K58 in AR3. Thus, no increase in activation is obse rved in the presence of another positive control substitution, EG58 (CRP HL 159 KE101 KN52 EG58). Similarly, both sigma(70) RE596 and sigma(70) RD596 c an restore activation by CRP EK58 but not CRP EG58, and they both decrease activation by wild-type CRP. We suggest that E596 and D596 in sigma(70) can positively interact with K58 in AR3, thereby enhancing activation, but neg atively interact with E58, thereby decreasing activation. The substitution, KA52 in AR3 increases Class II CRP-dependent activation by removing an inh ibitory lysine residue. However, this increase is not observed in the prese nce of either sigma(70) RE596 or sigma(70) RD596. We conclude that the inhi bitory side-chain, K52 in AR3, clashes with R596 in sigma(70). Finally, we show that the sigma(70) RE596 and RD596 substitutions affect CRP-dependent activation from Class II, but not Class I, promoters. (C) 2000 Academic Press.