A single aromatic residue in transcriptional repressor protein KorA is critical for cooperativity with its co-regulator KorB

Abstract

A central feature of broad host range IncP-1 plasmids is the set of regulatory circuits that tightly control plasmid core functions under steady-state conditions. Cooperativity between KorB and either KorA or TrbA repressor proteins is a key element of these circuits and deletion analysis has implicated the conserved C-terminal domain of KorA and TrbA in this interaction. By NMR we show that KorA and KorB interact directly and identify KorA amino acids that are affected on KorB binding. Studies on mutants showed that tyrosine 84 (or phenylalanine, in some alleles) is dispensable for repressor activity but critical for the specific interaction with KorB in both in vivo reporter gene assays and in vitro electrophoretic mobility shift and co-purification assays. This confirms that direct and specific protein-protein interactions are responsible for the cooperativity observed between KorB and its corepressors and lays the basis for determining the biological importance of this cooperativity.

Fig. 1

Alignment of conserved domains from KorA, KlcB and TrbA proteins. For KorA and TrbA proteins, only one representative sequence from each of the extant IncP-1 subgroups is included to improve clarity, along with the homologue of KorA from the IncU plasmid pFBAOT6 that is not thought to interact cooperatively with its cognate KorB. RK2, IncP-1; R751, IncP-1β; pEST4011, IncP-1δ; pQKH54, IncP-1γ; pFBAOT6 (IncU). The position of Tyr84 in RK2 KorA is indicated with an asterisk. Boxshade is used to indicate amino acids that are identical (black) and similar (grey) to the consensus. Amino acid substitutions created during the course of this study are indicated above the alignment.

Fig. 2

2D NMR analysis of KorA–KorB interaction. Overlay of ¹H-¹⁵N HSQC of a solution containing a 2:1 ratio of ¹⁵N-KorA to unlabelled KorB (green) on free ¹⁵N-KorA (red). The free KorA residues that were shifted to a new position in the titration are marked with their residue number: 65, 70, 72, 73, 74, 76, 77, 78, 82, 84, 85, 87, 89 and 90. New peaks that were clearly due to amino acids in new environments are indicated with boxes, but have not been unambiguously assigned.

Fig. 3

A. Scheme of the reporter plasmid pDM3.1 containing the korA promoter linked to the xylE reporter plasmid (Macartney et al., 1997). The sequence below shows the promoter region with KorB and KorA operators underlined (O_B and O_A respectively) as well as the the −35 and −10 boxes in bold. B. Phenotypes of KorA C-terminal domain mutants in reporter gene assays alone and in the presence of KorB. Bars indicate reporter gene activity from korA promoter relative to activity in the absence of any repressor. Black bars, in the presence of KorA; white bars, in the presence of KorA and KorB. Error bars indicate standard deviation. Mutants are grouped by similar phenotype.

Fig. 4

Co-purification of KorA with His-tagged KorB. Cell extracts with WT or mutant KorA and His-KorB alone were mixed, then Ni-NTA added and His-KorB purified. Samples of the extracts or the purified His-KorB were fractionated by SDS-PAGE, stained with Coomassie blue to detect His-KorB, Western-blotted and probed with purified anti-KorA antiserum. Labels on the horizontal axis indicate presence or absence of KorA/His-KorB in lysates. Negative controls were: extracts lacking KorA (−/− and −/+), which showed no KorA signal in either the presence or the absence of His-KorB; and extracts with KorA mixed with Ni-NTA in the absence of His-KorB (WT/–), which showed negligible retention of KorA. Band intensity values were normalized to those obtained with WT KorA/His-KorB.

Fig. 5

KorB electrophoretic mobility shift of a 220 bp korAp DNA fragment, in the absence and presence of KorA WT and Y84A. The korAp fragment was released from pSTM12 by EcoRI digestion – the 2.6 kb vector fragment runs at the top of the gel. The korA promoter includes KorA (O_A: 5′-GTTTAGCTAAAC-3′) and KorB (O_B: TTTAGCCGCTAAA-3′) sites separated by 20 bp. Each protein had previously been titrated separately to determine suitable concentrations. Proteins were added separately or together as described in Experimental procedures. The percentage of DNA fragment in the KorB–DNA complex, with or without KorA, is indicated on the gel. At 30 nM KorB the presence of 300 nM WT KorA results in > 80% retardation of the KorAp–KorA complex whereas the presence of 300 nM Y84A KorA has no effect.

Fig. 6

Scheme to summarize possible KorA–KorB interactions. For KorA the DNA binding (aa 1–61) and dimerization (aa 68–101) domains are shown as 1 and 2, respectively, while for KorB the apparently flexible N-terminal domain (aa 1–138), the DNA binding domain (aa 139–252), a putative flexible linker (aa 253–298) and the C-terminal dimerization (aa 299–358) domains are labelled 1, 2, 3 and 4 respectively. A. In solution, KorB can titrate the KorA NMR signals at a ratio of 1 KorB: 2 KorA so the complex is shown with this stoichiometry. However, on DNA the presence of specific binding sites constrains the location of adjacent KorA and KorB dimers and contact may be made by looping (B) or spreading (C) in a flexible way as rotation of the binding sites through 180° to each other does not interfere with observed cooperativity (Bingle et al., 2005).