Mutations upstream from sdaC and malT in Escherichia coli uncover a complex interplay between the cAMP receptor protein and different sigma factors

Abstract

In Escherichia coli, one of the best understood microorganisms, much can still be learned about the basic interactions between transcription factors and promoters. When a cAMP-deficient cya mutant is supplied with maltose as the main carbon source, mutations develop upstream from the two genes malT and sdaC. Here, we explore the regulation of the two promoters, using fluorescence-based genetic reporters in combination with both spontaneously evolved and systematically engineered cis-acting mutations. We show that in the cya mutant, regulation of malT and sdaC evolves toward cAMP-independence and increased expression in the stationary phase. Furthermore, we show that the location of the cAMP receptor protein (Crp) binding site upstream of malT is important for alternative sigma factor usage. This provides new insights into the architecture of bacterial promoters and the global interplay between Crp and sigma factors in different growth phases.IMPORTANCEThis work provides new general insights into (1) the architecture of bacterial promoters, (2) the importance of the location of Class I Crp-dependent promoters, and (3) the global interplay between Crp and sigma factors in different growth phases.

Keywords: cAMP receptor protein Crp/Cap; sigma factors; stationary phase.

Fig 1

malt and sdaC mutations occurring in experimental evolution of Escherichia coli. (a) Illustration of the experiment. E. coli K-12 MG1655 cya was plated on MacConkey maltose agar, small white colonies formed, and over time maltose fermenting (red) and non-fermenting (white) secondary colonies appeared. (b) Mutational hotspots found previously (4, 8, 28) mapped by genomic locations, and marked by mutant phenotype and mutation type; red, fermenting; white, non-fermenting; (c), cis-acting mutation; (T), trans-acting mutation. (c) Consensus sequence for RpoD, along with alignment of the cis-acting malT (SNP) and sdaC (insertions) promoter mutations (*) to their native (wt) sequences in E. coli K-12 MG1655. (d) Illustration of the sdaCB operon and its function. L-serine import is mediated by the sdaC-encoded L-serine-H⁺-symporter, and degradation of L-serine to pyruvate and NH₄⁺ is catalyzed by the L-serine deaminase SdaII. (e) Serial dilutions of E. coli K-12 MG1655 cya with the native or mutant PsdaC promoters on M9 minimal media at 37°C in the presence or absence of L-serine as the sole carbon source. Two independent isolates of each mutation type were found and are marked (1) and (2).

Fig 2

Characterization of the malT promoter and hybrid RpoD consensus mutants. (a) Illustration of the reporter plasmid according to SBOL standards, elements being; a truncated malT promoter, without the Mlc site, an RNA-stability element, a ribosomal binding site, the coding sequence of superfolder GFP, an ssrA tag for protein degradation, and a terminator. The reporter also contains an ampicillin resistance marker for plasmid maintenance and a kanamycin resistance marker with FRT sites for genomic integration. (b) Fluorescence from the PmalT reporter in E. coli K-12 MG1655 after 9 h of growth in the presence or absence (Ø) of glucose (glc, 1%). Data represent biological duplicates. (c)The activity of the reporter in E. coli K-12 MG1655 cya after 9 h of growth with different levels of cAMP. Data represent biological duplicates. (d) Fluorescence from the PmalT reporter in E. coli K-12 MG1655, cya, and cya crp* through 12 h of growth. Data represent biological triplicates. (e) Promoter layout of (top to bottom) RpoD consensus, the native PmalT sequence from E. coli K-12 MG1655, and hybrid promoters mutants where the −10 box, −35 box, or both are changed to the RpoD consensus (red font). (f) Fluorescence from the native and consensus PmalT reporter variants in E. coli K-12 MG1655 wt and cya after 12 h of growth, with no supplement (Ø, grey), 1% glucose (wt only, blue), or 0.5 mM cAMP (cya only, red). Data represent two or three biological replicates.

Fig 3

Effects of varying the Crp binding site location within the malT promoter. (a) RpoD consensus sequences and the malT promoter with annotated and alternative RpoD recognition sequences and TSS—see main text for further description. The identity between the recognition sites and the consensus sequences is indicated by the numbers above. (b) Fluorescence from the PmalT reporter variants after 12 h of growth. Indicated are the native distances between the Crp-binding site and +1, and the distances classically associated with Crp promoter classes I and II. Data represent two or three biological replicates. (c) Fluorescence from the native PmalT reporter (−70.5) compared to the −60.5 variant in E. coli K-12 MG1655 WT or a cya strain supplemented with 0.5 mM cAMP after 12 h of growth. Significance was based on two-sided unpaired t-tests between the groups indicated and designated as significant if P < 0.05. Data represent biological duplicates.

Fig 4

Expression from malT promoter reporters upon varying the central promoter motifs. (a) Fluorescence (top) and growth (bottom) of E. coli K-12 MG1655 with the native or consensus PmalT reporter variants during 12 h. Promoter variants: Solid line, native; dotted line, −10 consensus; stippled line, −35 consensus; dotted and stippled line, −10 and 35 consensus. Black, lag and log phase; red, stationary phase. (b) Alignment of PmalT to the reported consensus sequences of RpoD, RpoS, and RpoH, with potential recognition sites. RpoS and RpoH both recognize extended −10 boxes, although RpoH often is designated part of the −10 box (38). (c) Fluorescence with standard deviations from the PmalT reporter in E. coli K-12 MG1655 wt and rpoS during 48 h of growth. Solid line, wt; stippled line, rpoS. (d) Fluorescence from the native and −60.5 variant of PmalT in E. coli K-12 MG1655 cya crp pSEVA-crp with various rpoS genotypes after 12 h of growth. Gray, rpoS⁺; red, rpoS^-; blue, rpoS N98K. (e) Fluorescence with standard deviations (top) and growth (bottom) of E. coli K-12 MG1655 with the PmalT* reporter during 18 h. Solid line, native promoter; stippled line, evolved promoter; black lines, wild-type strain; gray lines, cya. Data represent biological triplicates.

Fig 5

Growth-phase dependent regulation of PsdaC. (a) Fluorescence with standard deviation (top) during 20 h of growth (bottom) of E. coli K-12 MG1655 with the PsdaC reporter mutants. Promoter variants: Solid line, native (wt); stippled line, mutant 1; dotted line, mutant 2. (b) Fluorescence with standard deviations from the native PsdaC (left), PsdaC*1 (middle), and PsdaC*2 (right) in E. coli K-12 MG1655 wt (grey) and rpoS (red) through 20 h of growth. Solid line, native (wt); stippled line, mutant 1; dotted line, mutant 2.