OUP user menu

The PecM protein is necessary for the DNA-binding capacity of the PecS repressor, one of the regulators of virulence-factor synthesis in Erwinia chrysanthemi

Thierry Praillet , Sylvie Reverchon , Janine Robert-Baudouy , William Nasser
DOI: http://dx.doi.org/10.1111/j.1574-6968.1997.tb12654.x 265-270 First published online: 1 September 1997


The pecS regulatory locus is responsible for the down-expression of many virulence genes in Erwinia chrysanthemi. This locus consists of two genes, pecS and pecM, divergently transcribed. Genetic evidence indicates that the PecM protein modulates the regulatory activity of PecS. Purification and characterization of PecS, expressed either from E. coli, from the wild-type E. chrysanthemi strain or from a pecM mutant, showed that the PecS protein produced in these three genetic backgrounds displays the same biochemical properties. Band-shift assay analysis with the three PecS isoforms confirmed the involvement of the PecM protein in modulating the PecS DNA-binding capacity. Moreover, determination of the Kdapp for operator regions of the PecS protein, produced either by the wild-type E. chrysanthemi or by E. coli, reveals similar affinities. Thus, in E. coli, there is likely to be at least one other PecM-like protein able to cross-react with the E. chrysanthemi PecS protein.

  • Erwinia chrysanthemi
  • Pectinolysis
  • Transcription
  • Protein interaction

1 Introduction

The phytopathogenicity of the enterobacterium Erwinia chrysanthemi is mainly due to its ability to produce and secrete a large array of depolymerizing enzymes (pectate lyases, cellulases, nucleases, polygalacturonase, pectin lyase, pectin esterases, …) [1]. Among these enzymes, the pectate lyases constitute the major virulence factors [2]. These enzymes allow the bacterium to grow on pectin as the sole carbon source.

This metabolic pathway – designated as pectinolysis – is regulated by a complex system of interconnected regulatory networks. Until now only one positive regulator, the CRP protein (cAMP receptor protein), has been identified [3] whereas at least three negative regulators (KdgR, PecT and PecS) have been characterized [46]. The KdgR protein is the main regulator mediating the induction by pectic compounds through its direct interaction with the regulatory regions of the target genes [7, 8]. The PecT protein is a LysR-like transcriptional regulator which controls the expression of the pectate lyase genes as well as the expression of the genes involved in the biosynthesis of exopolysaccharides. However, the nature of the PecT responsive signal has not yet been determined [6].

The third locus to be implicated in the down-expression of virulence factors in E. chrysanthemi is the pecS locus. It is composed of two divergently transcribed genes, pecS and pecM. The PecS protein belongs to the MarR family, which consists of bacterial proteins involved in transcriptional regulation in response to the presence of phenolic compounds [9]. PecM is an inner membrane protein which could be involved in the transduction of an extracellular signal [5]. Mutations, either in pecS or in pecM, lead to a similar phenotype, suggesting their involvement in the same regulatory network [5]. Purification and functional characterization of the PecS protein have shown that, in vitro, the PecS protein is able to specifically interact with the regulatory regions of the genes whose expression is controlled by PecS in vivo [10]. Moreover, the pecS gene is subjected to negative auto-regulation [11]. Further analysis of this auto-regulation has shown that the potency of the PecS repressor is dependent on the presence of the PecM protein. Thus, it seems that PecM is able to potentiate the PecS repressor activity in vivo [11].

In this paper, we describe the biochemical comparative analysis and functional characterization of the PecS isoforms purified from E. coli and from E. chrysanthemi strains harbouring a functional or mutated pecM gene.

2 Material and methods

2.1 Bacterial strains

The bacterial strains used in this study are the E. chrysanthemi parental strain A350 (laboratory collection), its pecM derivative (A2097) [5] and a PecS-overproducing E. coli strain (I1852) [10].

The E. chrysanthemi strains were grown in M63 synthetic medium supplemented with 2 g l−1 glycerol (M63Y). The I1852 E. coli strain was grown in pT7 medium described by Tabor and Richardson [12]. When required, antibiotics were added to obtain a final concentration of 50 μg ml−1.

2.2 Purification of the PecS isoforms

The PecS protein was overproduced in the E. coli I1852 strain, as previously described [10]. The E. chrysanthemi strains were grown in M63Y medium until the OD600 reached the value of 1.

Cells were then rinsed twice with lysis buffer (4 mM Tris-HCl pH 7.4, 12 mM Hepes-NaOH pH 7.4, 75 mM KCl, 1 mM DTT, 1 mM PMSF, 10% glycerol, 2 mM EDTA), and resuspended in 1/20 of the culture volume in the same buffer. Cells were disrupted in a French press exerting a pressure of 138 000 kPa. Residual insoluble material was eliminated by two centrifugation steps at 12 000 rpm for 15 min. Total cellular protein extract was then submitted to ammonium sulfate fractionation, using the following saturation conditions: 0–40%; 40–55%; 55–80% and 80–100%. Flocculated proteins at each stage were recovered by a 30 min centrifugation at 12 000 rpm. The protein pellets were resuspended in 5 ml of lysis buffer and dialysed three times against 2 l of the same buffer without KCl.

The PecS protein was identified in the different ammonium sulfate saturated fractions by immunoblots and purified in a two-step chromatography method, including separation on DEAE and Heparin columns as previously described [10, 11].

2.3 Structural characterization of the PecS isoforms

The structural characteristics of the PecS isoforms were determined by non-denaturing polyacrylamide gel electrophoresis, classical SDS-PAGE analysis or isoelectrofocusing, as previously described [10]. The native molecular mass was determined by gel filtration experiments using a standardized Superose 12 HR 10/30 column (Pharmacia) with a buffer containing 50 mM Tris-HCl pH 7.4 and 200 mM NaCl.

2.4 Band-shift assays

Probe labelling and purification, band-shift experiments and apparent dissociation constants (Kdapp) determination were carried out as previously described [10].

3 Results and discussion

3.1 Purification of the different PecS protein isoforms

Western blot analysis indicated that PecS, expressed in the I1852 E. coli strain, in the E. chrysanthemi A350 parental strain or in the A2097 pecM mutant, was precipitated under 55–80% ammonium sulfate saturating conditions. Purification of the PecS repressor from these ammonium sulfate precipitated fractions gave, per liter of culture, up to 6 mg of PecS protein from the E. coli strain and about 200 and 50 μg of PecS protein from the pecM and parental strain, respectively. The amount of PecS protein obtained from the two E. chrysanthemi strains was in accordance with our previous results. Indeed, the PecS cellular content in the parental strain is estimated to be about 100 monomers in the stationary phase and the PecS protein is about five times more abundant in the pecM mutant than in the A350 parental strain [10, 11].

3.2 Biochemical characterization of the different PecS isoforms

Preliminary comparisons of the PecS proteins purified from the PecS-overproducing E. coli strain (PecSEc) or from the pecM or pecM+ E. chrysanthemi strains (respectively PecSEchrM− and PecSEchrM+) show that these PecS isoforms present, essentially, the same biochemical characteristics (Table 1). However, the saline concentration needed to elute the PecS repressor produced by E. chrysanthemi from a DEAE column is approximately twice that required for PecS synthesized in E. coli (Table 1). SDS-PAGE, isoelectrofocusing and conventional polyacrylamide gel electrophoresis in non-denaturing conditions revealed that the PecS protein purified from E. chrysanthemi and E. coli displayed identical monomeric molecular mass and the same global charge (Fig. 1). Moreover, gel filtration experiments revealed that the three isoforms have a native molecular mass of about 40 kDa (Table 1). Any modification of methylation or removal of a small peptide by proteolytic attack was not revealed by our experimental approach. Such modification will be further investigated using mass spectrometry. The differential behaviour of the PecS isoforms produced in E. coli and in E. chrysanthemi could be attributed to a different protein environment in these two genetic backgrounds. Indeed, it is reasonable to imagine that a weak interaction between PecS and another protein, not yet identified, could temporarily modify the PecS global charge leading to the observed differences in the salt concentrations required for elution of PecSEc and PecSEchr from the DEAE column.

View this table:

Biochemical characteristics of the PecS isoforms

Ammonium sulfate fractionated precipitationa50–80%50–80%50–80%
DEAE elution saline concentrationsKCl 200 mMKCl 375 mMKCl 375 mM
TSK elution saline concentrationsKCl 425 mMKCl 425 mMKC 425 mM
Gel filtrationb40 kDa40 kDa40 kDa
  • PecS proteins were obtained from the I1852 E. coli PecS-overproducing strain, from the parental A350 E. chrysanthemi strain or from the A2097 E. chrysanthemi pecM derivative mutant. aThe PecS proteins extracted from the different strains flocculate in the indicated ammonium sulfate saturated fraction. bValues reported in this row correspond to native molecular masses determined in gel filtration experiments.

Figure 1

Determination of the global charge and the monomeric molecular mass of the different PecS isoforms. The different PecS isoforms extracted from the E. coli I1852 PecS-overproducing strain, from the parental A350 E. chrysanthemi strain or from the A2097 pecM mutant were loaded onto a native gel (A), onto a 17.5% acrylamide SDS-PAGE (B) or onto an isoelectrofocusing gel (C). After electrophoretic separation, PecS proteins were specifically revealed by immunoblots using anti-PecS antibodies [10]. Lane 1: 10 μg of the I1852 strain before PecS overproduction; lane 2: 10 μg of crude extract from the I1852 strain after PecS overproduction; lane 3 and 4: 5 and 10 μg of PecSEc purified protein respectively; lane 5: crude extract from the E. chrysanthemi A350 parental strain; lane 6: crude extracts from an E. chrysanthemi pecS mutant; lane 7: crude extract of the E. chrysanthemi pecM mutant.

3.3 Functional characterization of the PecS isoforms

In our previous work we showed, by a genetic approach, that the presence of the PecM protein is necessary for the full repressor activity of PecS. In order to elucidate whether this observation is related to a modification in the PecS DNA-binding ability, we performed band-shift assays with the different purified PecS isoforms. For this purpose, a 32P-labelled Nrul-SalI restriction fragment, harbouring the celZ regulatory region, was incubated with equivalent and increasing concentrations of the PecS purified isoforms (Fig. 2). PecSEchrM+ displayed a behaviour similar to PecSEc. At low PecS concentrations, only one complex (the C1 complex), corresponding to the binding of one PecS dimer, could be detected. Increasing the PecS concentration led to the formation of the C2 complex due to the binding of two PecS dimers. At high PecS concentration, a C3 complex with a highly retarded mobility could be detected (Fig. 2). The determination of the apparent affinity constants (Kdapp) [11] revealed that PecSEc and PecSEchrM+ display similar affinities for the celZ operator (about 4.0±0.8 nM). In contrast, the PecS protein purified from the pecM mutant displayed a very weak binding activity towards the celZ regulatory region, even under saturating PecS concentrations. Thus, the PecM protein appears to be necessary for the potentiation of the PecS DNA-binding activity. Since biochemical characterization has established that the PecS proteins purified from the E. chrysanthemi A350 parental strain, or its pecM derivative, display no difference in their global charge, molecular mass or dimerisation, it is reasonable to assume that the PecM protein does not potentiate the activation of PecS by significant proteolytic cleavage, the covalent binding of a charged radical (phosphate, adenylate, …) or by modification of the native form of the PecS protein. So, the modalities of the potentiation exerted by PecM on the PecS protein remain unknown.

Figure 2

Protein-DNA interactions obtained for the different PecS isoforms. PecS isoforms purified from E. coli, from the A350 E. chrysanthemi strain and from the E. chrysanthemi pecM mutant were obtained by ammonium sulfate fractionated precipitation followed by two chromatographic steps. Equivalent quantities of protein were then tested in mobility-shift assays with the celZ regulatory region. lane 1: 2 μM; lane 2: 1.3 μM; lane 3: 128 nM; lane 4: 64 nM; lane 5: 16 nM; lane 6: 8 nM; lane 7: no PecS protein. Cl, C2, C3: different DNA-protein complexes. F: free DNA probe.

As the PecSEchrM+ and the PecSEc proteins have similar affinities for DNA, there should be a PecM-like protein in E. coli able to cross-react with the E. chrysanthemi PecS protein. A search for PecM homologs in the data base reveals the existence of three proteins in E. coli, namely YdeD, YicL and YijE, that share significant similarity with PecM (25%). All these proteins, with unknown function, display ten transmembrane segments and a small hydrophilic region located at the C-terminal end. This hydrophilic region, characterized by a high proportion of basic amino-acids (arginine and lysine) (Fig. 3), could be involved in the modification of a specific regulator belonging to the PecS-MarR family. The ydeD gene is located in the mar locus and could define a regulatory couple with the marR gene. The potentiation of PecS protein in E. coli may be exerted by one of the PecM homologs detected in this bacteria. The PecS/PecM couple, like the MarR/YdeD proteins, could define a new type of regulatory couple.

Figure 3

Comparison of the hydropathy patterns and the amino-acid sequences of PecM homologs. Hydropathy was calculated according to Kyte and Doolitle [13] with a span of 9 amino-acids. The portions above the midpoint line indicate hydrophobic regions. For amino-acid alignment, gaps were introduced to maximize the homology. The consensus given indicates residues (bold letters) common to all four proteins or present in at least three proteins.

PecS potentiation by PecM may require specific protein-protein interaction, such as in the classical two-component systems. The PecS responsive signal must interfere with PecM conformation to initiate signal transduction to PecS. However, the intervention of an additional intermediate in this regulatory cascade is also a possibility. Characterization by in vitro and in vivo approaches of the PecS/PecM interactions, and of the potential cross-talking between PecS/PecM and the MarR/YdeD regulatory couples, needs to be investigated to give a more acute insight into the modalities of action of this new class of regulatory system.


We would like to thank N. Thomson, V. James, G. Condemine and N. Hugouvieux-Cotte-Pattat for their critical reading of the manuscript. This study was supported by grants from the Centre National de la Recherche Scientifique, from the Action Concertée Coordonnée-Sciences du Vivant 6 (ACC-SV6) n° 9506111 and from the Ministère de l'Education Nationale, de l'Enseignement Supérieur, de la Recherche et de l'lnsertion Professionnelle.


View Abstract