OUP user menu

Characterization of a periplasmic peptidyl-prolyl cis-trans isomerase in Erwinia chrysanthemi

Christine Pissavin, Nicole Hugouvieux-Cotte-Pattat
DOI: http://dx.doi.org/10.1111/j.1574-6968.1997.tb12753.x 59-65 First published online: 1 December 1997


The main determinant of the plant pathogen Erwinia chrysanthemi virulence is the production of extracellular enzymes, mainly pectate lyases. Adjacent to a pectate lyase encoding locus, we identified the gene rotA supposed to encode a folding catalyst. Overproduction of the protein and assay of activity using a synthetic substrate, confirmed that rotA encodes a periplasmic peptidyl-prolyl cis-trans isomerase. rotA disruption provokes no change in cell morphology, cell viability, growth rate or stability of the extracellular and periplasmic proteins. In addition, this mutation does not alter the activity of the pectate lyases, their stability in the periplasm during the transitory step of secretion or their recognition by the Out secretory system. rotA expression was followed using a rotA::uidA transcriptional fusion. Some environmental conditions, such as temperature variations and nitrogen starvation, modulate rotA expression. In contrast to the E. coli rotA gene, the E. chrysanthemi rotA possesses only one promoter and is not controlled by the CRP global regulator.

  • Peptidyl-prolyl cis-trans isomerase
  • Erwinia chrysanthemi
  • Protein secretion

1 Introduction

The phytopathogenic enterobacterium, Erwinia chrysanthemi, is the causal agent of soft rot disease on a broad host range. The pathogenesis results from the synthesis of extracellular virulence factors, such as pectinases and cellulases. The main pathogenic determinants are five major pectate lyases which are secreted extracellularly in three steps: (i) translocation through the inner membrane by the sec system, (ii) folding in the periplasm, (iii) translocation through the outer membrane by the Out system. In the periplasm, the presence of folding catalysts plays a key role in the steps of protein folding [1]. In E. chrysanthemi, the disulfide isomerases (Dsb) are required for the correct formation of disulfide bonds in the pectate lyases and cellulases and, consequently, for their recognition by the Out machinery [2, 3]. Another rate-limiting molecular event in the protein folding is the cis-trans isomerization of prolyl peptide bonds. This reaction is catalyzed by peptidyl-prolyl cis-trans isomerases (PPIase). PPIases are ubiquitous proteins which include the FKBPs, the cyclophilins and the parvulin-like proteins [4]. Several PPIases were identified in E. coli: two cyclophilin-like proteins, RotA and RotB [5]; two FKBP related proteins, SlyD and FkpA [68]; and two proteins of the parvulin family, PpiC and SurA [4, 9]. In some pathogenic bacteria, PPIases of the FKBP family are thought to be determinant in pathogenicity [1012]. The function of the PPIases of the cyclophilin family remains to be clarified. In E. coli[13] and Acinetobacter calcoaceticus[14], the disruption of rotA does not lead to a significant alteration of the phenotype. However, in Legionella pneumophila, the inactivation of the cyclophilin-like PPIase Cyp18, affects the intracellular survival of the pathogen in A. castellanii[15].

We identified the rotA gene of E. chrysanthemi at the vicinity of the pelB-pelC-pelZ cluster encoding three pectate lyases. This is the first example of cyclophilin-like protein identified in a phytopathogenic bacterium. To elucidate the role of RotA in E. chrysanthemi, we analyzed the characteristics of a rotA mutant, particularly concerning the putative involvement of RotA in the folding of pectate lyases and in their secretion by the Out system. In addition, construction of a rotA::uidA fusion allowed us to analyze its transcriptional regulation.

2 Materials and methods

2.1 Strains, media and growth conditions

Bacterial strains and plasmids used in this study are listed in Table 1. Cells were usually grown in LB or M63 media at 37°C for E. coli and 30°C for E. chrysanthemi. The derivative media used to test the different environmental conditions have been previously described [16].

View this table:
Table 1

Bacterial strains and plasmids

E. coli
NM522Δ(lac-proAB) Δ(mcrB-hsdSM)5 supE thi (F′, proAB lacIq, lacZΔM15)Laboratory collection
BL21(DE3)Fdcm ompT hsdS gal lonλ(DE3), T7 polymerase gene with the lacUV5 promoterStudier, W.F.
E. chrysanthemi
3937Wild-type strainLaboratory collection
A350lmrTclacZ2Laboratory collection
A2173lmrTclacZ2, rotA::uidA, Km®This work
A2712rotA::uidA, Km®This work
ISADoutD::Ω, Cm®Barras, F.
A2443outD::Ω, rotA::uidA, Cm®, Km®This work
pT7-5Φ10 promoter, Ap®Tabor, S.
pT7-6Φ10 promoter, Ap®Tabor, S.
pTPZpT7-6 derivative with a 2.3 kb NheI-EcoRI, pelZ+, Ap®[18]
pTRTpT7-5 derivative with a 2.3 kb NheI-EcoRI, rotA+, Ap®This work
pTRTPpT7-5 derivative with the 0.9 kb PstI-EcoRI, rotA+, Ap®This work
pC26pBluescript derivative with a 2.8 kb NruI-HindIII fragment, rotA+, Cm®This work
pC262insertion of uidA-Km in pC26, rotA::uidA, Cm®, Km®This work

2.2 Enzyme assays

Enzyme assays were usually performed on toluenized cultures after 24 h of growth. Pectate lyase and β-glucuronidase activities were determined spectrophotometrically [17, 18]. PPIase activity was assayed using 1 ml reaction mixture containing 50 mM HEPES pH 7.8, 200 μg ml−1α-chymotrypsin, 0.05 mM Succinyl-Ala-Ala-Pro-Phe-p-nitroanilide. The synthetic peptide is present in two forms: almost 80% of the most stable form, trans, and 20% of the cis form, which is the specific substrate of the PPIases. The α-chymotrypsin cleaves the trans isomer leading to the liberation of colored Phe-p-nitroanilide whose appearance was followed at 390 nm, for 2 min, at room temperature.

2.3 Cell fractionation and analytical procedures

The extracellular proteins of E. chrysanthemi were obtained in the supernatant after centrifugation. The periplasmic fraction was obtained after sphaeroplast formation resulting from lysozyme action combined with osmotic shock [2]. After separation by isoelectrofocusing, the pectate lyase isoenzymes were identified by a specific activity detection [18]. Pulse chase experiments followed by cell fractionation and proteinase K digestion were performed as previously described [13].

2.4 Molecular biology techniques

Preparations of plasmid DNA, restriction digestions, ligations, DNA electrophoresis and transformations were carried out as previously described [19]. To construct the rotA::uidA fusion, the uidA-Km cassette obtained by EcoRI digestion of the pUIDK3 plasmid [17] was introduced into the EcoRI site of the rotA gene. Only the orientation giving rise to β-glucuronidase activity was retained. The resulting plasmid pC262 was introduced in E. chrysanthemi cells by electroporation and a marker-exchange recombination was selected after successive cultures in low phosphate medium. Nucleotide sequence analysis was performed by the chain-termination method on double-stranded DNA template. The nucleotide sequence of the NarI-RsrII fragment will appear in EMBL, Genbank and DDBJ Nucleotide Sequence Databases under the accession number Y09804.

For Southern blotting, genomic DNA was extracted and transferred onto a nylon membrane [19]. The rotA probe was a 1.4-kb AflIII-SalI fragment labelled with digoxygenin-dUTP. The hybridization was realized in 50% formamide at 42°C. The membrane was washed twice in a 2×SSC, 0.1% SDS, at room temperature, and in 0.1×SSC, 0.1% SDS at 68°C. The DIG luminescent kit of Boehringer-Mannheim was used for detection. For the primer extension experiment, total RNAs were extracted using frozen phenol [18]. The 5′-end of the rotA mRNA was determined using a synthetic 27-mer oligonucleotide (Eurogenetec) 5′GGTGAAAATGCGGTCAGCGAGAGTACG3′, which is complementary to nucleotides +59 to +33 of the rotA ORF. The oligonucleotide was 5′-end-labelled with T4 polynucleotide kinase using [γ-32P]-ATP. After hybridization of the primer with total RNAs, the extension reaction was performed using avian myeloblastosis virus reverse transcriptase [19].

3 Results and discussion

3.1 Identification of the rotA gene

Insertion of the 2.3-kb EcoRI-NheI fragment adjacent to the pelB-pelC cluster (Fig. 1) in pT7-5 and pT7-6 vectors showed that in one orientation (pTPZ), it directed the production of the pectate lyase PelZ [18] while in the opposite direction (pTRT), it encodes a 19-kDa protein. This periplasmic protein is also synthesized from the 0.9-kb EcoRI-PstI fragment (pTRTP) (Fig. 1). Determination of the nucleotide sequence of the 858 nt NarI-RsrII fragment revealed the presence of an ORF which begins with an ATG at position 186 (preceded by the putative ribosome binding site AGGAA), and ends with a TAA codon at position 756. The predicted molecular mass of the corresponding protein is 20 565 Da. Comparison with sequence databases indicated that this protein is homologous to PPIases of the cyclophilin family (Fig. 2), particularly the periplasmic E. coli PPIase, RotA (77.2% of identity). Thus, the E. chrysanthemi gene adjacent to pelZ and transcribed in the opposite direction was named rotA. The N-terminal part of RotA presents the characteristics of signal sequences with a putative peptidase I cleavage site A21-X-A23/A24, conserved in E. coli and E. chrysanthemi (Fig. 2). The E. chrysanthemi RotA shares higher homology with the PPIases of Gram-negative bacteria than with those of Gram-positive bacteria (Fig. 2). In cyclophilins, one of the determinants of the cyclosporin A binding is the presence of a Trp residue [20]. Similarly to the other cyclophilins of Gram-negative bacteria, the E. chrysanthemi RotA may not be inhibited by cyclosporin A since it possesses a Phe residue at the corresponding position (Fig. 2). The E. chrysanthemi rotA gene is followed by a G-C rich inverted region (nt 801 to 827) followed by a stretch of T residues, typical of a Rho-independent transcriptional terminator (Fig. 1).

Figure 1

Genetic organization of the pelB-pelC-pelZ-rotA region. The arrows indicate the transcription orientation of the genes. The different plasmids are shown with the plasmid promoters indicated by plac, pΦ10. The location of the insertion of the uidA-Km cassette is indicated.

Figure 2

Alignment of E. chrysanthemi RotA with prokaryotic cyclophilin-like proteins. The proteins represented are E. coli RotA (Swiss-Prot accession number P20752), E. coli RotB (Swiss-Prot accession number P23869), A. calcoaceticus RotA (Swiss-Prot accession number P42693), S. typhimurium putative PPIase (partial sequence) (Swiss-Prot accession number P20753), H. influenzae putative PPIase (Swiss-Prot accession number P44499), L. pneumophila Cyp18 (EMBL accession number X83769), B. subtilis PPIB (Swiss-Prot accession number P35137), S. chrysomallus PPIase (Swiss-Prot accession number Q06118), Synechococcus sp. putative PPIase (Swiss-Prot accession number P29820). The potential cleavage site of the signal peptide is indicated with an arrow. The residues conserved in at least 70% of the proteins are shaded. The amino acid corresponding to the Trp residue involved in the cyclosporin A binding of human cyclophilin, is boxed.

3.2 Analysis of the E. chrysanthemi rotA mutant

Inactivation of the rotA gene by insertion of a uidA-Km cassette (strain A2173) (Fig. 1) was verified by Southern blotting experiment. In the wild-type strain, only one rotA hybridizing band was visualized, suggesting that rotA is a monocopy gene (data not shown). The PPIase activity detected in crude extracts and in periplasmic fractions of A2173 corresponded to 10% and 7%, respectively, of the activities observed in the parental strain A350. The low level of PPIase activity observed in the rotA mutant, confirmed the function of the RotA protein but suggested the presence of other PPIase(s) in E. chrysanthemi. The rotA mutation does not affect the growth rate or the viability of the cells, the mobility of the bacteria, the sensitivity to bacteriophages φEC2 or Mu, and the stability of periplasmic or extracellular proteins as shown using pulse chase experiments and proteinase K digestion.

A prolyl peptidic bond in the unusual cis conformation is present near the putative catalytic site of PelC and this Pro residue is conserved among all the members of the extracellular pectate lyase superfamily [21]. However, no difference was observed in the total pectate lyase activity of strains A350 and A2173 and isoelectrofocusing showed that the pI of the five major pectate lyases (PelA to PelE) is not modified in the rotA mutant (data not shown). These enzymes are secreted in strain A350 or A2173 while they are retained in the periplasm of the outD mutant ISAD. After electrofocusing, the pectate lyase profiles of outD and outD-rotA mutants were similar (data not shown). In conclusion, the RotA inactivation does not affect (i) the pectate lyase activity, (ii) the pectate lyase surface charges, (iii) the stability of pectate lyases accumulated in the periplasm of an outD mutant, (iv) the recognition of the pectate lyases by the Out system, and (v) the functionality of the Out machinery. Thus, the RotA action is either not essential or is compensated for by the presence of other periplasmic PPIases.

3.3 Transcription of the rotA gene

Little is known about the regulatory factors controlling the synthesis of the prokaryotic cyclophilin-like PPIases. Such studies could contribute to clarify their function in the cell. The rotA expression was analyzed after growth of E. chrysanthemi under different environmental conditions using a rotA::uidA transcriptional fusion. The rotA expression is constant whatever the growth phase (data not shown) and, consequently, it is not correlated with the cellular development. Increased temperature represents one condition in which PPIases may be necessary to catalyze limiting steps in the refolding of denatured proteins. In comparison with rotA expression at 30°C, we noticed a 50% increase at 25°C and a 40% decrease at 37°C (Table 2), suggesting that the rotA expression is sensitive to temperature variations. However, rotA is not induced under heat-shock conditions (Table 2) such as observed in E. coli[22]. Nitrogen starvation led to a three-fold decrease in rotA expression (Table 2) but phosphate starvation or increasing the medium osmolarity had no effect (data not shown). Although it is adjacent to a pel locus, the rotA expression is not inducible in the presence of polygalacturonate. A two-fold decrease in rotA transcription was observed in rich medium. However, the rotA gene appeared to be insensitive to the catabolite repression exerted by glucose since no difference was observed in minimal medium, with either glucose or glycerol as the sole carbon source. In addition, inactivation of the crp gene did not affect the rotA expression (data not shown). The E. chrysanthemi rotA is not regulated by CRP, in correlation with the absence of a CRP binding site in the promoter region.

View this table:
Table 2

Expression of the rotA::uidA fusion in different culture conditions

MediumCulture conditionsβ-Glucuronidase activitya
Heat shockb45°C, 20 min167±20
Heat shock10% ethanol, 20 min227±10
Glycerolnitrogen starvation82±4
Rich medium30°C103±16
  • aβ-glucuronidase activity is given as nmol of product per min per mg of bacterial dry weight. Each experiment was repeated three times and the average value is given with the standard error.

  • bHeat shock was performed at 45°C or by addition of 10% ethanol during 20 min.

To identify the rotA transcription start, a primer extension reaction was performed using total mRNAs extracted from the E. chrysanthemi strain A350 and from E. coli NM522/pC26. The 5′-end of the rotA mRNAs (Fig. 3) was similar for mRNA extracted from each strain, indicating that the E. chrysanthemi rotA promoter is recognized by the E. coli RNA polymerase. A sole transcriptional start was observed. It is preceded by a classical σ70 promoter with −10 (TAAAAT) and −35 (GTGGCG) boxes (Fig. 3). The E. coli rotA gene is expressed from four overlapping promoters among which P2, P3 and P4 are activated by CRP whilst the P1 promoter is not submitted to this regulation [22]. In contrast, the transcription of the E. chrysanthemi rotA gene is initiated from only one promoter, presenting a high homology with the P1 promoter of E. coli (Fig. 3). In this respect, E. chrysanthemi appears to be comparable with A. calcoaceticus where a single active promoter was identified [14]. Although the rotA genes of E. chrysanthemi and E. coli share strong homology, they are not submitted to the same transcriptional controls. This difference could reflect distinct roles of RotA in these two enterobacteria.

Figure 3

Characterization of the rotA promoter. A: Identification of the transcription initiation point by primer extension. Lane 1: Primer extension reaction product. Lanes A, C, G, T: DNA sequence using the same primer. B: Alignment of E. chrysanthemi (E. ch) and E. coli rotA promoters. The double points indicate identical bases. The Shine Dalgarno, −10 and −35 boxes are underlined. The transcription initiation points are indicated by arrows. The CRP binding sites are boxed.


We acknowledge J. Robert-Baudouy and the team of the laboratory for their interest in this work, V. James for reading the manuscript and F. Barras for providing strains and for helpful discussions. This work was supported by grants from the CNRS (UMR 5577), the DRED and an Action Concertées Coordonnées Sciences du Vivant 6.


View Abstract