Improved System for Construction and Analysis of Single-Copy {beta}-Galactosidase Operon Fusions in Yersinia enterocolitica

We report a significantly improved system for studying single-copylacZ operon fusions in Yersinia enterocolitica: a simple procedurefor the stable integration of lacZ operon fusions into the aralocus and a strain with a deletion mutation that abolishes thelow level of endogenous ß-galactosidase activity.

INTRODUCTION

Top

Introduction

Yersinia enterocolitica is a cause of foodborne human gastroenteritisand a favored model organism for studying bacterial virulence(2). Y. enterocolitica virulence gene regulation is studiedin many laboratories. This is especially true for the genesencoding the Ysc type III secretion system, which Y. enterocoliticashares in common with the other pathogenic Yersinia species,Y. pestis and Y. pseudotuberculosis (5).

In Escherichia coli K-12, the ability to construct single-copylacZ operon fusions greatly facilitates regulatory studies.Single-copy lacZ fusions overcome the problems of multicopyfusions, such as copy number variation and titration of transcriptionfactors. The only existing method for constructing single-copylacZ fusions in Y. enterocolitica relies on complete chromosomalintegration of a suicide plasmid by homologous recombination(see, e.g., reference 6). This approach has many limitations,including the following. (i) The integrated plasmid encodesan antibiotic resistance no longer available for subsequentmanipulations. (ii) Chromosomal integration is driven by thepromoter fragment, which limits the minimum size of the fragmentthat can be studied. (iii) Studying promoters internal to anoperon results in disruption of the operon after suicide plasmidintegration. (iv) Integration of fragments with promoter mutationsmay be tedious to verify (homologous recombination can occureither upstream or downstream of the mutation, and only theformer places the mutation upstream of lacZ). The use of lacZfusions in Y. enterocolitica is further compromised because,although it is known as a Lac^– species, there is a tracelevel of endogenous ß-galactosidase activity thatcauses the formation of pale blue colonies on agar containingX-Gal (5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside).This limits the use of X-Gal media for phenotypic screens andthe analysis of weakly expressed promoters in enzyme assays.These limitations motivated our development of an improved systemfor the construction and analysis of single-copy lacZ operonfusions in Y. enterocolitica.

Construction of ara integration plasmid pAJD905.

Top

Construction of ara integration...

We followed a strategy similar to that employed by Wu et al.,who targeted the rha locus of Klebsiella oxytoca for lacZ operonfusion integration (21). They used a suicide plasmid with anrpsL⁺ (streptomycin sensitivity) allele for counter selection,which requires the use of a strain with an rpsL mutation conferringstreptomycin resistance. Streptomycin resistance is a widelyused marker in Y. enterocolitica, which is naturally ampicillinresistant. Therefore, we chose a different approach, based onthe established technology of a suicide plasmid with a ${pi}$ protein-dependentR6K replication origin and the use of the sacB gene for selectionof plasmid-free segregants as sucrose-resistant colonies. Weextended this technology to develop a Y. enterocolitica site-specificintegration lacZY operon fusion vector, with several featuresthat increase its utility.

For the chromosomal integration site we chose the araFGHC-araBAlocus of a Y. enterocolitica strain 8081 derivative, which isthe highly pathogenic strain used in the genome sequencing project.Disruption of this locus abolishes arabinose catabolism, whichallows integrants to be easily identified by their Ara^–phenotype (e.g., formation of white colonies on MacConkey-arabinoseagar plates). It also protects the lacZY operon fusions fromflanking promoters by placing them between divergently transcribedoperons which have had their promoters deleted (10).

The allelic exchange vector pRE112 (Table 1) has a R6K replicationorigin and the sacB1 gene, which confers sucrose sensitivity(9). Two 1.5-kb fragments ('araGH' and 'araBA') were amplifiedby PCR and ligated into the pRE112 polylinker with a uniqueNotI site between them (pAJD898). Next, plasmid pAJD768 wasconstructed by joining the $~$ 7-kb BamHI lacZYA-cat fragment ofpFUSE to the $~$ 3.1-kb BglII pSC101 ori fragment of pWSK29 andthen inserting the StuI-SmaI pSL1180 polylinker upstream oflacZ. Finally, the polylinker-lacZY region of pAJD768 was amplifiedwith primers that incorporate E. coli rrnBT1 and rrnBT2 terminatorsupstream and downstream, respectively. This fragment was clonedinto the NotI site of plasmid pAJD898. Derivatives with thelacZY operon in either orientation were isolated. However, oneof these orientations produced somewhat higher basal lacZY expressionthan the other (data not shown). Therefore, a plasmid with thelacZY operon in the orientation that gave the lowest basal lacZYexpression was chosen. This was pAJD905, the ara-integrationlacZY operon fusion vector (Fig. 1A). The rrnB terminators furtherinsulate lacZY fusions from the influence of external chromosomalregions.

View this table:
[in this window]
[in a new window]
TABLE 1. Strains and plasmids

View larger version (22K):
[in this window]
[in a new window]
FIG. 1. Integration of lacZY operon fusions into the ara locus. (A) Partial restriction map of plasmid pAJD905. Restriction sites shown in the polylinker are unique (note that MfeI is compatible with EcoRI). (B) Summary of lacZY fusion integration procedure (see text for experimental details). Plasmid pAJD905 is integrated into the ara locus by homologous recombination between either the 'araGH' sequences or the 'araBA' sequences. The structures of both possible integrants are shown as structures 1 and 2. Next, sucrose-resistant segregants lose the integrated pAJD905 backbone by homologous recombination between the tandemly duplicated ara sequences. This either regenerates the ara⁺ locus or leaves the lacZY fusion on the chromosome, shown as structure 3. (C) Colony PCR analysis of the ara⁺ parental strain, integrants with structure 1 or 2, and segregant 3 is shown. In this case the lacZY operon from pAJD905 without a promoter cloned upstream was used. Binding sites of each primer are shown in panel B as w, x, y, and z (see also Table 2). For the ara⁺ strain, primers w and z generate a single product of $~$ 6 kb that is only weakly amplified (primers x and y do not anneal). Integrants 1 and 2 give only one PCR product, as indicated in panel B (the second pair of primers are too far apart for amplification). The desired segregant gives two PCR products as shown. Note that a promoter fragment inserted upstream of the lacZY operon would increase the size of the 1.8-kb product. M, DNA marker (kilobase sizes of some marker bands are indicated). The gel lane images have been rearranged from the original order, but the data are from the same gel and a single experiment.

Chromosomal integration of lacZY operon fusions.

Top

Chromosomal integration of laczy...

The method used to exchange lacZY operon fusions from pAJD905into the ara locus is summarized in Fig. 1B. First, pAJD905(or a derivative with a promoter insert) was transferred fromE. coli S17-1 ${lambda}$ pir (18) to Y. enterocolitica by conjugation.Plasmid integrants were identified by their Cm^r Ara^– phenotypeon MacConkey-arabinose agar (the 1.5-kb regions flanking lacZYensure that most integrations occur at ara). Segregants thenlost the integrated plasmid by homologous recombination betweenthe tandemly duplicated ara sequences. They were selected bystreaking integrants on LB agar containing 10% sucrose (lossof the plasmid sacB1 gene restores a sucrose-resistant phenotype).The sucrose-resistant colonies were replica plated onto MacConkey-arabinoseagar, which identified both Ara⁺ and Ara^– segregants asexpected. The authenticity of the desired Ara^– segregantswas confirmed by their Cm^s phenotype. Finally, the structureof segregants was confirmed by colony PCR analysis using anExpand PCR system (Roche; see Fig. 1C).

Application of the pAJD905 system.

Top

Application of the pajd905...

We tested the new lacZY fusion system with the pspA promoter(studied in our laboratory) and the yopH promoter (studied invarious laboratories). Approximately 500-bp control region fragmentswere amplified by PCR, confirmed by DNA sequencing, and ligatedbetween the XbaI and BglII sites of pAJD905. The ${Phi}$ (pspA-lacZ)and ${Phi}$ (yopH-lacZ) fusions were integrated into the ara locus asdescribed above (data not shown). In both cases, the initialintegration step occurred in the ara locus rather than the yopHor pspA locus in over 99% of cases (initially identified byan Ara^– phenotype). We have also done other studies withan approximately 2.2-kb fragment cloned into pAJD905. In thiscase the frequency of integration into the ara locus did decreasesomewhat. However, the majority of integrations were still inthe ara locus (data not shown).

Previous studies have shown that pspA promoter expression isinduced by the YscC and YsaC secretins, with YsaC being themost potent (6, 11, 15). The ${Delta}$ araGFB::[ ${Phi}$ (pspA-lacZY)] fusion demonstratedexactly the same pattern of regulation (Fig. 2A). The virulenceplasmid (pYV)-encoded yopH gene is induced by a low calciumconcentration at 37°C, and its expression requires the VirFtranscription activator (4, 13). Consistent with this, expressionof the ${Delta}$ araGFB::[ ${Phi}$ (yopH-lacZY)] fusion was induced approximately30-fold by these growth conditions (Fig. 2B). Furthermore, ina strain lacking pYV (and, therefore, VirF), ${Phi}$ (yopH-lacZY) expressionwas essentially abolished (7 Miller units; data not shown).

View larger version (21K):
[in this window]
[in a new window]
FIG. 2. Regulation of ${Delta}$ araGFB::[ ${Phi}$ (pspA-lacZY)] and ${Delta}$ araGFB::[ ${Phi}$ (yopH-lacZY)] fusions. (A) ${Phi}$ (pspA-lacZY) expression. Plasmid pVLT33 (–), pAJD509 (yscC), or pAJD554 (ysaC) encoding the indicated genes expressed from the tac promoter were transferred into a pYV-negative derivative of strain AJD957. Strains were grown exactly as described previously (11). (B) ${Phi}$ (yopH-lacZ) expression. Strain AJD1022 was grown at 26°C or 37°C in low- or high-calcium media as described previously (6). ß-Galactosidase activities were determined as described previously (14) and are expressed in Miller units (each value is the average of measurements from three independent cultures). Sp. Act. = specific activity.

These results demonstrate the functionality of the new lacZYfusion system. The ability to construct stable lacZY fusionswith pYV-encoded promoters is particularly important. The alternativeapproach of integrating suicide plasmids into the relativelysmall pYV could affect the stability or copy number of pYV orthe expression of pYV-encoded genes.

Elimination of Y. enterocolitica endogenous ß-galactosidase activity.

Top

Elimination of y. enterocolitica...

Y. enterocolitica is phenotypically Lac^– on MacConkey-lactoseagar but has a low level of ß-galactosidase activityin enzyme assays and forms blue colonies on X-Gal media. TheY. enterocolitica genome encodes a protein (YE2592) approximately60% identical to E. coli LacZ (data not shown). To determinewhether this gene is responsible for endogenous ß-galactosidaseactivity we constructed a strain with a YE2592 in-frame deletionmutation. The mutation was made using the sacB⁺ suicide plasmidpSR47S with a ${Delta}$ YE2592 insert, which was generated by PCR. Plasmidconstruction and mutagenesis procedures were identical to thosedescribed by others (19).

In enzyme assays, the YE2592⁺ Y. enterocolitica strain produced2 to 4 Miller units of ß-galactosidase activity, dependingon growth conditions, whereas the ${Delta}$ YE2592 mutant had no detectableactivity (data not shown). Furthermore, whereas the YE2592⁺strain was blue on agar containing X-Gal, the ${Delta}$ YE2592 mutantwas white (data not shown). These results confirm that YE2592is responsible for Y. enterocolitica endogenous ß-galactosidaseactivity. The ${Delta}$ YE2592 mutant strain provides a significant enhancementfor the use of ß-galactosidase enzyme assays and X-Galmedia in screens and to analyze weakly expressed lacZ operonfusions.

Use of the pAJD905 system with different Y. enterocolitica strains.

Top

Use of the pajd905...

There are different biotypes and serotypes of Y. enterocolitica(2). In particular, there are two groups of strains known asAmerican and non-American, which differ in their virulence properties(American strains are more virulent) (3). We were interestedto know whether plasmid pAJD905 could be used in different Y.enterocolitica strains, in addition to our strain 8081 derivative(an American strain of serotype O:8). We randomly chose oneAmerican strain (serotype O:4,32) and three non-American strains(serotypes O:5,27; O:3; and O:9), all of which are Ara⁺ (aswere all of the Y. enterocolitica strains in our collection).First, we attempted PCR amplification of the 'araBA' and 'araGH'fragments with the same primers used to construct pAJD898. Bothfragments were successfully amplified from the O:4,32; O:5,27;and O:3 strains (data not shown). However, only the 'araGH'fragment could be amplified from the O:9 strain (therefore,this strain was not used further). Next, we attempted the pAJD905integration and segregation procedure. For the American O:4,32strain, the process proceeded with speed and efficiency similarto that seen with our strain 8081 derivative. The procedurewas also successful for the non-American O:5,27 and O:3 strains.However, in these cases the segregation procedure heavily favoredreversion to Ara⁺. Only a few desired Ara^– segregantswere isolated. Even so, colony PCR analysis identified segregantswith the correct integration in both cases (data not shown).From these experiments we conclude that pAJD905 can be usedin different Y. enterocolitica strains, although not all. However,the system is likely to be most easily used in the highly pathogenicAmerican strains.

Summary.

Top

Summary.

The Y. enterocolitica single-copy lacZY operon fusion systemdescribed here provides many benefits, including the following.(i) Fusion strains are stable, are not marked with an antibioticresistance, and do not contain any suicide vector DNA, all ofwhich facilitate further manipulation. (ii) There is no lowerlimit to the size of promoter fragments that can be analyzed.(iii) There is low background (integration of lacZY withouta promoter fragment upstream causes only approximately 2 to4 Miller units of ß-galactosidase activity; data notshown). (iv) The native locus of the promoter under study isnot disrupted (polar effects are not a concern). (v) The integrationprocedure is simple and takes only a few days. (vi) CorrectlacZY fusion integrants are easily and unambiguously confirmedby their Ara^– phenotype and colony PCR analysis. Our identificationof the Y. enterocolitica gene responsible for endogenous ß-galactosidaseactivity, and construction of a deletion mutant, further improvesthe utility of lacZ fusions in this organism.

View this table:
[in this window]
[in a new window]
TABLE 2. Primers

ACKNOWLEDGMENTS

We thank Joe Vogel and Kimberly Walker for providing plasmidpSR47S and its DNA sequence. Y. enterocolitica genome sequencedata were produced by the Y. enterocolitica Sequencing Groupat the Sanger Institute and can be obtained online at http://www.sanger.ac.uk/Projects/Y_enterocolitica/.

This work was supported by Public Health Service grant AI-052148from the National Institute of Allergy and Infectious Diseasesand by a grant from the Speaker's Fund for Biomedical Research:Toward the Science of Patient Care, awarded by the City of NewYork.

FOOTNOTES

* Corresponding author. Mailing address: Department of Microbiology MSB 228, New York University School of Medicine, 550 First Avenue, New York, NY 10016. Phone: (212) 263-3223. Fax: (212) 263-8276. E-mail: darwia01{at}med.nyu.edu.

REFERENCES

Top