Employing Site-Specific Recombination for Conditional Genetic ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 2011, p. 3916–3922 0099-2240/11/$12.00 doi:10.1128/AEM.00544-11 Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Vol. 77, No. 12

Employing Site-Specific Recombination for Conditional Genetic Analysis in Sinorhizobium meliloti䌤† Clarice L. Harrison,1 Matthew B. Crook,1 Gledi Peco,1 Sharon R. Long,2 and Joel S. Griffitts1* Department of Microbiology and Molecular Biology, Brigham Young University, Provo, Utah 84602,1 and Department of Biological Sciences, Stanford University, Stanford, California 943052 Received 9 March 2011/Accepted 15 April 2011

The ability to remove a genetic function from an organism with good temporal resolution is crucial for characterizing essential genes or genes that act in complex developmental programs. The rhizobium-legume symbiosis involves an elaborate two-organism interaction requiring multiple levels of signal exchange. As an important step toward probing rhizobium genetic functions with temporal resolution, we present the development of a conditional gene deletion system in Sinorhizobium meliloti that employs Cre/loxP site-specific recombination. This system enables chemically inducible and irreversible gene deletion or gene upregulation. Recombinase-mediated excision events can be positively or negatively selected or monitored by a colorimetric assay. The system may be adaptable to various bacterial species, in which recombinase activity may be placed under the control of diverse user-defined promoters. This system also shows promise for uses in promoter trapping and biosensing applications. symbiosis with several leguminous plant species, including alfalfa. The process by which S. meliloti engages with a compatible plant host and invades root nodule tissue is complex, involving many levels of signal exchange and dramatic cellular differentiation on the part of both the host and the microsymbiont (10, 17, 18). A refined understanding of S. meliloti genetic functions involved in symbiotic development depends on conditional genetic techniques that are not currently available, even after decades of intensive research on this organism. The aim of the present study is to employ the Cre/loxP recombination system to enable temporally controlled gene deletion in Sinorhizobium meliloti.

In bacterial genetic analysis, there is a frequent need to work with conditional mutants, particularly if the gene under investigation is essential for viability. Historically, essential genes have been studied with the use of temperature-sensitive alleles or by placing the gene of interest under the control of a promoter that is inducible by an exogenous chemical (3, 4, 9, 15, 30, 31). In these systems, the manipulation of either the temperature or chemical inducer produces a reversible change in phenotype, while the genetic status of the organism remains unchanged. In bacteria that execute complex biological programs such as host infection, there is a need for new conditional genetic approaches in which a transient stimulus induces the irreversible deletion of a gene of interest. This way, genetic pathways can be manipulated with good temporal resolution as bacteria are functioning within a complex environment. Inducible gene deletion also has the potential to drive the design of biosensors that can obtain and store (through permanent genetic rearrangements) environmental information. Site-specific recombinases are naturally occurring enzymes that cause genetic rearrangements by aligning two copies of a specific sequence and catalyzing a crossover event. The enzymes Cre and Flp are members of the tyrosine recombinase family that catalyze crossovers between the 34-bp recognition sites loxP and FRT, respectively (14, 25). The Cre/loxP and Flp/FRT systems have been successfully used to cause programmed gene deletions and genome rearrangements in various model organisms (2, 5, 29, 33), but their use for conditional genetics in bacteria has been limited (19). Sinorhizobium meliloti is a Gram-negative bacterium that lives freely in the soil and can participate in a nitrogen-fixing

MATERIALS AND METHODS Bacterial strains and growth conditions. All bacterial strains used in this study were derivatives of either Escherichia coli strain DH5␣ or S. meliloti strain Rm1021 (Table 1). E. coli and S. meliloti cultures were grown at 37°C and 30°C, respectively. Strains were routinely grown in Luria-Bertani (LB) medium supplemented with the following antibiotics, as appropriate: ampicillin (Ap; 100 ␮g ml⫺1), chloramphenicol (Cm; 30 ␮g ml⫺1), gentamicin (Gm; 3 ␮g ml⫺1 for E. coli and 15 ␮g ml⫺1 for S. meliloti), kanamycin (Km; 30 ␮g ml⫺1), neomycin (Nm; 100 ␮g ml⫺1), spectinomycin (Sp; 80 ␮g ml⫺1), streptomycin (Sm; 200 ␮g ml⫺1), or tetracycline (Tc; 5 ␮g ml⫺1). For chemical induction of cre expression in S. meliloti, cells were grown in LB broth supplemented with 10 mM taurine or 10 mM ectoine. Plasmid and strain construction. Plasmids and strains used in this study are listed in Table 1. Plasmids were constructed with standard techniques, using enzymes purchased from New England Biolabs. The high-fidelity polymerase Pfx50 (Invitrogen) was used for insert amplification. All custom oligonucleotides were purchased from Invitrogen. For the construction of pJG181, the Ptrp promoter is a synthetic version of the Salmonella enterica tryptophan promoter that acts constitutively in S. meliloti. Wild-type loxP sites were created synthetically, the Gmr gene was derived from pJQ200sk (24), the thrA-rpoC dual terminator was derived from E. coli strain MG1655, the Spr gene was derived from pMB393 (1), and the gus gene was derived from pVO155 (23). These elements were ligated between the Tn5 inverted repeats of pJG110 (13), as shown in Fig. 1. A more detailed map of pJG181, as well as its entire sequence, is given in Fig. S1 in the supplemental material. To generate the general excision detection strain B033, pJG181 was mobilized into S. meliloti strain 1021 by triparental mating using E. coli strain B001 as a helper. Transposants were selected by growth on Gm and Sm. Strain B033 was confirmed to be able to grow on minimal medium and to stimulate

* Corresponding author. Mailing address: Department of Microbiology and Molecular Biology, Brigham Young University, Provo, UT 84602. Phone: (801) 422-7997. Fax: (801) 422-0519. E-mail: joelg@byu .edu. † Supplemental material for this article may be found at http://aem .asm.org/. 䌤 Published ahead of print on 22 April 2011. 3916

VOL. 77, 2011

CONDITIONAL GENETICS IN S. MELILOTI

3917

TABLE 1. Strains and plasmids used in this study Strain or plasmid

Relevant characteristics

Strains Rm1021 DH5␣ B001 B511 B033 C294 C296 C446 C450 C456 C369 C464

S. meliloti wild type (Smr) E. coli cloning strain and donor for S. meliloti modifications DH5␣/pRK600, helper strain for conjugation (Cmr) Rm1021, pJG298 integrated (Smr Gmr) Rm1021, transposon introduced from pJG181 (Smr Gmr) pJG468 integrated into B033 (Smr Gmr Tcr) pJG470 integrated into B033 (Smr Gmr Tcr) B511 containing transposon from pJG504 (Smr Gmr Nmr) C446, native copy of feuN removed (Smr Nmr) pJG468 integrated into C450 (Smr Nmr Tcr) pJG468 integrated into Rm1021 (Smr Tcr) loxP-flanked feuN from C456 transduced to C369 (Smr Nmr Tcr)

21 11 12 6 This This This This This This This This

Plasmids pJG153 pJG157 pJG179 pJG181 pJG298 pJG468 pJG470 pJG497 pJG504

Promoterless Cre expression vector (Tcr)a Promoterless Cre expression vector (Tcr)a Promoterless Cre expression vector (Tcr)a Excision module vector (Gmr)b feuN deletion plasmid, derived from pJQ200sk (Gmr) Ptau-cre integration vector derived from pJG157 (Tcr) Pehu-cre integration vector derived from pJG157 (Tcr) Excision module derived from pJG181 (Nmr) pJG497, feuN cloned into AvrII (floxed feuN; Nmr)

This This This This 6 This This This This

a b

Reference

study study study study study study study study study study study study study study study study

Fig. 2; see also Fig. S1B in the supplemental material. Fig. 1; see also Fig. S1A in the supplemental material.

nitrogen-fixing nodules on alfalfa plants. The transposon insertion in B033 was mapped by arbitrary PCR to hypothetical gene SMc01549 on the S. meliloti chromosome. A map of cre expression vector pJG157 is shown in Fig. 2. This vector is also derived from pJG110, with the following modifications: a synthetic transcriptional terminator based on the E. coli pyrL attenuator was cloned upstream of the left transposon end, cre was amplified directly from bacteriophage P1, and the Tcr region consists of the tetR-tetA genes from plasmid RK2. The cre ribosome binding region in pJG157 is shown in Fig. 2B. Plasmids pJG179 and pJG153 differ from pJG157 only in the cre ribosome binding regions, as shown in Fig. 2B. A more detailed map of pJG157, as well as its entire sequence, is given in Fig. S2

in the supplemental material. To generate cre promoter trap libraries, the three promoterless cre vectors (pJG179, pJG157, and pJG153) were mobilized into S. meliloti strain B033 by triparental mating. Transposants were selected by growth on Tc and Sm. Colonies were then tested for resistance to either Gm or Sp. Plasmids for integrating cre downstream of Pehu and Ptau were constructed by removing the HindIII/BamHI fragment from pJG157 and replacing it with appropriate homology regions. The Pehu homology region was amplified using primers 1115 (5⬘-CGCAAGCTTCTGGCAAGCCTTGGCCAC-3⬘) and 1116 (5⬘-CGCGGATCCCTGGGACATTCACGCCTCTC-3⬘). The Ptau homology region was amplified using primers 1105 (5⬘-CGCAAGCTTCGGTCTCGACGG AACAGATCC-3⬘) and 1106 (5⬘-CGCGGATCCGGCGGTCATGTCGAAATT

FIG. 1. Structure of the excision module. A map of the transposable region of plasmid pJG181 is depicted as an insertion in the S. meliloti chromosome. The predicted products of the Cre-mediated recombination reaction are shown, including the nonreplicative circle containing the floxed region. Binding sites for diagnostic primers 1248 and 1249 are also shown. TE, transposon ends; Ter, dual transcriptional terminator sequence derived from the E. coli thrA and rpoC genes. The AvrII restriction site is a convenient site for insertion of additional DNA into the floxed region.

3918

HARRISON ET AL.

APPL. ENVIRON. MICROBIOL.

FIG. 2. Structure and properties of the Cre expression module. (A) General structure of promoterless cre transposon delivery plasmids pJG153, pJG157, and pJG179. The HindIII/BamHI fragment may be replaced with a homology region for chromosomal integration by recombination rather than transposition. These plasmids differ in the RBS sequences diagrammed (B). Ter, E. coli pyrL terminator. (B) Comparison of tuned RBS variants. The BamHI site is shown in boldface. The cre start codon is shown in boldface and uppercase letters. Underlined regions are capable of forming RNA secondary structures with various folding energies (⌬G). Percent excision is determined by testing random cre transposants (recipient strain B081) for loss of Gm resistance.

CCTTG-3⬘). These cloning steps resulted in plasmids pJG470 and pJG468, respectively. Integration of these plasmids into the B033 genome was selected by growth on Tc and Gm after triparental mating, resulting in strains C296 (B033, Pehu-cre) and C294 (B033, Ptau-cre). To construct the feuN excision module, the Nmr gene derived from pJG110 was amplified using overlap extension PCR in order to eliminate a naturally occurring internal SphI site. The primers used to accomplish this were 1186 (CAGGCGCATGCCCCACAGCAAGCGAACCGG), 1187 (5⬘-TCGCCGTCG GGCATACGCGCCTTGAGCCT-3⬘), 1188 (5⬘AGGCTCAAGGCGCGTATG CCCGACGGCGA-3⬘), and 1189 (5⬘GTGCGGCATGCGGGTCAGAAGAAC TCGTCAAG-3⬘). The final product was used to replace the Gmr gene in pJG181 with Nmr using the SphI sites flanking the Gmr gene, resulting in plasmid pJG497. The feuN open reading frame was then cloned into the unique AvrII site of pJG497 by amplifying feuN from S. meliloti genomic DNA using primers 1214 (5⬘-CGCTCTAGATCCCAGTCGAAAGTGCGTTC-3⬘) and 1215 (5⬘-GCGTC TAGAGCCTTCGGGTACCCCGAAAG-3⬘) and digesting the PCR product with XbaI prior to ligation. This gave rise to plasmid pJG504. To make feuN deletable by chemical induction, the feuN allele exchange plasmid pJG298 (6) was first integrated into strain 1021 by selection with Gm after triparental mating, giving strain B511. The excision module containing floxed (flanked by loxP) feuN was introduced into B511 by transposition. This was done by mobilizing pJG504 into B511 by triparental mating and selecting with Nm and Gm, yielding strain C446. The allele exchange vector was evicted from the native feuN locus by selection on 1% sucrose and verification of feuN deletion at the native locus by PCR, giving strain C450. The Ptau-cre plasmid was then introduced into C450 by delivering pJG468 via triparental mating and selection on Nm and Tc, resulting in strain C456. The isogenic control strain containing intact feuN at the native feuN locus was constructed as follows: pJG468 was integrated into wild-type strain 1021 by triparental mating, followed by selection on Tc and Sm, resulting in strain C369. The floxed feuN module was then transduced from C456 into C369 using phage N3, giving the control strain C464. Expression of Cre using chemical induction. In vivo chemically induced excision reactions were performed using C296 (ectoine inducible) and C294 (taurine inducible). The strains were inoculated into fresh liquid medium containing Tc and 10 mM inducer at an initial cell density at 600 nm (OD600) of approximately 0.01. The culture was then allowed to grow with shaking for 48 h (approximately 8 generations). After this exposure period, cultures were diluted and plated to either LB-Gm or LB-Sp to evaluate the status of the excision module. Assays for detection of excision kinetics. Site-specific recombination events were monitored over time either by PCR or glucuronidase (GUS) assay. For PCR detection, primers 1248 (5⬘-AGGTCGACATCATAACGGTTCC-3⬘) and 1249 (5⬘-CTTACGGTCACCGTAACCAGC-3⬘) were used. These primers amplify a 1.5-kb fragment from unmodified excision modules and a 0.3-kb fragment from excision modules in which the floxed region has been removed. For PCR tests, culture samples were normalized to a uniform optical density before being briefly boiled in PCR lysis buffer (13). PCR was performed using Taq polymerase (NEB), according to the manufacturer-suggested protocol. GUS assays were performed by combining 50 ␮l of cell suspension with 600 ␮l of GUS buffer (100 mM sodium phosphate [pH 7.0], 7 mM ␤-mercaptoethanol,

0.1% Triton X-100, and 2.5 mM p-nitrophenyl-␤-D-glucuronide [PNPG]) and 50 ␮l of chloroform. This mixture was incubated at 37°C for 4 h, followed by addition of 600 ␮l of stop buffer (1 M sodium carbonate). Miller units were calculated as described previously for ␤-galactosidase assays (22), except that absorption at 405 nm was used to monitor PNPG hydrolysis. Analysis of the feuN deletion phenotype. Strains C464 and C456 were suspended in LB broth at equivalent densities (OD600 ⫽ 0.5). A total of 50 ␮l of these suspensions was then added to 40-ml LB cultures in the presence of Tc and the presence or absence of 10 mM taurine. Cultures were shaken at 30°C to induce the in vivo excision reactions. After 60 h of growth, taurine-induced cultures of strains C464 and C456 were normalized to an OD600 of 1.0, serially diluted, and spotted on LB-Sp plates to observe growth of single colonies.

RESULTS AND DISCUSSION Design and construction of a single-copy excision module. We previously reported the construction of an efficient miniTn5 delivery plasmid, pJG110 (13). The excision module for our Cre-mediated deletion system was constructed between the transposon ends of this plasmid, allowing delivery to the S. meliloti genome as a single copy. The structure of the excision module is shown in Fig. 1. It consists of a constitutive promoter (Ptrp) reading into the loxP-flanked (“floxed”) region, with transcription terminating within the floxed region due to a pair of strong intrinsic terminators derived from E. coli thrA and rpoC. To the right of the floxed region, promoterless reporter genes may be used to positively detect excision events due to read-through transcription from Ptrp after removal of the floxed terminators. In the pJG181 prototype construct, the floxed region contains a gentamicin (Gm) resistance gene, and the read-through reporter genes control spectinomycin (Sp) resistance and glucuronidase (GUS) activity. Figure 1 depicts in graphical format the Cre-mediated rearrangement that leads to removal of the floxed region and induction of the read-through reporters. With the excision module thus arranged, the researcher can select for either retention (using Gm) or excision (using Sp) of the floxed region. Excision may be monitored without selection by measuring GUS activity or by PCR. All promoter, terminator, and protein-encoding elements on the excision module are flanked by convenient restriction sites so that they can be customized according to researcher preferences (a detailed restriction map of pJG181 is

VOL. 77, 2011


3919

FIG. 3. Performance testing of chemically inducible excision. The Pehu-cre fusion (A) or the Ptau-cre fusion (B) was integrated into the genome of strain B081 by single-crossover homologous recombination. The resulting strains were grown in the presence or absence of 10 mM inducer for 48 h followed by plating of a 10⫺6 dilution of each culture on either LB-Gm or LB-Sp. Plates were incubated at 30°C for 3 days prior to imaging.

provided in Fig. S1 in the supplemental material). pJG181 contains a pUC replication origin and is therefore a high-copynumber plasmid in E. coli and nonreplicative in most other Gram-negative species. The construct also contains a conjugal transfer origin and the transposase (tnp) gene for Tn5 transposition in recipient cells. S. meliloti strain B033 is a derivative of the laboratory strain Rm1021, harboring the transposon from pJG181 at a chromosomal site that is not necessary for prototrophy or symbiosis. As expected, strain B033 is Gmr Sps and does not express detectable GUS activity. Design and construction of Cre expression modules. The Cre expression module is also designed to be introduced into cells as a single-copy entity. As illustrated in Fig. 2, this module resides on a pUC-derived plasmid that can be transferred to recipient cells by conjugation. A promoterless copy of cre may be randomly inserted into the genomes of recipient cells by Tn5 transposition to create a promoter trap library; alternatively, the HindIII/BamHI fragment may be removed and replaced with a fragment that can guide the integration of the cre gene into a specific site in the recipient genome by homologous recombination. In early studies on the functionality of our system, we transposed the Cre expression module randomly into strain B033 and found that ⬎95% of transposition events resulted in Gms Spr colonies. This suggested that Cre expression is sufficient for excision even when cre transcription occurs at extremely low levels. In these promoter trap experiments, the cre gene was accompanied by a canonical ribosome binding site (RBS) consisting of a purine-rich segment centered 9 bp upstream of the start codon. To “tune” the cre gene to be switched on only in response to more elevated levels of transcription, the RBS was manipulated (Fig. 2B). In the three cre variants shown, the sequence around the start codon was manipulated to form RNA secondary structures of various lengths without altering

the coding region of the gene. These variants were tested by random transposition into the reporter strain B033, and as expected, the percentage of Gms Spr transposants was decreased in a manner that was inversely proportional to the predicted folding energy of the RNA secondary structure. By generating this allelic series of cre variants, diverse regulated promoters can be productively fused to cre to enable controlled excision. Making Cre expression chemically inducible. The clone pJG157, which gave 10% excision frequency in the promoter trap assay described above, was used for making directed promoter fusions with cre. The following two endogenous S. meliloti promoters were tried: the taurine-inducible promoter located upstream of the tauABC genes on megaplasmid pSymB and the ectoine-inducible promoter located upstream of the ehuABCD genes on pSymB (16, 20, 32). These promoters will be referred to as Ptau and Pehu, respectively. DNA fragments designed to align cre appropriately with these promoters were cloned into the HindIII and BamHI sites of pJG157, resulting in the removal of most of the gene for the transposase (Fig. 2A). The plasmids thus modified were mobilized into the excision reporter strain B033 and selected for plasmid integration. The resulting B033/Pehu-cre and B033/Ptau-cre strains were maintained on medium containing Gm to ensure retention of the floxed region. Both strains were then grown in the presence or absence of the respective inducers to evaluate their performance. The induction period was carried out in liquid medium in the absence of Gm or Sp, and cells were then plated in the absence of inducer to evaluate relative frequencies of resistance to Gm and Sp. As shown in Fig. 3, both strains show high levels of retention (Gmr Sps) in the absence of inducer and high levels of excision (Gms Spr) in the presence of inducer. B033/Pehu-cre exhibited noticeable leakiness (some Spr colonies from the no-inducer control culture), but it showed a

3920

HARRISON ET AL.

FIG. 4. Kinetics of chemically inducible excision. Strain C294 (B081/Ptau-cre) was inoculated into liquid LB-Tc to an OD600 of 0.15 in the presence or absence of 10 mM taurine. Samples were removed for analysis every 3 h for 12 h. (A) Samples were brought to a uniform cell density in PCR lysis buffer and boiled for 3 min prior to amplification across the floxed region using primers 1248 and 1249. The 1.5-kb band indicates an intact excision module, and the 0.3-kb band indicates the product of the Cre-dependent excision reaction. (B) GUS activity was measured from the same samples used for the PCR analysis described in panel A.

high level of inducibility (all cells sampled were Gms Spr after exposure to inducer). B033/Ptau-cre behaved ideally, exhibiting no detectable leakiness and complete inducibility (Fig. 3B). Under the plating conditions used in Fig. 3B, we can conclude that greater than 99.9% of taurine-induced cells respond by excision of the floxed region. In a separate experiment, induced or uninduced cells were plated on nonselective medium followed by an assay of excision by patching individual colonies on Gm-supplemented medium. From this experiment, we determined that B033/Ptau-cre showed 100% retention in the absence of inducer (n ⫽ 295) and 100% excision in the presence of inducer (n ⫽ 296); B033/Pehu-cre showed 96% retention in the absence of inducer (292/305 Gmr) and 99.7% excision in the presence of inducer (1/341 Gmr). Analysis of excision kinetics. B033/Ptau-cre was then used to examine the kinetics of taurine-induced gene excision. Over a period of 12 h, samples were removed from B033/Ptau-cre cultures grown in the absence or presence of 10 mM taurine. These samples were then assayed for excision by both GUS assay and PCR. As shown in Fig. 4, neither method could detect excision in the absence of inducer. In the presence of 10 mM taurine, excision was detected by PCR as soon as 3 h after induction, with complete excision apparent by 9 h after induction. This corresponds to complete excision occurring across the population within two generation times after induction. GUS assays reflected what was observed by PCR but with a slight lag in response, probably due to the additional time required for transcription, translation, and folding of the GUS enzyme. Use of the conditional excision system to analyze loss of an essential S. meliloti gene. A simple application of the system we have developed is the direct observation of loss-of-function events that result in cessation of cell growth or cell death. A


FIG. 5. Demonstration of the importance of feuN for S. meliloti growth, using the chemically inducible gene excision system. (A) Genotypic descriptions of strains C464 and C456, followed by growth curves of strains in the presence or absence of 10 mM taurine. F, C464 without taurine (⫺tau); E, C464 with taurine (⫹tau); f, C456 without taurine; 䡺, C456 with taurine. Bars represent standard errors of the means. (B) Samples were removed from C464 and C456 taurineinduced cultures at the 60-h time point as indicated. Samples were normalized to an OD600 of 1.0 and used for dilution spotting on LB-Sp followed by incubation at 30°C for 3 days prior to imaging.

potential pitfall in such an experiment is that the chemically mediated excision reaction may not have sufficient penetrance, and cells that fail to respond to the excision inducer (ectoine or taurine in this case) may rapidly dominate a population in culture. The evidence we have presented above suggests that our system brings about gene excision with very high (⬎99.9%) penetrance. We recently characterized the S. meliloti protein FeuN, which acts as a negative regulator of signaling through the FeuP/FeuQ two-component system (6). At that time, we reported that feuN is required for viability since it cannot be deleted from the wild-type S. meliloti genome by a conventional allele exchange protocol unless a plasmid-borne copy of the gene is present. The technology presented here provides us with the opportunity to more directly observe the effects of feuN deletion in a population of living cells. For this experiment, the excision module on pJG181 was altered by first replacing the Gm resistance gene with one encoding resistance to neomycin (Nm). Then, a functional copy of feuN was cloned into the AvrII site in the floxed region. This module was transposed into S. meliloti. In the resulting strain, the native copy of feuN was deleted by allele exchange as described previously (6). The Ptau-cre construct was then introduced into this strain, and integration into the tauABC locus was selected for, resulting in strain C456. An isogenic strain (C464) was prepared in which the native copy of feuN remains intact (Fig. 5A). We predicted that growth of strain C456 in the presence of taurine would result in a deleterious phenotype due to Cre-mediated excision of feuN, while growth of C464 would continue unim-

VOL. 77, 2011

peded, owing to the intact native copy. As shown in Fig. 5A, C456 is markedly inhibited in its growth, and this inhibition is taurine dependent. The isogenic strain C464 was slightly affected only by taurine induction with respect to both the exponential growth rate and final cell density. In the presence of taurine, strain C456 reaches a considerable density, even though it is markedly less than that attained by the control strain C464. We hypothesized that the noticeable growth observed under the C456 plus taurine condition is due either to FeuN protein perdurance after gene excision or to feuN deletion, leading to a slow-growth but not an abolished-growth phenotype. To distinguish between these models, cells were removed from C464 plus taurine and C456 plus taurine cultures after 60 h (approximately 10 generations), when FeuN protein perdurance is presumed to be negligible. The suspensions were normalized to an OD600 of 1.0, and dilutions were spotted onto LB-Sp to ensure that only cells in which the floxed copy of feuN has been evicted can grow (Fig. 5B). Cells from the taurine-induced C464 culture yield visible colonies to a dilution of 10⫺6, which is consistent with near 100% viability based on the density of the undiluted suspension. Cells from the induced C456 culture, however, yielded colonies at only 10,000-fold-higher spotting densities than those yielded by the control. These colonies likely arise due to suppressor mutations that allow cells to survive the loss of feuN. We have shown previously that the lethality resulting from deletion of feuN can be suppressed by any mutation that decreases signaling through the FeuP/FeuQ two-component system (6). Potential applications of the system. This system has several potential uses. In promoter trap mode, random insertion of cre into the bacterial genome could allow the rapid identification of promoters that are activated under complex conditions, such as in soil or in a host organism. The use of recombinasemediated rearrangements to monitor gene expression in symbiotic nodules has been demonstrated previously (8). The advantage of our approach for promoter trapping comes in the ability to positively select for both the retention and excision of the floxed region. This attribute allows the library to be preselected against promoters that are constitutively active by simply growing on Gm. After subsequent exposure of the library to the complex environment in question, recovered library members in which cre expression had been induced can then be selected on medium containing Sp. As currently constructed, our system traps promoters based on random transposon insertions, which would be expected to frequently disrupt genes under the control of those promoters. To overcome this, it should be feasible to move the promoterless cre fragment to a low-copy-number replicative plasmid in which promoter-cre fusion libraries could be screened without disruption of genes at their native loci. Because this system is based on an inducible and irreversible DNA rearrangement, it may be developed for sensing and reporting specific conditions in the complex environment of the soil. S. meliloti is an opportune organism for soil biosensor development due to its genetic manipulability and its ability to persist in the soil as a free-living organism. Already, S. meliloti is being studied for potential soil bioremediation purposes (7, 26–28). We have shown that this system can be used to chemically


3921

induce a uniform loss-of-function phenotype in a large population of S. meliloti cells. This opens the door to more sophisticated genetic analysis by inducing gene excision events during symbiotic nodule infection. This may require that cre be transcriptionally fused to developmentally controlled promoters or a promoter that can be induced by a tissue-permeable compound. One of the properties of this system that could potentially complicate such experiments is that of the perdurance of gene function after gene excision. This property could be ameliorated by minimizing the expression of the gene product at the level of transcription or translation. The perdurance property may actually be of benefit for investigating early symbiotic events, because genetic excision could be induced under standard culture conditions at defined time points prior to inoculation of plants so as to cause the elimination of protein function at temporally defined points during symbiotic invasion. This system may be adaptable to other bacterial species in which tools for conditional genetics have not been developed. Our plasmids are mobilizable by RK2-based conjugation, and all of the critical genetic elements in our system are flanked by convenient restriction sites so that organism-specific requirements may be met. For both the excision module and the Cre expression module, delivery to the genome can be mediated by Tn5 transposition or by integration via homologous recombination. ACKNOWLEDGMENTS This work was supported by a postdoctoral fellowship from the Helen Hay Whitney Foundation and grant 1R15AI082504-01 from the National Institutes of Health. REFERENCES 1. Barnett, M. J., V. Oke, and S. R. Long. 2000. New genetic tools for use in the Rhizobiaceae and other bacteria. Biotechniques 29:240–242, 244–245. 2. Bischof, J., and K. Basler. 2008. Recombinases and their use in gene activation, gene inactivation, and transgenesis. Methods Mol. Biol. 420:175–195. 3. Boneca, I. G., et al. 2008. Development of inducible systems to engineer conditional mutants of essential genes of Helicobacter pylori. Appl. Environ. Microbiol. 74:2095–2102. 4. Boyd, D., R. Nixon, S. Gillespie, and D. Gillespie. 1968. Screening of Escherichia coli temperature-sensitive mutants by pretreatment with glucose starvation. J. Bacteriol. 95:1040–1050. 5. Branda, C. S., and S. M. Dymecki. 2004. Talking about a revolution: the impact of site-specific recombinases on genetic analyses in mice. Dev. Cell 6:7–28. 6. Carlyon, R. E., J. L. Ryther, R. D. VanYperen, and J. S. Griffitts. 2010. FeuN, a novel modulator of two-component signalling identified in Sinorhizobium meliloti. Mol. Microbiol. 77:170–182. 7. Fan, L. M., et al. 2011. Characterization of a copper-resistant symbiotic bacterium isolated from Medicago lupulina growing in mine tailings. Bioresour. Technol. 102:703–709. 8. Gao, M., and M. Teplitski. 2008. RIVET—a tool for in vivo analysis of symbiotically relevant gene expression in Sinorhizobium meliloti. Mol. Plant Microbe Interact. 21:162–170. 9. Garcia-Moreno, D., et al. 2009. A vitamin B12-based system for conditional expression reveals dksA to be an essential gene in Myxococcus xanthus. J. Bacteriol. 191:3108–3119. 10. Gibson, K. E., H. Kobayashi, and G. C. Walker. 2008. Molecular determinants of a symbiotic chronic infection. Annu. Rev. Genet. 42:413–441. 11. Grant, S. G., J. Jessee, F. R. Bloom, and D. Hanahan. 1990. Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants. Proc. Natl. Acad. Sci. U. S. A. 87:4645–4649. 12. Griffitts, J. S., et al. 2008. A Sinorhizobium meliloti osmosensory two-component system required for cyclic glucan export and symbiosis. Mol. Microbiol. 69:479–490. 13. Griffitts, J. S., and S. R. Long. 2008. A symbiotic mutant of Sinorhizobium meliloti reveals a novel genetic pathway involving succinoglycan biosynthetic functions. Mol. Microbiol. 67:1292–1306. 14. Grindley, N. D., K. L. Whiteson, and P. A. Rice. 2006. Mechanisms of site-specific recombination. Annu. Rev. Biochem. 75:567–605. 15. Hunt, A., J. P. Rawlins, H. B. Thomaides, and J. Errington. 2006. Functional

3922

16.

17.

18. 19.

20.

21.

22. 23. 24.

HARRISON ET AL.

analysis of 11 putative essential genes in Bacillus subtilis. Microbiology 152: 2895–2907. Jebbar, M., L. Sohn-Bosser, E. Bremer, T. Bernard, and C. Blanco. 2005. Ectoine-induced proteins in Sinorhizobium meliloti include an Ectoine ABCtype transporter involved in osmoprotection and ectoine catabolism. J. Bacteriol. 187:1293–1304. Jones, K. M., H. Kobayashi, B. W. Davies, M. E. Taga, and G. C. Walker. 2007. How rhizobial symbionts invade plants: the Sinorhizobium-Medicago model. Nat. Rev. Microbiol. 5:619–633. Kouchi, H., et al. 2010. How many peas in a pod? Legume genes responsible for mutualistic symbioses underground. Plant Cell Physiol. 51:1381–1397. Liu, Y., P. Gao, S. Banga, and Z. Q. Luo. 2008. An in vivo gene deletion system for determining temporal requirement of bacterial virulence factors. Proc. Natl. Acad. Sci. U. S. A. 105:9385–9390. Mauchline, T. H., et al. 2006. Mapping the Sinorhizobium meliloti 1021 solute-binding protein-dependent transportome. Proc. Natl. Acad. Sci. U. S. A. 103:17933–17938. Meade, H. M., S. R. Long, G. B. Ruvkun, S. E. Brown, and F. M. Ausubel. 1982. Physical and genetic characterization of symbiotic and auxotrophic mutants of Rhizobium meliloti induced by transposon Tn5 mutagenesis. J. Bacteriol. 149:114–122. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Oke, V., and S. R. Long. 1999. Bacterial genes induced within the nodule during the rhizobium-legume symbiosis. Mol. Microbiol. 32:837–849. Quandt, J., and M. F. Hynes. 1993. Versatile suicide vectors which allow


25. 26.

27.

28. 29.

30.

31. 32. 33.

direct selection for gene replacement in gram-negative bacteria. Gene 127: 15–21. Sadowski, P. 1986. Site-specific recombinases: changing partners and doing the twist. J. Bacteriol. 165:341–347. Slaveykova, V. I., N. Parthasarathy, K. Dedieu, and D. Toescher. 2010. Role of extracellular compounds in Cd-sequestration relative to Cd uptake by bacterium Sinorhizobium meliloti. Environ. Pollut. 158:2561–2565. Teng, Y., et al. 2011. Influence of Rhizobium meliloti on phytoremediation of polycyclic aromatic hydrocarbons by alfalfa in an aged contaminated soil. J. Hazard. Mater. 186:1271–1276. Tu, C., et al. 2011. Potential for biodegradation of polychlorinated biphenyls (PCBs) by Sinorhizobium meliloti. J. Hazard. Mater. 186:1438–1444. Van Ex, F., D. Verweire, M. Claeys, A. Depicker, and G. Angenon. 2009. Evaluation of seven promoters to achieve germline directed Cre-lox recombination in Arabidopsis thaliana. Plant Cell Rep. 28:1509–1520. Varadarajan, R., H. A. Nagarajaram, and C. Ramakrishnan. 1996. A procedure for the prediction of temperature-sensitive mutants of a globular protein based solely on the amino acid sequence. Proc. Natl. Acad. Sci. U. S. A. 93:13908–13913. Wechsler, J. A., and J. D. Gross. 1971. Escherichia coli mutants temperaturesensitive for DNA synthesis. Mol. Gen. Genet. 113:273–284. Wilson, J. J., and U. Kappler. 2009. Sulfite oxidation in Sinorhizobium meliloti. Biochim. Biophys. Acta 1787:1516–1525. Wirth, D., et al. 2007. Road to precision: recombinase-based targeting technologies for genome engineering. Curr. Opin. Biotechnol. 18:411–419.