Tools for Rapid Genetic Engineering of Vibrio fischeri

ABSTRACT

Vibrio fischeri is used as a model for a number of processes, including symbiosis, quorum sensing, bioluminescence, and biofilm formation. Many of these studies depend on generating deletion mutants and complementing them. Engineering such strains, however, is a time-consuming, multistep process that relies on cloning and subcloning. Here, we describe a set of tools that can be used to rapidly engineer deletions and insertions in the V. fischeri chromosome without cloning. We developed a uniform approach for generating deletions using PCR splicing by overlap extension (SOEing) with antibiotic cassettes flanked by standardized linker sequences. PCR SOEing of the cassettes to sequences up- and downstream of the target gene generates a DNA product that can be directly introduced by natural transformation. Selection for the introduced antibiotic resistance marker yields the deletion of interest in a single step. Because these cassettes also contain FRT (FLP recognition target) sequences flanking the resistance marker, Flp recombinase can be used to generate an unmarked, in-frame deletion. We developed a similar methodology and tools for the rapid insertion of specific genes at a benign site in the chromosome for purposes such as complementation. Finally, we generated derivatives of these tools to facilitate different applications, such as inducible gene expression and assessing protein production. We demonstrated the utility of these tools by deleting and inserting genes known or predicted to be involved in motility. While developed for V. fischeri strain ES114, we anticipate that these tools can be adapted for use in other V. fischeri strains and, potentially, other microbes.

IMPORTANCE Vibrio fischeri is a model organism for studying a variety of important processes, including symbiosis, biofilm formation, and quorum sensing. To facilitate investigation of these biological mechanisms, we developed approaches for rapidly generating deletions and insertions and demonstrated their utility using two genes of interest. The ease, consistency, and speed of the engineering is facilitated by a set of antibiotic resistance cassettes with common linker sequences that can be amplified by PCR with universal primers and fused to adjacent sequences using splicing by overlap extension and then introduced directly into V. fischeri, eliminating the need for cloning and plasmid conjugation. The antibiotic cassettes are flanked by FRT sequences, permitting their removal using Flp recombinase. We augmented these basic tools with a family of constructs for different applications. We anticipate that these tools will greatly accelerate mechanistic studies of biological processes in V. fischeri and potentially other Vibrio species.

INTRODUCTION

Vibrio fischeri is a model organism for the investigation of quorum sensing (1–5), bioluminescence (6–11), biofilm formation (12–19), and symbiosis (20–25). As a result, the ability to readily manipulate this organism genetically has the potential to impact studies of many important aspects of bacterial physiology. Historically, most genetic studies have used strain ES114, an isolate from V. fischeri's symbiotic host, the Hawaiian bobtail squid Euprymna scolopes, although other squid isolates have been sequenced (15, 26–28) and some are beginning to be manipulated through genetic approaches (7, 29), thus increasing the need for tools and approaches to facilitate manipulations of V. fischeri.

A variety of methods have been used to generate specific mutations in ES114, most of which depend on the delivery via conjugation of a plasmid that contains specific V. fischeri sequences. Both replication-deficient (e.g., oriR6K based, such as pEVS122 [30]) and unstable (e.g., pEVS79 based [31]) plasmids have been used with success. The replication-deficient plasmid approach has most commonly been used for generating insertional (Campbell-type integration) mutants, which involves cloning a small 5′ end region of the gene of interest. Selection for the plasmid results in mutants that have integrated the delivery vector into the chromosome; however, these mutants carry a duplication of the cloned and native sequences that can resolve, reverting the strain to the wild-type genotype. Gene replacement/gene deletion strategies have also been used. These approaches rely on cloning up- and downstream sequences into a delivery plasmid, either a replication-deficient or an unstable vector. To obtain the desired mutant, two recombination events are required. First, the plasmid must integrate into the chromosome, an event that can be selected using an antibiotic resistance marker encoded by the plasmid. The second recombination event resolves the partial duplication generated by plasmid integration but may result in either the desired mutation or restoration of the wild-type allele. The desired mutation can be identified by selection, if the target gene of interest is replaced by an antibiotic resistance cassette, or by screening (by phenotype or with PCR), if it is unmarked. The ability to capture strains that have undergone the second recombination event can be facilitated by inducing toxin production from the integrated plasmid (e.g., CcdB [32, 33]), thus eliminating cells that retain the integrated plasmid. Numerous marked and unmarked mutations have been generated with these approaches (e.g., see references 25, 34, and 35). However, these approaches are time-consuming and require cloning, conjugation, and selection for the initial plasmid integration, followed by additional selection or screening for the desired deletion.

More recently, natural transformation has been developed as a way to introduce DNA into V. fischeri (36). The original report documented the ability of ES114 to take up chromosomal, plasmid, and PCR-generated DNA following exposure to chitin derivatives or in strains that overexpressed the tfoX gene. Although the efficiency was lowest with PCR products, the ability to introduce DNA through natural transformation relieves the dependence on plasmid-borne DNA for genetic manipulation of V. fischeri. Since the first report, subsequent studies used natural transformation of chromosomal DNA to move marked transposon insertion mutations between strains of V. fischeri (37, 38) and to map and repair a large deletion that arose in a derivative of ES114 (39). The latter study demonstrated that fragments as large as 10 kb could be transferred between strains. In V. cholerae, natural transformation has been used successfully to coordinately introduce multiple deletions using PCR products (40). The ability to reliably introduce even single deletions into V. fischeri using a PCR-based approach would greatly accelerate the rate of genetic manipulation and thus propel discovery in this marine model.

In addition to the current cloning burden, another roadblock to rapid genetic manipulation of V. fischeri is the limited number of antibiotic resistance markers that have been developed for use in this organism, particularly with markers that function in single copy. Specifically, erythromycin (Em) and chloramphenicol (Cm) resistance markers are the primary cassettes, followed by kanamycin (Kn). To date, tetracycline (Tc) has been used exclusively for selecting multicopy plasmids. One solution for the relative lack of useful antibiotic resistance cassettes is to make use of the Flp recombinase technology, which permits the deletion of sequences between direct repeats of FRT (FLP recognition target) sequences (41–43). With this technology, antibiotic resistance cassettes flanked by FRT sites can be used, resolved, and used again elsewhere in the chromosome. Alternatively, or in addition, new antibiotic resistance cassettes can be developed.

Here, we sought to enhance our ability to genetically manipulate V. fischeri using two related approaches that take advantage of natural transformation. First, we developed tools for the rapid deletion of specific genes using PCR splicing by overlap extension (SOEing) (44) of antibiotic resistance cassettes that can be resolved by the FRT/Flp recombinase system. As part of this work, we developed and validated additional antibiotic resistance cassettes to increase this toolset for V. fischeri. Second, we developed a method for the rapid insertion of specific genes at a benign site in the chromosome to facilitate complementation. We then expanded this toolset to include a variety of promoters for constitutive and inducible gene expression. These deletion and insertion approaches were validated using the genes for FlrA, the master flagellar regulator (25, 45), and FliQ, an uncharacterized putative flagellar protein. These studies confirmed and demonstrated roles for flrA and fliQ in motility and uncovered an unexpected connection between motility and cellulose. While this toolset was developed specifically for V. fischeri strain ES114, we anticipate that the ease, consistency, and speed of the engineering will be attractive to a wide audience and will likely be adapted for compatibility in other V. fischeri strains and other organisms. These methods will permit researchers to quickly and reliably generate strains to probe the mechanisms involved in a variety of biological processes.

RESULTS AND DISCUSSION

Constructs to facilitate rapid deletion of V. fischeri genes.

To facilitate rapid gene deletion in V. fischeri, we developed a PCR-based approach that would permit reliable amplification of fragments of interest by gene SOEing. Specifically, we engineered constructs that contained two distinct linkers (L1 and L2) positioned on either side of an antibiotic resistance cassette (Fig. 1A). We also included flanking FRT sequences to provide an option for making unmarked deletions by resolving the antibiotic cassette. Initially, we engineered pKV494 (erythromycin resistant, Em^r) (Fig. 1A; see also Fig. S1 in the supplemental material). The antibiotic resistance cassette in this plasmid is flanked by BamHI sites, so it can be replaced with other antibiotic resistance cassettes or other genes or markers of interest. We subsequently replaced the Em^r gene in pKV494 with chloramphenicol (Cm^r), tetracycline (Tc^r), trimethoprim (Tm^r), spectinomycin (Sp^r), and zeocin (Zc^r) resistance markers to generate a family of cassettes that can be used for recombination (Fig. 1B). Each cassette can be amplified from these constructs using PCR with primers complementary to the linkers, cL1 and cL2 (Table 1), and then fused via PCR SOEing to sequences up- and downstream of a gene of interest (GOI; Fig. 1C); the up- and downstream sequences would be generated using primers with 5′-end tails that incorporate sequences complementary to L1 and L2 (Del-up-F and Del-Up-R as well as Del-Down-F and Del-Down-R, respectively) (Table 1).

FIG 1.

Scheme for rapid deletion of genes of interest. (A) Plasmid pKV494 was engineered to contain FRT sequences (black rectangles) and distinct linker sequences, L1 and L2, flanking the Em^r antibiotic resistance cassette. Em^r sequences can be removed using BamHI. (B) A family of cassettes, derived from pKV494, contain different antibiotic resistance markers. (C) Scheme for deleting a gene of interest. (i) Gene of interest, GOI, is shown with flanking up- and downstream regions. (ii) Three PCRs were carried out with the indicated primers and templates. For up- and downstream sequences, the template was typically chromosomal DNA. (iii) The PCR SOE product, at the top, recombines into the chromosome, at the bottom, as indicated by the crossed lines. (iv) The final strain with GOI replaced by the engineered sequences.

TABLE 1.

Primer design for generating PCR SOE products for making deletions and insertions

Primer	Sequencea (from 5′ end)	Purposeb
Deletions
Del-Up-F	GOI specific	Amplifying upstream sequences, F
Del-Up-R	TAGGCGGCCGCACTAAGTATGG-[20–24-nt GOI specific]	Amplifying upstream sequences, R
cL1	CCATACTTAGTGCGGCCGCCTA	Amplifying Abr cassette, F
cL2	CCATGGCCTTCTAGGCCTATCC	Amplifying Abr cassette, R
Del-Down-F	GGATAGGCCTAGAAGGCCATGG-[20–24-nt GOI specific]	Amplifying downstream sequences, F
Del-Down-R	GOI specific	Amplifying downstream sequences, R
Insertions
Ins-Up-F	CTTGATTTATACAGCGAAGGAG	Amplifying upstream sequences, F
Ins-Up-F*	AAGAAACCGATACCGTTTACG	Use instead of Ins-Up-F for recombining at Em′
Ins-Up-R (cL2)	Same as cL2, above	Amplifying upstream sequences, R
Ins-GOI-F	GGATAGGCCTAGAAGGCCATGG-[20–24-nt GOI specific]	Amplifying GOI sequences, F
Ins-GOI-R	TAGGCGGCCGCACTAAGTATGGA-[20–24-nt GOI specific]	Amplifying GOI sequences, R
Ins-Down-F (cL3)	TCCATACTTAGTGCGGCCGCCTA	Amplifying downstream sequences, F
Ins-Down-R	GGTCGTGGGGAGTTTTATCC	Amplifying downstream sequences, R

^aGOI specific means sequences specific to the gene of interest (GOI) or to sequences up- or downstream of the GOI; nt, nucleotide.

^bAb^r, antibiotic resistance; F, forward primer; R, reverse primer.

We then tested the system by following the scheme outlined in Fig. 1C. Specifically, we amplified ∼500-bp sequences up- and downstream of two different genes positioned at the top and bottom of the flagellar hierarchy: flrA, which has previously been shown to function as the master regulator of flagellar biosynthesis in V. fischeri (25), and fliQ (VF_1841), which is predicted, but not yet shown, to be involved in motility based on homology and its position within the fliL-R flagellar operon (46). We fused appropriate flanking sequences to different FRT-antibiotic cassettes and introduced the composite PCR products into V. fischeri using natural transformation and selection for antibiotic resistance (Table 2). We evaluated motility of representative colonies that arose and found that the strains were nonmotile (Fig. 2A and B). PCR using the outermost up- and downstream primers confirmed the presence of the expected gene replacement mutations (Fig. 2C and D). These data thus indicate that both flrA, as previously determined, and fliQ, as hypothesized, are required for motility of V. fischeri.

TABLE 2.

Antibiotics, resistance cassettes, and selection conditions

Antibiotic	Cassette sizea (bp)	Final concnb	Special conditionc	Notee
Erythromycin	960	2.5
Chloramphenicol	886	1		Longer incubation requiredf
Tetracycline	2,223	2.5	No Ca2+ or Low Mgd
Trimethoprim	691	10		Spontaneous Tmr occurs
Spectinomycin	1,091	200		Initial lawn gives rise to colonies
Zeocin	572	10	Recover in LB	Initial lawn gives rise to colonies

^aCassette size includes the flanking BamHI (12 bp) and FRT (112 bp) sequences.

^bConcentration in μg/ml in LBS, except for zeocin, for which LB is used.

^cFor natural transformation and selection.

^dFive mM Mg is sufficient to promote growth and transformation.

^eObservations of growth on selective media.

^fColonies arise about a day later than occurs for Em^r colonies.

FIG 2.

Assessment of ΔflrA and ΔfliQ mutants. (A and B) Motility of the indicated flrA (A) and fliQ (B) mutants and the wild-type control. (C and D) PCR amplification of the indicated flrA (C) and fliQ (D) mutants generated with primer sets 2215 and 2218 (flrA) or 2356 and 2359 (fliQ). Abbreviations for the antibiotic resistance cassettes are as indicated in the text. L, ladder; −, no template; +, ES114; Δ, deletion with the antibiotic resistance cassette resolved. Numbers on the left indicate the sizes (kb) of relevant ladder bands.

Of the antibiotic resistance cassettes tested, Em^r worked the best in our hands for natural transformation: selection for Em^r colonies resulted in relatively robust growth of resistant colonies and little background growth. Some of the other antibiotics, notably Cm^r, required longer incubation times than Em^r (up to a day longer, depending). We found that effective selection for Tc^r colonies required special conditions, namely, preparing competent cells in minimal medium that either lacked calcium chloride or contained only small amounts of magnesium sulfate (Table 2); incorporation of standard salt concentrations permitted lawns of growth for V. fischeri strains that lacked the Tc^r marker. Selection for either Sp^r or Zc^r initially gave rise to background growth (lawns), but with additional incubation time, the background diminished and larger colonies arose that contained the desired mutation. For Tm^r, we occasionally obtained colonies that appeared to be spontaneously resistant, but these did not grow well upon restreaking on selective medium. Thus, although not all antibiotic resistance markers worked equally well, it was possible to obtain deletion mutants using each of the cassettes presented here.

In the course of these experiments, we observed that the flrA::Tm mutant exhibited altered colony morphology. Rather than the smooth phenotype of the wild-type strain, colonies formed by the flrA::Tm mutant grown on trimethoprim were bumpy, a phenotype indicative of biofilm formation (Fig. S2). Two polysaccharide loci in V. fischeri have been associated with altered colony morphology, syp (symbiosis polysaccharide) and bcs (cellulose) (35, 47–49). Consistent with the possibility that bcs was responsible for this phenotype, we found that growth of either flrA or fliQ mutants on agar containing Congo red resulted in colonies with a darker red hue than ES114 colonies, suggesting the mutants produce more cellulose (Fig. 3A and B). In addition, whereas flrA syp and fliQ syp double mutants phenocopied the flrA and fliQ single mutants, flrA bcs and fliQ bcs double mutants behaved like the bcs mutant, lacking colony color (Fig. 3A and B). These results thus reveal a new function for FlrA and FliQ and, likely, flagella in inhibiting cellulose production by V. fischeri.

FIG 3.

Congo red staining of V. fischeri strains. Strains were streaked onto LBS containing Congo red dye and then transferred to paper and photographed. The indicated flrA (A and C) and fliQ (B) mutants and controls were evaluated.

In summary, the use of standardized linkers and a family of cassettes that can be amplified with the same set of primers makes it possible to rapidly and reliably generate PCR fusion products. Combined with the fact that this approach requires no cloning or subcloning, mutants can similarly be obtained very rapidly, within as few as 2 to 3 days, rather than a few weeks or more. In this study, we did not explore the optimal size of the flanking sequences necessary for recombination; thus, it is possible that sizes smaller than 500 bp would work as well. Finally, this work expands the number of antibiotic resistance markers available for generating mutations from the limited set (Em^r and Cm^r) routinely used in the past (e.g., see references 30 and 31).

Removal of antibiotic resistance cassette.

Despite the increased availability of useful antibiotic resistance cassettes, the presence of a specific marker limits subsequent manipulations. Thus, to facilitate the engineering of multiple deletions and/or introduction of other sequences as described below, we cloned the flp recombinase gene into a plasmid that could be introduced into V. fischeri to resolve antibiotic cassettes between FRT sites. We then introduced the resulting construct, pKV496, into the ΔflrA::FRT-Em and ΔfliQ::FRT-Cm mutants. Colonies that were Em^s and Cm^s, respectively, were obtained, and representative isolates retained the expected nonmotile phenotype (Fig. 2A and B). Sequencing of the junction revealed that the resulting unmarked mutants contained residual “scar” sequences in the chromosome of 78 bp (26 amino acids) (Fig. S3), confirming the resolution of the FRT sequences. The linker sequences were engineered such that the resulting scar could be in frame with the deleted gene to reduce the possibility of polarity on downstream genes. While neither the insertion nor the scar present in the flrA gene is predicted to impact any downstream genes, fliQ is embedded in an operon (fliOPQR). The potential polarity of the fliQ mutation was evaluated in experiments described below.

Rapid insertion of genes for complementation.

Because PCR can result in unwanted mutations, even with use of a high-fidelity polymerase, and spontaneous mutations can arise, it is necessary to apply an additional approach, such as complementation, to verify that the engineered mutation is the cause of any observed phenotypes. We thus sought to develop an equally rapid method for complementing V. fischeri mutants. We first considered where to insert the complementation sequences. In previous work using a plasmid-based recombination approach (50), we had inserted the lacI gene between the yeiR and glmS genes of V. fischeri (VF_2370-2372). This location is adjacent to, but does not disrupt, the position where Tn7 inserts (Fig. 4A); maintaining the Tn7 site allows for simultaneous use of the many tools that rely on site-specific insertion by the Tn7 transposase (51–53). The lacI insertion at the indicated position also had no demonstrable effect on important aspects of V. fischeri physiology, notably motility and biofilm formation (50). We therefore chose to use this location (between yeiR and glmS) as the basis for our chromosomal complementation system.

FIG 4.

Scheme for rapid insertion of genes of interest. A method was developed to introduce a gene of interest (GOI) with an Em^r cassette into a benign site in the chromosome. (A) In this study, insertions were targeted to the indicated locus, the intergenic region between yeiR (VF_2370) and glmS (VF_2372). The arrow indicates the site of insertion, while the arrowhead indicates the position of the Tn7 site. (B) The up- and downstream segments are amplified with the indicated primers using pKV502 and pKV503 as templates, as shown, or similar templates, while the GOI is typically amplified from chromosomal DNA. (C) The PCR SOE product (top) recombines into the chromosome (middle), as indicated by the crossed lines, resulting in the insertion (bottom).

We then developed an approach that facilitates reliable insertion at that location (Fig. 4B and C). Specifically, we engineered tools that would permit a three-piece PCR SOEing reaction similar to the deletion approach. We generated a construct, pKV502, that could be used as a template to amplify the upstream (3′ end of yeiR) sequences together with FRT-Em^r sequences in a single reaction using a forward primer complementary to yeiR and a reverse primer (cL2) complementary to the linker (Fig. 4B). We also generated a construct, pKV503, that fuses a linker to the downstream sequences (3′ end of glmS). Amplification of pKV502 and pKV503 with appropriate primers results in PCR fragments with ends compatible with a linker-containing middle piece consisting of the gene of interest (GOI). SOEing of the three PCR fragments results in a product that is poised to be inserted into the region between yeiR and glmS (Fig. 4C).

To test this approach, we used it to introduce cassettes containing flrA and fliQ. This approach was successful with the ∼0.3-kb fliQ gene: fliQ was readily inserted into the intergenic region, as determined by PCR. Unfortunately, for flrA, we found that the large size of the final cassette (∼3.7 kb, with ∼1.4 kb upstream [+ Em^r], ∼1.8 kb flrA [including the putative promoter region], and ∼0.5 kb downstream), made the PCR SOEing reaction suboptimal. Potentially, the further optimization of the PCR and/or the use of a distinct long-range PCR enzyme would permit a more efficient reaction. Rather than continue to optimize PCR SOEing, we devised an alternative approach that would be more feasible for larger genes. We reasoned that the total size of the PCR SOEing product could be diminished by directing recombination to occur downstream of yeiR at inserted Em^r sequences. We thus introduced the 5′ end of the Em^r gene (a 3′ end truncation; Em′) into the intergenic region between yeiR and glmS, selecting for this insertion using an adjacent Tm^r cassette, to generate strain KV8232 (Fig. 5A). The use of KV8232 as a recipient eliminated the need to amplify any yeiR sequence. In addition, the length of necessary Em^r sequence was reduced, diminishing the total size of the PCR SOE product (Fig. 5B). For flrA, the cassette was reduced to ∼3 kb, which we could readily obtain. By including the missing 3′ portion of the Em^r resistance gene (′Em) in the PCR SOEing product, the Em^r gene can be reconstituted by recombination, permitting selection for colonies with the desired insertion with Em (Fig. 5C). Recombination results in the concomitant loss of Tm^r.

FIG 5.

Solution to the problem of large PCR SOE sizes. A modified scheme was developed to facilitate insertion of larger genes/sequences. (A) A truncated Em^r (Em′) cassette was integrated into the chromosome, generating KV8232 and providing a region of homology for recombination and permitting shorter products to be generated. (B) The up- and downstream segments were amplified with the indicated primers using pKV502 and pKV503 as templates, as shown, or similar templates; amplification results in a 5′-truncated, nonfunctional EmR sequence (′Em). (C) Recombination occurs between the two truncated Em^r cassettes to restore functional Em^r. The PCR SOE product (top) recombines into the chromosome (middle), as indicated by the crossed lines, resulting in the insertion (bottom). (D) KV8292, a strain similar to KV8232 but also containing lacI^q, was engineered to facilitate IPTG-inducible gene expression.

To more easily evaluate the efficacy of this modified approach, we moved the truncated Em′ (Tm^r) sequences into a ΔflrA strain. This strain (KV8281) was transformed with a PCR SOE product containing flrA under the control of its native promoter. A representative Em^r/Tm^s colony that arose (KV8290) exhibited motility similar to that of the wild-type strain (Fig. 6A and B). PCR-based analysis of this motile strain confirmed both the ΔflrA deletion and an insertion of the expected size at the yeiR-glmS region (Fig. 6C). Together, these data indicate that the flrA cassette had recombined at the desired position in the chromosome and was responsible for the restored motility of this strain. In addition, the Congo red phenotype was similarly restored to wild-type levels in this complemented strain, a result that confirms that FlrA negatively controls this phenotype (Fig. 3C).

FIG 6.

Analysis of flrA complementation strain. (A) Motility of the WT (wild type; ES114), ΔflrA (ΔflrA Em′ Tm^r; KV8281), and ΔflrA PflrA-flrA (KV8290) strains. (B) Motility of the same strains as those shown for panel A, monitored over time. (C) PCR amplification of the flrA and yeiR-glmS regions of the same strains as those shown in panel A with primer set 2215 and 2218 (flrA) or 2185 and 1487 (yeiR-glmS). L, ladder; WT, ES114; Δ, ΔflrA strain KV8281; C, complemented strain KV8290. Ladder is the same as that indicated in Fig. 2.

In summary, this insertion approach can facilitate rapid introduction of sequences to complement specific mutations. As with the deletion approach, we engineered the insertion tools to be versatile: the antibiotic resistance marker can be removed from the strain of interest using Flp recombinase, and/or the Em^r cassette in the template plasmid(s) can be replaced by other antibiotic resistance cassettes by digesting with BamHI. Our modification to accommodate larger genes of interest further increases the utility of this approach by reducing the overall size of the required PCR SOE product. The truncated Em′ and associated Tm^r sequences can be readily introduced into other strains of interest by selection for Tm^r (as we did for the flrA mutant), and/or marked mutations can be moved into this parent strain. This modified approach can be adapted to introduce Em′ sequences at other positions in the chromosome, facilitating the use of these tools (and the ones described below) for insertion elsewhere. This adaptation may be necessary if this approach is used for other strains, as other sequenced V. fischeri isolates appear to differ somewhat in this region of the chromosome. While we used PCR SOEing to generate the strains and deletions/insertions described here, as we could rapidly and reliably generate a variety of products of interest, it is becoming increasingly cost-effective to obtain synthetic DNA fragments; such synthetic DNA could be modeled after the tools described here to facilitate introduction/removal of sequences at any desired location. Finally, although the maximum amount of DNA that can be readily recombined into the chromosome has not been specifically addressed, the insertion of flrA along with associated promoter and truncated Em^r sequences totaled about 2.5 kb and should support insertion of other nonnative sequences up to at least this size.

Additional tools for other outcomes and analyses.

While the tools described above were a great advance, they remained limited. For example, genes positioned in the middle of an operon, such as fliQ, rely on promoters located upstream. Thus, complementing the fliQ mutant with a single gene, rather than an operon, requires additional engineering, such as another PCR SOEing event, to include the native promoter. To overcome this problem, as well as to further increase the utility of this approach, we expanded these tools to support a variety of experimental outcomes (Fig. 7 and Fig. S4), as described below.

FIG 7.

Constructs designed to support a variety of outcomes. (A) A family of cassettes was designed to be similar to base plasmid pKV502. They contain the FRT-flanked Em^r gene and different promoters as indicated. (B) A derivative of pKV503 was engineered to contain an HA epitope tag (pKV505).

Constitutive promoters.

To facilitate expression of genes such as fliQ that lack immediately adjacent promoters, we engineered constructs with two different constitutive promoters fused to the yeiR-FRT-Em sequences: one (pKV506) incorporated the V. fischeri promoter for the nrdR gene, PnrdR, which has previously been shown to be a strong promoter (37), and one (pKV519) contained Pma, a modified version of PA1/34, a synthetic promoter generated by Bose et al. (6). When either promoter was fused to promoterless fliQ and introduced into a fliQ mutant, the resulting strains were motile, albeit less so than the wild-type strain (Fig. 8B). Surprisingly, however, when a similar approach was used to complement a flrA mutant, the strains were nonmotile (Fig. 8A). Further inspection revealed that flrA lacks an obvious ribosome-binding site (RBS) (Fig. S5); thus, we considered the possibility that the failure stemmed from a lack of protein production rather than a lack of transcription. To test this possibility, we engineered the upstream primer to contain RBS sequences upstream of the start codon for flrA and repeated the experiment. These RBS-containing flrA complementation cassettes restored motility to the flrA mutant, albeit not to the level of the wild type (Fig. 8A). It is unclear why neither promoter restored full motility to the flrA or fliQ mutants, but it is likely due to the relative strength of the engineered promoters and/or their activity under these conditions. Additional, stronger promoters could be generated with a similar approach. The sequences that naturally promote translation of flrA remain to be determined.

FIG 8.

Motility of strains expressing flrA or fliQ from constitutive promoters. (A) Motility of a flrA mutant that expresses flrA under the control of different promoters positioned in the yeiR-glmS intergenic region. We assessed the wild type (ES114; filled circles), ΔflrA strain (KV8281; filled squares), and ΔflrA strains expressing the following flrA alleles: Pma-flrA (KV8293; triangles), PnrdR-flrA (KV8294; inverted triangles), Pma RBS-flrA (KV8364, filled diamonds), and PnrdR-RBS-flrA (circles; KV8365). (B) Motility of a fliQ mutant that expresses fliQ under the control of different promoters positioned in the yeiR-glmS intergenic region. We assessed the wild type (ES114; filled circles), the ΔfliQ strain (KV8376; filled triangles), and ΔfliQ strains expressing the following fliQ alleles: promoterless fliQ (KV8397; filled squares), Pma-fliQ (KV8398; open circles), and PnrdR-fliQ (KV8294; gray squares). Data shown in panel A for wild-type and KV8281 strains are the same as those shown in Fig. 6B.

Inducible promoters.

We also developed tools for inducible expression by engineering constructs and strains to permit induction by (i) cellobiose, (ii) isopropyl-β-d-thiogalactopyranoside (IPTG), or (iii) 3-oxo-C6-homoserine lactone (HSL) (V. fischeri autoinducer, or AI). For induction via cellobiose addition, we engineered pKV518, which contains the cellobiose-inducible promoter Pcel (34). We fused this promoter upstream of flrA and fliQ and inserted it at the yeiR-glmS region of the flrA and fliQ mutants, respectively. We then evaluated motility in the presence of two concentrations of cellobiose. Motility was observed with as little as 0.001% cellobiose, and for flrA, additional cellobiose (0.1%) increased motility (Fig. 9). For the flrA construct, motility was not observed (within the time frame of the experiment) unless cellobiose was added, and the larger amount of cellobiose yielded an increase level of motility relative to that of the smaller amount (Fig. 9A). For fliQ, a low basal level of activity was observed in the absence of cellobiose, and both amounts of cellobiose promoted robust motility, approaching that of the wild-type control (Fig. 9B). These data indicate that the Pcel promoter can be used for inducible expression, but the basal level and range of induction may vary depending on the GOI and/or the exact construction. These data also demonstrate that the fliQ deletion was not substantially polar on the downstream fliR gene, which is required for motility (46).

FIG 9.

Motility of strains expressing flrA or fliQ from the cellobiose-inducible promoter. (A) Motility of a flrA mutant that expresses flrA under the control of the cellobiose-inducible promoter Pcel in medium lacking or containing cellobiose. The following strains were assessed: the wild type (ES114; filled circles), the ΔflrA strain (KV8281; filled squares), and the ΔflrA strain expressing Pcel-flrA in medium lacking cellobiose (open squares) or containing 0.001% (gray-filled squares) or 0.1% (black-filled squares) cellobiose. (B) Motility of a fliQ mutant that expresses fliQ under the control of the cellobiose-inducible promoter Pcel in medium lacking or containing cellobiose. The following strains were assessed: the wild type (ES114; filled circles), the ΔfliQ strain (KV8298; open circles), and the ΔfliQ strain expressing Pcel-fliQ in medium lacking cellobiose (open squares) or containing 0.001% (gray-filled squares) or 0.1% (black-filled squares) cellobiose. For both panels A and B, motility of wild-type strain ES114 was unaffected by cellobiose; shown are the results for the no-cellobiose control.

To develop a system for IPTG-inducible gene expression, we began with promoter Pma, which we used above for constitutive expression. This promoter contains two binding sites for the IPTG-inducible regulator LacI (6). We then engineered a new V. fischeri strain that contains lacI^q in the yeiR-glmS region, along with the 3′-truncated Em′ sequences, to facilitate recombination (Fig. 5D). Using this approach, we generated a LacI-expressing ΔflrA mutant strain that contains Pma-flrA. The resulting strain was nonmotile in the absence of IPTG and became motile upon IPTG addition (Fig. 10A). This tight control is similar to that observed by Bose et al. (6).

FIG 10.

Motility of strains expressing flrA from IPTG-inducible and autoinducer-inducible promoters. (A) Motility of a flrA mutant that expresses flrA under the control of the IPTG-inducible promoter Pma in medium lacking or containing IPTG. The following strains were assessed: the wild type (ES114; filled circles), the ΔflrA strain (KV8281; open circles), and the ΔflrA strain expressing Pma-flrA in medium lacking IPTG (open squares) or containing 0.1 mM (gray filled squares) or 10 mM (black filled squares) IPTG. (B) Motility of a flrA mutant that expresses flrA under the control of the lux promoter Plux. The ability of lux regulators LuxR and LuxI to induce Plux was assessed by overexpressing these regulators in ΔflrA Plux-flrA strain KV8416. Motility of KV8416 carrying vector pVO8 (open circles), plasmid pKV27 (luxI), and plasmid pKV31 (luxR; open squares) was assessed in medium that contained Cm.

Finally, to develop an AI-responsive promoter system, we engineered pKV509, which contains the AI-responsive lux promoter. Plux is activated by LuxR bound by AI, primarily but not exclusively the 3-oxo-C6-HSL produced by LuxI. V. fischeri is an underproducer of AI (54), thus it was not surprising that cells expressing flrA from Plux were only poorly motile. To determine if motility could be increased by overexpression of LuxR and/or induced by the LuxI-produced AI, we introduced a multicopy plasmid that encodes either LuxR or LuxI. Motility of these strains was greatly increased relative to that of the vector control (Fig. 10B). Thus, Plux is responsive to LuxR and AI. This particular tool could be useful for experiments designed to control/induce a gene during symbiotic colonization, as the LuxI-controlled luminescence phenotype is highly induced under those conditions (54, 55).

A tool for evaluating protein production.

Another useful tool is the ability to evaluate protein production using an epitope tag. To facilitate this application, we incorporated the sequences for the hemagglutinin (HA) epitope tag into the downstream construct to generate pKV505 (Fig. 7B). Fusion of a gene of interest lacking its stop codon to PCR generated using pKV505 as a template should result in an HA-tagged protein that can be evaluated by Western blotting. To test this tool, we amplified flrA without its stop codon sequences under the control of Pcel. We fused these sequences to the PCR product generated with template pKV505 and introduced the resulting products into the ΔflrA::FRT mutant, selecting for Em^r colonies. Motility of the resulting Pcel-RBS-flrA derivative depended on the presence of cellobiose, consistent with the motility of the untagged version (data not shown). Importantly, HA-tagged FlrA was produced in the presence of cellobiose, as detected by Western immunoblotting using an anti-HA antibody (Fig. 11). Thus, this tool permits the rapid generation of strains that produce an epitope-tagged protein either constitutively or upon induction. Additional epitope tags or even sequences for fusion proteins, such as green fluorescent protein (GFP), or transcriptional reporters, such as lacZ, could be incorporated into the downstream construct to further increase their versatility.

FIG 11.

Detection of FlrA protein. Western blot analysis was used to detect HA epitope-tagged FlrA protein produced from Pcel-RBS-flrA-HA. The amount of cellobiose added to overnight cultures of V. fischeri is indicated. The FlrA-HA band is indicated with an arrow. The nonspecific larger band observed in all lanes serves as a loading control.

Concluding remarks.

This work presents tools for the rapid generation of deletions and insertions and demonstrates their utility using flrA and fliQ. This toolbox includes a set of cassettes with identical linkers that makes genetic manipulation of V. fischeri an easy and reliable method that requires few primers and does not depend on plasmid conjugation. In addition, the incorporation of BamHI restriction sites outside the antibiotic resistance markers readily permits the construction of plasmids with additional antibiotic resistance markers, and the use of FRT sequences further increases the versatility of the system. These tools provide a foundation that can be adapted to include other promoters, epitope tags, and even reporters such as gfp to meet the experimental needs of the scientific community. Insertions are not limited to the yeiR-glmS region; they can be facilitated by the introduction of a truncated Em^r cassette anywhere in a strain of interest. We anticipate that these tools and approaches will propel genetic manipulation and, thus, discovery in V. fischeri. Furthermore, we expect that these tools, while developed for ES114, can also be used to facilitate the study of other V. fischeri isolates of interest and/or adapted for use in other species.

MATERIALS AND METHODS

Strains and media.

V. fischeri strains used or generated in this study are listed in Table 3. Genetic manipulations were performed in V. fischeri strain ES114 and its derivatives. ES114 was also used as a template for PCR designed to amplify V. fischeri genes of interest. Escherichia coli strains TAM1 (Active Motif), TAM1 λ pir (Active Motif), DH5α, and π3813 (33) were used for cloning. For routine culturing, V. fischeri strains were grown in LBS (56, 57), while E. coli strains were grown in LB (58), in some cases supplemented with thymidine. For selection of E. coli on Em, brain heart infusion agar (Difco) was used. LBS agar containing Congo red and Coomassie blue dyes (40 μg ml⁻¹ and 15 μg ml⁻¹, respectively) was used to evaluate polysaccharide production by motility mutants. For natural transformation of V. fischeri, Tris minimal medium (TMM) (100 mM Tris, pH 7.5, 300 mM NaCl, 0.1% ammonium chloride, 10 mM N-acetylglucosamine, 50 mM MgSO₄, 10 mM KCl, 10 mM CaCl₂, 0.0058% K₂HPO₄, 10 μM ferrous ammonium sulfate) was used. Soft-agar motility medium contained tryptone (1%), NaCl (2%), agar (0.25%), and MgSO₄ (35 mM). Antibiotics were added as appropriate at the following final concentrations: ampicillin, 100 μg ml⁻¹; Cm, 1 μg ml⁻¹; Em, 2.5 μg ml⁻¹ (V. fischeri) or 150 μg ml⁻¹ (E. coli); kanamycin (Kn), 100 μg ml⁻¹ (V. fischeri) or 50 μg ml⁻¹ (E. coli); Tc, 2.5 μg ml⁻¹; Tm, 10 μg ml⁻¹; Sp, 200 μg ml⁻¹; and Zc, 10 μg ml⁻¹. For Zc, LB was used as the medium instead of LBS for outgrowth and plating.

TABLE 3.

Strains generated or used in this study

Strain	Genotypea	Derivationb	Reference or source
ES114	WT	NA	54
KV4607	ΔbinA bcsA::Tn5	NA	47
KV6576	IG::lacIq	NA	50
KV8111	ΔflrA::FRT-Emr	NT ES114 with SOE using primers 2215 and 2216 (ES114), 2089 and 2090 (pKV494), and 2217 and 2218 (ES114)	This study
KV8140	ΔflrA::FRT-Cmr	NT ES114 with SOE using primers 2215 and 2216 (ES114), 2089 and 2090 (pKV495), and 2217 and 2218 (ES114)	This study
KV8148	ΔflrA::FRT	KV8111 with Emr cassette removed via pKV496	This study
KV8183	ΔflrA::FRT-Tmr	NT ES114 with SOE using primers 2215 and 2216 (ES114), 2089 and 2090 (pMLC2), and 2217 and 2218 (ES114)	This study
KV8191	ΔsypQ::FRT-Emr	NT ES114 with SOE using primers 443 and 1188 (ES114), 2089 and 2090 (pKV494), and 2174 and 2175 (ES114)	This study
KV8225	ΔfliQ::FRT-Spr	NT ES114 with SOE using primers 2356 and 2357 (ES114), 2089 and 2090 (pKV521), and 2358 and 2359 (ES114)	This study
KV8232	IG::Em′ Tmr	NT ES114 with SOE using primers 2185 and 2292 (pKV502), 2293 and 2294 (Tmr), and 2089 and 1487 (pKV503)	This study
KV8244	ΔflrA::FRT bcsA::Tn5	NT KV8148 with KV4607 chDNA	This study
KV8245	ΔflrA::FRT ΔsypQ::FRT-Emr	NT KV8148 with KV8191 chDNA	This study
KV8268	ΔflrA::FRT-Tcr	NT ES114 with SOE using primers 2215 and 2216 (ES114), 2089 and 2090 (pKV514), and 2217 and 2218 (ES114)	This study
KV8281	ΔflrA::FRT IG::Em′ Tmr	NT KV8148 with KV8232 chDNA	This study
KV8290	ΔflrA::FRT IG::PflrA-flrA	NT KV8281 with SOE using primers 2097 and 2098 (pEVS170), 2265 and 2254 (ES114), and 2196 and 1487 (pKV503)	This study
KV8292	IG::lacIq-Em′ Tmr	NT KV6576 with SOE using primers 2324 and 2325 (pKV507) and 2326 and 1487 (KV8232)	This study
KV8293	ΔflrA::FRT IG::Pma-flrA	NT KV8281 with SOE using primers 2290 and 2090 (pKV519) and 2219 and 1487 (KV8290)	This study
KV8294	ΔflrA::FRT IG::PnrdR-flrA	NT KV8281 with SOE using primers 2290 and 2090 (pKV506) and 2219 and 1487 (KV8290)	This study
KV8295	ΔflrA::FRT IG::Pcel-flrA	NT KV8281 with SOE using primers 2290 and 2090 (pKV518) and 2219 and 1487 (KV8290)	This study
KV8298	ΔfliQ::FRT-Emr	NT ES114 with SOE using primers 2356 and 2357 (ES114), 2089 and 2090 (pKV494), and 2358 and 2359 (ES114)	This study
KV8299	ΔfliQ::FRT-Cmr	NT ES114 with SOE using primers 2356 and 2357 (ES114), 2089 and 2090 (pKV495), and 2358 and 2359 (ES114)	This study
KV8300	ΔfliQ::FRT-Tmr	NT ES114 with SOE using primers 2356 and 2357 (ES114), 2089 and 2090 (pMLC2), and 2358 and 2359 (ES114)	This study
KV8364	ΔflrA::FRT IG::Pma-RBS-flrA	NT KV8281 with SOE using primers 2290 and 2090 (KV8293) and 2355 and 1487 (KV8290)	This study
KV8365	ΔflrA::FRT IG::PnrdR-RBS-flrA	NT KV8281 with SOE using primers 2290 and 2090 (KV8294) and 2355 and 1487 (KV8290)	This study
KV8366	ΔflrA::FRT IG::Pcel-RBS-flrA	NT KV8281 with SOE using primers 2290 and 2090 (KV8295) and 2355 and 1487 (KV8290)	This study
KV8368	ΔflrA::FRT IG::Pma-fliQ	NT KV8281 with SOE using primers 2290 and 2090 (KV8293), 2360 and 2361 (ES114), and 2196 and 1487 (pKV503)	This study
KV8369	ΔflrA::FRT IG::PnrdR-fliQ	NT KV8281 with SOE using primers 2290 and 2090 (KV8294), 2360 and 2361 (ES114), and 2196 and 1487 (pKV503)	This study
KV8370	ΔflrA::FRT IG::Pcel-fliQ	NT KV8281 with SOE using primers 2290 and 2090 (KV8295), 2360 and 2361 (ES114), and 2196 and 1487 (pKV503)	This study
KV8376	ΔfliQ::FRT	KV8299 with Cmr cassette removed using pKV496	This study
KV8377	IG::fliQ	NT ES114 with SOE using primers 2185 and 2090 (pKV502), 2360 and 2361 (ES114), and 2196 and 1487 (pKV503)	This study
KV8378	ΔflrA::FRT IG::Plux-RBS-flrA	NT KV8281 with SOE using primers 2290 and 2090 (pKV509) and 2355 and 1487 (KV8290)	This study
KV8380	IG::Pcel-RBS-flrA-HA	NT KV8281 with SOE using primers 2290 and 2220 (KV8366) and 2196 and 1487 (pKV505)
KV8384	ΔflrA::FRT IG::lacIq-Em′ Tmr	NT KV8148 with KV8292 chDNA	This study
KV8397	ΔfliQ::FRT IG::fliQ	NT KV8376 with KV8377 chDNA	This study
KV8398	ΔfliQ::FRT IG::Pma-fliQ	NT KV8376 with KV8368 chDNA	This study
KV8399	ΔfliQ::FRT IG::PnrdR-fliQ	NT KV8376 with KV8369 chDNA	This study
KV8400	ΔfliQ::FRT IG::Pcel-fliQ	NT KV8376 with KV8370 chDNA	This study
KV8407	ΔfliQ::FRT bcsA::Tn5	NT KV8376 with KV4607 chDNA	This study
KV8408	bcsA::Tn5	NT ES114 with KV4607 chDNA	This study
KV8411	ΔfliQ::FRT ΔsypQ::FRT-Emr	NT KV8376 with KV8191 chDNA	This study
KV8417	ΔflrA::FRT IG::Pcel-RBS-flrA-HA	NT KV8281 with KV8380 chDNA	This study
KV8430	ΔfliQ::FRT-Zcr	NT ES114 with SOE using primers 2356 and 2357 (ES114), 2089 and 2090 (pKV520), and 2358 and 2359 (ES114)	This study
KV8435	ΔflrA::FRT IG::lacIq Pma-RBS-flrA	NT KV8384 with KV8364 chDNA	This study

^aIG, intergenic region between yeiR and glmS; except for KV6576 and where indicated as Em′ Tm^r, all strains showing an IG designation contain the Em^r marker.

^bNatural transformation (NT) was performed using tfoX-expressing derivatives of the indicated strains; for the indicated PCR or PCR SOE reactions, templates are indicated in parentheses. chDNA, chromosomal DNA; NA, not applicable.

Plasmid construction.

Plasmids used or generated in this study are listed in Table 4. To generate the tools used in this study, PCR products were cloned into the pJET plasmid using the CloneJET PCR cloning kit (Fisher Scientific) according to the manufacturer's instructions and as described in Table 4. In some cases, PCR products were purified using the DNA Clean & Concentrator kit (Zymo Research) or extracted from a gel using a Zymoclean gel DNA recovery kit (Zymo Research). To generate pKV496, the flp gene was amplified using primers 2116 and 2117, using pCP20 as a template, and cloned into pJET. The resulting clone was digested with KpnI and SacI and ligated to KpnI/SacI-digested plostfoX-Kan (37), resulting in the insertion of flp in place of tfoX. Either plasmid pKV494, which contains the Em^r cassette flanked by BamHI, FRT, and linker sequences, or pKV495 (the Cm^r equivalent) was used as the base plasmid for the generation of other linker- and FRT-flanked cassettes with different antibiotic resistance markers. Plasmid pKV494 (or pKV495) was cut with BamHI, as were the donor pJET derivatives that contained BamHI-flanked antibiotic resistance cassettes. The resulting DNA fragments were gel purified with a Zymoclean gel DNA recovery kit (Zymo Research) and ligated using T4 DNA ligase (Promega).

TABLE 4.

Plasmids used in this study

Plasmid	Description	Derivationa	Reference
pCP20	flp+, Apr	NA	61
pEVS170	Vector, Emr	NA	62
pJMO13	lacIq, Apr	NA	50
pKV27	pVO8 + luxI	pVO8 SmaI + 1.8-kb EcoRV (luxI+) from pHV200 (63)	This study
pKV31	pVO8 + luxR	pVO8 SalI/SmaI + 1-kb SalI/PvuII (luxR+) from pHV200 (63)	This study
pKV69	Vector, Tcr, Cmr	NA	64
pKV494	pJET + FRT-Emr	pJET + SOE product generated with primers 2097 and 2098 (pEVS170)	This study
pKV495	pJET + FRT-Cmr	pKV494 BamHI + BamHI-digested pJET-Cmr generated with primers 2087 and 2088 (pKV69)	This study
pKV496	pEVS79-Knr + flp+	plostfoX-Kan KpnI/SacI + KpnI/SacI-digested pJET-flp+ generated with primers 2116 and 2117 (pCP20)	This study
pKV502	pJET + yeiR-FRT-Emr	pJET + SOE product generated with primers 2185 and 2186 (ES114) and 2187 and 2188 (pKV494)	This study
pKV503	pJET + glmS	pJET + PCR product generated with primers 2191 and 1487 (ES114)	This study
pKV505	pJET + HA-glmS	pJET + PCR product generated with 2198 and 1487 (pKV503)	This study
pKV506	pJET + yeiR-FRT-Emr-PnrdR	pJET + SOE product generated with primers 2185 and 2240 (pKV502) and 2189 and 2190 (ES114)	This study
pKV509	pJET + yeiR-FRT-Emr-Plux	pJET + SOE product generated with primers 2185 and 2240 (pKV502) and 2260 and 2261 (ES114)	This study
pKV510	pJET + Tcr	pJET + PCR product generated with primers 2159 and 2160 (pKV69)	This study
pKV514	pJET + FRT-Tcr	pKV495 BamHI + BamHI-digested pKV510	This study
pKV518	pJET + yeiR-FRT-Emr-Pcel	pJET + SOE product generated with primers 2185 and 2240 (pKV502) and 2251 and 2253 (ES114)	This study
pKV519	pJET + yeiR-FRT-Emr-Pma	pJET + SOE product generated with primers 2185 and 2240 (pKV502) and 2185 and 2327 (pKV506)	This study
pKV520	pJET + FRT-Zcr	pKV494 BamHI + BamHI-digested pJET-Zcr generated with primers 2343 and 2344 (Zcr)	This study
pKV521	pJET + FRT-Spr	pKV494 BamHI + BamHI-digested pJET-Spr generated with primers 2345 and 2346 (Spr)	This study
plostfoX	tfoX+, Cmr	NA	36
plostfoX-Kan	tfoX+, Knr	NA	37
pMLC1	pJET + Tmr	pJET + PCR product generated with primers 2278 and 2279 (Tmr)	This study
pMLC2	pJET + FRT-Tmr	pKV495 BamHI + BamHI-digested pMLC1	This study
pVO8	Vector, Emr, Cmr	NA	53

^aDerivation of plasmids generated in this study.

PCRs.

PCR was carried out using primers listed in Table 5. For making insertions and deletions, KOD high-fidelity polymerase (Fisher Scientific) was used per the manufacturer's instructions, with an annealing temperature of 55°C. The resulting products were examined using agarose gel electrophoresis, staining with ethidium bromide, and visualization with UV light. PCR products were separated from primers and other components of the PCR using a DNA Clean & Concentrator kit (Zymo Research) and quantified using a NanoDrop instrument. To splice PCR products together, 100 to 200 ng of each product was used as the templates in a PCR SOE reaction in the absence of primers for 20 cycles. An aliquot (5 to 10 μl) from the SOE reaction mixture was used in a subsequent amplification reaction in the presence of the outermost primers. In some cases, the concentration of primers was reduced from the recommended 0.4 μM to obtain a cleaner full-length product. PCRs were also used to evaluate the deletion or insertion of sequences of interest using whole cells or chromosomal DNA that was extracted using either the DNeasy blood and tissue kit (Qiagen) or the Quick-DNA miniprep plus kit (Zymo Research; purchased from Genesee) with Promega GoTaq Flexi DNA polymerase (Fisher Scientific) per the manufacturer's instructions.

TABLE 5.

Primers used in this study

^aLowercase letters indicate nonnative or tail sequences.

Natural transformation.

Natural transformation was carried out using established protocols (36, 37) with PCR products or chromosomal DNA. Briefly, V. fischeri strains carrying plostfoX (36) or plostfoX-Kan (37) were grown overnight in TMM containing Cm or Kan, respectively, subcultured into the same medium, and grown at 24°C to an optical density at 600 nm (OD₆₀₀) of between 0.2 and 1.1. Cells were incubated with DNA for 30 min at room temperature, and then 0.5 ml of LBS broth was added. The cultures were incubated with shaking for at least 90 min, and then 200-μl aliquots were spread onto LBS plates containing the appropriate antibiotics. In some cases, a second aliquot was spread onto plates following overnight incubation of the cultures statically at room temperature.

Conjugation.

Plasmids, such as pKV496 or plostfoX-Kan, were introduced into strains of interest using triparental conjugation as previously described (59).

Bioinformatics.

The locations of putative promoters were predicted using BPROM (60).

Motility assays.

Bacteria were grown in LBS at 28°C overnight and standardized to a uniform optical density at 600 nm (OD₆₀₀) of between 0.2 and 0.3. Aliquots (10 μl) were inoculated on the surface of fresh soft-agar motility plates and incubated at 28°C. Measurements of the outer diameter of the band of swimming cells were taken over time. Error bars represent standard deviations from three biological replicates; when no error bars can be observed, the error bars are smaller than the symbol. Data are representative of experiments performed on at least two separate days. Images were taken with a Canon XS170 IS camera.

Congo red assays.

Bacteria were grown overnight at 24°C on LBS plates containing Congo red and Coomassie blue as indicated above. To better visualize color differences of the resulting growth, cells were transferred onto white paper in a replica plating-like approach by briefly smoothing the paper onto the agar plate and then lifting it off. The result was photographed with a Canon XS170 IS camera.

Western immunoblotting.

V. fischeri cells were grown overnight in LBS lacking or containing cellobiose (0.01% or 0.1%). Cultures were normalized to an OD₆₀₀ of 3. One-ml aliquots were pelleted and lysed in 200 μl 2× sample buffer (4% SDS, 40 mM Tris, pH 6.3, 10% glycerol). Proteins were separated by SDS-PAGE (12% acrylamide) and transferred to a polyvinylidene difluoride (PVDF) membrane. After blocking with dry milk in PBS-T (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.47 mM KH₂PO₄, and 0.05% Tween 20), the membrane was treated with rabbit anti-HA antibody (Sigma-Aldrich), followed by exposure to a secondary antibody, goat anti-rabbit IgG antibody (Fisher Scientific) conjugated to horseradish peroxidase. Finally, to visualize HA-tagged FlrA, the membrane was incubated with SuperSignal West Pico plus chemiluminescent substrate (Thermo Fischer Scientific, Rockford, IL) and exposed to autoradiography film (Dot Scientific), which was developed in an autoprocessor.