|Home | About | Journals | Submit | Contact Us | Français|
Regulation of gene expression is critical for the ability of Borrelia burgdorferi to adapt to different environments during its natural infectious cycle. Reporter genes have been used successfully to study gene regulation in multiple organisms. We have introduced a lacZ gene into B. burgdorferi, and we show that B. burgdorferi produces a protein with detectable β-galactosidase activity in both liquid and solid media when lacZ is expressed from a constitutive promoter. Furthermore, when lacZ is expressed from the ospC promoter, β-galactosidase activity is detected only in B. burgdorferi clones that express ospC, and it accurately monitors endogenous gene expression. The addition of lacZ to the repertoire of genetic tools available for use in B. burgdorferi should contribute to a better understanding of how B. burgdorferi gene expression is regulated during the infectious cycle.
Borrelia burgdorferi sensu lato, the pathogen that causes Lyme disease (7), alternates between two distinct environments, an arthropod vector and a vertebrate host. As B. burgdorferi moves from one milieu to the other, its ability to adapt and survive requires dramatic changes in gene expression. Many studies have shown that different B. burgdorferi gene products are upregulated or downregulated at specific times during the infectious cycle (19, 31) and in response to host and environmental signals (6, 8a, 15, 24, 25). Although it is clear that B. burgdorferi alters gene expression to adapt to different environments, the genetic tools for studying gene regulation in B. burgdorferi are limited.
Within the last 2 decades, the complete genomic sequence of B. burgdorferi strain B31 was published (10, 14) and techniques for basic genetic manipulation of B. burgdorferi became available (5, 11, 13, 27-29, 36). A chloramphenicol acetyltransferase (CAT) gene was the first reporter gene that was fused to B. burgdorferi promoters for analysis of promoter strength (33). The development of luciferase (4) and multiple fluorescent proteins (9, 11, 30) as reporter systems in B. burgdorferi followed. Although these systems have value, there are limitations with each. β-Galactosidase, encoded by lacZ, has been used extensively as a convenient reporter gene in Escherichia coli and is still applicable to a broad range of organisms, both prokaryotic and eukaryotic, but has not yet been used with B. burgdorferi. β-Galactosidase activity can be monitored easily and quickly by simple colorimetric assays in both liquid and solid media, neither of which require expensive or specialized equipment. Additionally, a wide variety of substrates for β-galactosidase allow for different levels of sensitivity in either in vitro or in vivo detection formats (17). Having lacZ available as a genetic tool for B. burgdorferi would enhance investigation of the complex regulatory events that are integral to the spirochete's infectious cycle. To this end, we developed lacZ as a reporter gene in B. burgdorferi and demonstrated its utility.
Plasmids for cloning were transformed into electrocompetent or chemically competent Top10 E. coli (Invitrogen, Carlsbad, CA), and final constructs were transformed into Δlac E. coli strain MC4100 lamB-zjb::Tn10 (kindly provided by John Carlson, Rocky Mountain Laboratories, Hamilton, MT). E. coli cells were plated on LB agar with the appropriate antibiotic (spectinomycin at 300 μg/ml, kanamycin at 30 μg/ml, or gentamicin at 5 μg/ml). Liquid cultures were grown in LB broth supplemented with the appropriate antibiotic (spectinomycin or kanamycin at 100 μg/ml or gentamicin at 10 μg/ml).
BSK II medium (2) for culture of B. burgdorferi was made with CMRL 1060 lacking phenol red (US Biological, Swampscott, MA). B. burgdorferi strain B31 clones A34 (16) and B312 (26) were grown to mid-log phase (~5 × 107cells/ml). Spirochetes were enumerated using dark-field microscopy and a Petroff-Hausser counting chamber and prepared for electroporation as described previously (27). Transformed B31-A34 and B312 cells were selected using gentamicin at a concentration of 40 μg/ml. All B. burgdorferi cultures were grown at 35°C, and plates were incubated under 2.5% CO2.
The E. coli lacZ gene (lacZEc), the multiple cloning site (MCS), and the transcriptional terminator were amplified from pPBMB101 (8) using primers 1 and 2 (Table (Table2).2). The PCR product was cloned into the Gateway entrance vector pCR8/GW/TOPO (Invitrogen) and confirmed by direct sequencing with internal primers (Table (Table2,2, primers 1 through 10). pBSV2G_dvB2, kindly provided by James A. Carroll (Rocky Mountain Laboratories, Hamilton, MT), is an altered form of the B. burgdorferi shuttle vector pBSV2G, which has attR1 and attR2 sequences surrounding a chloramphenicol resistance cassette and the counterselectable gene ccdB, making it a suitable destination vector for Gateway cloning. The insert containing lacZEc, the MCS, and the transcriptional terminator was transferred from pCR8 to pBSV2G_dvB2 by using Clonase II enzyme (Invitrogen), creating pBH-lacZEc (Fig. (Fig.1A1A).
The flaB promoter was amplified from B. burgdorferi genomic DNA using primers 11 and 12 (Table (Table2).2). The PCR fragment was cloned into pCR2.1 (Invitrogen) and sequenced to confirm the insert. The flaB promoter was excised with BamHI and XhoI and ligated into appropriately digested pBH-lacZEc, creating pBHflaBp-lacZEc (Fig. (Fig.1A).1A). BamHI and XhoI were the enzymes used for creating promoter-lacZ fusions in pPBMB101, which has a ribosome binding site (RBS) between the XhoI site and the lacZ start codon (8).
pBH-lacZEc and pBHflaBp-lacZEc were transformed into electrocompetent B. burgdorferi cells (27). Spirochetes were allowed to recover overnight in BSK II medium without antibiotic selection and then plated in solid BSK medium supplemented with gentamicin. Individual colonies were screened by PCR for lacZEc and the gentamicin cassette.
E. coli lacZ codon usage in B. burgdorferi was analyzed using the Graphical Codon Usage Analyser version 2.0 (www.gcua.schoedl.de). More than one-third of the codons in the E. coli lacZ gene, lacZEc, are considered rare (used less than 20% of the time) in B. burgdorferi. A codon-optimized version of lacZ, lacZBb, which uses less than 1% of B. burgdorferi rare codons, was synthesized by GenScipt Corporation (Piscataway, NJ). A sequence alignment of the codon-optimized lacZBb gene with the original lacZEc gene is provided in Fig. S1 of the supplemental material. The synthesized gene was preceded by the same MCS and RBS as the original lacZEc. lacZBb (nucleotide sequence available upon request) was cloned into pBSV2G (13) with XbaI and KpnI, producing pBH-lacZBb (Fig. (Fig.1B).1B). The flaB promoter was added as described above to create pBHflaBp-lacZBb (Fig. (Fig.1B).1B). The intergenic region between guaA and ospC, which includes the ospC promoter and regulatory operator sequences up to the ospC start codon (12, 20, 21, 37-39), was amplified from B. burgdorferi genomic DNA with primers 13 and 14 (Table (Table2),2), cloned into pCR2.1, sequenced, and digested with BamHI and XhoI for ligation into pBH-lacZBb, creating pBHospCp-lacZBb. Constructs were transformed into electrocompetent B. burgdorferi (27) and Δlac E. coli cells.
For β-galactosidase assays, aliquots of 1 ml of overnight E. coli cultures and 5 to 10 ml of mid-log B. burgdorferi cultures were washed twice in HEPES-NaCl (HN) buffer (50 mM each; pH 7.6). Bacterial pellets were resuspended in Z-buffer (36 mM NaH2PO4, 67 mM NaHPO4, 0.1 mM MgCl2, 2 mM MgSO4, 2.7 ml/liter β-mercaptoethanol) at approximately 5 × 108 bacteria/ml before lysis with chloroform and SDS. β-Galactosidase assays were performed using a modified version of the Miller protocol (22). Briefly, aliquots (10 to 20 μl) of bacterial lysates were added to 96-well plates (Costar, Corning, NY) in triplicate. Z-buffer was added, increasing the volume to 160 μl. Fifty microliters of o-nitrophenyl-β-d-galactopyranoside (ONPG) dissolved in Z-buffer (4 mg/ml) was added, and the plate was incubated at room temperature for 10 to 15 min. After incubation, 90 μl of stop buffer (1 M Na2CO3) was added, and the absorbance was measured at 405 nm (Labsystems Multiskan Plus; Fisher Scientific, Pittsburgh, PA). β-Galactosidase activity units, or Miller units (nmoles/minute), were calculated as described before and reported as units per mg of protein (23). Background activity (units/mg of protein) of bacteria lacking a shuttle vector was subtracted from reported values.
Both E. coli and B. burgdorferi were grown on plates containing 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal; Roche Diagnostics, Indianapolis, IN) at 0.08 g/liter and gentamicin. Alternatively, ~0.5 ml X-Gal dissolved in dimethyl sulfoxide (DMSO; 20 mg/ml) was spread on B. burgdorferi plates after colony formation.
The concentration of protein in the bacterial lysates used for β-galactosidase assays was determined using the Bio-Rad protein assay (Hercules, CA) according to the manufacturer's protocol. Dilutions of bovine serum albumin (New England Biolabs [NEB], Ipswich, MA) were used as standards. Aliquots of 10 μl of the standards and 5 to 10 μl bacterial lysates were added to 96-well plates in triplicate. A standard curve was generated and used to determine protein concentrations of the bacterial lysates.
RNA was harvested from ~30 ml of a mid-log B. burgdorferi culture using a Masterpure RNA preparation kit (Epicentre Biotechnologies, Madison, WI) according to the manufacturer's protocol. Two to 5 μg of RNA was used as a template to synthesize cDNA using the high-capacity cDNA reverse transcriptase (RT) kit (Applied Biosystems, Life Technologies, Carlsbad, CA). Quantitative-PCR (q-PCR) was performed in triplicate on 100 ng cDNA using the TaqMan Universal PCR Mastermix (Applied Biosystems) and primer-probe combinations for lacZBb, flaB, and ospC (Table (Table2).2). Real-time q-PCR was performed using the ABI Prism 7900 HT sequence detection system.
To investigate whether lacZ could be used as a simple reporter gene in B. burgdorferi, we constructed four shuttle vectors, pBH-lacZEc, pBHflaBp-lacZEc, pBH-lacZBb, and pBHflaBp-lacZBb (see Materials and Methods and Fig. Fig.1).1). These constructs carry an E. coli lacZ gene (lacZEc) or a B. burgdorferi codon-optimized lacZ gene (lacZBb), with and without the constitutive B. burgdorferi flaB promoter. The lacZEc gene has 56% G+C base content, while the B. burgdorferi genome has only 28% G+C content (14). Previous studies have used codon optimization to express foreign genes in B. burgdorferi with much success (4, 13). Thus, we designed a synthetic lacZ gene, lacZBb, with reduced G+C content to better reflect the B. burgdorferi codon preference in order to enhance β-galactosidase synthesis in B. burgdorferi. The nucleotide sequences of lacZBb and lacZEc are aligned for comparison in Fig. S1 of the supplemental material. The flaB promoter was chosen for initial experiments because it is constitutively expressed in B. burgdorferi and E. coli and has been used to drive expression of antibiotic resistance markers in B. burgdorferi (5, 13, 33). We transformed all constructs into electrocompetent B. burgdorferi B31-A34 and a Δlac E. coli strain and assessed β-galactosidase activity in liquid and solid media.
In order to assess the β-galactosidase activity in B. burgdorferi and E. coli transformants carrying the lacZ constructs, we adapted the well-established Miller ONPG assay (22) and normalized β-galactosidase activity to the protein content of the bacterial lysates as described by Nielsen et al. (23). Both E. coli and B. burgdorferi harboring pBHflaBp-lacZEc or pBHflaBp-lacZBb displayed significant activity compared to bacteria containing the promoterless lacZ constructs (Fig. (Fig.2).2). β-Galactosidase activity was significantly higher in B. burgdorferi lysates containing pBHflaBp-lacZBb than in lysates harboring pBHflaBp-lacZEc (Fig. (Fig.2).2). While we cannot rule out the possibility that flipping the orientation of lacZ in the vector influenced expression, the same plasmids used in E. coli gave the opposite pattern (higher activity in bacteria containing pBHflaBp-lacZEc versus pBHflaBp-lacZBb), suggesting that the codon preferences of the respective bacteria produced the discrepancies in activity levels. B. burgdorferi harboring either pBHflaBp-lacZBb or pBHflaBp-lacZEc produced detectable β-galactosidase protein when assayed by immunoblotting (data not shown). These results demonstrate that lacZ produces a functional β-galactosidase enzyme in B. burgdorferi when expressed using the constitutive flaB promoter.
Since β-galactosidase activity could be detected in B. burgdorferi lysates, we wanted to determine if lacZ could be used as a reporter gene for B. burgdorferi colonies in solid medium. Typically, addition of X-Gal to solid medium allows for the detection of bacterial colonies that possess an active β-galactosidase enzyme, which cleaves X-Gal into a blue product. Therefore, we used BSK solid medium without phenol red to easily assay for any color change. When X-Gal was included in BSK solid medium (0.08 g/liter), blue color development was observed 2 to 5 days after colony formation (data not shown) for B. burgdorferi colonies harboring either pBHflaBp-lacZEc or pBHflaBp-lacZBb. However, the incorporation of X-Gal into the medium considerably slowed colony growth. The production of β-galactosidase was not intrinsically toxic, since colonies that produced β-galactosidase grown without X-Gal were of normal size and morphology. Interestingly, this growth defect seemed to be caused by the cleavage of X-Gal, as X-Gal or solvent incorporation into the medium did not affect colony formation in B. burgdorferi lacking a lacZ gene. Still, when 2% X-Gal in DMSO was spread on B. burgdorferi plates after colony formation, color development in colonies harboring the lacZ gene driven by the flaB promoter was evident within 15 min and continued to increase overnight as the substrate diffused throughout the plates (Fig. (Fig.3).3). Importantly, colonies without a promoter to drive lacZ remained unchanged (white) even after overnight incubation (Fig. (Fig.3).3). Furthermore, live spirochetes were recovered from colonies 48 h after X-Gal treatment, indicating that X-Gal treatment did not hinder bacterial viability. Colonies of B. burgdorferi harboring the optimized lacZ gene, lacZBb, when expressed by the flaB promoter developed a more uniform and greater color intensity than B. burgdorferi containing pBHflaBp-lacZEc, again suggesting that the optimized gene is better suited for B. burgdorferi (Fig. (Fig.3).3). These results demonstrated that lacZ can be used to monitor gene expression of B. burgdorferi colonies on plates. Furthermore, the codon-optimized B. burgdorferi lacZ gene (lacZBb) increased β-galactosidase activity, presumably through increased protein production. Thus, we used only the B. burgdorferi optimized lacZ gene (lacZBb) in further studies.
We demonstrated β-galactosidase protein production and activity in B. burgdorferi by expressing lacZ from a constitutive promoter, but we wished to confirm that lacZ expression reflects the expression of regulated genes in B. burgdorferi and therefore can be used as a reporter gene for transcriptional activity. For this objective, we used two strains of B. burgdorferi that differ in expression of the ospC gene: B. burgdorferi clones B31-A34 and B312. B31-A34 is derived from noninfectious clone B31-A (18) and does not produce OspC in vitro, while B312, a highly attenuated clone of B31 that lacks many plasmids, has been reported to produce abundant amounts of OspC in vitro (26). In order to determine lacZ expression driven by the ospC promoter in these two B. burgdorferi genetic backgrounds, we created pBHospCp-lacZBb. We transformed B31-A34 and B312 with pBHospCp-lacZBb and also introduced pBH-lacZBb and pBHflaBp-lacZBb into B312.
While performing these experiments, we noticed substantial variation in OspC synthesis among B312 transformants and between independent outgrowths of the same B312 derivative (data not shown). This observation has not been reported for B312 previously, although earlier studies did report variation in ospC expression on an individual cell basis in other B. burgdorferi backgrounds (12, 35). In order to use B312, we first confirmed OspC production in B312 transformants with immunoblotting before proceeding with further experiments on those clones (data not shown). Significant β-galactosidase activity was detected in B312 harboring the pBHospCp-lacZBb plasmid compared to clones with the promoterless construct (Fig. (Fig.4A).4A). In contrast, β-galactosidase activity was not detected in B31-A34 when the ospC promoter was used to express lacZBb (Fig. (Fig.4A),4A), consistent with the lack of OspC protein production in this strain. Notably, in B312 β-galactosidase activity controlled by the flaB promoter was 40-fold higher than activity from the ospC promoter (Fig. (Fig.4A).4A). Thus, we investigated the transcript levels of ospC and lacZBb in B312, normalizing the levels to flaB transcript levels. Endogenous ospC transcript levels in B312 lacking a shuttle vector and B312 transformants were roughly half of flaB transcript levels (56 ospC transcripts for every 100 flaB transcripts) (Fig. (Fig.4B,4B, solid blue bar). Similarly, lacZBb transcript levels in B312 were dramatically lower when controlled by the ospC promoter than by the flaB promoter (Fig. (Fig.4B,4B, hatched blue bars). The increased activity and transcript levels from the flaB promoter compared to the ospC promoter are consistent with previous results from transient CAT assays in B. burgdorferi B31 (33) and demonstrate different strengths of the flaB and ospC promoters. Additionally, since FlaB and OspC protein levels were similar when assayed by immunoblotting or Coomassie blue staining, these results also suggest posttranscriptional regulation in B. burgdorferi; further studies are needed to investigate this possibility. Notably, the lacZBb transcript levels driven by either the flaB or ospC promoters were 5- to 10-fold higher than the endogenous gene transcripts, presumably reflecting the difference in copy number between the shuttle vector (lacZBb) and genome (flaB or ospC) (3). This increase in gene expression from the shuttle vector may be beneficial when analyzing lacZ expression from weak B. burgdorferi promoters.
We also assessed β-galactosidase activity controlled from the ospC promoter in B. burgdorferi colonies grown on BSK solid medium without phenol red. As expected, colonies of B31-A34 harboring pBHospCp-lacZBb remained white after overnight treatment with X-Gal (Fig. (Fig.4C),4C), indicating these clones do not express ospC or lacZBb when under the control of the ospC promoter and thus did not synthesize the β-galactosidase protein. Even when OspC production was verified by immunoblotting and Coomassie blue staining (data not shown), not all of the B312 colonies harboring pBHospCp-lacZBb turned blue after X-Gal treatment, suggesting that ospC expression varied at the individual colony level (Fig. (Fig.4D).4D). This variation in expression of β-galactosidase appeared to be due to stochastic differences in expression of the ospC promoter in this genetic background rather than an intrinsic property of the B312 strain, as B312 colonies harboring the constitutively expressed pBHflaBp-lacZBb plasmid did not vary in color intensity (Fig. (Fig.4E).4E). Additionally, plasmid loss was not responsible for the variation seen with pBHospC-lacZBb, since at least 96% of colonies grown without antibiotic selection still contained the shuttle vector (data not shown). Together, these results demonstrate that β-galactosidase activity and lacZBb gene expression accurately reflect promoter strength and regulation of gene expression in B. burgdorferi (Fig. (Fig.44).
One advantage of using reporter gene fusions to assess the transcriptional activities of various promoters is that it eliminates differences in posttranscriptional regulation that may affect endogenous protein levels. The reporter gene transcript and encoded protein are the same irrespective of which promoter is used to express it. The lacZ gene is one of the most widely used reporter genes, and B. burgdorferi promoters have previously been fused to lacZ, but β-galactosidase activity was assessed in E. coli, not B. burgdorferi (8, 32). However, since many studies have detected significant differences in the expression profiles of B. burgdorferi and E. coli (1, 12, 33, 34), using lacZ as a reporter gene directly in B. burgdorferi will allow a more accurate analysis of native promoter activity. Additionally, there are many substrates for β-galactosidase, and this allows lacZ to be used in multiple applications, including simple in vitro assays to assess promoter activity and rapid screening for transformants on solid media. Thus, adding lacZ to the repertoire of genetic tools for B. burgdorferi should facilitate investigation of the regulatory mechanisms that allow B. burgdorferi to adapt to different environments.
We thank Mary Burtnick, Paul Brett, Frank Gherardini, John Carlson, and Jay Carroll for providing plasmids and strains. We thank Gary Hettrick and Anita Mora for assistance with graphics. We also thank members of the Rosa lab, Tom Schwan, Sonja Best, and Frank Gherardini for critical review of the manuscript.
This research was supported by the Intramural Research Program of the NIAID, NIH.
Published ahead of print on 17 September 2010.
†Supplemental material for this article may be found at http://aem.asm.org/.
‡The authors have paid a fee to allow immediate free access to this article.