|Home | About | Journals | Submit | Contact Us | Français|
Our knowledge of the genetics and molecular basis of the pathogenesis associated with Leptospira, in comparison to those of other bacterial species, is very limited. An improved understanding of pathogenic mechanisms requires reliable genetic tools for functional genetic analysis. Here, we report the expression of gfp and mRFP1 genes under the control of constitutive spirochetal promoters in both saprophytic and pathogenic Leptospira strains. We were able to reliably measure the fluorescence of Leptospira by fluorescence microscopy and a fluorometric microplate reader-based assay. We showed that the expression of the gfp gene had no significant effects on growth in vivo and pathogenicity in L. interrogans. We constructed an expression vector for L. biflexa that contains the lacI repressor, an inducible lac promoter, and gfp as the reporter, demonstrating that the lac system is functional in Leptospira. Green fluorescent protein (GFP) expression was induced by the addition of isopropyl-β-d-thiogalactopyranoside (IPTG) in L. biflexa transformants harboring the expression vector. Finally, we showed that GFP can be used as a reporter to assess promoter activity in different environmental conditions. These results may facilitate further advances for studying the genetics of Leptospira spp.
The genus Leptospira belongs to the order Spirochaetales and includes both saprophytic and pathogenic members, such as Leptospira biflexa and L. interrogans, respectively. Leptospirosis is the most widespread zoonosis worldwide, with more than one million severe cases annually (17, 21). This increasingly common disease occurs mostly in rural environments and poor urban centers subject to frequent flooding. Rodents are the main reservoir of the disease, excreting the bacteria in their urine (17, 21). Humans are usually infected through contact with water contaminated with the urine of infected animals.
Although our group has developed a number of tools for genetic manipulation of Leptospira in recent years (6, 11, 25), fewer tools are available for genetic studies of Leptospira than for various other bacteria. Further development and improvement of genetic tools are therefore necessary to improve understanding of the pathogenic mechanisms of Leptospira.
Green fluorescent protein (GFP) and its variants have become valuable tools in molecular biology. One advantage of GFP is that its autofluorescence does not require any cofactors for expression, enabling its detection in single cells and on agar plates. GFP was originally obtained from the jellyfish Aequorea victoria and has an excitation peak at 395 nm and a smaller peak at 475 nm. There are many derivatives of this wild-type GFP, which have increased levels of fluorescence emission and shifted excitation or emission spectra. One of these GFP variants, GFPuv, appears to have a higher fluorescence emission because it is more soluble than wild-type GFP (10). Site-directed mutagenesis of wild-type GFP has also been used to create F64L and S65T mutations to produce a series of GFPmut derivatives that have a red-shifted excitation spectrum (excitation maximum, 488 nm) giving them characteristics close to those of the fluorophore fluorescein isothiocyanate (FITC) (9). GFP mutants with blue, cyan, and yellowish-green emission spectra are also available (33). Another fluorescent protein, DsRed, originally isolated from corals, has an excitation maximum around 560 nm and an emission maximum around 580 nm; these values are significantly different from those for GFP fluorescence, and therefore DsRed can be used in conjunction with GFP. A disadvantage of wild-type DsRed is that it is tetrameric and matures more slowly than GFP. However, more rapidly maturing monomeric variants, for example, monomeric red fluorescent protein (mRFP1), have been developed (7). GFP and other fluorescent proteins have been used as reporters for studies of gene expression, protein localization, and bacterial localization and activities in infected animal/plant tissues. Many plasmids and transposons expressing the gfp gene have been used to label a broad variety of eukaryotic and prokaryotic cells. In this study, we generated replicative plasmid constructs containing constitutive and inducible promoters fused to genes encoding GFP and mRFP1. Plasmids were introduced into the saprophyte L. biflexa, and expression of fluorescent proteins in transformants was evaluated by epifluorescence microscopy and fluorometric microplate reader-based assay. Using a Himar1 transposon delivery system, we have also labeled the chromosome of the pathogen L. interrogans with gfp. These new genetic tools may help investigations of the virulence, and more generally the biology, of Leptospira.
Leptospira strains were cultivated in liquid Ellinghausen-McCullough-Johnson-Harris (EMJH) medium (13, 14) or on 1% agar plates at 30°C and counted in a Petroff-Hausser counting chamber (Fisher Scientific). The saprophyte Leptospira biflexa serovar Patoc strain Patoc I and the pathogens L. interrogans serovar Copenhageni strain Fiocruz L1-130 (16, 27) and L. interrogans serovar Lai strain Lai 56601 (30) were used. Escherichia coli was grown in Luria-Bertani (LB) medium. When appropriate, spectinomycin or kanamycin was added to the culture media at 50 μg ml−1. Leptospira strains containing the inducible system were grown with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG).
For infection experiments, 3-week-old Hartley male guinea pigs (Charles River Laboratories) or hamsters were inoculated intraperitoneally with 1 ml of various doses of L. interrogans as previously described (31). Protocols for animal experiments were prepared according to the guidelines of the Animal Care and Use Committees of the Institut Pasteur.
L. interrogans flgB and hsp10 promoters were amplified with FlgA/FlgC and HspA/HspC primer pairs, respectively, and inserted into the PvuII restriction site of the E. coli-L. biflexa shuttle vector pSLe94 (3) to generate plasmids pSLe94PF and pSLe94PH, respectively. The gfpuv (10) and gfpmut (23) alleles were amplified with flanking BglII and SmaI sites, using primers GfpC5 and GfpC3 (Table (Table1),1), from plasmids pSRKgfp (15) and pTM61 (24), respectively. The amplified gfp alleles were then digested with BglII and SmaI, purified, and inserted between the BamHI and SmaI restriction sites of pSLe94PF and pSLe94PH to generate pFA1/pFA4 and pFA2/pFA3, respectively. Similarly, the gene encoding the red fluorescent protein mRFP1 was amplified with primers MC5 and MC3 (Table (Table1)1) from pRSETb-mRFP1 (7). The PCR products were purified, digested with BamHI and XhoI, and inserted between the corresponding sites in pSLe94PF and pSLe94PH to generate pFA5 and pFA6 (Table (Table22).
To use a transposon delivery system, gfpuv and gfpmut were amplified from plasmids pSRKgfp (15) and pTM61 (24), respectively, with primer pairs FA/GfpAsc and FA/GAbis (Table (Table1).1). The PCR products were then purified, digested with AscI, and inserted into the AscI restriction site of the kanamycin-resistant transposon carried by pMKL (31), to generate pFA7 and pFA8.
To construct an inducible system, the DNA fragment containing the Borrelia burgdorferi flaB promoter and the lacI repressor gene was amplified from pJSB104 (4) with primers PFlaBF and LacIR and inserted into the PvuII restriction site of pSLe94. The hsp10 promoter was amplified with primers HspA and HspB to introduce a lacO site and a 6×His tag and inserted into pCR2.1 (Invitrogen) to generate pCR-Phsp/lacO. The gfpmut sequence was then amplified by GfpC5 and GfpC3 and inserted into the SmaI and BglII restriction sites of pCR-Phsp/lacO. The gfp allele under the control of Phsp/lacO was then amplified, digested with SmaI, and ligated to SmaI-digested pSLe94, thereby generating pGC1 (Table (Table2).2). Leptospira strains were electroporated with plasmid constructs as previously described (32).
To test for the use of gfp as a reporter gene, the promoterless gfp was amplified from plasmid pTM61 (24), using primers gfp1 and gfp2, digested with SmaI, and then ligated with AleI-digested pSW29TLe94Spc (29), resulting in the conjugative plasmid p-gfp. The promoter regions of hsp20 (LEPBIa1849) and groES (LEPBIa2343) were amplified from L. biflexa chromosomal DNA using primer pairs PLBa1849.1/PLBa1849.2b and PgroES1/PgroES2b, respectively, and then cloned between the AscI and NheI sites of p-gfp. The plasmids were introduced in L. biflexa by conjugation as previously described (29), resulting in strains Patoc HspG and Patoc GroG (Table (Table22).
Five-ml cultures were grown under agitation at 30°C until early to mid-exponential phase (optical density at 420 nm [OD420], 0.2 to 0.4). Half the culture was then transferred to 37°C under agitation for 4 h. For fluorescence measurements, 1 ml of cells was harvested by centrifugation at room temperature for 5 min at 5,000 × g, washed in phosphate-buffered saline (PBS), and resuspended in 200 μl of the same buffer. Then, 100 μl of each sample was aliquoted in duplicate in 96-well microplates and fluorescence was read as described below. The OD420 of the aliquots was then measured and used to normalize the fluorescence.
For RNA isolation and real-time reverse transcriptase PCR (RT-PCR) assay, 1 ml of cells was added to 3 ml of Trizol LS (Invitrogen) and RNA was purified according to the manufacturer's guidelines. Contaminating DNA was removed from RNA preparations using DNase I from Roche, and RNA was subsequently cleaned up using the RNeasy kit (Qiagen). cDNA was synthesized with the iScript kit (Bio-Rad), and 1/10 of the reaction was used for real-time RT-PCR assays with the SsoFast EvaGreen Supermix (Bio-Rad). The PCR mixture contained a 300 nM concentration of each primer in a total volume of 20 μl. The PCRs were performed and analyzed with a CFX96 real-time PCR detection system (Bio-Rad). Ultimately, the amount of cDNA of interest measured in each PCR assay was normalized to the amount of rpoB cDNA.
Leptospira strains carrying the GFP-inducible expression system were cultivated in the presence of spectinomycin until the cell density reached 107 bacteria ml−1. IPTG was then added to a concentration of 1 mM, and the culture was incubated for 1 week at 30°C. Approximately 109 spirochetes (1 ml) were collected and processed for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were then transferred to nitrocellulose membranes and probed with anti-6× histidine-tag antibody (mouse IgG, Clontech). Alkaline phosphatase-conjugated anti-mouse IgG and the 5-bromo-4-chloro-3-indolylphosphate-Nitro Blue Tetrazolium (BCIP-NBT) solution substrate (Uptima, France) were used to detect the bound primary antibodies.
Cells were washed in PBS, and the optical densities of the suspensions at 600 nm were adjusted to 0.3 before fluorescence measurements were taken. The measurements were performed on a Mithras LB940 microplate reader (Berthold Technologies, Bad Wildbad, Germany). The excitation and emission wavelengths used were 485 nm and 535 nm. The intensity of fluorescence is expressed in relative values (arbitrary units) obtained at emission maxima. Wild-type cells were used as controls to determine background fluorescence. All spectrofluorometer experiments were performed at least twice independently.
The cultures were spun down, and the resulting pellets were washed with 1× PBS to remove the culture medium. The pellets were resuspended in 1× PBS at an appropriate cell density. Aliquots of 10 μl of the bacterial suspensions were applied to poly-l-lysine slides. Images of Leptospira were acquired using an Axioplan 2 imaging microscope (Carl Zeiss) equipped with a Plan-Apochromat 63× 1.4-numerical-aperture (NA) oil objective and the AxioVision software (version 4.6). The GFP and mRFP1 filter sets used were FS44 (excitation, 455 to 495 nm; emission, 505 to 555 nm)/FS09 (excitation, 450 to 490 nm; emission, 515 nm) from Zeiss and XF43 (excitation, 563 to 587 nm; emission, 615 to 645 nm) from Omega Optical.
We chose the E. coli-L. biflexa spectinomycin-resistant plasmid pSLe94 (5) as the backbone for our system. The genes encoding GFPuv (10), a GFP variant optimized for excitation by UV light, GFPmut (23), a GFP variant previously used in the spirochete Borrelia burgdorferi (12, 24), and mRFP1 (7), a red fluorescent protein, were amplified by PCR and inserted into plasmid constructs carrying the flgB (flagellar basal body rod protein FlgB, LA0347) and hsp10 (GroES protein, LA2654) promoters from L. interrogans. The result was a set of plasmid vectors containing either gfp or mRFP1 under the control of the L. interrogans flgB and hsp10 promoters (Fig. (Fig.1).1). These plasmids were then introduced into the saprophyte L. biflexa, and the expression of fluorescent proteins in transformants was evaluated by epifluorescence microscopy (Fig. (Fig.2).2). All plasmid constructs proved to be functional in L. biflexa, such that it was possible to visualize transformants by fluorescence microscopy. In contrast, no fluorescent cells could be seen under the fluorescence microscope in the untransformed recipient wild-type L. biflexa, indicating that background fluorescence is insignificant. The strongest GFP fluorescence was from the gfpmut gene constructs. Transformants expressing the gfpuv allele presented weaker fluorescence as assessed both with the appropriate filter (excitation, 450 to 490 nm; emission, 515 nm) and with the FITC filter (excitation, 455 to 495 nm; emission, 505 to 555 nm). Subsurface colonies on solid medium of L. biflexa transformed with gfp variants were not fluorescent under UV light, whereas E. coli colonies harboring the same plasmids were highly fluorescent (data not shown). Finally, the fluorescence of L. biflexa transformants expressing mRFP1 was detectable by epifluorescence microscopy (data not shown).
As there is no replicative plasmid vector available for pathogenic Leptospira, exogenous genes can only be introduced with a transposon delivery system. The two gfp variants were amplified from the replicative plasmids and inserted into the Himar1 transposon carrying a kanamycin resistance gene (Fig. (Fig.1).1). The constructs were transferred into L. biflexa: the transposition frequency in cells that received the suicide delivery vector was satisfactory, and the transformants were fluorescent (data not shown). Next, we tested transformation of pathogenic L. interrogans strains with the Himar1 constructs. From three transformation experiments, only eight pFA8 transformants (gfpuv+) were obtained with the Lai strain and one pFA7 transformant (gfpmut+) with each of the Lai and Fiocruz strains. The transposon insertion site of each transformant was identified (Table (Table2).2). The gfpuv and gfpmut genes were efficiently expressed in L. interrogans transformants as assessed both by fluorescence microscopy and with a fluorometric plate reader (Fig. (Fig.22 and and3).3). The L. interrogans cells containing gfpmut, strain TKG1, were brightly fluorescent; the fluorescence of Lai TKG1 was stronger than that of Fiocruz TKG1 and even than that of Patoc TKG1 (Fig. (Fig.3).3). We determined the fluorescence of 10-fold serial dilutions of a suspension of GFPmut-transformed L. interrogans. The fluorescence of Lai TKG1 was 18,123, 2,221, and 815 arbitrary fluorescence units (AU) at a density of 108, 107, and 106 Leptospira cells respectively, in a volume of 100 μl. However, the fluorescence was 506 AU at a density of 105 Leptospira cells/100 μl, which is not substantially higher than the background fluorescence of untransformed L. interrogans (184 to 250 AU irrespective of cell density; data not shown).
We tested whether the growth of GFP-expressing Leptospira was impaired by the presence of the plasmid/transposon. We first assessed the growth of the transformed and parental control strains of L. interrogans in EMJH liquid medium. Under the experimental conditions used in this study, expression of the GFP gene had no significant effect on the growth rate of the transformants in liquid media (data not shown). We then compared the virulence of the GFP-transformed strains with that of the wild-type strains in animals by using the guinea pig model of leptospirosis. There was no difference in 50% lethal dose (LD50) values between gfp-transformed and parental strains: <102 Leptospira bacteria for the Fiocruz strain and 108 for the Lai strain (data not shown). Similar disease symptoms developed in animals infected with the GFP-transformed and parental strains, and immunohistochemistry detected similar high numbers of bacteria in both liver and kidney (data not shown). However, direct observation of frozen sections by fluorescence microscopy did not allow the identification of fluorescent bacteria in tissues of infected animals. In conclusion, no significant differences in growth or virulence were observed between the GFP-transformed and wild-type L. interrogans strains, suggesting that the expression of the gfp gene had no significant effects on growth in vivo and pathogenicity.
We used B. burgdorferi lacI (lacIBb) to construct an inducible expression system for L. biflexa; lacIBb is codon optimized to enhance the production of LacI in the spirochete B. burgdorferi (4). We amplified the sequence and inserted it into the E. coli-L. biflexa shuttle vector. A LacI-binding site (operator or lacO) was also introduced into the hsp10 promoter that controls gfp expression, between the −35 −10 regions and the Shine-Dalgarno sequence (Fig. (Fig.4A).4A). The resulting plasmid, pGC1, was then introduced into L. biflexa, and the fluorescence was measured after various times of growth in the presence of 0, 1, and 10 mM IPTG. Fluorescence was detected only in cultures to which IPTG had been added. There was no significant difference in fluorescence between cultures induced with 1 and 10 mM IPTG, and therefore 1 mM IPTG was used for subsequent studies. We used Western blotting to study GFP in IPTG-induced and noninduced cultures of transformed and untransformed L. biflexa (Fig. (Fig.4B).4B). Aliquots of both induced and noninduced cultures were collected 0, 1, 3, 6, 9, 12, and 24 h and 1 week postinduction (p.i.). The maximum of fluorescence in induced cultures was attained after 1 h of induction and did not change during the following days, except 1 week postinduction, when fluorescence declined. Nevertheless, the fluorescence measured 1 week p.i. was still more than 3-fold higher (average, 22,579 AU) than that in noninduced samples after 1 week (average, 7,301 AU) (Fig. (Fig.4C).4C). Fluorescence in the noninduced culture was higher than that in nontransformed culture, indicating leakage of the repression of protein production (Fig. (Fig.4C).4C). In the L. biflexa transformants, GFP expression driven by the LacI repressor-based system was not strong enough for the GFP to be visualized by epifluorescence microscopy (data not shown).
We then constructed reporters to explore the response of specific genes to a change in the environment. The hsp20 and groES genes encode heat stress proteins that have recently been shown to be at least 1.5-fold upregulated after a temperature upshift from 30°C to 37°C in L. interrogans (19). We examined the response of hsp20-gfp and groES-gfp transcriptional fusions to an increase in temperature in the saprophytic strain L. biflexa. GFP expression from the hsp20 and groES promoters was induced 2.6- and 1.3-fold, respectively, in cells grown at 37°C compared to cells grown at 30°C (Fig. (Fig.5A).5A). The higher expression levels at 37°C were correlated to an increase in transcript levels corresponding to the gfp gene, as well as to the specific hsp20 and groES genes, as measured by real-time RT-PCR assays (Fig. (Fig.5B5B).
The lack of genetic tools has hampered molecular analyses of Leptospira spp. In a recent study, the luxCDABE cassette from Photorhabdus luminescens was transferred into Leptospira spp. to generate luminescent bacteria (26). In this work, we constructed a set of plasmid vectors which contain either gfp or mRFP1 alleles. These plasmids confer spectinomycin resistance and have the LE1 origin of replication, so they can be stably maintained in L. biflexa (5). Moreover, the Himar1 transposon we used can deliver gfp to a large number of Leptospira strains, including pathogenic strains. The inclusion of a gfp gene within the transposon generated fluorescent Leptospira and thereby provided an additional method for screening mutants. More importantly, the fluorescence phenotype permits the in vitro, but not in vivo, observation of live mutants that retain full virulence.
GFP fluorescence in Leptospira spp. transformed with our plasmid constructs was lower than that in E. coli probably due to weaker expression of the gfp gene. This could be because in E. coli the gfp gene is present on a multicopy vector, whereas in our Leptospira strains it was present at only one copy per chromosome. In addition, the alleles we used were optimized for E. coli. To improve the translation efficiency, nucleotides could be altered to preferred codons for L. interrogans. Mutations in the gfp open reading frame may also improve the folding and the stability in Leptospira. The use of other strong promoters and/or multiple gfp genes may also allow stronger fluorescence intensity. Finally, the use of microscopy methods other than conventional epifluorescence may improve the visualization of fluorescent Leptospira (28).
GFP production did not affect the growth of pathogenic Leptospira either in liquid media or in animals. After passages in vivo, the strains retained the gfp gene and continued to exhibit uniform green fluorescence (data not shown). However, GFP expression in Leptospira was not sufficiently strong for the fluorescence to be visible in animal tissues. However, the use of fluorescent bacteria may facilitate the study of interactions between Leptospira and cell monolayers. Indeed, several studies have shown that Leptospira strains express surface proteins that interact with the extracellular matrix (1, 8, 22, 34) and that L. interrogans is an invasive pathogen that can rapidly translocate through the host cell (2).
We also showed that the lac system is functional in Leptospira and can be used to control the expression of gfp in this bacterium. The LacI repressor-based system can be simply regulated by the use of IPTG. This inducible expression system could therefore be a very useful tool to construct conditional mutants and to elucidate in more detail the role of essential genes in Leptospira spp. Finally, although the GFP fluorescence signal intensity was too low for characterization of the infection and colonization process in animal models, gfp can be used as a reporter for specific gene expression in Leptospira spp. The construction of transcriptional fusions of leptospiral promoters with gfp is a potentially powerful approach to the analysis of gene expression patterns under various in vitro conditions. For example, temperature and osmolarity are major environmental signals that can affect gene expression in Leptospira (18, 20). We showed that hsp20 and groES transcription levels are increased by a temperature shift from 30°C to 37°C.
Further development of genetic tools in Leptospira should provide major information about the roles of key components in the pathogenesis of leptospirosis.
We thank George Chaconas for kindly providing the gfpmut allele and Emmanuelle Perret (Plate-forme d'Imagerie Dynamique-Imagopole, Institut Pasteur) and Angélique Levoye for fluorescence studies. Albert Ko's group is thanked for animal experiments.
G.M.C. was supported by the CAPES Foundation, Brazilian Ministry of Education, Brazil. This work was supported by the Institut Pasteur, Paris, France, the French Ministry of Research ANR-08-MIE-018, and the Fiocruz-Pasteur Scientific Cooperation Agreement.
Published ahead of print on 29 October 2010.