Study of activity of cloned promoters in slow-growing Mycobacterium tuberculosis during long-term growth conditions in vitro or inside macrophages, requires a genome-integration proficient promoter probe vector, which can be stably maintained even without antibiotics, carrying a substrate-independent, easily scorable and highly sensitive reporter gene. In order to meet this requirement, we constructed pAKMN2, which contains mycobacterial codon-optimized gfpm2+ gene, coding for GFPm2+ of highest fluorescence reported till date, mycobacteriophage L5 attP-int sequence for genome integration, and a multiple cloning site. pAKMN2 showed stable integration and expression of GFPm2+ from M. tuberculosis and M. smegmatis genome. Expression of GFPm2+, driven by the cloned minimal promoters of M. tuberculosis cell division gene, ftsZ (MtftsZ), could be detected in the M. tuberculosis/pAKMN2-promoter integrants, growing at exponential phase in defined medium in vitro and inside macrophages. Stable expression from genome-integrated format even without antibiotic, and high sensitivity of detection by flow cytometry and fluorescence imaging, in spite of single copy integration, make pAKMN2 useful for the study of cloned promoters of any mycobacterial species under long-term in vitro growth or stress conditions, or inside macrophages.
Episomal promoter probe vectors that are stably maintained with antibiotic selection are useful experimental systems to study promoter activity during short durations. However, to study activity of cloned promoters of a slow-growing pathogen, such as Mycobacterium tuberculosis, on a long-term basis in vitro or in vivo inside macrophages, it is difficult to maintain stable concentration of antibiotics. Such experimental systems need genome-integration proficient promoter probe vectors that can be stably maintained even without antibiotics for long durations. Vectors that integrate into mycobacterial genome through site-specific recombination via attP of mycobacteriophage L5 or Ms6 and attB in the presence of phage integrase, int, have been reported (Lee et al., 1991; DasGupta et al., 1998; Freitas-Vieira et al., 1998; Vultos et al., 2006). As a reporter, GFP and GFP mutants of higher fluorescence used for mycobacterial studies (Dhandayuthapani et al., 1995; Kremer et al., 1995; Luo et al., 1996; Valdivia et al., 1996; Parker and Bermudez, 1997; Barker et al., 1998; Via et al., 1998; Teitelbaum et al., 1999; Cowley and Av-Gay, 2001) offer the advantages of a sensitive, non-invasive, substrate-independent rapid assay system over β-galactosidase (Rowland et al., 1999), catechol-2,3-dioxygenase (Curcic et al., 1994), chloramphenicol acetyl transferase (DasGupta et al., 1993), and bacterial luciferase (Roberts et al., 2005). However, being single copy in the genome, a highly fluorescent GFP variant is required for the accurate determination of activity of promoters of different strength. In order to meet these requirements, the present study describes construction and use of pAKMN2, which combines the sensitivity of gfpm2+ of highest fluorescence reported till date (Steinhauer et al., 2010) and L5attP-int-mediated stable genome-integration proficiency, to determine activity of cloned promoters of M. tuberculosis in M. tuberculosis on a long-term basis under in vitro growth conditions or inside macrophages even without antibiotics.
Construction and stability of, and expression from, genome-integrated pAKMN2
The construction of pAKMN2, from the source vectors, pMN406 (Roy et al., 2004) and pDK20 (DasGupta et al., 1998), through the intermediate episomal pAKMN1, is given in the self-explanatory Fig. 1. In order to verify the stability of genome-integrated pAKMN2-promoter constructs, M. tuberculosis/pAKMN2-PQ1K1 (Mt) and M. smegmatis/pAKMN2-PQ1K1 (Ms) integrants, carrying total promoter region, Q1-K1, of M. tuberculosis cell division gene, ftsZ, MtftsZ (Fig. 2A; Roy and Ajitkumar, 2005), were grown to mid-log phase without hygromycin and plated on hygromycin-containing and hygromycin-free plates. The colony-forming units (cfu) for both the integrants were comparable in the presence and absence of hygromycin (checked up to 30 and 60 generations for Mt and Ms integrants respectively) (Fig. 2B), with statistically insignificant values (two-sided P-values obtained by unpaired t-test: 0.3078 for Mt and 0.1374 for Ms). On the contrary, statistically significant reduction (two-sided P-values: 0.0008 for Mt and 0.0035 for Ms) in cfu was found in the absence of hygromycin for the episomal pMN406-PQ1K1 transformants of M. tuberculosis and M. smegmatis, grown under same conditions (Fig. 2B).
Flow cytometry of GFPm2+ fluorescence in pAKMN2-PQ1K1 integrants of M. tuberculosis and M. smegmatis did not show significant difference in the presence or absence of hygromycin (Fig. 2B and C). However, there was significant reduction in the level of GFPm2+ fluorescence in the absence of hygromycin for the episomal pMN406-PQ1K1, in both M. smegmatis (P = 0.0003) and M. tuberculosis (P = 0.0021) (Fig. 2B and C). Comparison of the coefficient of variation (CV) of GFPm2+ fluorescence intensities from Mt and Ms integrants, in the presence and absence of hygromycin, showed little difference (4% and 6% for Mt respectively; 5% and 7% for Ms respectively). Whereas, the CV of GFPm2+ fluorescence intensities of episomal pMN406-PQ1K1 transformants of M. tuberculosis and M. smegmatis showed about two- to threefold increase (11% for Mt and 9% for Ms), in the presence of hygromycin. The CV values were even higher in the absence of hygromycin (46% for Mtb and 32% for Ms).
Promoter activity from M. tuberculosis/pAKMN2 integrants in vitro
Activity of individual cloned minimal MtftsZ promoters, P1, P2, P3, P4, P6, and total promoter, PQ1K1 (Roy and Ajitkumar, 2005), was determined for the respective M. tuberculosis/pAKMN2-promoter integrants using flow cytometry and semi-quantitative RT-PCR. P5 was not considered in the present study, as it was found to be active only in M. smegmatis but not in M. tuberculosis (Roy and Ajitkumar, 2005). Flow cytometry analyses of mid-log phase M. tuberculosis/pAKMN2-P1, -P2, -P3, -P4, -P6 and -PQ1K1 (positive control) integrants showed expression of GFPm2+ (Fig. 2D). The vector control, M. tuberculosis/pAKMN2 (devoid of promoter), did not show GFPm2+ expression (Fig. 2D). In order to verify the validity of the GFP based quantification of promoter activity, levels of gfpm2+ mRNA in these integrants were determined using semi-quantitative RT-PCR, at the mid-linear range of amplification, taking PQ1K1 total promoter activity as 100% and sigA gene expression for normalization, as reported (Manganelli et al., 2001; Dubnau et al., 2002; Roy et al., 2004). Calculation of the % activities of P1, P2, P3, P4 and P6, obtained by flow cytometry and RT-PCR and comparison of the mean values themselves by one-way between-subjects anova, followed by Post hoc comparisons using Tukey HSD test showed that RT-PCR and flow cytometry data were mostly consistent with each other (Fig. 2E). P1 and P3 were the most significantly less active than P2, P4 and P6, whereas the latter three were active more or less at the same level as detected by RT-PCR and flow cytometry (Fig. 2E). Only P1 and P3 were found to be significantly different by RT-PCR (red line), but not by flow-cytometry (not shown) due to large standard deviations in P1 flow cytometry values. Flow cytometry based quantification in the steady-state condition in general reflected RT-PCR quantification of steady-state RNA levels. However, coefficients of variation (CV) of flow cytometry for % activity of the promoters were higher than the same in RT-PCR.
Promoter activity from M. tuberculosis/pAKMN2 integrants in macrophages
Mycobacterium tuberculosis H37Ra, which was used for the identification of MtftsZ promoters, infects macrophages (Falcone and Collins, 1997; Zhang et al., 2005). Therefore, the activity profile of the M. tuberculosis H37Ra/pAKMN2-P1, -P2, -P3, -P4 and -P6 integrants was examined in RAW 264.7 (ATCC TIB71) macrophages, in comparison with those under exponential phase of growth. Mycobacterium tuberculosis/pAKMN2-PQ1K1 (total promoter region) and M. tuberculosis/pAKMN2 (without promoter) integrants were the positive and negative controls respectively. At 36 h post infection, fluorescence microscopy of bacteria inside macrophage and flow cytometry analyses of macrophage-derived M. tuberculosis promoter integrants showed expression from all the individual promoters and PQ1K1 (Fig. 3A). Calculation of the % activity from flow cytometry data and pairwise comparison of mean of % activity (as in Fig. 2E) for each promoter between infection and exponential growth phase showed significantly low activity of P1 and P3 (two-sided P-value: 0.04 for P1 and 0.01 for P3; Fig. 3B). Statistically insignificant reduction was noticed in the % activity of P2, P4 and P6, compared with their activity under exponential phase. Similarly, % activity of the promoters were calculated from the semi-quantitative RT-PCR data for gfpm2+ mRNA from the pAKMN2-promoter-integrants from macrophages and from exponential cultures and statistical significance was evaluated (Fig. 3C), as described (see Fig. 2E). The results showed that there were no significant changes in the expression of gfpm2+ from P2, P4 and P6 promoter regions between exponential and infection phases. However, there was significant reduction in the activity of P1 (two-sided P = 0.04) and P3 (two-sided P = 0.02) in the infection phase. Again, flow cytometry data reflected promoter activity detected using RT-PCR on RNA from M. tuberculosis/pAKMN2-promoter integrants under infection phase.
The data presented in the study demonstrate that pAKMN2 possesses all the features that are required for a promoter probe vector for the study of promoters of M. tuberculosis on a long-term basis in vitro or inside macrophages, even without antibiotic. First of all, pAKMN2 showed stable genome integration in mycobacteria and stable expression of gfpm2+, even in the absence of antibiotic. Second, being genome-integrated single copy, the requirement for a highly sensitive reporter gene to detect and quantify low-activity promoters, such as P1 and P3, could be achieved with the use of gfpm2+. Since GFPm2+ has been demonstrated to be more fluorescent than the earlier higher fluorescent GFPm+ and twice more fluorescent than E-GFP, and stable in expression in both slow- and fast-growing mycobacteria (Steinhauer et al., 2010), its higher sensitivity of detection of MtftsZ promoters over other gfp variants or other conventional reporters was not addressed in this study. Third, in general, RT-PCR for gfpm2+ expression from the cloned promoter regions correlated with those from flow cytometry analyses under steady-state conditions. However, noticeably, the CV values of flow cytometry for % activity of some of the promoters were higher than those determined using RT-PCR. Therefore, it may not be completely reliable to use GFPm2+ for quantification of promoter activity in situations other than steady-state conditions, due to its stability. The data presented imply that pAKMN2 may be amenable for use in a multi-well plate format under fluorescence microscope and to screen for promoters of genes that are active in infection or dormancy or starvation or stationary phase. It also offers the possibility to transform pAKMN2-promoter integrants with a multicopy plasmid containing a candidate regulatory gene or a genomic/cDNA library under an inducible promoter, to find out or screen for regulatory proteins of the promoter that could be of therapeutic target value.
Bacteria, media and culture
Mycobacterium tuberculosis H37Ra and M. smegmatis mc2155 cells were grown in Middlebrook 7H9 (Difco) liquid medium supplemented with 0.2% glycerol and 0.05% Tween 80 or in Middlebrook 7H10 agar (Difco) medium. Mycobacterium tuberculosis cultures were supplemented with 10% albumin-dextrose-catalase (ADC) in liquid medium and 10% oleic acid-albumin-dextrose-catalase (OADC) in solid medium. Hygromycin was used at 50 µg ml−1 in mycobacteria.
Plasmid constructs and molecular cloning
pMN406-P1, -P2, -P3, -P4 and -PQ1K1, carrying the individual MtftsZ promoters, P1, P2, P3, P4 and PQ1K1 (the total promoter region encompassing all the promoters) cloned in place of Pimyc promoter of pMN406, upstream of gfpm2+, were already available (Roy and Ajitkumar, 2005). pMN406 contains mycobacterial plasmid ori, mycori, and ColE1 ori, hygr marker, and mycobacterial codon usage adapted gfpm2+ that codes for GFPm2+ (Steinhauer et al., 2010). pMN406-ΔPimyc is pMN406 devoid of Pimyc promoter (Roy et al., 2004). pMN406-P6 was constructed by the PCR amplification of P6 using primers Q1 and P1P2r (Table 1) to obtain a 545 bp PCR product, which was digested with EcoRV and BclI and the final 389 bp region was cloned between the SspI-BamHI sites in pMN406, in place of Pimyc promoter. Since pMN406-P1, -P2, -P3, -P4, -P6 and -PQ1K1 were already available, instead of cloning the respective promoters into pAKMN2, the 2.084 kb L5 attP-int region was subcloned from pBS(KS)-L5att-int as ClaI-NotI fragment and inserted into these vectors, replacing the mycori region from the respective constructs, to generate pAKMN2-P1, -P2, -P3, -P4, -P6 and -PQ1K1. The Shine–Dalgarno sequence of gfpm2+, downstream of the MCS, helps in the transcriptional fusion. Transcription termination sequence of T4g32 gene (terminator from gene 32 of phage T4) and transcription terminator rrnBT2 separate the cloned promoter-gfp fusion cassette from the rest of the vector backbone. Presence of the integrants in the genomic DNA was verified using PCR with vector-specific mgfp2 primer (Table 1) and forward primer specific for the respective promoter region (Roy and Ajitkumar, 2005).
The infection of mouse macrophage cell line RAW 264.7 (ATCC TIB71) with M. tuberculosis H37Ra cells was carried out, as described (Butcher et al., 1998). In brief, RAW 264.7 cells were grown to semi-confluence in either 175 cm2 flasks (for RNA isolation) or 24-well tissue culture plates containing #1 thickness 12 mm diameter glass cover glass (for microscopy) or 75 cm2 flasks (for flow cytometry) in complete RPMI-1640 medium, containing 10% foetal calf serum. Mycobacterium tuberculosis promoter-integrants were washed three times in PBS to remove traces of antibiotic and sonicated for 5 × 5 s at 50% output to disperse clumps. Macrophage cells were infected with the bacterial cells at a multiplicity of infection of 10 and allowed to phagocytose the bacilli for 6 h in humidified 5% CO2 incubator at 37°C. The cells were washed three times to remove un-phagocytosed bacteria and the medium was replaced with fresh RPMI-1640 medium containing 50 µg ml−1 gentamicin and incubated for a total period of 36 h post infection for analysis.
Preparation of cells for flow cytometry and fluorescence microscopy
Mycobacterium tuberculosis integrant cells were harvested at different stages of growth in vitro or from infected macrophages, washed once with PBS, finally resuspended in 1 ml PBS, for flow cytometry analysis. Mycobacterium tuberculosis cells from infected macrophage cells were harvested by osmolysing macrophage cells with sterile double-distilled water containing 0.1% Triton X-100. The bacterial cells were recovered by centrifugation at 8000 g for 15 min at 4°C, washed thrice with PBS, and resuspended in 500 ml PBS for flow cytometry and fluorescence microscopy. Flow cytometry was performed using Becton Dickinson FACScan machine. Dotplots were analysed using WinMDI software, version 2.8. GFPm2+ fluorescence of the bacilli in infected macrophages were observed using Leica Microscope (DMLB) and images were captured.
RNA isolation and semi-quantitative RT-PCR
Total RNA isolation from M. smegmatis and M. tuberculosis cells was carried out, as described earlier (Roy et al., 2004) or using Tri-reagent (Sigma) according to manufacturer's instructions. Isolation of bacteria from macrophages for RNA extraction was carried out, as described (Dubnau et al., 2002). In brief, the macrophage cells were lysed by suspending in Tri-reagent. The lysate was centrifuged at 5000 g for 20 min to pellet down intracellular mycobacteria. RNA from the pelleted bacteria was extracted using freshly added Tri-reagent. RNA samples were treated with DNaseI (USB), in the presence of 1 unit of Procine (RNase inhibitor) (USB) per µg of RNA, to remove DNA contamination. Integrity of the RNA was verified on formaldehyde agarose gel and quantified. Semi-quantitative RT-PCR reactions were performed on 50–200 ng RNA samples, using Qiagen One-Step RT-PCR Kit, as described (Roy et al., 2004), with the primers mgfp1 and mgfp2 (Table 1). As normalization control, sigA was amplified, using sigA1 and sigA2 (Table 1). RT-PCR was carried out at the linear range of amplification and the bands on agarose gel were quantified (ImageQuant software, V 2.54).
The percentage activity of each promoter was calculated based on the total activity of Q1K1 promoter region in the same species in both flow cytometry and RT-PCR from minimum three sets of independent experiments. Activity of promoter obtained by flow cytometry or RT-PCR in a given species was compared with each other by one-way analyses of variance (anova) with post-hoc adjustment for multiple comparisons after Tukey HSD test. For pairwise comparison unpaired t-tests were applied to obtain two-sided P-values.
This work was supported by a research grant [37(1361)/09/EMR-II] from CSIR and in part by the part grants from the DBT Centre of Excellence in Tuberculosis and DBT–IPSBBR/Pathogen Biology, and infrastructure support from the DST-FIST, UGC-CAS, The authors thank Dr M. Niederweis for pMN406, Dr A. K. Tyagi for pDK20, and Dr Omana Joy, Flow Cytometry Facility, Indian Institute of Science, for flow cytometry analyses.