|Home | About | Journals | Submit | Contact Us | Français|
The genetic aspects of DnaA mediated initiation of oriC replication in mycobacteria are largely unknown. To get insights into the replication initiation process in mycobacteria, we characterized Mycobacterium tuberculosis DnaA and its interactions with oriC. We show that the replacement of M. smegmatis dnaA with the M. tuberculosis counterpart expressed from its native promoter resulted in temperature-sensitive (TS) phenotype. However, the TS phenotype was abolished when the M. tuberculosis dnaA was expressed from the inducible amidase promoter, which produces elevated levels of DnaA. We provide evidence that M. tuberculosis dnaA promoter activity was unaffected at nonpermissive temperature, but the DnaA protein was found to be unstable indicating that protein factors stabilizing M. tuberculosis DnaA are absent in M. smegmatis. Finally, we show by surface plasmon resonance that the M. tuberculosis DnaA interacts with M. smegmatis oriC, similar to its cognate oriC indicating that the binding interactions between in vitro folded DnaA and oriC are unaffected. Our results suggest that Mtb DnaA functions as a partially active protein in M. smegmatis, hence is not as proficient as M. smegmatis counterpart in optimally driving the M. smegmatis oriC replication machinery.
Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis, is responsible annually for nearly two million deaths worldwide. Mtb is a slow grower with an average doubling time of 24 h and it is generally believed that Mtb modulates its growth rates in response to infection. Emergence of multiple-drug resistant and extremely drug resistant strains of Mtb globally necessitate an urgent need for the development of new drugs targeted to hitherto unexplored metabolic pathways 1. The genus Mycobacterium includes rapid growers such as M. smegmatis with average doubling time of 3 hours. The genetic elements responsible for growth rate differences between the two species are unknown. Despite the growth rate differences between the species, M. smegmatis serves as an excellent surrogate for evaluating Mtb gene functions.
DNA replication and cell division are essential aspects of cell cycle and both events are critical for cell duplication and multiplication. DNA replication is believed to be regulated at the level of initiation [reviewed in 2-4]. DnaA protein is the initiator of replication and binds to origin of replication, also known as oriC, to initiate replication. OriC contains several repeats of nine-nucleotide long DnaA protein recognition sequences, called DnaA-boxes. DnaA protein recognizes and binds to these DnaA-boxes independent of any nucleotide cofactor, but DnaA binding to ATP aids its interactions with DnaA-boxes 2-5. These initial events lead to rapid oligomerization of DnaA on oriC and organization of oriC-DnaA initiation complex that results in strand melting, unwinding and DNA synthesis. It has been shown that the oriC sequences from closely related bacterial species, viz. members of enterobacteriaceae, can be exchanged indicating a functional similarity within the same family members 6-9. Earlier studies related to mycobacterial replication initiation revealed that mycobacterial oriC sequences are essential and show species specificity, i.e. Mtb oriC cannot function in M. smegmatis and vice versa 8, 10. DnaA proteins of M. smegmatis and M. tuberculosis share significant sequence identity. Other data indicated that ATP-hydrolysis activity of DnaA protein is critical for DnaA oligomerization on oriC and viability of M. tuberculosis 11. The dnaA promoter (dnaAp) region contains several DnaA-boxes 8, 10, 12 and is the target of MtrA response regulator 13.
The present study was undertaken to gains insights into the DnaA mediated oriC replication initiation process of mycobacteria. The following questions were asked: Can Mtb dnaA serve as the sole source of DnaA protein in M. smegmatis and vice versa? Does Mtb DnaA protein interact with M. smegmatis oriC? We show that although Mtb DnaA binds to M. smegmatis oriC in vitro, the replacement of M. smegmatis dnaA gene with Mtb counterpart results in temperature sensitive (TS) phenotype. Our results favor a hypothesis that the native Mtb DnaA protein functions like a partially active protein in M. smegmatis.
Escherichia coli Top10 strain was used for cloning purposes and was propagated in Luria Bertani (LB) broth or agar. Recombinants were selected on LB agar containing appropriate antibiotics, viz. ampicillin (Amp) at 100 μg/ ml, kanamycin (Km) at 50 μg/ ml and hygromycin (Hyg) at 100 μg/ ml. M. smegmatis and M. tuberculosis strains were propagated in Middlebrook 7H9 broth or 7H10 agar supplemented with oleic acid, albumin, dextrose and catalase with appropriate antibiotics as needed (Hyg, 100 μg/ ml and Km, 15 μg/ ml). M. smegmatis strains RGM15 and RGM32 have been described earlier 14. RGM15 is M. smegmatis dnaA single crossover (SCO) strain carrying dnaA suicidal plasmid pMG25 whereas RGM32 is M. smegmatis dnaA mutant strain carrying functional dnaA gene expressed from its native promoter integrated at the attB locus as described earlier 14. Growth was monitored by measuring absorbance at 600 nm and viability determined by plating for colony forming units on appropriate media agar plates.
Two-step homologous recombination protocols essentially as described for the construction of M. smegmatis and Mtb dnaA mutant strains were followed 14. Briefly, pMG68 plasmid expressing Mtb dnaA from its native promoter 11 was integrated into RGM15 strain and screened for double crossovers (DCO) as described at indicated temperatures. In some experiments, pMBL45 plasmid expressing Mtb dnaA gene from amidase promoter (amip) 15 was integrated into RGM15 and processed for DCO at 37°C. Characterization of DCOs by PCR and Southern was as described earlier 11, 14.
M. smegmatis strains grown to exponential phase were spun briefly at 3000X g at 4°C, washed with 10 mM Tris-HCl pH 7.5 and 1 mM EDTA and cell free lysates were prepared by bead-beating as described 16. Total protein content was determined by BCA protein assay (Pierce) using bovine serum albumin standards. Equal amounts of various protein samples were separated by SDS-PAGE and proteins transferred to nitrocellulose membrane blots. DnaA protein levels in the blots were determined by immunoblotting with purified α-DnaA antibodies utilizing the ECF kit (GE Life sciences) followed by visualization of bands with a Molecular Imager using QuantityOne software.
Luciferase activity was measured as described earlier 17. The promoter-less lux reporter plasmid, pMH66, was generous gift from Dr. David Sherman 18. To create dnaAp::lux construct the Mtb dnaAp region spanning the rpmH-dnaA intergenic region was PCR amplified using primers MVM458 (5′-TGGCAACACTAGTATCCGGCCACTCACT-3′) and MVM459 (5′-GTCGGAAGCTTGTCTCGTTAGTTCCGG-3′). Similarly, to create sigAp::lux construct, the Mtb sigAp region was amplified using primers MVM480 (5′-CTACTAGACTAGTGCCGGAATTGT-3′) and MVM481 (5′-CCACCCAAGCTTTTACTCACTTCCGGT-3′). The PCR products were cloned as SpeI-HindIII fragments in pMH66 upstream of lux coding region. The reporter plasmids were transformed into RGM32 and RGM71 strains and luciferase activity at indicated time periods was measured by mixing 100 μl of culture with 100 μl of 1 mM Beetle luciferin (Promega Corporation) and measured in luminometer (Turner Designs). Luciferase activity was reported as Lux units per unit OD600 nm.
DNA synthesis was monitored as described using [5,6-3H] uracil 14. Briefly, actively growing cultures of RGM32 and RGM71 at 28°C were diluted to low cell density and shifted to 40°C for 6 h. Parallel cultures were grown at 28°C. At the end of 6 h incubation, radiolabeled uracil was added at a final concentration of 0. 5 μCi/ml, samples were removed at indicated time periods, incubated in 0.3 M KOH for 24 h at 37°C and the amount of radioactivity present was measured.
SPR experiments were carried out essentially as described earlier 11. Briefly, biotinylated oriC regions of Mtb and M. smegmatis were generated by PCR and immobilized on SA (streptavidine coated) sensor chip to 60 to 80 resonance units (RU). The dnaA-dnaN intergenic region was considered as oriC for each species. As a control, scrambled DNA sequence 5′-AAGTAAGTATATAAGTAAGT-3′ was coupled and all binding values were subtracted from oriC values. Typical binding experiments were carried in 50 mM Tris-acetate pH 8.25, 0. 5 mM magnesium acetate, 0. 3 mM EDTA, 5 mM beta-mercaptoethanol, 10 mM ammonium acetate, 0. 005% Tween-20, 50 ng poly(dA-dT) per ml, 50 ng sheared salmon sperm DNA/ml and 1 mM indicated nucleotides ATP, ADP or ATPγS. DnaA was incubated in the above buffer for 5 min and injected over the sensor surface pre-equilibrated in the same buffer. Typical flow rate of samples was 25 μl per min. Binding association and dissociation were measured as described 11.
Samples were prepared on ice in 40 μl of the above buffer containing radiolabelled M. tuberculosis oriC DNA fragment (0.034 μM) and different concentrations of DnaA protein. We prepared three probes each with a different DnaA-box sequence. Probe 1 sequence included 5′-CGGCGGTTCCGTTCACAACCCACGC-3′ and its complement 5′-GCGTGGGTTGTG-AACGGGAACCGCCG-3′; probe 2 sequence included 5′-GAGTGTCG-CTGTGCACAAACCGCGC-3′ and its complement sequence 5′-GCGCGGTTTGTGCACAGCGACACTC-3′, and probe 3 included 5′-CCCACGCCTCATCCCCACCGACCCA-3′ and its complement 5′-TGGGTCGGTGGG-GATGAGGCGTGGG-3′. The presumptive DnaA-box sequences were underlined and shown in boldface. ATP was used at a final concentration of 1 mM. All reactions were carried out in the presence of a 250-fold excess salmon sperm testis DNA. Reactions were incubated at 37 °C for 15 min, mixed with gel-loading buffer containing 50% glycerol and 0.02% bromophenol blue, resolved by PAGE in 5% acrylamide gels and visualized by autoradiography.
The primary amino acid sequences of Mtb and M. smegmatis DnaA show 55 % identity and 59% similarity whereas the nucleotide sequences of the dnaAp and the oriC show 46% and 49%, similarity, respectively [see also, 8, 10, 12, 19]. In an effort to understand why the oriC sequences from these two species cannot be exchanged, we first examined by homologous recombination if the Mtb dnaA gene can replace M. smegmatis counterpart and vice versa. The dnaA gene is essential for survival 14 , hence its disruption at the native locus requires the presence of a functional copy integrated elsewhere on the chromosome 14. Accordingly, we first integrated pMG68 plasmid expressing the Mtb dnaA gene from its native promoter at the attB locus in M. smegmatis dnaA SCO strain RGM15. Next, the SCO strain was processed for selection of DCOs at 37°C as described 14. When analyzed by PCR, 80/80 DCOs showed wild-type copy of dnaA (Fig. 1A) indicating that Mtb dnaA cannot replace M. smegmatis counterpart under these conditions. We considered a possibility that expression of Mtb dnaA in M. smegmatis could have resulted in toxicity, hence the potential mutant DCOs were nonviable. In such cases, selection of mutant DCOs at conditions that reduce growth rates, i.e. growth at 28°C, avoids issues related to toxicity. Therefore, we selected DCOs at 28°C and the resultant DCOs were replica patched at 42°C. Several DCOs showed TS phenotype and failed to grow at 42°C or 40°C (data not shown). The corresponding colonies growing at 28°C were processed and analyzed by PCR (Fig. 1B) and Southern hybridization (data not shown). All showed a pattern expected for mutant DCO indicating that Mtb dnaA can replace M. smegmatis counterpart provided the growth rate of the DCO was reduced. Presumably, expression of the wild-type Mtb dnaA in M. smegmatis from its native promoter produces a protein that is not proficient at driving M. smegmatis oriC replication machinery optimally, hence results in loss of viability. One DCO growing at 28°C, designated as RGM71, was characterized further.
To understand the molecular basis for the observed TS phenotype, we evaluated DNA synthesis at nonpermissive temperature (40°C) following 6 h temperature shift and compared results with the cultures growing at permissive temperature (28°C). The 6 h period corresponds to approximately two doubling times for M. smegmatis. Parallel experiments were carried out with control strain, RGM32, a DCO expressing M. smegmatis dnaA at the attB locus 14. As can be seen, the DNA synthesis rates of RGM32 growing at 28°C and 40°C were nearly comparable (Fig. 2A). The DNA synthesis at 40°C was linear as opposed to the exponential rate seen at permissive temperature (Fig. 2A). These results suggest that although Mtb DnaA functions in M. smegmatis, it behaves as a partially active protein, hence is unable to drive M. smegmatis oriC replication machine optimally at 40°C.
One explanation for the reduced DNA synthesis is that either Mtb DnaA protein is unstable at 40°C or that the Mtb dnaAp is not optimally active thereby results in reduced DnaA protein levels. Evaluation of DnaA protein levels by immunoblotting revealed a reduction in the DnaA protein levels in the cell lysates of cultures maintained for 1, 2, 4, 8 and 20 h at 40°C as compared to that grown at 28°C (see Fig. 2B) indicating that the protein is not stable at 40°C. In this context, it is unknown if Mtb DnaA protein folds properly in M. smegmatis.
Next, to evaluate dnaAp activity, we created a reporter plasmid expressing lux gene downstream of Mtb dnaAp (dnaAp-lux) and transformed into RGM71 and RGM32 strains. We measured the dnaAp activity as relative Lux units at different periods after temperature shift and compared results with the control growing at 28°C (Fig. 2C). Because RGM72 cultures filament and clump at 42°C, Lux activity was measured at 37°C. For control, lux gene was also placed downstream of promoter of house keeping gene, sigA. As can be seen, the Mtb dnaAp activity, expressed as relative Lux units, was same at 28°C and 37°C, like the sigAp, in both RGM32 and RGM71 strains (Fig. 2C).
Since DnaA protein levels were reduced upon temp shift, we considered a possibility that the TS phenotype can be reversed if elevated levels of DnaA were present. Accordingly, we expressed Mtb dnaA from inducible amip and selected for DCOs at 37°C on plates containing acetamide. Earlier studies revealed that expression of dnaA gene downstream of amip results in 8-fold DnaA accumulation 14. As expected, when Mtb dnaAp was expressed from amip, we obtained several viable DCOs (Fig. 1C) indicating that the TS phenotype associated with Mtb dnaA expression was not an issue in cells producing elevated levels of DnaA.
Together, these results suggest that the Mtb dnaAp is well expressed in M. smegmatis, but the translated product DnaA is unstable at 40°C. One explanation is that the Mtb DnaA protein might not fold properly in M. smegmatis and therefore becomes unstable (Fig. 2B). This could lead to a defect in oriC replication and bacterial multiplication. Expression from the amip results in large amounts of DnaA protein 14 that can counter the inefficient folding and the resulting instability issues at nonpermissive temperatures. It is also possible that hitherto unknown protein cofactors that otherwise promote the stabilization of Mtb DnaA protein are either missing or are in insufficient amounts in M. smegmatis. Presumably, the regulation of DnaA protein stability in both species could be different, and further studies are required to address this issue.
Earlier studies showed that optimal levels of DnaA protein are required for cell cycle progression 14, 20. Since DnaA protein levels were reduced following temperature shift, we expected that RGM71 cells would not be proficient for multiplication at 40°C. To evaluate this possibility, we examined the RGM71 cells growing at 28°C, and those following 8h and 12 h shift. The cells were stained with ethidium bromide-mithramycin dyes as described and visualized by brightfield and fluorescence microscopy (Fig. 3). As can be seen, actively dividing cells at 28°C were short (2 to 3 microns in length) and contained two nucleoids per dividing cell (Fig. 3 i and ii). In contrast, RGM71 cells at 40°C were elongated (Fig. 3 iii-vi) and contained multiple nucleoids (Fig. 3 iv and vi). Continued growth at 12 h (Fig. 3 v) and beyond (not shown) led to further increase in cell length and severe clumping. Some of filamentous cells under these growth conditions showed nucleoid-free zones. These results confirm earlier findings that that optimal levels of DnaA protein are required for normal cell cycle progression and reduction in DnaA protein levels lead to cell division blockage 14, 20.
In the above experiments we tested whether Mtb dnaA gene can replace the function of M. smegmatis counterpart. We carried out reciprocal experiments to test whether M. smegmatis dnaA can replace Mtb dnaA gene function. We transformed pMG25, a dnaA suicidal recombination plasmid 14 into Mtb expressing M. smegmatis dnaA from its own promoter. Screening of DCOs in this background at 37°C showed both wild-type and mutant patterns (data not shown). These results indicate that M. smegmatis DnaA is fully functional in Mtb. Presumably, the slow growth rate of Mtb made the selection of mutant DCOs feasible and this scenario is comparable to permissive temperature growth of M. smegmatis (Fig. 1B).
The above data tend to suggest that the Mtb DnaA protein can drive M. smegmatis DNA replication initiation machinery at 28°C. Using the SPR technique we showed earlier that Mtb DnaA protein binds ATP and oligomerizes on oriC utilizing its intrinsic ATPase activity 11. Furthermore, DnaA-binding to oriC in the presence of ATP is biphasic that shows rapid initial ‘on’ followed by slow association phase, and an initial rapid dissociation ‘off’ followed by a slow dissociation phase 11, whereas the binding in the presence of ADP and ATPγS is monophasic 11. Other data indicate that while both M. smegmatis and Mtb oriC sequences contain multiple DnaA-box repeats, sequences of all DnaA-boxes and their positions are not conserved. Hence, to get insights into the DnaA mediated oriC replication initiation process, we evaluated Mtb DnaA interactions with M. smegmatis oriC.
As a first step, we examined Mtb DnaA interactions with different DnaA-boxes by EMSA (Fig. 4A). We showed that Mtb DnaA binds and retards the mobility of three different DnaA boxes (Fig. 4A). These results indicate that DnaA protein shows broad sequence specificity, hence could likely interact with M. smegmatis oriC. Next, we examined Mtb DnaA interactions with M. smegmatis oriC by SPR technique in the presence of ATP, ADP or ATPγS. In these experiments, a biotinylated M. smegmatis oriC was coupled to streptavidine sensor chip to approximately 80 Response units. In parallel, Mtb oriC and a scrambled sequence were also coupled and DnaA interactions with oriC were examined. DnaA protein in the presence of indicated nucleotides was then flown over sensor surface and the association and dissociation data were measured. As can be seen, binding isotherms of DnaA with M. smegmatis oriC in the presence of different nucleotides were similar to that seen with Mtb oriC (Fig. 4B-i). For example, consistent with the earlier published data 11, Mtb DnaA showed biphasic binding to its oriC in the presence of ATP and monophasic binding in the presence of ADP and ATPγS 11. Essentially, similar association and dissociation profiles were observed with M. smegmatis oriC (Fig. 4B-ii). The biphasic nature of Mtb DnaA interaction with M. smegmatis oriC is consistent with a notion that Mtb DnaA rapidly oligomerizes on M. smegmatis oriC, as it does with its cognate oriC, presumably utilizing ATP-hydrolysis 11. It should be noted the DnaA protein used in these experiments was folded in vitro following denaturation to generate a functionally active protein 21. Nonetheless, these results indicate that the initial events associated with DnaA-oriC interactions between the two species are similar with the in vitro folded protein. It is unknown if these and those steps subsequent to the initial oriC-DnaA interaction are modulated in vivo. The differences in the extent of RU between the two species could reflect the different amounts of oriC coupled to the sensor chip.
Our results reveal that dnaA genes in both species can be exchanged, but with some limitations. The TS phenotype associated with the replacement of M. smegmatis dnaA function by Mtb counterpart suggests that the Mtb DnaA behaves like a partially active protein in M. smegmatis and becomes unstable at 37°C and higher. The observed TS phenotype is rather surprising as Mtb genes are often used to replace the functions of M. smegmatis counterparts and two examples relevant to cell cycle include ftsZ and crgA 22. Why might wild-type Mtb DnaA protein be partially active? One possibility is that the Mtb DnaA does not fold well in M. smegmatis, thereby making a partially active protein. It is known that DnaA protein interacts with DnaB helicase, another component of the oriC-initiation complex. It is also possible that optimal DnaA activity requires its association with hitherto unidentified cofactors, which are missing, or are present in limited amounts in M. smegmatis. A consequence would be inefficient activity and reduced stability of the DnaA protein. Optimal levels of DnaA have been shown to be required for regulated cell cycle progression 14, 20. Thus, a reduction in the intracellular levels of DnaA at non-permissive temperature would negatively impact cell cycle progression events. Further studies are required to resolve these issues. I
This work was supported by NIH grants, RO1AI48417 (MR) and RO1AI84734 and AI41406 (MM).
Competing interests: The authors have no conflicts of interest to declare.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.