Distinct properties of Mycobacterium tuberculosis single-stranded DNA binding protein and its functional characterization in Escherichia coli

Abstract

Single-stranded DNA binding proteins (SSBs) play an essential role in various DNA functions. Characterization of SSB from Mycobacterium tuberculosis, which infects nearly one-third of the world’s population and kills about 2–3 million people every year, showed that its oligomeric state and various in vitro DNA binding properties were similar to those of the SSB from Escherichia coli. In this study, use of the yeast two-hybrid assay suggests that the EcoSSB and the MtuSSB are even capable of heterooligomerization. However, the MtuSSB failed to complement a Δssb strain of E.coli. The sequence comparison suggested that MtuSSB contained a distinct C-terminal domain. The C-terminal domain of EcoSSB interacts with various cellular proteins. The chimeric constructs between the N- and C-terminal domains of the MtuSSB and EcoSSB exist as homotetramers and demonstrate DNA binding properties similar to the wild-type counterparts. Despite similar biochemical properties, the chimeric SSBs also failed to complement the Δssb strain of E.coli. These data allude to the occurrence of a ‘cross talk’ between the N- and the C-terminal domains of the SSBs for their in vivo function. Further, compared with those of the EcoSSB, the secondary/tertiary interactions within MtuSSB were found to be less susceptible to disruption by guanidinium hydrochloride. Such structural differences could be exploited for utilizing such essential proteins as crucial molecular targets for controlling the growth of the pathogen.

INTRODUCTION

Single-stranded DNA binding proteins (SSBs) are an important class of proteins (1,2) that play an essential role in DNA replication (3–5), repair (6–8) and general recombination (9,10). The SSBs exist in monomeric to tetrameric forms (11–15). The SSB from Escherichia coli (EcoSSB) has been the most vigorously studied and it has served as a paradigm to understand the biochemical, biophysical and the structure–function aspects of the related SSBs (16). EcoSSB is a homotetramer consisting of monomeric subunits of 177 amino acids (17). The secondary structure prediction suggests that EcoSSB can be divided into two parts, an N-terminal domain (∼120 amino acids) rich in α-helices and β-sheets and the C-terminal domain (∼60 amino acids) which has no definite structure (18). Studies with EcoSSB N-terminal fragments obtained by cleavage at Arg115 by trypsin (SSB_T) or at Trp135 by chymotrypsin (SSB_C) have established that the tetramerization and DNA binding sites are contained within the first 115 amino acids (19).

The C-terminal of EcoSSB can also be further divided into two regions, a spacer region of ∼50 amino acids, rich in proline and glycine residues, and a region of ∼12 residues predominantly consisting of negatively charged amino acids towards the end (16,20). The acidic tail is highly conserved in the bacterial SSBs. The spacer region is thought to modulate the strength of DNA binding, possibly by distancing the highly negatively charged C-terminal end from the DNA binding N-terminal domain. The acidic tail has been suggested to be important in the interaction of SSBs with various proteins in E.coli and it is not essential for DNA binding (16,20). In fact, its deletion results in a protein of higher affinity (19,20).

Recently, we cloned and overexpressed SSB from Mycobacterium tuberculosis (MtuSSB). Biochemical characterization showed it to be a tetrameric protein and comparative analyses of DNA binding properties such as site size requirement, modes of binding etc. between MtuSSB and EcoSSB suggested that the two SSBs were biochemically similar to one another (21). We now show that Mtu- and EcoSSBs are also capable of heterooligomerization. In spite of these parallels between their biochemical properties in vitro, it is noteworthy that MtuSSB failed to complement a Δssb strain of E.coli.

MATERIALS AND METHODS

Subcloning of MtuSSB and EcoSSB into pTrc99C

The MtuSSB gene was excised as a NcoI–HindIII fragment from pETMtuSSB (21) and subcloned into similarly digested pTrc99C (LKB-Pharmacia Biotech, Uppsala, Sweden) to generate pTrcMtuSSB. The open reading frame of EcoSSB was PCR amplified from pTL119 (22) by using a forward primer containing an NcoI site (5′-GGAATTCACCATGGCCAGCAGAGG-3′) and a reverse primer containing a BclI site (5′-GACTGATCAGAACGGAATGTC-3′) and cloned into pTrc99C into the NcoI and BamHI sites to generate pTrcEcoSSB.

Generation of MtuEcoSSB chimeric construct

MtuSSB contains a unique and conveniently located BamHI site (corresponding to amino acid position 129 of EcoSSB), in a region that separates the N- and the C-terminal domains of the protein. To generate a chimeric MtuEcoSSB gene, a BamHI site was engineered into a DNA oligomer (forward primer) that annealed with the EcoSSB gene at the corresponding position. The PCR reaction was carried out using pTrcEcoSSB as a template, the forward primer (5′-ATCGGTGGATCCCAGCCGC-3′) and a reverse primer (5′-CTTTGATCATCCGCCAAAACAGCC-3′) that annealed to the pTrc99C vector sequence downstream of the multiple cloning site. The PCR was carried out for 25 cycles of temperature shifts at 95°C for 1 min, 55°C for 45 s, 68°C for 1 min after initial denaturation of the template at 95°C for 4 min, using Pfu DNA polymerase (Promega Corporation, Madison, WI). The last cycle of PCR was followed by a further incubation at 68°C for 10 min. The resulting PCR product (220 bp) was digested with BamHI and HindIII and used to replace the equivalent DNA fragment from pETMtuSSB (21) and pTrcMtuSSB vectors by standard recombinant DNA methods (23) to generate pETMtuEcoSSB and pTrcMtuEcoSSB, respectively. The pTrcMtuEcoSSB was verified by DNA sequence analysis.

Cloning of EcoMtuSSB chimeric construct

The 220 bp PCR product (∼100 ng) obtained above was also used for inverse PCR to generate a BamHI site in the pTrcEcoSSB in the position corresponding to that of MtuSSB. Conditions for inverse PCR (QuikChange site-directed mutagenesis, Stratagene, La Jolla, CA) were as follows: initial denaturation at 95°C for 5 min, followed by 19 cycles of temperature shifts at 95°C for 1 min, 50°C for 1 min, 68°C for 10 min, followed by a final extension of 68°C for 10 min. The PCR reaction was subjected to DpnI digestion prior to transformation into E.coli TG1 (23). The NcoI–BamHI fragment from this construct (pTrcEcoSSBG129S) was used to replace an equivalent DNA fragment from pTrcMtuSSB using standard recombinant DNA methods to generate pTrcEcoMtuSSB (23) and the construct verified by restriction mapping, a diagnostic PCR and DNA sequence analysis. Further, NcoI–HindIII fragment from pTrcEcoMtuSSB was subcloned into pET11d vector to generate an overexpression construct (pETEcoMtuSSB).

Yeast two-hybrid assay

Yeast reporter strains, SFY526 (24) and HF7c (Feilotter,H., Hannon,G. and Beach,D., cited in the MATCHMAKER Two-Hybrid System manual, Clontech Laboratories Inc., Palo Alto, CA) and the plasmid vectors pGBT9 (carrying nutritional marker TRP1), encoding GAL4 binding domain and pGAD424 (carrying the LEU2 marker), encoding GAL4 activation domain, were used as fusion protein expression systems (25). The ORFs corresponding to the EcoSSB and MtuSSB genes were amplified from pTrcEcoSSB and pTrcMtuSSB vectors, respectively, using Pfu DNA polymerase, using forward (5′-GGAATTCCACAGGAAACAGACCAT-3′) and reverse (5′-CTTTGATCATCCGCCAAAACAGCC-3′) primers. The PCR products were digested with EcoRI and PstI enzymes and cloned into the corresponding sites of pGBT9 and pGAD424 vectors. The recombinant clones (pGBT9EcoSSB, pGBT9MtuSSB, pGAD424EcoSSB and pGAD424MtuSSB) were confirmed by restriction digestions and DNA sequence analysis. The fusion constructs were co-transformed into the aforementioned strains in different combinations and the transformants were obtained on minimal medium plates lacking leucine and tryptophan. The β-galactosidase assays (26) using cells from 1.5 ml cultures of the transformants obtained in the strain SFY526 were carried out in 1.16 ml reaction vol for 120 min using the Clontech MATCHMAKER two-hybrid system protocol. The expression of HIS3 reporter was assessed by growth of the double transformants (in strain HF7c) on synthetic minimal media plates lacking Trp, Leu and His but containing 10 mM 3-aminotriazole (Sigma Chemical Co., St Louis, MO).

Complementation analysis (‘plasmid bumping’)

Escherichia coli RDP317 strain, a derivative of AB1157 (27) carrying a deletion in its chromosomal ssb gene (ssb::Kan^r) and a wild-type copy of the ssb gene on plasmid, pRPZ150 (ColE1 ori, Tc^R), was used. Plasmids (ColE1 ori, Amp^R) harboring the various tester ssb genes were transformed into this strain and the transformants were grown in the presence of ampicillin (100 µg/ml) and kanamycin (25 µg/ml). After four consecutive subculturings in 2 ml of 2YT (1.6% tryptone, 1% yeast extract and 0.5% NaCl) containing the same antibiotics, the cultures were streaked on agar plates to obtain isolated colonies. The colonies were patched on 2YT agar plates containing ampicillin (100 µg/ml) and on plates containing both tetracycline (15 µg/ml) and ampicillin (100 µg/ml). An Amp^R and Tc^S phenotype shows that the incoming plasmid harbors a gene that complements E.coli. On the other hand, maintenance of Tc^R, Amp^R phenotype indicates that the tester plasmid is unable to complement for EcoSSB (20,28,29).

Purification and quantification of the proteins

The MtuSSB, EcoSSB and the two chimeric SSBs were purified exactly according to the protocol previously described (21). Proteins were estimated by a modified Bradford’s method using BSA as a standard (30). Molecular weight and pI were determined using the ExPASy programme (SwissProt).

Gel electrophoresis

Proteins were electrophoresed on 15% polyacrylamide gels containing 0.1% SDS under reducing conditions and visualized by Coomassie brilliant blue R-250 staining (31). Non-denaturing polyacrylamide gel electrophoresis (native PAGE) was performed in the same way but lacked SDS. Loading dye for native gels consisted of 50 mM Tris–HCl (pH 6.8), 10% (v/v) glycerol and 0.01% bromophenol blue.

Electrophoretic mobility shift assays (EMSA)

Ten picomoles of a 27mer DNA (5′-CACCTGTATCATATTCGTCGGCGAGCT-3′) was 5′-end-labeled with [γ-³²P]ATP (32,33) and purified by the spin column method using Sephadex G-50 minicolumn (23). The radiolabeled oligonucleotide was diluted with a known amount of cold oligonucleotide such that the contribution from the labeled counterpart was <1%. Aliquots (1 pmol) of oligomer from this mix (∼20 000 c.p.m.) were incubated with 1 pmol (tetramer) of EcoSSB, MtuSSB, MtuEcoSSB and EcoMtuSSB, at 25°C for 30 min, electrophoresed on an 8% native polyacrylamide gel (21) and autoradiographed.

Fluorescence measurements

All fluorescence emission measurements were performed in a Jasco FP777 spectrofluorimeter with a thermostat cell holder using a cuvette of 1 cm path length and slit widths of 5 nm for excitation and emission. The SSBs (1 µM) were either not treated or treated with 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, 1.4, 1.6, 1.8, 2.0, 2.4, 2.8, 3.0, 3.5, 4.0, 4.5 or 5.0 M guanidinium hydrochloride in 20 mM Tris–HCl (pH 7.4) at 30°C for 4 h before recording the changes in the fluorescence intensities at 343 nm for EcoSSB and 332 nm for MtuSSB upon excitation at 280 nm. The wavelengths of 343 and 332 nm corresponded to λ_em at which maximum difference in fluorescence intensities was observed between the native (0 M guanidinium hydrochloride) and unfolded (6 M guanidinium hydrochloride treated) proteins. The emission intensities were converted to relative fluorescence changes and plotted against the corresponding denaturant concentrations. The relative fluorescence changes were calculated as (F_f – F_u)/F_f, where F_f corresponds to the fluorescence intensity of the native proteins and F_u corresponds to the fluorescence intensity of the guanidinium hydrochloride treated proteins.

RESULTS

In vivo heterooligomerization

Closely related SSBs have been shown to form heterotetramers in vitro (20,35). Comparative biochemical analysis showed EcoSSB and MtuSSB to be similar. However, our in vitro experiments failed to detect heterooligomers of EcoSSB and MtuSSB (21). Therefore, it was of interest to analyze if MtuSSB and EcoSSB heterooligomerized in vivo. For this purpose, we used an in vivo approach employing the yeast two-hybrid assay system (Fig. 1). As expected, the host (HF7c) alone or the transformant containing a single plasmid, pGBT9EcoSSB (sectors 1 and 4, respectively) showed weak growth on the His⁺ plate but no growth on the His^– plate containing 3-aminotriazole (Fig. 1A and B, respectively). The growth of the transformants, harboring various combinations of EcoSSB and MtuSSB constructs, on the His^– plate in the presence of 3-aminotriazole (Fig. 1B, sectors 2, 3, 5 and 6) suggested interaction between the two SSBs. Quantitative analysis of the interactions between the two SSBs, as assayed by lacZ expression using the strain SFY526, is shown in Table 1. Irrespective of whether the EcoSSB or MtuSSB were fused to the DNA binding or the activation domains of GAL4, they led to expression of lacZ suggesting that the two SSBs heterooligomerized in vivo. The β-galactosidase values for the heteroligomers was intermediate (46.41 and 33.75 U) to those obtained for EcoSSB homooligomers (25.76 U) and MtuSSB homooligomers (52.04 U). As seen from the lacZ expression for Mtu-, MtuSSB combination (52.04 U), and that for Eco-, EcoSSB combination (25.76 U), it appears that the affinity between the MtuSSB subunits is much higher than that of the EcoSSB subunits.

Figure 1.

Interaction between MtuSSB and EcoSSB using the yeast two-hybrid system. The host strain alone and the transformants were streaked on medium containing histidine, tryptophan and leucine (A) or on medium lacking histidine, tryptophan and leucine but containing 10 mM 3-aminotriazole (B). Sector 1, HF7c; sector 2, pGBT9EcoSSB and pGAD424EcoSSB; sector 3, pGBT9EcoSSB and pGAD424MtuSSB; sector 4, pGBT9EcoSSB; sector 5, pGBT9MtuSSB and pGAD424MtuSSB; sector 6, pGBT9MtuSSB and pGAD424EcoSSB.

Table 1. Interactions of various SSBs as indicated by β-galactosidase activity.

Protein fused to GAL4 binding domain in pGBT9	Protein fused to GAL4 activation domain in pGAD424	β-galactosidase (Miller units)
EcoSSB	EcoSSB	25.76
EcoSSB	MtuSSB	46.41
MtuSSB	EcoSSB	33.75
MtuSSB	MtuSSB	52.04
–	EcoSSB	0.855
–	MtuSSB	0.8
EcoSSB	pGAD424	1.3
Positive control, GAL4 (in pCL1)		120

Two transformants of each of the combinations were grown in liquid medium till mid-log phase (OD₆₀₀ ∼1.1) and assayed for lacZ activity (see Materials and Methods). Average values are shown. Variation between the two values for each combination was <10%. Full-length GAL4 activator expressed from pCL1 (LEU2) was used as a positive control (40).

In this experiment, the single plasmid controls (pGAD424EcoSSB, pGAD424MtuSSB) or a double plasmid control (pGBT9EcoSSB and pGAD424) showed a basal level of lacZ activity (0.8–1.3 U). On the other hand, when the GAL4 activator itself was expressed from pCL1 (LEU2) as positive control, an activity of 120 U was obtained.

Complementation of E.coli Δssb strain with MtuSSB

In the ‘plasmid bumping’ experiments, we attempted to replace the resident plasmid (pRPZ150, ColE1 ori, Tc^R) harboring a wild-type ssb gene from E.coli RDP317 (Δssb::Kan), with the plasmids harboring test ssb genes (pTrcEcoSSB or pTrcMtuSSB, ColE1 ori, Amp^R). Since SSB is an essential protein, success in replacement of the original Tc^R plasmid by the incoming Amp^R plasmid, resulting in a Tc^S, Amp^R phenotype, shows that the test SSB complements the Δssb strain of E.coli.

The pTrcEcoSSB was used as a control to assess the efficacy of the ‘plasmid bumping’ assay. Expectedly, culturing of the cells in a medium containing Amp and Kan but lacking Tc, resulted in the replacement of the original plasmid (Tc^R) with the incoming test plasmid (Amp^R) and 90% of the cells in the culture became Tc^S (Table 2). Interestingly, under the same assay conditions, the pTrcMtuSSB could not substitute for the pRPZ150 (Tc^R), suggesting that MtuSSB failed to complement the Δssb strain of E.coli (Table 2). The failure of complementation was not because of the lack of the expression of MtuSSB, as the pTrcMtuSSB afforded overproduction of MtuSSB in E.coli (data not shown).

Table 2. Complementation analysis of SSBs.

Test ssb genes	No. of colonies (AmpR)	No. of colonies (AmpR, TcR)	Efficiency of plasmid replacement (TcR→TcS) in the Δssb strain (%)
EcoSSB	136	14	90
EcoSSB(G129S)	50	2	96
MtuSSB	50	50	0
MtuEcoSSB	50	50	0
EcoMtuSSB	50	50	0

Transformants of the various SSB containing test plasmids (Amp^R) in Δssb strain of E.coli RDP 317 were grown in liquid medium containing Amp and Kan and streaked to obtain isolated colonies. The colonies were patched on 2YT agar plates containing Amp alone, as well as on the plates containing Amp and Tc. The ability of the patches to grow on the plates was recorded (see Materials and Methods).

Generation of Mtu- and Eco- chimeric SSBs

Comparison of EcoSSB and MtuSSB shows maximal variation in the C-terminal domains of the two SSBs (Fig. 2A). Since this region in EcoSSB has been implicated in interaction with various cellular proteins (20), we generated two chimeric constructs between the N- and C-terminal domains of the two SSBs (Fig. 2B). The DNA sequence at position 129 (EcoSSB numbering) constitutes a BamHI site in MtuSSB. This position is located within the region where trypsin and chymotrypsin cleavage sites have been mapped and used to define the N-terminal domain of SSB (SSB_T and SSB_C, respectively) for the study of DNA binding properties in EcoSSB (19). As a first step in generating the chimeric constructs, BamHI site was introduced into the EcoSSB gene by site-directed mutagenesis (G129S mutation) and then the DNA sequences corresponding to the N- and C-terminal domains were swapped between the two SSBs. The construct MtuEcoSSB contains the N-terminal portion from MtuSSB and the C-terminal domain from EcoSSB. Similarly, the chimeric construct EcoMtuSSB, comprises the N-terminal domain from EcoSSB and the C-terminal domain from MtuSSB (Fig. 2B).

Figure 2.

(A) Comparison of EcoSSB and MtuSSB sequences. Shading indicates residues that are identical/conserved between the two SSBs. (B) Schematic representation of the chimeric constructs as shown with respect to EcoSSB and MtuSSB (shaded and hatched regions, respectively). Dashed box, linker region missing in MtuSSB and EcoMtuSSB; unfilled box, the corresponding region in EcoSSB, EcoSSBG129S and MtuEcoSSB; downward arrow, BamHI site, found in MtuSSB, EcoSSBG129S and the fusion constructs. Sizes of the SSBs (amino acids) are shown on the right.

Complementation analyses with the chimeric SSBs

The results of the complementation analyses with the EcoSSB, the G129S mutant (EcoSSBG129S) and the chimeras MtuEcoSSB and EcoMtuSSB, as obtained from the ‘plasmid bumping’ experiments are shown in Table 2. The EcoSSBG129S complements the Δssb strain of E.coli, as effectively as the wild-type SSB, suggesting that engineering of the BamHI site into the EcoSSB gene, which resulted in the G129S mutation, did not affect its in vivo function. However, as was the case with MtuSSB, neither the MtuEcoSSB nor the EcoMtuSSB complemented the E.coli Δssb strain, even though both the pTrcMtuEcoSSB and the pTrcEcoMtuSSB constructs support overproduction of respective proteins in E.coli (data not shown).

Overexpression and purification of the chimeric SSBs

To determine the DNA binding properties and the oligomerization status of the two chimeric proteins, they were overproduced in E.coli BL21 (DE3) (36) using the T7-based expression constructs, pETMtuEcoSSB and pETEcoMtuSSB, and purified to apparent homogeneity. As shown in Figure 3, the wild-type and the chimeric SSBs migrated according to their calculated molecular weights (lane 2, MtuSSB, M_r 17.35 kDa; lane 3, EcoSSB, M_r 18.9 kDa; lane 4, MtuEcoSSB, M_r 19.16; lane 5, EcoMtuSSB, M_r 17.2 kDa).

Figure 3.

Analysis of purified SSBs on 15% SDS–PAGE (∼1.5 µg protein was analyzed). Lane 2, MtuSSB; lane 3, EcoSSB; lane 4, MtuEcoSSB; lane 5, EcoMtuSSB. Prestained molecular weight markers (lane 1, BioRad) correspond to bovine serum albumin (∼79.5 kDa), ovalbumin (∼49.5 kDa), carbonic anhydrase (∼34.8 kDa), soybean trypsin inhibitor (∼28.3 kDa), lysozyme (∼20.4 kDa) and aprotinin (7.2 kDa).

Oligomerization status of the chimeric SSBs

MtuSSB (pI 5.12, M_r 17.35 kDa) and EcoSSB (pI 5.44, M_r 18.9 kDa) have been shown to be homotetramers. Their relative mobility on native polyacrylamide gel has been shown to be characteristic of their homotetrameric nature (21). Using the native polyacrylamide gels as our assay system, we show that both the chimeric constructs migrate with the mobilities intermediate to those of the EcoSSB and MtuSSB (Fig. 4), suggesting that both the MtuEcoSSB (pI 5.12; M_r 19.16 kDa) and the EcoMtuSSB (pI 5.18; M_r 17.2 kDa) formed homotetramers.

Figure 4.

Analysis of the SSBs on 15% native PAGE (∼7.5 µg protein was analyzed). Lane 1, MtuSSB; lane 2, EcoSSB; lane 3, MtuEcoSSB; lane 4, EcoMtuSSB.

EMSAs with the chimeric SSBs

The ability of the chimeric SSBs to bind to DNA were analyzed by EMSA, after incubating the 27mer DNA and the various SSB tetramers in 1:1 molar stoichiometries. As shown in Figure 5, compared with the free oligomer (lane 1), the MtuEcoSSB and the EcoMtuSSB (lanes 4 and 5, respectively) formed complexes with DNA which migrated with mobilities similar to those of the parent SSBs (lanes 2 and 3).

Figure 5.

Electrophoretic mobility shift assay (see Materials and Methods). Lane 1, free 27mer DNA; lanes 2–5, complexes of the 27mer DNA with MtuSSB (lane 2), EcoSSB (lane 3), MtuEcoSSB (lane 4) and EcoMtuSSB (lane 5).

Fluorescence studies

In these experiments we have used the changes in the intrinsic fluorescence upon addition of the denaturant (guanidinium hydrochloride), as a correlate for the loss of tertiary/secondary structure of the SSBs. Changes in the relative fluorescence of the isothermally denatured proteins as a function of increasing denaturant concentrations are shown in Figure 6. As is evident, ∼1.5 and 2.5 M guanidinium hydrochloride concentrations were required for a loss of 50% of the initial fluorescence intensity for EcoSSB and MtuSSB, respectively. Consistent with the results of yeast two-hybrid assay (Table 1), these data reflect greater stability of MtuSSB when compared with EcoSSB.

Figure 6.

Relative fluorescence intensity changes in EcoSSB (open circles) and MtuSSB (filled circles) upon isothermal denaturation with varying concentrations of guanidinium hydrochloride (see Materials and Methods).

DISCUSSION

Mycobacterium tuberculosis, the causative agent of tuberculosis, is an important organism. Approximately one-third of the world’s population is infected with M.tuberculosis and nearly 2–3 million people die of tuberculosis every year. To develop the much needed prophylactics and newer drugs against this organism, it is important to unravel its biology.

Recently, we cloned and overexpressed SSB from M.tuberculosis. Biochemical characterization showed it to be a tetrameric protein, and its comparative analyses with EcoSSB for DNA binding properties such as site size requirement, relative DNA affinity and modes of DNA binding etc. suggested that the two SSBs were similar to one another (21). In this study, using a yeast two-hybrid assay, we observed the formation of the heterooligomers between the MtuSSB and EcoSSB. Surprisingly, in spite of its biochemical similarity with EcoSSB, the MtuSSB failed to complement a Δssb strain of E.coli.

The sequence comparison of MtuSSB with the EcoSSB revealed several differences. While some of these are conserved, others are strikingly different (21). Of particular interest is the region corresponding to the amino acids 159–171 of EcoSSB which is absent from MtuSSB (Fig. 2B). Notably, this gap overlaps with the C-terminal 12 amino acid acidic stretch (166–177) of EcoSSB implicated in interaction with various cellular proteins. The C-terminal acidic residues are extremely well conserved across the prokaryotic SSBs. Interestingly, MtuSSB lacks the glycine-proline-rich linker region which connects the acidic domain with the rest of the protein. The χ subunit of DNA polymerase has been shown to interact with C-terminal tail residues of EcoSSB (37,38). Since MtuSSB lacks part of the region needed for its interaction with the χ protein in E.coli, it was thought that replacement of the C-terminal domain of MtuSSB with that of EcoSSB might result in a chimera which would complement the Δssb strain of E.coli. Thus, the chimeric constructs (MtuEcoSSB and EcoMtuSSB) were generated by interchanging the N- and C-terminal domains of the two SSBs. Mobility of these proteins on native gels showed that they existed as tetramers, and were efficient in DNA binding (Figs 4 and 5). However, neither the EcoMtuSSB nor the MtuEcoSSB rescued the Δssb strain of E.coli. This failure to complement reflects the inability of these domains to function independently and highlights the need of the N- and C-terminal domains to communicate with each other as well as with the other cellular proteins. It appears that such ‘cross-talk’ between the two domains, which is likely to be vital for the biological functions of SSBs, does not occur in the context of the heterologous C-terminal domains.

Interestingly, the recently completed genome sequencing data from M.tuberculosis suggest that it lacks the χ homolog of E.coli (39). Since the homolog of χ subunit of DNA polymerase is most likely not there and the C-terminal tail of EcoSSB which interacts with it is also missing in MtuSSB, it is possible that the mechanism by which SSB functions in M.tuberculosis is distinct. Furthermore, the conserved residues such as lysines at positions 49 and 62 and arginines at positions 21, 41, 56, 72, 86 and 96 have been shown to be important in determining the path of the DNA in EcoSSB (14). It was observed that in MtuSSB, at least three of these positions have been replaced by non-conserved amino acids such as Tyr, Glu and Thr (21) which suggested that while the DNA binding properties of the two SSBs are similar, the molecular mechanisms underlying the interactions of SSB with DNA are probably distinct in MtuSSB (21).

The yeast two-hybrid analysis and the intrinsic fluorescence quenching studies (Table 1; Fig. 6) suggest a more firm structure for MtuSSB when compared to that of EcoSSB. How these structural differences may contribute to the lack of complementation of E.coli by MtuSSB is unclear. Interpretation of many of these biochemical properties and the structure function relationship of the SSBs, necessitates the determination of the three-dimensional structure of MtuSSB. Differences in the mechanism of action of the essential proteins may form the basis in establishing them as useful targets for controlling the growth of M.tuberculosis.