DNA Clasping by Mycobacterial HU: The C-Terminal Region of HupB Mediates Increased Specificity of DNA Binding

Background HU a small, basic, histone like protein is a major component of the bacterial nucleoid. E. coli has two subunits of HU coded by hupA and hupB genes whereas Mycobacterium tuberculosis (Mtb) has only one subunit of HU coded by ORF Rv2986c (hupB gene). One noticeable feature regarding Mtb HupB, based on sequence alignment of HU orthologs from different bacteria, was that HupBMtb bears at its C-terminal end, a highly basic extension and this prompted an examination of its role in Mtb HupB function. Methodology/Principal Findings With this objective two clones of Mtb HupB were generated; one expressing full length HupB protein (HupBMtb) and another which expresses only the N terminal region (first 95 amino acid) of hupB (HupBMtbN). Gel retardation assays revealed that HupBMtbN is almost like E. coli HU (heat stable nucleoid protein) in terms of its DNA binding, with a binding constant (Kd) for linear dsDNA greater than 1000 nM, a value comparable to that obtained for the HUαα and HUαβ forms. However CTR (C-terminal Region) of HupBMtb imparts greater specificity in DNA binding. HupBMtb protein binds more strongly to supercoiled plasmid DNA than to linear DNA, also this binding is very stable as it provides DNase I protection even up to 5 minutes. Similar results were obtained when the abilities of both proteins to mediate protection against DNA strand cleavage by hydroxyl radicals generated by the Fenton's reaction, were compared. It was also observed that both the proteins have DNA binding preference for A:T rich DNA which may occur at the regulatory regions of ORFs and the oriC region of Mtb. Conclusions/Significance These data thus point that HupBMtb may participate in chromosome organization in-vivo, it may also play a passive, possibly an architectural role.


Introduction
The Escherichia coli (E.coli) protein HU (heat stable nucleoid protein) is an abundant DNA binding protein, which is a major component of the bacterial nucleoid [1].HU is a small, basic, histone like protein, initially called factor U, isolated for the first time from E. coli strain U93 [1,2].In E. coli HU activity is comprised of a hetero-dimer of HupA (Hu a) and HupB (Hu b) coded by hupA and hupB genes, respectively.Besides this, homodimeric forms (Hu a2 and Hu b2) are also observed with different forms (mainly Hu a2 and Hu ab) predominating in distinct phases of growth [3].HU is highly conserved [4] and unlike most DNA binding proteins, binds to both single stranded (ss) and double stranded (ds) DNA in a sequence independent manner [5,6].The other major nucleoid associated proteins are FIS (factor for inversion stimulation), H-NS (histone like nucleoid protein), IHF (integration host factor) and a stationary phase specific DNA binding protein, DPS (DNA binding protein from starved cells [2,7,8].HU, which resembles eukaryotic proteins of the high mobility group (HMG) class in terms of its DNA binding properties e.g. it binds dsDNA with low affinity and negligible sequence specificity; also displays high affinity for some altered DNA structures such as junctions, nicks, gaps, forks, and overhangs even under stringent salt conditions [9,10].In E. coli HU action is fairly pleiotropic in the sense that HU plays a role in many cellular processes such as site specific recombination [11,12], initiation of DNA replication [13,14], phage Mu transposition [15] and introducing negative supercoiling into relaxed DNA molecule in presence of Topoisomerase I [2].Furthermore, cells deficient for HU are known to be highly sensitive to gamma and UV irradiation and it is thought that HU might assist in the processes of recombinational repair [11,12].
HU also mediates ring closure of linear DNA enhancing DNA cyclization rates [16].Crystal structure of Anabaena HU bound to DNA revealed that binding of dimers of HU to linear DNA with cohesive ends produces an overall bend of ,105-140u thus stimulating rate of ring formation [17].Once cohesive DNA ends are brought together they can be sealed by T 4 DNA ligase [16,18,19].E. coli HU binds to linear DNA fragments in a weakly cooperative fashion with one dimer occupancy per 9 bp and this binding is observed only under low salt conditions [20].The binding of HU to a DNA four way junction is several orders of magnitude higher in comparison to binding to linear DNA even under high salt conditions [21] and is not inhibited by a large (100 fold) excess of competitor linear DNA.This binding preference is thought to be commensurate with the role of HU in DNA inversion [22].
Although in E. coli the heterodimeric state of HU is quite preponderant, in many bacterial species HU exists as a homodimer like in Bacillus subtilis [23,24], Mycobacterial family etc.The annotated genome of Mycobacterium tuberculosis strain H37Rv, (Mtb) bears the potential to encode only one subunit of HU -the product of ORF Rv2986c (hupB Mtb ).The product of ORF Rv2986c is also referred to as mycobacterial DNA binding protein (MDP-1) or histone like protein (HLP Mt ) [25,26].HupB Mtb is 214 amino acids long, has a high content of alanine (23.78%) and lysine (18.93%) and has a calculated pI of 12.4 with the ratio of basic to acidic amino acid residues being 12.This ratio in HupA and HupB of E. coli is 1.4, while for histone H1 or H5 this ratio is approximately 7 [25].The N-terminal portion of HupB Mtb exhibits significant homology to histone like proteins of E. coli while the Cterminal part displays homology to eukaryotic H1 histone [25].Sequence alignment of HU homolog's from different members of the Mtb complex shows that N terminal end of HupB is conserved, but the C-terminal end is variable (Supplementary Figure S2) and this feature has therefore been used as a diagnostic marker for differentiating members of Mtb complex [27].HupB Mtb , like other bacterial HU proteins, lacks tryptophan, cysteine and tyrosine residues.These properties and an overall homology to bacterial HU suggest that HupB Mtb could be involved in the packaging of mycobacterial DNA and functions as a nucleoid associated protein.Interestingly HupB Mtb interacts with the immune system, an interaction that may occur due to release of protein during natural cell lysis or release resulting from interaction with the host immune system [25,28].Recently HupB Mtb ortholog in M. smegmatis has been shown to possess DNA end joining/ring closure protein activity [19].
In this report we describe biochemical studies on recombinant, purified HupB Mtb and show that it displays properties of a nucleoid associated protein in terms of non-specific DNA binding activity.By undertaking comparative studies on HupB Mtb and a derivative that lacks 119 amino acids located at the C-terminal end, we provide evidence to show that this C-terminal region (CTR) of HupB Mtb is required for providing increased specificity of DNA binding and for specific recognition of altered nucleic acid structures under stringent (high salt) conditions.DNA binding by HupB Mtb is stable and, in contrast to its derivative lacking the CTR, protects DNA from attack by reactive oxygen species produced by the Fenton's reagent and also affords substantial protection against cleavage by DNaseI.HupB Mtb displayed substantial preference in binding to AT rich DNA.These observations suggest that HupB Mtb action is different from enterobacterial HU and the CTR of HupB Mtb may act like a DNA clasp.Given its increased preference to bind to AT rich DNA it is speculated that the site(s) of HupB Mtb action on the Mtb genome could be AT rich regions that are found in the replication origin and regulatory regions of genes.

Presence of a basic CTR in HupB orthologs from the mycobacterial clade
As noted previously a very visible feature of the polypeptide sequence of HupB orthologs belonging to the mycobacterial clade is the presence of a basic CTR [25,27].Member proteins of this clade are roughly 214 amino acids long whereas E. coli HupA and HupB are 90 amino acids long.The N terminal region of HupB Mtb shows significant homology to individual subunits of enterobacterial HU namely HupA, HupB from E. coli, HupB from S. enterica and HU from B. subtilis (Supplementary Figure S1A), whereas the CTR despite retaining its overall basic character displays some variation in amino acid composition within members of the mycobacterial clade (Supplementary Figure S2).Amino acid sequence of the HupB Mtb CTR shows that it is rich in lysine and alanine.It has six PAKK and one KAAK repeats (Supplementary Figure S1A and Supplementary S1B,) which are also present in histone H1, and are known to facilitate DNA binding [29].The CTR of HupB Mtb also shows significant homology to sea urchin histone H1 (Supplementary Figure S1B) primarily in terms of the presence of these tetrapeptide repeats.In order to determine the specific function related to this CTR, two hupB Mtb ORFs were generated by PCR; one bearing the full length hupB Mtb DNA sequence (encoding HupB Mtb ) and another encoding a hupB Mtb variant bearing only the N-terminal 95 amino acids (HupB MtbN ).

The CTR of HupB Mtb imparts increased specificity in DNA binding
Under low salt conditions (10 mM NaCl or KCl) HupB Mtb bound to linear ds DNA (ds oligonucleotide A1) in a weakly cooperative manner with a Kd of 42 nM whereas even at 1200 nM HupB MtbN barely displayed 50% binding to A1. (Figure 1A and 1B).For HupB Mtb linear ds DNA interaction three retarded complexes (Figure 1A shown by arrowheads) were apparent with complex 2 being the major species.However with HupB MtbN only a single band could be observed (Figure 1C) which shows that the binding is non-cooperative.Based upon the length of ds DNA used herein (that is 48 bp) and assuming that retardation is provoked by dimeric HupB Mtb , one can roughly estimate that HupB Mtb binds to linear DNA with one dimer occupying DNA ranging from 16 to 24 bp.Binding of HupB Mtb (at 50 nM) to A1 under increasing salt concentrations revealed that HupB Mtb binding was greatly impaired at salt concentrations above 150 mM (NaCl or KCl; Figure 2A).E. coli HU is known to bind with duplex DNA containing a nick or a gap of one or two nucleotides with high affinity [30].A specific DNA protein complex (Kd 65 nM), under high salt conditions was detected when the interaction of HupB Mtb with linear DNA bearing a nick (ds oligonucleotide A2) was studied (Figure 2B).On the other hand, binding of higher concentrations of HupB MtbN (employed at 450 nM) promoted binding to A2 under low salt conditions, which was impeded under high salt conditions (Figure 2C).Under low salt HupB MtbN bound to nicked DNA with a Kd of 1000 nM (Table 1).HU is known to bind specifically to DNA containing either nick or a gap or DNA junction (cruciform DNA) without sequence preference [9,10,30].These structures are associated with DNA damage and repair.
In order to address the question whether HupB Mtb and HupB MtbN displayed differential binding we studied their interaction with linear ds DNA bearing one or two nucleotide gap and cruciform DNA, represented by ds oligonucleotides A3, A4 and A5 respectively, in the presence of low salt (10 mM KCl) (Supplementary Figure S3 and Supplementary Figure S4; for details refer Supplementary Methods S1).It is apparent (Figure 3; Table 1) that the CTR of HupB Mtb is required for specific recognition and high affinity recognition of a duplex DNA containing a nick or a gap of one or two nucleotides and cruciform DNA.

Interaction of HupB Mtb with supercoiled plasmid DNA
Despite a lack of sequence specificity in DNA binding, HU is known to bind preferentially to negatively supercoiled DNA [17,31].The interaction of HupB Mtb and HupB MtbN with supercoiled pBluescriptSK (pBSK) plasmid DNA was studied in buffer conditions such that the molar ratio of protein to DNA ranged from 15:1 to 180:1.It could be seen that HupB MtbN displayed a markedly reduced interaction with supercoiled plasmid DNA (Figure 4A, lanes 1-7).Even at the highest protein to DNA molar ratio sufficient amount of unbound pBSK plasmid was visible (Figure 4A, lane 7).HupB Mtb on the other hand bound very proficiently with supercoiled DNA, in the sense that at intermediate protein to DNA molar ratio the intensity of DNA protein complex was large and the DNA remained lodged in the wells Figure 4A, lanes 11 to 14).This property is reminiscent of proteins that can induce DNA compaction [32].In order to test the preference of HupB Mtb towards supercoiled/linear DNA, pBluescript SK (pBSK) plasmid was linearized using EcoR1.EMSA was carried out with supercoiled pBSK (Figure 4B, lanes 1-8) and with linear pBSK (lanes 9-16).It was observed that interaction of HupB Mtb with supercoiled pBSK at a molar ratio (protein: DNA) of 75:1 the intensity of DNA protein complex was large and the DNA remained lodged in the well (lane 4).However, the interaction of HupB Mtb with linear pBSK was observed at a molar ratio of 112.5:1 (lane 14).It therefore appears that HupB Mtb exhibits ,1.5-fold higher affinity towards supercoiled plasmid DNA.Furthermore, the interaction of HupB Mtb with supercoiled DNA was more avid than with linear DNA of the same size (Figure 4B  HupB Mtb displays greater affinity in binding to AT rich DNA Since the average GC content of the Mtb H37Rv genome works out to around 65%, we wondered whether the CTR of HupB Mtb would endow it with the ability to interact more proficiently with GC rather than AT rich DNA.To assess whether HupB Mtb bore this attribute, interaction of this protein was studied with ds DNA oligonucleotides A6, (GC content 84%) and A7 (AT content 64%).Surprisingly, HupB Mtb interacted proficiently with AT rich DNA forming two gel-retarded complexes (C1, C2; Figure 5A marked by arrowheads) where as its interaction with GC rich DNA was negligible.On the other hand HupB MtbN while retaining its ability to interact more proficiently with AT rich DNA, forming a single gel retarded complex (Figure 5B, lanes 9-14), displayed weak interaction with GC rich DNA (Figure 5B, lanes 2-7).We then determined by circular dichroism spectrometry the secondary structure parameters of the two proteins, which revealed that the two proteins exhibited  some differences in the content of alpha helix and beta sheet structures (Supplementary Figure S5; for details refer Supplementary Methods S1).Perhaps these differences may account for the ability of HupB MtbN to display binding to GC rich DNA.
Enhanced DNA protection mediated by the CTR of HupB Mtb Given that the CTR of HupB Mtb imparts upon it increased avidity with respect to DNA binding we tested whether the said region could protect DNA from enzymatic or non-enzymatic DNA strand breakage.In the first instance we tested the abilities of HupB Mtb and HupB MtbN to mediate protection of DNA from cleavage by DNaseI.Supercoiled pBSK plasmid was exposed to 1 unit of DNase I for 30 seconds at 25uC in the presence of varying concentrations of HupB Mtb and HupB MtbN (0.2-2.5 mM).Further-more, timed digestions with DNase I were carried out with 2.5 mM of HupB Mtb and HupB MtbN (Figure 6A and 6B).While both proteins at 2.5 mM gave substantial protection to DNA from DNase I cleavage (data not shown), analysis of timed digestion with DNaseI revealed that the binding of HupB Mtb was stable enough to provide protection from DNase I even up to 5 min (Figure 6A, lane 8).However HupB MtbN was significantly impaired in mediating protection from DNase I, and could not provide protection beyond 30 sec (Figure 6A, lanes 11-15), as could be seen from the absence of DNA band (lanes 11-15).Similar results were obtained when the abilities of both proteins to mediate protection against DNA strand cleavage by hydroxyl radicals generated by the Fenton's reaction, were compared.HupB Mtb was significantly more proficient than HupB MtbN in preventing DNA strand cleavage upon exposure to the Fenton system (Figure 6B).These studies thus suggest that the CTR of HupB Mtb promotes associations with DNA that are stable enough to promote substantial DNA protection.

Discussion
In bacteria the chromosomal DNA is present in a highly compacted form, called the bacterial nucleoid, in which DNA exists in association with basic low molecular weight proteins.Unlike that seen for eukaryotic chromatin, bacterial chromosome lacks perceptible DNA protein organization.In the well-studied case of E. coli, low molecular weight proteins such as H-NS, HU, IHF and FIS are thought to be the key protein constituents of the bacterial nucleoid [7,8,33,34].While both HU and H-NS are very abundant, HU has been considered to be a prokaryotic histone, Figure 4. HupB Mtb binds strongly to supercoiled and linear plasmid DNA in comparison to HupB MtbN as evident from agarose gel electrophoresis of pBluescript SK (linear and supercoiled).A) 500 ng of supercoiled pBluescript SK plasmid was incubated with increasing amounts (0, 1.2, 0.5, 1.0, 1.5, 2.0 and 2.5 mM) of either HupB MtbN (lanes 1-7) or HupB Mtb (lanes 8-14) in low salt conditions (10 mM KCl).The DNAprotein interaction products were resolved by electrophoresis on a 1% agarose gel buffered with 0.56 TBE at 35 V, 16-20 hrs.B) 500 ng of either supercoiled (lanes 1-8) or linear (lanes 9-16) pBluescript SK DNA was incubated with increasing amount of HupB Mtb .The DNA-protein interaction products were resolved as described above.doi:10.1371/journal.pone.0012551.g004primarily based upon its resemblance to eukaryotic histones in terms of its non-specific DNA binding capacity and amino acid sequence composition [2,8].HU is highly conserved and in enterobacteriacae HU exists as a heterodimer comprising of two homologous protein subunits, Hua and Hub.The heterodimeric state of HU appears to be restricted to enterobacteriacae, whereas in many other bacteria HU exists as a homodimer.
In this study we have examined the biochemical properties of the HU ortholog from M. tuberculosis (Mtb) strain H37Rv, the product of ORF Rv2986c, designated as HupB Mtb .One noticeable feature regarding HupB Mtb , based upon sequence alignment of HU orthologs from different bacteria, was the presence at its Cterminal end a highly basic extension (Supplementary Figure S1), designated as the C-terminal region (CTR), which prompted us to examine its role.Using purified recombinant HupB Mtb and HupB MtbN , a variant of the former that lacks the CTR, comparative biochemical studies were undertaken.HupB Mtb and HupB MtbN exhibited HU-like DNA binding properties, in that both displayed non-specific DNA binding with the difference that HupB Mtb was more proficient in binding to DNA than HupB MtbN in several DNA binding assays employed.These results therefore suggest that the CTR of HupB Mtb imparts upon it greater specificity in DNA binding (Figure 1 and Figure 2).Recently, HU ortholog from M.smegmatis has been shown to possess similar properties [19].It is worth noting that HupB MtbN is almost E. coli HU-like in terms of its DNA binding, its DNA binding constant (K d ) with linear dsDNA is greater than 1000 nM, a value comparable to that obtained for the HUaa and HUab forms [10].Furthermore, HupB MtbN like E. coli HU displayed comparatively greater affinity towards DNA bearing a nick or gap and cruciform DNA (Supplementary Figure S4, Table 1).In all these instances DNA binding affinities of HupB Mtb were comparatively greater.These observations suggest that the basic or skeletal DNA binding property of HupB Mtb resides in its first 90 amino acid residues and that the basic CTR acts to enhance DNA binding.
Recent crystal structure studies of Anabaena HU (homodimeric protein of size similar to and homologous to E. coli HU) have shown that in the HU dimer, two monomers associate to form an a-helical base extending outwards from each monomer and contribution from its C-terminal region is a largely b-sheeted structure (extends as two b-ribbon arms that contact the DNA within its minor groove) [17,35].The N-terminal half of each monomer contributes the a-helical base.The presence of a highly conserved proline residue at position 63 (present in HupB Mtb and HupB MtbN , Supplementary Figure S1) at the tip of each arm is thought to be critical for DNA binding.Given these observations, it is possible to place the presence of the basic DNA binding property of HupB Mtb in its N-terminal 90 amino acids.HupB Mtb displayed more stable DNA associations than HupN MtbN , HupB Mtb bound more proficiently to supercoiled plasmid DNA and, commensurate with this property, mediated enhanced protection of DNA from enzymatic or non-enzymatic DNA strand breakage (Figure 4).From these studies it appears that the CTR of HupB Mtb may act like a DNA clasp promoting more stable DNA interactions.It is apparent that though presence of CTR lowers the K d of dsDNA binding by at least one order of magnitude in comparison to HupB MtbN , (and E. coli HU), the K d (42 nM for HupB Mtb ) is not lowered to the extent expected for a site specific DNA binding protein.Thus, HupB Mtb bears the propensity to function as a nucleoid-associated protein in its natural setting.
An interesting finding was that HupB Mtb interacted much more proficiently with substantially A:T rich than with G:C rich DNA.The G:C content of the Mtb genome is around 65% and is fairly uniform.Admittedly the G:C content of the dsDNA oligonucleotide A13 used in these studies is high (84%), nevertheless it is present in the Mtb genome located upstream of the ORF Rv494.On the other hand HupB Mtb bound proficiently with dsDNA oligonucleotide A1 that constitutes a DNA sequence located within the origin of chromosomal replication on the Mtb genome, present between the dnaA and the dnaN genes.Recently we showed that Mtb DnaA in-vitro can mediate DNA duplex unwinding in the vicinity of this sequence [36].At the E. coli origin of replication oriC, strand opening by DnaA is known to require the presence of either HU or IHF [14].It is thus possible that the site(s) of biological HupB Mtb activity in Mtb may be restricted to A:T rich regions, at least one that is present in the origin of replication where HupB Mtb could promote efficient DnaA mediated helix unwinding and/or to other regions of limited A:T richness that may occur at the regulatory regions of ORFs.It is interesting to note that HupB Mtb is considered to be essential for mycobacterial growth, an observation that is apparently compatible with its suspected role in DNA replication [37].Future studies in this regard, directed towards exploring the possible requirement of HupB Mtb in the initiation of chromosomal DNA replication are underway.

Materials and Methods
Cloning and purification of recombinant His tagged HupB Mtb and HupB MtbN protein The M. tuberculosis ORF Rv2986c coding for HupB Mtb and HupB MtbN protein was PCR amplified using genomic DNA from H37Rv as a template and primers HupB Mtb _F, HupB Mtb _R and HupB MtbN _R, carrying specific restriction enzyme sites (Table 2), by Accutaq DNA polymerase (Sigma).The amplicons thus generated were digested with Nde1/HindIII restriction enzymes and cloned into the corresponding sites of pET28a expression vector.The resultant plasmids were labeled as pETHupB Mtb and pETHupB MtbN .Restriction analysis and DNA sequencing confirmed the authenticity of all constructs.Recombinant HupB Mtb and HupB MtbN , coded by M.tb ORF Rv2986c, were purified from the soluble fraction of BL21 (DE3) cells transformed with pETHupB Mtb and pETHupB MtbN .The transformants were grown overnight at 18uC and induced with 0.15 mM IPTG at an OD 600 of 0.2 for the expression of recombinant protein as described earlier [38,39].The recombinant protein was purified in buffer having 30 mM Tris (pH 8.5), 5 mM MgCl 2 , 250 mM NaCl, 100 mM potassium glutamate and 7% glycerol.SDS PAGE confirmed the purity of the protein.The concentration of the protein was estimated by BCA (Bichinconic acid) and protein was stored at 220uC until further use.

Electrophoretic mobility shift assays
The interaction of increasing amounts of protein (0.2-2.5 mM) (HupB Mtb and HupB MtbN ) with 480 ng of supercoiled or linearized plasmid DNA (pBSK linearized with EcoR1) was carried out in buffer A (20 mM HEPES-KOH pH 7.5, 5 mM magnesium acetate, 1 mM EDTA, 0.05 mM bovine serum albumin and 7% glycerol).The reactions were incubated at 25uC for 30 min and the products resolved by electrophoresis on a 1% agarose gel in 0.56 TBE at 35 V for 16-20 hrs.The gels were visualized by staining with ethidium bromide.
The interaction of HU (HupB Mtb and HupB MtbN ) protein with dsDNA oligonucleotide probes was characterized by native PAGE.Increasing amounts of protein was incubated with 59 32 P-labelled DNA for 30 min at room temperature in 20 ml of binding buffer A. The DNA protein complex were resolved on a 5% polyacrylamide gel (29:1) buffered with 0.256 TBE as mentioned in figure legends.The binding of HU protein (HupB Mtb and HupB MtbN ) to nicked, 1-nt gap, 2-nt gap or cruciform DNA was carried out in buffer A, either in low salt (10 mM KCl) or in high salt (200 mM KCl) conditions.Any variations apart from the above-mentioned conditions are specified in the figure legends.

DNaseI protection Assay
The DNaseI protection assays were carried out in presence of 500 ng of pBluescriptK plasmid either in presence of increasing protein concentrations (HupB full length and mutant protein) (0.2-2.5 mM) or constant protein concentration (2.5 mM).The reactions were performed in buffer B (20 mM HEPES-KOH pH 7.5, 5 mM magnesium chloride, 1 mM EDTA, 0.05 mM bovine serum albumin and 7% glycerol) and were incubated for 30 min at 25uC.The DNaseI digestion was initiated by the addition of 1 unit of DNaseI (PROMEGA, USA) for 30 sec or an increasing time interval ranging from 30 sec to 5 minutes as mentioned in figure legends.The reactions were stopped by the addition of 3 ml of 20% SDS and 2 ml of 0.5 M EDTA.The reaction products were subsequently resolved by electrophoresis on 1% agarose gel in 16 TAE.

DNA protection from metal catalyzed oxidation system
The ability of HupB Mtb and HupB MtbN to protect 500 ng of plasmid pBluescript II SK (+) from Metal catalyzed oxidation system (MCO) was tested in presence of either 1 mM of HupB Mtb or HupB MtbN in a reaction volume of 15 ml for 30 minutes at room temperature.MCO (0.4 mM FeCl 3 , 10 mM DTT, 100 mM Ethanol and 2 mM H 2 O 2 ) was added to the protein DNA complex and the reaction was further continued for 30 min.The reaction was terminated by the addition of 10 mM EDTA, phenol extracted and analyzed on a 1% agarose gel in 16 TAE at 7 V/cm for 30 minutes.

Supporting Information
Methods S1 A detailed description of method employed for multiple sequence alignment, calculation of dissociation constant and estimation of secondary structure of protein.

Figure 1 .
Figure 1.HupB Mtb shows higher binding affinity to double stranded DNA.Comparative gel retardation analysis showing the binding of increasing amounts of (A) HupB Mtb (0-120 nM, lanes 1-12) and (C) HupB MtbN (0-1200 nM, lanes 1-15) to radioactively labeled ds DNA under low salt conditions (10 mM KCl) (sequences shown in Table 2).Dissociation constants (Kd) were also calculated (B) for HupB Mtb and (D) for HupB MtbN binding to ds DNA.The Kd values are indicated as inset in the respective panels.Sequence of the oligonucleotide A1 used here is provided at the bottom of the figure.doi:10.1371/journal.pone.0012551.g001 lanes 9-16).

Figure 6 .
Figure 6.CTR of HupB Mtb provides protection from DNaseI.(A) Timed digestion of supercoiled pBSK DNA with DNase I was carried out with 2.5 mM of HupB Mtb (A, lanes 1-8) or HupB MtbN (lanes 9-15).Lanes 1, 2 and 9 have no protein while lanes 3-8 and lanes 10-15 have 2.5 mM of HupB Mtb and HupB MtbN , respectively.The reaction was initiated by the addition of one unit of DNaseI to all the lanes except lane 1. Incubation time with DNaseI in lanes 2-3 and 9-10 was 30 sec; lanes 4 and 11 for 1 min; lanes 5 and 12 for 2 min; lanes 6 and 13 for 3 min, lanes 7 and 14 for 4 min and lanes 8 and 15 for 5 min respectively.Area quantitated in this figure is marked by bracket and the amount of DNA (%) in each lane is indicated at the bottom of figure.(B) Protection of DNA from H 2 O 2 mediated damage.Lane 1, pBSK DNA alone; lane 2, pBSK treated with FeCl 3 and H 2 O 2 ; lane 3, first incubated with HupB Mtb and then treated with FeCl 3 and H 2 O 2 and lane 4, first incubated with HupB MtbN and then treated with FeCl 3 and H 2 O 2 as indicated in material and methods.The lower band in Figure 6B represents supercoiled DNA while the upper band represents relaxed DNA.The band marked by arrowhead was quantitated and the amount of DNA (%) in each lane is indicated at the bottom of figure.doi:10.1371/journal.pone.0012551.g006 Figure S1 A) Sequence alignment of HU protein from M. tuberculosis (M._tb_HU), E. coli (E._coli), S. enterica (S._enterica_Hup) and B. subtilis (B._subtilis_HU) using Clustal W programme shows conservation of N terminal region.Completely conserved residues are shaded green, identical residues are shaded yellow, similar residues are shaded cyan and different residues are white.B) Sequence alignment between M. tuberculosis HupB (M._tb_HU) and sea urchin Histone H1 (Sea_Urchin_H1-g) revealed conservation of amino acids at C terminal region.The rectangular shaded boxes represent conserved PAKK and KAAK amino acid residues.Found at: doi:10.1371/journal.pone.0012551.s002(0.66 MB PDF) Figure S2 Amino acid sequence alignment of HupB Mtb from M. tuberculosis (M.tb) with M. bovis, M. bovis BCG, M. ulcerans, M. marinum, M. leprae, M. kansasii, M. smegmatis, M. avium, M. parascrofula, M. gilvium, M. vanbaalenii and M. abscessus.Completely conserved residues are shaded green, identical residues are shaded yellow, similar residues are shaded cyan and different residues are white.Found at: doi:10.1371/journal.pone.0012551.s003(1.35 MB PDF) Figure S3 Plots indicating the half maximal saturation of A) 2nucleotide gap DNA (high salt), B) 2-nucleotide gap DNA (low salt), C) Cruciform DNA (high salt), D) 1-nucleotide gap DNA (low salt), E) nick DNA (high salt), in presence of indicated HupB Mtb .Respective Kd values are shown as inset in each panel.Found at: doi:10.1371/journal.pone.0012551.s004(1.89 MB EPS)

Table 2 .
Sequence of oligonucleotides used for Electrophoretic Mobility Shift Assays.Figure S4 Plots indicating the half maximal saturation of A) nick DNA B) 1-nucleotide gap DNA, C) 2-nucleotide gap DNA, D) Cruciform DNA, in presence of indicated HupB MtbN in low salt conditions.Respective Kd values are shown as inset in each panel.Found at: doi:10.1371/journal.pone.0012551.s005(1.54 MB EPS) Figure S5 Far UV CD spectrum analysis of HupB Mtb and HupB MtbN was carried out as described in supplementary methods and shown by arrowheads.HupB Mtb has 8.3% as a helix, 50.8% as b sheet, 7.6% as turn and 33.3% as random, however HupB MtbN has 16.9% as a helix, 35.2% as b sheet, 16.9% as turn and 31.1% as random.Found at: doi:10.1371/journal.pone.0012551.s006(1.36 MB EPS)