Molecular characterization of secretory proteins Rv3619c and Rv3620c from Mycobacterium tuberculosis H37Rv Anjum Mahmood 1 , Shubhra Srivastava 1 , Sarita Tripathi 1 , Mairaj Ahmed Ansari 2 , Mohammad Owais 2 and Ashish Arora 1 1 Molecular and Structural Biology Division, Central Drug Research Institute, Lucknow, India 2 Interdisciplinary Biotechnology Unit, Aligarh Muslim University, Aligarh, India Keywords binding constant; Mycobacterium tuberculosis; Rv3619c; thermal unfolding; vaccine Correspondence A. Arora, Molecular and Structural Biology Division, Central Drug Research Institute, Lucknow 226001, India Fax: +91 522 223405 Tel: +91 522 2612411 ext: 4329 E-mail: ashishcdri@yahoo.com (Received 5 September 2010, revised 1 November 2010, accepted 9 November 2010) doi:10.1111/j.1742-4658.2010.07958.x Rv3619c and Rv3620c are the secretory, antigenic proteins of the ESAT- 6 ⁄ CFP-10 family of Mycobacterium tuberculosis H37Rv. In this article, we show that Rv3619c interacts with Rv3620c to form a 1 : 1 heterodimeric complex with a dissociation constant (K d ) of 4.8 · 10 )7 M. The thermal unfolding of the heterodimer was completely reversible, with a T m of 48 °C. The comparative thermodynamics and thermal unfolding analysis of the Rv3619c–Rv3620c dimer, the ESAT-6–CFP-10 dimer and another ESAT family heterodimer, Rv0287–Rv0288, revealed that the binding strength and stability of Rv3619c–Rv3620c are relatively lower than those of the other two pairs. Molecular modeling and docking studies predict the structure of Rv3619c–Rv3620c to be similar to that of ESAT-6–CFP-10. Spectroscopic studies revealed that, in an acidic environment, Rv3619c and Rv3620c lose their secondary structure and interact weakly to form a com- plex with a lower helical content, indicating that Rv3619c–Rv3620c is destabilized at low pH. These results, combined with those of previous studies, suggest that unfolding of the proteins is required for dissociation of the complex and membrane binding. In the presence of membrane mimetics, the a-helical contents of Rv3619c and Rv3620 increased by 42% and 35%, respectively. In mice, the immune response against Rv3619c pro- tein is characterized by increased levels of interferon-c, interleukin-12 and IgG 2a , indicating a dominant Th1 response, which is mandatory for protec- tion against mycobacterial infection. This study therefore emphasizes the potential of Rv3619c as a subunit vaccine candidate. Structured digital abstract l MINT-8056093: Rv0288 (uniprotkb:P0A568) and Rv0287 (uniprotkb:O53692) bind (MI:0407) by isothermal titration calorimetry ( MI:0065) l MINT-8055978: Rv3620c (uniprotkb:O07932) and Rv3619c (uniprotkb:P96364) bind ( MI:0407)bycircular dichroism (MI:0016) l MINT-8055964: Rv3620c (uniprotkb:O07932) and Rv3619c (uniprotkb:P96364) bind ( MI:0407)byisothermal titration calorimetry (MI:0065) Abbreviations ASA, accessible surface area; BCG, bacille Calmette–Gue ´ rin; DMPC, dimyristoylphosphatidylcholine; DPC, dodecylphosphocholine; HRP, horseradish peroxidase; IFN, interferon; IL, interleukin; ITC, isothermal titration calorimetry; MRE, mean residual ellipticity; MTT,3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; RD, region of deletion; SI, stimulation index; TFE, trifluoroethanol. FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS 341 Introduction Comparative genomic studies based on whole genome DNA microarrays have led to the identification of 16 regions of deletion (RDs) in Mycobacterium bovis bacille Calmette–Gue ´ rin (BCG), which is currently used as a vaccine, with respect to Mycobacterium tuberculosis, and five RDs with respect to M. bovis. The RD encom- passing ORFs Rv3619c and Rv3620c is absent from all vaccine strains of M. bovis. This region has been classi- fied as RD9 by Behr et al. and as RD8 by Gordon et al. [1,2]. It is a stretch of 5516 bp encompassing seven ORFs (Rv3617 to Rv3623). Rv3619c and Rv3620c are the ESAT-6 ⁄ CFP-10 family members. ORFs Rv3621c and Rv3622c belong to the family encoding proteins containing sequence motif Pro-Pro-Glu (PPE) and Pro- Glu (PE), respectively. The region also contains an epox- ide hydrolase encoded by Rv3617, which may be involved in detoxification, catabolism and regulation of signaling molecules [3]. Rv3618 and Rv3623 encode a probable monooxygenase and lipoprotein, respectively. Rv3619c and Rv3620c are secretory proteins of 94 and 98 amino acids, respectively, reported in culture filtrates of M. tuberculosis [4–6]. In silico studies have predicted their presence in Mycobacterium leprae, Mycobacterium avium and Mycobacterium marinum [7]. They belong to the ESAT-6 family, which comprises 23 members. However, they share only 20% sequence identity with ESAT-6 and CFP-10. Within the ESAT-6 family, Rv3619c and Rv3620c are constituted within a subfamily comprising Rv1037c ⁄ Rv1038c, Rv1197 ⁄ Rv1198, Rv1792 ⁄ Rv1793 and Rv2346c ⁄ Rv2347c [8]. The members within this subfamily share > 90% amino acid sequence identity with Rv3619c ⁄ Rv3620c. Overlapping synthetic peptide studies have demon- strated that the Rv3619c ⁄ Rv3620c subfamily consists of potent T-cell antigens [9]. Detailed studies of ESAT-6 and CFP-10 have revealed that they interact strongly to form a 1 : 1 heterodimeric stable complex with a four-helix bundle in which each protein bears a central WXG motif [10,11]. Similar results have been obtained with Rv0287–Rv0288 [12]. However, deviations from the basic prototype structure of the ESAT-6–CFP-10 com- plex have also been reported. Recent crystallographic studies have suggested that Rv3019c–Rv3020c (ESAT-6 ⁄ CFP-10 homologs) exists as a heterotetramer. Rv3020c contains histidine in place of tryptophan in the WXG motif, which induces the formation of a small helix that joins the N-terminal and C-terminal domains [13]. The ESAT-6 ⁄ CFP-10 homologs in Staphylococcus aureus (Sa EsxA and SaEsxB) do not form a hetero- dimer complex; rather, they homodimerize. The crystal structure of the homodimer of SaEsxA has been deter- mined. Furthermore, sequence analysis has predicted that SaEsxB (CFP-10 homolog) will have a structure similar to that of SaEsxA; on the basis of this, it has been suggested that the two proteins may work inde- pendently [14]. These variations among complexes indi- cate that different homologs and paralogs of the ESX family may have different structural properties that may lead to further functional dissimilarities. There- fore, a separate detailed analysis for each pair is required, structurally as well as functionally. One of the major functions associated with ESAT-6 is its cytolytic activity. Hsu et al. have shown the cytolysis of host cells by ESAT-6 secreted by intracel- lular mycobacteria [15]. We have previously shown that ESAT-6 adopts significant helical structure in the presence of dimyristoylphosphatidylcholine (DMPC) vesicles and dodecylphosphocholine (DPC) micelles, indicating membrane binding. Furthermore, only ESAT-6, and not CFP-10 or the complex, was found to interact with lipid membranes [16]. Moreover, Smith et al. have shown that ESAT-6 induces pore formation in M. marinum in a dose-dependent manner, enabling the bacterium to escape from the vacuole to the host cell cytosol [17]. The membrane-binding ability was related to disruption of the complex by Jonge et al. [18]. These authors demonstrated that, in the acidic phagosomal environment, the 1 : 1 ESAT-6–CFP-10 complex dissociated to free ESAT-6 to exert its cyto- lytic activity. However, no studies have been per- formed on other ESAT-6 ⁄ CFP-10 paralogs to determine whether other members have similar mem- brane-destabilizing functions. The ESAT-6 family is a potential source of T-cell antigens, which can be exploited for the development of suitable vaccines against mycobacteria. ESAT-6 and CFP-10 are the most widely studied proteins of this family. ESAT-6 and CFP-10 activate the Th1 response which is marked by T-cell proliferation and interferon (IFN)-c release [19,20]. Subunit-based or DNA-based ESAT-6 vaccines have been prepared, and their protec- tive efficacy has been evaluated. The first DNA-based vaccine involving ESAT-6 was reported by Kamath et al. [21]. Although this vaccine provided a significant level of protection against mycobacteria, its potency was found to be lower than that of BCG. Brandt et al. prepared the first single protein subunit tuberculosis vaccine, using ESAT-6 with dioctadecylammonium bromide and monophosphoryl lipid A, that conferred protection similar to that provided by BCG [22]. A recombinant chimeric fusion protein of ESAT-6 and Characterization of Rv3619c and Rv3620c A. Mahmood et al. 342 FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS Ag85B was found to provide better protection than individual proteins or a mixture of them, and is cur- rently being evaluated as a promising vaccine candi- date [23,24]. Besides ESAT-6, other members evaluated for immune response were Rv0288, Rv3019c, and Rv3017c [25]. Rv3019c demonstrated its potential to be used as a heterologous prime booster in conjunction with BCG. In the present study, we characterized the forma- tion of the complex between the ESAT-6 family protein Rv3619c and its pairing partner Rv3620c, using isothermal titration calorimetry (ITC) and CD spectroscopy. We also examined the structural prop- erties of the proteins by spectroscopy and molecular modeling. Furthermore, we evaluated the immune response of Rv3619c in free antigenic form in mice by determining lymphocyte proliferation, cytokine levels, and antigen-specific antibody levels. Our study provides insights into the properties of these secre- tory proteins. Results and Discussion Interaction of Rv3619c and Rv3620c ITC experiments were performed to determine the thermodynamic parameters governing formation of the complex between Rv3619c and Rv3620c. The raw ITC data and integrated areas under each peak versus Rv3619c ⁄ Rv3620c molar ratio are shown in Fig. 1A. The binding isotherm was fitted to a single-site binding model for determination of thermodynamic parame- ters. The parameters used in fitting were the stoichiom- etry of association (n), the binding constant (K b ), and the change in enthalpy (DH b ). The values of these parameters obtained from the nonlinear least-squares fit to the binding curve are as follows: n = 1.0, K b = (2.05 · 10 6 ) ± (3.24 · 10 5 ) m )1 and DH b = )3.35 · 10 4 ± 760.1 calÆmol )1 . The saturation of heat released at a molar ratio of 1.0 strongly suggests that the proteins form a 1 : 1 heterodimeric complex. The Fig. 1. Thermodynamic and spectroscopic studies on Rv3619c–Rv3620c. (A) ITC measurements of the interaction between Rv3619c and Rv3620c in phosphate buffer at 25 °C; raw data of heat effect (lcalÆs )1 ) of 30 injections (10 lL each) of 0.1 mM Rv3619c into 1.43 mL of 0.01 m M Rv3620c. The data points (j) were obtained by integration of heat signals plotted against the Rv3619c ⁄ Rv3620c molar ratio in the reaction cell. The solid line represents a calculated curve using the best-fit parameters obtained by a nonlinear least square fit. The heat of dilution was subtracted from the raw data of titration of Rv3619c with Rv3620c. (B) Far-UV CD spectra of Rv3619c, Rv3620c, and the 1 : 1 complex. CD spectra of 5 l M Rv3619c (j), Rv3620c (•) and the 1 : 1 complex (m) in phosphate buffer (pH 6.5, 25 °C. (C) Normalized transi- tion curves for temperature-induced transition of the complex monitored in the far-UV CD region at 222 nm. The thermal unfolding (j) and thermal refolding (•) profiles of the complex were plotted as fraction of protein folded versus temperature in °C. A. Mahmood et al. Characterization of Rv3619c and Rv3620c FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS 343 dissociation constant of the complex ( K d =1⁄ K b ) was 4.8 · 10 )7 m. The free energy change (DG) and entropy change (DS) associated with complex formation were )8.55 kcalÆmol )1 and )83.7 calÆmol )1 ÆK )1 , respectively, at 25 °C. A comparative thermodynamic analysis of ESAT-6⁄ CFP-10 family members is shown in Table 1. Although Rv0287–Rv0288 is well characterized [12], we performed ITC experiments with Rv0287 and Rv0288 to generate the thermodynamic data for their interaction. The ITC data for ESAT-6–CFP-10 have already been obtained [16]. Data analysis revealed that Rv0288–Rv0287 has the strongest binding affinity (K d $ 10 nm), followed by ESAT-6–CFP-10 (K d $ 50 nm). Rv3619c–Rv3620c showed the weakest binding (K d $ 480 nm). The bind- ing free energy (DG) of Rv3619c–Rv3620c revealed dif- ferences of 2.26 kcalÆmol )1 relative to Rv0287–Rv0288 and 1.4 kcalÆmol )1 relative to ESAT-6–CFP-10. This suggests that binding of Rv0287 and Rv0288 is energeti- cally more favored. The difference in entropy values indicates that the conformational freedom of side chains in Rv3619c–Rv3620c is comparatively greater. The dif- ferences in binding affinity among ESAT-6 ⁄ CFP-10 par- alogs suggest that the binding equilibrium in loosely bound complexes may be shifted to the reactant side, resulting in the release of unbound proteins under cer- tain specific conditions, such as those in the acidic phag- osomal environment. The conformational changes associated with com- plex formation were estimated by recording far-UV CD spectra. The CD spectra of Rv3619c, Rv3620c and the 1 : 1 complex were recorded (Fig. 1B) at 25 °C, and data were analyzed by the k2d server. The CD spectra showed that the two proteins and their 1 : 1 complex adopt a predominantly a-helical conforma- tion. The a-helical contents of Rv3619c, Rv3620c and the 1 : 1 complex were approximately 33%, 34% and 70%, respectively. The secondary structure of Rv3619c–Rv3620c was very similar to that of ESAT- 6–CFP-10 and Rv0288–Rv0287. However, unlike CFP- 10 and Rv0287, which are unstructured, Rv3620c had a significant a-helical content, even in the uncomplexed state. The lower values of DH and DS (Table 1) observed for Rv3619c–Rv3620c than for ESAT-6– CFP-10 and Rv0287–Rv0288 could result from the dif- ference in the folding state of Rv3620c from that of CFP-10 and Rv0287. Thermal unfolding of proteins The presence of any stable tertiary structure in Rv3619c and Rv3620c was ruled out by thermal unfolding experiments (data not shown). The two pro- teins were denatured when the temperature was increased from 25 °Cto80°C, following non-coopera- tive unfolding, and the structure was not regained on cooling, indicating that they lack any stable tertiary structure. The complex, however, demonstrated signifi- cant resistance to denaturation before melting. The unfolding started at 38 °C, following a cooperative pathway with a denaturation midpoint (T m )of48°C (Fig. 1C). Rv3619c–Rv3620c showed lower thermal stability than ESAT-6–CFP-10 (T m =54°C) and Rv0287–Rv0288 (T m =70°C) [12]. This indicates that Rv3619c–Rv3620c has a smaller intermolecular hydro- phobic overlapping interface. Overall, the thermal denaturation profile and ITC data suggest that Rv3619c forms a loose complex with its genomic part- ner Rv3620c, because of a smaller protein–protein interaction surface area than that in ESAT-6–CFP-10 and Rv0287–Rv0288. The thermal renaturation profile of the 1 : 1 com- plex was recorded by cooling the sample from 80 °C to 25 °C (Fig. 1C). On reversal of the temperature, the complex completely regained its secondary structure, retracing a similar path. This characteristic feature was also observed for ESAT-6–CFP-10 [16]. The two pro- teins complement each other to attain a folded struc- ture. However, in the absence of the other partner, they lose their structure irreversibly. This implies that Rv3620c, like CFP-10, functions to keep Rv3619c in a structured and soluble form under physiological conditions. Modeling and docking The secondary structure was also analyzed by molecu- lar modeling and docking experiments. On the basis of Table 1. A comparative analysis of thermodynamic parameters of ESAT–CFP-10, Rv0287–Rv0288, and Rv3619c–Rv3620c. The stoichiome- try of interaction and values of K b and DH were determined by ITC. DG and DS were calculated from the thermodynamic formula DG = )RT ln K b = DH ) TDS. ESX pairs NK b (M )1 ) K d (M) DH (kcalÆmol )1 ) DS (calÆmol )1 ÆK )1 ) DG (kcalÆmol )1 ) Rv3619c–Rv3620c 1.0 2.0 · 10 6 4.8 · 10 )7 )33.5 )83.7 )8.55 ESAT-6–CFP-10 1.0 2.0 · 10 7 5.0 · 10 )8 )40.3 )101 )9.95 Rv0287–Rv0288 1.0 9.2 · 10 7 1.0 · 10 )8 )40.8 )100 )10.81 Characterization of Rv3619c and Rv3620c A. Mahmood et al. 344 FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS the solution structure of the 1 : 1 ESAT-6–CFP-10 complex, we generated molecular models of Rv3619c and Rv3620c using i-tasser and swiss model, respec- tively, and molecules were docked using the patch dock server. The model Rv3619c–Rv3620c is shown in Fig. 2. We determined the accessible surface area (ASA) for the Rv3619c–Rv3620c model and the ESAT-6–CFP-10 solution structure by using the dis- covery studio 2.1 package. The ASA for docked Rv3619c–Rv3620c was 12 190 A 2 , and the buried surface area was 1211 A 2 . The ASA for the ESAT-6– CFP-10 solution structure was 11 512 A 2 , and the buried surface area was 1675 A 2 . The buried surface area for docked Rv3619c–Rv3620c was smaller than that for the ESAT-6–CFP-10 solution structure, which is in complete agreement with our ITC data. The model suggests that the two proteins interact with 1 : 1 stoichiometry, lying antiparallel to each other, with each one having two helixes separated by a loop containing the WXG motif. The predicted struc- ture resembles the four-helix bundle packing of ESAT- 6–CFP-10. The crystal structure of Rv3019c–Rv3020c suggested that replacement of tryptophan by histidine in the WXG motif induced the formation of a small helix in Rv3020c, joining the N-terminal and C-termi- nal helices, conferring a tetramer structure. However, as the WXG motif in the Rv3620c is strictly con- served (Trp45 and Gly47), it is highly unlikely that Rv3619c–Rv3620c would form a tetrameric complex like Rv3019c–Rv3020c. Despite the predicted structure being similar to that of ESAT-6–CFP-10, a consider- able difference at the C-terminus of Rv3620c was noticed. The C-terminal end of CFP-10 is unstruc- tured, whereas the modeling and docking results pre- dicted that Rv3620c would possess a C-terminal helix. The C-terminal end of SaEsxA from S. aureus also contains a folded region [14]. In order to identify the residues forming the inter- molecular contact surface of Rv3619c–Rv3620c, we generated a contact map for docked Rv3619c and Rv3620c, using the acclerys discovery studio 2.1 software package. The contact map analysis predicted that Val10, Ile17, Ala21, Leu24, Ala26, Ala30, Ile31, Ile32, Val35, Leu36, Ala38, Phe41, Cys50, Phe53, Leu57, Phe61, Val63, Ile64, Ala68, Ala70, Val75, Ala77, Ala78, Met82, Val89 and Ala94 of Rv3619c and Met12, Met15, Ala16, Phe19, Val21, Ala23, Val26, Ala30, Met33, Ala35, Ala37, Ile40, Ala43, Met48, Ala49, Leu54, Met57, Met60, Phe64, Ile67, Val68, Met70, Leu71, Val74, Leu78, Val79 and Ala82 of Rv3620c are likely to form hydrophobic contact surfaces. We plotted the N-terminal and C-terminal helices of Rv3619c and Rv3620c on a heptad repeat helical wheel, on the basis of optimum sequence alignment (Figs S1 and S2). Hydrophobic residues suggested by contact map analysis occupied predomi- nantly ‘a’ and’d’ positions, suggesting that they form the core of helix bundle packing. This is in full agree- ment with the models suggested for ESAT-6–CFP-10 and Rv0287–Rv0288. The complex was predicted to be stabilized by formation of a single salt bridge between Glu25 of Rv3619c and Arg31 of Rv3620c. The residues highlighted in Fig. 2, in the Rv3619c helix, represent the nonconserved, semiconserved or substituted conserved residues at the corresponding positions among the related members Rv1037c, Rv1198, Rv1793 and Rv2346c. Substitution of residues at these positions severely alters antigenic recognition by the cell. Alderson et al. demonstrated that a T-cell line specific for Rv1198 failed to recognize peptides from Rv1793 and Rv3619c with amino acid substitu- tions at the 22nd and 23rd positions [9]. Effect of pH on complex stability We analyzed pH-induced conformational changes in Rv3619c, Rv3620c and the 1 : 1 complex by recording the far-UV CD spectra at different pH values. The mean residual ellipticity (MRE) at 222 nm was plotted against different pH values, as shown in Fig. 3A. On G22 / A22 S23 / L23 S33 / R33 T37 / A37 S39 / G39 A 48 / V48 G52 / E52 I32 / V32 N Rv3620c C Rv3620c N Rv3619c C Rv3619c Fig. 2. In silico modeling and docking of Rv3619c and Rv3620c. The two proteins form a 1 : 1 complex. Rv3619c is shown in blue and Rv3620c is shown in gray. The nonconserved, semiconserved and substituted conserved regions of paralogs of Rv3619c (Rv1037, Rv1198, Rv1793, and Rv2346c) are highlighted in blue, orange, and yellow. A. Mahmood et al. Characterization of Rv3619c and Rv3620c FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS 345 lowering of the pH from 6.5 to 4.5, Rv3619c and Rv3620c showed considerable loss of conformation, although Rv3619c resisted conformational change until pH 5.5. However, on mixing of the two proteins at pH 4.5, the resultant spectra, shown in Fig. 3B, dem- onstrated significant loss of structure. This could be attributable to weak interactions between two unfolded or partially folded proteins. Furthermore, it indicates that the structural unfolding of individual proteins could be the event that results in dissociation of the complex at acidic pH. Recently, Arbing et al. also reported the pH-induced dissociation of Rv3019c– Rv3020c (unpublished data) [13]. Effect of trifluoroethanol (TFE) and DPC micelles on the conformation of Rv3619c and Rv3620c Previous work suggested that ESAT-6 adopts a helical structure, and is sequestered and induces pore formation in the cell membrane. Furthermore, only ESAT-6, and not CFP-10, was associated with cytolytic activity [16,26]. To investigate the role of Rv3619c, Rv3620c and the 1 : 1 complex in membrane binding, CD spectros- copy experiments were performed in the presence of TFE and DPC micelles, as shown in Fig . 4A. In 40% TFE, Rv3619c, Rv3620c and the 1 : 1 complex adopted a highly folded structure. In 20 mm DPC, Rv3619c and Rv3620c demonstrated 42% and 35% increases, respec- tively, in helical content. However, the 1 : 1 complex did not show any significant change in conformation in 20 mm DPC. This implies that the two proteins bind to the membrane individually, but not after they form a complex, as also observed for ESAT-6 and CFP-10. Furthermore, we recorded the intrinsic tryptophan fluo- rescence of Rv3619c and Rv3620c in 20 mm DPC at wavelengths ranging from 300 to 400 nm (Fig. 4B). The k max values of Rv3619c and Rv3620c shifted to lower wavelengths by 5 nm and 10 nm, respectively, and this was accompanied by enhancements of fluorescence intensity, suggesting the relocation of tryptophans in the hydrophobic environment because of membrane bind- ing. We also checked the binding of the two proteins to small unilamellar vesicles of DMPC, using CD spectros- copy. Our preliminary results suggested that Rv3619c, but not Rv3620c, underwent a change in helicity in the presence of DMPC small unilamellar vesicless. This does not correlate with the change in CD spectra observed in the presence of DPC micelles. However, interestingly, it does correlate with the binding of the fluorescent dye 8- anilinonapthalene-1-sulfonate, which was observed for Rv3619c but not for Rv3620c (data not shown). Previous membrane-binding studies with ESAT-6 suggested that the deeper integration of the protein into phospholipids is related to its unfolding and struc- tural transition [16]. We also found unfolding of Rv3619c and Rv3620c at pH 4.5. In some cases, phagosomes containing mycobacteria may advance to the phagolysosomal stage, with concomitant lowering of the pH from 5.1 to 4.8–4.5. It has been suggested that, under such conditions, ESAT-6 could dissociate from the complex and bind the lipid membranes [17]. Considering the two events together, it can be inferred that unfolding of proteins and formation of a struc- tural intermediate is associated with membrane bind- ing. The acidic phagolysosome may provide an environment that stimulates dissociation of the com- plex and structural reorientation of proteins, allowing them to assemble and penetrate the membrane deeply. Taking into account that mycobacteria in infected cells inhibit the fusion of lysosomes with phagosomes, Rv3619c–Rv3620c would most likely exist as a com- plex. However, under certain conditions when the infected cells also contain phagolysosomes [27,28], the 1 : 1 complex may be disrupted, and Rv3619c and Fig. 3. pH-induced conformational changes in Rv3619c, Rv3620c, and the 1 : 1 complex. (A) MREs of Rv3619c (j), Rv3620c (d) and the 1 : 1 complex (m) at 222 nm were plotted against different pH values. (B) Far UV-CD spectra of the 1 : 1 complex at pH values of 6.5 (j), 5.5 (d), and 4.5 (m). Characterization of Rv3619c and Rv3620c A. Mahmood et al. 346 FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS Rv3620c may execute their functions independently and bind to lipid membranes. Although no pH-based structural studies have so far been performed with ESAT-6, our study supports the hypothesis of Simeone et al. [26] that the biological activity of ESAT-6 depends on the pH of phagosomal compartments. Proteins of the ESAT-6 family are not only involved in spreading virulence by exhibiting cytolytic activity; they also confer protection against M. tuberculosis through the antimicrobial Th1 host immune response. ESAT-6, Rv0288 and Rv3019c are known to stimulate T-cells to proliferate and protect against mycobacteria [25,29,30]. Th1 activation results in IFN-c and inter- leukin (IL)-12 release, whereas Th2 activation releases IL-4. A balanced Th1 ⁄ Th2 response is important for clearance of mycobacteria. To determine whether Rv3619c also activates T-cells to release cytokines, we examined its role in the immune response in mice. Analysis of immune response of Rv3619c We evaluated the immune response of Rv3619c by injecting the free antigen in NaCl ⁄ P i in Balb ⁄ c mice, with NaCl ⁄ P i as a control. Initially, the antigen-induced lymphocyte proliferation activity of Rv3619c was assessed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl- tetrazolium bromide (MTT) assay. The proliferative response of Rv3619c was demonstrated by stimulation Fig. 4. Membrane binding of Rv3619c, Rv3620c, and the 1 : 1 complex. (A) Far-UV CD spectra of 5 lM Rv3619c, Rv3620c and the 1 : 1 complex in the presence of phosphate buffer (j), 40% TFE (•), and 20 m M DPC (m) are shown. All of the spectra were recorded at 25 °C. (B) Fluorescence emission spectra of Rv3619c and Rv3620c in phosphate buffer (j) and 20 m M DPC (•). A. Mahmood et al. Characterization of Rv3619c and Rv3620c FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS 347 index (SI), as shown in Fig. 5A; it was determined to be 4.695 ± 0.24 pgÆmL )1 postbooster, with a statistically significant difference (P < 0.05) from the NaCl ⁄ P i group. The significant augmentation of antigen-specific proliferation clearly demonstrates the presence of immunologically active lymphocytes in immunized mice. Statistically significant levels of IFN-c (832 ± 30.61 pgÆmL )1 ; P < 0.001) and IL-12 (481 ± 5.46 pgÆmL )1 ; P < 0.05) were produced by spleenocytes of mice immunized with Rv3619c (Fig. 5B); IL-4, a Th2 cytokine, was also detected in immunized groups, with a significant difference, but the level of IFN-c was very high when compared with that of IL-4. The humoral response was determined by measuring the Rv3619c- specific serum IgG level (Fig. 6A). Class switching showed a predominant IgG 2a response, as indicated by a postbooster IgG 2a ⁄ IgG 1 ratio greater than 1 (Fig. 6B) The high levels of IFN-c and IL-12 secretion, suggesting a biased Th1-type response of Rv3619c, is consistent with the IgG 2a ⁄ IgG 1 ratio. The study clearly demon- strates that Rv3619c is a potent T-cell antigen that may provide protection against mycobacterial infection if used in combination with a suitable adjuvant or in a mixture with other antigens. Rv3619c paralogs, i.e. Rv1037c, Rv1198, Rv1793 and Rv2346c and their genomic partners, share greater than 90% amino acid sequence identity. Structurally, they may form similar complexes to those observed for Rv3619c, Rv3620c, and other pairs, but the immune response may vary, as they display unique epitopes. The presence of similar structure and function but different immune responses has been related to the A B Fig. 5. Study of immune response to Rv3619c antigen. Analysis of the Rv3619c-specific cytokine profile of 6–8-week-old female Balb ⁄ c mice immunized with 20–25 lg of antigen. (A) Lymphocyte proliferation response of Rv3619c expressed in terms of SI. Each bar represents mean ± standard deviation. (B) IFN-c, IL-12 and IL-4 levels determined 2 weeks postimmunization and 2 weeks post- booster. Three mice per group were used, and the data obtained were statistically significant different, with ***P < 0.001, **P < 0.01, or *P < 0.05, from those obtained with NaCl ⁄ P i . A B Fig. 6. Estimation of humoral response to Rv3619c antigen. The antibody response against Rv3619c in mice (three mice per group) is shown. Serum was collected 2 weeks postimmunization and 2 weeks postbooster. The antibody level was estimated by record- ing the absorbance at 490 nm. (A) Total IgG content. (B) IgG 2a ⁄ IgG 1 ratio, indicating the biased Th1 response. Characterization of Rv3619c and Rv3620c A. Mahmood et al. 348 FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS mycobacterial strategy for escaping the host immune recognition system [30]. Studies with tuberculosis reac- tor animals also revealed that immunodominant epi- topes of the ESAT-6 family come from variable regions more than homologous regions, suggesting that mycobacteria may vary their antigenic load according to requirements, leading to antigenic drift that enables mycobacterial escape [31]. The presence of several pairs of ESAT family proteins within the M. tuberculo- sis genome suggests that they might be expressed under different physiological conditions, and are able to sub- stitute for each other functionally; this strategy enables them to survive longer in host cells. However, at the same time, sequence variations among the family mem- bers provide a pool of antigens that can generate the effector molecules that restrict mycobacterial growth. Evaluation of the formation of complexes and their protective efficacy will help in the development of suit- able vaccine candidates. Conclusion The ESAT-6 ⁄ CFP-10 family has 23 members, consisting of 11 pairs and one unpaired member. Although these 11 pairs are likely to form complexes in a manner simi- lar to the formation of ESAT-6–CFP-10, they may exhi- bit subtle variations in affinity, stability and immune response, which in turn may further define their individ- ual functional roles. A systematic study of various com- plex-forming pairs would improve our understanding of the evolutionary and functional relevance of the whole ESAT family. Our study further consolidates the hypothesis that the structural unfolding of individual proteins under conditions of acidic pH may be the key factor triggering the dissociation of complexes. How- ever, we believe that this aspect still needs more elegant, unambiguous, quantitative and time-resolved character- ization. The sequence variation in the ESAT-6 ⁄ CFP-10 family not only determines the stability of the complex, but also provides an antigenic pool in which a change in a single amino acid can change the host immune response. The efficacy of peptides and proteins incorpo- rating these antigens must be evaluated so that those conferring protection can be developed further. We are currently working in this direction. Experimental procedures Materials pET expression vectors were obtained from Novagen (Darmstadt, Germany). Oligonucleotides for gene isolation were from BIO Serve (Hyderabad, Andhra Pradesh, India). Restriction endonucleases, T4 DNA ligase and DNA size markers were from New England Biolabs (Beverly, MA, USA). Taq polymerase and other reagents for PCR, the plasmid miniprep kit and the gel extraction kit were from Qiagen. The Ni 2+ –nitrilotriacetic acid superflow metal- affinity chromatography matrix was from Qiagen. For protein concentration, Centricon membrane were used (molecular mass cut-off 3KDa: Millipore (India) Pvt. Ltd, Bangalore, India). The rest of the chemical reagents were from Sigma (New Delhi, India). Cloning, expression and purification Genomic DNA of M. tuberculosis H37Rv was prepared as described by Kremer et al. [32]. The genes encoding Rv3619c and Rv3620c were PCR-amplified with oligonu- cleotide primers and pfu DNA polymerase, and cloned into pET-NH6. This cloning strategy added an additional 30 residues at the N-terminus, including the six residues of the His-tag. The vectors containing the genes encoding Rv3619c and Rv3620c were then transformed into BL21(kDE3) Escherichia coli cells, which were grown in LB medium supplemented with ampicillin (100 lg Æ mL )1 ). BL21(kDE3) cells containing the plasmids pET-NH6– Rv3619c and pET-NH6–Rv3620c were grown in LB med- ium supplemented with ampicillin (100 lgÆmL )1 ) and induced at D 600 nm = 1.0 with a final concentration of 0.5 mm isopropyl thio-b-d-galactoside. The Rv3619c culture was grown for a further 12–14 h at 27 °C, and the Rv3620c culture for 6 h at 37 °C. All proteins were purified over a Ni 2+ –nitrilotriacetic acid matrix with a standard protocol under denaturing conditions, according to the manufac- turer’s instructions, except that NaCl and guanidine hydro- chloride were excluded from the buffer. The column fractions were checked for purity by SDS ⁄ PAGE (15% gel). The proteins were refolded by dialysis, with a buffer containing 25 mm NaH 2 PO 4 , 100 mm NaCl, and 1 mm EDTA (pH 6.5). Refolded Rv3619c and Rv3620c were dia- lyzed against buffer containing 20 mm NaH 2 PO 4 ,50mm NaCl, and 0.1% NaN 3 (pH 6.5). The pET-NH6–Rv3619c- encoded and pET-NH6–Rv3620c-encoded proteins con- tained 30 extra N-terminal residues with a His-tag. ITC The ITC experiments were performed at 25 °Cona VP-ITC calorimeter from Microcal (Northampton, MA, USA). The calorimeter was calibrated according to the user manual of the instrument. The proteins were dialyzed against buffer containing 20 mm NaH 2 PO 4 and 50 mm NaCl (pH 6.5). Samples were degassed prior to titration at 20 °C. The ITC experiments were performed by adding aliquots of Rv3619c to Rv3620c. The sample cell was filled with 1.43 mL of 0.01 mm Rv3620c and titrated against 0.1 mm Rv3619c. Thirty injections of 10 lL each were A. Mahmood et al. Characterization of Rv3619c and Rv3620c FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS 349 made at intervals of 180 s. The ITC data were analyzed with origin version 7. The amount of heat produced per injection was calculated by integration of the area under each peak with a baseline selected by origin. CD spectroscopy CD measurements were performed to determine the second- ary structure of Rv3619c, Rv3620c, and the 1 : 1 complex. The experiments were performed on a Jasco Spectropola- rimeter model J-810. The instrument was calibrated with (+)-10-camphorsulfonic acid. The protein spectra were recorded with the protein samples in buffer containing 20 mm NaH 2 PO 4 and 50 mm NaCl (pH 6.5). The protein concentration used was in the range 5–10 lm . Three scans were averaged for each spectrum. Isothermal wavelength scans were recorded in the range 250–200 nm, with a path- length of 2 mm, a response time of 1 s, a scan speed of 20 nmÆmin )1 , and a data pitch of 0.5. The CD results were expressed as MRE, in degree cm )2 Ædmol )1 , calculated as follows: MRE ¼ðh  100  M r Þ=ðcdN A Þ where h is the observed ellipiticity (°), c is the protein con- centration (mgÆmL )1 ), d is the pathlength (cm), and N A is the number of amino acids. Percentage secondary structure was calculated with the online k 2D server (http://www.embl. de/ ~ andrade/k2d/). For thermal denaturation studies, the spectra were recorded in the temperature range 25–80 °Cat a speed of 1 °CÆmin )1 . For refolding, the temperature was reversed at the same speed. The fraction of protein folded corresponding to the MRE at 222 nm was calculated from the equation ½h obs Àh den =½h nat Àh den  where h nat and h den are the MREs at 222 nm when proteins are in the native state at 25 °C, and in the denatured state at 80 °C. h obs is the observed MRE. To study the effect of membrane mimetic conditions on conformation of proteins, far-UV CD spectra were acquired in the presence of either 40% TFE or 20 mm DPC. DPC stock (200 mm) was prepared in buffer containing 20 mm NaH 2 PO 4 and 50 mm NaCl (pH 6.5), and centrifuged at 18 500 g to remove any suspended particles. DPC was added to 5 lm protein to a final concentration of 20 mm. For TFE experiments, 5 lm protein was added to 40% TFE. The spec- tra were recorded in the wavelength range 250–200 nm and analyzed on the k 2D server. Protein modeling and docking The protein sequences of Rv3619c and Rv3620c were taken from the TB Structural Genomics Consortium (http:// www.doe-mbi.ucla.edu/TB/), and sequences were aligned using a server (http://www.ebi.ac.uk/clustalw). Models of Rv3619c and Rv3620c were generated using online servers (http://zhang.bioinformatics.ku.edu/I-TASSER/ and http:// swissmodel.expasy.org//SWISS-MODEL.html, respectively). The modeled structures of Rv3619c and Rv3620c were docked using patchdock (http://bioinfo3d.cs.tau.ac.il/ PatchDock), a geometry-based molecular docking algo- rithm. The docked complex was analyzed with accelrys discovery studio 2.0. Fluorescence spectroscopy Fluorescence spectra were acquired to record the intrinsic tryptophan fluorescence changes of Rv3619c, Rv3620c and the 1 : 1 complex in the presence of 40% TFE and 20 mm DPC. Fluorescence spectra were acquired at 25 °Cona Perkin-Elmer Life Sciences LS 50B spectroluminescence meter, with a 5-mm-pathlength quartz cell. Protein (1 lm) was mixed with either 40% TFE or 20 mm DPC. The maxi- mum intrinsic fluorescence was monitored to record the wavelength shift. The fluorescence emission spectra were recorded in the range 300–400 nm, with an excitation wave- length of 280 nm. Animals and immunization Female Balb ⁄ c mice (6–8 weeks old) were purchased from the JALMA Institute for Leprosy and Other Microbial Diseases, Agra, India. Mice were maintained in the animal facility, and the techniques used for injection and bleeding of animals were performed in strict accordance with the mandates approved by the Animal Ethics Committee (CPCSEA, Gov- ernment of India). Ten mice in two groups were taken for study. One group was immunized with antigen in NaCl ⁄ P i , and the other group (control) was injected with NaCl ⁄ P i only. The immunization volume was 100 lL per animal, and the antigen dose was 20–25 lg per injection. The mice were immunized by subcutaneous injection in the lower abdominal region. Two weeks postimmunization, a single booster with same amount of antigen was given to each animal. Lymphocyte proliferation Cells were grown in RPMI-1640 with 10% fetal bovine serum and 1% antimycotic solution for 48 h in a CO 2 incu- bator at 37 °C, 5% CO 2 and appropriate humidity under aseptic conditions, with 20 lg of antigen and 2 lgÆmL )1 concanavalin A. Four to five hours before completion of incubation, 25 lLof55mgÆmL )1 of MTT stock solution was added to each well in 96-well plates to attain a final concentration of 1 mgÆmL )1 . One well was kept blank; that is, before addition of the MTT, 100 lL of lysis buffer was added to the well. When dark crystals appeared, 50 l Lof tissue culture-grade dimethylsulfoxide was added to each well. Plates were then incubated for 2 h in a CO 2 incubator Characterization of Rv3619c and Rv3620c A. Mahmood et al. 350 FEBS Journal 278 (2011) 341–353 ª 2010 CDRI. Journal compilation ª 2010 FEBS [...]... absorbance was then measured at 540 nm The data are presented here in the form of SI An SI > 2 was considered to be significant: SI = (D540 nm of activated cells – D540 nm of inactivated cells) ⁄ D540 nm of inactivated cells Characterization of Rv3619c and Rv3620c at 37 °C for 40 min The reaction was terminated by the addition of 50 lL of 1 m H2SO4 The absorbance was read at 490 nm with a microtiter plate reader... T-cell antigens of the Mycobacterium tuberculosis complex ESAT-6 and CFP-10 form a tight, 1:1 complex and characterization of the structural properties of ESAT-6, CFP-10, and the ESAT-6:CFP-10 complex J Biol Chem 277, 21598– 21603 12 Lightbody KL, Ilghari D, Waters LC, Carey G, Bailey MA, Williamson RA, Renshaw PS & Carr MD (2008) Molecular features governing the stability and specificity of functional... HG(2007) Comprehensive analysis of exported proteins from Mycobacterium tuberculosis H37Rv Proteomics 7, 1702–1718 351 Characterization of Rv3619c and Rv3620c 7 Marmiesse M, Brodin P, Buchrieser C, Gutierrez C, Simoes N, Vincent V, Glaser P, Cole ST & Brosch R (2004) Macro-array and bioinformatic analyses reveal mycobacterial ‘core’ genes, variation in the ESAT-6 gene family and new phylogenetic markers... 37 °C for 1 h After the washing steps, 100 lL of (1 : 5000 dilution of stock) HRP-conjugated rabbit antigoat antibody was added to each well, and the plate was incubated at 37 °C for 1 h The plate was washed again before addition of 100 lL of substrate solution (6 mg of o-phenylenediamine dihydrochloride in 12 mL of substrate buffer with 5 lL of 30% H2O2), and was finally incubated FEBS Journal 278 (2011)... Immunogenicity and protective efficacy of a tuberculosis DNA vaccine Nat Med 2, 893–898 24 Olsen AW, Pinxteren LAHV, Okkels LM, Rasmussen PB & Anderson P (2001) Protection of mice with a tuberculosis subunit vaccine based on a fusion protein of antigen 85B and ESAT-6 Infect Immun 69, 2773–2778 25 Skjøt RLV, Brock I, Arend SM, Munk ME, Theisen M, Ottenhoff THM & Andersen P (2002) Epitope mapping of the immunodominant... tuberculosis operon encoding ESAT-6 and a novel low-molecularmass culture filtrate protein (CFP-10) Microbiology 144, 3195–3203 21 Kamath AT, Feng CG, MacDonald M, Briscoe H & Britton WJ (1999) Differential protective efficacy of DNA vaccines expressing secreted proteins of Mycobacterium tuberculosis Infect Immun 67, 1702–1707 22 Brandt L, Elhay M, Rosenkrands I, Lindblad EB & Andersen P (2000) ESAT-6 subunit... samples tested IL-12 and IL-4 levels were estimated similarly Determination of total IgG and isotyping The production of antigen-specific total IgG and isotype antibodies was measured in the sera of the immunized mice bled 2 weeks postimmunization and 2 weeks postchallenge, as described previously [33] Briefly, ninety-six-well microtiter plates were incubated overnight with 100 lL of antigen (0.5 ngÆmL)1)... FEBS Characterization of Rv3619c and Rv3620c 33 Mallick AI, Singha H, Khan S, Anwar T, Ansari MA, Khalid R, Chaudhuri P & Owais M (2007) Escheriosome-mediated delivery of recombinant ribosomal L7 ⁄ L12 protein confers protection against murine brucellosis Vaccine 25, 7873–7884 Supporting information The following supplementary material is available: Fig S1 Helical wheel projection of N-terminal and. .. England P et al (2007) ESAT-6 from Mycobacterium tuberculosis dissociates from its putative chaperone CFP-10 under acidic conditions and exhibits membrane-lysing activity J Bacteriol 189, 6028–6034 19 Andersen P, Andersen AB, Sorensen AL & Nagai S (1995) Recall of long-lived immunity to Mycobacterium tuberculosis infection in mice J Immunol 154, 3359– 3372 20 Berthet FX, Rasmussen PB, Rosenkrands I, Andersen... (0.05 m, pH 9.6) at 4 °C After washing and blocking steps, test and control sera were serially diluted, and plates were then incubated at 37 °C for 1 h After several washings of the plate, total IgG was determined by using HRP-tagged rabbit anti(mouse IgG); in isotyping, plate was further incubated with 100 lL of (1 : 5000 dilution of stock) goat anti-mouse (IgG1) and goat anti-mouse (IgG2a) The plates . V32 N Rv3620c C Rv3620c N Rv3619c C Rv3619c Fig. 2. In silico modeling and docking of Rv3619c and Rv3620c. The two proteins form a 1 : 1 complex. Rv3619c. disrupted, and Rv3619c and Fig. 3. pH-induced conformational changes in Rv3619c, Rv3620c, and the 1 : 1 complex. (A) MREs of Rv3619c (j), Rv3620c (d) and the