The Mycobacterium tuberculosis membrane protein Rv2560 ) biochemical and functional studies David F. Plaza 1 , Hernando Curtidor 2,4 , Manuel A. Patarroyo 1,4 , Julie A. Chapeton-Montes 1 , Claudia Reyes 3 , Jose Barreto 3 and Manuel E. Patarroyo 1,5 1 Molecular Biology Department, Fundacion Instituto de Inmunologia de Colombia, Bogota, Colombia 2 Receptor–Ligand Department, Fundacion Instituto de Inmunologia de Colombia, Bogota, Colombia 3 Chemical Synthesis Department, Fundacion Instituto de Inmunologia de Colombia, Bogota, Colombia 4 Universidad del Rosario, Bogota, Colombia 5 Universidad Nacional de Colombia, Bogota, Colombia Tuberculosis (TB) is a major public health problem throughout the world, affecting almost nine million people [1] and causing more than three million deaths per year. An increasing incidence of TB, related to the high risk of developing the disease in immuno- suppressed individuals and the increasing proportion of Mycobacterium tuberculosis drug-resistant strains, has contributed to this problem [2,3]. This makes the Keywords high-activity binding peptide; invasion inhibition; Mycobacterium tuberculosis – host cell interaction; Rv2560 membrane protein Correspondence M. E. Patarroyo, Carrera 50 # 26-00, Bogota, Colombia Fax: +57 (1) 4815269 Tel: +57 (1) 4815219 or +57 (1) 3244672 Ext 125 E-mail: mepatarr@mail.com (Received 31 July 2007, revised 11 October 2007, accepted 17 October 2007) doi:10.1111/j.1742-4658.2007.06153.x The characterization of membrane proteins having no identified function in Mycobacterium tuberculosis is important for a better understanding of the biology of this pathogen. In this work, the biological activity of the Rv2560 protein was characterized and evaluated. Primers used in PCR and RT-PCR assays revealed that the gene encoding protein Rv2560 is present in M. tuberculosis complex strains, but transcribed in only some of them. Sera obtained from rabbits inoculated with polymer peptides from this pro- tein recognized a 33 kDa band in the M. tuberculosis lysate and a mem- brane fraction corresponding to the predicted molecular mass (33.1 kDa) of this protein. Immunoelectron microscopy analysis found this protein on the mycobacterial membrane. Sixteen peptides covering its entire length were chemically synthesized and tested for their ability to bind to A549 and U937 cells. Peptide 11024 (121VVALSDRATTAYTNTSGVSS140) showed high specific binding to both cell types (dissociation constants of 380 and 800 nm, respectively, and positive receptor–ligand interaction cooperativity), whereas peptide 11033 (284LIGIPVAALIHVYTYRKLS GG304) displayed high binding activity to A549 cells only. Cross-linking assays showed the specific binding of peptide 11024 to a 54 kDa membrane protein on U937. Invasion inhibition assays, in the presence of shared high-activity binding peptide identified for U937 and A549 cells, presented maximum inhibition percentages of 50.53% and 58.27%, respectively. Our work highlights the relevance of the Rv2560 protein in the M. tuberculosis invasion process of monocytes and epithelial cells, and represents a funda- mental step in the rational selection of new antigens to be included as components in a multiepitope, subunit-based, chemically synthesized, anti- tuberculosis vaccine. Abbreviations GRAVY, grand average of hydropathicity; HABP, high-activity binding peptide; SPf66, synthetic Plasmodium falciparum 66; SSP, simple sequence protein; TB, tuberculosis. 6352 FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS development of new therapies and ⁄ or vaccines, or the improvement of existing ones, necessary to control this disease. The identification and characterization of the pro- teins involved in the interaction of M. tuberculosis with its target cells and the mechanisms for evading an immune response produced against this pathogen are thus of crucial interest. Of special interest are those proteins found on the cell membrane and in the secreted fraction, as they represent the ‘face’ shown by the bacterium to its host, thereby leading to its identifi- cation and elimination. Although TB is mainly a pulmonary disease, it can also affect other organs. Few microorganisms can sur- vive within macrophages because of the abundance of hydrolytic enzymes contained in acid phagocytic com- partments [4]. M. tuberculosis has developed evolution- ary mechanisms that allow it to penetrate, survive and grow within these compartments [5]. Adhesion and phagocytosis are the initial steps in the entry of the bacterium into macrophages, processes triggered by receptor–ligand interactions. These interactions are complex and have an important influence on the intra- cellular fate of M. tuberculosis [6,7], being mediated mainly by membrane proteins present on the bacillus and its host [8]. The M. tuberculosis cell membrane, and that of other prokaryotes, contains proteins func- tioning as enzymes, ligands, signal transducers or viru- lence factors, all of great importance for bacterial survival [9]. M. tuberculosis invades alveolar epithelial cells, thereby inducing macropinocytosis, using secreted or membrane components [10]. The identification of the bacterial peptides involved in this process has been poorly studied, and represents an objective of this work. Membrane proteins represent a bank of antigens available for the design of a subunit-based, multiepi- tope vaccine, as they may be involved in the patho- genicity of the bacterium and may also be highly susceptible to immune recognition. The M. tuberculosis membrane proteome has revealed the presence of mul- tiple immunogenic components with unknown function [11–13], which could be used for this purpose. Bioinformatics represents a preliminary method for the evaluation of the presence of a determined gene or protein sequence in different organisms, and for pre- dicting (with some degree of confidence) the existence of membrane anchoring or signal export regions within them, making it a fundamental tool when screening and ⁄ or selecting for a vaccine candidate [14,15]. Fur- thermore, work carried out on the proteomics of the bacterium has led to the identification of a large number of membrane proteins with an as yet undeter- mined function [11–13]. A previous study has shown the importance of the melA gene (encoding an Rv2560 homologous protein) in Mycobacterium marinum invasion of THP-1 cells, indicating that its homologous gene might be relevant in M. tuberculosis–host cell interaction [16]. This work has led to the identification of the pres- ence of the Rv2560 proline- and glycine-rich trans- membrane protein encoding gene and its transcripts in the M. tuberculosis complex and clinical isolate strains, as well as the characterization of the high-activity binding peptide (HABP) involved in the binding to and invasion of monocytes (U937) and type II alveolar epithelial cells (A549), using synthetic peptides. The protein encoded by the Rv2560 gene was synthesized in 20-mer-long, nonoverlapping peptides, which were analysed by a highly specific and robust methodology (widely used in Plasmodium falciparum research) for the identification of the HABPs interacting with U937 and A549 cells (in this case), their binding dissociation constants, their critical binding residues and their role in cell invasion inhibition, with the specific purpose of employing them as probable components of a multiepi- tope, subunit-based, synthetic vaccine. This has impor- tant biological significance and implications for the development of strategies to control this disease. Results Bioinformatic analysis Different bioinformatics’ tools were used to predict the cell location of Rv2560 and its transmembrane topol- ogy. Although signalp did not predict the existence of a signal peptide, the high grand average of hydropath- icity (GRAVY) value (0.488) and the existence of four possible transmembrane helices in the complete Rv2560 amino acid sequence suggested its presence on the bacterial surface. Membrane topology prediction proposed four transmembrane helices (i103–125o, o145–167i, i197–228o and o271–293I, where i denotes inside and o denotes outside), suggesting that the iden- tified HABPs are located in the protein’s solvent- exposed region (Fig. 4, topology diagram, see later). Genomic PCR assay Two specific primers, flanking an HABP region (encoding amino acids 80–290), were designed and syn- thesized to determine the presence or absence of the gene encoding the Rv2560 protein in M. tuberculosis complex strains. A single 631 bp amplification band D. F. Plaza et al. M. tuberculosis Rv2560 protein characterization FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS 6353 was observed in the following strains: M. tuberculosis H37Rv, M. tuberculosis H37Ra, Mycobacterium bovis, M. bovis BCG, Mycobacterium africanum and Myco- bacterium microti (Fig. 1A, lanes 2–7). tblastn analy- sis showed similar (although not identical) genes in M. marinum and Mycobacterium ulcerans. DNA sequencing The amplified fragments in the different clinical iso- lates studied were purified, and their DNA was sequenced to determine this region’s polymorphism. It was observed that this region was completely con- served in all M. tuberculosis clinical isolates analysed, independent of geographical origin. This result was confirmed by sequencing each fragment’s complemen- tary chain (data not shown). RT-PCR assay Two specific primers were designed and synthesized to determine the transcriptional pattern of the Rv2560 gene in M. tuberculosis complex strains. A single 308 bp amplification band was observed in some M. tuberculosis complex strains (Fig. 1B, lanes 2–7). This gene’s transcription was exclusive to M. tuberculo- sis H37Rv, M. tuberculosis H37Ra, M. bovis BCG and M. africanum strains, but not M. bovis and M. microti, suggesting that it was not transcribed, in spite of this gene being present as assessed by PCR. Figure 1C shows a 360 bp conserved fragment from a mycobacte- rial genus constitutive gene (the rpoB gene encoding the RNA polymerase b-subunit) used as transcription control. Western blot Polymerized synthetic peptides corresponding to the Rv2560 protein peptide 11027 (181ADGKPVTIATFF RPRNLGLV200) and 11031 (242SPIDSVKASIETV GSNIGGSVY262) amino acid sequences, located in the extracellular solvent-exposed region (as predicted by the TMHMM server), were used as immunogens in rabbits, employing a 0, 20 and 40 day inoculation scheme. Western blot (Fig. 2) was performed with preimmune and post-third inoculation rabbit sera against M. tuberculosis sonicate and membrane frac- tion. Sera had previously been adsorbed on columns with Escherichia coli and Mycobacterium smegmatis lysates, as well as synthetic Plasmodium falcipa- rum 66 (SPf66), prior to western blotting to eliminate the cross-reactivity against bacterial (E. coli), myco- bacterial (M. smegmatis) and polymerization (SPf66) neo-antigens. Lanes 1 and 3 in Fig. 2 show the lack of preimmune rabbit sera protein recognition in the M. tuberculosis sonicate and membrane fraction, respectively. Rabbits immunized with 11031 polymerized peptide developed polyclonal antibodies specifically recognizing a 33 kDa molecular mass band in the M. tuberculosis sonicate, close to the protein’s molecular mass (33.1 kDa) in lane 2. Preimmune serum (lane 3) did not recognize bands in the membrane fraction; however, postimmune serum from 11027 immunized rabbits strongly recog- nized a band close to 33 kDa in this fraction (lane 4), suggesting that this protein becomes enriched in the cell membrane. Fig. 1. Genomic PCR and RT-PCR assays. (A) A 631 bp PCR prod- uct from the gene encoding M. tuberculosis H37Rv Rv2560 protein was only amplified on DNA from M. tuberculosis complex strains (lanes 2–7). Lane 1, 1 kb molecular weight marker (Gibco); lane 2, M. tuberculosis H37Rv; lane 3, M. tuberculosis H37Ra; lane 4, M. bovis; lane 5, M. bovis BCG; lane 6, M. africanum; lane 7, M. microti; lane 8, PCR negative control. (B) M. tuberculosis Rv2560 protein H37Rv 308 bp RT-PCR product obtained from M. tuberculosis complex strain cDNA. Lane 1, 1 kb molecular weight marker (Gibco); lane 2, M. tuberculosis H37Rv; lane 3, M. tuberculosis H37Ra; lane 4, M. bovis; lane 5, M. bovis BCG; lane 6, M. africanum; lane 7, M. microti; lane 8, negative control [M. tuberculosis H37Rv DNA treated with DNAse Q (Promega)]; lane 9, PCR positive control (M. tuberculosis H37Rv DNA); lane 10, PCR negative control. A single 308 bp amplification band was observed in some M. tuberculosis complex strains. (C) PCR 360 bp product from the same strains as (B) but with amplification of the Mycobacterium rpoB gene as positive control for mRNA expression and cDNA synthesis. M. tuberculosis Rv2560 protein characterization D. F. Plaza et al. 6354 FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS Immunoelectron microscopy Rabbit serum raised against peptide 11027 was used for immunolabelling. Immunoelectron microscopy showed that colloidal gold particles (5 nm) were mainly located on the surface of M. tuberculosis H37Rv (Fig. 3), supporting the concept that this expressed gene is located on the surface and is recog- nized by serum directed against one of its forming pep- tides. No labelling was observed in negative controls with preimmune serum (data not shown). High specific binding peptides A highly specific, sensitive and robust receptor–ligand binding assay was developed, based on previous stud- ies with erythrocytes and hepatocytes [17–19]. This methodology was adapted to other cell types, such as monoblastic U937 and epithelial A549 cells, as they are useful models for M. tuberculosis invasion studies [20]. The binding assay was performed at four 125 I- radiolabelled peptide concentrations (10–200 nm)in the presence or absence of at least a 400 times excess of unlabelled peptide, as described previously [21–23]. The binding activity was defined as the slope of the specific binding curve in the added peptide range. Pep- tides showing binding activity greater than or equal to 1% were considered to be U937 or A549 cell HABPs. A previously reported U937 and A549 cell HABP, 11095 (from the Rv1510c protein), was used as binding assay control [23]. Figure 4 shows the Rv2560 synthetic peptide sequences and the corresponding U937 or A549 cell binding activities. The black bars represent the binding activity. Those peptides having a binding activity ‡ 1.0 slope (dotted line) were considered to be cell line HABPs. Peptide 11024 (121VVALSDRATTAYT NTSGVSS140) was found to be the only U937 and A549 cell HABP identified for the Rv2560 protein. This peptide is predicted to be in the extracellular sol- vent-exposed region, as determined by tmhmm and tmpred online software. Peptide 11033 (284LIGI PVAALIHVYTYRKLSGG304) showed high binding activity to the epithelial A549 cell line only. Affinity constants Saturation assays and Hill analysis [24,25] were carried out for shared 11024 HABPs based on initial screening results, using a greater 125 I-radiolabelled peptide Fig. 2. Immunoblotting assay. Western blotting was performed with rabbit preimmune and post-third inoculation sera against a M. tuberculosis sonicate (lanes 1, 2 and 5) and a membrane frac- tion (lanes 3 and 4). Lanes 1 and 3, absence of recognition of M. tuberculosis proteins by rabbit preimmune sera; lane 2, serum from rabbit inoculated with polymerized peptide 11031; lane 4, serum from rabbit inoculated with polymerized peptide 11027; lane 5, serum from rabbit inoculated with M. tuberculosis total son- icate. Molecular weight markers are shown on the left-hand side. Fig. 3. Immunoelectron microscopy. Immunogold location of Rv2560 protein on M. tuberculosis H37Rv membrane, as assessed by 5 nm gold-labelled anti-rabbit IgG particles. Intact cells were incubated with a 1 : 10 dilution (1 and 2) or pure (3) rabbit sera directed against peptide 11027, and prepared for immunoelectron microscopy by negative contrast, as described. No labelling of the surface was obtained in control experiments from which these anti- bodies had been omitted, or with preimmune serum (data not shown). The arrows indicate the location of the Rv2560 protein on the mycobacterial surface. D. F. Plaza et al. M. tuberculosis Rv2560 protein characterization FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS 6355 concentration range (500–2000 nm) (Fig. 5A). Scat- chard and Hill analyses were then performed and dis- sociation constants (K d ) were calculated for Rv2560 HABP 11024 for U937 and A549 cell lines (K d of 380 and 800 nm, respectively). The Hill coefficient (n h ) was 1.5 for U937 and 1.2 for A549 cells, suggesting positive ligand–receptor cooperativity. The numbers of binding sites per cell calculated for peptide 11024 were 200 000 and 3 000 000 for U937 and A549 cells, respectively. Critical residues for cell binding in U937 cells Glycine scanning analogue peptides were synthesized, and competition cell binding assays were performed between original radiolabelled peptides and their ana- logue peptides to identify HABP critical amino acids for U937 cell binding. Critical residues were those which, on replacement with glycine, rendered an invariable decrease of at least 50% in the ability to compete with original radiolabelled peptide in a bind- ing assay at two concentrations (16 and 400 nm). This assumed dramatic changes in peptide binding activity. Figure 5B shows that critical residues in peptide 11024 binding were VVALSDRATTAYTNTSGVSS (shown in bold italic in the sequence). Cross-linking assays HABP 11024 bound to a protein located on the U937 cell membrane, specifically recognizing a 54 kDa hypo- thetical receptor. Radiolabelled peptide binding to this receptor protein was inhibited in the presence of non- radiolabelled peptide; this was clearly shown by the band intensities (Fig. 6). Peptide 11024 binding was target cell specific, as it did not show any binding to HepG2 cells or erythrocytes (data not shown). Invasion inhibition assay using HABPs A549 and U937 cell lines were preincubated in the presence of HABP 11024 in two independent experi- ments to evaluate the effect of HABP on bacterial invasion of target lung cells. It was found that invasion was inhibited in both cell types. The inhibitory power of HABP 11024 was greater in assays carried out using the A549 cell line, where it reached close to 60% inhi- bition when such cells were preincubated in the pres- ence of 1 lm peptide. Monocyte invasion assays revealed around 50% maximum inhibition only when higher peptide concentrations were used than those employed for A549 cell assays (Fig. 7). It was also Fig. 4. Cell binding activity of Rv2560 peptides. Amino acid sequence and specific U937 and A549 cell binding activity for 20-mer long chem- ically synthesized Rv2560 peptides from the M. tuberculosis H37Rv strain. The peptides are given on the left-hand side, with numbers indi- cating their position within the native protein; in peptides that did not contain tyrosine, it was added to the carboxy-terminal end. To the right, the black bar represents each peptide binding activity, determined as the specific binding ⁄ total added peptide ratio. The dotted line separates peptides having ‡ 1% binding activity. HABP 11095 was used as a binding activity control. Rv2560 transmembrane topology is presented on the left-hand side. Grey shows the transmembrane helices present in the sequence according to the TMHMM prediction server. nd, not done because of solubilization problems. M. tuberculosis Rv2560 protein characterization D. F. Plaza et al. 6356 FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS observed that the inhibitory activity of HABP 11024 decreased at concentrations equal to or greater than 100 lm. The secondary structure of HABP 11024 in solution at different concentrations (including those in which activity was seen to decrease) was verified by CD, and it was shown that this HABP underwent conformational changes critical for its function at the concentrations already established in functional assays (data not shown). Discussion Although the elucidation of the M. tuberculosis gen- ome sequence has revealed the presence of many gene families which can be classified according to their homology to genes having a known function in other microorganisms, no functional counterparts have been found for many of these mycobacterial genes. Pro- teome studies have led to the determination of some of the proteins present on the membrane for which no role in M. tuberculosis invasion of the host cell has yet been determined [12,13]. The infection of epithelial cells in the alveoli requires the interaction of the bacterium with several surface receptors, which have been poorly studied to date [26]. Studies of nonpolarized pulmonary epithelial cells (HEp-2) have revealed the existence of at least five M. tuberculosis proteins that bind to biotinylated cell fractions, including the already characterized adhesion heparin-binding haemagglutinin [27]. A B Fig. 5. Saturation curves and critical residues of cell binding. (A) Saturation binding curve for HABP 11024 binding to U937 and A549 cells. The saturation curves resulted from plotting the specifically bound 125 I-HABP concentration versus free 125 I-HABP. The affinity constants and maximum number of sites per cell were obtained from these curves. Inset: the abscissa is log F in the Hill plot and the ordinate is log [B ⁄ B m ) B], where B m is the maximum bound peptide, B is the bound peptide and F is free peptide. (B) Competition binding assay with analogous peptides. Specific original radiolabelled peptide binding inhibited by analogous peptide (at 16 and 400 n M) is shown. Amino acids underlined in bold represent the critical binding residues, as their binding activity decreased by 50% or more when modified and assayed at the two concentrations used. Fig. 6. Cross-linking assay for peptide 11024. Autoradiographs for U937 proteins specifically cross-linked with radiolabelled peptide 11024. Lanes 1 and 2, U937 cells; lanes 3 and 4, HepG2 cells; lanes 1 and 3, total radiolabelled peptide binding; lanes 2 and 4, inhibited binding. The cross-linking assays show that peptide 11024 specifically binds to a 54 kDa membrane protein on U937 cells. D. F. Plaza et al. M. tuberculosis Rv2560 protein characterization FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS 6357 Moreover, it has been observed that M. tuberculosis can promote monocyte differentiation into dendritic cells that lack CD1, have low CD80 and produce inter- leukin-10; this altered phenotype is unable to prime effector T cells, thus becoming an escape mechanism [28]. The identification of proteins specifically binding to receptors on monocytes or other M. tuberculosis infection-susceptible cells has thus become of interest in order to determine precisely the regions in such pro- teins that are directly involved in mycobacterial inva- sion or immune response modulation. Our group has characterized three mycobacterial membrane proteins (Rv2004c [21], Rv2536 [22] and Rv1510c [23]) in recent studies, and has determined their target cell binding profiles. These studies repre- sent an initial step in the rational selection of com- ponents for a subunit-based, multiepitope, anti-TB vaccine candidate. Rv2560 is a proline- and glycine-rich transmembrane protein, and may thus be classified as a simple sequence protein (SSP). This type of sequence is fre- quent in organisms having G + C-rich genomes, such as M. tuberculosis, where several SSPs are organized into large-sized families having a role in host–pathogen interaction and are membrane-associated [29]. tblastn analysis of the Rv2560 amino acid sequence revealed the existence of homologous genes in M. marinum and M. ulcerans. The homologous gene in M. marinum (melA) was shown to be important in the adhesion, invasion and intracellular survival of this bacterium in THP-1 cells after insertion into the noninvasive M. smegmatis species [16]. This work shows that the Rv2560 encoded protein, classified as a putative membrane protein [9] in the M. tuberculosis genome, is present in M. tuberculosis complex strain bacilli, is transcribed, expressed and present in greater concentration on the membrane- enriched fraction and the mycobacterial surface, and is involved in the mycobacterial invasion of host cells. On PCR amplification of a 631 bp fragment, includ- ing the sequence of nucleotides encoding a high specific binding Rv2560 protein region (amino acids 80–290) (Fig. 1), it was found that this fragment was amplified in M. tuberculosis complex strains. The presence of this gene was also determined in clinical M. tuberculosis isolates obtained from patients having different types of TB (pulmonary, pleural, osseous, meningeal, abdominal, genitourinary and renal), with the 631 bp fragment being amplified in all clinical isolates. All of the foregoing suggests that the Rv2560 gene is present in different M. tuberculosis complex strains, including clinical isolates, and has no genetic variability, as shown by DNA sequencing (data not shown). The transcription of this gene was also confirmed after the presence of the highly conserved gene had been established in different M. tuberculosis complex strains. cDNA was therefore obtained from M. tuber- culosis complex strains and amplified by PCR, using specific primers for the 308 bp fragment (using more internal primers to amplify smaller fragments in order to improve RT-PCR efficiency). The results (Fig. 1B) led to the conclusion that the Rv2560 gene was transcribed in M. tuberculosis H37Rv, M. tuberculosis H37Ra, M. bovis BCG and M. africanum strains, but not in M. bovis or M. microti. The two polymerized peptides used in rabbit immu- nization produced an antibody response (differing in intensity) evident in western blot results (Fig. 2); a Fig. 7. Invasion inhibition assays with HABP 11024. Inhibition percentages with different HABP 11024 concentrations using U937 and A549 cells. A peptide having low specific binding ability to A549 cells was used as negative control for this cell line (11021). Colchicine (a vesicular traffic inhibitor) was used as positive invasion inhibition control. The results correspond to the average inhibition percentage calculated for each treatment ± standard deviation. *P £ 0.05, **P £ 0.01, ***P £ 0.001. M. tuberculosis Rv2560 protein characterization D. F. Plaza et al. 6358 FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS 33 kDa band corresponding to the theoretical molecu- lar mass of the Rv2560 protein (33.1 kDa) was recog- nized by these sera, suggesting that this protein is transcribed and expressed in M. tuberculosis sonicate, and at much higher concentration in the membrane fraction, as inferred from the band intensity. When using serum from rabbits immunized with poly- merized peptide 11027, immunoelectron microscopy confirmed the presence of this protein on the mycobac- terial surface (which has not been reported to date). Receptor–ligand assays led to the identification of two high specific binding peptides, one (11024) being shared by both U937 monoblast and A549 epithelial cells. Peptide 11024 binding to both cell types was sat- urable, presenting strong positive cooperativity, as shown by the dissociation constant and Hill coefficient. The corresponding analyses for the determination of the receptor for 11024 HABPs and the identification of the critical binding residues, and the invasion inhibi- tion assays, were thus performed exclusively with this peptide. This peptide is located in the solvent-exposed central region of the protein between residues 121 and 140, displaying the VVALSDRATTAYTNTSGVSS sequence (where the amino acids shown in bold italic correspond to those identified as being critical in 11024 binding to U937 cells). These residues may be involved directly in target cell binding or may be important for the structural characteristics of the peptide, thereby allowing specific binding. BLAST analysis of this small region led to the identification of 78% homology with Burkholderia cenocepacia-dependent AMP synthase, which causes respiratory tract infection and is cyto- toxic for alveolar epithelial cells [30]. A 54 kDa protein was found to be a receptor on U937 cells; its binding by peptide 11024 was inhibited by an excess of nonradiolabelled peptide. This receptor was specific for the U937 cell line, but was not seen in hepatocytes or erythrocytes. A receptor having a simi- lar molecular mass (51 kDa) has been described for the Rv1510c protein HABP 11095; however, further studies are required to determine whether the same membrane protein is involved. Peptide 11024 (for which high monocyte and alveo- lar epithelial cell binding has already been determined) induced a 36.96–50.53% decrease in M. tuberculosis invasion of U937 cells and a 30.98–58.27% decrease in M. tuberculosis invasion of A549 cells (Fig. 7). This indicates that this HABP is important in the recogni- tion and invasion of monocytes and type II alveolar epithelial cells by the mycobacterium. It was observed in preliminary assays that HABP concentrations of less than 100 lm had a greater inhibitory effect than larger concentrations. Therefore, additional CD secondary structural studies were per- formed to verify possible conformational changes at concentrations greater than 100 lm which were critical in the inhibitory activity of this peptide. It was found that the structure of the peptide changed at concentra- tions equal to or greater than 100 lm (not shown), thereby explaining the loss of inhibitory potential in the assay carried out for alveolar epithelial cells. The above results indicate that the Rv2560 protein is conserved within the M. tuberculosis complex. It is also transcribed and expressed in low abundance on the mycobacterial surface. The protein contains a spe- cific high binding peptide for U937 and A549 cells, which plays an important role in mycobacterial inva- sion. Therefore, it is concluded that the Rv2560 pro- tein is highly relevant in pathogen–host interaction, and that the identification of the region directly involved in such interaction makes it an important candidate for inclusion in the development of a subunit-based, multiepitope, chemically synthesized vaccine. Experimental procedures Bioinformatics’ analysis M. tuberculosis Rv2560 protein sequences were obtained from the Tuberculist Webserver (http://genolist.pasteur.fr/ TubercuList/). Proteins were aligned by BLAST (http:// www.ncbi.nlm.nih.gov/BLAST/) and clustalw (http:// npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=npsa_clu stalw.html) multiple alignments. Protein features and GRAVY scores for the Rv2560 amino acid sequence were calculated using the protparam tool (http://us.expasy. org ⁄ tools⁄ protparam.html), for which a score of greater than ) 0.4 (mean score for the cytosolic proteins) suggests membrane association probability; the higher the score, the greater the probability [13]. Transmembrane regions were predicted by tmhmm (http://www.cbs.dtu.dk/services/ TMHMM) and tmpred (http://www.ch.embnet.org/ software/TMPRED_form.html). Cell localization, lipid attachment sites and signal sequences were predicted using psort (http://psort.nibb.ac.jp). Mycobacterial species and strains The following M. tuberculosis complex species and strains were used: M. tuberculosis H37Rv (ATCC27294), M. tuber- culosis H37Ra (ATCC25177), M. bovis (ATCC19210), M. bovis BCG (ATCC27291, Pasteur sub strain), M. africa- num (ATCC25420) and M. microti (kindly donated by F. Portaels, Prince Leopold Institute of Tropical Medicine). D. F. Plaza et al. M. tuberculosis Rv2560 protein characterization FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS 6359 Ten M. tuberculosis clinical strains were isolated from speci- mens obtained from different patients attending the TB pro- gramme at either the San Juan de Dios Hospital or Santa Clara Hospital, both in Bogota, Colombia. Patients came from several geographical areas of Colombia. All mycobacte- rial strains were grown for 5–15 days in 7H9 Middlebrook broth (Difco Laboratory, Detroit, MI, USA) with 0.04% Tween 80 and in Middlebrook agar 7H10 (Difco), both sup- plemented with oleic acid, albumin, dextrose and catalase (BBL, Becton Dickinson, Mountain View, CA, USA). Genomic DNA extraction Cells were harvested in Tris–EDTA buffer (TE, pH 8.0) and suspended in 25% sucrose in 10 · TE. Lysozyme (100 lgÆmL )1 ) was added and the preparation was incubated at 37 °C for 2 h with subsequent addition of 100 lgÆmL )1 proteinase K and N-lauroylsarcosine (Sigma, St Louis, MO, USA) at 1% final concentration. Following 18 h incubation at 65 °C, the mixture was extracted with chloroform–iso- amyl alcohol (24 : 1, v ⁄ v), once with phenol–chloroform (1 : 1, v ⁄ v) and twice more with chloroform–isoamyl alco- hol. DNA was precipitated with 2-propanol, washed with 70% (v ⁄ v) ethanol and suspended in 1 · TE [31,32]. PCR assay PCR amplifications were performed in a thermal cycler GeneAmp PCR System 9600 (Perkin-Elmer Life Sciences Inc., Boston, MA, USA), using 100 ng mycobacterial geno- mic DNA for amplifications. The mixture contained 50 mm KCl, 10 mm Tris ⁄ HCl (pH 8.3), 1.5 mm MgCl 2 , 0.1 mm of each dNTP, 0.4 mm of both direct 5¢-CAATCGTCG GGTTTTAGCG-3¢ and reverse 5¢-GTGTAGACGTG GATGAGCG-3¢ oligonucleotide primers and 1.5 U Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA). Twenty-five cycles of the following thermal profile were car- ried out after DNA had been denatured for 5 min at 94 °C: 56 °C for 30 s, 72 °C for 40 s and 95 °C for 40 s. A final 5 min extension cycle was then performed at 72 °C. The amplification products were loaded onto 1% agarose gel, electrophoresed at 70 V in 1 · Tris/acetate/EDTA, stained with ethidium bromide and photographed. DNA sequencing The dideoxy chain termination method was used for sequencing reactions with a Taq FS DyeDeoxy Terminator cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) on a GeneAmp PCR System 9600 (Perkin-Elmer Life Sciences Inc.), and run on a 373 DNA sequencer model (Applied Biosystems). The sequencing strategy involved for- ward and reverse specific primers for the 631 bp fragment of the Rv2560 gene. RNA isolation The bacilli were harvested by spinning at 12 000 g for 15 min at 4 °C. Sodium azide (10 mm) was added to the culture just before harvesting. The cell pellet was suspended in 2 mL of cold lysis buffer for each 200 mg wet weight of cells [33], and sonicated twice for 15 min. Two volumes of Trizol (Gibco BRL, Gaithersburg, MD, USA) were then added and extracted according to the manufacturer’s instructions. The pellet was suspended in 100 lL of distilled water and stored in aliquots at ) 80 °C. RT-PCR Total RNA was quantified by a GeneQuant spectropho- tometer (Pharmacia Biotech, Piscataway, NJ, USA), treated with RNAse-free DNAse RQ1 at 37 °C for 3 h, precipi- tated with isopropanol, washed with 70% ethanol and sus- pended in distilled water. M. tuberculosis H37Rv DNA was included as DNAse Q activity control (1 UÆlg )1 DNA). Target RNA (500 lg) was reverse transcribed in a tube containing distilled water and 10 lgÆmL )1 random primers (Gibco BRL). This mixture was incubated for 10 min at 70 °C; 1 · RT buffer, 10 mm dithiothreitol, 0.5 mm dNTPs and 40 U human placenta ribonuclease inhibitor (Promega, Madison, WI, USA) were then added on ice; 200 U M-MLV reverse transcriptase (Gibco-Life Technologies, Grand Island, NY, USA) was then added at a final volume of 30 lL. This mixture was kept at 37 °C for 1 h. The enzyme was finally denatured for 5 min at 95 °C. PCR was carried out as described above. The rpoB gene was used as positive transcription control. This gene, encoding the RNA polymerase b-subunit, is present in all mycobacterial species [33]. DNAse-Q-treated M. tuberculosis H37Rv was used as cDNA synthesis nega- tive control. Distilled water and M. tuberculosis H37Rv DNA were used as negative and positive PCR controls, respectively. Rabbit immunization Two New Zealand strain rabbits per peptide (previously determined to be nonreactive to M. tuberculosis sonicate, as assessed by western blot) were injected with 500 lgof peptide 11027 (181ADGKPVTIATFFRPRNLGLV200) or 500 lg of peptide 11031 (242SPIDSVKASIETVGSNIG GSVY262) [polymerized via the addition of cysteine at the amino and carboxy termini and mixed with Freund’s Incomplete Adjuvant (Sigma)] on days 0, 20 and 40. Final bleeding was carried out on day 60 and sera were collected in accordance with the Fundacion Instituto de Inmunolo- gia’s Bioethics Committee recommendations. The Rv2560 protein peptide sequences chosen for the immunization of rabbits were obtained using T-epitope prediction software M. tuberculosis Rv2560 protein characterization D. F. Plaza et al. 6360 FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS syfpeithi (15-mer T-epitope for major histocompatibility complex class II only) downloaded from http://www. syfpeithi.de/Scripts/MHCServer.dLl/EpitopePrediction.htm. SDS-PAGE and immunoblotting M. tuberculosis sonicate or membrane fraction (1 mg) was loaded in a discontinuous SDS-PAGE system, using a 10– 20% (w ⁄ v) acrylamide gradient, and then transferred to nitrocellulose paper using the semidry blotting technique [34]. The filters were incubated with a 1 : 100 (v ⁄ v) dilution of the sera obtained from rabbits immunized with polymer- ized 11027 or 11031 peptides. Sera were diluted in NaCl ⁄ Tris ⁄ Tween 20 (0.02 m Tris-HCl, pH 7.5, 0.05 m NaCl, 1% Tween 20) and 5% skimmed milk. Incubation for 1 h with 1 : 3000 (v ⁄ v) alkaline phosphatase-conjugated anti-rabbit IgG (ICN Biomedicals, Costa Mesa, CA, USA) followed five NaCl ⁄ Tris ⁄ Tween 20 washes. The reaction was devel- oped with Nitro Blue tetrazolium ⁄ 5-bromo-4-chloroindol- 2-yl phosphate (KPL, Gaithersburg, MD, USA). Immunoelectron microscopy Immunoelectron microscopy studies were carried out on a Philips CM 10 transmission electron microscope (Philips, Suresne, Hauts-de-Seine, France). Briefly, a wet pellet (50 lL) of M. tuberculosis H37Rv was fixed with a 4% paraformaldehyde)0.5% glutaraldehyde solution for 2 h at 4 °C. After fixation, the pellet was dehydrated in graded ethanol and then embedded in LR-white hard-grade acrylic resin (Sigma) for 4 days at 4 °C. Thin slices (400 nm) were cut and mounted on 300 mesh nickel grids. The slices were incubated in a saturated sodium metaperiodate solution for 1 h at 20–22 °C for antigen retrieval [35]. The grids were then floated, section down, in a beaker containing 0.01 m sodium citrate buffer for 15 min at 80 °C. After 1 h of blocking in Tris-buffered saline (NaCl ⁄ Tris) (0.05 m Tris in isotonic saline, pH 7.6) containing 0.05% BSA, the grids were incubated in either 1 : 10 (v ⁄ v) or pure rabbit polyclonal serum for 1 h at 37 °C. Following an NaCl ⁄ Tris ⁄ 0.025% Tween 20 wash, the grids were immersed in a 1 : 50 (v ⁄ v) dilution of 5 nm gold-labelled anti-rabbit IgG (Sigma) for 1 h at 20–22 °C. The grids were then washed with NaCl ⁄ Tris ⁄ Tween 20 and fixed in 2.5% glutaralde- hyde, followed by 15 min of incubation in 1% uranyl ace- tate. The grids were washed with distilled water and dried at 20–22 °C before observation. Cell culture A549 human lung epithelial (ATCC CCL-185) and U937 monocyte-like (ATCC CRL-1593.2) cells were grown at 37 °C and 5% CO 2 in RPMI 1640 medium (Gibco-BRL) supplemented with 10% fetal bovine serum (Hyclone, Logon, UT, USA). A549 cells were dislodged with 0.1% EDTA–NaCl ⁄ P i . Cell lines were collected in 50 mL tubes, centrifuged at 1000 g for 5 min and washed with RPMI 1640. Peptide synthesis Sixteen sequential 20-mer peptides, corresponding to the Rv2560 amino acid sequence [9], were synthesized in this study by a solid-phase multiple peptide system [36,37]. 4-Methylbenzhydrylamine resin (0.7 meqÆg )1 ), t-Boc amino acids and low–high cleavage techniques were used [38]. Pep- tide identity and purity were analysed by MALDI-TOF MS and analytical reverse phase HPLC. An extra tyrosine resi- due was added to any peptide C-terminus that did not contain it to enable radiolabelling. Peptide radiolabelling Radiolabelling with 125 I was performed according to previ- ously described techniques [24,25,39,40], in which chlor- amine-T (2.25 mgÆmL )1 ) and 3.2 lLNa 125 I (100 mCiÆmL )1 ) were added to 5 lL peptide solution (1 lgÆlL )1 ); 15 lL sodium bisulfite (2.75 mgÆmL )1 ) and 50 lL NaI (0.16 m) were added after 5 min of reaction at 18 °C. The radiola- belled peptide was then separated from reaction by-prod- ucts on a Sephadex G-10 column (Pharmacia Biotech, St Albans, UK) (80 · 5.0 mm). Binding assay A549 and U937 cells (1 · 10 6 ) were incubated with increas- ing quantities of each Rv2560 m. tuberculosis putative mem- brane protein 125 I-radiolabelled peptide (100–2000 nm)at 100 lL total volume for 90 min at 4 °C, in the presence or absence of 40 lm unlabelled peptide, to determine the bind- ing specificity. After incubation, unbound peptide was removed from the cells by sedimentation through a dioctyl- phthalate–dibutylphthalate cushion (d ¼ 1.015 gÆmL )1 ) and centrifuged at 9000 g for 2 min [18,19]. The assay was car- ried out in triplicate in identical conditions; bound and free peptides were determined by measuring the cell-associated radioactivity on a gamma counter (Gamma Counter Cobra II, Packard Instrument Co., Meriden, CT, USA); the average results of the assays performed in triplicate are reported and shown graphically in Fig. 4. Saturation assay A549 and U937 cells (1 · 10 6 ) were incubated with increas- ing concentrations (100–2000 nm) of radiolabelled peptide at 120 lL total volume for 90 min at 4 °C, in the presence or absence of 40 lm unlabelled peptide, to determine the binding specificity. After incubation, unbound peptide was D. F. Plaza et al. M. tuberculosis Rv2560 protein characterization FEBS Journal 274 (2007) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS 6361 [...]... incubation, the cells were washed with NaCl ⁄ Pi and cross-linked with 25 lm bis(sulfosuccinymidyl suberate) (Pierce, Rockford, IL, USA) for 20 min at 4 °C The reaction was stopped with 40 nm Tris ⁄ HCl (pH 7. 4), and washed again with NaCl ⁄ Pi The cells were then treated with lysis buffer (5% SDS, 10 nm iodoacetamide, 1% Triton X-100, 100 mm EDTA, 10 mm phenylmethanesulfonyl fluoride) The membrane proteins... U937 cells and 10 0)1 05 nm or 200 lm HABP 11024 for A549 cells before the addition of 1 · 107 M tuberculosis H37Rv bacilli (multiplicity of infection, 1 : 1 0) at a final volume of 200 lL They were incubated at 37 °C in 5% CO2 for 2 h with constant shaking, and then placed in 20 lgÆmL)1 amikacin-supplemented medium (ICN Biomedicals) for 30 min to kill off the extracellular bacteria, whilst the intracellular... (199 8) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence Nature 393, 537–544 10 Garcia-Perez BE, Mondragon-Flores R & Luna-Herrera J (200 3) Internalization of Mycobacterium tuberculosis by macropinocytosis in non-phagocytic cells Microb Pathog 35, 49–55 M tuberculosis Rv2560 protein characterization 11 Chakravarti DN, Fiske MJ, Fletcher LD & Zagursky RJ (200 0) Application... macrophages and A549 epithelial cells Protein Sci 14, 2767–2780 22 Garcia J, Puentes A, Rodriguez L, Ocampo M, Curtidor H, Vera R, Lopez R, Valbuena J, Cortes J, Vanegas M, et al (200 5) Mycobacterium tuberculosis Rv2536 protein implicated in specific binding to human cell lines Protein Sci 14, 2236–2245 FEBS Journal 274 (200 7) 6352–6364 ª 2007 The Authors Journal compilation ª 2007 FEBS 6363 M tuberculosis Rv2560. .. (200 5) Immunogenic membrane- associated proteins of Mycobacterium tuberculosis revealed by proteomics Microbiology 151, 2411– 2419 14 Cardoso FC, Pinho JM, Azevedo V & Oliveira SC (200 6) Identification of a new Schistosoma mansoni membrane- bound protein through bioinformatic analysis Genet Mol Res 5, 609–618 15 Kyte J & Doolittle RF (198 2) A simple method for displaying the hydropathic character of a protein. .. Weiland GA & Molinoff PB (198 1) Quantitative analysis of drug–receptor interactions I Determination of kinetic and equilibrium properties Life Sci 29, 313–330 26 Bermudez LE & Goodman J (199 6) Mycobacterium tuberculosis invades and replicates within type II alveolar cells Infect Immun 64, 1400–1406 27 Reddy VM & Hayworth DA (200 2) Interaction of Mycobacterium tuberculosis with human respiratory epithelial... Global Tuberculosis Control: Surveillance, Planning, Financing WHO, Geneva 2 Russell DG (200 7) Who puts the tubercle in tuberculosis? Nat Rev Microbiol 5, 39–47 3 Snider DE Jr & Castro KG (199 8) The global threat of drug-resistant tuberculosis N Engl J Med 338, 1689– 1690 4 Pillay CS, Elliott E & Dennison C (200 2) Endolysosomal proteolysis and its regulation Biochem J 363, 417– 429 5 Pieters J (200 1) Entry... J, Li YJ & Bermudez LE (200 3) Mycobacterium tuberculosis infection causes different levels of apoptosis and necrosis in human macrophages and alveolar epithelial cells Cell Microbiol 5, 649–660 21 Forero M, Puentes A, Cortes J, Castillo F, Vera R, Rodriguez LE, Valbuena J, Ocampo M, Curtidor H, Rosas J, et al (200 5) Identifying putative Mycobacterium tuberculosis Rv2004c protein sequences that bind... interaction at the level of individual amino acids Proc Natl Acad Sci USA 82, 5131–5135 37 Merrifield RB (196 9) Solid-phase peptide synthesis Adv Enzymol Relat Areas Mol Biol 32, 221–296 38 Tam JP, Heath WF & Merrifield RB (198 3) SN 1 and SN 2 mechanisms for the deprotection of synthetic peptides by hydrogen fluoride Studies to minimize the tyrosine alkylation side reaction Int J Pept Protein Res 21,...M tuberculosis Rv2560 protein characterization D F Plaza et al removed from the cells by sedimentation through a dioctylphthalate–dibutylphthalate cushion (d ¼ 1.015 gÆmL) 1) and centrifuged at 9000 g for 2 min As before, each assay was performed in triplicate; bound and free peptides were determined by measuring the cell-associated radioactivity on a gamma . The Mycobacterium tuberculosis membrane protein Rv2560 ) biochemical and functional studies David F. Plaza 1 , Hernando Curtidor 2,4 ,. mediated mainly by membrane proteins present on the bacillus and its host [8]. The M. tuberculosis cell membrane, and that of other prokaryotes, contains proteins