* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Protein expression by a Beijing strain differs from that of another
Survey
Document related concepts
Signal transduction wikipedia , lookup
G protein–coupled receptor wikipedia , lookup
Protein (nutrient) wikipedia , lookup
List of types of proteins wikipedia , lookup
Protein phosphorylation wikipedia , lookup
Magnesium transporter wikipedia , lookup
Protein structure prediction wikipedia , lookup
Intrinsically disordered proteins wikipedia , lookup
Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup
Protein moonlighting wikipedia , lookup
Protein–protein interaction wikipedia , lookup
Transcript
Microbiology (2005), 151, 1139–1150 DOI 10.1099/mic.0.27518-0 Protein expression by a Beijing strain differs from that of another clinical isolate and Mycobacterium tuberculosis H37Rv Carmen Pheiffer,1 Joanna C. Betts,2 Helen R. Flynn,2 Pauline T. Lukey2 and Paul van Helden1 1 MRC Centre for Molecular and Cellular Biology, Department of Medical Biochemistry, University of Stellenbosch Medical School, PO Box 19063, Tygerberg, 7505, South Africa Correspondence Paul van Helden 2 [email protected] GlaxoSmithKline Research and Development, Stevenage, Hertfordshire SG1 2NY, UK Received 27 July 2004 Revised 10 December 2004 Accepted 23 December 2004 The Beijing strain family has often been associated with tuberculosis (TB) outbreaks and drug resistance worldwide. In this study the authors have compared the protein expression and antigen recognition profiles of a local Beijing strain with a less prevalent clinical isolate belonging to the family 23 strain lineage, and the laboratory strain Mycobacterium tuberculosis H37Rv. Using two-dimensional electrophoresis, liquid chromatography tandem mass spectrometry and Western blot analysis several proteins were identified as quantitatively increased or decreased in both clinical strains compared to H37Rv. Remarkably, the Beijing strain showed increased expression of a-crystallin and decreased expression of Hsp65, PstS1, and the 47 kDa protein compared to the other clinical strain and H37Rv. One- and two-dimensional Western blot analysis of antigens expressed by the three strains, using plasma from TB patients, confirmed differential antigen expression by strains and patient-to-patient variation in humoral immunity. These observed protein differences could aid the elucidation of mechanisms underlying the success of the Beijing strain family, measured by global dissemination, compared to other M. tuberculosis strains. INTRODUCTION The Beijing strain family is a family of Mycobacterium tuberculosis strains, speculated to have originated in Asia (van Soolingen et al., 1995), of which the strain W and strain W-like families responsible for many cases of drug resistance are a subset (Bifani et al., 2002). Molecular epidemiological studies have shown that these strains (both drug-susceptible and -resistant) are distributed worldwide and are able to spread in large clonal clusters (Glynn et al., 2002). The worldwide occurrence of Beijing strains and their frequent association with outbreaks and drug resistance makes study of this strain family important in the understanding of M. tuberculosis pathogenesis. It is speculated that strains belonging to the Beijing family have a genetic advantage to cause disease and that the wide dispersion of this family compared to other less prevalent clinical isolates may be related to differential protein expression (Bifani et al., 2002). Proteomics, the study of the protein complement of the genome, has been used extensively to identify differences Abbreviations: 1D, one-dimensional; 2D, two-dimensional; BCG, Mycobacterium bovis bacille Calmette–Guérin; BCIP, 5-bromo-4chloro-indolyl-phosphatase; CF, culture filtrate; F11, family 11; F23, family 23; ICD, isocitrate dehydrogenase; IPG, immobilized pH gradient; LC/MS/MS, liquid chromatography tandem mass spectrometry; NBT, nitro blue tetrazolium; TB, tuberculosis; WCL, whole-cell lysate. 0002-7518 G 2005 SGM with respect to virulence and pathogenesis between mycobacterial strains. Comparative proteome analysis has revealed numerous differences in the cellular protein composition of the laboratory strains M. tuberculosis H37Rv and Erdman to that of the vaccine strain Mycobacterium bovis bacille Calmette–Guérin (BCG) (Jungblut et al., 1999; Mattow et al., 2001). In addition, a number of differences have also been observed in the expression of proteins in culture supernatants of H37Rv and BCG (Mattow et al., 2003), and H37Rv and the attenuated strain H37Ra (He et al., 2003). Analysis of the proteome of M. tuberculosis H37Rv and a clinical strain CDC1551, believed to elicit a more vigorous host immune response than H37Rv (Manca et al., 1999), identified several quantitative differences in the cellular protein composition of these strains (Betts et al., 2000). In a previous study we compared protein expression of two M. tuberculosis clinical strains, originally isolated from patients from a community in Cape Town, South Africa, with a very high tuberculosis (TB) incidence (Warren et al., 2000), to the laboratory strain H37Rv using one-dimensional (1D) PAGE, ELISA and Western blotting (Pheiffer et al., 2002). Based on the number of IS6110 insertions and spoligotyping, the two clinical strains were classified as belonging to the Beijing family and family 23 (F23) (Pheiffer et al., 2002). Results from that study (Pheiffer et al., 2002) showed that protein expression by M. tuberculosis strains Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 Printed in Great Britain 1139 C. Pheiffer and others was mainly growth phase dependent, although some differences between the strains were observed. However, due to the low resolving power of 1D gel electrophoresis, the identity of the differentially expressed proteins could not be ascertained. Here, we have extended our previous study through the use of two-dimensional (2D) PAGE coupled with identification of protein spot differences by MS and Western blot analysis of the expression of 14 M. tuberculosis antigens. Furthermore we have used plasma from patients infected with Beijing, F23 and family 11 (F11) strains to investigate differences in antigen recognition. F11 strains constitute another highly prevalent strain family in the study community. This study has revealed a number of differences in cellular and culture supernatant protein composition between M. tuberculosis H37Rv, the Beijing strain and F23, particularly between the two clinical strains compared to H37Rv. Careful analysis of these differences and the differences in the antibody recognition profiles could possibly explain the different frequencies of the clinical strains in our study population, where the Beijing strain is at least tenfold more frequent than F23 strains (Warren et al., 2000). In addition to providing clues as to the differences in pathogenicity and prevalence of strains, identification of protein expression differences between strains will aid the development of vaccines, serodiagnostic tests and the choice of drug targets. This is believed to be the first study to profile expression patterns of cellular and culture supernatant proteins of a Beijing strain by 2D gel electrophoresis and the first study attempting to correlate strain prevalence with protein expression. METHODS M. tuberculosis strains and culture. M. tuberculosis H37Rv (ATCC 25618) and two local clinical strains, Beijing (SAWC 1524) and F23 (SAWC 1296), were grown in modified Sauton medium (Andersen et al., 1991). Briefly, mycobacterial strains at late exponential growth (OD600>0?6) were diluted 1 : 10 into 200 ml Sauton medium and grown at 37 uC without stirring. After 4 weeks, cultures were harvested by centrifugation (1900 g, 20 min) and mycobacterial pellets were washed twice with PBS containing 1 % (v/v) Tween-20. Proteins used in Western blots with plasma from TB patients were prepared from strains during exponential growth in Middlebrook 7H9 medium (Becton Dickinson) supplemented with 10 % (v/v) albumin-dextrose-catalase and 0?05 % (v/v) Tween-80 as previously described (Pheiffer et al., 2002). Morphology and acid-fastness were checked by Ziehl–Nielsen staining (Heifets & Good, 1994). All strains were sensitive to isoniazid, rifampicin and ethambutol. Isocitrate dehydrogenase (ICD) assay. The amount of ICD in culture supernatants was measured using the Sigma Diagnostics ICD kit. In this procedure, ICD catalyses the oxidative decarboxylation of L-isocitrate to 2-oxoglutarate, and the reduced NADPH produced is measured at 340 nm. Protein extraction. Whole-cell lysate (WCL) proteins were extracted by bead disruption of the mycobacterial pellet in lysis buffer [0?3 % (w/v) SDS, 200 mM DTT, 50 mM Tris/HCl pH 7?0, 1 mM PMSF and complete protease inhibitor cocktail] (Roche Molecular Biochemicals) as previously described (Pheiffer et al., 1140 2002). Culture filtrate (CF) proteins were prepared from culture supernatants sterilized by sequential filtration of the culture supernatant through 0?45 mm and 0?22 mm filter units (Corning). Filtrates were then concentrated using Centricon Plus-80 filter units (Biomax-PB membrane, 5000 MWCO), followed by Centricon units (YM membrane, 3000 MWCO) (Millipore). WCL and CF protein concentrations were estimated using the Bradford assay (Bradford, 1976). 2D gel electrophoresis. Proteins were separated by 2D gel electrophoresis as previously described (Betts et al., 2000), with minor modifications. Briefly, 20 mg (60 mg for Western blotting) protein was resuspended in rehydration buffer [8 M urea, 2 % (w/v) CHAPS, 10 mM DTT, 2 % (v/v) immobilized pH gradient (IPG) buffer (pH 4–7), trace bromophenol blue] and applied to pH 4–7 IPG strips (Amersham Biosciences) for overnight rehydration. IEF was performed using a Multiphor II system (Amersham Biosciences) as follows: 100 V for 2 h, 300 V for 2 h, 1000 V for 1 h, 3500 V for 20 h (3500 V for 27?5 h for Western blotting). Second-dimension separation was carried out by placing equilibrated IPG strips (Bjellqvist et al., 1993, http://www.expasy.ch/ch2d/protocols/) onto 12 % PAGE gels and sealing with 0?5 % (w/v) agarose in cathode buffer (defined below), containing a trace amount of bromophenol blue. Anode (375 mM Tris/HCl, pH 8?8) and cathode (192 M glycine, 0?1 %, w/v, SDS, pH 8?3) buffers as described by Herbert et al. (1998) were used. Proteins were visualized by silver staining using an ammoniacal stain (Bjellqvist et al., 1993, http://www. expasy.ch/ch2d/protocols/). Gels were air-dried (Bio-Rad GelAir Dryer) and compared by visualization on a light box. When protein spot differences were noted, gels were rerun, stained with a silver stain compatible with MS (Betts & Smith, 2001) and spots of interest excised from the gel. Sample preparation for MS. Excised gel pieces were washed with double-distilled H2O for 10 min, followed by 100 % acetonitrile for 5 min, centrifuged and then dehydrated by vacuum centrifugation. Automated robotic digestion was carried out using a MassPREP station (Micromass), equipped with four probes for aspirating and dispensing reagent and washing solutions, and a heated incubation platform. Samples were destained with 50 mM ammonium bicarbonate/ acetonitrile (1 : 1, v/v), then reduced and alkylated with 10 mM DTT and 55 mM iodoacetamide respectively. This was followed by in-gel digestion with porcine trypsin (Promega) (6 ng ml21) in 50 mM ammonium bicarbonate (25 ml) for 5 h at 37 uC. The resulting peptides were then extracted with 1 % (v/v) formic acid/acetonitrile (98 : 2, v/v). Nanoscale liquid chromatography tandem mass spectrometry (LC/MS/MS) and database searching. Samples were introduced using the Micromass CapLC system (Micromass), comprising a low flow capillary HPLC pump and autosampler. A 10 port valve was configured with a pre-concentration column (300 mm ID65 mm C18 PepMap, LC Packings) and a nanoscale analytical column (75 mm ID615 cm C18 PepMap, LC Packings). Peptides were eluted using a reverse-phase gradient of 5–50 % buffer B over 30 min [A=5 % (v/v) acetonitrile, 0?1 % (v/v) formic acid; B=95 % (v/v) acetonitrile, 0?1 % (v/v) formic acid] at a flow rate of approximately 200 nl min21. All data were acquired using a Q-Tof Ultima API (Micromass) hybrid quadrupole orthogonal acceleration time of flight mass spectrometer equipped with a nanospray source. Up to eight precursor ions were automatically selected from the time of flight (TOF)/MS survey scan for MS/MS per cycle, and collision energies were chosen automatically based on the m/z value and the charge state of the selected precursor ions. Peptide sequence data generated were searched against the non-redundant protein database using MASCOT (http://www.matrixscience.com). Search parameters included Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 Microbiology 151 Mycobacterium tuberculosis protein expression Table 1. Mouse mAbs mAb IT-4 IT-7 IT-13 IT-19 IT-23 IT-38 IT-42 IT-43 IT-49 IT-51 IT-52 IT-53 IT-58 HYB76-8 Corresponding antigen* Dilution of mAbD 16 kDa (a-crystallin) 40 kDa (L-alanine dehydrogenase) 65 kDa (Hsp65) 19 kDa 38 kDa (PstS1) 20 kDa 82 kDa (KatG) 56 kDa 32–33 kDa (Ag85 complex) 17 kDa 25 kDa (MPT51) 96 kDa 47 kDa ESAT6 1 : 1000 1 : 2500 1 : 500 1 : 100 1 : 100 1 : 100 1 : 100 1 : 100 1 : 100 1 : 100 1 : 100 1 : 100 1 : 100 1 : 100 Table 1) and plasma (1 : 100 and 1 : 500) and screened for antibody binding with a horseradish-peroxidase-conjugated goat anti-mouse IgG (1 : 10 000, CALTAG laboratories) as the secondary antibody for mAbs and an alkaline-phosphatase-conjugated goat anti-human IgG (1 : 2000, Kirkegaard & Perry Laboratories) for plasma. Bound antigens were detected using chemiluminescence detection reagents (ECL Western blotting detection reagents, Amersham Biosciences) or with the 5-bromo-4-chloro-indolyl-phosphatase (BCIP)/nitro blue tetrazolium (NBT) substrate (Kirkegaard & Perry Laboratories) for plasma samples. 2D Western blotting. For Western blotting, 2D gels were transferred to nitrocellulose membranes (S&S Protran BA85, Schleicher & Schuell) using semi-dry blotting. Blots obtained from 2D gels were probed with mAbs IT-4 and IT-23 using the same conditions as for 1D Western blotting. For blots using plasma from TB patients, plasma was diluted 1 : 150 and screened for antibody binding using alkaline-phosphatase-conjugated goat anti-human IgG (Kirkegaard & Perry Laboratories) as described above. Immunoreactive antigens were detected using the BCIP/NBT substrate (Kirkegaard & Perry Laboratories) as described above. Patients. Patients were recruited from a high TB incidence com- *Antigen against which the monoclonal antibody was raised. DDilution of mAb used in Western blots. the fixed modification carboamidomethyl, due to the alkylation of cysteine residues by iodoacetamide and the variable modification, oxidation of methionine residues. Monoclonal antibodies (mAbs). Mouse mAbs against a range of M. tuberculosis proteins (Table 1) (Engers et al., 1986, KhanolkarYoung et al., 1992; Sonnenberg & Belisle, 1997) were made available by the Department of Microbiology, Colorado State University, through funds from the National Institutes of Health, National Institute of Allergy and Infectious Disease, Contract NO1-AI-75320. HYB76-8 directed to ESAT6 (Sorensen et al., 1995) was received from Karin Weldingh (Statens Serum Institut, Copenhagen, Denmark). 1D SDS-PAGE and Western blotting. Proteins were separated by SDS-PAGE using a 4 % stack over a 12 % resolving gel and transferred to PVDF membranes (Amersham Biosciences) by tank blotting. Blots were incubated with mAbs (dilutions are listed in munity and were homogeneous with respect to social class and ethnicity. All patients were culture-positive for M. tuberculosis drug-sensitive organisms. The smear status, disease episode, antituberculosis chemotherapy status and IS6110 genotype of the infecting M. tuberculosis strain were documented for all patient samples (Table 2). Blood was collected by clinical staff; plasma was collected by centrifugation at 2500 g for 5 min, and stored at 220 uC. Ethical approval for this study was obtained from the University of Stellenbosch Faculty of Health Sciences ethics committee and samples were only taken after informed consent was given. RESULTS 2D PAGE and protein identification Mycobacterial strains were cultured in modified Sauton medium, a synthetic medium without protein enrichment (Andersen et al., 1991), for 4 weeks. This enabled CF proteins to be analysed free of BSA contamination. Pilot Table 2. TB patient data Patient no. 97 484 522 672 726 880 881 941 Strain family* Smear statusD Anti-TB treatmentd Disease episode§ F11 F11 F11 F11 F23 F29/Beijing F29/Beijing F29/Beijing positive negative negative positive negative positive positive positive yes (6 months) no no no yes (4 days) no no no second first first first first first first first *M. tuberculosis strains are grouped into families based on the number of IS6110 insertion elements identified by DNA fingerprinting. The strain family of the M. tuberculosis strain isolated from the patient sputum is shown. DSmear status refers to whether a patient’s sputum stained positive or negative for acid-fast bacilli. dAnti-TB treatment refers to the 6-month TB treatment regimen prescribed by DOTS (directly observed therapy shortcourse: isoniazid, pyrazinamide, rifampicin, streptomycin). The time blood was taken after the commencement of treatment is shown. §Disease episode refers to the TB episode the patient was experiencing. http://mic.sgmjournals.org Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 1141 C. Pheiffer and others studies showed that after 4 weeks growth all three strains were in late exponential growth, thereby minimizing growth-related differences between strains. Whole-cell lysate and CF proteins were separated by 2D PAGE using pH 4–7 IPG strips and 12 % PAGE gels. Pilot experiments demonstrated that many M. tuberculosis proteins focus between pH 4 and 7, and the best resolution was achieved using this pH range. The 2D gels of WCL and CF proteins derived from M. tuberculosis H37Rv, the Beijing strain and F23 are shown in Fig. 1. Some overlap between the proteins detected in the WCL and CF was observed, suggesting leakage of cell wall proteins or limited bacterial lysis. Fig. 1. 2D gels of WCL and CF proteins extracted from the three M. tuberculosis strains. (a) H37Rv, WCL; (b) H37Rv, CF; (c) Beijing strain, WCL; (d) Beijing strain, CF; (e) F23, WCL; (f) F23, CF. Protein spots with increased intensity in clinical strains are ringed and those with increased intensity in H37Rv are boxed. Their identities are listed in Table 3. 1142 Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 Microbiology 151 Mycobacterium tuberculosis protein expression Quantification of ICD in culture supernatants showed no significant difference in the extent of autolysis in the three strains (data not shown). The protein expression profiles were highly similar between the strains, enabling visual comparison. Spot differences consistent between quadruplicate gels, and obvious by visual examination, were analysed further. Protein spots of interest were analysed by LC/MS/MS. All proteins identified had their counterparts in all three strains, although quantitative expression differences were observed. Protein spots identified by LC/MS/MS are indicated in Fig. 1 and their identities listed in Table 3. Eight protein spots were increased in the WCLs of both clinical isolates relative to H37Rv. Seven of these were identified as acrystallin, and two proteins, Rv2005c and the 35 kDa antigen Rv2744c, were identified within the eighth spot. Previous studies have also reported the appearance of multiple a-crystallin species with different molecular masses and isoelectric points (Betts et al., 2000; Sonnenberg & Belisle, 1997). One of the factors contributing to mobility variants of a-crystallin is the oxidation of methionine residues, as recently demonstrated by Abulimiti et al. (2003). Three spots, identified as Ag85A, had decreased intensity in the CFs of both clinical strains compared to H37Rv. Spot 9 also contained Rv1096 and spot 10 three other proteins: Ag85C, Rv0831c and CysA3. Spot 12, identified as containing both PstS1 and Ag85B, was decreased in CFs of the Beijing strain compared to both H37Rv and F23. Two of the protein spots identified by LC/MS/MS contained novel proteins, Rv1096, Rv0831c and Rv3117, not previously identified by proteomic studies (Jungblut et al., 1999; Rosenkrands et al., 2000a, b; http://www.mpiib-berlin. mpg.de/2D-PAGE). Reactivity of mAbs with proteins extracted from M. tuberculosis strains In some cases, spots excised from 2D gels were identified as containing two or more proteins by LC/MS/MS. This included proteins of the Ag85 complex and PstS1. Western blot analysis was therefore employed to identify the proteins responsible for the differences in spot intensities observed on 2D gels and to investigate the expression of 12 other antigens, selected on the basis of mAb availability. 1D and 2D Western blotting identified PstS1 as the protein responsible for the decreased expression of spot 12 (Fig. 1, Table 3. Protein spot differences observed for whole-cell lysate and culture filtrate protein fractions of M. tuberculosis H37Rv, the Beijing and the F23 strain Spot Fraction ORF 1–7 WCL/CF Rv2031c 8 WCL Rv2005c Rv2744c 9 CF Rv1096D Rv3804c 10 CF Rv0129c Rv0831cD Rv3117D Rv3804c 11 CF Rv3804c 12 CF Rv0934 Rv1886c Gene name Protein Matched peptides Increased expression in both clinical strains compared to H37Rv hspX Heat-shock protein, HspX SEFAYGSFVR TVSLPVGADEDDIK – Conserved hypothetical protein YANAIGSAELAESSVQGR 35kd_ag Conserved 35 kDa alanine-rich protein GLLGSVSSSLVR Decreased expression in both clinical strains compared to H37Rv – Possible glycosyl hydrolase ANDVIAAATGR LVTVSELLGPR fbpA Antigen 85A FLEGFVR ASDMWGPK NDPLLNVGK ALGATPNTGPAPQGA fbpC Antigen 85C NDPMVQIPR EMPAWLQANK – Conserved hypothetical protein NQAIVVETTAYR cysA3 Probable thiosulphate sulphurtransferase DFVDAQQFSK fbpA Antigen 85A NDPLLNVGK ALGATPNTGPAPQGA fbpA Antigen 85A FLEGFVR NDPLLNVGK ALGATPNTGPAPQGA Decreased expression in Beijing strain only pstS1 Periplasmic phosphate-binding lipoprotein VLAAMYQGTIK fbpB Antigen 85B AADMWGPSSDPAWER Mr* pI* 16 096 5?00 30 985 29 257 5?53 5?71 31 109 6?52 35 686 6?08 36 771 5?92 33 885 31 014 35 686 5?88 5?14 6?08 35 686 6?08 38 243 34 580 5?14 5?62 *Predicted Mr and pI values calculated using the Expasy compute pI/Mw tool (http://www.expasy.ch/tools/pi_tool.html). DNovel proteins not previously identified by proteomics (Jungblut et al., 1999; Rosenkrands et al., 2000a, b; http://www.mpiib-berlin.mpg.de/ 2D-PAGE). http://mic.sgmjournals.org Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 1143 C. Pheiffer and others Fig. 2. Protein expression in M. tuberculosis strains. Ten micrograms of protein extracted from WCL (lanes 1, 3, 5) and CF (lanes 2, 4, 6) of M. tuberculosis H37Rv (lanes 1, 2), the Beijing strain (lanes 3, 4) and F23 (lanes 5, 6) was separated by 12 % SDS-PAGE, transferred onto PVDF membranes and probed with mAbs directed to (a) a-crystallin, (b) L-alanine dehydrogenase, (c) Hsp65, (d) PstS1, (e) KatG, (f) Ag85 complex, (g) 96 kDa protein, (h) 47 kDa protein, (i) ESAT6, ( j ) 19 kDa protein, (k) 17 kDa protein, (l) 20 kDa protein, (m) 56 kDa protein and (n) 25 kDa protein. Table 3) in the Beijing strain compared to the F23 strain and H37Rv (Fig. 2d, Fig. 3b). 2D Western blotting showed the existence of more than one species of PstS1, which has been previously reported by Sonnenberg & Belisle (1997) and is probably due to post-translational modifications such as glycosylation, phosphorylation or acetylation. The (a) a-Crystallin M. tuberculosis H37Rv WCL CF Beijing WCL F23 CF WCL CF WCL CF (b) PstS1 M. tuberculosis H37Rv WCL CF Beijing WCL F23 CF Fig. 3. a-Crystallin and PstS1 expression in M. tuberculosis strains. Sixty micrograms of protein extracted from M. tuberculosis H37Rv WCL (1, 7) and CF (2, 8), the Beijing strain WCL (3, 9) and CF (4, 10) and the F23 strain WCL (5, 11) and CF (6, 12) was separated by 12 % 2D PAGE, transferred to nitrocellulose membranes and probed with mAbs directed to (a) a-crystallin and (b) PstS1. 1144 Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 Microbiology 151 Mycobacterium tuberculosis protein expression absence of PstS1 on 2D blots of CFs of the F23 strain (Fig. 3b) could possibly be due to decreased expression of PstS1 in CFs compared to WCLs, as observed on 1D Western blots (Fig. 2d). Additionally, proteins may be more concentrated in bands on 1D gels than in spots on 2D gels and proteins may transfer more efficiently from 1D gels than from 2D gels. Several species corresponding to a-crystallin were observed and were expressed more highly in the Beijing strain compared to F23 and H37Rv (Fig. 2a and 3a). Consistent with the 2D results, expression of a-crystallin was increased in the WCL compared to the CF in all the strains. Expression of the 47 kDa protein was decreased in the Beijing strain compared to H37Rv and F23 (Fig. 2h). Two proteins, Hsp65 and the Ag85 complex, were downregulated in both clinical strains relative to H37Rv (Fig. 2c, f), while Hsp65 was also downregulated in the Beijing strain compared to F23. Multiple bands of Hsp65, possibly different isoforms or degradation products, were observed in both WCLs and CFs. Hsp65 observed in CFs is possibly due to bacterial lysis. L-Alanine dehydrogenase and ESAT6 (Fig. 2b, 2i) showed decreased expression in H37Rv compared to both clinical strains. The expression of the other proteins was approximately similar in all the strains analysed, although some quantitative differences between their expression levels in WCLs and CFs were observed. For example, expression of KatG was increased in the CF of H37Rv compared to the clinical strains, but decreased in the WCL of H37Rv compared to the clinical strains (Fig. 2e). The mAb directed to KatG (IT42) reacted with two bands in both WCLs and CFs, with the lower band corresponding to the correct molecular mass of 82 kDa. Sonnenberg & Belisle (1997) have also reported that IT42 reacts with more than one protein. The upregulation of a-crystallin, L-alanine dehydrogenase and ESAT6, and the downregulation of Hsp65 and Ag85 in the clinical strains compared to H37Rv, as well as the decreased expression of PstS1 in the Beijing strain compared to H37Rv and F23, were confirmed by ELISA (data not shown). Reactivity of plasma samples to proteins extracted from the three M. tuberculosis strains To establish whether there was differential antigen expression by the M. tuberculosis strains, plasma-derived antibodies of TB patients were tested against WCL proteins extracted from M. tuberculosis H37Rv, the Beijing strain and F23. After 1D PAGE separation and transfer to PVDF membranes, proteins were probed with plasma from TB patients infected with Beijing, F23 or F11 strains (Table 2, Fig. 4). As expected, plasma from TB patients reacted with a wide variety of M. tuberculosis proteins, ranging from 10 kDa to over 105 kDa (Fig. 4). Plasma samples were tested on antigens fractionated on different gels; therefore figures were aligned according to molecular mass markers, which were included on each gel, and mobility of immunodominant antigens. As noted previously by others, patientto-patient variation, where each patient had a characteristic http://mic.sgmjournals.org banding pattern (Fig. 4), was evident. However, some prominent differences between the antigens expressed by the strains were observed; these are circled in Fig. 4 and listed in Table 4. A band of approximately 60 kDa (circle 1) was recognized by patient 880 in H37Rv but not in F23 or the Beijing strain (Fig. 4f). Patient 880 also reacted with a protein of about 40 kDa (circle 2) in protein extracts of F23 and the Beijing strain only (Fig. 4f), with more prominent expression in the Beijing strain. Similarly, a band of approximately 15 kDa (circle 5) was recognized more strongly in the Beijing strain compared to the other strains by patient 880. Proteins of approximately 28 and 25 kDa which were recognized strongly by patient 97 in H37Rv are indicated by circles 3 and 4 (Fig. 4a). 2D Western blotting of patient plasma samples 97, 726, 880 and 941 against WCL proteins extracted from M. tuberculosis H37Rv, the Beijing strain and F23 was used to further investigate the differences seen using the 1D approach through the increased resolution obtainable (Fig. 5). Many differences in antigen expression levels between the strains were identified, extending the findings using 1D Western blotting (Table 4). Again, each patient showed a characteristic antigen recognition pattern. Patients 97, 880 and 941 reacted with fewer proteins in the Beijing strain compared to the other strains, indicated by boxes 1, 4 and 7, respectively, in Fig. 5, perhaps suggesting lower levels of expression of these antigens in the Beijing strain. Interestingly, patients 880 and 941 were infected with strains belonging to the Beijing family. Patient 97 also recognized a spot at approximately 10 kDa more strongly in the clinical strains than H37Rv (Fig. 5, spot 2). Many of the other differences observed in antigen recognition between the strains for patient 97 (Fig. 4a) were obscured by a broad, diffuse smear on 2D blots (Fig. 5). The smear is probably lipoarabinomannan (LAM), an immunodominant glycolipid which migrates between 30 and 40 kDa, as demonstrated previously (Hunter et al., 1986). Patient 726 recognized more proteins in the Beijing strain compared to H37Rv and F23, which is surprising as this patient is infected with an F23 strain (Fig. 5, box 3). The 40 kDa and 60 kDa changes observed on 1D blots using plasma from patient 880 (Fig. 4f) are in a region containing many spots (Fig. 5, box 4) on the 2D blots and therefore cannot be uniquely identified. However, notable differences in protein expression within this region, particularly between the clinical strains and H37Rv, were evident. Similarly, patient 941 recognized a spot migrating at about 32 kDa more strongly in the clinical strains (Fig. 5, spot 9). In contrast, patient 880 showed decreased recognition of a spot at approximately 30 kDa (Fig. 5, spot 5) in the Beijing strain, as did patient 941 with a spot at 35 kDa (Fig. 5, spot 8). Although it is difficult to speculate which proteins these spot differences represent, comparison with silver-stained 2D gels suggests that spot number 6, which is recognized by patient 880 only, represents a-crystallin. Supporting this Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 1145 C. Pheiffer and others Fig. 4. Antigen expression by M. tuberculosis strains. Plasma from TB patients (a) 97, (b) 484, (c) 522, (d) 672, (e) 726, (f) 880, (g) 881, (h) 941 was probed against whole-cell lysate proteins extracted from M. tuberculosis H37Rv (lane 1, 1 : 100 plasma dilution; lane 2, 1 : 500 plasma dilution), the Beijing strain (lane 3, 1 : 100 plasma dilution; lane 4, 1 : 500 plasma dilution) and the F23 strain (lane 5, 1 : 100 plasma dilution; lane 6, 1 : 500 plasma dilution). Proteins were separated by 12 % SDS-PAGE under reducing conditions and then transferred onto PVDF membranes. Molecular mass markers are indicated on the left. Prominent differences between the strains are circled. Table 4. Antigen expression differences detected between strains Band/spot 1D Western blots 1 2 3 4 5 2D Western blots 1 2 3 4 5 6 7 8 9 Patient Approx. mol. mass (kDa) H37Rv* Beijing F23 880 880 97 97 880 60 40 28 25 14 + 2 ++ ++ + 2 ++ 2 + ++ 2 + 2 + + 97 97 726 880 880 880 941 941 941 35250 10 15225 35250 30 14 35250 35 32 +++ 2 + +++ ++ +++ ++ + 2 + + +++ + + ++ + 2 + ++ + ++ ++ ++ ++ ++ + + *Relative expression levels are indicated by +. Protein bands/spots absent are indicated by 2. 1146 Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 Microbiology 151 Mycobacterium tuberculosis protein expression Fig. 5. Antigen expression by M. tuberculosis strains. WCL proteins extracted from cultures of M. tuberculosis H37Rv, the Beijing strain and F23 were separated by 12 % 2D PAGE, transferred to nitrocellulose membranes and probed with a 1 : 150 dilution of plasma samples from TB patients 97, 726, 880 and 941. Molecular mass markers are indicated on the right. Differential antigen recognition is indicated by circles and boxes. speculation is the fact that a band migrating at approximately the molecular mass of a-crystallin was prominently recognized by patient 880 on 1D blots (Fig. 4f, circle 5). Analysis of a-crystallin expression across the three strains using mAbs (Fig. 3a) showed increased expression of different protein species in the Beijing strain, not reflected using patient anti-sera. This may suggest that these forms are not expressed in vivo or are not recognized by patient antibodies. Several of the proteins for which expression differences were observed using serology have similar molecular masses to proteins identified as different between the strains on the silver-stained 2D gels. However, the serology approach has identified several other antigens not apparent from comparison of the silver-stained gels, possibly due to low abundance compared to other proteins. As such, the serology has proved a complementary approach in identifying antigens differentially expressed http://mic.sgmjournals.org between the strains and further investigation of methods to identify these proteins would be warranted. DISCUSSION The availability of the genome sequence of M. tuberculosis H37Rv (Cole et al., 1998) has facilitated the study of the proteome of the bacillus. Here, we describe a comparison of the proteome of M. tuberculosis H37Rv and two local clinical strains, one belonging to the Beijing strain family and ten-fold more prevalent than the other clinical strain (F23) in the study community. The Beijing strain family is responsible for at least 18?7 % of all TB cases in the study community (Warren et al., 2000). It was hoped that comparison of protein expression and antigen recognition patterns of these strains would give some clues as to why the Beijing strain is more prevalent than the other clinical strain. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 1147 C. Pheiffer and others The high prevalence of the Beijing genotype worldwide is indicative of the success of this M. tuberculosis strain type as a human pathogen (Glynn et al., 2002). It has been suggested that the global dissemination of this strain genotype may be linked to an altered phenotype, thereby conferring an advantage over M. tuberculosis strains belonging to other genotypes (Lopez et al., 2003). This is supported by the observation that Beijing strains are more virulent and elicit a non-protective immune response compared to other genotypes, during experimental disease in mice (Lopez et al., 2003). Infection with Beijing genotypes has also been linked to the development of fever during the early phase of treatment (van Crevel et al., 2001). These findings highlight the need to elucidate the mechanisms underlying the success of Beijing strains compared to other M. tuberculosis strains. It has been suggested that the wide dispersion of the Beijing strain family compared to other less prevalent clinical isolates may be related to differential protein expression (Bifani et al., 2002). In agreement with previous studies of protein expression across different M. tuberculosis strains (Betts et al., 2000; Jungblut et al., 1999; Mattow et al., 2001), this study revealed that a large portion of the proteome is similarly expressed. Despite this, we have been able to detect several differences in protein expression and shown distinct differences in antigen expression levels across the three strains, with several changes being particular to the Beijing strain. Several species of a-crystallin, a M. tuberculosis virulence factor (Monahan et al., 2001; Sherman et al., 2001; Yuan et al., 1998), were more highly expressed in the Beijing strain compared to H37Rv and F23. Altered expression of this protein may be a factor contributing to the virulence of Beijing strains. The success of the Beijing family as a human pathogen may in part be due to decreased expression of Hsp65, PstS1 and the 47kDa protein as it has previously been suggested that reduced expression of certain major antigens may allow strains to evade the host immune response (Stewart et al., 2001). Moreover, differences in antigen expression levels between the strains, as observed from 1D and 2D Western blots with patient plasma, may be significant and could assist the Beijing strain to minimize recognition by the host immune response, thereby facilitating the increased prevalence of Beijing strains. In order to further assess the importance of the antigens identified here in general strain prevalence or whether these differences are unique to the Beijing family, expression levels could be determined in other prevalent strains. Importantly, certain antigen expression differences between strains may only become evident when the bacteria are growing within the host environment. Therefore, it will be necessary to analyse proteins extracted from mycobacteria isolated from conditions that mimic the host environment, for example macrophages or animal models. In addition, to further validate the antigens found to be differentially expressed here and determine whether they also show expression differences in vivo, sera from patients infected with the Beijing strain could be used to probe against 1148 purified recombinant proteins of the antigens of interest, with the caveat that patient variation and extent of disease may influence antigen recognition. Differential protein expression may explain the heterogeneous host humoral immune response and why no serodiagnostic test for TB has yet been developed (Lyashchenko et al., 1998). The most marked difference in protein expression between the Beijing and F23 was observed for PstS1. PstS1 is a 38 kDa M. tuberculosis complex-specific phosphate-binding lipoprotein, and a known B- and T-cell stimulant (Harboe & Wiker, 1992). To date, PstS1 has shown promise for serological diagnosis of TB, with sensitivities of 70 % and 73 % for smearnegative pulmonary TB and extrapulmonary TB patients, respectively (Wilkins & Ivanyi, 1990). However, in a recent evaluation of commercially available tests the highest sensitivity achieved with PstS1 was 55 % (Pottumarthy et al., 2000). The absence of PstS1 antibodies in some TB patients may be due to decreased expression of PstS1 in some clinical strains, as observed for the Beijing strain in this study. These findings suggest that the development of a serodiagnostic test for TB may be hindered by variable protein expression by M. tuberculosis strains, and support the development of a test measuring the levels of antibodies to a panel of antigens that are common to different M. tuberculosis genotypes (Al Zahrani et al., 2000; Amicosante et al., 1999). Furthermore, differential expression of PstS1 by different M. tuberculosis genotypes suggests that this antigen may not be a good choice for a vaccine, and may partially explain the discrepant results obtained when using PstS1 as a subunit or DNA vaccine candidate (Falero-Diaz et al., 2000). This study has demonstrated extensive humoral heterogeneity between patients, even though they were homogeneous with respect to social class and ethnicity. Humoral heterogeneity has previously been shown (Lyashchenko et al., 1998) and attributed, in part, to differential protein expression by M. tuberculosis genotypes. Variation in antigen recognition was observed even between patients infected with the same genotype, highlighting the importance of other factors, in addition to differential protein expression, in dictating the host humoral immune response. These factors could be host factors (Bothamley et al., 1989), extent of disease (Jackett et al., 1988), or even multiple infections with different M. tuberculosis genotypes (Warren et al., 2004). Taken together, these results suggest that detection of mycobacterial antigens (Wallis et al., 1998) may be more valuable than antibody detection, since antigen detection will not be influenced by the significant humoral heterogeneity between patients. In conclusion, this study has shown that proteome analysis of M. tuberculosis genotypes may contribute to our understanding of the pathogenesis of tuberculosis by identifying differentially expressed proteins, and potentially helping us to understand why certain strain families, such as the Beijing family, are more successful than others. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 Microbiology 151 Mycobacterium tuberculosis protein expression In addition to aiding an understanding of pathogenicity and strain prevalence, identification of antigens differentially expressed between strains provides important information for consideration in the design of serodiagnostic tests, vaccines and even drug target selection. Falero-Diaz, G., Challacombe, S., Banerjee, D., Douce, G., Boyd, A. & Ivanyi, J. (2000). Intranasal vaccination of mice against infection with Mycobacterium tuberculosis. Vaccine 18, 3223–3229. Glynn, J. R., Whiteley, J., Bifani, P. J., Kremer, K. & van Soolingen, D. (2002). Worldwide occurrence of Beijing/W strains of Mycobacterium tuberculosis: a systematic review. Emerg Infect Dis 8, 843–849. Harboe, M. & Wiker, H. G. (1992). The 38-kDa protein of Myco- bacterium tuberculosis: a review. J Infect Dis 166, 874–884. ACKNOWLEDGEMENTS He, X. Y., Zhuang, Y. H., Zhang, X. G. & Li, G. L. (2003). Com- We thank the National Institutes of Health and the National Institute of Allergy and Infectious Diseases for mAbs made available through the Department of Microbiology at Colorado State University. We would also like to thank Karin Weldingh of the Statens Serum Institut, Denmark, for kindly providing us with HYB76-8. We are grateful to Eileen Hoal van Helden for kindly supplying TB plasma samples. This work was funded by the GlaxoSmithKline Action TB initiative and the National Research Foundation, South Africa. parative proteome analysis of culture supernatant proteins of Mycobacterium tuberculosis H37Rv and H37Ra. Microbes Infect 5, 851–856. Heifets, L. B. & Good, R. C. (1994). Current laboratory methods for the diagnosis of tuberculosis. In Tuberculosis: Pathogenesis, Protection, and Control, pp. 85–110. Edited by B. R. Bloom. Washington, DC: American Society for Microbiology. Herbert, B. R., Molloy, M. P., Gooley, A. A., Walsh, B. J., Bryson, W. G. & Williams, K. L. (1998). Improved protein solubility in two- dimensional electrophoresis using tributyl phosphine as reducing agent. Electrophoresis 19, 845–851. REFERENCES Hunter, S. W., Gaylord, H. & Brennan, P. J. (1986). Structure and Abulimiti, A., Qiu, X., Chen, J., Liu, Y. & Chang, Z. (2003). Reversible antigenicity of the phosphorylated lipopolysaccharide antigens from the leprosy and tubercle bacilli. J Biol Chem 261, 12345–12351. methionine sulfoxidation of Mycobacterium tuberculosis small heat shock protein Hsp16.3 and its possible role in scavenging oxidants. Biochem Biophys Res Commun 305, 87–93. Jackett, P. S., Bothamley, G. H., Batra, H. V., Mistry, A., Young, D. B. & Ivanyi, J. (1988). Specificity of antibodies to immunodominant Al Zahrani, K., Al Jahdali, H., Poirier, L., Rene, P., Gennaro, M. L. & Menzies, D. (2000). Accuracy and utility of commercially available mycobacterial antigens in pulmonary tuberculosis. J Clin Microbiol 26, 2313–2318. amplification and serologic tests for the diagnosis of minimal pulmonary tuberculosis. Am J Respir Crit Care Med 162, 1323–1329. Jungblut, P. R., Schaible, U. E., Mollenkopf, H. J. & 7 other authors (1999). Comparative proteome analysis of Mycobacterium tubercu- Amicosante, M., Houde, M., Guaraldi, G. & Saltini, C. (1999). Sensi- losis and Mycobacterium bovis BCG strains: towards functional genomics of microbial pathogens. Mol Microbiol 33, 1103–1117. tivity and specificity of a multi-antigen ELISA test for the serological diagnosis of tuberculosis. Int J Tuberc Lung Dis 3, 736–740. Andersen, P., Askgaard, D., Ljungqvist, L., Bennedsen, J. & Heron, I. (1991). Proteins released from Mycobacterium tuberculosis during Khanolkar-Young, S., Kolk, A. H. J., Andersen, A. B. & 11 other authors (1992). Results of the third immunology of leprosy/ growth. Infect Immun 59, 1905–1910. immunology of tuberculosis antimycobacterial monoclonal antibody workshop. Infect Immun 60, 3925–3927. Betts, J. C. & Smith, M. A. (2001). Proteomics. Methods Microbiol 54, Lopez, B., Aguilar, D., Orozco, H. & 8 other authors (2003). A 315–334. Betts, J. C., Dodson, P., Quan, S., Lewis, A. P., Thomas, P. J., Duncan, K. & McAdam, R. A. (2000). Comparison of the proteome of marked difference in pathogenesis and immune response induced by different Mycobacterium tuberculosis genotypes. Clin Exp Immunol 133, 30–37. Mycobacterium tuberculosis strain H37Rv with clinical isolate CDC 1551. Microbiology 146, 3205–3216. Lyashchenko, K., Colangeli, R., Houde, M., Al Jahdali, H., Menzies, D. & Gennaro, M. L. (1998). Heterogeneous antibody Bifani, P. J., Mathema, B., Kurepina, N. E. & Kreiswirth, B. N. (2002). responses in tuberculosis. Infect Immun 66, 3936–3940. Global dissemination of the Mycobacterium tuberculosis W-Beijing family strains. Trends Microbiol 10, 45–52. Manca, C., Tsenova, L., Barry, C. E., III, Bergtold, A., Freeman, S., Haslett, P. A., Musser, J. M., Freedman, V. H. & Kaplan, G. (1999). Bjellqvist, B., Pasquali, C., Ravier, F., Sanchez, J. C. & Hochstrasser, D. (1993). A nonlinear wide-range immobilized pH Mycobacterium tuberculosis CDC1551 induces a more vigorous host response in vivo and in vitro, but is not more virulent than other clinical isolates. J Immunol 162, 6740–6746. gradient for two-dimensional electrophoresis and its definition in a relevant pH scale. Electrophoresis 14, 1357–1365. Bothamley, G. H., Beck, J. S., Schreuder, G. M., D’Amaro, J., de Vries, R. R., Kardjito, T. & Ivanyi, J. (1989). Association of Mattow, J., Jungblut, P. R., Schaible, U. E., Mollenkopf, H. J., Lamer, S., Zimny-Arndt, U., Hagens, K., Muller, E. C. & Kaufmann, S. H. (2001). Identification of proteins from Mycobacterium tuberculosis and M. tuberculosis-specific antibody levels with HLA. J Infect Dis 159, 549–555. tuberculosis missing in attenuated Mycobacterium bovis BCG strains. Electrophoresis 22, 2936–2946. Bradford, M. M. (1976). A rapid and sensitive method for the Mattow, J., Schaible, U. E., Schmidt, F. & 7 other authors (2003). quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248–254. Comparative proteome analysis of culture supernatant proteins from virulent Mycobacterium tuberculosis H37Rv and attenuated M. bovis BCG Copenhagen. Electrophoresis 24, 3405–3420. Cole, S. T. R., Brosch, J., Parkhill, T. & 39 other authors (1998). Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544. Engers, H. D., Houba, V., Bennedsen, J. & 19 other authors (1986). Results of a World Health Organization-sponsored workshop to characterize antigens recognized by Mycobacterium-specific monoclonal antibodies. Infect Immun 51, 718–720. http://mic.sgmjournals.org Monahan, I. M., Betts, J., Banerjee, D. K. & Butcher, P. D. (2001). Differential expression of mycobacterial proteins following phagocytosis by macrophages. Microbiology 147, 459–471. Pheiffer, C., Betts, J., Lukey, P. & van Helden, P. (2002). Protein expression in Mycobacterium tuberculosis differs with growth stage and strain type. Clin Chem Lab Med 40, 869–875. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 1149 C. Pheiffer and others Pottumarthy, S., Wells, V. C. & Morris, A. J. (2000). A comparison of seven tests for serological diagnosis of tuberculosis. J Clin Microbiol 38, 2227–2231. van Crevel, R., Nelwan, R. H., de Lenne, W., Veeraragu, Y., van der Zanden, A. G., Amin, Z., van der Meer, J. W. & van Soolingen, D. (2001). Mycobacterium tuberculosis Beijing Rosenkrands, I., Weldingh, K., Jacobsen, S., Hansen, C. V., Florio, W., Gianetri, I. & Andersen, P. (2000a). Mapping and genotype strains associated with febrile response to treatment. Emerg Infect Dis 7, 880–883. identification of Mycobacterium tuberculosis proteins by twodimensional gel electrophoresis, microsequencing and immunodetection. Electrophoresis 21, 935–948. van Soolingen, D., Qian, L., de Haas, P. E. & 7 other authors (1995). Rosenkrands, I., King, A., Weldingh, K., Moniatte, M., Moertz, E. & Andersen, P. (2000b). Towards the proteome of Mycobacterium Wallis, R. S., Perkins, M., Phillips, M. & 11 other authors (1998). tuberculosis. Electrophoresis 21, 3740–3756. Sherman, D. R., Voskuil, M., Schnappinger, D., Liao, R., Harrell, M. I. & Schoolnik, G. K. (2001). Regulation of the Mycobacterium tuberculosis hypoxic response gene encoding alpha-crystallin. Proc Natl Acad Sci U S A 98, 7534–7539. Sonnenberg, M. G. & Belisle, J. T. (1997). Definition of Myco- Predominance of a single genotype of Mycobacterium tuberculosis in countries of east Asia. J Clin Microbiol 33, 3234–3238. Induction of the antigen 85 complex of Mycobacterium tuberculosis in sputum: a determinant of outcome in pulmonary tuberculosis treatment. J Infect Dis 178, 1115–1121. Warren, R. M., Sampson, S. L., Richardson, M., Van Der Spuy, G. D., Lombard, C. J., Victor, T. C. & van Helden, P. D. (2000). Mapping of IS6110 flanking regions in clinical isolates of Mycobacterium tuberculosis demonstrates genome plasticity. Mol Microbiol 37, 1405–1416. bacterium tuberculosis culture filtrate proteins by two-dimensional polyacrylamide gel electrophoresis, N-terminal amino acid sequencing, and electrospray mass spectrometry. Infect Immun 65, 4515–4524. Warren, R. M., Victor, T. C., Streicher, E. M., Richardson, M., Beyers, N., van Pittius, N. C. & van Helden, P. D. (2004). Patients Sorensen, A. L., Nagai, S., Houen, G., Andersen, P. & Andersen, A. B. (1995). Purification and characterization of a low-molecular-mass with active tuberculosis often have different strains in the same sputum specimen. Am J Respir Crit Care Med 169, 610–614. T-cell antigen secreted by Mycobacterium tuberculosis. Infect Immun 63, 1710–1717. Wilkins, E. G. & Ivanyi, J. (1990). Potential value of serology for Stewart, G. R., Snewin, V. A., Walzl, G. & 7 other authors (2001). Overexpression of heat-shock proteins reduces survival of Yuan, Y., Crane, D. D., Simpson, R. M., Zhu, Y. Q., Hickey, M. J., Sherman, D. R. & Barry, C. E., III (1998). The 16-kDa alpha-crystallin Mycobacterium tuberculosis in the chronic phase of infection. Nat Med 7, 732–737. (Acr) protein of Mycobacterium tuberculosis is required for growth in macrophages. Proc Natl Acad Sci U S A 95, 9578–9583. 1150 diagnosis of extrapulmonary tuberculosis. Lancet 336, 641–644. Downloaded from www.microbiologyresearch.org by IP: 88.99.165.207 On: Sat, 12 Aug 2017 04:26:35 Microbiology 151