Cloning and Expression of Immunoreactive Antigens from Mycobacterium tuberculosis

M. tuberculosis total-protein extraction.

M. tuberculosis cells (ATCC 27294) were cultured in MycoFlasks (Gibco, BRL) containing Lowenstein-Jensen medium at 37°C with 10% CO₂ in a humidified incubation chamber (Jouan IG/50 model). Confluent cells from six culture flasks were harvested by adding 3 ml of Middlebrook 7H9 medium (Difco Laboratories, Detroit, Mich.) into each flask and gently flushing the surface of the flask. Dislocated cells were placed into sterile plastic tubes (Falcon), and cells were pelleted by centrifugation at 1,100 × g for 5 min. Cells were washed once in an equal volume of distilled H₂O before being resuspended in an equal volume of distilled H₂O. The cell suspension was then heated to 90°C for 2 h and frozen at −20°C overnight. Cells were then thawed on ice and pelleted by centrifugation at 20,000 × g for 10 min. Extraction of total protein was performed by adding 500 μl of 8 M urea solution to 0.5 g of cell pellet, vortexing the cell suspension at room temperature for 20 min, and heating it at 90°C for 2 min. Insoluble cellular debris was removed by centrifugation at 20,000 × g for 10 min, and the supernatant containing the extracted total protein was kept at −20°C until further use.

Western blot analysis.

The total-protein extract of M. tuberculosis was fractionated on a sodium dodecyl sulfate–7.5% polyacrylamide gel electrophoresis (SDS–7.5% PAGE) gel (20) and transferred onto a nitrocellulose membrane by Western blotting (33). Strips from the immunoblot were probed against pooled positive (tested positive by skin PPD and culture) and negative sera from nine individuals with active TB and seven healthy individuals, respectively. Incubation with pooled sera (1:100 in 1% skim milk-TBST [10 mM Tris, pH 7.5, 300 mM NaCl, 0.005% Tween 20]) was carried out with rocking for 1 h at room temperature. The blots were then washed four times in TBST before incubation with alkaline phosphatase-conjugated goat anti-human immunoglobulin (Ig) (Harlan Sera Lab, Loughborough, United Kingdom) (1:1000 in 1% skim milk-TBST) for another 1 h. The strips were again washed four times in TBST followed by incubation in 1 ml of bromochloroindolyl phosphate-nitroblue tetrazolium substrate (NBT-BCIP; Bio-Rad) for 4 min. The reaction was stopped by washing four times in distilled H₂O.

N-terminal sequencing.

Individual protein bands (which were shown to react positively in the immunoscreening experiment) were excised from several preparative SDS–7.5% polyacrylamide gels and concentrated by reelectrophoresis (constant current of 18 mA at 8°C) on a long stacking gel (7 cm of 4% stacking gel, 5 cm of 10% resolving gel). The concentrated protein bands were blotted onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad), stained with Coomassie brilliant blue R-250 (Sigma, St. Louis, Mo.), and excised for N-terminal microsequencing. The protein bands were also blotted onto Hybond-C nitrocellulose membranes (Amersham Life Science, Little Chalfont, United Kingdom) for validation by immunoscreening using the same pooled sera samples as described above.

Screening of a phage expression library.

An expression library of EcoRI-restricted genomic DNA of M. tuberculosis was constructed in lambda ZAP Phage expression vector, according to the protocol by Stratagene (ZAP Express cDNA Synthesis kit manual, Stratagene Cloning Systems, 1998). The resultant library has 98% recombinants (2 × 10⁶ PFU/μg arms) and insert sizes ranging from 0.7 to 2 kb. A lawn of XL1-MRF′ host cells infected with about 2 × 10⁴ PFU of the phage stock was prepared on a 150-mm plate and incubated for 6 to 7 h at 42°C. The lawn was then overlaid with a Hybond-C nitrocellulose membrane (Amersham) presoaked in 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for induction of protein expression by further incubation at 37°C for 4 h. The plate and membrane were indexed for matching corresponding plate and membrane position. Approximately 0.4 × 10⁶ plaques were screened, and as a negative control, a lawn of host cells infected with 2 × 10⁴ PFU of the nonrecombinant lambda ZAP phage was used instead.

After transfer, membranes were washed twice in TBST buffer and blocked in 5% skim milk-TBST. Pooled sera from four individuals with active TB and from four healthy individuals were preabsorbed overnight, against negative-control membranes. Membranes from the expression library were incubated with the preabsorbed human sera for 2 h with rocking at room temperature, followed by three washes in TBST. A secondary antibody of alkaline phosphatase-conjugated goat anti-human Ig (diluted 1:1,000 1% skim milk-TBST) was added to the membranes and allowed to incubate for 1 h. After a final wash, colorimetric detection was performed using NBT-BCIP substrate as described previously. Positive plaques were cored out, and recombinant phage was eluted in SM buffer containing 2% chloroform (Stratagene manual). These were replated at about 100 to 200 PFU on 82-mm plates for secondary and tertiary screenings using the same preabsorbed pooled sera. Positive recombinant phage clones from tertiary screenings were subjected to pBlueScript SK plasmid excision using helper phage, and recombinant plasmids were DNA sequenced using the forward T3 primer (Stratagene manual).

Cloning and expression of M. tuberculosis antigens.

Synthetic oligonucleotides were designed for PCR and cloning of the B.4, B.6, B.10, B.M, and C17 genes, including a known 38-kDa antigen (2). M. tuberculosis genomic DNA was extracted as previously described with a few modifications (7). The PCR mixture contained 0.1 μg of M. tuberculosis genomic DNA, 50 pmol of synthetic oligonucleotides (Genset, Singapore), 10 μl of GC-melt (from the Advantage-GC Genomic PCR kit [Clontech, Inc., Palo Alto, Calif.], 5 μl of 10× Expand High-Fidelity DNA Polymerase buffer (containing MgCl₂), 0.2 mM deoxynucleoside triphosphates (dNTPs), 0.75 U of Expand High-Fidelity DNA Polymerase (Boehringer GmbH, Mannheim, Germany), and distilled H₂O to a total reaction volume of 50 μl. PCR amplications were carried out in a DNA PTC-100 thermal cycler (MJ Research, Watertown, Mass.) under the following conditions: 94°C for 2 min; 9 cycles of 94°C for 15 s, 55°C for 30 s, and 72°C for 1 min; 19 cycles of 94°C for 15 s, 55°C for 30 s, and 72°C for 1 min plus 20 s per cycle; and a final extension step at 72°C for 7 min.

PCR products were initially cloned into pGEM-T Easy vector (Promega, Madison, Wis.) and subsequently subcloned into pQE30 expression vector (Qiagen, Inc., Valencia, Calif.). DH5α and Epicurian Coli XL10-Gold (Stratagene) E. coli cells were used for cloning and maintenance of plasmid DNA vector and recombinant plasmid DNA, whereas M₁₅ E. coli cells (Qiagen) were used for expression studies. Transformation of plasmid DNA into E. coli cells was done by a heat shock method as described by Crouse et al. (13). Plasmid DNA extraction was performed using the Wizard Midi and Miniprep DNA purification systems (Promega). DNA sequencing was carried out on an Automated DNA Sequencer, model 373A (Perkin-Elmer Applied Biosystems, Foster City, Calif.) using the forward (5′-GAATTCATTAAAGAGGAGAAA-3′) and reverse (5′-GTTCTGAGGTCATTACTGG-3′) primers of the pQE30 vector and primers synthesized based on internal sequences obtained from DNA sequencing of the TB genes. Recombinant proteins were expressed as NH₂-terminally polyhistidine-tagged fusion proteins and purified from the M₁₅ E. coli cell extracts to near-homogeneity by Ni-nitrilotriacetic acid (NTA) affinity chromatography (Qiagen, The QIAexpressionist, 3rd ed., July 1997). Quantitation of recombinant proteins was carried out using the DC Protein Assay Kit (Bio-Rad). SDS-PAGE gels for visualization were stained either with Coomassie brilliant blue R-250 or with silver stain (27).

SDS-PAGE and Western blotting of M. tuberculosis recombinant antigens.

To maintain consistency, Tris-HCl two-dimensional (2D) preparative ready gels (Bio-Rad) were used for Western blotting. A total of 10 μg of purified recombinant antigen was subjected to SDS-PAGE and Western blotted onto Hybond-C nitrocellulose membranes (Amersham) using the Bio-Rad TransBlotter (according to the manufacturer's protocol). After transfer, the membrane was blocked in 5% skim milk-TBST, air dried, and stored at 4°C until further use.

Immunoblot analysis of recombinant antigens.

Each membrane containing Western-blotted antigen was cut into strips of 3-mm width (a total of 23 strips from each blot), and each strip was used for screening with a serum specimen. One strip was used as an internal positive control probed with a positive serum specimen that is reactive to the recombinant protein antigens. A second strip was probed with the commercially available anti-RGS His probe (Qiagen). Screening was carried out by incubating each strip in trays with 1 ml of diluted serum specimen (1:100 in 1% skim milk-TBST) per well for 1 h with rocking at room temperature. The strips were then washed four times in TBST, followed by incubation with alkaline phosphatase-conjugated goat anti-human Ig (Harlan Sera Lab) for 1 h with rocking at room temperature. The strips were again washed four times in TBST and then allowed to develop in 1 ml of NBT-BCIP substrate (Bio-Rad) for 4 min. The reaction was stopped by washing four times in distilled H₂O.

Western blot score.

The reactivities of recombinant proteins to serum specimens were interpreted based on the intensities of bands obtained on an X-Rite 400 densitometer (X-Rite Inc., Grandville, Mich.). The cutoff values (expressed as densities) for each individual recombinant antigen were determined based on the range observed for normal serum specimens.

Serum specimens.

A total of 139 human sera were used in this study, of which 119 serum specimens were purchased from BioClinical Partners, Inc., Franklin, Mass., whereas 20 were donated by healthy laboratory workers. The control groups consisted of two panels, (i) a panel of 50 serum specimens from healthy individuals, of whom 20 (laboratory workers) had been Mycobacterium bovis BCG vaccinated previously and 30 (BioClinical) had unknown BCG status, and (ii) a panel of 19 serum specimens (BioClinical) from individuals with non-TB respiratory disease (lung cancer).

The test group consisted of 48 serum specimens (BioClinical) from bacteriologically confirmed active TB patients and 22 serum specimens (BioClinical) from patients with inactive TB. The active-TB serum specimens comprised 28 from pulmonary-TB and 20 from extrapulmonary-TB patients. Serum specimens from the inactive-TB panel were from patients with positive PPD skin tests but negative acid-fast stains of sputum and bacterial culture. All sera were aliquoted and stored at −70°C before use.

Databases and software.

Nucleotide and protein sequence analysis was carried out using the basic BLAST 2.0 search program from the National Center for Biotechnology Information (NBCI) and the M. tuberculosis BLAST server at the Sanger Centre (Cambridge, United Kingdom).

Statistical analysis.

Sensitivities, specificities, and positive and negative predictive values were calculated using the Win Episcope 1.0 (Borland International Inc.).

Identification and isolation of M. tuberculosis antigens.

Immunoblot analysis of M. tuberculosis total proteins revealed protein bands which reacted with the pooled active sera but not with the pooled normal sera. When the respective bands were concentrated on a long stacking gel, excised, and Western blotted, these bands were reactive with the pooled active sera but not with pooled normal sera, thus confirming the authenticity of these excised proteins as those initially observed in the primary screening (Fig. 1). These proteins were identified by homology searches against protein sequence databases, which gave a high percentage of homology to Mycobacterium proteins (Table 1). PCR primers were designed to isolate and clone the genes coding for four proteins (B.4, B.6, B.10, and B.M) which gave the highest matches based on a BLAST homology search against the SwissProt database.

Concurrently, primary screening of the phage expression library gave eight reactive phage recombinants, of which six were further confirmed positive by secondary and tertiary screenings (Fig. 2). These clones were subjected to plasmid DNA excision, and restriction enzyme digestions with EcoRI indicated that all the clones contained a 2-kb insert. DNA sequencing revealed that all the clones were identical, having a 1.161-kb open reading frame (in frame with the vector's ATG initiation codon) which coded for a proline-rich protein. A summary of all five TB antigens, with the respective gene sizes, theoretical molecular masses, and pI values is shown in Table 2.

Expression and purification of recombinant antigens.

Expression was detected by probing immunoblots containing these antigens using the commercial anti-RGS His antibody. The levels of expression observed were high for the B.4, B.M, and 38-kDa proteins, moderate for the B.6 and B.10 proteins, and low for the C17 protein (Fig. 3). All of these recombinant proteins, except for C17, were present in the insoluble fraction of an SDS-PAGE analysis (data not shown), indicating that these proteins formed inclusion bodies and were insoluble. As such, these recombinant proteins were purified by Ni-NTA affinity chromatography in 8 M urea, and the SDS-PAGE profile of the purified antigens subsequently used for immunoblots is shown in Fig. 4. The approximate yields of recombinant antigens purified through Ni-NTA were 36 mg/liter for B.4, 0.5 mg/liter for B.6, 0.2 mg/liter for B.10, 15 mg/liter for B.M, <0.1 mg/liter for C17, and 10 mg/liter for the 38-kDa protein.

Reactivities of the recombinant antigens to TB serum specimens.

The different reactivities of recombinant TB antigens on immunoblot strips probed with serum specimens and anti-RGS His are shown in Fig. 5. The respective cutoff values for determining reactivity to the different recombinant antigens were obtained based on the mean densities observed in sera from the control group of healthy individuals (n = 50). The cutoff values were densities of >0.04 for the B.4 and B.6 bands, ≥0.04 for the B.10 and B.M bands, ≥0.15 for the C17 band, and >0.15 for the 38-kDa band.

Based on the Western blot assay, the reactivities of these antigens to a panel of active-TB serum specimens are shown in Table 3. Percentages of reactivities and positive and negative predictive values for each antigen were calculated based on sera from infected individuals (n = 48; pulmonary and extrapulmonary TB) and sera from the control group of healthy individuals (n = 50), with a P value of <0.05. The specificities for the B.4, B.6, B.10, B.M, C17, and 38-kDa recombinant antigens are 94, 88, 100, 96, 90, and 98%, respectively. All of the recombinant TB antigens showed substantial reactivity to active-TB specimens, both pulmonary and extrapulmonary. The B.4 antigen was reactive with 58.3% of the active-TB panel, compared to 37.5% detected by the known 38-kDa antigen. In addition, the B.4 antigen showed reactivity to 27.3% of the inactive-TB specimens compared to other TB antigens, which exhibited lower percentages of reactivity to these specimens (Table 3). The B.6 antigen was found to exhibit specific reactivity to pulmonary-TB specimens (46.4%) compared to extrapulmonary specimens (5%) (Fig. 6). All the other antigens were able to detect antibodies in both pulmonary- and extrapulmonary-TB specimens.