All patients were treated in accordance with the Helsinki Declaration on the participation of human subjects in medical research. The study protocol was approved by the Tongji University Ethics Committee (permit number: K08-018), and informed consent was obtained from each participant. Animal studies were approved by the Second Military Medical University Animal Care and Ethics Committee [approval ID: SYXK/2007-0003].

Based on the sequence of the H37Rv TB reference strain encoding the MPT64 gene (accession number: NC_000962), we designed forward (5'-CCCATATGCGCATCAAGATCTTCAT-3') and reverse (5'-CCAAGCTTGGCCAGCATCGAGTCG-3') primers, containing Nde I and Hind III restriction sites, respectively. PCR amplification with these primers was performed using H37Rv nucleic acid as template. Subsequently, the purified PCR product was inserted into the T vector using ligase. After digestion with restriction enzyme, pMD18-MPT64 was to pET21a using T4 ligase and the pET21a-MPT64 expression plasmid was used to transform BL21 (DE3) competent cells. Positive recombinant clones were identified by DNA sequencing.

A recombinant clone was inoculated overnight into the Luria Bertani medium at 37°C. The resultant bacteria were lysed with ultrasound and the medium was centrifuged. The sediment was resuspended in 10 % Triton-X 100 buffer, and MPT64 protein was purified on a nickel-agarose column. MPT64 protein was identified by sodium dodecyl sulfate polyacrylamide gel electrophoresis [¹²] [Figure 1.

Rabbits were immunized with MPT64 protein by subcutaneously injecting a suspension of 0.5 ml phosphate-buffered saline (PBS] containing 0.5 g/L MPT64 protein and 0.5 ml Freund's complete adjuvant into six sites on the back of each rabbit. On days 14, 24, 34, and 44, each rabbit was injected with the same dose of MPT64 protein and Freund's incomplete adjuvant. Rabbit serum was collected, and MPT64 antibody was purified on a CNBr-activated Sepharose 4B column, with the purified antibody identified by western blotting [Figure 2. The details of the method have been described previously [¹³]. Purified antibody was quantified by the Bradford method [¹⁴].

Using the primers 5’-GGGAGCTCAGAATAAACGCTCAA-3’ (forward) and 5’-Biotin-TTCGACATGAGGCCCGGATC-3’ (reverse), a 78-mer ssDNA library was constructed with the sequences, 5’ -GGGAGCTCAGAATAAACGCTCAA-AN35-CGACATGAGGCCCGGATC-3’, with the middle 35 sequences being random. The random ssDNA library and primers were synthesized commercially by Shanghai Sangon Company.

The ssDNA library was optimized by PCR amplification to obtain a double-stranded DNA (dsDNA) library. Furthermore, the ssDNA library was constructed by using asymmetric PCR. The amplified product was purified with a phenol-chloroform-isoamyl alcohol mixture (volume ratios, 25:24:1)]. ELISA plates coated with bovine serum were used to remove non-specific nucleotides from the ssDNA library and the purified ssDNA library added to microplates coated with antibody to MPT64. The plates were washed six times with SELEX buffer. SELEX eluate (20.0 mM Tris-HCl, 4.0 mM guanidinium isothiocyanate, 1.0 mM 1,4-dithiothreitol; pH 8.3) was added and allowed to incubate at 80°C for 10 min. The ssDNA was extracted with a phenol-chloroform-ethanol mixture. Using the extracted ssDNA as a template, asymmetric PCR was performed. Then, the next round of screening was performed. The parameters of the SELEX assay are shown in Table 1.

After 12 rounds of aptamer selection, the absorbance of the ssDNA reached its maximum value and a single electrophoretic stripe was displayed. This indicated that the purified ssDNA library with high affinity to the anti-MPT64 antibody was saturated. The PCR products were purified using the TIANgen Midi Purification Kit (TIANgen Biotech [Beijing] Co., Ltd.) and subcloned into a pMD 18-T vector with a TA cloning kit (TaKaRa, Dalian, China). The resultant bacteria were used to transform Escherichia coli DH5α strains, and individual bacterial clones were selectively sequenced by a commercial company (Sangon, Shanghai, China).

The affinity of the selected ssDNA aptamers was measured by ELISA. Polystyrene microplates were coated overnight at 4°C with 10 μg/well of anti-MPT64 antibody in 100 μl 0.1 M NaHCO₃ (pH 9.4), washed four times with washing buffer (PBS containing 0.05 % Tween 20, pH 7.4) and incubated for 1 hour at room temperature with 200 μl of blocking buffer. The microplates were washed once with washing buffer and 1.0 μg/well of biotin-labeled DNA aptamer in a binding buffer (SHCMK) containing 20.0 mM Hepes, pH 7.35, 1.0 mM CaCl₂, 120 mM KCl and 1.0 mM MgCl₂ was added. The aptamers and MPT64 antibody were allowed to react at 37°C for 40 min. The plates were washed six times with washing buffer (SHCMK + 0.05 % Tween 20) and incubated at 37°C for 30 min with 100 μl/well of streptavidin-horseradish peroxidase (Sigma, USA) diluted 1:1000 in PBS. Finally, the plate was washed six times with PBST and 100 μl of 1.0 mM tetramethylbenzidine in citrate buffer (0.1 M, pH 4.25), with 2.0 mM H₂O₂ added as substrate in a ratio of 1:20. The enzymatic reaction was stopped 5 min later by the addition of 50 μl 1.0 M H₂SO₄ and optical density (OD) of 450 nm was measured spectrophotometrically.

We assessed the secondary structures of the aptamers using sequencing results and the DNAMAN version 4.0 (Lynnon Corporation, Quebec, Canada). The aptamer with highest affinity to the anti-MPT64 antibody was regarded as the capture and detection aptamer by ELISA. Each microplate well was coated with 0.1 μg of the capture aptamer, and the 5 ' end of the detection aptamer was labeled with biotin. Based on the detection of purified anti-MPT64 antibody at different dilution ratios, the minimum limit of detection and the linear range of this ELISA method were determined.

We obtained 328 serum samples, 160 from patients with definite PTB and 168 from non-tuberculous controls including 78 healthy volunteers and 90 non-tuberculous patients. All participants were confirmed as being serologically HIV-negative. Of the 90 non-tuberculous patients, 35 had pneumonia, 40 had lung cancer, and 15 had lung abscesses and other conditions. Lowenstein-Jensen (L-J) culture was performed on all of these subjects to rule out TB.

The median ages of the PTB (31 years; range, 12–73 years; IQR [interquartile range], 17–45 years) and non-tuberculous controls (28 years; range, 14–69 years; IQR, 18–42 years) did not differ significantly P > 0.05). Patients were diagnosed with PTB by physicians based on the criteria of the Chinese Anti-tuberculosis Association, mainly including continuous fever and cough for more than 3 weeks, abnormal chest X-ray, weight loss and the demonstration of TB in sputum by L-J culture and standard biochemical identification tests. Sputum smear examinations for PTB and non-tuberculous patients were performed using the Ziehl-Neelsen method at the same time as L-J culture inoculation.

The sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR) and area under the curve (AUC), each with 95 % confidence interval (CI), were calculated. All statistical analyses were performed using Stata version 9 (Statacorp, Texas, USA).

The OD values of ssDNA aptamers against MPT64 antibody (reflecting affinity) ranged from 0.55 to 1.62 [Table 2]. We elected MPT64-A1, which had a higher affinity (OD value: 1.62) than did the capture aptamer for ELISA. The sequences of aptamer included forward fixed sequences (5’- GGGAGCTCAGAATAAACGCTCAAA-3’), random sequences (5’- AACGCTCAAGAGGCCCGGATC-3’) and reverse fixed sequences (5’-TTCGACATGAGGCCCGGATC-3’). The DNAMAN package predicted that the secondary structure of MPT64-A1 is a stem-loop structure, similar to a large pocket [Figure 3].

Based on the detection of the purified MPT64 antibody at different dilution ratios, we found that the limit of detection of this method was 2.5 mg/L and the linear range varied from 10 mg/L to 800 mg/L [Figure 4].

In diagnosing PTB, this method had a sensitivity, specificity, PLR, NLR, and AUC, with 95 % CI, of 64.4 % (56.7 %–71.4 %), 99.4 % (96.7 %–99.9 %), 108.2 (15.3–765.9), 0.35 (0.291–0.442), and 0.819 (0.770–0.868), respectively [Table 3]. Its sensitivity was similar for sputum smear positive (73/112, 65.2 %) and negative (30/48, 62.5 %) individuals [Table 4].

PLR: positive likelihood ratio. NLR: negative likelihood ratio. CI: confidence interval. PTB: pulmonary tuberculosis. AUC: area under the curve.