Journal of Microbiological Methods 115 (2015) 104–111

Contents lists available at ScienceDirect

Journal of Microbiological Methods journal homepage: www.elsevier.com/locate/jmicmeth

Fusion of selected regions of mycobacterial antigens for enhancing sensitivity in serodiagnosis of tuberculosis Madeeha Afzal a, Sana Khurshid a, Ruqyya Khalid a, Rehan Zafar Paracha b, Imran H. Khan c, M. Waheed Akhtar a,⁎ a b c

School of Biological Sciences, University of the Punjab, Lahore 54590, Pakistan Atta-Ur-Rahman School of Applied Biosciences, National University of Sciences and Technology, Islamabad 44000, Pakistan Department of Pathology and Laboratory Medicine, University of California, Davis 95616, USA

a r t i c l e

i n f o

Article history: Received 29 April 2015 Received in revised form 4 June 2015 Accepted 6 June 2015 Available online 9 June 2015 Keywords: Epitopes Fusion antigens Serodiagnosis HSP FbpC1 PstS1

a b s t r a c t Serodiagnosis of tuberculosis requires detection of antibodies against multiple antigens of Mycobacterium tuberculosis, because antibody profiles differ among the patients. Using fusion proteins with epitopes from two or more antigens would facilitate in the detection of multiple antibodies. Fusion constructs tn1FbpC1-tnPstS1 and tn2FbpC1-tnPstS1 were produced by linking truncated regions of variable lengths from FbpC1 to the N-terminus of the truncated PstS1. Similarly a truncated fragment of HSP was linked to the N-terminus of a truncated fragment from FbpC1 to produce tnHSP-tn1FbpC1. ELISA analysis of the plasma samples of TB patients against tn2FbpC1-tnPstS1 showed 72.2% sensitivity which is nearly the same as the expected combined value for the two individual antigens. However, the sensitivity of tn1FbpC1-tnPstS1 was lowered to 60%. tnHSPtn1FbpC1 showed 67.7% sensitivity which is slightly less than the expected combined value for the two individual antigens, but still significantly higher than that of each of the individual antigen. Data for secondary structure analysis by CD spectrometry was in reasonable agreement with the X-ray crystallographic data of the native proteins and the predicted structure of the fusion proteins. Comparative molecular modeling suggests that the epitopes of the constituent proteins are better exposed in tn2FbpC1-tnPstS1 as compared to those in tn1FbpC1tnPstS1. Therefore, removal of the N-terminal non-epitopic region of FbpC1 from 34–96 amino acids seems to have unmasked at least some of the epitopes, resulting in greater sensitivity. The high level of sensitivity of tn2FbpC1-tnPstS1 and tnHSP-tn1FbpC1, not reported before, shows that these fusion proteins have great potential for use in serodiagnosis of tuberculosis. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The antibody response against Mycobacterium tuberculosis antigens is variable in tuberculosis (TB) patients. It was realized quite early that a combination of antigens, rather than a single antigen would be required to develop satisfactory serodiagnosis (Hewitt et al., 1982; Kaplan and Chase, 1980). First generation immunoassays of TB had low specificity because they were based on crude antigenic materials of ill-defined composition e.g., purified protein derivatives (PPD) from M. tuberculosis cultures (Straus and Wu, 1980) or BCG sonicate (Wang et al., 1989), which often comprised antigens common to all mycobacteria and sometimes even other bacterial genera, such as

⁎ Corresponding author at: School of Biological Sciences, University of the Punjab, Lahore 54590, Pakistan. E-mail addresses: [email protected] (M. Afzal), [email protected] (S. Khurshid), [email protected] (R. Khalid), [email protected] (R.Z. Paracha), [email protected] (I.H. Khan), [email protected] (M.W. Akhtar).

http://dx.doi.org/10.1016/j.mimet.2015.06.003 0167-7012/© 2015 Elsevier B.V. All rights reserved.

Nocardia and Corynebacterium. Successful production of monoclonal antibodies (Coates et al., 1981; Morris and Ivanyi, 1985) and the development of recombinant DNA systems for production of M. tuberculosis antigens in Escherichia coli (Young et al., 1985) paved way for identification and utilization of purified antigens specific for MTB complex (Daniel and Debanne, 1987) leading to improved specificity. Whole genome sequencing of M. tuberculosis led to the identification of several ORFs encoding antigenic proteins (Cole et al., 1998) and during the past two decades numerous antigens have been evaluated for their serodiagnostic potential (Abebe et al., 2007; Weldingh et al., 2005). In a policy statement (WHO, 2011), regarding 19 commercially available serodiagnostic tests, WHO strongly recommended that they should not be used for TB diagnosis but encouraged further research to develop new tests with improved accuracy. Recent advances in epitope-serology can be helpful in improving diagnostic accuracy. Epitopes can be predicted or mapped by use of several algorithms (Scarabelli et al., 2010; Singh et al., 2013). The development of antigen microarray technology comprising of overlapping peptides spanning

M. Afzal et al. / Journal of Microbiological Methods 115 (2015) 104–111

the entire antigenic sequence has enabled the simultaneous measurement of antibody reactivity to thousands of peptides, and hence easy identification of B-cell epitopes of diagnostic value (Nahtman et al., 2007). It has been widely demonstrated that improvements in sensitivity of serodiagnostic tests can be achieved by using either a panel of antigens or using fusion molecules containing several antigens (Gennaro, 2000; Tong et al., 2005). Genes encoding immunodominant antigens have often been tandemly linked to encode a single polyprotein (Hoff et al., 2007; Houghton et al., 2002; Khurshid et al., 2014). Fusion proteins consisting of epitopes from two or more antigens are likely to offer a cheaper and more reliable serodiagnosis. Antibody based serological assays are attractive for resource limited countries for being relatively simple and inexpensive. Many serum samples can be tested in parallel and the process can be completely automated. They also offer a chance to detect childhood TB, latent TB and sputum smear negative cases which are difficult to be diagnosed with other techniques. PstS1, previously known as phoS1 or phoS, is one of the earliest known immunodominant antigens (Davidow et al., 2005; Espitia et al., 1989). It is specific only to the cavitary TB patients (Samanich et al., 2001; Sartain et al., 2006). Several overlapping epitopes of PstS1 have been analyzed (Baassi et al., 2009; Gaseitsiwe et al., 2008; Harris et al., 1996; Jackett et al., 1988; Landowski et al., 2001; Lopez-Vidal et al., 2004) and epitope-specific antibody detection correlates well with antibody levels to the purified antigen (Bothamley, 2014). The immunodominance of fibronectin binding protein FbpC1, also known as the MPT51, has also been widely investigated, and it is reported to elicit antibody responses during early and advanced stages of TB in both HIV-negative and HIV-positive patients (Achkar et al., 2006; Bethunaickan et al., 2007; Samanich et al., 2001). Through the use of peptide microarray comprising of 15 mer peptides, with 5 aa overlaps, spanning entire FbpC1 sequence, two B-cell epitopes have been identified (Nahtman et al., 2007). So a fusion molecule comprising of the epitopes from FbpC1 and PstS1 proteins, could diagnose both cavitary and non-cavitary TB patients, during early and advanced stages. The heat shock protein HSP, also known as the hrpA or acr2, is also an immunodominant antigen (Zhang et al., 2009) and a B-cell epitope sequence has been identified through peptide microarray technique (Gaseitsiwe et al., 2008). We have shown previously that the truncation of PstS1, by removing 96 and 14 amino acid residues from the N- and C-terminals respectively, improves its diagnostic efficiency (Khurshid et al., 2013). This study describes the sensitivity and specificity of the novel fusion proteins constructed from the truncated versions of PstS1, FbpC1 and HSP for diagnosis of active TB patients.

105

2. Materials and methods 2.1. Molecular cloning of recombinant proteins Immunodominant B cell epitope sequences in HSP, FbpC1 and PstS1 were identified from immune epitope database http://www. immuneepitope.org (Vita et al., 2010). One epitope from HSP, two epitopes from FbpC1 and 21 overlapping epitopes from PstS1 were selected for constructing fusion proteins. Relative position of the epitopes on HSP, FbpC1 and PstS1 is shown in Fig. 1. tn2FbpC1-tnPstS1 was designed to contain an FbpC1 fragment of 97–111 amino acid residues comprising of two epitopes, and PstS1 fragment of 97–360 amino acid residues comprising of 21 epitopes. Similarly tn1FbpC1-tnPstS1 was designed to contain the same PstS1 fragment, but a longer fragment of FbpC1 i.e., of 34–111 amino acid residues, comprising of an N-terminal flanking region in addition to the two epitopes. tnHSP-tn1FbpC1 was designed to contain 1–99 amino acid residues of HSP, comprising of an epitope and its Nterminal flanking region and 34–111 amino acid residues of FbpC1. PCR was done to amplify 318 bp fragment of Rv3803c (which encodes tn1FbpC1 fragment i.e., 34–111 amino acid residues of FbpC1) using the forward primer F1 and the reverse primer R1. Similarly, 126 bp fragment of Rv3803c (which encodes tn2FbpC1 fragment i.e., 97–111 amino acid residues of FbpC1) was amplified using the forward primer F2 and the reverse primer R1. PCR was also done to amplify 792 bp fragment of Rv0934 (which encodes tnPstS1 i.e., 97–360 amino acid residues of PstS1) as described previously (Khurshid et al., 2013). This fragment was ligated into EcoRI and HindIII sites of pET28a(+). Thereafter, the DNA fragments encoding tn1FbpC1 and tn2FbpC1 were ligated into NdeI and EcoRI sites of pET28-tnPstS1 separately. PCR was done to amplify 294 bp fragment of Rv0251c (which encodes tnHSP fragment i.e., 1–99 amino acid residues of HSP), using the forward primer F5 and the reverse primer R4. The same 318 bp fragment of Rv3803c was also amplified using the forward primer F3 and the reverse primer R1. These two fragments were ligated together first, and then the chimeric DNA was ligated into NdeI and EcoRI sites in pET28a(+). Full length Rv3803c, encoding FbpC1, was amplified using forward primer F1 and reverse primer R2. Full length Rv0251c, encoding HSP, was amplified using the primers F5 and R5. Full length Rv0934, encoding PstS1 was amplified as described previously (Khurshid et al., 2013). All the full length genes were initially cloned in pTZ57R/T and then subcloned in the vector pET28a(+). All the constructs were analyzed by colony PCR, restriction fragment analysis and sequencing using the Beckman Coulter SEQ800 Genetic analyzer. Primer sequences are shown in Table 1 (supplementary data).

Fig. 1. Diagrammatic representation of the scheme for construction of multi-epitope fusion constructs. F1, F2, F3, F4, and F5 are the forward primers, whereas R1, R2, R3, R4 and R5 are the reverse primers. Truncated stretches of FbpC1 i.e., tn1FbpC1 and tn2FbpC1 were fused with truncated PstS1 i.e., tnPstS1 to produce tn1FbpC1-tnPstS1 and tn2FbpC1-tnPstS1, respectively. Truncated HSP i.e., tnHSP was fused with tn1FbpC1 to produce the fusion tnHSP-tn1FbpC1.

106

M. Afzal et al. / Journal of Microbiological Methods 115 (2015) 104–111

2.2. Expression and purification of recombinant proteins E. coli BL21-CodonPlus (DE3)-RIPL competent cells were transformed with the recombinant DNAs prepared as above. Expression of the recombinant fusion proteins was carried out in LB medium supplemented with kanamycin at a final concentration of 50 μg/ml and induced with 0.5 mM IPTG at 37 °C for 4 h. FbpC1 and HSP were expressed in E. coli grown in Terrific broth at 18 °C for 16 h and 37 °C for 4 h, respectively. Proteins were also expressed by auto-induction using M9NG medium (Sadaf et al., 2007). The E. coli cells were resuspended in 20 mM Tris–Cl (pH 8.0) containing 0.5 M NaCl and 10 mM imidazole and lysed ultrasonically. The proteins recovered in the soluble fraction of E. coli cell lysates were purified using Ni-affinity chromatography. The proteins expressed as inclusion bodies were harvested after cell-lysis by sonication and washed with 0.5% Triton X-100. These were then solubilized in 20 mM Tris–Cl (pH 8.0) containing 0.5 M NaCl and 8 M urea and were purified in the denatured form using Ni-affinity chromatography, followed by refolding in a dialysis bag by sequential removal of urea. The proteins were further purified by anion exchange FPLC using HiTrap Q FF 5 ml column (GE Healthcare). PstS1 and tnPstS1 were purified using single step purification and refolding method as described previously (Khurshid et al., 2013). 2.3. Reactivity with rabbit polyclonal antibodies Polyclonal anti-PstS1, anti-FbpC1, and anti-HSP antisera were produced by immunizing rabbits with the purified antigens. Institutional animal experimentation guidelines and EU Directive 2010/63/EU for animal experiments were followed. All the procedures were the same as reported previously (Khurshid et al., 2013). The antigens tnPstS1 and FbpC1 and the fusion molecules i.e., tn1FbpC1-tnPstS1 and tn2FbpC1-tnPstS1 were studied for immunoreactivity against the rabbit polyclonal anti-FbpC1 and anti-PstS1 using both western blotting and ELISA methods. Similarly FbpC1, HSP and the fusion molecule tnHSP-tn1FbpC1 were also analyzed for immunoreactivity against the rabbit polyclonal anti-FbpC1 and anti-HSP. The procedures followed for these analyses were generally the same as described previously (Khurshid et al., 2013). The pre-immune sera were used as controls in ELISA. 2.4. ELISA with plasma samples Ethical approval for the study was obtained from Ethical Committee of School of Biological Sciences, University of the Punjab, Lahore, Pakistan (SBS/987/11). Plasma samples from 180 patients, diagnosed with pulmonary TB and confirmed by growth of M. tuberculosis on LJ medium, were collected from Gulab Devi Hospital, Lahore, which specializes in chest diseases. Informed consent was obtained from all individual participants included in the study, and a brief history of disease was recorded. All the samples were collected before starting treatment for TB. The control group included samples from 100 healthy individuals. The antigens PstS1, FbpC1, HSP and the multi-epitope fusion molecules i.e., tn1FbpC1-tnPstS1, tn2FbpC1-tnPstS1 and tnHSP-tn1FbpC1 were used to detect antibodies in the plasma samples of TB patients. The optimum coating concentration was found to be 100 ng/100 μl/well for HSP, and 200 ng/100 μl/well for all other antigens. Also, horseradish peroxidase-conjugated anti-human antibody (MP biomedicals 55221) was used at a working concentration of 1:5000 for 1 h at 37 °C. The rest of the procedures were the same as described previously (Khurshid et al., 2013). 2.5. Statistical analysis The mean absorbance and standard deviation of the healthy control samples were calculated. Mean OD450/630 + 2.576 SD was considered as

a cut-off value. The samples with an absorbance higher than the cutoff were scored as positive. Since all the six antigens had different cutoff values, the OD450/630 values for each test antigen were normalized by dividing with the respective cutoff value (a value greater than 1 indicated positive results). Sensitivity was determined by dividing the number of positive cases with the total number of TB patients. Specificity was determined by dividing the number of controls found negative by the total number of healthy controls. Combined sensitivity for two individual antigens was calculated by adding the number of samples positive for the two individual antigens minus the number of samples positive for both the antigens. Statistical analysis was done using GraphPad Prism 6 (Graph-Pad Software Inc., San Diego, CA). The results from the TB patients and the healthy controls were compared using the NonParametric Mann–Whitney Test (Henry B. Mann and Whitney, 1947). Area under curve (AUC) for each antigen was determined from receiver operating characteristics (ROC) curves. 2.6. Circular dichroism spectrometry Circular dichroism (CD) data were collected on a Chirascan Plus CD spectrophotometer (Applied Photophysics) equipped with a peltier thermal-controlled cuvette holder. The protein solution containing 0.4 mg/ml in 5 mM Tris–Cl (pH 8.0) was scanned over wavelength 190 to 260 nm at 20 °C, using a quartz cell of 1 mm path length. For tnHSP-tn1FbpC1, the spectra were obtained with 0.14 mg/ml solution using a cell of 0.5 mm path length. Each wavelength spectrum was the result of averaging two consecutive scans with a bandwidth of 1.0 nm. The wavelength spectra were refined by subtracting a blank spectrum with buffer only. The secondary structure content of protein was calculated using the CD spectrum deconvolution software CDNN (Bohm et al., 1992). 2.7. Molecular modeling Comparative modeling of tn2FbpC1-tnPstS1 and tn1FbpC1-tnPstS1 was performed using the crystal structures of FbpC1 PDB: 1R88 resolved at 1.71 Å (Wilson et al., 2004) and PstS1 PDB: 1PC3 resolved at 2.16 Å (Vyas et al., 2003). Since the X-ray structure of HSP is not known, the sequence was submitted to QUARK ab initio modeling server (Xu and Zhang, 2012) and the resulting model was energy minimized and used as a template for comparative modeling of tnHSP-tn1FbpC1. Structural alignment between the query and template sequences was performed using 3D-Coffee (Poirot et al., 2004). The three dimensional structures of fusion proteins were generated as described previously (Sajjad et al., 2012). Briefly, models were generated using MODELLER (Eswar et al., 2006), and were clustered by NMRCLUST (Kelley et al., 1996). Models from each cluster were analyzed by Ramachandran plot (Ramachandran et al., 1963), QMEAN (Benkert et al., 2008), ERRAT (Colovos and Yeates, 1993), PROSA (Wiederstein and Sippl, 2007) and

Fig. 2. SDS PAGE gel stained with Coomassie brilliant blue. M, molecular weight markers; lanes 1, 3 and 5, proteins of E. coli cells expressing tn1FbpC1-tnPstS1, tn2FbpC1-tnPstS1 and tnHSP-tn1FbpC1 respectively; lanes 2, 4 and 6, purified tn1FbpC1-tnPstS1, tn2FbpC1-tnPstS1 and tnHSP-tn1FbpC1, respectively.

M. Afzal et al. / Journal of Microbiological Methods 115 (2015) 104–111

107

Table 1 Summary of expression and purification of native and fusion antigens of M. tuberculosis. Antigenic proteins

FbpC1

PstS1a

HSP

tn1FbpC1-tnPstS1

tn2FbpC1-tnPstS1

tnHSP-tn1FbpC1

Expression level Yield (mg/l/OD600) Purity (%) Recovery (%) Recovered protein (mg/l/OD600)

39% 70 N90 25 17.5

15% 27 N90 28.9 7.8

36% 64.8 N90 25 16.2

35% 63 80 30 18.9

30% 54 84 33 17.8

32% 57.6 N90 29 16.7

a

As reported previously (Khurshid et al., 2013).

PROCHECK (Laskowski et al., 1993), and a final model was selected for each fusion protein. In order to observe the flexible nature of tn2FbpC1 fragment based on RMSD between corresponding residues, the tn2FbpC1-tnPstS1 models were superimposed and structurally aligned using the CHIMERA software (Pettersen et al., 2004). Similarly, the models of tn1FbpC1-tnPstS1 were superimposed using the same region of tnPstS1 to observe flexibility of tn1FbpC1 fragment and RMSDs between the corresponding residues. Differential antigen–antibody contact regions of the fusion antigens were analyzed through solvent accessibility analysis (Lebeda and Olson, 1997). Possible surface accessible residues of FbpC1 protein fragments in three fusion proteins were predicted using CPORT facility (de Vries and Bonvin, 2011) which uses a combination of six interface prediction algorithms including PINUP (Liang et al., 2006), PIER (Kufareva et al., 2007), WHISCY (de Vries et al., 2006), ProMate (Neuvirth et al., 2004), cons-PPISP (Chen and Zhou, 2005) and SPPIDER (Porollo and Meller, 2007) for determination of the interface residues. 3. Results and discussion 3.1. Expression, refolding and purification Upon induction of E. coli cells harboring the recombinant plasmids, the fusion proteins tn1FbpC1-tnPstS1, tn2FbpC1-tnPstS1 and tnHSPtn1FbpC1 were expressed and analyzed by SDS-PAGE as shown in Fig. 2. The levels of expression as determined densitometrically using Syngene GeneTools were found to be 35%, 30% and 32% of the total cell proteins, respectively. tn1FbpC1-tnPstS1 and tn2FbpC1-tnPstS1 were expressed in the cells as insoluble inclusion bodies, whereas tnHSP-tn1FbpC1 was recovered in the soluble fraction of E. coli cell lysate. The fusion molecules expressed as inclusion bodies were solubilized and refolded as described above. All the fusion antigens were eluted from the Ni-Sepharose column, with 250 mM Imidazole. After purification, percentage recoveries of tn1FbpC1-tnPstS1, tn2FbpC1tnPstS1 and tnHSP-tn1FbpC1 were 30%, 33%, and 29%, respectively. Their yields were 18.9, 17.8 and 16.7 mg per liter per OD600, respectively as shown in Table 1. 3.2. Reactivity with rabbit polyclonal antisera As expected, the polyclonal anti-PstS1 antisera reacted with tnPstS1 protein and the fusion antigens tn2FbpC1-tnPstS1 and tn1FbpC1-

tnPstS1, but it did not react with FbpC1 (Fig. 1b (supplementary data)). Similarly polyclonal anti-FbpC1 antisera reacted with FbpC1 and the fusion antigens tn2FbpC1-tnPstS1 and tn1FbpC1-tnPstS1 but did not react with the tnPstS1 (Fig. 1c (supplementary data)). Similarly tnHSP-tn1FbpC1 reacted well with polyclonal anti-FbpC1 and anti-HSP (Fig. 1e and f (supplementary data)) respectively. Same results were observed in ELISA (Fig. 1g (supplementary data)). 3.3. ELISA analysis of plasma samples Humoral immune response of TB patients against all the antigenic proteins is shown in Table 2 and their scatter plot is shown in Fig. 3. Out of the 180 TB patient plasma samples analyzed, 66 were positive for PstS1 and 108 were positive for FbpC1. However, out of these, 43 samples showed the presence of antibodies against both the antigens. Therefore the expected number of samples testing positive for a fusion molecule between PstS1 and FbpC1 showing maximum sensitivity should be 131 (sensitivity 72.77%). Whereas tn2FbpC1tnPstS1 showed a positive result for 130 samples (sensitivity 72.22%), tn1FbpC1-tnPstS1 was able to detect the antibodies in only 108 samples (sensitivity 60%). It appears that the fragment consisting of 34–96 amino acid residues of FbpC1, which contains no epitopes, if present as in tn1FbpC1-tnPstS1 results in masking of some of the epitopes, and hence lowering the sensitivity. Removal of this fragment restores nearly the same sensitivity as expected when compared with the total value for the two individual antigens. Similarly, out of 180 TB patients' plasma samples, 62 were positive for HSP and 108 for FbpC1. Out of these, 35 samples were positive for both the individual proteins. Therefore the maximum expected number of samples testing positive for the fusion molecule should be 135 (sensitivity 75%). Although tnHSP-tn1FbpC1 was able to detect only 122 samples (sensitivity 67.77%), this sensitivity level is significantly higher than that of each of the individual antigens. High sensitivity of tnHSP-tn1FbpC1 would facilitate in diagnosing TB patients even when they were found smear negative in the preliminary screening. Non-Parametric Mann–Whitney Test indicated that plasma from TB patients had a statistically significant (P value b0.0001) response against all the native and multiepitope fusion proteins, when compared with that of the healthy controls. The area under ROC curve was used to determine the overall ability of the test to discriminate between patients and healthy individuals. AUC represents the probability that a randomly selected plasma sample

Table 2 Sensitivity and specificity of the recombinant antigens using ELISA method. Antigens

FbpC1 PstS1 HSP tn1FbpC1-tnPstS1 tn2FbpC1-tnPstS1 tnHSP-tn1FbpC1 FbpC1 + PstS1 FbpC1 + HSP

Total (n = 180)

Smear positive (n = 94)

Smear negative (n = 86)

% specificity

No. of samples positive

% sensitivity

No. of samples positive

% sensitivity

No. of samples positive

% sensitivity

108 66 62 108 130 122 131 135

60 36.66 34.44 60 72.22 67.77 72.77 75

67 33 20 58 75 60 74 68

71.27 35.10 21.27 61.70 79.78 63.82 78.72 72.34

41 33 42 50 55 62 57 67

47.67 38.37 48.83 58.13 63.95 72.09 66.27 77.90

100 100 98 99 100 100 100 98

108

M. Afzal et al. / Journal of Microbiological Methods 115 (2015) 104–111

3.4. CD spectrometry CD spectra of the three fusion proteins and their native counterparts are shown in Fig. 2 (supplementary data). The spectra of tn1FbpC1tnPstS1 and tn2FbpC1-tnPstS1 showed the characteristic positive band at 193 nm and negative bands at 222 nm and 208 nm (Holzwarth and Doty, 1965) indicating predominant α-helices. Proteins with welldefined antiparallel β-pleated sheets also have negative bands at 218 nm and positive bands at 195 nm (Greenfield and Fasman, 1969). Data for the secondary structures of proteins as obtained by CD was in reasonable agreement with the X-ray crystallographic data of the native proteins and the predicted structure of the fusion proteins as shown in Table 2 (supplementary data). 3.5. Molecular modeling Fig. 3. ELISA analysis for antibody levels determined against native and fusion antigens of M. tuberculosis. GraphPad Prism 6 was employed to draw scatter dot plots of the normalized OD450/630 values for 100 healthy controls (HC) and 180 TB patients (TB) plasma samples. Mean OD450/630 for HC samples + 2.576 standard deviation was considered as a cutoff value. Normalized OD450/630 values greater than 1 indicate positive samples.

from TB patient will have a higher test result than a randomly selected control. As a general rule, an AUC value of 0.5 indicates that the diagnostic test cannot discriminate between the diseased and healthy individuals at all, whereas AUC value of 1 indicates full discriminatory power. Values closer to 1 indicate better diagnostic capability. A test with perfect discrimination has a ROC curve that passes through the upper left corner (100% sensitivity, 100% specificity). Therefore the closer the ROC curve is to the upper left corner, the higher the overall accuracy of the test (Zweig and Campbell, 1993). The ROC curves for all the antigens are shown in Fig. 4. The fusion protein tn2FbpC1-tnPstS1 had the highest AUC value 0.9585 with a standard error 0.01, and a P value 0.0001. The greater AUC of tn2FbpC1-tnPstS1 as compared to that of the individual antigens and the other fusion constructs shows its better diagnostic capability.

MODELLER predicted a wide conformational space of tn2FbpC1 in the models generated for the fusion protein tn2FbpC1-tnPstS1. In contrast, the comparative models of tn1FbpC1-tnPstS1 showed relatively less conformational flexibility of tn1FbpC1 against tnPstS1. A selection of 17 representative models of tn2FbpC1-tnPstS1 from each cluster provided by NMRCLUST showed that these models have an RMSD difference of ≈ 15 Å between positions of corresponding residues in the tn2FbpC1 portion of the model. Similarly, a selection of 10 representative models of tn1FbpC1-tnPstS1 from each cluster showed an RMSD difference of ≈3 Å between positions of corresponding residues in the tn1FbpC1 portion of the model. In the case of tn2FbpC1-tnPstS1 ERRAT scores of representative models ranged from 85–72 with a major disturbance area of loop between tn2FbpC1 and tnPstS1 related residues. Loop refinement and proper rotamer selection of the selected model of tn2FbpC1-tnPstS1 increased the ERRAT score to 91. In the case of tn1FbpC1-tnPstS1, the ERRAT score was in the range of 85–60, and loop refinement of the selected model increased the ERRAT score to 95. All of the residues in the models of the fusion proteins were in the allowed regions of the Ramachandran plot. QMEAN, PROSA and

Fig. 4. ROC curves for ELISA against native and fusion antigens of M. tuberculosis, constructed using GraphPad Prism 6. The true-positive rate (sensitivity) and false-positive rate (1-specificity) of each antigen, as determined using normalized OD450/630 values of TB patients and healthy controls, were plotted. The AUC values were obtained with a P value ˂0.0001.

M. Afzal et al. / Journal of Microbiological Methods 115 (2015) 104–111

PROCHECK values indicated that these selected models were as accurate as NMR derived structures. Comparative molecular models of tn2FbpC1-tnPstS1, tn1FbpC1-tnPstS1 and tnHSP-tn1FbpC1 are shown in Fig. 5a, b and c, respectively. RMSD between the corresponding residues of tn2FbpC1 region in the comparative models of tn2FbpC1-tnPstS1 was found within a range of 14.494 Å whereas in the case of tn1FbpC1 region in tn1FbpC1-tnPstS1 models, this value was found within a range of 11.148 Å. FbpC1 epitopes had the highest solvent accessibility in tn2FbpC1-tnPstS1 as shown in Fig. 5aI. tn1FbpC1 fragment had a lesser solvent accessibility in tn1FbpC1-tnPstS1 as compared to that in tnHSPtn1FbpC1 as shown in Fig. 5bI and cI, respectively. Antibody molecules may use conformational selection to recognize preexisting sub-states of antigen conformational ensembles and the binding to the antigens is stabilized by small local rearrangements of the interface residues (Boehr et al., 2009). Both the paratope of antibody and the epitope of antigen undergo slight conformational changes in each other's presence. The epitope must therefore be readily accessible to the paratopes and it must be somewhat flexible to undergo the conformational changes required for interaction with the antibody molecule. Comparative molecular modeling of tn2FbpC1-tnPstS1 and tn1FbpC1-tnPstS1 showed that the folding pattern of truncated PstS1 was highly similar in both the molecules, but

109

the FbpC1 epitopes were more accessible in tn2FbpC1-tnPstS1 molecule as compared to the tn1FbpC1-tnPstS1 molecule. A higher degree of conformational flexibility in tn2FbpC1-tnPstS1 seemed to be due to the long loop present between tn2FbpC1 and tnPstS1 portions of the models, whereas reduced conformational flexibility in tn1FbpC1tnPstS1 is perhaps due to the interference from the N-terminal flanking region. The regions of an antigen with higher solvent accessible surface area often account for antibody contact regions (Lebeda and Olson, 1997). The data obtained for solvent accessible surface area analysis as well as the high RMSD value support the view that tn2FbpC1-tnPstS1 protein has the two components arranged in a more flexible manner and the epitopes are better available for interaction with the antibodies, which seem to be responsible for its higher sensitivity. Reliability and accuracy of serodiagnostic assay can be improved by including other M. tuberculosis antigens in the protocol. The use of fusion molecules constructed from epitopes of multiple antigens would facilitate in achieving this objective more efficiently and economically. The real value of these fusion antigens also needs to be ascertained by cross-sectional studies in clinical settings where the TB suspects include patients with a variety of respiratory diseases (Achkar et al., 2010). This serodiagnostic assay must also be evaluated further to determine its diagnostic capacity for extrapulmonary TB and pediatric TB. The assay can

Fig. 5. Comparative molecular models and interface prediction of fusion antigens of M. tuberculosis. Models of (a) tn2FbpC1-tnPstS1, (b) tn1FbpC1-tnPstS1 and (c) tnHSP-tn1FbpC1. The tn2FbpC1 and tn1FbpC1 fragments are shown in yellow color and the two epitopes are shown in black color. The tnPstS1 fragment is shown in dark green color. The tnHSP is shown in gray color and its epitope is shown in black color. Interface prediction (solvent assessable surface area) models of FbpC1 residues of (aI) tn2FbpC1-tnPstS1, (bI) tn1FbpC1-tnPstS1 and (cI) tnHSP-tn1FbpC1. The residues which can take part in any type of interaction are colored red; the residues which can support the interaction and surround the active residues are colored green, whereas non-accessible residues are colored blue. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

110

M. Afzal et al. / Journal of Microbiological Methods 115 (2015) 104–111

also be extended to high throughput analysis using multiplex microbead assays (Khan et al., 2008). 4. Conclusion The high level of sensitivity of tn2FbpC1-tnPstS1 and tnHSPtn1FbpC1 shows that these fusion proteins have good potential for use in serodiagnosis of tuberculosis. Using chimeras of multiple antigens with the epitopes available for interaction with the corresponding antibodies seems to have potential in developing more reliable and economical serodiagnostic procedures for tuberculosis. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.mimet.2015.06.003. Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments. Institutional animal experimentation guidelines and EU Directive 2010/63/EU for animal experiments were followed for the care and use of animals. Acknowledgments This work was supported by the Higher Education Commission, Pakistan (Grant No. 106-1926-BM6-144). We are thankful to Miss Aasia Khaliq for her assistance in the collection of human plasma samples and Mr. Muhammad Ali for his assistance in raising the antisera in rabbits. References Abebe, F., Holm-Hansen, C., Wiker, H.G., Bjune, G., 2007. Progress in serodiagnosis of Mycobacterium tuberculosis infection. Scand. J. Immunol. 66, 176–191. Achkar, J.M., Dong, Y., Holzman, R.S., Belisle, J., Kourbeti, I.S., Sherpa, T., Condos, R., Rom, W.N., Laal, S., 2006. Mycobacterium tuberculosis malate synthase- and MPT51-based serodiagnostic assay as an adjunct to rapid identification of pulmonary tuberculosis. Clin. Vaccine Immunol. 13, 1291–1293. Achkar, J.M., Jenny-Avital, E., Yu, X., Burger, S., Leibert, E., Bilder, P.W., Almo, S.C., Casadevall, A., Laal, S., 2010. Antibodies against immunodominant antigens of Mycobacterium tuberculosis in subjects with suspected tuberculosis in the United States compared by HIV status. Clin. Vaccine Immunol. 17, 384–392. Baassi, L., Sadki, K., Seghrouchni, F., Contini, S., Cherki, W., Nagelkerke, N., Benjouad, A., Saltini, C., Colizzi, V., El Aouad, R., Amicosante, M., 2009. Evaluation of a multiantigen test based on B-cell epitope peptides for the serodiagnosis of pulmonary tuberculosis. Int. J. Tuberc. Lung Dis. 13, 848–854. Benkert, P., Tosatto, S.C., Schomburg, D., 2008. QMEAN: a comprehensive scoring function for model quality assessment. Proteins 71, 261–277. Bethunaickan, R., Baulard, A.R., Locht, C., Raja, A., 2007. Antibody response in pulmonary tuberculosis against recombinant 27 kDa (MPT51, Rv3803c) protein of Mycobacterium tuberculosis. Scand. J. Infect. Dis. 39, 867–874. Boehr, D.D., Nussinov, R., Wright, P.E., 2009. The role of dynamic conformational ensembles in biomolecular recognition. Nat. Chem. Biol. 5, 789–796. Bohm, G., Muhr, R., Jaenicke, R., 1992. Quantitative analysis of protein far UV circular dichroism spectra by neural networks. Protein Eng. 5, 191–195. Bothamley, G.H., 2014. Epitope-specific antibody levels in tuberculosis: biomarkers of protection, disease, and response to treatment. Front. Immunol. 5, 243. Chen, H., Zhou, H.X., 2005. Prediction of interface residues in protein-protein complexes by a consensus neural network method: test against NMR data. Proteins 61, 21–35. Coates, A.R., Hewitt, J., Allen, B.W., Ivanyi, J., Mitchison, D.A., 1981. Antigenic diversity of Mycobacterium tuberculosis and Mycobacterium bovis detected by means of monoclonal antibodies. Lancet 2, 167–169. Cole, S.T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris, D., Gordon, S.V., Eiglmeier, K., Gas, S., Barry III, C.E., Tekaia, F., Badcock, K., Basham, D., Brown, D., Chillingworth, T., Connor, R., Davies, R., Devlin, K., Feltwell, T., Gentles, S., Hamlin, N., Holroyd, S., Hornsby, T., Jagels, K., Krogh, A., McLean, J., Moule, S., Murphy, L., Oliver, K., Osborne, J., Quail, M.A., Rajandream, M.A., Rogers, J., Rutter, S., Seeger, K., Skelton, J., Squares, R., Squares, S., Sulston, J.E., Taylor, K., Whitehead, S., Barrell, B.G., 1998. Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544. Colovos, C., Yeates, T.O., 1993. Verification of protein structures: patterns of nonbonded atomic interactions. Protein Sci. 2, 1511–1519. Daniel, T.M., Debanne, S.M., 1987. The serodiagnosis of tuberculosis and other mycobacterial diseases by enzyme-linked immunosorbent assay. Am. Rev. Respir. Dis. 135, 1137–1151.

Davidow, A., Kanaujia, G.V., Shi, L., Kaviar, J., Guo, X., Sung, N., Kaplan, G., Menzies, D., Gennaro, M.L., 2005. Antibody profiles characteristic of Mycobacterium tuberculosis infection state. Infect. Immun. 73, 6846–6851. de Vries, S.J., Bonvin, A.M., 2011. CPORT: a consensus interface predictor and its performance in prediction-driven docking with HADDOCK. PLoS One 6, e17695. de Vries, S.J., van Dijk, A.D., Bonvin, A.M., 2006. WHISCY: what information does surface conservation yield? Application to data-driven docking. Proteins 63, 479–489. Espitia, C., Cervera, I., Gonzalez, R., Mancilla, R., 1989. A 38-kD Mycobacterium tuberculosis antigen associated with infection. Its isolation and serologic evaluation. Clin. Exp. Immunol. 77, 373–377. Eswar, N., Webb, B., Marti-Renom, M.A., Madhusudhan, M.S., Eramian, D., Shen, M.Y., Pieper, U., Sali, A., 2006. Comparative protein structure modeling using Modeller. Curr. Protoc. Bioinformatics (Chapter 5, Unit 5.6). Gaseitsiwe, S., Valentini, D., Mahdavifar, S., Magalhaes, I., Hoft, D.F., Zerweck, J., Schutkowski, M., Andersson, J., Reilly, M., Maeurer, M.J., 2008. Pattern recognition in pulmonary tuberculosis defined by high content peptide microarray chip analysis representing 61 proteins from M. tuberculosis. PLoS One 3, e3840. Gennaro, M.L., 2000. Immunologic diagnosis of tuberculosis. Clin. Infect. Dis. 30 (Suppl. 3), S243–S246. Greenfield, N., Fasman, G.D., 1969. Computed circular dichroism spectra for the evaluation of protein conformation. Biochemistry 8, 4108–4116. Harris, D.P., Vordermeier, H.M., Arya, A., Bogdan, K., Moreno, C., Ivanyi, J., 1996. Immunogenicity of peptides for B cells is not impaired by overlapping T-cell epitope topology. Immunology 88, 348–354. Hewitt, J., Coates, A.R.M., Mitchison, D.A., Ivanyi, J., 1982. The use of murine monoclonalantibodies without purification of antigen in the serodiagnosis of tuberculosis. J. Immunol. Methods 55, 205–211. Hoff, S.T., Abebe, M., Ravn, P., Range, N., Malenganisho, W., Rodriques, D.S., Kallas, E.G., Soborg, C., Mark Doherty, T., Andersen, P., Weldingh, K., 2007. Evaluation of Mycobacterium tuberculosis-specific antibody responses in populations with different levels of exposure from Tanzania, Ethiopia, Brazil, and Denmark. Clin. Infect. Dis. 45, 575–582. Holzwarth, G., Doty, P., 1965. The ultraviolet circular dichroism of polypeptides. J. Am. Chem. Soc. 87, 218–228. Houghton, R.L., Lodes, M.J., Dillon, D.C., Reynolds, L.D., Day, C.H., McNeill, P.D., Hendrickson, R.C., Skeiky, Y.A., Sampaio, D.P., Badaro, R., Lyashchenko, K.P., Reed, S.G., 2002. Use of multiepitope polyproteins in serodiagnosis of active tuberculosis. Clin. Diagn. Lab. Immunol. 9, 883–891. Jackett, P.S., Bothamley, G.H., Batra, H.V., Mistry, A., Young, D.B., Ivanyi, J., 1988. Specificity of antibodies to immunodominant mycobacterial antigens in pulmonary tuberculosis. J. Clin. Microbiol. 26, 2313–2318. Kaplan, M.H., Chase, M.W., 1980. Antibodies to mycobacteria in human tuberculosis. II. Response to nine defined mycobacterial antigens with evidence for an antibody common to tuberculosis and lepromatous leprosy. J. Infect. Dis. 142, 835–843. Kelley, L.A., Gardner, S.P., Sutcliffe, M.J., 1996. An automated approach for clustering an ensemble of NMR-derived protein structures into conformationally related subfamilies. Protein Eng. 9, 1063–1065. Khan, I.H., Ravindran, R., Yee, J., Ziman, M., Lewinsohn, D.M., Gennaro, M.L., Flynn, J.L., Goulding, C.W., DeRiemer, K., Lerche, N.W., Luciw, P.A., 2008. Profiling antibodies to Mycobacterium tuberculosis by multiplex microbead suspension arrays for serodiagnosis of tuberculosis. Clin. Vaccine Immunol. 15, 433–438. Khurshid, S., Khalid, R., Afzal, M., Waheed Akhtar, M., 2013. Truncation of PstS1 antigen of Mycobacterium tuberculosis improves diagnostic efficiency. Tuberculosis (Edinb.) 93, 654–659. Khurshid, S., Afzal, M., Khalid, R., Khan, I.H., Akhtar, M.W., 2014. Improving sensitivity for serodiagnosis of tuberculosis using TB16.3-echA1 fusion protein. Tuberculosis (Edinb.) 94, 519–524. Kufareva, I., Budagyan, L., Raush, E., Totrov, M., Abagyan, R., 2007. PIER: protein interface recognition for structural proteomics. Proteins 67, 400–417. Landowski, C.P., Godfrey, H.P., Bentley-Hibbert, S.I., Liu, X., Huang, Z., Sepulveda, R., Huygen, K., Gennaro, M.L., Moy, F.H., Lesley, S.A., Haak-Frendscho, M., 2001. Combinatorial use of antibodies to secreted mycobacterial proteins in a host immune systemindependent test for tuberculosis. J. Clin. Microbiol. 39, 2418–2424. Laskowski, R.A., Macarthur, M.W., Moss, D.S., Thornton, J.M., 1993. Procheck — a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26, 283–291. Lebeda, F.J., Olson, M.A., 1997. Predicting differential antigen–antibody contact regions based on solvent accessibility. J. Protein Chem. 16, 607–618. Liang, S., Zhang, C., Liu, S., Zhou, Y., 2006. Protein binding site prediction using an empirical scoring function. Nucleic Acids Res. 34, 3698–3707. Lopez-Vidal, Y., de Leon-Rosales, S.P., Castanon-Arreola, M., Rangel-Frausto, M.S., Melendez-Herrada, E., Sada-Diaz, E., 2004. Response of IFN-gamma and IgG to ESAT-6 and 38 kDa recombinant proteins and their peptides from Mycobacterium tuberculosis in tuberculosis patients and asymptomatic household contacts may indicate possible early-stage infection in the latter. Arch. Med. Res. 35, 308–317. Mann, Henry B., Whitney, D.R., 1947. On a test of whether one of two random variables is stochastically larger than the other. Ann. Math. Stat. 18, 50–60. Morris, J.A., Ivanyi, J., 1985. Immunoassays of field isolates of Mycobacterium bovis and other mycobacteria by use of monoclonal antibodies. J. Med. Microbiol. 19, 367–373. Nahtman, T., Jernberg, A., Mahdavifar, S., Zerweck, J., Schutkowski, M., Maeurer, M., Reilly, M., 2007. Validation of peptide epitope microarray experiments and extraction of quality data. J. Immunol. Methods 328, 1–13. Neuvirth, H., Raz, R., Schreiber, G., 2004. ProMate: a structure based prediction program to identify the location of protein-protein binding sites. J. Mol. Biol. 338, 181–199. Pettersen, E.F., Goddard, T.D., Huang, C.C., Couch, G.S., Greenblatt, D.M., Meng, E.C., Ferrin, T.E., 2004. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605–1612.

M. Afzal et al. / Journal of Microbiological Methods 115 (2015) 104–111 Poirot, O., Suhre, K., Abergel, C., O'Toole, E., Notredame, C., 2004. 3DCoffee@igs: a web server for combining sequences and structures into a multiple sequence alignment. Nucleic Acids Res. 32, W37–W40. Porollo, A., Meller, J., 2007. Prediction-based fingerprints of protein–protein interactions. Proteins 66, 630–645. Ramachandran, G.N., Ramakrishnan, C., Sasisekharan, V., 1963. Stereochemistry of polypeptide chain configurations. J. Mol. Biol. 7, 95–99. Sadaf, S., Khan, M.A., Akhtar, M.W., 2007. Production of bubaline somatotropin by autoinduction in Escherichia coli. Biotechnol. Appl. Biochem. 47, 21–26. Sajjad, M., Khan, M.I., Zafar, R., Ahmad, S., Niazi, U.H., Akhtar, M.W., 2012. Influence of positioning of carbohydrate binding module on the activity of endoglucanase CelA of Clostridium thermocellum. J. Biotechnol. 161, 206–212. Samanich, K., Belisle, J.T., Laal, S., 2001. Homogeneity of antibody responses in tuberculosis patients. Infect. Immun. 69, 4600–4609. Sartain, M.J., Slayden, R.A., Singh, K.K., Laal, S., Belisle, J.T., 2006. Disease state differentiation and identification of tuberculosis biomarkers via native antigen array profiling. Mol. Cell. Proteomics 5, 2102–2113. Scarabelli, G., Morra, G., Colombo, G., 2010. Predicting interaction sites from the energetics of isolated proteins: a new approach to epitope mapping. Biophys. J. 98, 1966–1975. Singh, H., Ansari, H.R., Raghava, G.P., 2013. Improved method for linear B-cell epitope prediction using antigen's primary sequence. PLoS One 8, e62216. Straus, E., Wu, N., 1980. Radioimmunoassay of tuberculoprotein derived from Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. U. S. A. 77, 4301–4304. Tong, M., Jacobi, C.E., van de Rijke, F.M., Kuijper, S., van de Werken, S., Lowary, T.L., Hokke, C.H., Appelmelk, B.J., Nagelkerke, N.J., Tanke, H.J., van Gijlswijk, R.P., Veuskens, J., Kolk, A.H., Raap, A.K., 2005. A multiplexed and miniaturized serological tuberculosis assay identifies antigens that discriminate maximally between TB and non-TB sera. J. Immunol. Methods 301, 154–163. Vita, R., Zarebski, L., Greenbaum, J.A., Emami, H., Hoof, I., Salimi, N., Damle, R., Sette, A., Peters, B., 2010. The immune epitope database 2.0. Nucleic Acids Res. 38, D854–D862.

111

Vyas, N.K., Vyas, M.N., Quiocho, F.A., 2003. Crystal structure of M tuberculosis ABC phosphate transport receptor: specificity and charge compensation dominated by ion–dipole interactions. Structure 11, 765–774. Wang, C.R., Liu, M.F., Chen, M.Y., Lin, T.P., Cheng, C.S., Chuang, C.Y., 1989. Enzyme-linked immunosorbent assay with BCG sonicate antigen for diagnostic potential of mycobacterial infection in Taiwan. Zhonghua Min. Guo Wei Sheng Wu J. Mian Yi Xue Za Zhi 22, 97–104. Weldingh, K., Rosenkrands, I., Okkels, L.M., Doherty, T.M., Andersen, P., 2005. Assessing the serodiagnostic potential of 35 Mycobacterium tuberculosis proteins and identification of four novel serological antigens. J. Clin. Microbiol. 43, 57–65. WHO, 2011. Commercial serodiagnostic tests for diagnosis of tuberculosis. Policy Statement. Geneva, World Health Organisation. Wiederstein, M., Sippl, M.J., 2007. ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res. 35, W407–W410. Wilson, R.A., Maughan, W.N., Kremer, L., Besra, G.S., Futterer, K., 2004. The structure of Mycobacterium tuberculosis MPT51 (FbpC1) defines a new family of non-catalytic alpha/beta hydrolases. J. Mol. Biol. 335, 519–530. Xu, D., Zhang, Y., 2012. Ab initio protein structure assembly using continuous structure fragments and optimized knowledge-based force field. Proteins 80, 1715–1735. Young, R.A., Bloom, B.R., Grosskinsky, C.M., Ivanyi, J., Thomas, D., Davis, R.W., 1985. Dissection of Mycobacterium tuberculosis antigens using recombinant DNA. Proc. Natl. Acad. Sci. U. S. A. 82, 2583–2587. Zhang, G., Zhang, L., Zhang, M., Pan, L., Wang, F., Huang, J., Li, G., Yu, J., Hu, S., 2009. Screening and assessing 11 Mycobacterium tuberculosis proteins as potential serodiagnostical markers for discriminating TB patients from BCG vaccines. Genomics Proteomics Bioinformatics 7, 107–115. Zweig, M.H., Campbell, G., 1993. Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin. Chem. 39, 561–577.