Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis

Tuberculosis xxx (2013) 1e7

Contents lists available at ScienceDirect

Tuberculosis journal homepage: http://intl.elsevierhealth.com/journals/tube

IMMUNOLOGICAL ASPECTS

Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis Salam Bhopen Singh a, Debasis Biswas a, **, Jagdish Rawat b, Girish Sindhwani b, Abhishek Patras a, Suraj Devrani a, Pronoti Sarkar a, Soumik Mitra a, Shailendra K. Gupta c, * a b c

Department of Microbiology, Himalayan Institute of Medical Sciences, Swami Ram Nagar, Jolly Grant, Dehradun 248140, Uttarakhand, India Department of Pulmonary Medicine, Himalayan Institute of Medical Sciences, Dehradun 248140, India System Toxicology Group, CSIR e Indian Institute of Toxicology Research, PO Box e 80, M. G. Road, Lucknow 226001, UP, India

a r t i c l e i n f o

s u m m a r y

Article history: Received 15 December 2012 Received in revised form 30 July 2013 Accepted 4 August 2013

Differentiation between active and latent TB is a diagnostic challenge in TB-endemic regions. The commercially available IFN-g-release assays are unsuitable for achieving this discrimination. We, therefore, screened ESAT-6 and CFP-10 proteins through population coverage analysis to identify minimal sets of peptides that can discriminate between these two forms of TB in a North Indian population. Comparing the diagnostic performance of a set of 2 ESAT-6 peptides (positions: 16e36; 59e79) to that of the QuantiFERONÒ-TB Gold IT (QFTGIT) assay, we observed significant difference in IFN-g and TNF-a levels between patients (n ¼ 15) and their age- and sex-matched healthy household contacts (n ¼ 15). While the mean (SD) IFNg titer was 241.8 (219.24) IU/ml for patients, the same in controls was 564.2 (334.82) IU/ml (p ¼ 0.039). Similarly the TNFa response was significantly higher in patients, compared to controls (796.47  175.21 IU/ml vs. 481.81  378.72 IU/ml; p ¼ 0.047). IL-4 response to these peptides was non- discriminatory between the two groups. The QFTGIT Assay, however, elicited no significant difference in IFN-g, TNF-a or IL-4 levels. Hence we conclude that IFN-g or TNF-a response to these ESAT-6 peptides has the potential to differentiate between active and latent TB in our population. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: ESAT-6 Peptides Active tuberculosis Latent tuberculosis Immunodiagnosis

1. Introduction Tuberculosis (TB) is a “global emergency”, accounting for about 8.7 million new cases and 1.4 million deaths annually [1]. Though a third of the global population is infected with Mycobacterium tuberculosis (Mtb), 10% of this population progresses to develop active disease in absence of immuno-compromising conditions [2,3]. The discrimination between latently infected and actively diseased subjects is important for better understanding of the pathogenesis of TB and identification of potential diagnostic markers for active TB, particularly in TB-endemic countries.

Abbreviations: ATB, active tuberculosis; CFP-10, culture filtrate protein 10; ESAT6, early secretory antigenic target 6; HHC, healthy household contact; IEDB, immune epitope database; IU, international unit; Mtb, Mycobacterium tuberculosis; OPD, outdoor patients department; PHA, phytohaemagglutinin; PPD, purified protein derivative; QFTGIT, QuantiFERON TB gold in tube; SD, standard deviation; SPSS, statistical package for the social sciences; TST, tuberculin skin test. * Corresponding author. Tel.: þ91 522 2284591, þ91 9839533951; fax: þ91 522 2628227. ** Corresponding author. Tel.: þ91 135 2113545, þ91 135 2471575, þ91 9411362351 fax: þ91135 2471317. E-mail addresses: [email protected] (D. Biswas), [email protected], [email protected] (S.K. Gupta).

The discovery of Mtb-specific immunodominant antigens (ESAT6 and CFP-10), encoded by the RD1 region of the pathogen, have generated a growing interest in the immunodiagnosis of TB [4]. Though used in commercial IFN-g release assays, viz. QuantiFERONÒ-TB Gold IT Assay (QFTGIT) and T SPOT-TB, the diagnostic utility of these antigens remains to be established in highendemicity situations. Recent studies have shown that 62% and 80% of the urban population in Vietnam [5] and India [6] is latently infected with Mtb, as revealed by the presence of IFN-g response to peptides derived from ESAT-6 and CFP-10. Though this suggests a low likelihood of these commercial assays being diagnostically useful in endemic regions, there have been recent independent reports showing that the discrimination between active and latent TB can be achieved using two multi-epitopic peptides from ESAT-6 [7] and another RD1-encoded protein, viz. TB37.6 [8]. Despite being based on relatively small sample sizes drawn from non-endemic regions, these findings led us to hypothesize that use of selected peptides might perform better in specifically diagnosing active TB, compared to previous alternative approaches using the entire ESAT-6/CFP-10 protein [9e13] or a pool of overlapping peptides derived from the entire protein [14e16]. In addition to the potential utility associated with selected peptides, we further hypothesized that since these diagnostic assays

1472-9792/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tube.2013.08.001

Please cite this article in press as: Singh SB, et al., Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis, Tuberculosis (2013), http://dx.doi.org/10.1016/j.tube.2013.08.001

2

S.B. Singh et al. / Tuberculosis xxx (2013) 1e7

were based on host cell-mediated immune response, the selection of the peptides ought to be guided by the spectrum of HLA haplotypes prevalent in the target population. As HLA-DR alleles are highly polymorphic and their expression frequencies vary between different ethnicities, we considered it unlikely that a single pool of peptides, marketed globally in a given commercial assay, would be useful in delivering a specific immunological assay across ethnicity barriers. The present study was, therefore, carried out with the objectives of identifying a selected pool of ESAT-6/CFP-10 peptides, which offered the potential for differentiating between active and latent TB in TB-endemic Indian population and to examine if the diagnostic performance of this set of peptides in our TB patients varied significantly from that of a set of peptides from ESAT-6 protein of Mtb reported to be diagnostically useful in population belonging to a different ethnicity by Vincenti and coworkers [7]. We also compared the performance of this set of peptides with the QFTGIT Assay and further explored if any other cytokine response, apart from IFN-g, could be used as a diagnostic marker for active TB. 2. Materials and methods 2.1. Prediction of CD4þ T-cell epitopes and probable immunogenic peptide fragments Novel CD4þ T-cell epitopes from ESAT-6 and CFP-10 proteins of Mtb were predicted using NetMHCIIpan server (http://www.cbs. dtu.dk/services/NetMHCIIpan/) [17]. ESAT-6 and CFP-10 sequences were individually analyzed against all the HLA-DR alleles for potential epitopes. All epitopes with identical core region were merged to obtain the highly immunogenic peptide fragments from both the proteins as we did in our previous work [18]. In total, we have identified 2 peptide fragments from ESAT-6 and 4 peptide fragments from CFP-10 proteins of Mtb. 2.2. Population coverage analysis We performed population coverage analysis for potential immunogenic peptide fragments predicted from ESAT-6 and CFP-10. Population coverage analysis tool (http://tools.immuneepitope.org/tools/ population/) was used for this purpose [19]. The tool uses peptideMHC binding information and the pre-defined alleles frequency data of several populations to predict the maximum population coverage for a given set of epitopes. Currently, the tool covers 78 populations grouped into 11 different geographical locations. As all our cases were from North India, we have specifically used the previously published frequency data of 42 HLA-DR alleles from North Indian population. 2.3. Peptides Synthetic peptides (Table 1), of >95% purity (Peptides-1, 2, 3, 4, 5, 6 and 8) and >90% purity (Peptide-7), were procured commercially from BioConcept Lab Pvt. Ltd, Gurgaon, India. All the peptides were 21-mer, except peptide-7 which was 18-mer. The 8 peptides (Peptide 1e6 identified in this study and peptide 7, 8 identified by Vincenti and coworkers [7]) were used in 4 peptide pools, as depicted in Table 1. The PCT-1 pool comprised of peptides which were found to have maximum population coverage in Indian ethnicity, based on computational analysis from ESAT-6 and CFP-10, and PCT-2 comprised of the two peptides found to be capable of discriminating between active and latent TB in a previous study [7]. Both PCT-1 and PCT-2 were used at concentrations of 1 and 10 mg/ ml. In order to increase the specificity of PCT-1 pool, we further subdivided it into PCT-A (ESAT-6 peptides) and PCT-B (CFP-10 peptides) by selecting only non overlapping peptide fragments

Table 1 Synthetic peptides used in the study. Peptide pool

Constituent peptide

Origin

Amino acid sequence

PCT 1

Peptide-1

ESAT-6 (16 e36) ESAT-6 (59 e79) CFP-10 (1e18) CFP-10 (15 e35) CFP-10 (49 e69) CFP-10 (66 e86) ESAT-6 (6e28) ESAT-6 (66 e79) ESAT-6 (16 e36) ESAT-6 (59 e79) CFP-10 (1e18) CFP-10 (49 e69)

SAIQGNVTSIHSLLDEGKQSL

Peptide-2 Peptide-3 Peptide-4 Peptide-5 Peptide-6 PCT 2

Peptide-7 Peptide-8

PCT A

Peptide-1 Peptide-2

PCT B

Peptide-3 Peptide-5

DATATELNNALQNLARTISEA MAEMKTDAATLAQEAGNF AGNFERISGDLKTQIDQVEST TAAQAAVVRFQEAANKQKQEL KQELDEISTNIRQAGVQYSRA WNFAGIEAAASAIQGNVTSIHSL NALQNLARTISEA SAIQGNVTSIHSLLDEGKQSL DATATELNNALQNLARTISEA MAEMKTDAATLAQEAGNF TAAQAAVVRFQEAANKQKQEL

from both the proteins. The PCT-A and PCT-B pools were used at the concentrations of 10 mg/ml. The positions of all 8 peptides from ESAT-6 and CFP-10 protein used in the formation of various peptide pools are shown in Figure 1. 2.4. Patients and healthy household contacts In the first phase of the study, we compared the diagnostic performance of peptide pools, PCT-1 and PCT-2 with the commercial kit QFTGIT. For this, we recruited 15 consecutive patients with smear-positive Pulmonary TB attending the Pulmonary Medicine OPD of Himalayan Institute of Medical Sciences, Dehradun, India. We also recruited 15 of their age- and sex- matched healthy household contacts (HHCs). All patients with active TB had cough and expectoration of sputum for at least 2 weeks, were treatmentnaive fresh cases and were HIV-negative. Patients with relapse of TB, cases of treatment failure, patients suffering from co-morbid conditions or those receiving immunosuppressive medication were excluded from the study. All the controls were clinically asymptomatic, TST þve and had no past history of TB. None of the controls developed any symptom suggestive of active TB during the follow- up period of 6 months. All study participants were BCGvaccinated, as evidenced by the presence of BCG scar. In the next phase, we evaluated the performance of peptide pools PCT-A and PCT-B as well as of QFTGIT. For this, we freshly recruited 15 new cases of active TB and 15 HHCs, using similar inclusion and exclusion criteria. The control subjects were similarly found to remain asymptomatic during the follow- up period of 6 months. The study protocol was duly approved by the institutional ethics committee and written informed consent was taken from each study participant. 2.5. Comparison of IFN-g response to peptide pools PCT-1, PCT-2 and QFTGIT assay Nil tubes provided with QFTGIT kit, were coated with either of the two peptide pools at concentrations of 1 and 10 mg/ml. Three additional tubes (Nil tube, TB antigen tube and Mitogen tube) were used for every subject, in accordance to the instructions of kit manufacturers. Seven milliliters of whole blood samples were collected from each subject and divided equally in each of these 7 tubes. After overnight incubation at 37  C, the supernatants were

Please cite this article in press as: Singh SB, et al., Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis, Tuberculosis (2013), http://dx.doi.org/10.1016/j.tube.2013.08.001

S.B. Singh et al. / Tuberculosis xxx (2013) 1e7

3

Figure 1. Position of peptides on their respective proteins and preparation of peptide pools. Positions of all the 8 peptides used in this study are shown on their respective proteins. Numbers indicate the start and end position of the peptides. Two peptides (yellow in color) from ESAT-6 protein were included in PCT-2 pool and also previously used by Vincenti and coworkers. Color of the peptides indicates their inclusion in various peptide pools. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

harvested, aliquoted at 20  C and estimated for IFN-g titers, using ELISA kits. 2.6. Ex-vivo estimation of IFN-g, TNF-a and IL-4 response to PCT-A, PCT-B and QFTGIT peptides Nil tubes provided with the kit were again coated with either selected ESAT-6 or CFP-10 peptides, included in PCT-A and PCT-B respectively, at concentrations of 10 mg/ml. One milliliter of whole blood was added to each of these two tubes and to each of three tubes provided with QFTGIT kit, viz. a Nil tube, a TB antigen tube and a Mitogen tube. After overnight incubation at 37  C, the supernatants were harvested and aliquoted at 20  C. Quantitative estimation of IFN-g (Gen-Probe Diaclone SAS), TNF-a and IL-4 (both from Ani Biotech Oy, Orgenium Laboratories) was done using commercial kits, according to the instructions of kit manufacturers. 2.7. Statistical analysis Sensitivity, specificity, positive and negative likelihood ratios were calculated as per conventional definitions. For the calculation of sensitivity, the proportion of individuals testing positive among the cases while for specificity, the proportion testing negative among the controls was considered. Positive likelihood ratio was defined as the probability of a person who has the disease testing positive (sensitivity) divided by the probability of a person who does not have the disease testing positive (1- specificity). Similarly negative likelihood ratio was defined as the probability of a person who has the disease testing negative (1- sensitivity) divided by the probability of a person who does not have the disease testing negative (specificity). The sensitivity and specificity of the PCT-1 peptide pool was compared with PCT-2 peptide pool and QFTGIT Assay using McNemar’s test. The cytokine levels were compared between the patients and HHCs using independent sample t test. SPSS 17.0 was used and p value <0.05 was considered significant. 3. Results 3.1. CD4þ T-cell epitope prediction A total of 63 and 86 putative CD4þ T-cell epitopes were predicted for ESAT-6 and CFP-10 respectively using NetMHCIIpan

server for more than 500 known HLA-DR alleles. All predicted epitopes with similar core binding region were merged to screen out 6 peptide fragments from ESAT-6 and 17 peptide fragments from CFP-10 with potential CD4þ T-cell responses [Supplementary Tables 1 and 2].

3.2. Population coverage analysis Population coverage analysis is important to predict the sensitivity and specificity of HLA-DR allelic response for a given set of immunogenic peptides as the frequency of expression of these molecules varies across ethnicities. The combination of two peptide fragments from ESAT-6 (Position: 16e36; 59e79) and 4 from CFP10 (Position: 1e18; 15e35; 49e69; 66e86) gave maximum population coverage of 91.17% and 78.66% respectively in our population. Compared to this, the immunogenic peptide fragments of ESAT-6 suggested by Vincenti and co-workers [7] (Position: 6e28; 66e 80), showed 87.53% population coverage (Table 2). The combined pool of six peptide fragments (2 from ESAT-6 and 4 from CFP-10) showed maximum population coverage of 96.68% in North Indian ethnicity. This large North Indian population coverage with the cocktail of 6 peptide fragments motivated us to validate the results in clinical settings.

Table 2 The predicted population coverage rate (%) for the peptide pools PCT-1, PCT-2, PCT-A and PCT-B. Population

Population coverage (%) PCT-1

PCT-2

PCT-A

PCT-B

North India Australia Europe North Africa North America North East Asia Oceania South America South-East Asia South-West Asia Sub Saharan Africa

96.68 87.24 58.97 80.59 96.88 81.70 76.71 91.48 51.76 95.75 78.98

87.53 52.98 51.32 63.70 76.00 68.86 63.92 88.61 17.64 83.84 65.53

91.17 84.85 53.82 71.40 96.55 79.89 73.23 91.33 51.41 90.50 67.19

78.66 50.50 45.09 49.58 73.98 51.04 60.94 88.43 8.27 75.00 70.60

Please cite this article in press as: Singh SB, et al., Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis, Tuberculosis (2013), http://dx.doi.org/10.1016/j.tube.2013.08.001

4

S.B. Singh et al. / Tuberculosis xxx (2013) 1e7

3.3. Evaluation of IFN-g response to various peptide combinations in active and latent TB

low North Indian population coverage (21.05% and 13.14%) in comparison to other peptides in the group.

To examine if the difference in population coverage analysis is reflected in a differential immune response, we compared the IFN-g levels elicited by a combination of these 6 peptide fragments (PCT1) with the peptides used by Vincenti et al. [7] (PCT-2) and the combination of peptides provided in the commercial QFTGIT assay. For this, we recruited 15 patients with active TB and 15 age- and sex- matched HHCs. The mean (SD) age of the patients and the controls was 30 (7) and 40 (15) years respectively. With the commercial QFTGIT Assay, we observed false positive reaction in 11 of the 15 healthy contacts, thereby demonstrating specificity of 26.7%. Minimal reactivity was observed among both cases and controls to PCT-1 and PCT-2 at 1 mg/ml concentration. However at 10 mg/ml concentration, positive reaction was observed against PCT-1 and PCT-2 in 11 and 7 patients respectively. Though the sensitivity of the PCT-1 IFN-g assay (73.3%) was higher, compared to the PCT-2 (46.7%) and QFTGIT Assay (53.3%), the difference was not statistically significant compared to QFTGIT assay (p ¼ 0.375) or PCT-2 (p ¼ 0.289). Similarly the specificity of PCT-1 peptide pool was not significantly higher compared to QFTGIT Assay (p ¼ 0.453) or PCT-2 (p ¼ 0.754) (Figure 2). We also compared the likelihood ratios of the three peptide combinations, in order to understand how much the odds of disease changed based on positive or negative assay results, and observed that the likelihood ratios were similar for both QFTGIT Assay and PCT-2 (at 10 mg/ml concentration). However PCT-1 demonstrated higher positive and lower negative likelihood ratio, leading us to conclude that PCT-1 was more suited for ruling in or ruling out active TB than the other peptide combinations. But despite showing better diagnostic potential, the low specificity (46.7%) observed with PCT-1, precluded its application as a clinically useful diagnostic test and prompted us to further modify our choice of peptides with the aim of reducing falseepositive reactions (Figure 3). With this aim, we further subdivided the PCT-1 peptides pool into PCT-A (peptides 1 and 2 from ESAT-6) and PCT-B (peptides 3 and 5 from CFP-10) by selecting only non overlapping peptide fragments with maximum population coverage. In the process, we excluded peptide 4 and 6 belonging to CFP-10 because of the overlapping regions with peptide 3 and 5 and also because of their

3.4. Comparison of IFN-g, TNF-a and IL-4 response to selected ESAT-6 vs. selected CFP-10 and QFTGIT peptides

Figure 2. Comparison of sensitivity and specificity of QFTGIT with two different concentrations of PCT 1 and PCT 2. Fifteen patients with active TB and 15 age- and sexmatched healthy household contacts were recruited. Whole Blood ELISA was performed using the commercial QFTGIT assay and the peptide cocktails PCT 1 and PCT 2 (individual peptide sequences mentioned in Table 1) at concentrations of 1 mg/ml and 10 mg/ml. Sensitivity and specificity were calculated as per standard definitions. Though the sensitivity and specificity of the PCT 1 IFN-g assay was higher, compared to the PCT 2 and QFTGIT Assay, the difference was not statistically significant.

For this part of the study, we treated nil tubes provided with the QFTGIT Assay kit with either peptides 1 and 2 belonging to the ESAT-6 protein (PCT-A) or with peptides 3 and 5 belonging to the CFP-10 protein (PCT-B) (Table 1). These tubes were inoculated with whole blood from 15 new patients with active TB and 15 of their HHCs. The mean (SD) age of the patients and the controls was 32 (10) and 38 (8) years respectively. Along with these tubes, TB antigen tubes containing peptides used in the commercial QFTGIT Assay were similarly inoculated. Supernatants from these tubes were harvested after overnight incubation and IFN-g, TNF-a and IL4 levels were estimated using commercial ELISA kits. In contrast to the response obtained with PCT-B and QFTGIT assay, we observed significant difference between patients and HHCs in IFN-g levels produced in response to PCT-A (p ¼ 0.039). The mean (SD) IFN-g titer in patients and controls was 241.8 (219.24) IU/ml and 564.2 (334.82) IU/ml respectively (Figure 4). Similarly, unlike PCT-B and QFTGIT assay, PCT-A alone elicited significantly higher (p ¼ 0.047) TNF-a levels in patients (796.47(175.21) IU/ml) than in HHCs (481.81 (378.72) IU/ml) (Figure 5). There was no difference in IL-4 levels between patients and controls with either of the peptide combinations. 4. Discussion We have shown that though the commercially available QFTGIT IFN-g release assay cannot discriminate between active and latent TB in TB-endemic Indian population, the same can be accomplished

Figure 3. Comparison of the likelihood ratios achieved with peptide pools PCT 1 and PCT 2 and QFTGIT assay. Positive and negative likelihood ratios were calculated, as per standard definitions, for the whole blood ELISA experiments using QFTGIT assay and peptide cocktails PCT 1 (10 mg/ml) and PCT 2 (10 mg/ml). PCT 1 demonstrated higher positive likelihood ratio and lower negative likelihood ratio, compared to the other peptide combinations, thereby indicating its improved suitability for ruling in or ruling out the diagnosis of TB.

Please cite this article in press as: Singh SB, et al., Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis, Tuberculosis (2013), http://dx.doi.org/10.1016/j.tube.2013.08.001

S.B. Singh et al. / Tuberculosis xxx (2013) 1e7

Figure 4. Comparison of IFN-g response between ATB patients and HHCs using peptide pools, PCT A, PCT B and QFTGIT assay. The IFN-g response to QFTGIT antigens and peptide pools PCT A and PCT B was compared between individuals with active and latent TB. For this, nil tubes provided with the QFTGIT Assay kit were treated with the respective peptides. Whole blood was added to these tubes, along with TB antigen tubes containing peptides used in the commercial QFTGIT Assay, and culture supernatant samples were analysed for IFN-g response after overnight incubation. The bars indicate mean IFN-g values in the culture supernatants of the respective groups in ELISA experiments.

by an alternative combination of ESAT-6 peptides identified through analysis of broadly reactive HLA-DR-restricted CD4þ T cell epitopes with high population coverage in Indian ethnicity. We have also demonstrated that apart from IFN-g, TNF-a response to these peptides is also different in the two groups. In order to increase the specificity of PCT-1 pool of 6 peptides from ESAT-6 and CFP-10, we subdivided the pool into PCT-A and PCT-B by selecting non-overlapping peptides from the proteins. The strategy resulted in two peptides from ESAT-6 in PCT-A (position: 16e36 and 59e79) and two peptides from CFP-10 in PCT-B (position: 1e18 and 49e69). The combination of the two peptides in PCT-A, have

Figure 5. Comparison of TNF-a response between ATB patients and HHCs using peptide pools PCT A, PCT B and QFTGIT assay. Using the PCT-A, PCT-B and peptides provided in the commercial QFTGIT assay as stimulants, whole blood ELISA for TNF-a levels was performed on cell culture supernatants. PCT-A peptides were found to evoke significantly higher TNF-a levels in patients with active TB, compared to latently infected healthy household contacts. Significant difference in TNF-a titers was not observed with PCT-B peptide pool and the commercial assay. The bars indicate mean values of the respective groups.

5

maximum population coverage (91.17%) in North Indian ethnicity, which may be due to their potential binding affinity with 4 unique HLA-DR alleles (DRB1*0403, DRB1*0406, DRB1*1401 and DRB1*1405) expressed in North Indian population. We compared the populationspecific HLA haplotype coverage of two ESAT-6 peptides, previously suggested by Vincenti et al. [7], with other potential immunogenic peptide fragments. Though the previous study also explored the potential of ESAT-6 peptides to discriminate between active and latent TB, it had limited representation of Asian population. By failing to reciprocate the high sensitivity and specificity values (74% and 100% respectively) using these peptides in the Indian population, we demonstrate that the selection of peptides for such assays need to be guided by consideration of the HLA haplotypes prevalent in the ethnicity of interest. Though such an approach opens up an interesting and novel diagnostic path that needs to be explored further, we understand that the development of ethnic group- specific diagnostic tests for TB might have logistic and feasibility issues. Majority of studies on the diagnostic utility of IFN-g release assays have used these assays as indicators of Mtb infection [14,20e25]. While it is important to identify latently infected individuals for adequate control of TB in industrialized nations, the prime clinical requirement in TB-endemic regions is to differentiate between latent infection and active disease. Such differentiation has important therapeutic implications because only the latter group qualifies for anti-tubercular treatment in high TB-burden countries. In absence of a gold standard, we considered household contact without manifestation of active TB and positive tuberculin skin test result as indicative of latent infection. By demonstrating that this differentiation between active and latent TB is achieved better in the Indian population by use of the two selected ESAT-6 peptides (PCT-A) than with the commercial QFTGIT Assay, we underscore the importance of tailoring immunological diagnostic assays for TB with locally- relevant antigenic peptides. As HLA-DR alleles are highly polymorphic in nature and their expression frequencies vary in different ethnicities, we surmise that it would be unlikely that a single pool of peptides included in a particular commercial assay would be globally useful in delivering a specific immunological assay across ethnicity barriers. Previous papers using ESAT-6/CFP-10 derivatives in IFN-g release assays have mostly used three kinds of diagnostic reagents: (a) recombinant individual proteins or fusion protein [9,13,26e28] (b) pool of overlapping peptides [14e16,28] and (c) selected nonoverlapping multi-epitopic peptides [7,29e31]. Of them, similar to our findings, only the use of selected non-overlapping peptides have led to successful differentiation between active and latent TB. In this study, the QFTGIT kit which is based on a pool of 14 peptides (7 from ESAT-6; 6 from CFP-10 and 1 from TB7.7) has most of the peptides with 10e11 amino acid residues overlapping regions covering entire ESAT-6 and CFP-10 protein and the part of TB7.7 protein. To identify any similarity/overlaps between the peptides used in PCT-A and PCT-B pool with those used in the QFTGIT kit, we performed sequence alignment. The peptide-1 from PCT-A is a subset of the peptide-3 of ESAT-6 included in QFTGIT. However, the peptide-2 from PCT-A has partial overlap with either of the peptide-5 and 6 from ESAT-6 protein used in the QFTGIT kit. In case of PCT-B, both the peptide (peptide 3 and 5) are the subset of peptide-1 and 4 from CFP10 protein used in the QFTGIT kit. The sequence alignment data reveals that the peptide pool PCT-A and PCT-B has small subset of non overlapping peptides from ESAT-6 and CFP-10 proteins in comparison to the long overlapping peptides used in the QFTGIT kit. Since the nature and magnitude of protective host response to Mtb should be a major differentiating factor between active and latent TB, we hypothesized that a diagnostic test aiming to discriminate between these two groups must target the host response evoked by Mtb antigens in the population of effector immune cells. Compared to long-term cultures which can generate a memory

Please cite this article in press as: Singh SB, et al., Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis, Tuberculosis (2013), http://dx.doi.org/10.1016/j.tube.2013.08.001

6

S.B. Singh et al. / Tuberculosis xxx (2013) 1e7

response as well, overnight cultures with appropriate antigens are best suited for detecting immediate effector immune response [30,32]. Accordingly, we used relatively short overnight incubation in our assay in order to achieve our diagnostic objective of differentiation between active and latent TB. Moreover, an assay with shorter incubation period offers obvious clinical advantages. IFN-g, being the signature cytokine of protective Th1 response, might be expected to be present in lower titers in patients with active TB, as shown in the present study. However, contrary to our findings, some previous workers have observed higher IFN-g levels in actively diseased individuals than in their HHCs [7,9,10,21]. The lower IFN-g levels in our study could be because its production from peripheral blood cells is decreased with advanced TB [33e35], owing to the sequestration of the cells at the site of infection [36,37]. We further explored the possibility of estimating multiple cytokine responses to our selected peptide pool (PCT-A) and observed that TNF-a levels were significantly higher in active TB patients. This is in agreement with the recently published study by Harari et al. [38] who demonstrated that the proportion of singlepositive TNF-a specific CD4 T-cells was the strongest predictor of diagnosis of active versus latent infection [38]. While this study used polychromatic flow-cytometry and ELISPOT assay following stimulation with ESAT6 or CFP10 peptide pools, we observed similar TNF-a response in active TB patients with an ELISA-based assay. To the best of our knowledge, this is the first study demonstrating that the TNF-a response to ESAT-6 peptides can be used as a diagnostic marker for active TB in a technically simple assay that has the potential of being implemented in clinical laboratories in resource-constrained settings. Differential TNF-a response between subjects with Mtb infection and healthy individuals has been reported by other authors using alternative antigens, like PPD, PHA, TB10.4, Ag85A, Ag85B, ESAT-6/CFP-10 fusion protein, ESAT-6 and overlapping peptides of ESAT-6 [10,21,39]. One of the limitations of our study was the failure to explore if titers of cytokines, other than IFN-g, TNF-a and IL-4, also varied between latently infected and diseased individuals in response to the selected pool of ESAT-6 peptides. However, most authors reporting on the levels of cytokines, like IL-2, IL-4 and IL-18, have failed to detect a significant difference between TB patients and healthy subjects [21,40]. To conclude, our results demonstrate that while the commercially available QFTGIT assay did not discriminate between active and latent TB, an alternative combination of 2 ESAT-6 peptides elicited significantly different IFN-g and TNF-a response in patients with active TB compared to their HHCs. While the findings of this pilot study hint at the possibility of using these peptides in a simple diagnostic assay for active TB in the Indian population, they need to be further validated in more elaborate studies conducted on a larger sample size representing the diverse ethnic sub-groups within the Indian population. Ethical approval:

Not required.

Funding: This work was supported by financial assistance from the Dept of Biotechnology, Ministry of Science & Technology, Govt. of India [Project no. BT/PR/9785/GBD/27/72/2007]. SKG acknowledges the financial support from Council of Scientific and Industrial Research (CSIR) Network Projects GENESIS (BSC0121) and INDEPTH (BSC0111). Competing interests:

None declared.

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.tube.2013.08.001.

References [1] World Health Organization. Global tuberculosis report. Available at, http:// www.who.int/tb/publications/global_report/gtbr12_executivesummary.pdf; 2012 [accessed on 30.04.13]. [2] Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO global surveillance and monitoring project. JAMA 1999;282: 677e86. [3] Styblo K. Recent advances in epidemiological research in tuberculosis. Adv Tuberc Res 1980;20:1e63. [4] Connell T, Bar-Zeev N, Curtis N. Early detection of perinatal tuberculosis using a whole blood interferon-gamma release assay. Clin Infect Dis 2006;42:e82e5. [5] Simmons CP, Thwaites GE, Quyen NT, Torok E, Hoang DM, Chau TT, Mai PP, Lan NT, Dung NH, Quy HT, Bang ND, Hien TT, Farrar J. Pretreatment intracerebral and peripheral blood immune responses in Vietnamese adults with tuberculous meningitis: diagnostic value and relationship to disease severity and outcome. J Immunol 2006;176:2007e14. [6] Lalvani A, Nagvenkar P, Udwadia Z, Pathan AA, Wilkinson KA, Shastri JS, Ewer K, Hill AV, Mehta A, Rodrigues C. Enumeration of T cells specific for RD1encoded antigens suggests a high prevalence of latent Mycobacterium tuberculosis infection in healthy urban Indians. J Infect Dis 2001;183:469e77. [7] Vincenti D, Carrara S, De Mori P, Pucillo LP, Petrosillo N, Palmieri F, Armignacco O, Ippolito G, Girardi E, Amicosante M, Goletti D. Identification of early secretory antigen target-6 epitopes for the immunodiagnosis of active tuberculosis. Mol Med 2003;9:105e11. [8] Leyten EM, Mulder B, Prins C, Weldingh K, Andersen P, Ottenhoff TH, van Dissel JT, Arend SM. Use of enzyme-linked immunospot assay with Mycobacterium tuberculosis-specific peptides for diagnosis of recent infection with M. tuberculosis after accidental laboratory exposure. J Clin Microbiol 2006;44: 1197e201. [9] Kobashi Y, Shimizu H, Ohue Y, Mouri K, Obase Y, Miyashita N, Oka M. Comparison of T-cell interferon-gamma release assays for Mycobacterium tuberculosis-specific antigens in patients with active and latent tuberculosis. Lung 2010;188:283e7. [10] Macedo GC, Bozzi A, Weinreich HR, Bafica A, Teixeira HC, Oliveira SC. Human T cell and antibody-mediated responses to the Mycobacterium tuberculosis recombinant 85A, 85B, and ESAT-6 antigens. Clin Dev Immunol 2011;2011: 351573. [11] Marei A, Ghaemmaghami A, Renshaw P, Wiselka M, Barer M, Carr M, ZieglerHeitbrock L. Superior T cell activation by ESAT-6 as compared with the ESAT6-CFP-10 complex. Int Immunol 2005;17:1439e46. [12] Hasan Z, Jamil B, Ashraf M, Islam M, Yusuf MS, Khan JA, Hussain R. ESAT6induced IFN gamma and CXCL9 can differentiate severity of tuberculosis. PLoS One 2009;4:e5158. [13] Ravn P, Munk ME, Andersen AB, Lundgren B, Lundgren JD, Nielsen LN, KokJensen A, Andersen P, Weldingh K. Prospective evaluation of a whole-blood test using Mycobacterium tuberculosis-specific antigens ESAT-6 and CFP-10 for diagnosis of active tuberculosis. Clin Diagn Lab Immunol 2005;12:491e6. [14] Lalvani A, Pathan AA, McShane H, Wilkinson RJ, Latif M, Conlon CP, Pasvol G, Hill AV. Rapid detection of Mycobacterium tuberculosis infection by enumeration of antigen-specific T cells. Am J Respir Crit Care Med 2001;163:824e8. [15] Mustafa AS, Al-Attiyah R, Hanif SN, Shaban FA. Efficient testing of large pools of Mycobacterium tuberculosis RD1 peptides and identification of major antigens and immunodominant peptides recognized by human Th1 cells. Clin Vaccine Immunol 2008;15:916e24. [16] Pathan AA, Wilkinson KA, Klenerman P, McShane H, Davidson RN, Pasvol G, Hill AV, Lalvani A. Direct ex vivo analysis of antigen-specific IFN-gamma-secreting CD4 T cells in Mycobacterium tuberculosis-infected individuals: associations with clinical disease state and effect of treatment. J Immunol 2001;167:5217e25. [17] Lin HH, Zhang GL, Tongchusak S, Reinherz EL, Brusic V. Evaluation of MHC-II peptide binding prediction servers: applications for vaccine research. BMC Bioinformatics 2008;9(Suppl. 12):S22. [18] Gupta SK, Srivastava M, Akhoon BA, Smita S, Schmitz U, Wolkenhauer O, Vera J, Gupta SK. Identification of immunogenic consensus T-cell epitopes in globally distributed influenza-A H1N1 neuraminidase. Infect Genet Evol 2011;11:308e19. [19] Bui HH, Sidney J, Dinh K, Southwood S, Newman MJ, Sette A. Predicting population coverage of T-cell epitope-based diagnostics and vaccines. BMC Bioinformatics 2006;7:153. [20] Mori T, Sakatani M, Yamagishi F, Takashima T, Kawabe Y, Nagao K, Shigeto E, Harada N, Mitarai S, Okada M, Suzuki K, Inoue Y, Tsuyuguchi K, Sasaki Y, Mazurek GH, Tsuyuguchi I. Specific detection of tuberculosis infection: an interferon-gamma-based assay using new antigens. Am J Respir Crit Care Med 2004;170:59e64. [21] Sutherland JS, de Jong BC, Jeffries DJ, Adetifa IM, Ota MO. Production of TNFalpha, IL-12(p40) and IL-17 can discriminate between active TB disease and latent infection in a West African cohort. PLoS One 2010;5:e12365. [22] Scarpellini P, Tasca S, Galli L, Beretta A, Lazzarin A, Fortis C. Selected pool of peptides from ESAT-6 and CFP-10 proteins for detection of Mycobacterium tuberculosis infection. J Clin Microbiol 2004;42:3469e74. [23] Abramo C, Meijgaarden KE, Garcia D, Franken KL, Klein MR, Kolk AJ, Oliveira SC, Ottenhoff TH, Teixeira HC. Monokine induced by interferon

Please cite this article in press as: Singh SB, et al., Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis, Tuberculosis (2013), http://dx.doi.org/10.1016/j.tube.2013.08.001

S.B. Singh et al. / Tuberculosis xxx (2013) 1e7

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

gamma and IFN-gamma response to a fusion protein of Mycobacterium tuberculosis ESAT-6 and CFP-10 in Brazilian tuberculosis patients. Microbe. Infect 2006;8:45e51. Brock I, Munk ME, Kok-Jensen A, Andersen P. Performance of whole blood IFN-gamma test for tuberculosis diagnosis based on PPD or the specific antigens ESAT-6 and CFP-10. Int J Tuberc Lung Dis 2001;5:462e7. Ulrichs T, Anding R, Kaufmann SH, Munk ME. Numbers of IFN-gammaproducing cells against ESAT-6 increase in tuberculosis patients during chemotherapy. Int J Tuberc Lung Dis 2000;4:1181e3. Sargentini V, Mariotti S, Carrara S, Gagliardi MC, Teloni R, Goletti D, Nisini R. Cytometric detection of antigen-specific IFN-gamma/IL-2 secreting cells in the diagnosis of tuberculosis. BMC Infect Dis 2009;9:99. Hill PC, Jackson-Sillah D, Fox A, Franken KL, Lugos MD, Jeffries DJ, Donkor SA, Hammond AS, Adegbola RA, Ottenhoff TH, Klein MR, Brookes RH. ESAT-6/CFP10 fusion protein and peptides for optimal diagnosis of Mycobacterium tuberculosis infection by ex vivo enzyme-linked immunospot assay in the Gambia. J Clin Microbiol 2005;43:2070e4. Arend SM, Geluk A, van Meijgaarden KE, van Dissel JT, Theisen M, Andersen P, Ottenhoff TH. Antigenic equivalence of human T-cell responses to Mycobacterium tuberculosis-specific RD1-encoded protein antigens ESAT-6 and culture filtrate protein 10 and to mixtures of synthetic peptides. Infect Immun 2000;68:3314e21. Goletti D, Vincenti D, Carrara S, Butera O, Bizzoni F, Bernardini G, Amicosante M, Girardi E. Selected RD1 peptides for active tuberculosis diagnosis: comparison of a gamma interferon whole-blood enzyme-linked immunosorbent assay and an enzyme-linked immunospot assay. Clin Diagn Lab Immunol 2005;12:1311e6. Goletti D, Butera O, Bizzoni F, Casetti R, Girardi E, Poccia F. Region of difference 1 antigen-specific CD4þ memory T cells correlate with a favorable outcome of tuberculosis. J Infect Dis 2006;194:984e92. Goletti D, Carrara S, Vincenti D, Saltini C, Rizzi EB, Schininà V, Ippolito G, Amicosante M, Girardi E. Accuracy of an immune diagnostic assay based on RD1 selected epitopes for active tuberculosis in a clinical setting: a pilot study. Clin Microbiol Infect 2006;12:544e50.

7

[32] Lalvani A, Brookes R, Hambleton S, Britton WJ, Hill AV, McMichael AJ. Rapid effector function in CD8þ memory T cells. J Exp Med 1997;186:859e65. [33] Morosini M, Meloni F, Uccelli M, Marone Bianco A, Solari N, Fietta AM. Ex vivo evaluation of PPD-specific IFN-gamma or IL-5 secreting cells in the peripheral blood and lungs of patients with tuberculosis. Int J Tuberc Lung Dis 2005;9: 753e9. [34] Dlugovitzky D, Torres-Morales A, Rateni L, Farroni MA, Largacha C, Molteni O, Bottasso O. Circulating profile of Th1 and Th2 cytokines in tuberculosis patients with different degrees of pulmonary involvement. FEMS Immunol Med Microbiol 1997;18:203e7. [35] Sharma SK, Mitra DK, Balamurugan A, Pandey RM, Mehra NK. Cytokine polarization in miliary and pleural tuberculosis. J Clin Immunol 2002;22: 345e52. [36] Scriba TJ, Kalsdorf B, Abrahams DA, Isaacs F, Hofmeister J, Black G, Hassan HY, Wilkinson RJ, Walzl G, Gelderbloem SJ, Mahomed H, Hussey GD, Hanekom WA. Distinct, specific IL-17- and IL-22-producing CD4þ T cell subsets contribute to the human anti-mycobacterial immune response. J Immunol 2008;180:1962e70. [37] Brock I, Weldingh K, Leyten EM, Arend SM, Ravn P, Andersen P. Specific T-cell epitopes for immunoassay-based diagnosis of Mycobacterium tuberculosis infection. J Clin Microbiol 2004;42:2379e87. [38] Harari A, Rozot V, Enders FB, Perreau M, Stalder JM, Nicod LP, Cavassini M, Calandra T, Blanchet CL, Jaton K, Faouzi M, Day CL, Hanekom WA, Bart PA, Pantaleo G. Dominant TNF-aþ Mycobacterium tuberculosis-specific CD4þ T cell responses discriminate between latent infection and active disease. Nat Med 2011;17:372e6. [39] Fortes A, Pereira K, Antas PR, Franken CL, Dalcolmo M, Ribeiro-Carvalho MM, Cunha KS, Geluk A, Kritski A, Kolk A, Klatser P, Sarno EN, Ottenhoff TH, Sampaio EP. Detection of in vitro interferon-gamma and serum tumour necrosis factor-alpha in multidrug-resistant tuberculosis patients. Clin Exp Immunol 2005;141:541e8. [40] Hughes AJ, Hutchinson P, Gooding T, Freezer NJ, Holdsworth SR, Johnson PD. Diagnosis of Mycobacterium tuberculosis infection using ESAT-6 and intracellular cytokine cytometry. Clin Exp Immunol 2005;142:132e9.

Please cite this article in press as: Singh SB, et al., Ethnicity-tailored novel set of ESAT-6 peptides for differentiating active and latent tuberculosis, Tuberculosis (2013), http://dx.doi.org/10.1016/j.tube.2013.08.001