Journal of Infection (2016) 73, 455e467

www.elsevierhealth.com/journals/jinf

T cell recall response of two hypotheticalproteins (Rv2251 and Rv2721c) fromMycobacterium tuberculosis in healthyhousehold contacts of TB e Possible subunitvaccine candidates

D. Santhi, Alamelu Raja*

Department of Immunology, National Institute for Research in Tuberculosis (ICMR), (FormerlyTuberculosis Research Centre), No.1, Mayor Sathyamoorthy Road, Chetpet, Chennai 600 031, India

Accepted 28 June 2016Available online 9 July 2016

KEYWORDSM. tuberculosis;Latent and activetuberculosis;T cell response;Whole blood culture;Multicolur flowcytometry

* Corresponding author. Fax: þ91 (0E-mail address: alamelur@nirt.res.

http://dx.doi.org/10.1016/j.jinf.20160163-4453/ª 2016 Published by Elsevi

Summary The demonstrated variable efficacy of the only licensed TB vaccine Mycobacte-rium bovis bacillus CalmetteeGuerin (M. bovis BCG) encourages the need for new vaccine can-didates against TB. Antigen specific cellular immune response is often considered imperativeduring Mycobacterium tuberculosis (M. tuberculosis) infection and antigens that are stronglyassociated with the latent phase of infection are drawing increasing attention for anti-TB vac-cine development. Here, we investigated the phenotypic and functional profiles of two novelmycobacterial antigens Rv2251 and Rv2721c during T cell recall response via multi-color flowcytometry. Healthy household contacts of TB (latent/HHC) and active pulmonary TB (PTB) pa-tients were recruited to investigate the difference in antigen specific T cell recall response.These two antigens induced expansion of CD45RA� CCR7þ central memory subtypes andCD45RA� CCR7� effector memory cells in latent population which suggests their possible asso-ciation with HHC. Rv2251 and Rv2721c antigen specific IFN-g, TNF-a and IL-2 response was alsosignificantly high in HHC when compared to the PTB (p < 0.005, p < 0.05 and p < 0.05 respec-tively). The frequency of multifunctional T cells also was high in HHC compared to the PTB withstatistical significance only for the antigen Rv2251. Often, the dominant Th1 immune responsein HHC is correlated with the protection against the active TB disease. Collectively, we reportthe first insights into Rv2251 and Rv2721c antigen specific immune response in human donors ofTB and provide the immunologic rationale for selecting them for vaccine development againstTB.ª 2016 Published by Elsevier Ltd on behalf of The British Infection Association.

44) 2836 2528.in (A. Raja).

.06.012er Ltd on behalf of The British Infection Association.

456 D. Santhi, A. Raja

Introduction

Despite the availability of the only licensed vaccine BCG,tuberculosis (TB) is still the leading microbial cause ofdeath.1 Though, the administration of BCG conferred pro-tection against TB from its first oral administration in1921, it later showed little or no protection during largefield trials.2 BCG might diminish the most severe forms ofTB, such as TB meningitis in children. But, it has had varyingimpacts on the occurrence of pulmonary TB in adults, whichrepresents the transmissible form of this disease. Hence, animproved vaccine as a replacement to BCG or a subunitboosting vaccine to administer after BCG vaccination isneeded. Subunit vaccine approaches hold a number of ad-vantages, like increased safety, stability and ability toboost prior BCG immunization.3 Pools of proteins obtainedfrom Mycobacterium tuberculosis, the causative agent oftuberculosis, are often screened for subunit vaccine devel-opment and several promising candidates have beentargeted.

The role of CD4þ Th1-cells in TB is best understood andmany researchers concluded that anti-TB immunity is pre-dominantly mediated by CD4þ Th1 cells.4,5 To date, manymycobacterial antigens have been screened for their abilityto induce cellular immune responses with the goal of sub-unit vaccine application. These include Ag85 complex,6

ESAT-6,7 CFP-10,8 Rv0577,9 HspX5 etc., but their T cell re-sponses are not homogenous.10 We hypothesize that, oneof the reasons for variation in T cell response to these anti-gens might be due to the absence of specific MHC allelesamongst different ethnicities for presentation of these anti-genic peptides. This hypothesis is well supported by ourearlier published work that showed comparatively low pop-ulation coverage and MHC binding for ESAT-6 and CFP-10peptides relative to other screened antigens.11 Speculationsabout the reasons for the differences in BCG efficacy alsolinked the role of genetic differences in different ethnic-ities.12 In this light, we have selected two mycobacterial an-tigens, predicted to have high MHC binding affinity and highpercentage of population coverage in lieu of immunodomi-nant antigens ESAT-6 and CFP-10. We evaluated T cell im-mune responses with the aim of subunit vaccine targetings.

These two antigens are Rv2251 (Possible flavoprotein)and Rv2721c (Possible conserved transmembrane alanineand glycine rich protein). Rv2251 was predicted to be anouter membrane protein13 and the Rv2721c was identifiedby mass spectrometry in the membrane fractions of M.tuberculosis.14 In general, membrane localized proteinsare targeted for the vaccination studies since they arereadily available for immune processing within the host.15

In the field of TB, we can consider latently infected individ-uals who remain healthy as protected against active TB dis-ease due to their ability to control the infection, unlikeactive TB patients. Their cellular immune response wouldreflect the type of immunity responsible for their efficientdisease control and serve as a good experimental model.16

Thus, we are interested in evaluating these two antigens’specific T cell responses to make a better correlation eitherwith protection or pathology. This human based in vitrostudy would minimize the variations with other existing an-imal based in vitro TB models. Our analysis is bringing the

first insight into Rv2251 and Rv2721c antigens’ specific im-mune response in human models.

Materials and methods

In vitro cloning of Rv2251 and Rv2721c, over-expression and purification

The recombinant plasmid encoding Rv2251 antigen wasobtained by in vitro cloning.11 The same methodology wasadapted for cloning and purification of Rv2721c. Briefly, en-coding gene sequences of Rv2721c were amplified from M.tuberculosis H37Rv genomic DNA with gene specific primersby Phusion High Fidelity DNA polymerase (New England Bio-labs, MA). The amplified Rv2721c gene and pRSET-A plasmidDNA were digested with restriction enzymes Sac-I and Hind-III (NEB, MA). The digested gene and plasmid DNA wereligated by T4 DNA ligase (NEB, MA) and transformed into acloning strain of Escherichia coli (E. coli) DH5a. Positive re-combinants confirmed by DNA sequencing were then trans-formed into the E. coli BL21 (DE3) (Invitrogen, USA) strainfor protein induction and purification. The purifiedRv2721c antigen was used for in vitro blood stimulationalong with recombinant Rv2251 antigen.

Study participants

Informed written consent was obtained from all the donorsand the study was approved by an Institutional Ethicalcommittee of the National Institute for Research in Tuber-culosis, Chennai, India. Thirteen milliliters (mL) peripheralblood samples were collected from a total of thirty nineindividuals. Among the thirty nine donors, twenty two werelatently infected individuals in our case healthy householdcontacts (HHCs) of TB and seventeen had active pulmonarytuberculosis (PTB). The criteria for HHC recruitment wasbased on sharing the living shelter minimum for threemonths with at least one sputum positive active TB patients(index TB case) who are naive for anti-tubercular therapy.HHCs had more than ten hours per day of close contact withadult smear-positive PTB patients. They were identified byvisiting the households of adult smear positive pulmonaryTB patients, who were enrolled for treatment at theGovernment Thiruvatteeswarar Hospital of Thoracic Medi-cine, Otteri, Chennai, Tamil Nadu, India. HHCs had noactive TB symptoms which were ruled out by chest X-rayand smear microscopy. PTB individuals were selected basedon positive sputum microscopy results of three sputumsamples.

The infection state of all the study participants wasassessed by the Interferon gamma release assay (IGRA). Noactive TB symptoms were found in HHCs which wasconfirmed by negative sputum smear microscopy. PTB studyparticipants who had symptoms of immunosuppressive dis-ease such as HIV infection were excluded from this study.

Interferon gamma release assay (IGRA)

QuantiFERON-TB Gold In-Tube (QFT-GIT) kit (Cellestis,Qiagen) contains three pre-coated tubes coated with TB-

T-cell response of Rv2251 and Rv2721c 457

antigens (ESAT-6, CFP-10, and TB7.7) as a test, phytohe-magglutinin (PHA) as a positive control, the nil antigen(coated with saline) as a negative control. Out of thirteenmilliliters (mL) of collected blood, one mL was added toeach tube of the kit and incubated for 16e24 h at 37 �C, 5%atmospheric CO2. The supernatant was collected aftercentrifugation of the three tubes and stored at �80 �C untilthe assay was done. The test results were interpreted asper kit guidelines, using software provided by themanufacturer.

Whole blood culture with Rv2251 and Rv2721 fromthe blood of HHCs and PTB

The collected blood was diluted 1:1 with RPMI1640 (SigmaAldrich, St. Louis, MO, USA) medium with penicillin/streptomycin (100 U/100 mg/mL), L-glutamine (2 mM),and HEPES (10 mM) and distributed as two mL into tissueculture plates. Then, the cultures were stimulated withM. tuberculosis ESAT-6 (E6), CFP-10 (C-10), Rv2251 andRv2721c antigens at the final concentration of 5 mg/mL asdetermined previously.16 Phytohemagglutinin (PHA) wasused as mitogen control, at the final concentration of1 mg/mL, to test the proliferating capacity of the cellsfrom donors. The recombinant plasmids of immunodomi-nant antigens ESAT-6 and CFP-10 were received from Colo-rado State University (CSU), USA. Diluted blood withoutany added antigen was set as an unstimulated culture con-trol. To all the stimulations, purified co-stimulatory mole-cules CD49d/CD28 (BD biosciences, San Diego, CA, USA)were added at a final concentration of 0.5 mg/mL. Theseculture plates were incubated for 4 h at 37 �C, 5% CO2

and removed for the addition of Brefeldin A (10 mg/mL)then continued the incubation for 12 h strictly. Uponcompletion of 16 h incubation, plates were removed forcell harvesting with PBS. The harvested cells were pelleteddown by centrifuging at 2600 rpm for 10 min and thentreated with a BD FACS lysing solution (BD, San Diego,CA, USA) to lyse RBCs as per manufacturer’s instruction.The cells were fixed using BD cytofix/cytoperm bufferand cryopreserved with 10% DMSO at �80 �C until intracel-lular staining was done.

Surface markers and intracellular cytokine staining

Fixed cultured cells were rapidly thawed from �80 �C andwashed with PBS. These cells were stained for T cellsurface markers and intracellular markers after single cellsuspensions were made. The following antibodies wereused for surface staining at several different panels: APC-anti-CD69 (clone FN 50), PerCP 5.5 - anti-CD3 (cloneUCHT1), APC-Cy7-CD4 (clone RPA-T4), PE-Cy7-CD8 (cloneRPA-T8), FITC-CD197 (CCR7) (clone 3D12), APC-C45RA(clone HI100) at the final concentration of 5 ml/1 millioncells. The following antibodies were used for intracellularstaining: FITC-IFN-g (clone B27), PE-TNF-a (clone MAb11),APC-IL-2 (clone 5344.111), FITC-IL-17A (clone N49-653), PE-IL-22 (clone BG/IL-22) and APC-IL-21 (clone 3A3-N2.1). Allantibodies were purchased from BD Pharmingen, USA orBioLegend, USA. All staining was done at 4� C for 30 min in

the dark followed by washing with PBS for surface stainingand with perm wash buffer after intracellular staining.

FACSCanto II flow cytometer with FACSDiva software,version 6 (Becton Dickinson and Company, Cockeysville,MD) was used for acquisition of stained single cell suspen-sions. Acquired data was analyzed in Flow Jo software(TreeStar). All data is depicted as the percentage of CD4þ Tcells expressing cytokine(s).

Positive and negative boundaries were defined by settingFluorescence minus one (FMO) controls for all the usedantibodies and for all the positive results obtained, FMOcontrol is shown in supplementary figures. CompBeadspurchased from BD Biosciences, USA and stained with theall used fluorochrome monoclonal antibodies as compensa-tion control.

Data analysis

Statistical analyses were performed using Graphpad Prismsoftware version 5 (GraphPad software, CA, USA). Inter-group comparisons were performed using the nonpara-metric Mann Whitney U and with Holm’s correction formultiple comparisons. For all the analysis, differences wereconsidered as significant if p value was less than 0.05.

Results

Table 1 summarizes the clinical characteristics of thehealthy household contacts (HHC) and the active TB pa-tients (PTB). QFT-IT test confirmed the mycobacterialinfection status of all the study participants who were pos-itive (>3.5) for QFT-IT. The cellular response, was similar inboth HHC and PTB, �0.5 IU/mL, in the mitogen tube ofQFT-IT kit showed no defects in immune response in allstudy subjects. All PTB subjects were naive to anti-TBtreatment at the time of enrolling in our study. Percentageof CD4þ cells in both HHC and PTB ranged between 20% and65% and expression of early T cell activation marker CD69þ

was similar also between the two groups (SupplementaryFig. 1) showing no defects in immune cell responsivenessin all the study subjects. Fluorescence minus One (FMO)for all the results obtained (memory T cell and Th1 analysis)is given in Supplementary Fig. 2.

Cloning, expression and purification of Rv2721c

The encoding gene sequence of Rv2721c (2.1 kb) wasamplified from M. tuberculosis H37Rv with gene specificprimers (forward primer sequence (Sac I) is 50CTG GGAGAG CTC GTG AAC GGG CAG AG 30 and Reverse primersequence (Hind-III) 50 CC GCG TGC AAG CTT TCA ATC CGCCC30) by Phusion high fidelity DNA polymerase whichyielded a single 2100 bp fragment at an optimal tempera-ture of 60.8 �C (Fig. 1a, lane 4). The plasmid DNA pRSET-A was digested with Sac I and Hind-III to produce cohesivetermini (Fig. 1b). Digested plasmid pRSET-A and Rv2721cgene were ligated by T4 DNA ligase and transformed intoE. coli DH5a. Then, these recombinant colonies (taken astemplate) were screened for presence of Rv2721c gene bycolony PCR with gene specific primers and PCR

Table 1 Demographic details of all study participants.

Characteristic of study participants Healthy household contacts (HHC) Active pulmonary TB individuals (PTB)

Sample size 22 17Mean age (yr) (range) 39 (27e55) 41 (25e56)Sex

Female, N (%) 12 (54%) 8 (47%)Male, N (%) 10 (45%) 9 (55%)Sputum smear positivity, N (%) 0 (0) 17 (100)Semear grade, þþþ, N (%) 0 (0) 8 (50)Semear grade, þþ, N (%) 0 (0) 5 (27)Semear grade, þ, N (%) 0 (0) 4 (22)QFT-GIT

Positive, N (%) 22 (100) 17 (100)Negative, N (%) 0 (0) 0 (0)Indeterminate, N (%) 0 (0) 0 (0)CD4 count e range in % 20e65 21e62

Figure 1 In vitro cloning, over-expression and purification of Rv2721c from H37Rv genome. a) Rv2721c gene, 2100 bp, was ampli-fied by PCR. Lane 1e7 shows amplification at various temperatures ranging from 55 �C to 70 �C. Maximum amplification with no non-specific amplification was considered as optimal and 60.8 �C was selected. Lane “M” is 10 kb DNA ladder. b) pRSET-A plasmid DNAwas digested with Sac-I and Hind-III restriction enzyme and complete linearization was obtained at 37 �C for 4 h s (lane 2), lane 1shows undigested plasmid pRSET-A. Lane “M” is 10 kb DNA ladder. c) PCR analysis with recombinant colonies showed the amplifi-cation of Rv2721c (2.1 kb) which confirmed the presence of gene. Lane M is 10 kb DNA ladder, lane “1” is positive control (H37Rvgenomic DNA) and lane 2e12 various colonies screened. d) The release of Rv2721c (insert release) from the positive recombinantcolonies confirms the presence of Rv2721c gene. Lane 1 and 2 shows the insert release and 3 is negative control. Lane M is 10 kb DNAladder. e) IPTG induction of positive recombinant colonies. Lane UI denotes culture aliquoted before adding IPTG and lane 1e3indicates expression of Rv2721c at 25 �C, 30 �C and 37 �C. f) Rv2721c protein was purified by His-tag Ni-NTA affinity chromatographyand lane 4e8 shows purified protein fractions. M e indicates protein ladder. g) Western blot with anti-His antibody confirms therecombinant Rv2721c.

458 D. Santhi, A. Raja

amplification to confirm the presence of the insert geneRv2721c (Fig. 1c). Colonies with amplification of Rv2721cwere marked as “positive recombinant colonies”.

Plasmid DNA was isolated from positive recombinantcolonies and subjected to restriction digestion with thesame set of enzymes used for cloning (Sac-I and Hind-III).

Release of Rv2721c (insert release) from the recombinantplasmid DNA confirms the presence of gene of interest(Fig. 1d). The presence of the Rv2721c gene in recombinantplasmid DNA was also confirmed by DNA sequencing.Sequencing result showed 100% homology identity withthe original gene sequence when searched against M.

T-cell response of Rv2251 and Rv2721c 459

tuberculosis genomic DNA in NCBI BLAST. Positive recombi-nant plasmids were transformed into E. coli expression sys-tem BL21 DE3 and protein expression was tested at varioustemperatures (25 �C, 30 �C and 37� C) (Fig. 1e). Cultureswere grown at 37� C until reaching OD600 nm Z 0.5, inducedwith 1 mM Isopropyl b-D-1-thiogalactopyranoside (IPTG) andthe eluted fraction of Rv2721 was analyzed by SDS-PAGE(Fig. 1f). The recombinant Rv2721c was confirmed by West-ern blot with His-Tag Antibody (Novagen, Germany)(Fig. 1g).

Antigen specific central memory phenotypeCD45RAL CCR7D in healthy mycobacterial infectedindividuals

Gating strategy for the selection of CD4þ cells from wholeblood cells to analyze surface and intracellular T cellmarkers is given in Fig. 2a. The most commonly used sur-face markers to define memory T cell subsets are CD45RA,a protein tyrosine phosphatase regulating src-family ki-nases, and the chemokine receptor CCR7.17e20 Based onthe expression of these surface markers, memory subsetsare classified as naive cells (CD45RAþ CCR7þ) central mem-ory (CD45RA� CCR7þ), effector memory (CD45RA� CCR7�),and terminal effector cells or CD45RAþ cells (CD45RAþ

CCR7�). Expression levels of these phenotypic markers inour cultured cells were measured at baseline (unstimulatedcontrol) and mycobacterial antigen stimulations by poly-chromatic flow cytometry. Antigen specific T cell responsewas calculated by subtracting the unstimulated valuesfrom the test (Test-nil). Density plot is given to representthe memory T cell population in our antigens stimulationsas mentioned19e21 and representative plots from HHC andPTB are given in Fig 2b.

In our 16 h whole blood cultures, CD45RA� CCR7�

effector memory types were predominantly presentcompared to other memory subtypes in both HHC andPTB. The percentage of central memory subtypes (CD45RA�

CCR7þ) was high in HHC compared to the PTB with all anti-gen stimulations except for CFP-10. ManneWhitney U testanalysis showed highest significance for Rv2721c(p Z 0.0004) and followed by Rv2251 (p Z 0.002) whencompared to PTB. The levels of significance for these twonovel antigens were comparatively higher than the immu-nodominant antigen ESAT-6. Based on these observations,we conclude that the antigen specific memory cells arereadily detectable and high in healthy household contacts.The level of significance was low for the standard immuno-dominant antigens ESAT-6 and CFP-10 as these antigensspecific memory cells were also present in PTB (Fig.2c).

Like central memory cells, effector memory T cellsubtypes (CD45RA� CCR7�) were also high in HHC againstall the antigen stimulation used when compared to activeTB, with the exception of CFP-10. But, the significancewas observed only for Rv2251 and Rv2721 (p < 0.01)Fig. 2d. The other two memory cell types, naive (CD45RAþ

CCR7þ) and terminal effector cells (CD45RAþ CCR7�) re-vealed no significant in differences between HHC andPTB, with any of the stimulations used Fig. 2e and f.

Antigen specific single positive Th1 cytokinesecreting cells were higher than multifunctional Tcells in healthy household contacts

Th1 polarized immune responses are considered importantfor anti-TB immunity. Thus, the circulating levels of Th1immunomodulatory cytokines (IFN-g, TNF-a and IL-2) spe-cific to Rv2251 and Rv2721 antigens in addition to immu-nodominant antigens ESAT-6 and CFP-10 were measured. Asobserved by several studies, the circulating levels of Th1cytokines were high in HHCs compared to PTB.

Elevated levels of CD4þ IFN-gþ was observed in HHC forRv2251 and Rv2721 antigen stimulations (p < 0.005) andwith ESAT-6 (p < 0.05) showing the presence of these anti-gen specific IFN-g secreting cells in circulation of HHC. InPTB, the percentage of CD4þ IFN-gþ cells were detectablebut less than HHCs. However, except CFP-10 stimulationhad a high frequency of IFN-g cells in PTB Fig. 3a and b.Like CD4þ IFN-g, the frequency of CD4þ TNF-a cell washigh in HHC compared to PTB with statistical significanceonly for the novel test antigens Rv2251 and Rv2721c(p < 0.05). In PTB, Rv2251 and Rv2721c specific expansionof CD4þ TNF-aþ cell was minimal when compared toESAT-6 and CFP-10 specific CD4þ TNF-aþ Fig. 3c and d. Like-wise, the percentage of IL-2 secreting T cells were high inHHC than PTB and the antigen specific recall responsewas found to be significant (p < 0.05) in Rv2251 andRv2721c antigen stimulations over other stimulations whereno statistical significance was found Fig. 3e and f.

T cells that coexpress IFN-g, TNF-a and IL-2 are desig-nated polyfunctional T cells (PFT) with a possible role inprotection against TB. The responding CD4þ T cells wereclassified as triple positive (IFN-gþ/IL-2þ/TNF-aþ), doublepositive (IFN-gþ/IL-2þ, IFN-gþ/TNF-aþ or TNF-aþ/IL-2þ)or single positive (IFN-gþ, IL-2þ or TNF-aþ). Boolean gateplatform in FlowJo analysis software was used to calculateproportions of PFT and are given in Fig. 4a. As shown inFig. 4b the Rv2251 had potency to induce high levels ofpoly functional T cells in HHCs compared to the PTB(p < 0.05). Although Rv2721c specific poly functional T cellswere high in HHC, this difference was not statistically sig-nificant. ESAT-6 and CFP-10 poly functional T cells weredetectable in both HHC and PTB with no significant increasein either of the study groups. Among the double positivecells, CD4þ that co-expresses IFN-gþ and TNF-aþ werehigh in HHCs with Rv2251 and Rv2721c antigen stimulations(p < 0.05).

Variable antigen specific Th17 responses

Among the analyzed Th17 cytokines, IL-22 was predomi-nantly present in our 16 h whole blood cultures. Baselinevalues did not vary between HHC and PTB for IL-17 and IL-21, whereas IL-22 was present high in PTB at baseline (datanot shown). Elevated levels of IL-17 were observed in HHCafter all antigen stimulations except for CFP-10 and tosome extent in PTB Supplementary Fig. 3a. A similar type ofrecall response was also observed with IL-21 levels in HHCin all stimulations except for Rv2721c Supplementary

460 D. Santhi, A. Raja

Figure 3 Intracellular staining for Th1 cytokines expression. CD4þ T cells were gated as described before. Intracellular Th1cytokines in circulation of healthy household contacts (n Z 22) and pulmonary TB subjects (n Z 17). Frequency of IFN-g secretingCD4þ cells (a), TNF-a secreting CD4þ cells (c) and IL-2 secreting CD4þ cells (e) against all antigenic stimulation is shown and repre-sentative flow diagram from HHC and PTB is given. Box whisker plot showing the percentage of cells secreting IFN-g (b), TNF-a (d)and IL-2 (e) from HHC and PTB is given. The whisker graph illustrates minimum to maximum range and middle line in the box in-dicates median value. p values less than 0.05 was considered as significant.

T-cell response of Rv2251 and Rv2721c 461

Fig. 3b, but both of these cytokines did not vary signifi-cantly between the groups.

IL-22 levels were high in PTB compared to HHC for all theantigen stimulations except for Rv2251 Supplementary

Figure 2 Antigen specific CD4D memory T cell subtypes in HH

acquired in flow cytometry. Single cells were selected by taking FSCthe expression of CD3þ, a specific T cell marker (1 to 3). CD3þ cellsthe expression of memory cell markers. b) Representative flow diaggiven. CD4þ cells were classified according to the expression of CDtive percentage of naive cells (CD45RAþ CCR7þ) central memory (Cminal effector cells (CD45RAþ CCR7�) were determined. The expanagainst Rv2251 and Rv2721c antigens when compared to other antigHHC and PTB indicating their immune reactivity is not compromisednaive cells (e) and terminal effector cells (f) found in 16 h whole bloscatter plots and line in the middle depicts median. Central memoHHC against Rv2251 and Rv2721c stimulations with statistical signifi0.02 by ManneWhitney U test with Holm’s correction. p values les

Fig. 3c. Comparatively high levels of IL-22 were observedfor both the immunodominant antigens ESAT-6 and CFP-10in PTB. Th17 cells secreting both IL-22 and IL-17 (IL-17/22) cytokines were high in HHC for all antigens’ stimulation

C and PTB. a) In vitro stimulated whole blood was stained and-H vs FSC-A and then lymphocyte populations were analyzed forwere analyzed for the expression of CD4 and CD8 (4) to analyseram of memory markers CD45RA and CCR7 from HHC and PTB is45RA and CCR7 into various memory cell subtypes. The respec-D45RA� CCR7þ), effector memory (CD45RA� CCR7�), and ter-sion of central memory cells was high in HHC compared to PTBens stimulations. PHA mitogen control showed response in both. The percentage of central memory (c), effector memory (d),od culture from 22 HHC and 17 PTB subjects are represented asry and effector memory population were found to be higher incance of p < 0.005 for Rv2251 and for Rv2721c and ESAT-6 p <

s than 0.05 was considered as significant.

Figure 3 Continued

462 D. Santhi, A. Raja

compared to PTB. And in a few PTB patients Th17 dual cellswere absent completely with no statistical differences.Supplementary Fig. 3d.

Discussion

Circulating antigen specific T cells in a protected popula-tion of TB would reflect the possible dominant immuneresponse necessary for effective disease control and arewell measured by antigen specific recall responses. Thisapproach has been used for the selection of potentialantigens for anti-TB vaccine development.5 Effective TBvaccines are often aimed at evoking T cell-mediated im-mune responses, since that confers protective immunityto TB.22

To assess the T-lymphocyte recall responses, assaysbased on peripheral blood mononuclear cells (PBMC) andwhole blood (WB) were widely used with the similaroutcome. But, WB assays hold an advantage as compara-tively less blood samples are required than PBMC assays.Hence, we preferred to use WB assays in our study anddiluted the blood sample to minimize volume use as well asto screen many antigens with the same blood sample.

Various blood sample dilutions of 1:1,23 1:2,16 1:5 and1:1024 were used by different studies. Though, a few re-ports have mentioned that peripheral blood might notnecessarily address the immune response at the site ofinfection,25 recent evidence has shown that disease activityat the site of the infection may be reflected in peripheralblood. Blood serves as a reservoir of trafficking immunecells that travel to and from sites of active disease andlymphoid organs. As it is readily accessible, it could be anappropriate test tissue in humans for studying immunemechanisms26e28 and to evaluate vaccine efficacy in termsof T cell response.29e31 But, for a better understanding ofantigen or vaccine mediated protective immunity to M.tuberculosis, a closer look at the immunological events inthe lung is also essential.

The difference between protection and the immunopa-thology of any infectious disease is conferred by the qualityof memory responses.32 The quality of two selected anti-gens Rv2251 and Rv2721c in eliciting memory responsesare demonstrated first in our results. They induced highlevels of central and effector memory phenotypes, predom-inantly in healthy M. tuberculosis infected individuals. Ourobservation suggests that these two antigens could bepossibly associated with latently infected subjects, who

Figure 3 Continued

T-cell response of Rv2251 and Rv2721c 463

are presumed to be protected against active TB disease,and provides the immunological rationale for evaluatingthese antigens for vaccine development against TB. Centralmemory subtypes are highly correlated as mediators of pro-tection (high proliferative capacity and rapidly transform toeffector cells upon re-exposure to antigen) and their pres-ence is often found with latent TB rather than active TB.21

The same was observed in our study. Emerging reports sug-gest the necessity of effector memory subtypes in additionto central memory cells for effective infection control.33

This is well documented from our study where the levelsof effector memory cells were high in HHCs showing theirpossible role in protection against active TB disease. Ourobservations open a new path to explore the effector cyto-kines secreted by Rv2251 and Rv2721c specific memorycells. The two standard antigens ESAT-6 and CFP-10 specificmemory cells were also detected at significant levels in theactive TB population showing their lack of uniqueness tothe latent TB population. Hence, protection against TBfrom these standard antigens might not be guaranteed.

Rv2251 and Rv2721c antigens specific Th1 response IFN-g, TNF-a and IL-2 were also dominant in latently infectedsubjects. Resistance to intracellular pathogens and

macrophage activation is thought to be mediated by IFN-g.34 The role of TNF-a which synergizes with IFN-g for intra-cellular pathogen elimination, is supported by otherstudies.35 TNF-a is also important in orchestrating granu-loma formation, the hallmark of tuberculosis, and is neededfor active disease control. IFN-g is often used as a marker ofprotective immunity against M. tuberculosis infection.36e38

The marked increase in Th1 immune response against thesetwo antigens supports our hypothesis for these antigens aspromising vaccine candidates against TB. Our resultsdemonstrated minimal expansion of antigen specific TNF-a cells in PTB with Rv2251 and Rv2721c, in contrast toESAT-6 and CFP-10, which shows that these antigens arenot predominantly detected by active TB patients hence,these antigens could be latently associated. Compared tothe test antigens (Rv2251 and Rv2721c), CD4þ IFN-gþ levelswere minimal against the standard antigen ESAT-6 stimula-tion. This was already observed with diluted whole blood16

and PBMC39 when ESAT-6 was used at 5 mg/mL with anagreement to our observation.

We showed that frequency of Rv2251 and Rv2721cspecific poly functional T cells were high in HHC comparedto other antigen stimulations with statistical significance

Figure 4 Mycobacterial antigen specific poly functional CD4D Th1 cells. Whole blood cells were cultured with ESAT-6 (E6),CFP-10 (C-10) and Rv2251, Rv2721c mycobacterial antigens and percentage of poly functional T cells were determined. a) Boolean

464 D. Santhi, A. Raja

T-cell response of Rv2251 and Rv2721c 465

only for the former antigen. This could be attributed to theelevated levels of IFN-g and TNF-a with these two antigenicstimulations. These multifunctional T cells are capable of abroader repertoire of T cell functions, hence are associatedwith enhanced protection.40 In humans, high levels of cyto-kine secretion by triple-positive antiviral T cells have beenshown rather than single positive cells.41 Thus, the qualitiesof T cells induced by these two antigens could be superior inproducing of Th1 cytokines, which could be correlated withprotective immunity and the development of a preventivevaccine.

The frequency of Rv2251 and Rv2721c specific Th1cytokine T cell profiles also supports evidence for the Tcell co-expressing either TNF-aþ or TNF-aþ/IFN-gþ. Themaintenance of this population in the late stage of infec-tion was associated with enhanced control of bacterialgrowth.42 Interestingly, these two antigens’ specific doublepositive cells that secrete IFN-gþ/TNF-aþ were high inlatent study subjects. The role of these IFN-gþ/TNF-aþ

secreting cells as a latent marker and as a correlate of pro-tection during PCV2 vaccination43 imply the possible poten-tial of these two antigens to elicit protective immunity.

New evidence supporting the role of Th17 cells invaccine mediated immunity against TB are emerging. Ourstudy implied the detection of IL-17A in both groups, unlikeother studies where insufficient or absent IL-17A wasreported.44 Though, the statistical differences were not sig-nificant, our results showed high antigen specific Th17 re-sponses in HHC suggesting that, Th17 response might beassociated with latent study subjects. Distinctly IL-22levels, despite belonging to Th17 group of cells, werehigh in PTB, but during Rv2251 and Rv271c stimulations.Comparative high levels of IL-22 in PTB were further sup-ported by our analysis on Th17 dual cells where a majorityof the Th17 dual cells were secreting IL-22 rather thansecreting both IL-17 and IL-22. This suggestion of a distinctTh17 response to Rv2251 and Rv2721c antigens are to bestudied further for better understanding.

Antigen specific Th1 response was observed also inperipheral blood of active TB patients and used for TBdiagnosis.45e47 Increase in Th1 cell response during MVA85A(modified Vaccinia Ankara virus expressing antigen 85A)vaccine trial did not confer protection against M. tubercu-losis infection. But, MVA85A vaccine efficacy trial was car-ried in infants aged 4e6 months and their immune system isstill immature, but it could protect adults against PTB.48

These reports suggested Th1 response alone might not besufficient for controlling TB infection and contradict the hy-pothesis of correlating antigen specific Th1 response in pe-ripheral blood for protection. Notably, these observationsused ESAT-6, CFP-10 and Ag85A standard antigens of M.tuberculosis to assess peripheral blood Th1 response. Thissupports our observation that ESAT-6 and CFP-10 specificT cell response is not unique to latency, but also to TB,hence novel antigens of M. tuberculosis that are highly spe-cific to latency might be needed for assessing their immune

gating strategy to assess the co-expression of Th1 cytokines is givenpositive (IFN-g/IL-2, IFN-g/TNF-a or TNF-a/IL-2) Th1 cytokines in peManneWhitney U test and p values less than 0.05 was considered

status and to correlate it with protection against active TBdisease.

The lack of reliable correlates to immune protectionduring TB infection or a biomarker to predict vaccineefficacy poses challenges for TB vaccine development. Foroptimal protection against TB CD4þ T cells, effector cyto-kines such as IFN-g and TNF-a are requisite, but are not anexclusive component of protective immunity.49e51 Majorhistocompatibility complex (MHC) class II-restricted CD4þ

T cells play an essential role in protective immunityagainst M. tuberculosis. Identification of immunogenic Tcell epitopes is also essential for the design of peptide/protein-based vaccines. Our earlier in silico analysis pre-dicted high class I and II MHC binding affinity (97% and100%, respectively) and a high percentage of populationcoverage for these two novel antigens which were compar-atively higher than the standard antigens, ESAT-6 and CFP-10.11

Taken together, the first comprehensive analysis ofmycobacterial antigens Rv2251 and Rv2721c from humandonors showed dominant central and effector memorysubtypes, with Th1 dominant response alluding to itsimportance for vaccine development. The frequency andphenotypes of memory cells against these antigens are welldocumented in our study. Assessing the quality of theseantigens specific memory cells for their effector cytokinesand whether the T cell responses against these two antigensis sufficient to control the TB infection are yet to bestudied. Further evidence from large study populations,animal based experiments to evaluate the efficacy of thesetwo antigens, and tolerance assessments would shed morelight on the potential of these two antigens in protectionagainst the TB infection.

Acknowledgments

The authors wish to thank all the study subjects whoparticipated in this study. Authors also thank the cliniciansof Govt. Thiruvatteeswarar Hospital of ThoracicMedicine (GTHTM) hospital, Otteri, Chennai, India. SanthiDevasundaram expresses her gratitude to Indian Council ofMedical Research (ICMR), New Delhi, India for providingSenior Research Fellowship. ESAT-6, CFP-10 recombinantproteins were kind gift from Colorado State University. Theauthors wish to thank Dr. Subash Babu, National Institutesof HealthdNIRTdInternational Center for Excellence inResearch, Chennai, India for allowing to use their FACSfacility.

Appendix A. Supplementary data

Supplementary data related to this article can be found athttp://dx.doi.org/10.1016/j.jinf.2016.06.012.

. b) The frequency of triple positive IFN-g/IL-2/TNF-a and dualrcentages are shown as bar graphs. p values were calculated byas significant.

466 D. Santhi, A. Raja

Competing interests

The authors declare that they have no competing interests.

References

1. McShane H. Tuberculosis vaccines: beyond bacille Calmet-teeGuerin. Philosophical Trans R Soc Lond Ser B Biol Sci2011;366(1579):2782e9.

2. Baily GV. Tuberculosis prevention trial, Madras. Indian J MedRes 1980;72(Suppl.):1e74.

3. Dietrich J, Andersen C, Rappuoli R, Doherty TM, Jensen CG,Andersen P. Mucosal administration of Ag85B-ESAT-6 protectsagainst infection with Mycobacterium tuberculosis and boostsprior bacillus CalmetteeGuerin immunity. J Immunol 2006;177(9):6353e60.

4. Kaufmann SH. How can immunology contribute to the controlof tuberculosis. Nat Rev Immunol 2001;1:20e30.

5. Bertholet S, Ireton GC, Kahn M, Guderian J, Mohamath R,Stride N, et al. Identification of human T cell antigens forthe development of vaccines against Mycobacterium tubercu-losis. J Immunol 2008;181(11):7948e57.

6. Horwitz MA, Lee BW, Dillon BJ, Harth G. Protective immunityagainst tuberculosis induced by vaccination with major extra-cellular proteins of Mycobacterium tuberculosis. Proc NatlAcad Sci USA 1995;92:1530e4.

7. Weinrich Olsen A, van Pinxteren LA, Meng Okkels L, BirkRasmussen P, Andersen P. Protection of mice with a tubercu-losis subunit vaccine based on a fusion protein of antigen 85band esat-6. Infect Immun 2001;69:2773e8.

8. Luo Y, Wang B, Hu L, Yu H, Da Z, Jiang W, et al. Fusion proteinAg85B-MPT64 (190-198)-Mtb8.4 has higher immunogenicitythan Ag85B with capacity to boost BCG-primed immunityagainst Mycobacterium tuberculosis in mice. Vaccine 2009;27:6179e85.

9. Byun EH, Kim WS, Kim JS, Jung ID, Park YM, Kim HJ, et al.Mycobacterium tuberculosis Rv0577, a novel TLR2 agonist, in-duces maturation of dendritic cells and drives Th1 immuneresponse. FA Seb J 2012;26:2695e711.

10. Arlehamn CS, Sidney J, Henderson R, Greenbaum JA,James EA, Moutaftsi M, et al. Dissecting mechanisms of immu-nodominance to the common tuberculosis antigens ESAT-6,CFP10, Rv2031c (hspX), Rv2654c (TB7.7), and Rv1038c(EsxJ). J Immunol 2012;188(10):5020e31.

11. Devasundaram S, Deenadayalan A, Raja A. In silico analysis ofpotential human T Cell antigens from Mycobacterium tuber-culosis for the development of subunit vaccines against tuber-culosis. Immunol Invest 2014;43(2):137e59.

12. Kollmann TR. Variation between populations in the innate im-mune response to vaccine adjuvants. Front Immunol 2013;4:81. http://dx.doi.org/10.3389/fimmu.2013.00081.

13. Song H, Sandie R, Wang Y, Andrade-Navarro MA,Niederweis M. Identification of outer membrane proteins ofMycobacterium tuberculosis. Tuberculosis (Edinb) 2008;88(6):526e44.

14. de Souza GA, Leversen NA, Malen H, Wiker HG. Bacterial pro-teins with cleaved or uncleaved signal peptides of the generalsecretory pathway. J Proteomics 2011;75(2):502e10.

15. Kunnath-Velayudhan S, Salamon H, Wang HY, Davidow AL,Molina DM, Huynh VT, et al. Dynamic antibody responses tothe Mycobacterium tuberculosis proteome. Proc Natl AcadSci USA 2010;107(33):14703e8.

16. Kumar M, Meenakshi N, Sundaramurthi JC, Kaur G, Mehra NK,Raja A. Immune response to Mycobacterium tuberculosis spe-cific antigen ESAT-6 among south Indians. Tuberculosis 2010;90(1):60e9.

17. Takahashi A, Hanson MG, Norell HR, Havelka AM, Kono K,Malmberg KJ, et al. Preferential cell death of CD8þ effectormemory (CCR7eCD45RA-) T cells by hydrogen peroxide-induced oxidative stress. J Immunol 2005;174(10):6080e7.

18. D’Asaro M, Dieli F, Caccamo N, Musso M, Porretto F, Salerno A.Increase of CCR7- CD45RAþ CD8 T cells (T(EMRA)) in chronicgraft-versus-host disease. Leukemia 2006;20(3):545e7.

19. Mojumdar K, Vajpayee M, Chauhan NK, Singh A, Singh R,Kurapati S. Loss of CD127 & increased immunosenescence ofT cell subsets in HIV infected individuals. Indian J Med Res2011;134(6):972e81.

20. Dintwe OB, Day CL, Smit E, Nemes E, Gray C, Tameris M, et al.Heterologous vaccination against human tuberculosis modu-lates antigen-specific CD4þ T-cell function. Eur J Immunol2013;43(9):2409e20.

21. Lindestam Arlehamn CS, Gerasimova A, Mele F, Henderson R,Swann J, Greenbaum JA, et al. Memory T cells in latent Myco-bacterium tuberculosis infection are directed against threeantigenic islands and largely contained in a CXCR3þCCR6þTh1 subset. PLoS Pathog 2013;9(1):e1003130.

22. Flynn JL, Chan J. Immunology of tuberculosis. Annu Rev Im-munol 2001;19:93e129.

23. Kumar NP, Sridhar R, Banurekha VV, Jawahar MS, Nutman TB,Babu S. Expansion of pathogen-specific T-helper 1 and T-help-er 17 cells in pulmonary tuberculosis with coincident type 2diabetes mellitus. J Infect Dis 2013;208(5):739e48.

24. Deenadayalan A, Maddineni P, Raja A. Comparison of wholeblood and PBMC assays for T-cell functional analysis. BMCRes Notes 2013;6:120.

25. Tully G, Kortsik C, Hohn H, Zehbe I, Hitzler WE, Neukirch C,et al. Highly focused T cell responses in latent human pulmo-nary Mycobacterium tuberculosis infection. J Immunol 2005;174:2174e84.

26. Beamer GL, Flaherty DK, Vesosky B, Turner J. Peripheralblood gamma interferon release assays predict lung responsesand Mycobacterium tuberculosis disease outcome in mice.Clin Vaccine Immunol 2008;15:474e83.

27. BerryMP,GrahamCM,McNabFW,Xu Z, Bloch SA,Oni T, et al. Aninterferon-inducible neutrophil-driven blood transcriptionalsignature in human tuberculosis. Nature 2010;466:973e7.

28. Cliff JM, Kaufmann SH, McShane H, van Helden P, O’Garra A.The human immune response to tuberculosis and its treat-ment: a view from the blood. Immunol Rev 2015;264:88e102.

29. Harris SA, Meyer J, Satti I, Marsay L, Poulton ID, Tanner R,et al. Evaluation of a human BCG challenge model to assessantimycobacterial immunity induced by BCG and a candidatetuberculosis vaccine, MVA85A, alone and in combination. JInfect Dis 2014;209:1259e68.

30. Xu Y, Liu W, Shen H, Yan J, Yang E, Wang H. RecombinantMycobacterium bovis BCG expressing chimaeric protein ofAg85B and ESAT-6 enhances immunostimulatory activity of hu-man macrophages. Microbes Infect 2010;12:683e9.

31. Lahey T, Sheth S, Matee M, Arbeit R, Horsburgh CR, Mtei L,et al. Interferon gamma responses to mycobacterial antigensprotect against subsequent HIV-associated tuberculosis. JInfect Dis 2010;202:1265e72.

32. Prezzemolo T, Guggino G, La Manna MP, Di Liberto D, Dieli F,Caccamo N. Functional signatures of human CD4 and CD8 Tcell responses to Mycobacterium tuberculosis. Front Immunol2014;5:180.

33. Farber DL, Ahmadzadeh M. Dissecting the complexity of thememory T cell response. Immunol Res 2002;25(3):247e59.

34. Fenton MJ, Vermeulen MW, Kim S, Burdick M, Strieter RM,Kornfeld H. Induction of gamma interferon production in hu-man alveolar macrophages by Mycobacterium tuberculosis.Infect Immun 1997;65(12):5149e56.

35. Kannanganat S, Ibegbu C, Chennareddi L, Robinson HL,Amara RR. Multiple-cytokine-producing antiviral CD4 T cells

T-cell response of Rv2251 and Rv2721c 467

are functionally superior to single-cytokine-producing cells. JVirol 2007;81(16):8468e76.

36. Beveridge NE, Fletcher HA, Hughes J, Pathan AA, Scriba TJ,Minassian A, et al. A comparison of IFN gamma detectionmethods used in tuberculosis vaccine trials. Tuberculosis2008;88(6):631e40.

37. Pollock L, Basu Roy R, Kampmann B. How to use: interferongamma release assays for tuberculosis. Archives Dis ChildEduc Pract Ed 2013;98(3):99e105.

38. Thillai M, Pollock K, Pareek M, Lalvani A. Interferon-gammarelease assays for tuberculosis: current and future applica-tions. Expert Rev Respir Med 2014;8(1):67e78. http://dx.doi.org/10.1586/17476348.2014.852471.

39. Cardoso FL, Antas PR, Milagres AS, Geluk A, Franken KL,Oliveira EB, et al. T-cell responses to the Mycobacteriumtuberculosis-specific antigen ESAT-6 in Brazilian tuberculosispatients. Infect Immun 2002;70(12):6707e14.

40. Seder RA, Darrah PA, Roederer M. T-cell quality in memoryand protection: implications for vaccine design. Nat Rev Im-munol 2008;8:247e58.

41. Precopio ML, Betts MR, Parrino J, Price DA, Gostick E,Ambrozak DR, et al. Immunization with vaccinia virus inducespolyfunctional and phenotypically distinctive CD8(þ) T cellresponses. J Exp Med 2007;204:1405e16.

42. Nemeth J, Winkler HM, Karlhofer F, Selenko-Gebauer N,Graninger W, Winkler S. T cells co-producing Mycobacteriumtuberculosis-specific type 1 cytokines for the diagnosis oflatent tuberculosis. Eur cytokine Netw 2010;21(1):34e9.

43. Koinig HC, Talker SC, Stadler M, Ladinig A, Graage R,Ritzmann M, et al. PCV2 vaccination induces IFN-gamma/TNF-alpha co-producing T cells with a potential rolein protection. Vet Res 2015;46:20.

44. Matthews K, Wilkinson KA, Kalsdorf B, Roberts T, Diacon A,Walzl G, et al. Predominance of interleukin-22 overinterleukin-17 at the site of disease in human tuberculosis.Tuberculosis 2011;91(6):587e93.

45. Jafari C, Ernst M, Kalsdorf B, Greinert U, Diel R, Kirsten D,et al. Rapid diagnosis of smear-negative tuberculosis by bron-choalveolar lavage enzyme-linked immunospot. Am J RespirCrit Care Med 2006;174:1048e54.

46. Nemeth J, Winkler HM, Zwick RH, Muller C, Rumetshofer R,Boeck L, et al. Peripheral T cell cytokine responses for diag-nosis of active tuberculosis. PLoS One 2012;7:e35290.

47. Goletti D, Vincenti D, Carrara S, Butera O, Bizzoni F,Bernardini G, et al. Selected RD1 peptides for active tubercu-losis 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.

48. Tameris MD, Hatherill M, Landry BS, Scriba TJ, Snowden MA,Lockhart S, et al. Safety and efficacy of MVA85A, a new tuber-culosis vaccine, in infants previously vaccinated with BCG: arandomised, placebo-controlled phase 2b trial. Lancet 2013;381:1021e8.

49. Kaveh DA, Carmen Garcia-Pelayo M, Hogarth PJ. PersistentBCG bacilli perpetuate CD4 T effector memory and optimalprotection against tuberculosis. Vaccine 2014;32:6911e8.

50. Serbina NV, Lazarevic V, Flynn JL. CD4(þ) T cells are requiredfor the development of cytotoxic CD8(þ) T cells during Myco-bacterium tuberculosis infection. J Immunol 2001;167:6991e7000.

51. Cooper AM. Cell-mediated immune responses in tuberculosis.Annu Rev Immunol 2009;27:393e422.