INFECTION AND IMMUNITY, Feb. 2011, p. 716–723
Copyright © 2011, American Society for Microbiology. All Rights Reserved.
Vol. 79, No. 2
Three Protein Cocktails Mediate Delayed-Type Hypersensitivity Responses
Indistinguishable from That Elicited by Purified Protein Derivative in the
Guinea Pig Model of Mycobacterium tuberculosis Infection?
Hongliang Yang,1JoLynn Troudt,1Ajay Grover,1Kimberly Arnett,1Megan Lucas,1Yun Sang Cho,1,2
Helle Bielefeldt-Ohmann,3Jennifer Taylor,1Angelo Izzo,1and Karen M. Dobos1*
Colorado State University, Department of Microbiology, Immunology, and Pathology, Fort Collins, Colorado 805231; Bacteriology and
Parasitology Division, National Veterinary Research and Quarantine Service, Anyang 430-824, Republic of Korea2; and
School of Veterinary Science, University of Queensland, Gatton Campus, Gatton Qld 4343, Australia3
Received 10 May 2010/Returned for modification 29 June 2010/Accepted 21 November 2010
Purified protein derivative (PPD) is a widely used reagent for the diagnosis of Mycobacterium tuberculosis
infection. Recently, the molecular composition of PPD was defined, with hundreds of mycobacterial protein
representatives making up PPD. Which, if any, of these specific products drive the potency of PPD remains in
question. In this study, two proteins (DnaK and GroEL2) previously identified as dominant proteins in PPD
were tested for the capacity to induce delayed-type hypersensitivity (DTH) responses in H37Rv-infected or
BCG-vaccinated guinea pigs. These two proteins were used in pull-down assays to identify interacting PPD
products. Six proteins were identified as interacting partners with DnaK and GroEL2, i.e., Rv0009, Rv0475,
Rv0569, Rv0685, Rv2626c, and Rv2632c. These six proteins were tested alone and in combination with DnaK
and GroEL2 for the capacity to induce a DTH response in the guinea pig model. From these studies, two
cocktails, DnaK/GroEL2/Rv0009 and DnaK/GroEL2/Rv0685, were found to induce DTH responses in H37Rv-
infected or BCG-vaccinated guinea pigs that were indistinguishable from DTH responses driven by a PPD
injection. The mechanism by which DTH responses were induced was elucidated by histologic examination,
analysis of activated CD4?/CD8?T cells, and cytokine mRNA expression at the site of the DTH response. PPD
and the protein cocktails tested induced strong DTH responses in H37Rv-infected guinea pigs. Ex vivo
phenotyping of T cells at the DTH site indicated that this response is mediated by activated CD4?and CD8?
T cells, with increases in gamma interferon and tumor necrosis factor alpha, but not interleukin-10, at the site
of the DTH response. Our results demonstrate for the first time that the PPD response can be mimicked at the
molecular level with defined protein cocktails. The use of this defined product will allow a more thorough
understanding of the DTH response and may provide a platform for more rapid and sensitive second-
generation skin test reagents for the diagnosis of M. tuberculosis infection.
Tuberculosis remains one of the largest single causes of
disease and death from an infectious agent, particularly in
developing nations. With nearly 2 million deaths each year and
an estimated 8 to 9 million new cases annually, tuberculosis is
the second leading cause of death worldwide among commu-
nicable diseases (8). Further, it is estimated that one-third of
the world’s population is latently infected with Mycobacterium
tuberculosis (9). The epidemic of human immunodeficiency
virus infection (27) and the increase in the incidence of mul-
tidrug-resistant M. tuberculosis (10) have further complicated
tuberculosis control efforts.
The most common way persons infected with M. tuberculosis
are identified is through use of the tuberculin skin test (TST).
Over the last century, the TST has been proven to be an impor-
tant diagnostic tool for tuberculosis infection (25, 35). It is also an
important component of epidemiologic studies to evaluate the
prevalence of latent tuberculosis in various populations.
The TST, which is also called the Mantoux test after the
French physician Charles Mantoux (1877 to 1947), was devel-
oped by Florence B. Seibert (1897 to 1991) (33). The TST is a
modification of Koch’s old tuberculin, resulting in purified
protein derivative (PPD). Several PPD standards exist, includ-
ing PPD-S (1, 16), the first FDA standard; PPD-RT23 (7, 30),
the standard produced by the Statens Serum Institut; and
PPD-S2 (37), the new FDA standard of PPD. The basic meth-
ods used to produce these PPD standards are similar. Proteins
are purified by repeated precipitation with trichloroacetic acid,
ammonium sulfate, or both after cultures of M. tuberculosis are
grown to stationary phase and sterilized (34). The preparations
are dominated by M. tuberculosis proteins and peptides and
contain few polysaccharides, nucleic acids, and lipids. The elu-
cidation of the molecular composition of PPD demonstrated
that PPD is a complex mixture of hundreds of denatured an-
tigens (3, 19; Y. S. Cho et al., unpublished data). It is not
surprising, then, that little is known about the active compo-
nents of PPD responsible for its potency. The presence of so
many antigens likely contributes to the poor specificity ob-
served with PPD testing; it cannot differentiate active tubercu-
losis from latent infection or vaccination, complicating TST
use as a tool for epidemiologic studies and diagnosis of tuber-
culosis. PPD from different sources may vary in the delayed-
type hypersensitivity (DTH) response because there is no way
to standardize the hundreds of antigenic components in PPD
* Corresponding author. Mailing address: 1682 Campus Delivery,
Department of Microbiology, Immunology, and Pathology, Colorado
State University, Fort Collins, CO 80523-1682. Phone: (970) 491-6549.
Fax: (970) 491-1815. E-mail: firstname.lastname@example.org.
?Published ahead of print on 6 December 2010.
(23). In addition to issues arising from the number of proteins
present in PPD, three of the most dominant proteins in PPD,
Rv3418 (GroES), Rv0440 (GroEL2), and Rv0350 (DnaK) (3,
19; Cho et al., unpublished), are highly conserved chaperones;
reactivity to these proteins may also explain the diminished
specificity of PPD. Lastly, the complexity of PPD limits our
capacity to tease out the molecular interactions involved in the
DTH response and, in particular, understand which of these
interactions, if any, discern chronic infections from protective
immunity. Skin test reagents using defined antigens may be
able to be used to address these questions.
Over the past 20 years, major efforts have focused on the
identification of highly purified M. tuberculosis recombinant
proteins that may be used to develop a more specific diagnostic
skin test (31). Some protein candidates, including ESAT-6
(28), CFP10 (39), MPT64 (24), DPPD (6), and Rv0934 (17),
were tested and found to induce strong DTH reactions in
guinea pigs, but none of them were developed as second-
generation skin test reagents. Several studies indicate that a
combination of several purified antigens (antigen cocktails) or
a combination of several DTH-inducing epitopes may be re-
quired for second-generation skin test reagents to effectively
replace PPD (21, 24, 32).
The definition of the molecular composition of PPD allows
for the development of a more refined product. Recently, four
heat shock proteins (GroES, GroEL2, HspX, and DnaK) were
identified as dominant products in PPD-S2, the U.S. FDA
standard PPD (Cho et al., unpublished). In this study, DnaK
and GroEL2, two of the major chaperones in PPD-S2 (Cho et
al., unpublished) and bovine and avian PPD (3), were tested
for the capacity to induce a DTH response and were found to
elicit DTH responses in H37Rv-infected and BCG-vaccinated
guinea pigs. As these proteins are conserved among other
mycobacteria, their use in a second-generation TST may be
limited. Thus, these proteins were used to pull out additional
protein representatives from PPD, resulting in the identifica-
tion of six highly interacting proteins. When these proteins
were tested in various formulations for the potential to elicit a
DTH response, two cocktails, DnaK/GroEL2/Rv0685 and
DnaK/GroEL2/Rv0009, were found to induce DTH responses
that were indistinguishable from that induced by PPD in the
guinea pig model of tuberculosis, as determined by skin test
responses and at the molecular level.
MATERIALS AND METHODS
Expression of recombinant proteins in Escherichia coli. The four chaperones,
DnaK (Rv0350), GroEL2 (Rv0440), GroES (Rv3418c), and HspX (Rv2031c),
and M. tuberculosis protein Rv2626c were expressed using plasmids available
from the Tuberculosis Vaccine Testing and Research Material contract
Gateway cloning-based (Invitrogen, Carlsbad, CA) entry plasmids containing the
open reading frames for M. tuberculosis proteins Rv0009, Rv0475, Rv0569, Rv0685,
Rv2626c, and Rv2632c were obtained from the Pathogen Functional Genomics
Resource Center (http://pfgrc.jcvi.org/fir_index.php/fir/available_libraries.html). LR
recombination between each entry plasmid and the pDEST17 destination vector
(Invitrogen), thus creating an expression plasmid for the gene of interest. The
resulting expression clones were transformed to DH5? competent cells (Invitrogen).
The sequence fidelity of expression clones was confirmed by DNA sequencing
(Proteomics and Metabolomics Facility, Colorado State University).
E. coli BL21(DE3)pLysS or BL21(DE3) host cells (Invitrogen) were trans-
formed with the expression clones. The pDEST17 destination vector contains an
N-terminal His tag allowing for affinity purification of expressed proteins. Re-
combinant His-tagged proteins were expressed following induction with IPTG
(isopropyl-?-D-thiogalactopyranoside; Calbiochem, Darmstadt, Germany) and
then purified by affinity chromatography using nickel (Ni)-charged His-bind resin
(Novagen, Darmstadt, Germany). Resin containing bound recombinant protein
was washed with 10 mM Tris-HCl containing 0.5% ASB-14 (Calbiochem, Darm-
stadt, Germany) to remove endotoxin prior to elution of proteins with imidazole.
Purified proteins were dialyzed against 10 mM ammonium bicarbonate overnight
at 4°C. The purity of recombinant proteins was assessed by sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by silver
staining. The identification of His-tagged recombinant proteins was performed
by Western blotting with mouse monoclonal IgG1 anti-DnaK and anti-GroEL2
primary antibodies (IT41 and IT13, respectively, provided by the Tuberculosis
Vaccine Testing and Research Materials contract) or mouse monoclonal IgG1
Penta-His antibody (Qiagen, Valencia, CA). The reaction was detected with anti-
mouse antibody conjugated to allophycocyanin, followed by development with 5-
MO). The concentration of recombinant proteins was determined with the bicin-
choninic acid protein assay (Pierce, Rockford, IL). The endotoxin level for all the
lysate QCL-1000 kit (Cambrex, Walkersville, MD), performed according to the
Identification of GroEL2 and DnaK interacting proteins. Purified recombi-
nant DnaK or GroEL2 or a mixture of the two proteins was used in pull-down
assays to identify protein representatives in PPD which bound to DnaK and
GroEL2. Briefly, 0.5 mg of recombinant DnaK, GroEL2, or a 1:1 mixture of
DnaK and GroEL2 was incubated with 5.0 mg of PPD overnight at 4°C. The
protein-PPD mixture was then passed separately over Ni2?affinity resin (His
band Quick 300 cartridges; Novagen, Darmstadt, Germany) to capture the re-
combinant protein and any PPD products bound to it, the resin was washed
extensively with 10 mM Tris-HCl to remove nonspecifically bound material, and
the bound material was eluted with imidazole as described above.
Mass spectrometry. The pull-down assay-eluted samples were reduced, alkyl-
ated, trypsin digested, and dissolved in 5% acetonitrile–0.1% acetic acid for
protein identification by liquid chromatography-mass spectrometry (MS) as de-
scribed previously (22). Tandem mass spectra were extracted, charge state de-
convoluted, and deisotoped by BioWorks version 3.2. All tandem MS (MS/MS)
samples were analyzed using Sequest (version 2.7; ThermoFinnigan, San Jose,
CA) set up to search the M. tuberculosis database (GenBank accession no.
AL123456, R9, 3913 entries) assuming the digestion enzyme trypsin and a max-
imum of four missed cleavages. Sequest was searched with a fragment ion mass
tolerance of 1.00 Da and a parent ion tolerance of 2.5 Da. Scaffold (version
Scaffold_2_05_02; Proteome Software Inc., Portland, OR) was used to validate
MS/MS-based peptide and protein identifications. Peptide identifications were
accepted if they could be established at greater than 95.0% probability as spec-
ified by the Peptide Prophet algorithm. Protein identifications were accepted if
they could be established at greater than 90.0% probability and contained at least
two identified peptides.
Guinea pig model. Specific-pathogen-free, outbred Hartley guinea pigs (450 to
500 g at the beginning of the experiment) from Charles River Laboratories
(Wilmington, MA) were maintained under biosafety level 3 conditions. Guinea
pigs were randomly assigned to one of three groups: BCG vaccinated, H37Rv
infected, or negative control. All animal experiment procedures were approved
by the Animal Care and Use Committee of Colorado State University.
BCG vaccination and aerosol challenge. Vaccinations were performed by
subcutaneous injection of 0.1 ml (103CFU) of BCG Pasteur.
Guinea pigs were infected using a Madison chamber aerosol generation device
to deliver M. tuberculosis H37Rv at a low aerosol dose of 20 bacilli. Subcutaneous
injection with 100 ?l endotoxin-free phosphate-buffered saline (PBS) was used as
a negative control.
Skin testing and biopsy. BCG-vaccinated or H37Rv-infected guinea pigs were
examined for a DTH response induced by different formulations (PPD-S2, DnaK,
GroEL2, GroES, HspX, DnaK/GroEL2, DnaK/GroEL2/Rv0009, DnaK/GroEL2/
Rv0475, DnaK/GroEL2/Rv0569, DnaK/GroEL2/Rv0685, DnaK/GroEL2/Rv2626c,
DnaK/GroEL2/Rv2632c, and PBS as a negative control). The skin test, according to
the Mantoux technique, was performed 10 weeks after aerosol infection or BCG
vaccination. The hair on the ventral side was shaved off, and the guinea pigs were
injected intradermally with 0.1 ml antigen solution diluted in endotoxin-free PBS.
The PPD-S2 and formulations containing single proteins were used at a dose of 0.4
?g. The DnaK/GroEL2 and three-protein formulations all used 0.13 ?g of each
Erythema and induration were measured with a micrometer after 24 and 48 h.
VOL. 79, 2011NOVEL SECOND-GENERATION TUBERCULOSIS SKIN TEST REAGENT717
The diameter (in millimeters) was defined as the mean of vertical and transverse
Skin samples were obtained from infected group guinea pigs. The guinea pigs
were euthanized by injection with sodium pentobarbital (Sleepaway; Fort Dodge
Laboratories, Inc., Fort Dodge, IA) 48 h after antigen stimulation. The injection
site skin was rinsed with 70% ethanol, removed, and then fixed with 10%
formalin for histological examination. Skin samples were prepared for RNA
extraction (30- to 50-mg skin sample in 0.6 ml RNAlater [Qiagen]) or cell
isolation (the same area of skin sample in 1.5 ml of incomplete RPMI medium).
The samples for RNA extraction were homogenized in RNAlater and immedi-
ately stored at ?20°C. The samples for cell isolation were kept on ice and used
immediately after biopsy.
Histological examination. Following fixation in formalin, the skin tissues were
routinely processed for paraffin embedding. Sections were cut at a 4- to 5-?m
thickness, stained with hematoxylin and eosin (H&E), and examined on a Nikon
Eclipse 50i microscope. Microphotographs were taken with a DS-Fi1 camera and
associated software (Nikon, Japan).
Flow cytometric analysis of cell surface markers. Flow cytometry was per-
formed to investigate the subgroups of T cells in infected guinea pigs stimulated
with different formulations. Briefly, after the removal of subcutaneous fat and
muscle tissues by scraping, the skin samples were minced into small pieces and
digested for 1 h at 37°C with 2 mg/ml collagenase type IV and 1.2 U/ml dispase
(Worthington, Lakewood, NJ). The digested skin samples were further disrupted
by gently pushing the tissue through a 70-?m cell strainer (BD Biosciences, San
Jose, CA). Red blood cells were lysed with ACK buffer (0.14 M NH4Cl, 1.0 mM
KHCO3, pH 7.2 to 7.4) and then washed and resuspended in complete RPMI
medium. The viability of digested skin cells was determined by trypan blue
staining. Single-cell suspensions from each individual guinea pig were stained
with fluorescently labeled monoclonal antibodies against CD4 (clone FITC CT7;
Serotec Inc., Raleigh, NC) (36), CD8 (clone FITC CT6; Serotec Inc.) (36), pan-T
cells (clone APC CT5; Serotec Inc.) (36), and CD62L (clone PE lam1-116; Santa
Cruz Biotechnology, Santa Cruz, CA) (13) at 4°C for 30 min. Antibodies were
used at 0.2 ?g/106cells. Cells were gated on lymphocytes by forward and side
scatter according to their characteristic scatter profiles and further gated based
on pan-T and CD4 or CD8 expression.
All analyses were performed with an acquisition of at least 100,000 events
on a Becton Dickinson FACScalibur flow cytometer (BD Biosciences, San
Jose, CA), and the data were analyzed using FlowJo software (Tree Star Inc.,
Total RNA isolation and real-time quantitative reverse transcription (RT)-
PCR assay. To determine the quantities of cytokine mRNA in skin samples,
real-time RT-PCR was performed for the H37Rv-infected guinea pig group.
First, the skin samples were homogenized and released skin RNA was extracted
using the RNeasy mini kit with columns (Qiagen, Valencia, CA). Contaminating
DNA was removed by on-column treatment with RNase-free DNase (Qiagen,
Valencia, CA) according to the manufacturer’s instructions. The yield and qual-
ity of total RNA were examined using an ND1000 spectrophotometer (Nano-
Drop Technologies, Wilmington, DE). The RNA was then stored at ?80°C until
required for gene expression studies. RT was performed using an iScript cDNA
synthesis kit (Qiagen, Valencia, CA). Real-time RT-PCR, performed with iQ
CYBR green supermix (Bio-Rad, Hercules, CA), was carried out using primers
for guinea pig ?-actin, gamma interferon (IFN-?), tumor necrosis factor alpha
(TNF-?), and interleukin-10 (IL-10) as described previously (15). Results were
recorded as relative gene expression after normalization for ?-actin threshold
cycle (CT) values by the 2???CTmethod (20). All real-time RT-PCR data are
presented such that the values for stimulated samples using different formula-
tions are expressed relative to the value for the PBS-stimulated site on the same
Statistical analysis. Descriptive analyses were carried out for histopathologi-
cal features. All other data were expressed as the mean ? the SEM (standard
error of the mean). The DTH response data were analyzed by one-tailed Student
t test. The Mann-Whitney test was used to characterize changes in cytokine
mRNA expression and cell surface markers between groups of guinea pigs. All
statistical comparisons were carried out using GraphPad Prism version 4.00
(GraphPad Software, San Diego, CA). P values of ?0.05 were considered sta-
DTH responses of the dominant chaperones in PPD. Four
chaperones (DnaK, GroEL2, GroES, and HspX) were ex-
pressed as recombinant proteins in E. coli BL21(DE3)pLysS
cells and purified using Ni-charged His-bind resin (data not
shown). M. tuberculosis-infected, BCG-vaccinated, and naïve
guinea pigs were used to evaluate DTH responses to these
chaperones, with PPD and saline as positive and negative con-
trols, respectively. The results demonstrated that among the
four recombinant proteins used, only DnaK and GroEL2 elic-
ited DTH responses ranging from weak to significant (one
study demonstrated mean responses of 3.36 ? 4.53 and
10.04 ? 2.29 mm for DnaK and GroEL2, respectively). All
four chaperones were combined in various formulations and
tested for the capacity to induce a DTH response. Only the
combination of DnaK and GroEL2 augmented the DTH re-
sponse in infected and vaccinated animals (one study demon-
strated mean responses of 10.68 ? 2.32 mm for DnaK/GroEL2
versus 9.46 ? 1.83 mm for PPD). None of the naïve guinea pigs
developed a reaction to any of the antigens (data not shown).
Dose-response assays were performed with injection doses
ranging from 0.1 to 1.2 ?g total protein/site. Positive DTH
responses were observed for all doses, with 0.25 to 0.4 ?g total
protein/site being the optimum dose (data not shown).
Pull-down assays. Ni chromatography was used to pull down
protein representatives and peptides in PPD that bound to
recombinant DnaK and GroEL2. From this, DnaK and
GroEL2 were found to interact with each other, and six addi-
tional protein representatives were identified in the pull-down
assay elution buffer, indicating an interaction with these two
chaperones (Table 1). All six proteins are small-size proteins
with native molecular masses ranging from 9.5 to 22 kDa,
except Rv0685, with a molecular mass of 44 kDa. As expected
from previous SDS-PAGE analyses of PPD (3; Cho et al.,
unpublished), only recombinant DnaK and GroEL2 and none
of the six new products were observed as discrete protein
bands upon SDS-PAGE analysis of eluted samples (data not
shown), due to the fact that PPD is composed mostly of protein
representatives or aggregates of partially denatured proteins
and peptides rather than discrete protein molecules (3; Cho et
DTH responses to protein cocktails. All of the proteins
identified as interacting partners in the pull-down assay were
expressed and purified as soluble recombinant products, ex-
cept Rv0685, which was purified from inclusion bodies. To
establish the diagnostic potential of these proteins, we evalu-
ated DTH responses to these proteins, individually and com-
TABLE 1. Proteins identified in PPD-S2 through DnaK and
GroEL2 pull-down assay
Rv0009 Probable iron-regulated peptidyl-
prolyl cis-trans isomerase A
Probable chaperone protein DnaK
60-kDa chaperone protein GroEL2
Elongation factor Tu
Rv2626c Hypothetical protein
Rv2632c Hypothetical protein
718YANG ET AL.INFECT. IMMUN.
bined with DnaK and GroEL2, in M. tuberculosis H37Rv-in-
fected, BCG-vaccinated, and naïve guinea pigs. None of the
naïve guinea pigs developed a reaction to any of the formula-
tions (data not shown). The results showed that there was no
DTH response when the six proteins were injected individually
(data not shown); however, distinct DTH responses could be
measured when proteins were combined with DnaK and
GroEL2 (Fig. 1). As expected, PPD did not discriminate be-
tween the infected and BCG-vaccinated guinea pigs, with DTH
responses of 8.68 ? 0.34 mm (mean responses of erythema ?
SEM) and 8.94 ? 0.4 mm, respectively. Like PPD, all of the
other formulations tested elicited almost identical DTH reac-
tions in M. tuberculosis-infected and BCG-vaccinated guinea
pigs. However, both DnaK/GroEL2/Rv0009 and DnaK/GroEL2/
Rv0685 induced stronger DTH responses than did the DnaK/
GroEL2 formulation alone (P ? 0.05); the DTH response to
DnaK/GroEL2/Rv0569 was equivalent to that to DnaK/GroEL2
alone. Interestingly, the DnaK/GroEL2/Rv0475, DnaK/GroEL2/
Rv2626c, and DnaK/GroEL2/Rv2632c formulations weakened
the DTH responses in both groups (P ? 0.05) compared to that
elicited by DnaK/GroEL2 (data not shown for DnaK/GroEL2/
Histological examination of DTH sites from infected guinea
pigs. To assess if the same type of cellular and inflammatory
response was induced by using different formulations and PPD,
microscopic examination of H&E-stained sections was per-
formed on skin test sites from infected guinea pigs (Fig. 2).
This revealed that the different protein formulations induced
pathological changes similar to that of PPD. In all cases, the
inflammation in the dermis was characterized by vasodilation,
edema, degranulation of mast cells, and infiltration of granu-
locytes, mainly neutrophils, as well as lymphocytes and mono-
cytes, at the site of antigen injection. These results suggest that
the different formulations induced a DTH response that
closely mimics that of PPD. No significant response was ob-
served at the site of the saline injection in infected animals.
Flow cytometry. To determine the phenotype of the T cells
recruited to the DTH reaction site, flow cytometry was per-
formed on skin samples from infected guinea pigs stimulated
with four different protein cocktails and PPD. The PBS injec-
tion site was used to measure nonspecific T-cell accumulation.
The absolute number and percentage of T subgroup cells pos-
itive for each marker studied was determined by comparison to
a sample stained with an isotype control antibody. We ob-
served a 3- to 4-fold increase in both the absolute number and
the percentage of CD3?CD4?CD62Llowand CD3?CD8?
CD62LlowT cells in animals treated with the different formu-
lations compared to those of animals treated with the PBS
control (P ? 0.05 or P ? 0.001, Fig. 3A and B). An increase
was also observed in terms of the percentage of CD3?CD4?
CD62LlowT cells in the DnaK/GroEL2 and DnaK/GroEL2/
Rv0685 groups compared to that in the PPD group (P ? 0.05,
Fig. 3B); however, there were no differences between different
formulations when the data on the absolute numbers of CD3?
CD4?CD62Llowand CD3?CD8?CD62LlowT cells in the
samples were analyzed (Fig. 3A).
Cytokine expression. To investigate further the potential
differences in the cytokine secretion pattern of skin cells at the
DTH sites, we analyzed cytokine expression by real-time RT-
PCR in infected animals using four protein cocktails and PPD.
The results demonstrate that IFN-? and TNF-? were signifi-
cantly upregulated in response to all of the formulations over
PBS (Fig. 4). While there were no significant differences in
cytokine induction between different formulations, there was a
trend of higher TNF-? expression in three of the protein cock-
tails and PPD versus the cocktail containing DnaK and
GroEL2 only. The results also showed an absence of detect-
able IL-10 mRNA in any of the samples.
In this study, we tested four dominant proteins found in PPD
for their potential to induce a DTH response in M. tuberculo-
sis-infected or BCG-vaccinated guinea pigs. A baseline DTH
response was established for two of these proteins, DnaK and
GroEL2. Products in PPD that interacted with these two chap-
erones were identified by pull-down assays and mass spectrom-
etry, produced as recombinant antigens, and tested alone or in
combination with DnaK and GroEL2 for the capacity to in-
duce a DTH response. Three cocktails, each containing three
proteins, were as potent as PPD in the guinea pig model of
tuberculosis. The primary molecular mechanisms responsible
for inducing a DTH response were indistinguishable when
differences between these cocktails and PPD were evaluated.
Thus, the biological potency of PPD can be reproduced using
a simplified, three-protein cocktail mixture of DnaK/GroEL2/
Rv0009, DnaK/GroEL2/Rv0569, or DnaK/GroEL2/Rv0685.
These well-defined cocktails can be used to further define the
FIG. 1. Skin test responses in M. tuberculosis H37Rv-infected or
BCG-vaccinated guinea pigs following intradermal injection with PPD
(0.4 ?g/site), DnaK/GroEL2 (0.13 ?g of each protein/site), DnaK/
GroEL2/Rv0009 (0.13 ?g of each protein/site), DnaK/GroEL2/Rv0569
(0.13 ?g of each protein/site), DnaK/GroEL2/Rv0685 (0.13 ?g of each
protein/site), DnaK/GroEL2/Rv0475 (0.13 ?g of each protein/site),
DnaK/GroEL2/Rv2632c (0.13 ?g of each protein/site), and PBS (data
not shown). The 48-h DTH (skin erythema measurement) response is
shown. No DTH reactions were found in naïve guinea pigs stimulated
with any of the antigens or in any animals stimulated with PBS alone.
The results shown are from one representative experiment (eight an-
imals per group; all experiments were duplicated) and are expressed as
the mean and SEM.*, P ? 0.05 (compared to the DnaK/GroEL2
VOL. 79, 2011NOVEL SECOND-GENERATION TUBERCULOSIS SKIN TEST REAGENT 719
molecular mechanisms of the DTH response in tuberculosis,
including the kinetics of the response and the role of sensiti-
zation. PPD contains hundreds of antigens that cumulatively
induce the DTH response (3; Cho et al., unpublished). Prior to
this study, it was unknown which, or which combination, of
these products, were responsible for the specificity or sensitiv-
ity of the DTH response. This is further complicated by the
abundance of denatured, aggregated, and lysed proteinaceous
products in PPD, complicating analysis of the PPD proteome,
such that only two publications describe the PPD proteome in
detail (3; Cho et al., unpublished). Both the work by Mc-
Fadden and our own identified the chaperones DnaK,
GroEL2, and GroES as dominant products in PPD. In this
study, these three chaperones and an additional chaperone,
HspX, were tested using the guinea pig model of M. tubercu-
losis infection. DTH responses similar to that induced by PPD
were observed when DnaK and GroEL2 were used separately or
in combination in M. tuberculosis-infected or BCG-vaccinated
guinea pigs. HspX and GroES, on the other hand, were negative
or gave a poor response, whether injected separately or as a
cocktail, in the guinea pig model of tuberculosis. Based on the
that these two proteins significantly contribute to both the po-
tency and lack of specificity of PPD in the DTH response.
Because both DnaK and GroEL2 are chaperone proteins, it
was proposed that other proteins in PPD may be interacting
with DnaK and GroEL2, and addition of these may enhance
and engender a more specific DTH response than the use of
these two proteins alone. To test this hypothesis, a pull-down
experiment was performed to identify these proteins in PPD.
Six novel proteins in PPD were identified by this method
(Rv0009, Rv0475, Rv0569, Rv0685, Rv2626c, Rv2632c). None
of these proteins induced a DTH response when tested indi-
vidually. However, the addition of some of these proteins to
DnaK and GroEL2 elicited a potent DTH response. Specifi-
cally, the protein cocktails DnaK/GroEL2/Rv0009 and DnaK/
GroEL2/Rv0685 induced stronger DTH responses in H37Rv-
infected and BCG-vaccinated guinea pigs than did DnaK/
GroEL2 only, while the addition of other proteins to the
DnaK/ GroEL2 mixture resulted in either an equally active or
a weakened DTH response. This illustrates that protein-pro-
tein interactions either enhance or abrogate the DTH re-
sponse. In addition, it is possible that the proteins contributing
to a weakened DTH response may be inducing an alternative
type of immune response. An anti-inflammatory or regulatory
T-cell-type response may be occurring, leading to an overall
weakening of the DTH-mediated immune response. Com-
bined, these factors explain, in part, the variability seen in
DTH responses, further demonstrating the challenges faced
when defining the DTH response in M. tuberculosis infection or
FIG. 2. Microphotographs of inflammatory changes at injection sites in skin from guinea pigs infected with M. tuberculosis H37Rv. Skin biopsy
specimens were taken 48 h following intradermal injection of PPD, protein cocktails, and PBS as a negative control. (A) No changes were apparent
at the PBS injection site. (B to D) Following injection of PPD (0.4 ?g protein/site) (B), DnaK/GroEL2 (0.13 ?g each protein/site) (C), or
DnaK/GroEL2/Rv0685 (0.13 ?g each protein/site) (D), there was vasodilation, mild edema, and dermal infiltration of neutrophils, lymphocytes,
and monocyte/macrophages. Final magnification, ?200.
720YANG ET AL.INFECT. IMMUN.
after BCG vaccination. The specific cellular attributes respon-
sible for the DTH response in this model of tuberculosis are
currently being studied using these defined protein cocktails.
Discovery of these essential components during infection with
M. tuberculosis or protection by BCG vaccination may lead to
the development of rapid, inexpensive second-generation skin
test antigens for tuberculosis.
Many protein candidates have been described as tuberculo-
sis-specific skin test antigens over the past 2 decades (5, 6, 11,
24); however, none have reached an optimal diagnostic sensi-
tivity. In this study, the protein mixtures described elicited a
response that was equal to or greater than the response to PPD
and was indistinguishable from PPD at the molecular level.
While these cocktails did not demonstrate significantly differ-
ent responses between infected and vaccinated animals, they
do provide a tool to define the DTH response during the
course of infection. Once defined, additional protein formula-
tions that induce a strong and tuberculosis-specific DTH re-
sponse can be rapidly tested.
To elucidate the mechanism of the DTH response in M.
tuberculosis-infected guinea pigs, histological examination and
cellular analyses were performed at PPD or protein formula-
tion injection sites on the skin of infected animals. The analysis
FIG. 3. Flow cytometry of the DTH response in M. tuberculosis-
infected guinea pigs. Skin samples were obtained from M. tuberculosis
H37Rv-infected guinea pigs stimulated with PPD (0.4 ?g protein/site),
DnaK/GroEL2 (0.13 ?g each protein/site), DnaK/GroEL2/Rv0009
(0.13 ?g each protein/site), DnaK/GroEL2/Rv0569 (0.13 ?g each pro-
tein/site), DnaK/GroEL2/Rv0685 (0.13 ?g each protein/site), and PBS.
Proportions of T cells and their subsets were determined by flow
cytometry analysis after staining the cells with monoclonal antibodies
directed against the surface T cells (MCA751APC), CD4?T cells
(MCA749F and MCA749PE), CD8?T cells (MCA752F), and CD62L
cells (SC-13505PE). Results are expressed as the mean number of cells
of each sample ? the SEM (A) and the mean percentage of cells ? the
SEM (B). Eight animals were used for each assay.*, P ? 0.05;**, P ?
0.001 (compared to the PBS group).
FIG. 4. Cytokine mRNA expression at the site of the DTH re-
sponse in M. tuberculosis-infected guinea pigs. Skin cells were obtained
from M. tuberculosis H37Rv-infected guinea pigs stimulated with PPD
(0.4 ?g protein/site), DnaK/GroEL2 (0.13 ?g each protein/site),
DnaK/GroEL2/Rv0009 (0.13 ?g each protein/site), DnaK/GroEL2/
Rv0569 (0.13 ?g each protein/site), DnaK/GroEL2/Rv0685 (0.13 ?g
each protein/site), or PBS. IFN-? (A), TNF-? (B), and IL-10 (data not
shown) mRNA expression was quantified by real-time RT-PCR. The
n-fold induction of mRNA was calculated from the CTvalues normal-
ized to ?-actin CTvalues and then to PBS-stimulated skin samples. The
results are expressed as the mean and SEM. Five animals were used for
VOL. 79, 2011 NOVEL SECOND-GENERATION TUBERCULOSIS SKIN TEST REAGENT 721
of the flow cytometry data demonstrated that almost the same
number of activated CD4?and CD8?T cells are involved in
the DTH responses. Previous studies identified more CD4?T
cells than CD8?T cells in DTH sites (14, 29, 38). There are
two probable reasons for the discrepancy. One is the differ-
ences between the human experience and the guinea pig model
of tuberculosis. The other is that, in this study, activated CD4?
and CD8?T-cell populations were measured directly at the
site of the DTH response. To the best of our knowledge, this
study is the first in which flow cytometry was used to look at
DTH responses in the guinea pig model and may more accu-
rately reflect the essential mechanisms of a productive DTH
DTH is classically defined as the recruitment of T cells to be
activated by antigen-presenting cells, mainly Langerhans cells
in skin, to produce cytokines that mediate local inflammation.
Although the role of CD8?T cells was previously poorly rec-
ognized, there is now considerable evidence that during the
process of DTH, the antigen is processed and presented to
both CD4?and CD8?T cells (2). It is well known that phago-
cytosed antigens enter the exogenous pathway and are pro-
cessed for presentation on major histocompatibility complex
(MHC) class II molecules to CD4?T cells, while cytoplasmic
antigens are processed by the endogenous pathway for presen-
tation on MHC class I molecules to CD8?T cells. However,
exogenous antigen can also enter the endogenous pathway to
be presented to CD8?T cells (18). Our data demonstrate the
activation of CD8?T cells in DTH responses using PPD and
different formulations, indicating that after proteins are di-
gested in vivo, the specific peptides are processed and pre-
sented on both MHC class I and II molecules. The expression
of cytokines during the DTH response was determined by
measuring specific mRNA expression levels at the sites of the
DTH response in infected guinea pigs. Our data demonstrated
that IFN-? and TNF-? are actively expressed at the DTH site,
and the same immune response pattern is induced by defined
protein cocktails and PPD. Further, the T-cell response ap-
pears to be biased toward a Th1 T-cell response, based on the
increases in the Th1 cytokines IFN-? and TNF-? and the
absence of IL-10 at the DTH sites. Injection of PPD into
the footpads of H37Rv-infected mice is also characterized by
high expression of Th1 cytokines like IL-2 and IFN-? and the
absence of Th2 cytokines such as IL-4 detected by RT-PCR
(26). Similarly, the expression of IFN-?, the major cytokine
produced by Th1 cells at sites of DTH responses, has been
confirmed by others (4, 12) and is consistent with our results.
Real-time RT-PCR can thus be used for validation of candi-
date PPD alternatives.
In summary, in our study, DnaK/GroEL2/Rv0009 and
DnaK/GroEL2/Rv0685 were found to induce stronger re-
sponses than single protein or DnaK/GroEL2 protein cocktail
injections and DTH responses indistinguishable from that elic-
ited by PPD in the guinea pig model of tuberculosis. This is the
first time that mimicking the PPD response has been shown at
the molecular level with defined protein cocktails. The use of
defined formulations will not only help provide a more thor-
ough understanding of the DTH response but also provide a
platform for standardized, defined second-generation skin test
reagents for the diagnosis of M. tuberculosis infection and pos-
sibly novel skin test reagents for the evaluation of vaccine
This work was supported by the Tuberculosis Vaccine Testing and
Research Materials contract (HHSN266200400091C) from the NIH.
We thank Linda Izzo for technical assistance with real-time PCR
1. Affronti, L. F. 1959. Purified protein derivatives (PPD) and antigens pre-
pared from atypical acid-fast bacilli and Nocardia asteroides. Am. Rev.
2. Black, C. A. 1999. Delayed type hypersensitivity: current theories with an
historic perspective. Dermatol. Online J. 5:7.
3. Borsuk, S., J. Newcombe, T. A. Mendum, O. A. Dellagostin, and J. Mc-
Fadden. 2009. Identification of proteins from tuberculin purified protein
derivative (PPD) by LC-MS/MS. Tuberculosis (Edinb.) 89:423–430.
4. Cher, D. J., and T. R. Mosmann. 1987. Two types of murine helper T cell
clone. II. Delayed-type hypersensitivity is mediated by TH1 clones. J. Im-
5. Colangeli, R., et al. 2000. MTSA-10, the product of the Rv3874 gene of
Mycobacterium tuberculosis, elicits tuberculosis-specific, delayed-type hyper-
sensitivity in guinea pigs. Infect. Immun. 68:990–993.
6. Coler, R. N., et al. 2000. Cloning of a Mycobacterium tuberculosis gene
encoding a purified protein derivative protein that elicits strong tuberculosis-
specific delayed-type hypersensitivity. J. Infect. Dis. 182:224–233.
7. Comstock, G. W., L. B. Edwards, R. N. Philip, and W. A. Winn. 1964. A
comparison in the United States of America of two tuberculins, Ppd-S and
Rt 23. Bull. World Health Organ. 31:161–170.
8. Donald, P. R., and P. D. van Helden. 2009. The global burden of tubercu-
losis—combating drug resistance in difficult times. N. Engl. J. Med. 360:
9. Dye, C., S. Scheele, P. Dolin, V. Pathania, and M. C. Raviglione. 1999.
Consensus statement. Global burden of tuberculosis: estimated incidence,
prevalence, and mortality by country. WHO Global Surveillance and Mon-
itoring Project. JAMA 282:677–686.
10. Dye, C., B. G. Williams, M. A. Espinal, and M. C. Raviglione. 2002. Erasing
the world’s slow stain: strategies to beat multidrug-resistant tuberculosis.
11. Elhay, M. J., T. Oettinger, and P. Andersen. 1998. Delayed-type hypersen-
sitivity responses to ESAT-6 and MPT64 from Mycobacterium tuberculosis in
the guinea pig. Infect. Immun. 66:3454–3456.
12. Fong, T. A., and T. R. Mosmann. 1989. The role of IFN-gamma in delayed-
type hypersensitivity mediated by Th1 clones. J. Immunol. 143:2887–2893.
13. Fryer, A. D., et al. 1997. Antibody to VLA-4, but not to L-selectin, protects
neuronal M2 muscarinic receptors in antigen-challenged guinea pig airways.
J. Clin. Invest. 99:2036–2044.
14. Gibbs, J. H., et al. 1984. Histometric study of the localisation of lymphocyte
subsets and accessory cells in human Mantoux reactions. J. Clin. Pathol.
15. Grover, A., et al. 2009. Kinetics of the immune response profile in guinea pigs
after vaccination with Mycobacterium bovis BCG and infection with Myco-
bacterium tuberculosis. Infect. Immun. 77:4837–4846.
16. Guld, J., et al. 1958. Standardization of a new batch of purified tuberculin
(PPD) intended for international use. Bull. World Health Organ. 19:845–
17. He, X. Y., et al. 2008. Clinical evaluation of the recombinant 38 kDa protein
of Mycobacterium tuberculosis. Scand. J. Infect. Dis. 40:121–126.
18. Kalish, R. S., and P. W. Askenase. 1999. Molecular mechanisms of CD8? T
cell-mediated delayed hypersensitivity: implications for allergies, asthma,
and autoimmunity. J. Allergy Clin. Immunol. 103:192–199.
19. Klausen, J., M. Magnusson, A. B. Andersen, and C. Koch. 1994. Character-
ization of purified protein derivative of tuberculin by use of monoclonal
antibodies: isolation of a delayed-type hypersensitivity reactive component
from M. tuberculosis culture filtrate. Scand. J. Immunol. 40:345–349.
20. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression
data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method.
21. Lyashchenko, K., et al. 1998. Use of Mycobacterium tuberculosis complex-
specific antigen cocktails for a skin test specific for tuberculosis. Infect.
22. Mehaffy, C., et al. 2010. Descriptive proteomic analysis shows protein vari-
ability between closely related clinical isolates of Mycobacterium tuberculo-
sis. Proteomics 10:1966–1984.
23. Mustafa, A. S. 2001. Biotechnology in the development of new vaccines and
diagnostic reagents against tuberculosis. Curr. Pharm. Biotechnol. 2:157–
24. Oettinger, T., A. Holm, I. M. Mtoni, A. B. Andersen, and K. Hasloov. 1995.
722 YANG ET AL.INFECT. IMMUN.
Mapping of the delayed-type hypersensitivity-inducing epitope of secreted
protein MPT64 from Mycobacterium tuberculosis. Infect. Immun. 63:4613–
25. Pai, M., K. Dheda, J. Cunningham, F. Scano, and R. O’Brien. 2007. T-cell
assays for the diagnosis of latent tuberculosis infection: moving the research
agenda forward. Lancet Infect. Dis. 7:428–438.
26. Pais, T. F., R. A. Silva, B. Smedegaard, R. Appelberg, and P. Andersen. 1998.
Analysis of T cells recruited during delayed-type hypersensitivity to purified
protein derivative (PPD) versus challenge with tuberculosis infection. Im-
27. Pillay, T., M. Khan, J. Moodley, M. Adhikari, and H. Coovadia. 2004.
Perinatal tuberculosis and HIV-1: considerations for resource-limited set-
tings. Lancet Infect. Dis. 4:155–165.
28. Pollock, J. M., et al. 2003. Specific delayed-type hypersensitivity responses to
ESAT-6 identify tuberculosis-infected cattle. J. Clin. Microbiol. 41:1856–
29. Poulter, L. W., G. J. Seymour, O. Duke, G. Janossy, and G. Panayi. 1982.
Immunohistological analysis of delayed-type hypersensitivity in man. Cell.
30. Rangel-Frausto, M. S., S. Ponce-De-Leon-Rosales, C. Martinez-Abaroa, and
K. Haslov. 2001. Tuberculosis and tuberculin quality: best intentions, mis-
leading results. Infect. Control Hosp. Epidemiol. 22:481–484.
31. Reed, S. G., and A. Campos-Neto. 2003. Vaccines for parasitic and bacterial
diseases. Curr. Opin. Immunol. 15:456–460.
32. Rhodes, S. G., et al. 2000. Antigen specificity in experimental bovine tuber-
culosis. Infect. Immun. 68:2573–2578.
33. Seibert, F. B. 1934. The isolation and properties of the purified protein
derivative of tuberculin. Am. Rev. Tuberc. 30:713–720.
34. Seibert, F. B., and J. T. Glen. 1941. Tuberculin purified protein derivative:
preparation and analyses of a large quantity of standard. Am. Rev. Tuberc.
35. Shingadia, D., and V. Novelli. 2008. The tuberculin skin test: a hundred, not
out? Arch. Dis. Child. 93:189–190.
36. Tan, B. T., et al. 1985. Production of monoclonal antibodies defining guinea
pig T-cell surface markers and a strain 13 Ia-like antigen: the value of
immunohistological screening. Hybridoma 4:115–124.
37. Villarino, M. E., et al. 2000. Comparison testing of current (PPD-S1) and
proposed (PPD-S2) reference tuberculin standards. Am. J. Respir. Crit. Care
38. Vukmanovic-Stejic, M., J. R. Reed, K. E. Lacy, M. H. Rustin, and A. N.
Akbar. 2006. Mantoux test as a model for a secondary immune response in
humans. Immunol. Lett. 107:93–101.
39. Weldingh, K., and P. Andersen. 2008. ESAT-6/CFP10 skin test predicts
disease in M. tuberculosis-infected guinea pigs. PLoS One 3:e1978.
Editor: J. L. Flynn
VOL. 79, 2011 NOVEL SECOND-GENERATION TUBERCULOSIS SKIN TEST REAGENT723