Cell Host & Microbe

Review

Tuberculosis in Africa: Learning from

Pathogenesis for Biomarker Identification
Stefan H.E. Kaufmann1,* and Shreemanta K. Parida1
1Max Planck Institute for Infection Biology, Department of Immunology, Charitéplatz 1, D-10117 Berlin, Germany

*Correspondence: kaufmann@mpiib-berlin.mpg.de
DOI 10.1016/j.chom.2008.08.002

In Africa, more than 4 million people suffer from active tuberculosis (TB) resulting in an estimated 650,000
deaths every year. The etiologic agent of TB, Mycobacterium tuberculosis, survives in resting macrophages,
which control the pathogen after activation by specific T lymphocytes. Here, we describe the basic mecha-
nisms underlying the host response to TB with an emphasis on immunity and discuss diagnostics, drugs, and
vaccines for TB. Moreover, we outline our attempts to develop biomarkers, which could help the monitoring
of TB clinical trials, provide the basis for new diagnostics, and allow prognosis of outcome of infection and of
drug treatment.

Introduction research leading to the discovery of Mtb as the etiologic agent

Although tuberculosis (TB) is a health threat of global dimension,
it is most on the rampage in Africa with a focus in the Sub-
Saharan part of the continent (WHO, 2008). Between 1990 and
2006, TB prevalence in Africa rose from 1.7 million to 4.2 million
active cases with no signs of relief, as dramatically illustrated by
the more than three-fold increase in new cases from 800,000 in
1990 to 2.8 million in 2006. Much of this problem is due to the
human immunodeficiency virus/acquired immune deficiency
syndrome (HIV/AIDS) plague as illustrated by the enormous
HIV prevalence among TB patients of 22% (Table 1). In today’s
Africa, Mycobacterium tuberculosis (Mtb) is the number one killer
of HIV-infected individuals with one third of the 640,000 deaths
due to TB occurring in HIV-coinfected sufferers. Better health
care ranging from prevention to therapy for TB patients is ur-
gently required. Because of the high coincidence of TB and
AIDS, Mtb and HIV can no longer be viewed exclusively by them-
selves but must also be seen as two diseases in a single patient
(Kaufmann and McMichael, 2005). The sufferers of HIV–Mtb
coinfection need particular care since, historically, they have
often been treated at different health facilities by different health
care providers. To make the situation worse, we are faced with
an increasing occurrence of immune reconstitution inflammatory
syndrome (IRIS) in HIV–Mtb-coinfected individuals under antire-
troviral therapy (ART), which exacerbates TB reactivation.
It has occasionally been proposed that efforts should be con-
centrated on prevention and therapy of HIV/AIDS and that TB
would disappear once HIV control has been successfully imple-
mented. Unfortunately, the different modes of transmission of
HIV and Mtb make the success of such a proposal unlikely.
HIV is primarily transmitted by sexual intercourse and thus can
be prevented to a large degree by changes in behavior such as
abstinence or use of condoms accompanied by appropriate
education. In contrast, behavioral changes are largely inefficient
for TB because the pathogen is transmitted by aerosols, and it is
simply impossible to stop breathing.
In some regions of Sub-Saharan Africa, TB has reached a level
reminiscent of that of European cities at the turn of the last cen-
tury. The European TB situation improved not only due to better
living conditions. These were also the times of highly active TB
more than 125 years ago, the development of the tuberculin
skin test almost 120 years ago, both by Robert Koch, and the de-
velopment of a vaccine against TB named after its discoverers,
bacille Calmette-Guérin (BCG) ca. 80 years ago (Kaufmann
and Winau, 2005; Kaufmann and Parida, 2007). All these mea-
sures are still in use today. It is no wonder that they cannot satisfy
our state-of-the-art demands. The acid-fast staining of Mtb is the
basis of sputum diagnosis, which in Africa detects less than half
of all patients with active TB (Box S1 available online). The tuber-
culin skin test forms the basis for population-wide assessment of
infection with Mtb but cannot distinguish between latent infec-
tion and active disease (Lalvani, 2007) (Box S1). The vaccine
BCG protects newborns and young children from developing
severe disseminated TB (miliary TB) but cannot protect against
adult pulmonary TB, which today is the most prevalent form of
the disease (Kaufmann, 2006; Fine, 1995). Accordingly, this vac-
cine has virtually no impact on transmission of TB (Corbett et al.,
2003). Most drugs used to treat TB were developed between
1950 and 1970 (Box S2). The need for 3 to 4 drugs over a period
of 6 to 9 months—combined with poor compliance—has led to
the emergence of single drug-resistant, multidrug-resistant
(MDR), and extensively drug-resistant (XDR) strains of Mtb. An
effective response to this increasing threat requires new drugs
targeting novel structures and pathways in Mtb that are not con-
quered by current drugs. Without doubt, better housing condi-
tions for those living in townships and densely crowded parts
of cities in Sub-Saharan Africa and improved food and nutrient
supply are needed to raise socioeconomic standards (Schaible
and Kaufmann, 2007). Yet, in a continent where a large propor-
tion of the inhabitants live on less than $2 US a day, it is unlikely
that these goals will be achieved in the near future. They need to
be complemented by better medical intervention measures.
In the short-term, broad coverage of BCG vaccination of
newborns and complete coverage of TB drug supply, along
with better diagnostic tools in the field, are needed to face the
threat. In the long-term, satisfactory control will depend on an
efficacious vaccine to prevent pulmonary TB in adults and trans-
mission of Mtb (Kaufmann, 2006; Kaufmann and McMichael,
2005). A similar long-term strategy is hoped for HIV, namely

Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc. 219

Table 1. TB and HIV/AIDS-TB in Africa and in African States, Where GC6-74 Studies Are Being Performed
Incidence of All Incidence of All
Forms of TB 2006 Forms of TB 2006
Nation (Numbers) (Rate per 100,000)
Ethiopia 306,330 379
Gambia 4,278 257
Malawi 51,172 377
South Africa 453,929 940
Uganda 106,037 355
Total Africa 2,807,688(921,746)a 363
Total Global 9,157,021 139
Source: (WHO, 2008).
a
Total of five countries depicted in Table 1.
development of a vaccine that provides protection (Kaufmann
and McMichael, 2005). However, despite major efforts to de-
velop an efficacious HIV vaccine, this goal seems to be more dis-
tant than ever (Lagakos and Gable, 2008; Walker and Burton,
2008; Boasso et al., 2008). Although the situation for developing
a successful TB vaccine is improving, much still needs to be
done (Kaufmann, 2006).
Why Biomarkers for TB?
We have argued that the vast majority (approx. 80% in Sub-
Saharan Africa) of all who become infected with Mtb will control
the pathogen lifelong without developing clinical signs of dis-
ease. Thus, the host immune system has mechanisms at hand
to control the pathogen effectively although it only rarely seems
to achieve sterile eradication of Mtb (McKinney et al., 1998). Will
it be possible to learn from this naturally induced host response
and better understand those mechanisms that control infection
and exploit their value as potential indicators of protection? If
so, then a biomarker of protection against TB could be defined
and serve as a guideline for monitoring vaccine-induced immu-
nity. Such a biomarker of protection against disease could
provide valuable information about vaccine efficacy prior to the
clinical end-point of TB disease outbreak. In the area of clinical
trials, a biomarker is defined as a characteristic feature that is
objectively measured and evaluated as an indicator of a normal
or a pathological process or of the response to an intervention
(Biomarkers Definitions Working Group, 2001). Thus, our goals,
to identify biomarkers of protection and susceptibility to natural
infection with Mtb leading to the adoption of a marker of
vaccine-induced immunity against TB, fall into this definition
(Table 2).
With support from the Bill & Melinda Gates Foundation under
the Grand Challenges in Global Health program (http://www.
gcgh.org/NewVaccines/Challenges/LearnaboutImmunological
Responses/Pages/default.aspx), we attempt to define bio-
markers for TB in the context of HIV/AIDS in five African nations,
namely, Ethiopia, The Gambia, Malawi, South Africa, and
Uganda (http://www.biomarkers-for-tb.net/). This collaborative
project falls under the broader thematic goal ‘‘To Create New
Vaccines’’ among seven major goal-oriented themes and specif-
ically under Grand Challenge 6 (GC6) out of the 14 defined Grand
Challenges topics: ‘‘Learn which immunological responses pro-
vide protective immunity’’; referred as GC6-74 hereafter (Box
220 Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc.
Cell Host & Microbe
Review
Prevalence of All HIV Prevalence
Forms of TB in Incident TB Population
(per 100,000) Cases (%) (Millions)
643 6.3 81
423 7.5 2
322 70 14
998 44 50
561 16 30
547 22 775
219 7.7 6,590
S3). The countries involved are all severely affected by Mtb/TB
and HIV/AIDS. TB incidences range from 940 per 100,000 in
South Africa as the highest to 257 per 100,000 in The Gambia
as the lowest, and from 70% HIV in TB patients as the highest
in Malawi to 6.3% in Ethiopia (See Table 1). Newly diagnosed
acute pulmonary TB patients are the entry point of our studies,
and their healthy household contacts infected with Mtb as evi-
denced by positive tuberculin skin testing are the focus of our
studies. Several cohorts have been included in longitudinal pro-
spective case control studies spanning a follow-up period of
2 years from the time of recruitment. An integrated interdisciplin-
ary immunologic and multiomics approach has been adopted
for blood leukocytes from both HIV ve and HIV+ve subjects.
HIV+ve cohorts (with and without active TB disease) are con-
founded by several factors such as the use of chemoprophylaxis
and ART (particularly in subjects without active TB) that would in-
fluence the outcome of TB (re)activation. Nevertheless, they pro-
vide a trend in comparison to the HIV ve cohorts in the same
population.
A major goal of our strategy is the definition of a surrogate
end-point for a phase II/III clinical trial with novel TB vaccines
and perhaps also with new drugs against TB. Several TB vac-
cines are due to enter, or have already completed, clinical phase
I trials and a few candidates have reached phase II trials (Box
S4). Some are expected to enter phase III clinical trials in the
near future. The greatest obstacle will thus come at the end of
the research-and-development pipeline with a phase III clinical
trial aimed at definitively determining protective vaccine efficacy
in adults. Currently, the success of a vaccine in a phase III clinical
trial is measured by the prevention of active TB outbreak in study
group participants. In other words, clinical diagnosis of TB dis-
ease serves as the clinical end-point of such a trial. Even with
a TB incidence rate of 1 in 100 as seen in the worst-affected parts
of South Africa and Swaziland, large groups of individuals would
need to be enrolled to reach adequate power in the study to ob-
tain conclusive evidence. Added to this is the long incubation
time of latent Mtb infection ranging from months to lifelong.
About 10%–20% of those who will become infected with Mtb
during a phase III trial in Africa will develop active TB at a later
time point with 5%–10% probably occurring within the first
2 years. Thus, it has been estimated that clinical trials for TB vac-
cines may last several years (possibly a decade if not longer)
comprising thousands of study subjects. In view of this grave
Cell Host & Microbe
Review
Table 2. Biomarker Definitions
Term Definition
Biomarker Characteristic that is objectively measured
and evaluated as an indicator of normal
biological processes, pathological processes,
or physiological/pharmacological responses
to an intervention.
Correlate of Measurable sign in a host in response to
protection an infectious agent indicating whether the
individual is being protected against
becoming infected and/or developing
disease.
Surrogate of Validated marker of correlate of protection.
protection
Clinical endpoint Characteristic or variable that reflects the
final outcome of disease in terms of function,
effect, progress, recovery, survival, or death.
Surrogate endpoint Biomarker that is intended to substitute for a
clinical endpoint, predicting clinical outcome
in terms of benefit, or harm, or lack of benefit
or harm.
Source: (Biomarkers Definitions Working Group, 2001; Colburn, 2000).
situation, biomarkers could help to accelerate the development
process at different stages of clinical trials.
Even before a phase III clinical trial, biomarkers that provide
robust information for educated decisions about entry into, or
termination of, phase II clinical vaccine trials as well as informa-
tion for decisions on combinations of different vaccine candi-
dates in heterologous prime-boost schemes would be extraordi-
narily helpful. As an added value, information on biomarkers for
TB can provide the basis for (1) novel diagnostic tools to more
reliably diagnose active TB (Box S1), (2) prognostic markers of
infection outcome, i.e., TB disease reactivation or latent infec-
tion, and (3) biomarkers that can predict drug treatment out-
come, i.e., treatment success, drug failure, or relapse (Figure 1).
From the Site of Infection and Disease
to the Peripheral Blood
In the following we will describe our rationale for gathering infor-
mation about the host response in naturally infected individuals
for definition of biomarkers in TB.
In Africa, almost 60% of all TB cases are sputum smear neg-
ative (WHO, 2008). This not only represents a severe impediment
to active case finding and hence for efficacious drug treatment,
but it also suggests that there is a high proportion of extrapulmo-
nary TB (WHO, 2008). Extrapulmonary TB can progress to a gen-
eralized form such as miliary TB in AIDS patients or be confined
to distinct organs with the most prevalent and deadly form being
meningeal TB (Maher, 2008). Yet, in the majority of cases, includ-
ing all sputum smear-positive and a significant proportion of
smear-negative cases, Mtb resides in the lung, which therefore
serves as the origin of transmission via aerosols. Enclosed in
minute droplets, bacteria are expelled by coughing of an active
pulmonary TB patient and in this way enter lung alveoli of other
individuals. There, alveolar macrophages, as well as interstitial
dendritic cells (DCs), engulf Mtb (Figure 1). Some bacteria are
transported to draining lymph nodes while others are trapped
Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc. 221
in the lung parenchyma. Monocytes and granulocytes are at-
tracted to the site of bacterial deposition in the lung parenchyma
where accumulating cells form loosely packed lesions. In the
draining lymph nodes, T lymphocytes with specificity for myco-
bacterial antigens are stimulated. Antigen-specific T cells, by
way of their recirculation and tissue surveillance, recognize
Mtb-infected macrophages and DCs in lung lesions and induce
formation of productive granulomas where Mtb is contained
and thus prevented from causing active disease (Ulrichs and
Kaufmann, 2006; Russell, 2007). Different combinations of che-
mokines produced by cells in the inflamed and infected area
attract different leukocyte populations to the site of action (Charo
and Ransohoff, 2006; Li and Flavell, 2008; Wolf et al., 2008). Re-
ciprocally, cells that respond to the chemokine gradient need to
express the homologous set of chemokine receptors (Charo and
Ransohoff, 2006).
T Cell Subsets and Their Cytokines
Activated T lymphocytes comprise CD4 T cells and CD8 T cells
as well as unconventional T cells, such as the gd T lymphocytes
and the CD1-restricted T lymphocytes (Flynn and Chan, 2001).
The CD4 T cells that are restricted by gene products of the major
histocompatibility complex (MHC) class II represent major
players in protection against TB (Figure 2). Of greatest relevance
are the T helper (Th) cells that produce interferon gamma (IFN-g),
which have been termed Th1 cells. IFN-g activates macro-
phages (Nathan et al., 1983). As a consequence, the resting
macrophage, which primarily serves as a fertile habitat for Mtb,
is transformed into an activated macrophage that can control
replication of Mtb. Still, even activated macrophages fail to erad-
icate their predators (Vandal et al., 2008). Thus, the macrophage
serves as both host and effector cell. Recent evidence suggests
that interleukin (IL) 17-producing Th cells (so-called Th17 cells)
participate in protection against TB, although their role is less
clear (Khader et al., 2007). Probably they provide a first line of
defense against Mtb and participate in the early steps of granu-
loma formation (Ouyang et al., 2008).
The CD8 T cells recognize mycobacterial antigens in the con-
text of MHC class I and contribute to protection first by produc-
ing cytokines, notably IFN-g, and second by direct cytolytic and
mycobacteriolytic mechanisms. Accordingly, they have been
termed cytolytic T lymphocytes (CTLs). Molecules involved in
these CTL mechanisms include granzymes and perforins. The
gd T cells recognize phosphorylated ligands in apparent ab-
sence of restriction elements and the CD1-restricted T cells
are specific for mycobacterial glycolipids, such as lipoarabino-
mannan (LAM) and mycolic acids (Figure 2). The role of these
unconventional T cells in protective immunity is unclear. It is
most likely that they serve as back-up cells producing similar
cytokines as the CD4 Th1 and Th17 cells as well as the
CD8 CTLs.
The Th1 cells produce additional cytokines, notably IL-2,
which promotes stimulation of both Th1 cells and CTLs, lympho-
toxin (LT), tumor necrosis factor alpha (TNF-a), which synergizes
with IFN-g in macrophage activation, and granulocyte-macro-
phage colony-stimulating factor (GM-CSF), which stimulates
granulocyte and macrophage maturation (Figure 2). The Th17
cells also produce IL-6 as proinflammatory mediator as well as
IL-22, which stimulates defensin production in epithelial cells

Figure 1. From the Site of Action in the Lung to Biomarkers in the Peripheral Blood
The figure depicts the major pathologic and protective mechanisms in the lung during latent Mtb infection and active TB disease. In addition, the picture indicates
determination of immunologic, transcriptomic, and metabolomic biomarkers.
(McGeachy and Cua, 2008; Ouyang et al., 2008). Another Th cell
population, the Th2 cells, produce B cell-stimulating factors, no-
tably IL-4 and IL-5, and hence promote antibody production; yet,
the role of antibodies produced by plasma cells as progeny of
activated B cells in protection against TB is minor. CD4 T cells
can also produce cytokines that impair immunity to TB such as
IL-4, produced by Th2 cells as well as IL-10 and tumor growth
factor beta (TGF-b) produced by regulatory T (Treg) cells (Belkaid
and Rouse, 2005). Some cytokines such as IL-21 that are pro-
duced by different CD4 T cell sets express pleiotropic activities
on various T cell populations. Phenotype studies using knockout
mutant mice or blocking agent(s) for cytokines represent a reduc-
tionist approach in a well-defined model system. In contrast,
in vivo activities of multiple intertwined processes, which pro-
ceed concurrently, have to be addressed using a global systems
biology approach. Obviously, characterization of these different
Th cell populations and their cytokine patterns provide promising
candidates for biomarker research. Recent findings suggest that
polyfunctional T cells, which produce multiple cytokines such
as IFN-g, IL-2, TNF-a, LT, and/or GM-CSF, are best suited as
biomarker candidates for protection in TB (Seder et al., 2008;
Mueller et al., 2008). These polyfunctional T cells probably repre-
sent T memory (Tm) cells, which sustain protective immunity
(Sallusto et al., 2004). Treg cells are interesting candidates for
biomarker research as well because their increase may correlate
with disease reactivation as a consequence of dampened
immune responses (Belkaid and Rouse, 2005).
222 Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc.
Cell Host & Microbe
Review
Antigen Processing/Presentation
Mtb resides in the early phagosome and prevents it from further
maturation (Armstrong and Hart, 1971). Thus, mycobacterial an-
tigens can be loaded on MHC-II molecules, which then present
the antigenic peptides to CD4 T cells. The MHC-I molecules
are loaded with antigens in the cytosolic compartment (Figure 2).
Because Mtb persists within the phagosome, efficient loading of
MHC-I molecules remained a mystery until recently. Recent find-
ings indicate egress of Mtb into the cytosol (van der Wel et al.,
2007); however this issue is still controversial. Evidence from
our laboratory suggests that macrophages infected with Mtb
undergo apoptosis. This leads to the formation of vesicles
containing mycobacterial antigens. DCs in the vicinity of such
apoptotic macrophages engulf these vesicles and present the
antigens in the context of MHC-II, MHC-I, and CD1 in a highly
efficacious way (Schaible et al., 2003; Winau et al., 2006). This
process, termed crosspriming, probably plays a major role in
T cell stimulation during Mtb infection.
T Cell Polarization by Cytokines
Polarization into different T cell sets is mostly dictated by distinct
cytokines (Figure 2) (Kaufmann, 2007; Keir et al., 2008). Thus,
TGF-b, together with IL-6, promotes development of Th17 cells
in mice while in humans IL-21 replaces IL-6 (Yang et al., 2008).
IL-1b, IL-6, and IL-23 can further support growth of Th17 cells.
IL-12, IL-18, and IFN-g promote Th1 cell development. IL-4 as
well as IL-25 and IL-33 induce Th2 cell development. TGF-b, in
Cell Host & Microbe

Review

Figure 2. Three Steps to T Cell Activation

The figure shows three critical steps toward T cell activation and polarization.
(1) T cells recognize antigen not in its natural form; rather the antigen is presented by reference structures encoded by the major histocompatibility complex
(MHC). Antigens that reside in phagosomes are presented by MHC-II molecules to CD4 T cells. Because Mtb resides in phagosomes, this is the main antigen
presentation/recognition pathway in TB. Antigens in the cytosol are loaded onto MHC-I molecules that present their antigens to CD8 T cells. Unconventional
T cells comprising gd T cells and CD1-restricted T cells are activated during Mtb infection, as well. Human gd T cells recognize small molecules containing
pyrophosphate residues, also termed phospholigands. The CD1-restricted T cells recognize mycobacterial glycolipids in the context of CD1 molecules.
(2) CD4 T cells can be further subdivided into different T helper (Th) cells according to their cytokine expression pattern. The Th1 cells are critical for protection
against TB. They typically produce interferon gamma (IFN-g), tumor necrosis factor alpha (TNF-a), lymphotoxin (LT), interleukin (IL)-2, and granulocyte-macro-
phage colony-stimulating factor (GM-CSF). The Th2 cells stimulate humoral immune responses via secretion of IL-4 and IL-5. The Th17 cells produce IL-6, IL-17,
IL-21, and IL-22, which probably contribute to early protection. The regulatory T cells (Treg) produce tumor growth factor beta (TGF-b) and IL-10, which suppress
immune responses. Cytokines are also important for polarization of CD4 Th cells. Thus, IL-12, IL-18, and IFN-g promote Th1 cell development; IL-33 together
with IL-4 and IL-25 induce Th2 cell development; TGF-b, together with IL-6, promotes Th17 cell polarization in mice and TGF-b and IL-21 do so in humans, and
TGF-b together with vitamin A derivates (retinoic acid, RA) favors Treg cell maturation.
(3) For complete activation, T cells need to be costimulated by surface molecules. Many of these molecules belong to the B7 family. Thus, interactions of CD28
on T cells with B7.1 and/or B7.2 on antigen-presenting cells (APCs) promote T cell stimulation as do interactions of the inducible T cell costimulator (ICOS) on
activated T cells with ICOSL (B7.h) on APCs. Other mechanisms are inhibitory such as interactions of B7.1 and B7.2 with CTLA-4 and interactions with PD-1 on
activated T cells with PD-L on APCs.

combination with molecules other than cytokines, such as vita-
min A derivates, favors development of Treg cells that can be fur-
ther subdivided into regulatory subsets. Additional control is
achieved by counteractive effects of cytokines: IFN-g promotes
Th1 cells and blocks Th2 cells whereas IL-4 promotes Th2 cells
but blocks Th1 cells. IL-27 favors Treg cells and impairs Th17
cells (Bettelli et al., 2008).
T Cell Activation by Costimulatory Molecules
T cell activation is regulated by costimulatory molecules
(Hutloff et al., 1999; Keir et al., 2008) (Figure 2). Thus, members
of the B7 family, notably B7.1 (CD80) and B7.2 (CD86)
expressed on the surface of antigen-presenting cells (APCs),
are costimulatory molecules that interact with CD28 on T cells
and in this way initiate T cell activation. In contrast, CTLA-4,
which is also recognized by B7.1 and B7.2, inhibits T cell acti-
vation. The crosstalk between CD40 on APCs and CD40 ligand
(CD40L, CD154) on T cells can contribute to T cell activation,
and the inducible T cell costimulator (ICOS) expressed on acti-
vated T cells sustains the activation status when it interacts
with ICOS-L (B7.h) on APCs. In contrast, the programmed
death 1 molecule (PD-1) causes exhaustion of activated

Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc. 223

T cells (primarily CD8 T cells) when it reacts with its ligand PD-L
on APCs.
Instruction of Acquired Immunity by Innate Immunity
Surface expression of the different accessory molecules and
production of the various cytokines that determine activation
and polarization of T cell populations are strongly influenced
by the type of the invading pathogen (Pulendran and Ahmed,
2006). Thus, the innate immune system is critical for instructing
the acquired immune response so that it is best tailored to coun-
teract the unique survival stratagem of the invading pathogen. In
the case of Mtb, members of the toll-like receptor family (TLR)
recognize CpG motifs of bacterial DNA as general molecular
patterns for bacteria, and lipoproteins and glycolipids as more
specific characteristic patterns for mycobacteria. The respective
TLR members, TLR-9, TLR-2, and TLR-4, are central to the
buildup of an effective immune response against Mtb (Flynn
and Chan, 2001).
The Life Cycle of Mtb in Infected Individuals
During latent infection, Mtb is contained in productive granulo-
mas where it transforms into a dormant state (Figure 1). It is be-
lieved that the bacilli are in a state of nonreplicating persistence
although recent observations suggest that they may indeed
slowly replicate (Manabe and Bishai, 2000; Munoz-Elias et al.,
2005). In vitro studies have provided compelling evidence that
in this state, Mtb expresses a distinct gene expression program
characterized by upregulation of DosR genes and starvation
genes (Voskuil et al., 2003; Rosenkrands et al., 2002; Betts
et al., 2002).
TB becomes clinically apparent either when dormant Mtb
organisms are resuscitated to cause disease reactivation or sub-
sequent to reinfection with Mtb. Disease reactivation is paral-
leled by transformation of the Mtb gene-expression profile
from a dormant to a metabolically active type. It is likely that
this latter expression profile resembles that of Mtb when it enters
the host. With the onset of full-blown TB disease, granulomas liq-
uefy and Mtb organisms flourish in the caseous cellular detritus
to numbers as high as 1012 or more bacteria per lesion (Canetti,
1965). Yet, it is likely that even during active disease some
lesions remain productive and succeed in containing Mtb in
a dormant state. Hence, the lung of a TB patient may comprise
different independent entities of infection ranging from minute
productive granulomas containing few dormant mycobacteria
to large cavities housing an enormous amount of highly active
Mtb microorganisms. Probably, even a single lesion can contain
Mtb in different metabolic states.
The global gene-expression profile of Mtb residing in different
sites of lungs from patients suffering from severe TB has been
determined (Fenhalls et al., 2002; Rachman et al., 2006). Overall,
in our study more than 20% of the genome of Mtb was differen-
tially regulated in vivo as compared to in vitro-cultured samples
of the same Mtb microbes (Rachman et al., 2006). These genes
were primarily involved in (1) cell wall synthesis, (2) energy pro-
duction through fatty acid degradation under microaerophilic
conditions, and (3) active synthesis of proteins involved in eva-
sion and fortification of Mtb. Subsequent analyses compared
Mtb (1) in the center of caseous lesions, (2) in the pericavity,
and (3) at distant sites of the lung (Rachman et al., 2006). These
224 Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc.
Cell Host & Microbe
Review
gene-expression profiles revealed overlaps as well as unique
expression profiles at the different sites. Our findings strongly
support the notion that in patients suffering from active TB, dif-
ferent disease entities exist, ranging from dormant Mtb in pro-
ductive granulomas to active Mtb in cavitary lesions. Currently
available drugs primarily affect metabolically active Mtb, and it
has been speculated that after eradication of these active popu-
lations the dormant population is resuscitated and continues
the vicious cycle: a new population of Mtb organisms grows to
enormous numbers and forms new lesions. This differential
drug susceptibility of Mtb during different stages of its life cycle,
which coexist in a patient with active TB disease, is one reason
for the prolonged treatment time of at least 6 months required
to cure TB (Box S2).
It will be important to understand whether during latent infec-
tion, protective T cells respond to antigens representing the met-
abolically active stage of Mtb, the dormant stage, or both stages.
It will be equally important to determine the immune mechanisms
that precede reactivation. Do Treg cells arise prior to reactivation
and if so, which antigens do they recognize? Do effector and/or
Tm cells become exhausted, e.g., through PD-1–PD-L interac-
tions? There is experimental evidence for both types of immuno-
suppression in TB (Kursar et al., 2007; Jurado et al., 2008; Hegasy
et al., unpublished data). Moreover, it will be interesting to know
whether latent Mtb infection is sustained by effector T cells that
recognize the antigens expressed during latency, i.e., dormancy
and starvation gene products, or whether protection is sustained
by Tm cells that, for example, recognize antigens expressed early
after host invasion by Mtb such as the secreted antigens Antigen
85 and ESAT-6 (Demissie et al., 2006). Ottenhoff and coworkers
have determined human T cell responses to DosR-controlled
antigens (Leyten et al., 2006; Lin et al., 2007). Recent work from
our group using sensitive T cell stimulation assays revealed re-
sponses to subdominant latency antigens of Mtb allowing dis-
crimination of latently infected individuals from active TB patients
(Schuck et al., unpublished data). Within GC6-74, a pilot study
has analyzed T cell responses at different sites in Africa to dor-
mancy and starvation antigens and selected the most potent
ones for further longitudinal studies (GC6 Consortium, unpub-
lished data). Such analyses will determine the value of antigens
expressed during dormancy for monitoring of immune responses
during progression of latent infection to active TB disease.
Determination of antigen-specific responses is further compli-
cated by prevalence of distinct lineages of Mtb in different parts
of the world (Gagneux and Small, 2007). These lineages seem
to express different degrees of virulence. Accordingly, different
lineages and even sublineages or strains may express different
antigen expression patterns. Preliminary evidence obtained
within GC6-74 indeed indicates that the different Mtb isolates
prevalent at field sites of our collaborative project show marked
differences at the transcription level under different in vitro con-
ditions mimicking the natural history of the disease (Schoolnik
et al., unpublished data). Further studies are in progress to
determine the implications of these observations in terms of
Mtb-specific immune responses.
Immunologic Biomarkers in TB
All of these questions are important for biomarker research
focusing on immunologic mechanisms and phenotypes.
Cell Host & Microbe
Review
Obviously, concomitant determination of multiple cytokines (no-
tably IL-2, IFN-g, TNF-a, GM-CSF) and killer molecules (perforin,
granulysin), the phenotype of the T cells (CD4, CD8), expression
of memory markers, and T cell homing (various chemokine
receptors notably CCR7) and adhesion molecules (notably
CD62L) should be considered for multiplex profiling of immuno-
logic biomarkers (Figure 1). This functional assessment should
be combined with concomitant determination of specificity for
a large panel of mycobacterial antigens characteristic of dor-
mant and metabolically active Mtb. In an ideal world, these
assays would be performed by multicolor fluorescence-acti-
vated cell sorting (FACS) analysis using antibodies for critical
cell-surface molecules and cytokines combined with tetramer
constructs presenting defined mycobacterial epitopes. A more
realistic approach would be the combination of whole blood
cell assays using a panel of mycobacterial antigens for T cell
restimulation (Black et al., unpublished data) with multiplex cyto-
kine determinations by commercially available assay systems,
such as the Luminex system, which allows determination of up
to 100 different cytokines (Box S5) (Vignali, 2000).
Regional or Central Immune Regulation in TB?
Recent studies have revealed the emergence of lymphoid struc-
tures in the vicinity of granulomas in the lungs of patients suffer-
ing from TB (Ulrichs and Kaufmann, 2006) (Figure 1). In these
experiments, tissue from patients who had undergone surgical
resection of affected lung areas due to excessive multidrug-
resistant (MDR)-TB were analyzed by immunohistology (Ulrichs
et al., 2004). These analyses revealed evidence for tertiary lym-
phoid structures in the vicinity of granulomatous lesions com-
prising DCs, B cells, T cells, and macrophages. Consistent
with these observations, experimental studies with influenza
virus-infected mice have revealed that the infected lung can as-
sume functions of secondary lymphoid organs (Moyron-Quiroz
et al., 2006; Randall et al., 2008). Most recent studies from our
laboratory provide strong evidence that granulomas from Mtb-
infected mice serve as tertiary lymphoid structures capable of
inducing and sustaining T cell responses to mycobacterial and
bystander antigens (Walker et al., unpublished data). It will be
important to determine whether such local immune responses
operate in the affected organ independently from the central im-
mune system or not. The more independent the affected tissue
is, the smaller the likelihood will be that relevant T lymphocytes
recirculate through the bloodstream. This question therefore
has a direct impact on the value of peripheral blood leukocytes
for immunologic biomarker characterization.
Transcriptomics in TB
Measurement of immune biomarkers as described above is
biased by the panels of antigens, cytokines, and cell-surface
receptors preselected for analysis. This approach should be
complemented by an unbiased approach, namely, global gene-
expression profiling of peripheral blood leukocytes (Jacobsen
et al., 2005, 2008) (Figure 1). Before embarking on transcriptomic
biomarker studies within GC6-74, we have performed proof-of-
principle experiments among healthy latently infected individ-
uals, as evidenced by positive tuberculin skin test responses,
and patients suffering from active TB disease. Differentially reg-
ulated genes belong to the following four groups: (1) antimicro-
Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc. 225
bial host molecules, (2) molecules involved in inflammation, (3)
chemokines, and (4) molecules responsible for vesicle trafficking
(Jacobsen et al., 2007). Examples of differentially expressed
genes are formyl peptide receptor, bacteriocidal permeability
increasing protein, lactoferrin, a-defensin 1, a-defensin 4,
CD64 (Fcg receptor 1 with high affinity), and Rab33A (Jacobsen
et al., 2007, 2005). After appropriate deconfounding, linear dis-
criminant analysis identified three genes among the ca. 30,000
genes of the human genome that allow discrimination between
healthy latently infected individuals and active TB patients
(Jacobsen et al., 2006). These were: Rab33A, lactoferrin, and
CD64. Hence, design of a custom-made transcriptome chip or
multiplex ligation-dependent probe amplification (MLPA) for TB
measuring a restricted number of highly differentially expressed
genes appears feasible (Box S6). We considered this observa-
tion sufficiently convincing to pursue a large transcriptome
analysis within GC6-74.
Two of the three genes identified, lactoferrin and CD64, may
be of little surprise for those involved in TB research. Lactoferrin
is a transporter molecule with high affinity for iron that allows the
host to efficiently compete with Mtb for iron (Schaible and Kauf-
mann, 2005). In a mouse model, lactoferrin has been shown to
reverse a high susceptibility to TB by correcting the iron overload
(Schaible et al., 2002). CD64, the Fcg receptor with high affinity,
is upregulated by IFN-g stimulation and may contribute to host
defense against TB by stimulating the respiratory burst during
uptake of Mtb (Perussia et al., 1983; van de Winkel and Capel,
1993).
More surprising was the identification of Rab33A as a discrim-
inating factor. This molecule is a member of the Rab family of
small GTPases. Although we were originally puzzled by the func-
tional role of this molecule in TB, more recent experiments (Tsu-
boi and Fukuda, 2006) identified Rab33A, together with other
Rab molecules, in dense core vesicles. In our studies, differential
expression of Rab33A was restricted to CD8 T cells (Jacobsen
et al., 2007). Our current working hypothesis, therefore, is that
Rab33A is involved in the release of lytic molecules by CTLs.
As described above, such molecules comprise perforin and
granulysin, which have been implicated in lysis of host cells
and killing of Mtb (Stenger et al., 1998).
The exclusive identification of Rab33A in CD8 T cells also
emphasizes the importance of deconfounding for transcriptome
analysis of data obtained with heterogenous surrogate cell pop-
ulations such as peripheral blood leukocytes (Jacobsen et al.,
2006; Repsilber et al., 2005). The peripheral blood cell composi-
tion of latently infected individuals and patients with active TB
differs profoundly. While monocytes and natural killer (NK) cells
are present at higher proportions in TB patients, T cells and B
cells are present in higher proportions in latently infected individ-
uals (Jacobsen et al., 2007). Appropriate integration of bioinfor-
matics is needed to correct for these differences if determination
of the individual leukocyte populations in all samples is not
feasible as is the case within the GC6-74 project.
Glycolipids are well-known stimulators of immune-suppres-
sive mechanisms with mannose-capped LAMs being of particu-
lar importance. This molecule is known to stimulate suppressive
mechanisms by signaling through DC-SIGN (van Kooyk and Ra-
binovich, 2008; Geijtenbeek et al., 2003). Gene-expression pro-
filing of DCs stimulated by LAM cognates revealed upregulation

of PD-L1 and of molecules involved in tryptophan catabolism on
the level of transcription (Hegasy et al., unpublished data). In the
human system, PD-L1 is expressed at medium cell-surface den-
sity on DCs and activated T cells and at high density on mono-
cytes (Figure 2). Interestingly, the lung has also been found to ex-
press high levels of PD-L1 (Keir et al., 2008). PD-1 is expressed
on activated and exhausted T cells, DCs, and monocytes. It is
thus possible that during active TB, PD-1–PD-L interactions
lead to T cell exhaustion (Jurado et al., 2008; Hegasy et al., un-
published data). Moreover, our metabolomic studies revealed
a low abundance of tryptophan in serum of patients with active
TB as compared to serum of latently infected individuals (unpub-
lished data). Tryptophan degradation is a well-described cata-
bolic pathway of immune suppression with indoleamine 2, 3-di-
oxygenase (IDO) as a crucial enzyme (Munn and Mellor, 2007).
This enzyme is critically involved in activation of Treg cells, as
well as in CD4 and CD8 T cell suppression.
Metabolomic Biomarkers in TB
Metabolomics measures the global spectrum of metabolic reac-
tions of an organism during physiological or pathological condi-
tions (Box S7) (Mitchell and Carmichael, 2005). It can be as-
sumed that Mtb strongly influences metabolism of affected
tissues. It is likely that the productive lung containing dormant
bacteria during latent infection and the lung granuloma associ-
ated with excessive tissue destruction during full-blown TB
disease produce different types of metabolites. Similarly, Mtb
at different stages of its life cycle may well produce different me-
tabolites. Some of the host- and pathogen-derived metabolites
enter the circulation where they can be measured. This is the ba-
sis for metabolomic biomarker studies in TB (Idle and Gonzalez,
2007) (Figure 1). Although not part of GC6-74 we have decided to
perform proof-of-principle experiments in the area of metabolo-
mics of TB in collaboration with Metabolon Inc, USA (http://
www.metabolon.com), and we were gratified to observe that
metabolomics allowed distinction between healthy latently in-
fected individuals and patients with active TB (unpublished
data). Currently, most of the 350 identified metabolites are of
host origin. Of these almost two-thirds are unknown and one-
third has been identified unequivocally. Yet, both types of com-
pounds can be used for metabolomic discrimination between
healthy and diseased individuals. We are currently assembling
components of Mtb to broaden the analytical spectrum. We
assume that identification of unique Mtb molecules will allow
us to define one component of a biosignature useful for definitive
distinction between TB and other diseases unrelated to Mtb
infection such as Sarcoidosis or Crohn’s disease. It is also pos-
sible that volatile compounds leave lung lesions and reach the al-
veolar space from where they are released into the environment
via breathing, coughing, etc. (Figure 1). At this stage, although
measurement of volatile substances is not part of our research
strategy, we consider this an interesting avenue to pursue.
Indeed, recent evidence suggests the feasibility of such an
approach (Syhre and Chambers, 2008).
Concluding Remarks
We are aware that a long and winding road lies ahead of us
before biomarkers can be used for clinical trial monitoring and
226 Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc.
Cell Host & Microbe
Review
prognostic diagnosis in TB. Yet, we consider this a worthwhile
undertaking for the following reasons:
(1) Two billion inhabitants of this globe are infected with Mtb
and more than 10% of them are at risk of developing
active TB disease after weakening of the immune system.
The risk of developing active TB disease could be pre-
vented by prophylactic chemotherapy after diagnosis
using prognostic biomarkers.
(2) Several vaccine candidates have entered the pipeline and
hence, biomarkers could accelerate clinical trials by pro-
viding valuable information on vaccine efficacy prior to the
clinical end-point of active TB outbreak (Table 2).
(3) None of the vaccine candidates in the pipeline has pro-
vided evidence for inducing sterile eradication of Mtb,
making heterologous prime-boost stratagems composed
of novel prime vaccines (improved BCG) and new subunit
vaccines (protein/adjuvant formulations) a valid option
(Kaufmann, 2006). Biomarkers could help to provide
relevant information for educated selection of the most
promising combined vaccination against TB.
(4) Biomarkers could be equally helpful for monitoring of clin-
ical trials assessing anti-Mtb activities of existing drugs
already approved for infectious diseases other than TB
as well as for novel drug candidates that are refilling
the dried-out pipeline of TB drug development, notably,
because combinations with existing drug treatment
schemes need to be assessed (Box S2).
For basic TB researchers, biomarkers provide an added value,
namely a holistic view of in vivo functions during latent Mtb infec-
tion and active TB disease, which will open up new research
avenues (Young et al., 2008). Global biomarker studies have
sometimes been criticized as being non-hypothesis driven. We
consider the identification of novel molecules and pathways in
TB highly fertile soil for propagating research of the hypothe-
sis-driven kind. For example, differential expression of Rab33A
in CD8 T lymphocytes in the peripheral blood from patients
with active TB versus latently infected healthy individuals has
prompted us not only to reassess the functional role of CD8
T cells in TB but also to study more precisely the role of the sus-
pected CTL mechanisms expressed by CD8 T cells in TB. This is
only one of many examples that have fueled new research ave-
nues. We consider it rewarding that findings obtained in the field
are being transferred back to the academic laboratory thus re-
translating field observations into hypothesis-driven elucidation
of basic mechanisms. We are optimistic that collaborations like
ours in GC6-74 between scientists in the northern and southern
hemispheres will benefit both partners equally by producing the
basis for novel control measures as well as for innovative
research on basic disease mechanisms.
SUPPLEMENTAL DATA
Supplemental Data include seven boxes and can be found with this article online
at http://www.cellhostandmicrobe.com/cgi/content/full/4/3/219/DC1/.
ACKNOWLEDGMENTS
The authors are grateful to M.L. Grossman for editorial help, J. Maertzdorf for
preparation of the transcriptomics box, and D. Schad and L. Fehlig for graphic
Cell Host & Microbe
Review
support. We also thank members of our department for sharing unpublished
work. Studies on TB in our department receive financial support from the fol-
lowing: Bill & Melinda Gates Foundation Grand Challenges in Global Health
Grant No. 37772, the European Union Framework Program 6 Projects, Design
and Testing of Vaccine Candidates against Tuberculosis (TBVAC) and
MUVAPRED (Mucosal Vaccines for Poverty Related Diseases); Sonderfor-
schungsbereich Protektive und pathologische Folgen der Antigen-Verarbei-
tung (SFB 421); and Bundesministerium für Bildung und Forschung Networks
PathoGenoMikPlus and Nationales Genomforschungsnetz (NGFN). The
authors are particularly grateful for the collaborative efforts of GC6-74 consor-
tium members.
REFERENCES
Armstrong, J.A., and Hart, P.D. (1971). Response of cultured macrophages to
Mycobacterium tuberculosis, with observations on fusion of lysosomes with
phagosomes. J. Exp. Med. 134, 713–740.
Belkaid, Y., and Rouse, B.T. (2005). Natural regulatory T cells in infectious dis-
ease. Nat. Immunol. 6, 353–360.
Bettelli, E., Korn, T., Oukka, M., and Kuchroo, V.K. (2008). Induction and
effector functions of T(H)17 cells. Nature 453, 1051–1057.
Betts, J.C., Lukey, P.T., Robb, L.C., McAdam, R.A., and Duncan, K. (2002).
Evaluation of a nutrient starvation model of Mycobacterium tuberculosis
persistence by gene and protein expression profiling. Mol. Microbiol. 43,
717–731.
Biomarkers Definitions Working Group (2001). Biomarkers and surrogate end-
points: preferred definitions and conceptual framework. Clin. Pharmacol. Ther.
69, 89–95.
Boasso, A., Shearer, G.M., and Clerici, M. (2008). The hunt for an HIV vaccine:
time to rethink recent failures. Lancet 371, 1897–1898.
Canetti, G. (1965). Present aspects of bacterial resistance in tuberculosis. Am.
Rev. Respir. Dis. 92, 687–703.
Charo, I.F., and Ransohoff, R.M. (2006). The many roles of chemokines and
chemokine receptors in inflammation. N. Engl. J. Med. 354, 610–621.
Colburn, W.A. (2000). Optimizing the use of biomarkers, surrogate endpoints,
and clinical endpoints for more efficient drug development. J. Clin. Pharmacol.
40, 1419–1427.
Corbett, E.L., Watt, C.J., Walker, N., Maher, D., Williams, B.G., Raviglione,
M.C., and Dye, C. (2003). The growing burden of tuberculosis: global trends
and interactions with the HIV epidemic. Arch. Intern. Med. 163, 1009–1021.
Demissie, A., Leyten, E.M., Abebe, M., Wassie, L., Aseffa, A., Abate, G.,
Fletcher, H., Owiafe, P., Hill, P.C., Brookes, R., et al. (2006). Recognition of
stage-specific mycobacterial antigens differentiates between acute and
latent infections with Mycobacterium tuberculosis. Clin. Vaccine Immunol.
13, 179–186.
Fenhalls, G., Stevens, L., Moses, L., Bezuidenhout, J., Betts, J.C., Helden,
P.P., Lukey, P.T., and Duncan, K. (2002). In situ detection of Mycobacterium
tuberculosis transcripts in human lung granulomas reveals differential gene
expression in necrotic lesions. Infect. Immun. 70, 6330–6338.
Fine, P.E. (1995). Variation in protection by BCG: implications of and for
heterologous immunity. Lancet 346, 1339–1345.
Flynn, J.L., and Chan, J. (2001). Immunology of tuberculosis. Annu. Rev.
Immunol. 19, 93–129.
Gagneux, S., and Small, P.M. (2007). Global phylogeography of Mycobacte-
rium tuberculosis and implications for tuberculosis product development.
Lancet Infect. Dis. 7, 328–337.
Geijtenbeek, T.B., van Vliet, S.J., Koppel, E.A., Sanchez-Hernandez, M.,
Vandenbroucke-Grauls, C.M., Appelmelk, B., and van Kooyk, Y. (2003). Myco-
bacteria target DC-SIGN to suppress dendritic cell function. J. Exp. Med. 197,
7–17.
Hutloff, A., Dittrich, A.M., Beier, K.C., Eljaschewitsch, B., Kraft, R., Anagnosto-
poulos, I., and Kroczek, R.A. (1999). ICOS is an inducible T-cell co-stimulator
structurally and functionally related to CD28. Nature 397, 263–266.
Idle, J.R., and Gonzalez, F.J. (2007). Metabolomics. Cell Metab. 6, 348–351.
Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc. 227
Jacobsen, M., Mattow, J., Repsilber, D., and Kaufmann, S.H. (2008). Novel
strategies to identify biomarkers in tuberculosis. Biol. Chem. 389, 487–495.
Jacobsen, M., Repsilber, D., Gutschmidt, A., Neher, A., Feldmann, K., Mollen-
kopf, H.J., Ziegler, A., and Kaufmann, S.H.E. (2005). Ras-associated small
GTPase 33A, a novel T cell factor, is down-regulated in patients with tubercu-
losis. J. Infect. Dis. 192, 1211–1218.
Jacobsen, M., Repsilber, D., Gutschmidt, A., Neher, A., Feldmann, K., Mollen-
kopf, H.J., Kaufmann, S.H., and Ziegler, A. (2006). Deconfounding microarray
analysis - independent measurements of cell type proportions used in a
regression model to resolve tissue heterogeneity bias. Methods Inf. Med. 45,
557–563.
Jacobsen, M., Repsilber, D., Gutschmidt, A., Neher, A., Feldmann, K., Mollen-
kopf, H.J., Ziegler, A., and Kaufmann, S.H. (2007). Candidate biomarkers for
discrimination between infection and disease caused by Mycobacterium
tuberculosis. J. Mol. Med. 85, 613–621.
Jurado, J.O., Alvarez, I.B., Pasquinelli, V., Martinez, G.J., Quiroga, M.F., Ab-
bate, E., Musella, R.M., Chuluyan, H.E., and Garcia, V.E. (2008). Programmed
death (PD)-1:PD-ligand 1/PD-ligand 2 pathway inhibits T cell effector functions
during human tuberculosis. J. Immunol. 181, 116–125.
Kaufmann, S.H. (2007). The contribution of immunology to the rational design
of novel antibacterial vaccines. Nat. Rev. Microbiol. 5, 491–504.
Kaufmann, S.H., and Parida, S.K. (2007). Changing funding patterns in tuber-
culosis. Nat. Med. 13, 299–303.
Kaufmann, S.H.E. (2006). Envisioning future strategies for vaccination against
tuberculosis. Nat. Rev. Immunol. 6, 699–704.
Kaufmann, S.H.E., and McMichael, A.J. (2005). Annulling a dangerous liaison:
vaccination strategies against AIDS and tuberculosis. Nat. Med. 11, S33–S44.
Kaufmann, S.H.E., and Winau, F. (2005). From bacteriology to immunology:
the dualism of specificity. Nat. Immunol. 6, 1063–1066.
Keir, M.E., Butte, M.J., Freeman, G.J., and Sharpe, A.H. (2008). PD-1 and its
ligands in tolerance and immunity. Annu. Rev. Immunol. 26, 677–704.
Khader, S.A., Bell, G.K., Pearl, J.E., Fountain, J.J., Rangel-Moreno, J., Cilley,
G.E., Shen, F., Eaton, S.M., Gaffen, S.L., Swain, S.L., et al. (2007). IL-23 and
IL-17 in the establishment of protective pulmonary CD4(+) T cell responses
after vaccination and during Mycobacterium tuberculosis challenge. Nat.
Immunol. 8, 369–377.
Kursar, M., Koch, M., Mittrucker, H.W., Nouailles, G., Bonhagen, K., Kamradt,
T., and Kaufmann, S.H. (2007). Cutting edge: regulatory T cells prevent effi-
cient clearance of Mycobacterium tuberculosis. J. Immunol. 178, 2661–2665.
Lagakos, S.W., and Gable, A.R. (2008). Challenges to HIV prevention–seeking
effective measures in the absence of a vaccine. N. Engl. J. Med. 358, 1543–
1545.
Lalvani, A. (2007). Diagnosing tuberculosis infection in the 21st century: new
tools to tackle an old enemy. Chest 131, 1898–1906.
Leyten, E.M., Lin, M.Y., Franken, K.L., Friggen, A.H., Prins, C., van Meijgaar-
den, K.E., Voskuil, M.I., Weldingh, K., Andersen, P., Schoolnik, G.K., et al.
(2006). Human T-cell responses to 25 novel antigens encoded by genes of
the dormancy regulon of Mycobacterium tuberculosis. Microbes Infect. 8,
2052–2060.
Li, M.O., and Flavell, R.A. (2008). Contextual regulation of inflammation: a duet
by transforming growth factor-beta and interleukin-10. Immunity 28, 468–476.
Lin, M.Y., Geluk, A., Smith, S.G., Stewart, A.L., Friggen, A.H., Franken, K.L.,
Verduyn, M.J., van Meijgaarden, K.E., Voskuil, M.I., Dockrell, H.M., et al.
(2007). Lack of immune responses to Mycobacterium tuberculosis DosR reg-
ulon proteins following Mycobacterium bovis BCG vaccination. Infect. Immun.
75, 3523–3530.
Maher, D. (2008). Tuberculosis Control: Good Clinical Care and Good Public
Health, S.H.E. Kaufmann and P.D. van Helden, eds. (Weinheim: Wiley-VCH),
pp. 115–130.
Manabe, Y.C., and Bishai, W.R. (2000). Latent Mycobacterium tuberculosis-
persistence, patience, and winning by waiting. Nat. Med. 6, 1327–1329.
McGeachy, M.J., and Cua, D.J. (2008). Th17 cell differentiation: the long and
winding road. Immunity 28, 445–453.

McKinney, J.D., Jacobs, W.R., and Bloom, B.R. (1998). Persisting problems in
tuberculosis. In Emerging Infections, R. Krause, J.I. Gallin, and A.S. Fauci, eds.
(New York: Academic Press), pp. 51–146.
Mitchell, S.C., and Carmichael, P.L. (2005). Metabonomics and the endocrine
system. Mol. Cell. Endocrinol. 244, 10–14.
Moyron-Quiroz, J.E., Rangel-Moreno, J., Hartson, L., Kusser, K., Tighe, M.P.,
Klonowski, K.D., Lefrancois, L., Cauley, L.S., Harmsen, A.G., Lund, F.E., and
Randall, T.D. (2006). Persistence and responsiveness of immunologic memory
in the absence of secondary lymphoid organs. Immunity 25, 643–654.
Mueller, C., Muller, B., and Perruchoud, A.P. (2008). Biomarkers: past, pres-
ent, and future. Swiss Med. Wkly. 138, 225–229.
Munn, D.H., and Mellor, A.L. (2007). Indoleamine 2,3-dioxygenase and tumor-
induced tolerance. J. Clin. Invest. 117, 1147–1154.
Munoz-Elias, E.J., Timm, J., Botha, T., Chan, W.T., Gomez, J.E., and
McKinney, J.D. (2005). Replication dynamics of Mycobacterium tuberculosis
in chronically infected mice. Infect. Immun. 73, 546–551.
Nathan, C.F., Murray, H.W., Wiebe, M.E., and Rubin, B.Y. (1983). Identification
of interferon-gamma as the lymphokine that activates human macrophage
oxidative metabolism and antimicrobial activity. J. Exp. Med. 158, 670–689.
Ouyang, W., Kolls, J.K., and Zheng, Y. (2008). The biological functions of
T helper 17 cell effector cytokines in inflammation. Immunity 28, 454–467.
Perussia, B., Dayton, E.T., Fanning, V., Thiagarajan, P., Hoxie, J., and
Trinchieri, G. (1983). Immune interferon and leukocyte-conditioned medium
induce normal and leukemic myeloid cells to differentiate along the monocytic
pathway. J. Exp. Med. 158, 2058–2080.
Pulendran, B., and Ahmed, R. (2006). Translating innate immunity into immu-
nological memory: implications for vaccine development. Cell 124, 849–863.
Rachman, H., Strong, M., Ulrichs, T., Grode, L., Schuchhardt, J., Mollenkopf,
H., Kosmiadi, G.A., Eisenberg, D., and Kaufmann, S.H.E. (2006). Unique
transcriptome signature of Mycobacterium tuberculosis in pulmonary tubercu-
losis. Infect. Immun. 74, 1233–1242.
Randall, T.D., Carragher, D.M., and Rangel-Moreno, J. (2008). Development of
secondary lymphoid organs. Annu. Rev. Immunol. 26, 627–650.
Repsilber, D., Fink, L., Jacobsen, M., Blasing, O., and Ziegler, A. (2005). Sam-
ple selection for microarray gene expression studies. Methods Inf. Med. 44,
461–467.
Rosenkrands, I., Slayden, R.A., Crawford, J., Aagaard, C., Barry, C.E., III, and
Andersen, P. (2002). Hypoxic response of Mycobacterium tuberculosis studied
by metabolic labeling and proteome analysis of cellular and extracellular
proteins. J. Bacteriol. 184, 3485–3491.
Russell, D.G. (2007). Who puts the tubercle in tuberculosis? Nat. Rev. Micro-
biol. 5, 39–47.
Sallusto, F., Geginat, J., and Lanzavecchia, A. (2004). Central memory and
effector memory T cell subsets: function, generation, and maintenance.
Annu. Rev. Immunol. 22, 745–763.
Schaible, U.E., and Kaufmann, S.H.E. (2005). A nutritive view on the host-path-
ogen interplay. Trends Microbiol. 13, 373–380.
Schaible, U.E., and Kaufmann, S.H. (2007). Malnutrition and infection: com-
plex mechanisms and global impacts. PLoS Med. 4, e115. 10.1371/journal.
pmed.0040115.
Schaible, U.E., Collins, H.L., Priem, F., and Kaufmann, S.H.E. (2002). Correc-
tion of the iron overload defect in beta-2-microglobulin knockout mice by
lactoferrin abolishes their increased susceptibility to tuberculosis. J. Exp.
Med. 196, 1507–1513.
Schaible, U.E., Winau, F., Sieling, P.A., Fischer, K., Collins, H.L., Hagens, K.,
Modlin, R.L., Brinkmann, V., and Kaufmann, S.H.E. (2003). Apoptosis facili-
228 Cell Host & Microbe 4, September 11, 2008 ª2008 Elsevier Inc.
Cell Host & Microbe
Review
tates antigen presentation to T lymphocytes through MHC-I and CD1 in tuber-
culosis. Nat. Med. 9, 1039–1046.
Seder, R.A., Darrah, P.A., and Roederer, M. (2008). T-cell quality in memory
and protection: implications for vaccine design. Nat. Rev. Immunol. 8,
247–258.
Stenger, S., Hanson, D.A., Teitelbaum, R., Dewan, P., Niazi, K.R., Froelich,
C.J., Ganz, T., Thoma-Uszynski, S., Melian, A., Bogdan, C., et al. (1998). An
antimicrobial activity of cytolytic T cells mediated by granulysin. Science
282, 121–125.
Syhre, M., and Chambers, S.T. (2008). The scent of Mycobacterium tubercu-
losis. Tuberculosis (Edinb.) 88, 317–323.
Tsuboi, T., and Fukuda, M. (2006). Rab3A and Rab27A cooperatively regulate
the docking step of dense-core vesicle exocytosis in PC12 cells. J. Cell Sci.
119, 2196–2203.
Ulrichs, T., and Kaufmann, S.H.E. (2006). New insights into the function of
granulomas in human tuberculosis. J. Pathol. 208, 261–269.
Ulrichs, T., Kosmiadi, G.A., Trusov, V., Jörg, S., Pradl, L., Titukhina, M., Mis-
henko, V., Gushina, N., and Kaufmann, S.H.E. (2004). Human tuberculous
granulomas induce peripheral lymphoid follicle-like structures to orchestrate
local host defence in the lung. J. Pathol. 204, 217–228.
van de Winkel, J.G., and Capel, P.J. (1993). Human IgG Fc receptor hetero-
geneity: molecular aspects and clinical implications. Immunol. Today 14,
215–221.
van der Wel, N., Hava, D., Houben, D., Fluitsma, D., van Zon, M., Pierson, J.,
Brenner, M., and Peters, P.J. (2007). M. tuberculosis and M. leprae trans-
locate from the phagolysosome to the cytosol in myeloid cells. Cell 129,
1287–1298.
van Kooyk, Y., and Rabinovich, G.A. (2008). Protein-glycan interactions in
the control of innate and adaptive immune responses. Nat. Immunol. 9,
593–601.
Vandal, O.H., Pierini, L.M., Schnappinger, D., Nathan, C.F., and Ehrt, S. (2008).
A membrane protein preserves intrabacterial pH in intraphagosomal Myco-
bacterium tuberculosis. Nat. Med. 14, 849–854.
Vignali, D.A. (2000). Multiplexed particle-based flow cytometric assays.
J. Immunol. Methods 243, 243–255.
Voskuil, M.I., Schnappinger, D., Visconti, K.C., Harrell, M.I., Dolganov, G.M.,
Sherman, D.R., and Schoolnik, G.K. (2003). Inhibition of respiration by nitric
oxide induces a Mycobacterium tuberculosis dormancy program. J. Exp.
Med. 198, 705–713.
Walker, B.D., and Burton, D.R. (2008). Toward an AIDS vaccine. Science 320,
760–764.
WHO (2008). Global tuberculosis control: surveillance, planning, financing.
http://www.who.int/tb/publications/global_report/en/index.html. Geneva, World
Health Organization.
Winau, F., Weber, S., Sad, S., de Diego, J., Hoops, S.L., Breiden, B., Sandhoff,
K., Brinkmann, V., Kaufmann, S.H.E., and Schaible, U.E. (2006). Apoptotic
vesicles crossprime CD8 T cells and protect against tuberculosis. Immunity
24, 105–117.
Wolf, M., Albrecht, S., and Marki, C. (2008). Proteolytic processing of chemo-
kines: implications in physiological and pathological conditions. Int. J. Bio-
chem. Cell Biol. 40, 1185–1198.
Yang, L., Anderson, D.E., Baecher-Allan, C., Hastings, W.D., Bettelli, E.,
Oukka, M., Kuchroo, V.K., and Hafler, D.A. (2008). IL-21 and TGF-beta are
required for differentiation of human T(H)17 cells. Nature 454, 350–352.
Young, D., Stark, J., and Kirschner, D. (2008). Systems biology of persistent
infection: tuberculosis as a case study. Nat. Rev. Microbiol. 6, 520–528.