Cellular Microbiology (2010) 12(3), 301–309
Microreview
cmi_1424
doi:10.1111/j.1462-5822.2009.01424.x
First published online 20 January 2010
301..309
Penitentiary or penthouse condo: the tuberculous
granuloma from the microbe’s point of view
Carleitta Paige and William R. Bishai*
Department of Medicine, Johns Hopkins University
School of Medicine, Baltimore, MD, USA.
Summary
Granuloma formation represents a pivotal point
during human infection with Mycobacterium tuberculosis, for this structure may limit mycobacterial
spread and prevent active disease, while at the
same time allow for the survival and persistence of
viable mycobacteria within the host. The current
therapeutic regimens for treating tuberculosis
disease have proven effective in developing countries. However, in countries with large populations,
limited access to health care, and high incidence of
HIV co-infection, tuberculosis disease continues
to represent a major global health emergency. Particularly, the emergence of extensively and multidrug-resistant forms of tuberculosis underscores
the need develop new treatment strategies. Recent
mechanistic studies have identified bacterial virulence mechanisms that subvert host responses and
lead to an inappropriate upregulation of host
factors such as tumour necrosis factor-a (TNF-a)
and matrix metalloproteinases (MMPs). Paradoxically, then, part of the mycobacterial virulence
programme may be to promote granuloma development and maturation. These observations suggest that together with appropriate anti-microbials
host-based therapeutics directed at TNF-a and
MMP inhibition may counteract the microbial subterfuge, reduce the pro-granulomatous response,
and offer an enhanced therapeutic effect. Hostdirected therapy that alters the immune response
may offer an alternative approach towards reducing
treatment duration, the risk of anti-microbial resistance and improving patient outcome.
Received 9 October, 2009; revised 22 November, 2009; accepted 23
November, 2009. *For correspondence. E-mail wbishai@jhmi.edu;
Tel. (+1) 410 955 3507; Fax (+1) 410 614 8173.
The tuberculous granuloma: a mycobacterial
perspective
Mycobacterium tuberculosis infection elicits a robust
immunological response, which results in an aggregation
of host and mycobacterial components, otherwise
referred to as a granuloma (Russell, 2007). Granuloma
formation is commonly referred to as a host-protective
strategy to limit mycobacterial replication and prevent the
spread of infection. However, within the granulomatous
environment, M. tuberculosis may be shielded from
immune-based killing mechanisms, and anti-tuberculous
drugs may have limited penetration into this compartment.
Recent molecular studies indicate that M. tuberculosis
possesses mechanisms that deliberately promote cellular
recruitment to the nascent granuloma, suggesting that
granuloma formation is part of a pathogen-directed virulence programme. Therefore, while granuloma formation
appears to function as a host defence mechanism, there
are apparent mycobacterial advantages to this process
(Bold and Ernst, 2009; Rubin, 2009).
Tumour necrosis factor (TNF-a), a key proinflammatory cytokine produced by macrophages, dendritic cells and T cells, promotes immune cell migration to
the site of infection and contributes to the dynamic nature
of the granulomatous environment (Flynn and Chan,
2005). TNF-a is thought to mediate a successful host
response to mycobacteria through increasing the cytotoxicity of M. tuberculosis-infected macrophages (Keane
et al., 1997). In addition, TNF-a appears to maintain the
overall structural integrity of the granuloma as observed in
a murine model, where TNF-a neutralization therapy
resulted in fatal reactivation of persistent tuberculosis that
was characterized by increased tissue bacillary burden
and severe histopathological deterioration in the mouse
lungs (Mohan et al., 2001). More recent studies demonstrate that TNF-a may regulate the granulomatous
response by exerting an anti-inflammatory effect through
modulation of the expression of pro-inflammatory mediators (Chakravarty et al., 2008). Given the host TNF-a
responses, it has also been demonstrated that M. tuberculosis contributes to granuloma formation via TNF-a
stimulation. Specifically, the lipid-rich mycobacterial cell
© 2010 Blackwell Publishing Ltd
cellular microbiology
302 C. Paige and W. R. Bishai
Fig. 1. Tumour necrosis factor-a (TNF-a) balancing act and granuloma formation during primary infection, latency/containment, and
post-primary reactivation TB. M. tuberculosis has recently been shown to inject cAMP into host macrophages using a bacterial adenylate
cyclase, Rv0386 (Agarwal et al., 2009). The burst of cAMP in macrophages drives the secretion of TNF-a via a protein kinase A,
CREB-dependent mechanism. It is hypothesized that during the early days of primary infection, M. tuberculosis drives an excessive TNF-a
response (together with other mechanisms such as MMP secretion) to deliberately promote granuloma formation, and that the resulting
recruitment of additional naïve macrophages and the tissue pathology foster increased bacterial proliferation in the days prior to the
development of a complete acquired immune response (right). Later, in the face of acquired cell-mediated antimicrobial responses, stable
granulomas may form in which the tubercle bacilli enter a latent state of limited or non-replication (bottom). Later in life, due to immune
perturbation, abnormally low TNF-a levels may be associated with loss of granuloma integrity, failed anti-microbial containment and
subsequent post-primary reactivation tuberculosis disease (left).
wall composed of trehalose dimycolate is known to induce
a granulomatous inflammatory response through induction of TNF-a (Flynn and Chan, 2005).
A recent study indicates that beyond antigenic stimulation, M. tuberculosis may increase TNF-a levels secreted
from infected macrophage by direct interference with
intracellular second messenger signalling. An M. tuberculosis mutant lacking the Rv0386 adenylate cyclase
elicited a reduced intracellular cyclic adenosine monophosphate (cAMP) response following infection (Agarwal
et al., 2009). This was linked to reduced protein kinase A
(PKA) and cAMP response-element-binding protein
(CREB) activation, and a significant reduction of macrophage TNF-a secretion by mutant-infected macrophages. Importantly, the mutant survived poorly in mouse
infections and gave reduced lung pathology. Radiolabelling studies indicated that wild-type M. tuberculosis uses
the Rv0386 adenylate cyclase to deliver cAMP to the
macrophage cytoplasm in order to initiate this cascade of
events ending in TNF-a upregulation. Therefore, Rv0386
is a bacterial enzyme that acts to subvert host-cell signal
transduction in a manner that leads to a progranulomatous response with excess TNF-a secretion;
loss of this enzyme is associated with reduced TNF-a
levels in mouse lungs and poorer bacterial survival
(Agarwal et al., 2009). Although TNF-a is generally con-
sidered to be required for host containment of tuberculosis as discussed below, this recent study raises the
possibility that at certain phases of the infection cycle, the
microbe elicits an excessive TNF-a response as part of a
virulence strategy (Fig. 1).
As M. tuberculosis establishes an infection, the local
cellular organization of the lung is modified to facilitate
macrophage infiltration and initiation of granuloma formation. While the mechanisms of tissue remodelling by the
tuberculous granuloma are poorly understood, it has been
hypothesized that the matrix metalloproteinases (MMPs)
may play a role (Elkington and Friedland, 2006). The
MMPs form a family of zinc-dependent proteases, which
have the unique ability to degrade the extremely stable
collagen backbone of the lung extracellular matrix (ECM)
(Page-McCaw et al., 2007). The MMPs are inhibited by
specific endogenous tissue inhibitors of metalloproteinases (TIMPs), which comprise a family of four protease
inhibitors: TIMP1, TIMP2, TIMP3 and TIMP4. Infection by
M. tuberculosis has been shown to induce the secretion of
MMP-1, -2, -7 and -9 (Chang et al., 1996; QuidingJarbrink et al., 2001; Elkington et al., 2005; Rand et al.,
2009), as well as decrease expression of TIMP-1, -2 and
-3 (Brew et al., 2000) from peripheral blood mononuclear
cells and human airway epithelial cells. While the tissuespecific immune cell responses of MMP secretion have
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
Anti-granuloma strategies for tuberculosis 303
yet to be evaluated, these studies suggest a similar
mechanism. Moreover, the expression of several MMPs,
including MMP-1 and MMP-3, has been shown to be
dependent upon cAMP through PGE2 stimulation of host
G-protein-coupled receptor adenylate cyclase (RangelMoreno et al., 2002; Lai et al., 2003; Chen et al., 2008).
This suggests that bacterial mechanisms that modulate
cellular cAMP pools (i.e., the M. tuberculosis Rv0386
pathway) could influence the expression of these MMPs.
MMP-9 has been demonstrated to be involved in macrophage recruitment and granuloma development in an in
vivo mouse model of tuberculosis (Taylor et al., 2006).
Studies of M. tuberculosis infection in MMP-9-deficient
mice revealed that host MMP-9 is required for full bacterial
proliferation, as these mice showed reduced bacterial
burden and reduced lung macrophage recruitment compared with wild type. In addition, the MMP-9 knockout
mice displayed reduced ability to produce well-formed
granulomas.
Given the contributions of M. tuberculosis to stimulate
and facilitate granuloma formation, there is recent evidence to suggest that M. tuberculosis also exploits this
process for local expansion and systemic dissemination.
The Mycobacterium marinum-zebrafish model utilizes
intravital imaging to observe consequences of the infection in real time, evaluating the fate of infected and uninfected phagocytes. Davis et al. found that infected cells
recruit uninfected macrophages to nascent zebrafish
granulomas, and that the recruited cells ingest bacteria
and apoptotic bodies (Davis and Ramakrishnan, 2009).
The newly infected macrophages sustain continued proliferation of the mycobacteria enabling the granuloma to
expand. These data suggest that M. marinum directs host
signalling in a manner leading to naïve cell chemoattraction, and that newly arriving cells foster granuloma formation and expansion.
Together, these recent developments in M. tuberculosis
pathogenesis suggest that pathogenic mycobacteria
possess specific mechanisms to subvert host signalling in
a pro-inflammatory manner that favours the development
of granulomas, and that these same mechanisms facilitate bacterial survival within the host. These findings
support the concept of tissue-based remodelling by the
pathogen towards granuloma formation. Paradoxically,
the granuloma may be less like a penitentiary, confining
the microbe, and more akin to luxury penthouse condo
facilitating the survival of a manipulative intruder (Fig. 2).
Fig. 2. Pathogen-driven granuloma expansion. Specific M.
tuberculosis virulence mechanisms exploiting the cAMP signal
transduction pathways may promote recruitment of naïve
macrophages to the nascent granuloma. cAMP intoxication of
infected macrophages via an Rv0386 adenylate cyclase is
associated with increased expression of TNF-a and several MMPs.
During the early phase of infection, this aberrant signalling – driven
by the pathogen – may represent a virulence scheme to fuel
bacterial proliferation by attracting naive cells which may
subsequently be infected.
phenotypic resistance to standard chemotherapy in
active tuberculosis. Host-directed therapy may modulate
host responses that are important in both LTBI and
active tuberculosis. Strategies for host-based therapy
include the targeting of intracellular mycobacterial
growth within macrophages, enhancement of cytotoxic
elimination of infected host cells, or alternatively, modification of the immune response in order to reduce interference with chemotherapy.
Biologic disease modifying agents
Early efforts in tuberculosis immunotherapy
A major dilemma in treating tuberculosis is effectively
eliminating dormant bacilli that may exist within the
granulomatous areas in latent tuberculosis infection
(LTBI) and in more rapid killing of persisters that display
Immune intervention therapy against tuberculosis disease
is not a new concept; cytokine and other immunomodulatory therapies have been evaluated for their effectiveness against tuberculosis disease (Rook et al., 2007).
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
304 C. Paige and W. R. Bishai
Briefly, interferon (IFN)-g, essential for anti-mycobacterial
host defences, was investigated in several small trials of
aerosolized IFN-g treatment of multi-drug-resistant pulmonary tuberculosis. IFN-g, which plays key role for the activation of macrophages, is produced by cytotoxic T
lymphocytes as well as NK cells. As a therapeutic, it has
been hypothesized that this immune modulator will
enhance the cytotoxic activity of macrophages to render a
more robust immune response and reduce pulmonary
inflammation. Dawson et al. performed a randomized,
controlled clinical trial of directly observed therapy
(DOTS) versus DOTS supplemented with IFN-g administered subcutaneously or by nebulizer over 4 months to 89
patients with cavitary pulmonary tuberculosis. They concluded that recombinant IFN-g in combination with directly
observed therapy can reduce inflammatory cytokines at
the site of infection, improve bacterial clearance from the
sputum, and improve constitutional symptoms of tuberculosis disease (Dawson et al., 2009). However, earlier
studies did not show a long-term microbiological effect,
indicating that additional studies are needed to determine
the durability and long-term impact of IFN-g therapy
(Giosue et al., 2000; Koh et al., 2004; Suarez-Mendez
et al., 2004).
Interleukin 2 (IL-2) is another cytokine evaluated for its
immunotherapeutic potential. IL-2 promotes T-cell replication and is essential for cellular immune function and
granuloma formation. In 1997, low-dose recombinant
human IL-2 adjunctive immunotherapy was administered
to multi-drug-resistant tuberculosis patients daily or in
5-day pulses (Johnson et al., 1997). The daily regimen
exhibited reduced or cleared sputum bacterial load;
however, this treatment also stimulated immune activation, likely enhancing the antimicrobial response. A
follow-up larger study assessed IL-2 in a randomized,
placebo-controlled, double-blinded trial of 110 HIVnegative active tuberculosis patients who received IL-2
adjunctive immunotherapy (Johnson et al., 2003). While
IL-2 was found to be safe in the latter study, in contrast to
the earlier finding it did not enhance bacillary clearance or
show differences in symptom resolution rates.
Anti-TNF therapy
The TNF-a-dependent homeostasis between host and M.
tuberculosis during infection and disease may represent
an understudied therapeutic opportunity for tuberculosis
treatment. Initial efforts in anti-TNF-a therapy involved the
use of thalidomide or its analogues to determine their
ability to modify the granulomatous environment and
render the tubercle bacilli more responsive to the bactericidal action of chemotherapy. This hypothesis was tested
in a tuberculous meningitis rabbit model comparing the
efficiency of anti-tuberculosis antibiotics in the presence
or absence of thalidomide (Tsenova et al., 1998). Those
receiving antibiotics alone experienced inflammation
despite a decrease in mycobacterial burden, and 50%
succumbed to infection. Conversely, the addition of thalidomide to anti-TB drug therapy led to a reduction in
TNF-a levels, leukocytosis, brain pathology, and was
associated with 100% survival of all rabbits. The thalidomide analogue, IMiD3, was also proven effective as
adjunctive therapy using the same model (Tsenova et al.,
2002). Unfortunately, thalidomide has an infamous tendency to induce birth defects, which has limited its availability and use. However, these data do support immune
modulation with TNF-a inhibitors, as they are associated
with accelerated bacillary clearance by antibiotics and
reduced pathological sequellae.
Approximately 10 years ago, TNF-a antagonists etanercept, infliximab, and later adalimumab, certolizumab
and golimumab were approved for treating chronic inflammatory conditions, including rheumatoid arthritis and in
some settings psoriasis and Crohn’s disease. During the
first 6 months after initiating anti-TNF-a therapy, there
was an increased risk of reactivation tuberculosis (Keane
et al., 2001). About half of these patients developed extrapulmonary or disseminated forms of tuberculosis. This
association of tuberculosis reactivation with decreased
TNF-a activity suggests that this cytokine has a key role in
the control of latent tuberculosis. Indeed, when patients
are screened for LTBI and appropriately treated prior to
initiation of TNF-a antagonist therapy, the risk of reactivation TB drops to near baseline levels (Schiff et al., 2006).
Certainly, the outcome of these experiences indicates an
essential role for TNF-a in the adequate maintenance of
LTBI.
These observations have prompted studies as to
whether adjunctive therapy with TNF-a antagonists may
reduce granuloma integrity and/or stimulate bacterial replication and thereby potentiate anti-TB drug therapy in
active TB. Indeed clinical trials evaluating two cohorts of
HIV-seropositive patients with drug-susceptible active
tuberculosis were treated with standard anti-TB drug
therapy with or without etanercept. Patients who received
adjunctive TNF-a inhibitor therapy with etanercept for the
first 4 weeks of treatment demonstrated an acceleration of
sputum conversion to culture negativity (Wallis et al.,
2004; Wallis, 2005). Recently, case reports have also
documented the therapeutic potential of adding TNF-a
antagonists to an effective drug regimen for patients with
extensive tuberculosis disease who were failing standard
anti-TB drug therapy. In one instance, a South African
rheumatoid arthritis patient developed extensive pulmonary tuberculosis after 8 months on adalimumab. Immediately, adalimumab was withdrawn and standard
tuberculosis therapy was initiated. Despite having
fully susceptible M. tuberculosis at diagnosis, her status
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
Anti-granuloma strategies for tuberculosis 305
deteriorated, and she required mechanical ventilation. As
a last-hope effort her physicians added back adalimumab
on day 17, and her status promptly improved allowing
mechanical ventilation to be stopped and paving the way
for the patient ultimately to make a full recovery (Wallis
et al., 2009; Wallis, 2009).
While the aforementioned studies suggest that TNF-a
could be used as an effective adjuvant with standard
tuberculosis treatment, there are several concerns associated with its use. It is frequently difficult to know a priori
whether patients have drug-susceptible or drug-resistant
TB, and use of TNF-a antagonists with unsuspected drugresistant disease could be a dangerous combination.
Furthermore, while the data available to date support a
possible acceleration of early response to treatment, there
is inadequate knowledge on whether adjunctive TNF-a
antagonists would also be associated with sterilization and
durable cure. Nevertheless, the limited studies and case
reports available to date are supportive of the concept that
in the presence of active anti-TB drugs, interference with
TNF-a-mediated granuloma containment may have beneficial effects on patient response rates.
Anti-MMP therapy
During active tuberculosis disease, pulmonary cavitation
is an important contributor for effective transmission of M.
tuberculosis. Cavity formation occurs as a result of liquefaction of solid caseous necrotic tissue within granulomas,
but little is known regarding the mechanistic events
responsible for this process. It has been suggested that
host proteases such as the MMPs may contribute to the
liquefaction and subsequent cavitation process and that
therapies to prevent of cavitation may reduce TB transmission and possibly the tissue destruction associated
with advanced disease (Yoder et al., 2004; Elkington and
Friedland, 2006).
There are several approaches towards inhibiting aberrant host MMP activity that may contribute to lung pathology including cavities. Tetracycline antimicrobials have
been demonstrated to inhibit MMP activity at sub-MIC
doses (Hurewitz et al., 1993; Smith et al., 1999). As the
only MMP inhibitors with wide clinical availability, tetracyclines, such as doxycycline and minocycline, have been
used in recalcitrant recurrent corneal erosions and
periodontal disease (Golub et al., 1998; Bezerra et al.,
2002). To evaluate the effectiveness of tetracyclines in the
treatment of tuberculosis, Kawada et al. successfully
treated a patient with XDR-TB with a regimen of minocycline and gatifloxacin even though the isolate showed in
vitro resistance to gatifloxacin (Kawada et al., 2008). Tetracyclines are rapidly bactericidal for Mycobacterium
leprae and are widely used in managing infection by
several other significant non-tuberculous mycobacteria
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
including Mycobacterium chelonae, M. fortuitum and M.
marinum (De Groote and Huitt, 2006).
Rationally designed MMP inhibitors have been developed by pharmaceutical companies for inflammatory diseases like arthritis and chronic obstructive pulmonary
disease (COPD); these include marimastat (BB-2516), a
broad spectrum MMP inhibitor, and trocade (Ro
32-3555), an MMP-1-selective inhibitor (Steward and
Thomas, 2000; Hemmings et al., 2001). A study of batimastat (BB-94) in M. tuberculosis-infected mice revealed
that MMP inhibition early in infection reduced hematogenous spread and dissemination to extrapulmonary
organs (Taylor et al., 2006). Surprisingly, these agents
have performed poorly in human clinical trials, in part
due to toxicity (particularly musculoskeletal toxicity in the
case of broad-spectrum inhibitors) but also lack of efficacy (i.e. for trocade, promising results in rabbit arthritis
models were not replicated in human trials). The reasons
for the discrepancies between the disappointing human
clinical results and the promising activity in animal
models are not clear. Currently, there are FDA-approved
anti-TNF-a agents, but as yet no approved anti-MMP
compounds beyond tetracyclines, suggesting the need to
further pursue this line of investigation.
Revitalizing vitamin therapy
Prior to the identification of antibiotics, tuberculosis was
commonly treated with rest, nutrition and sunlight exposure as part of the sanatorium movement. One key
element of therapy was vitamin supplementation including
attention to vitamin D (Martineau et al., 2007). While the
molecular mechanisms were not understood at the time,
vitamin D therapy has been associated with substantial
benefits in tuberculosis patients (Zasloff, 2006). Recent
studies on the molecular mechanisms of vitamin D in
tuberculosis disease reveal that activated macrophages
are effective in eliminating M. tuberculosis through a tolllike receptor (TLR)-2/1-mediated mechanism resulting in
the expression of the gene encoding LL-37, an antimicrobial peptide known as cathelicidin (Liu et al., 2006). Specifically, the binding of mycobacterial antigens to TLR2/1
activates circulating monocytes and macrophages. This
leads to the induction of genes encoding vitamin D
receptor (VDR) and Cyp27B1, an enzyme that catalyses
the conversion of inactive provitamin D3 hormone
25-hydroxyvitamin D3 [25(OH)D3] into the active form
1,25-dihydroxyvitamin D3 [1,25(OH)2D3]. The VDR and
1,25(OH)2D3 interact to stimulate expression of the LL-37
gene, leading to an increased intracellular concentration of
LL-37 and enhanced microbicidal activity of the phagocyte
(Liu et al., 2007). In addition, a recent study evaluated the
consequences of 1,25(OH)2D3 on MMP and TIMP activities (Coussens et al., 2009). The authors report that
306 C. Paige and W. R. Bishai
1,25(OH)2D3 significantly attenuates M. tuberculosisinduced increases in MMP-1, MMP-7, MMP-9 and
MMP-10 expression, while the effects on TIMPs were
negligible. Therefore, these observations suggest that in
addition to stimulating anti-microbial peptide production,
vitamin D therapy may also reduce the expression and
resulting activity of MMPs, thereby decreasing tissue
damage, cavitation, and ultimately transmission.
Various studies have been conducted to evaluate the
efficacy and therapeutic potential of vitamin D in tuberculosis patients. Initially, large bolus doses of vitamin D2
were used to treat tuberculosis as monotherapy, and later
it was used as an adjunct to anti-TB drug therapy (Martineau et al., 2007). Unfortunately, hypercalcaemia and
paradoxical upgrading reactions developed as a result. In
an effort to determine the effects of a single dose of
vitamin D2, tuberculosis patients receiving standard antimycobacterial therapy were treated with an oral dose of
2.5 mg vitamin D2 (Martineau et al., 2009). The results
demonstrated that single-dose vitamin D2 resulted in
elevated serum levels of 25-hydroxyvitamin D, the principal circulating metabolite of vitamin D, following 1 week,
but not 8 weeks post-dose. In addition, no patient developed hypercalcaemia or paradoxical upgrading reaction,
suggesting that moderate levels of vitamin D dosing
may be safe in tuberculosis treatment. Furthermore,
in a double-blinded, randomized, placebo-controlled trial,
tuberculosis patients were given vitamin D supplements
to evaluate its ability to improve the clinical outcome and
reduced mortality (Wejse et al., 2009). The authors report
that vitamin D supplementation was not associated with
serious adverse effects among tuberculosis patients; particularly there was no evidence of tuberculosis-associated
risk of hypercalcaemia, consistent with the previous study.
Unfortunately, the study did not identify a significant
benefit in clinical outcomes, although it should be noted
that relatively low vitamin D doses were used.
The need for improved animal models
Optimal evaluation of the efficacy of new therapies
against tuberculosis including new vaccines, drugs, and
host-directed therapies requires appropriate animal
models. In light of the importance of granuloma formation
and maintenance in the pathogenesis of tuberculosis, it is
clear that animal models to study host-directed, antigranuloma therapies should display characteristics that
closely resemble those observed in humans. Although the
current animal models are not exact replicas of disease
pathogenesis in humans, each model offers certain
attributes. The small animal models most commonly
pursued in preclinical studies are mice, guinea pigs and
rabbits. The murine model represents the most costeffective method to evaluate therapeutic potential of a
given compound. Chronic infection models in mice have
been developed to study the role of the host immune
response in controlling the initial stage of infection and
maintaining the bacillary load at a stable level; specifically
these include the low-dose infection model (Kelly et al.,
1996) and the Cornell model (McCune et al., 1966a,b).
Weaknesses associated with these models are the relatively high bacillary burden in the spleen and lungs of the
infected mice in the low-dose strategy, as well as potential
confounding factors from anti-mycobacterial therapy in
the Cornell model. A newer approach, the paucibacillary
tuberculosis model, may allay some of these concerns.
The model employs aerosol BCG vaccination prior to
aerosolized M. tuberculosis challenge, and mice display
unusually low lung bacterial counts and no apparent
dissemination suggesting an improved ability to contain
infection (Nuermberger et al., 2004). Despite these
efforts, major drawbacks to the murine model are the
absence of tuberculin responsiveness, the lack of
caseation necrosis in its granuloma-like lesions, and the
fact that the mouse does not progress to true latent infection. However, data from the murine model of tuberculosis
continues to provide enormous insight and remains a
useful model for the initial screening of drugs.
Most commonly used for predicting the protective efficacy of vaccine candidates, the guinea pig and rabbit are
also susceptible to M. tuberculosis infection. Unlike the
mouse, guinea pigs do possess necrotic granulomas
(McMurray, 1994). In addition, the rabbit model of tuberculosis pathogenesis has been in use since the early 20th
century as it maintains innate resistance to tuberculosis,
mimicking that seen in humans (Dannenberg, 2006). Following M. tuberculosis infection, rabbits display a 4- to
6-week period of bacillary replication followed by a bactericidal phase with containment, but incomplete sterilization of viable bacilli. In order to evaluate reactivation of
tuberculosis disease, dexamethasone-induced immunosuppression of infected rabbits has been shown to
produce worsening of disease akin to reactivation
(Kesavan et al., 2005; Manabe et al., 2008). In addition,
the rabbit model may also be manipulated towards the
development of pulmonary cavitary lesions. Using presensitization and relatively high-dose bronchoscopic
inocula, the reproducible formation of cavitary lung
lesions in rabbits has been recently described
(Nedeltchev et al., 2009). Overall, the guinea pig and
rabbit models offer the advantage of granuloma histopathology that closely resembles that of humans, but they
are limited by a lack of immunological reagents.
As previously indicated, small animal tuberculosis
models provide various strengths in their resemblance to
human disease; however, there is no comprehensive
model to render ideal extrapolations. Instead, non-human
primates represent another model that resembles human
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
Anti-granuloma strategies for tuberculosis 307
immune response. In light of the magnitude of the extensively and multi-drug-resistant tuberculosis crisis, alternative approaches including host-directed therapies warrant
renewed attention. Host-directed therapy is a promising
avenue towards the acceleration and enhancement of
traditional chemotherapy for tuberculosis.
tuberculosis more closely than any other model (Flynn
et al., 2003). The spectrum of disease and resulting
pathology is directly comparable. In addition, active tuberculosis in non-human primates is fatal, and during this
stage animals can effectively transmit disease. Unlike the
guinea pig and rabbit, there is an abundance of immunological agents available for non-human primates. A limitation of non-human primate use is their high cost and the
extensive bio-safety containment requirements for their
housing.
The support of NIAID grants and contracts 37856, 36973, 79590
and 30036 are gratefully acknowledged.
Conclusions
References
Tuberculosis remains one of the major threats to global
public health as the number of cases continues to rise.
The current anti-tuberculosis chemotherapy lasts 6–9
months and provides the basis for controlling active tuberculosis disease. The granuloma is tissue-based host
immune response in tuberculosis that is widely considered to be required for host containment of the infection.
However, recent mechanistic studies indicate that the
pathogen may possess virulence mechanisms that deliberately promote granuloma formation. This suggests that
the granuloma response may offer a survival benefit to the
pathogen, and indeed, there is evidence that bacterial
mutants lacking granuloma promoting genes also show
reduced survival in immunocompetent animals. Granuloma formation is also associated with entry of tubercle
bacilli into a persister state and the structure itself may
restrict the penetration of antimicrobial agents – both
factors that may foster the emergence of multi- and extensively drug-resistant strains of M. tuberculosis.
In view of the urgent need for new drugs against tuberculosis, therapies that target granuloma formation are
beginning to receive attention as adjunctive, host-directed
therapy to be given with antimicrobials. Given that
immune-targeted therapies have been approved to treat
other chronic diseases and further studies of these agents
have been recommended by the World Health Organization for the treatment of tuberculosis (WHO, 2007), it
stands to reason that adjunctive use of immunotherapy
with the current standard chemotherapy may offer novel
approaches towards shortening the duration of treatment
and preventing the emergence of resistance. The growing
number of these therapeutic agents entering clinical use
may facilitate additional opportunities to seek hostdirected, adjunctive treatments for tuberculosis. As
described above, several past evaluations of immunebased therapies have demonstrated safety and efficacy in
both animal models and in humans. Two distinct types of
immunomodulatory agents have been considered: (i)
those that facilitate the access of chemotherapeutic
agents to the tubercle bacilli by disrupting granuloma
integrity, and (ii) those that enhance the protective host
Agarwal, N., Lamichhane, G., Gupta, R., Nolan, S., and
Bishai, W.R. (2009) Cyclic AMP intoxication of macrophages by Mycobacterium tuberculosis adenylate cyclase.
Nature 460: 98–102.
Bezerra, M.M., Brito, G.A., Ribeiro, R.A., and Rocha, F.A.
(2002) Low-dose doxycylcine prevents inflammatory bone
resorption in rats. Braz J Med Biol Res 35: 613–616.
Bold, T.D., and Ernst, J.D. (2009) Who benefits from granulomas, mycobacteria or host? Cell 136: 17–19.
Brew, K., Dinakarpandian, D., and Nagase, H. (2000) Tissue
ihibitors of metalloproteinase: evolution, structure and
function. Biochem Biophys Acta 2: 267–283.
Chakravarty, S.D., Zhu, G., Tsai, M.C., Mohan, V.P., Marino,
S., Kirschner, D.E., et al. (2008) Tumor necrosis factor
blockade in chronic murine tuberculosis enhances granulomatous inflammation and disorganizes granulomas in the
lungs. Infect Immun 76: 916–926.
Chang, J.C., Wysocki, A., Tchou-Wong, K.M., Moskowitz, N.,
Zhang, Y., and Rom, W.N. (1996) Effect of Mycobacterium
tuberculosis and its components on macrophages and the
release of matrix metalloproteinases. Thorax 51: 306–311.
Chen, M., Divangahi, M., Gan, H., Shin, D.S., Hong, S., Lee,
D.M., et al. (2008) Lipid mediators in innate immunity
against tuberculosis: opposing roles of PGE2 and LXA4 in
the induction of macrophage death. J Exp Med 205: 2791–
2801.
Coussens, A.K., Timms, P.M., Boucher, B.J., Venton, T.R.,
Ashcroft, A.T., Skolimowska, K.H., et al. (2009) 1,
25-dihydroxyvitamin D3 inhibits matrix metalloproteinases
induced by Mycobacterium tuberculosis infection. Immunology 127: 539–548.
Dannenberg, A.M., Jr (2006) Pathogenesis of Human Pulmonary Tuberculosis. Insights from the Rabbit Model. Washington, DC: American Society for Microbiology.
Davis, J.M., and Ramakrishnan, L. (2009) The role of the
granuloma in expansion and dissemination of early tuberculosis infection. Cell 136: 37–49.
Dawson, R., Condos, R., Tse, D., Huie, M.L., Ress, R.,
Tseng, C.H., et al. (2009) Immunomodulation with recombinant interferon-1b in pulmonary tuberculosis. PLoS One
4: e6984. [PMID: 19753300].
De Groote, M.A., and Huitt, G. (2006) Infectious due to
rapidly growing mycobacteria. Clin Infect Dis 42: 1756–
1763.
Elkington, P.T., and Friedland, J.S. (2006) Matrix metalloproteinases in destructive pulmonary pathology. Thorax
61: 259–266.
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
Acknowledgements
308 C. Paige and W. R. Bishai
Elkington, P.T., Emerson, J.E., Lopez-Pascua, L.D., O’Kane,
C.M., Hornecastle, D.E., Boyle, J.J., and Friedland, J.S.
(2005) Mycobacterium tuberculosis up-regulates matrix
metalloproteinase-1 secretion from human airway epithelial cells via a p38 MAPK switch. J Immunol 175: 5333–
5340.
Flynn, J.L., and Chan, J. (2005) What’s good for the host is
good for the bug. Trends Microbiol 13: 98–102.
Flynn, J.L., Capuano, S.V., Croix, D., Pawar, S., Myers, A.,
Zinovik, A., and Klein, E. (2003) Non-human primates:
a model for tuberculosis research. Tuberculosis 83: 116–
118.
Giosue, S., Casarini, M., Ameglio, F., Zangrilli, P., Palla, M.,
Altieri, A.M., et al. (2000) Aerosolized interferon-treatment
in patients with multidrug-resistant pulmonary tuberculosis.
Eur Cytokine Netw 11: 99–104.
Golub, L.M., Lee, H.M., Ryan, M.E., Giannobile, W.V., Payne,
J., and Sorsa, T. (1998) Tetracyclines inhibit connective
tissue breakdown my multiple-antimicrobial mechanisms.
Adv Dent Res 12: 12–26.
Hemmings, F.J., Farhan, M., Rowland, J., Banken, L., and
Jain, R. (2001) Tolerability and pharamacokinetics of the
collagenase-selective inhibitor Trocade in patients with
rheumatoid arthritis. Rheumatology (Oxford) 40: 537–
543.
Hurewitz, A.N., Wu, C.L., Mancuso, P., and Zucker, S. (1993)
Tetracycline and doxycyline inhibit pleural fluid metalloproteinases. A possible mechanism for chemical pleurodesis.
Chest 103: 1113–1117.
Johnson, B.J., Bekker, L.G., Rickman, R., Brown, S., Lesser,
M., Ress, S., et al. (1997) rhuIL-2 adjunctive therapy in
multidrug resistant tuberculosis: a comparison of two treatment regimens and placebo. Tuber Lung Dis 78: 195–
203.
Johnson, J.L., Ssekasanvu, E., Okwera, A., Mayanja, H.,
Hirsch, C.S., Nakibali, J.G., et al. (2003) Randomized trial
of adjunctive interlukin-2 in adults with pulmonary tuberculosis. Am J Respir Crit Care Med 168: 185–191.
Kawada, H., Yamazato, M., Shinozawa, Y., Suzuki, K., Otani,
S., Ouchi, M., and Miyairi, M. (2008) Achievement of
sputum culture negative conversion by minocycline in a
case with extensively drug-resistant pulmonary tuberculosis. Kekkaku 83: 725–728.
Keane, J., Balcewicz-Sablinska, M.K., Remold, H.G., Chupp,
G.L., Meek, B.B., Fenton, M.J., and Kornfeld, H. (1997)
Infection by Mycobacterium tuberculosis promotes human
alveolar macrophage apoptosis. Infect Immun 65: 298–
304.
Keane, J., Gershon, S., Wise, R.P., Mirabile-Levens, E.,
Kasznica, J., Schwieterman, W.D., et al. (2001) Tuberculosis associated with Infliximab, a tumor necrosis factor
-neutralizing agent. N Engl J Med 345: 1098–1104.
Kelly, B.P., Furney, S.K., Jessen, M.T., and Orme, I.M. (1996)
Low-dose aerosol infection model for testing drugs for
efficacy against Mycobacterium tuberculosis. Antimicrob
Agents Chemother 40: 2809–2812.
Kesavan, A.K., Mendez, S.E., Hatem, C.L., Lopez-Molina, J.,
Aird, K., Pitt, M.L., Dannenberg, A.M., Jr, and Manabe, Y.C.
(2005) Effects of dexamethasone and transient malnutrition on rabbits infected with aerosolized Mycobacterium
tuberculosis CDC1551. Infect Immun 73: 7056–7060.
Koh, W.J., Kwon, O.J., Suh, G.Y., Chung, M.P., Kim, H., Lee,
N.Y., et al. (2004) Six-month therapy with aerosolized
interferon-gamma for refractory multidrug-resistant pulmonary tuberculosis. J Korean Med Sci 19: 167–171.
Lai, W.C., Zhou, M., Shankavaram, U., Peng, G., and Wahl,
L.M. (2003) Differential regulation of lipopolysaccharideinduced monocyte matrix metalloproteinase (MMP)-1 and
MMP-9 by p38 and extracellular signal-regulated kinase
1/2 mitogen-activated protein kinases. J Immunol 170:
6244–6249.
Liu, P.T., Stenger, S., Li, H., Wenzel, L., Tan, B.H., Krutzik,
S.R., et al. (2006) Toll-like receptor triggering of a vitamin
D-mediated human antimicrobial response. Science 311:
1770–1773.
Liu, P.T., Stenger, S., Tang, D.H., and Modlin, R.L. (2007)
Cutting edge: vitamin D-mediated human antimicrobial
activity against Mycobacterium tuberculosis is dependent
on the induction of cathelicidin. J Immunol 179: 2060–
2063.
McCune, R.M., Feldman, F.M., Lambert, H.P., and McDermott, W. (1966a) Microbial persistence I. The capacity of
tubercle bacilli to survive sterilization in mouse tissues.
J Exp Med 123: 445–468.
McCune, R.M., Feldman, F.M., and McDermott, W. (1966b)
Microbial persistence II. Characteristics of the sterile state
of tubercle bacilli. J Exp Med 123: 469–486.
McMurray, D.N. (1994) Guinea pig model of tuberculosis.
In Tuberculosis: Pathogenesis, Protection and Control.
Bloom, B. (ed.). Washington, DC: American Society for
Microbiology, pp. 135–147.
Manabe, Y.C., Kesavan, A.K., Lopez-Molina, J., Hatem, C.L.,
Brooks, M., Fujiwara, R., et al. (2008) The aerosol rabbit
model of TB latency, reactivation and immune reconstitution inflammatory syndrome. Tuberculosis (Edinb) 88: 187–
196.
Martineau, A.R., Honecker, F.U., Wilkinson, R.J., and Griffiths, C.J. (2007) Vitamin D in the treatment of pulmonary
tuberculosis. J Steroid Biochem Mol Biol 103: 793–798.
Martineau, A.R., Nanzer, A.M., Satkunam, K.R., Packe, G.E.,
Rainbow, S.J., Maunsell, Z.J., et al. (2009) Influence of a
single oral dose of vitamin D2 on serum 25-hydroxyvitamin
D concentrations in tuberculosis patients. Int J Tuberc Lung
Dis 13: 119–125.
Mohan, V.P., Scanga, C.A., Yu, K., Scott, H.M., Tanaka, K.E.,
Tsang, E., et al. (2001) Effects of tumor necrosis factor on
host immune response in chronic persistent tuberculosis:
possible role for limiting pathology. Infect Immun 69: 1847–
1855.
Nedeltchev, G.G., Raghunand, T.R., Jassal, M.S., Lun, S.,
Cheng, Q.J., and Bishai, W.R. (2009) Extrapulmonary dissemination of Mycobacterium bovis but not Mycobacterium
tuberculosis in a bronchoscopic rabbit model of cavitary
tuberculosis. Infect Immun 77: 598–603.
Nuermberger, E.L., Yoshimatsu, T., Tyagi, S., Bishai, W.R.,
and Grosset, J.H. (2004) Paucibacillary tuberculosis in
mice after prior aerosol immunization with Mycobacterium
bovis BCG. Infect Immun 72: 1065–1071.
Page-McCaw, A., Ewald, A.J., and Werb, Z. (2007) Matrix
metalloproteinases as modulators of inflammation and
innate immunity. Nat Rev Immunol 4: 221–233.
Quiding-Jarbrink, M., Smith, D.A., and Bancroft, G.J. (2001)
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
Anti-granuloma strategies for tuberculosis 309
Production of matrix metalloproteinases in response to
mycobacterial infection. Infect Immun 69: 5661–5670.
Rand, L., Green, J.A., Saraiva, L., Friedland, J.S., and
Elkington, P.T. (2009) Matrix metalloproteinase-1 is
regulated in tuberculosis by a p38 MAPK-dependent,
p-aminosaliciylic
acid-sensitive
signaling
cascade.
J Immunol 182: 5865–5586.
Rangel-Moreno, J., Estrada-Garcia, I., De La Luz Garcia
Hernandez, M., Agular-Leon, D., Marquez, R., and Hernandez Pando, R. (2002) The role of prostaglandin E2
in the immunopathogenesis of experimental pulmonary
tuberculosis. Immunology 106: 257–266.
Rook, G.A.W., Lowrie, D.B., and Hernandez-Pando, R.
(2007) Immunotherapeutics for tuberculosis in experimental animals: is there a common pathway activated by effective protocols? J Infect Dis 196: 191–198.
Rubin, E.J. (2009) The granuloma in tuberculosis – friend or
foe? N Engl J Med 360: 2471–2473.
Russell, D.G. (2007) Who puts the tubercle in tuberculosis?
Nat Rev Microbiol 5: 39–47.
Schiff, M.H., Burmester, G.R., Kent, J.D., Pangan, A.L.,
Kupper, H., Fitzpatrick, S.B., and Donovan, C. (2006)
Safety analyses of adalimumab (HUMIRA) in global
clinical trials and US postmarketing surveillance of
patients with rheumatoid arthritis. Ann Rheum Dis 65: 889–
894.
Smith, G.N., Jr, Mickler, E.A., Hasty, K.A., and Brandt, K.D.
(1999) Specificity of inhibition of matrix metalloproteinase
activity by doxycycline: relationship to structure of the
enzyme. Arthritis Rheum 42: 1140–1146.
Steward, W.P., and Thomas, A.L. (2000) Marimastat: the
clinical development of matrix metalloproteinase inhibitor.
Expert Opin Invest Drugs 9: 2913–2922.
Suarez-Mendez, R., Garcia-Garcia, I., Fernandez-Olivera,
N., Valdes-Quintana, M., Milanes-Virelles, M., Carbonell,
D., et al. (2004) Adjuvant interferon gamma in patients with
drug-resistant pulmonary tuberculosis: a pilot study. BMC
Infect Dis 4: 44.
Taylor, J.L., Hattle, J.M., Dreitz, S.A., Troudt, J.M., Izzo,
© 2010 Blackwell Publishing Ltd, Cellular Microbiology, 12, 301–309
L.S., Basaraba, R.J., et al. (2006) Role for matrix metalloproteinase 9 in granuloma formation during pulmonary
Mycobacterium tuberculosis infection. Infect Immun 74:
6135–6144.
Tsenova, L., Sokol, K., Freedman, V.H., and Kaplan, G.
(1998) A combination of thalidomide plus antibiotics protects rabbits from mycobacterial meningitis-associated
death. J Infect Dis 177: 1563–1572.
Tsenova, L., Mangaliso, B., Muller, G., Chen, Y., Freedman,
V.H., Stirling, D., and Kaplan, G. (2002) Use of IMiD3,
a thalidomide analog, as an adjunct to therapy for
experimental tuberculosis meningits. Antimicrob Agents
Chemother 46: 1887–1895.
Wallis, R.S. (2005) Reconsidering adjuvant immunotherapy
for tuberculosis. Clin Infect Dis 41: 201–208.
Wallis, R.S. (2009) Infectious complications of tumor necrosis
factor blockade. Curr Opin Infect Dis 22: 403–409.
Wallis, R.S., Kyambadde, P., Johnson, J.L., Horter, L., Kittle,
R., Pohle, M., et al. (2004) A study of the safety, immunology, virology, and microbiology of adjunctive etanercept
in HIV-1-associated tuberculosis. AIDS 18: 257–264.
Wallis, R.S., van Vuuren, C., and Potgieter, S (2009) Adalimumab treatment of life-threatening tuberculosis. Clin
Infect Dis 48: 1429–1432.
Wejse, C., Gomes, V.F., Rabna, P., Gustafson, P., Aaby, P.,
Lisse, I.M., et al. (2009) Vitamin D as supplementary
treatment for tuberculosis. A double-blind, randomized,
placebo-controlled trial. Am J Respir Crit Care Med 179:
843–850.
World Health Organization (2007) Report of the expert
consultation on immunotherapeutic interventions for
tuberculosis. Geneva, Switzerland. Available from:
http://apps.who.int/tdr/publications/tdr-research-publications/
immunotherapeutic-interventions-tb/pdf/interventions-tb.pdf
Yoder, M.A., Lamichhane, G., and Bishai, W.R. (2004) Cavitary tuberculosis: the holey grail of disease transmission.
Curr Sci 86: 101–108.
Zasloff, M. (2006) Fighting infectious with vitamin D. Nat Med
12: 388–390.