| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Pathology and Immunology,* University of Geneva, Geneva, Switzerland; the Department of Pathology, University of Bern, Bern, Switzerland; and the Laboratoire du Bacillus Calmette-Guerin and Unité de Physiopathologie de l’Infection, Institut Pasteur, Paris, France
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
–/– mice expressing a noncleavable transgenic TmTNF (TmTNF tg) to those of TNF/LT-–/– and wild-type mice. Susceptibility of TNF/LT-–/– mice to BCG infection was associated with impaired induction of systemic RANTES but not of monocyte chemoattractant protein 1 (MCP-1), the development of excessive local and systemic Th1-type immune responses, and a substantially reduced inducible nitric oxide synthase (iNOS) activity. Resistance of TmTNF tg mice to BCG infection was associated with efficient activation of iNOS in granulomas and with the regulated release of local and systemic chemokines and Th1-type cytokines. However, M. tuberculosis infection of TmTNF tg mice resulted in longer survival and enhanced resistance compared to TNF/LT-–/– mice but higher sensitivity than wild-type mice. TmTNF tg mice exhibited reduced pulmonary iNOS expression and showed an exacerbated cellular infiltration in the lungs despite a modest bacillary content. Our data thus indicate a role for TmTNF in host defense against mycobacteria by contributing to induction and regulation of Th1-type cytokine and chemokine expression leading to development of bactericidal granulomas expressing iNOS, which critically determines susceptibility versus resistance of the host to mycobacterial infections.
Murine models of experimental tuberculosis have allowed the understanding of part of the mechanisms implicated in mycobacterial infections, as the induction of Th1 cell-mediated immune responses, characterized by the presence of interferon (IFN)-
.3,4
Mycobacteria are isolated in protective granulomas,5
consisting of differentiated macrophages and T cells, and killed by bactericidal mechanisms such as inducible nitric oxide synthase (iNOS), the enzyme producing nitric oxide (NO). The iNOS plays a crucial role in host defense against mycobacterial infections,6,7
and is synergistically activated by tumor necrosis factor (TNF) and IFN-.8
TNF is first synthesized as a 26-kd transmembrane (TmTNF) precursor and cleaved by membrane-bound metalloprotease-disintegrin TACE (TNF--converting enzyme) generating a soluble 17-kd TNF molecule.9,10
Most TNF activities were attributed to the secreted or soluble (sTNF) form, particularly those related to resistance to intracellular bacterial infections such as mycobacteria. Animal studies, including our previous works, have provided compelling evidence implicating sTNF as a key factor in host defense against mycobacterial infections. Impaired granuloma formation and bactericidal mechanisms, and alteration of mycobacterium-induced Th1-type immune responses were observed in animals unable to use TNF.11-21
The differential contribution of TmTNF and sTNF in mycobacterial host defense has not been studied in detail. We previously showed that TmTNF, in the absence of sTNF and lymphotoxin (LT)-, induces an efficient cell-mediated immune response to M. bovis BCG infection.20
This study was performed using TmTNF transgenic mice in which expression of the transgene is controlled by the TNF promoter and the 3' AU-rich elements of the TNF genomic locus are maintained to assure adequate gene regulation in mice that lack both the TNF and the LT- genes.22
Studies on mice expressing a noncleavable TmTNF mutant that replaces the endogenous TNF allele demonstrated an important role for TmTNF in the development of lymphoid structures and in the production of chemokines.23
LT-, as well as sTNF, is implicated in host defense against mycobacterial infections.19,24
In this study, we examine the role of TmTNF in the development of the immune response to M. bovis BCG and M. tuberculosis infections. We show that TmTNF tg mice were able to efficiently induce a cell-mediated immune response with development of granulomas containing iNOS-expressing cells as also observed in wild-type mice after M. bovis BCG infection. In contrast, TNF/LT-–/– mice showed high sensitivity to BCG infection associated with reduced iNOS activity in infected organs. Although, TmTNF tg mice showed higher survival and enhanced resistance than TNF/LT-–/– mice to M. tuberculosis infection, they died from infection and failed to activate iNOS-producing cells within granulomas. The differential production of iNOS within granulomas may thus represent a key factor in conferring resistance, or susceptibility, to M. bovis BCG and M. tuberculosis infections.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The generation and characterization of TNF/LT-–/– mice (C57BL/6 x 129/SVEV) transgenic for a noncleavable TmTNF gene have already been described.22,20
As reported, three mutations (at positions –10, +1, and +10 of the TNF molecule) were introduced into the coding sequence of a genomic clone of mouse TNF. The resulting transgene was under the control of the mouse TNF promotor, and the AU-rich elements at the 3' end were maintained. TmTNF construct was microinjected into the male pronucleus of TNF/LT-–/– mice to generate transgenic mice expressing TmTNF in the absence of TNF and LT-.22
Mice were maintained under conventional conditions in the animal facility of the Medical Faculty, University of Geneva (Geneva, Switzerland), or under P3 conditions at the Pasteur Institute (Paris, France).
Experimental Infections and Determination of BCG and M. tuberculosis Colony-Forming Units (CFUs)
Mice were infected intravenously with 107 living BCG Pasteur strain 1173 P2 or with 105 CFU of virulent M. tuberculosis strain H37Rv. Mortality and weight loss were monitored during the infection. The numbers of viable bacteria were determined by CFU as previously described.20
Histological Analyses
Histological analyses of organs were performed at 4 weeks after BCG and M. tuberculosis infections and at necropsy. Liver, lungs, and spleen were fixed in 4% buffered formaldehyde and embedded in paraffin for subsequent hematoxylin and eosin (H&E) and Ziehl-Neelsen staining.
In Situ Hybridization
An 1108-bp cDNA fragment of the murine TNF gene (positions 1 to 1108; obtained from Genentech Inc., San Francisco, CA) was subcloned into pGEM-2. After linearization of the plasmid, sense and anti-sense RNA probes were prepared using the appropriate RNA polymerase as described previously.22 In situ hybridizations of paraformaldehyde-fixed cryostat sections of lungs were performed with 2 x 105 cpm of 35S-labeled RNA probe per µl of hybridization solution for 18 hours at 48°C.20
Spleen Cell Recovery
Spleen cells from 4-week BCG-infected mice were suspended in Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum, treated for 5 minutes with a 0.155 mol/L ammonium chloride/0.010 mol/L potassium bicarbonate solution to lyse the erythrocytes, washed, and resuspended in Dulbecco’s modified Eagle’s medium plus 10% fetal calf serum as described previously.20
Bronchoalveolar Lavage (BAL)
Mice were anesthetized with Nembutal (0.5 mg/mouse) and the thoracic cavity was opened. The trachea was cannulated and the lungs of each mouse were washed five times with 1 ml of phosphate-buffered saline (PBS) each by gentle instillation. The BAL was spun and the supernatant was removed and stored at –20°C for cytokine analyses.
Flow Cytometric Analyses
Flow cytometry was performed using one- or two-color staining of spleen cells 4 weeks after BCG infection, and analyzed with a FACSCalibur (BD Biosciences, Mountain View, CA). The antibodies and reagents used were phycoerythrin-conjugated rat anti-CD8 (clone H-35), biotin-labeled rat anti-CD4 (clone H-129), biotin-labeled rat anti-Mac-1 (clone M1/70), biotin-labeled hamster anti-CD11c (clone HL3), and avidin-phycoerythrin (all from Becton Dickinson-Pharmingen, Heidelberg, Germany). Stainings were performed in the presence of a saturating concentration of 2.4 G2 anti-Fc RII/III monoclonal antibody.
Evaluation of Chemokines and Cytokines in Serum, Bronchoalveolar Fluid, and Lung Protein Extracts
Blood samples were obtained from retro-orbital sinuses at different time points after infection and bronchoalveolar fluids were obtained by BAL at 4 weeks of BCG infection as described above. Lungs from BCG-infected mice were homogenized in 0.04% Tween 80/saline buffer (125 mg of tissue/ml of buffer). Systemic RANTES regulated on activation normal T cell expressed and secreted monocyte chemoattractant protein (MCP)-1, IFN-, interleukin (IL)-12p70, and IL-12p40 amounts were evaluated by enzyme-linked immunosorbent assay with a sensitivity of 2 to 1000 pg/ml (R&D System, Abingdon, UK, and Immunotec, Belgium).
Inducible NOS Activity in Lung and Spleen Extracts
Evaluation of iNOS activity was done on crude frozen lung and spleen extracts of infected and uninfected mice. Spleens were homogenized in 25 mmol/L Tris-HCl (pH 7.4), 1 mmol/L EDTA, and 1 mmol/L EGTA (125 mg of tissue/ml of buffer). Crude supernatant was obtained by centrifugation of the homogenate at 10,000 x g for 5 minutes. iNOS activity was measured by the ability of supernatant to convert radioactive L-(14C)-arginine (Amersham Life Science) to L-(14C)-citrulline as previously described.20
TNF and iNOS Immunohistochemistry
Lung sections from paraffin-embedded tissue were rehydrated in PBS and processed for 2 x 5 minutes in citrate buffer (0.01 mol/L, pH 6.0) in a microwave for iNOS staining and in a pressure cooker for TNF staining. Nonspecific binding sites were blocked in PBS and 1% bovine serum albumin for 30 minutes. For iNOS staining, the sections were incubated with rabbit anti-mouse iNOS (Calbiochem, San Diego, CA) for 60 minutes followed by incubation with biotinylated donkey anti-rabbit antibody (Amersham, Freiburg Germany) revealed by streptavidin-alkaline-phosphatase (Boehringer Mannheim, Rotkreuz, Switzerland). For TNF staining, sections were incubated with a rabbit anti-mouse TNF (Bioreba, Reinach BL, CH) followed by an incubation with a goat anti-rabbit antibody (DakoCytomation, Carpinteria, CA) and revealed with diaminobenzidine and counterstained with H&E. For these two stainings, the preparation incubated without the primary antibody served as the control.
Morphometric Analysis
The sections from immunohistochemistry staining were analyzed by the Metamorph 5.0 software (Universal Imaging, West Chester, PA). An automatic threshold corresponding to the histogram average was applied to generate binary images, and a median filter was used. Results were given in counts (integrating the intensity and the surface of the immunostaining) of positive pixels for iNOS. Objects smaller than a perimeter of 20 pixels were not taken into account.
Statistical Analyses
The unpaired Student’s t-test was used for all analyses with the exception of survival data which were analyzed by Kaplan-Meier using the log rank test. P values <0.05 were considered as statistically significant.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
We have previously shown that mice deficient in TNF/LT- were highly sensitive to BCG infection and die 3 to 8 weeks after infection and that TNF/LT-–/– mice expressing a transgenic TmTNF form (TmTNF tg) survived for more than 24 weeks.19,20
In the present study we have examined lung lesions of wild-type, TmTNF tg, and TNF/LT-–/– mice 4 weeks after infection with BCG (107 living bacilli). Wild-type and TmTNF tg mice exhibited a similar pattern of pulmonary pathology showing differentiated granulomas containing large macrophages and multinucleated giant cells (Figure 1, A and B)
. In contrast, TNF/LT-–/– mice displayed very large lesions and lung necrosis (Figure 1C)
. As previously described, TNF/LT-–/– mice developed very few granulomas that did not differentiate and contained high numbers of acid-fast bacilli.19,20
Lung granulomas from wild-type and TmTNF tg mice contained cells expressing TNF mRNA as evaluated by in situ hybridization. Lung tissues showed strong signal for TNF mRNA in cells forming granulomas in wild-type and TmTNF tg mice but not in TNF/LT-–/– mice (Figure 1; D to F)
. Moreover, the TNF protein was detectable in the lungs of both wild-type and TmTNF tg mice, preferentially in granulomas and in multinucleated giant cells (Figure 1, G and H)
.
|
–/– Mice and Effect of TmTNF
Mycobacterial infections induce chemokine activation involved in migration of cells to the site of the infection favoring granuloma formation.25
It has been shown that TNF is required for regulation of chemokine expression essential for cell recruitment to the granuloma.26
We have previously examined chemokine activation at the level of RNAs in TNF/LT-–/– mice and observed that MCP-1, MCP-5, MIP-1, and MIP-1ß are produced in the lung and liver during BCG infection.19
We have now evaluated protein levels of MCP-1 and RANTES in the three groups of BCG-infected mice. Indeed, MCP-1 is produced by and acts on monocytes and macrophages, and deficiency in MCP-1 receptor (CCR2) was shown to inhibit granuloma formation after M. tuberculosis infection.27,28
BCG infection induced MCP-1 secretion in serum of wild-type, TmTNF tg, and TNF/LT-–/– mice indicating that this was independent of TNF release. Wild-type and TNF/LT-–/– mice had a similar pattern at 2 weeks of infection, however, the presence of TmTNF enhanced MCP-1 secretion (Figure 2A)
. At 4 weeks of infection, MCP-1 was augmented in both the serum and the BAL fluid of TNF/LT-–/– mice and this increase was attenuated in TmTNF tg mice, and similar amounts of MCP-1 in BAL fluid of TmTNF tg and wild-type mice were observed (Figure 2B)
. RANTES, which is produced by a wide variety of cells, including lymphocytes and macrophages,29,30
and acts on a large group of inflammatory cells such as lymphocytes and macrophages,31
was measured. RANTES was not released in the serum of TNF/LT-–/– mice and the amount was lower in the serum of TmTNF tg mice as compared to wild-type mice at 2 weeks after infection, this indicating that RANTES was influenced by TNF. At 4 weeks of infection, systemic RANTES was also decreased in TmTNF tg and TNF/LT-–/– mice (Figure 2C)
. In addition, RANTES protein was increased in the lungs of sick TNF/LT-–/– mice and this was attenuated in the presence of TmTNF (Figure 2D)
.
|
–/– and Effect of TmTNF
BCG infection induces a Th1-type immune response characterized by the production of IFN- and IL-12 required for host defense. We have previously shown that TNF/LT-–/– mice are unable to activate systemic IFN- and IL-12p40 at early infection which is restored by the presence of TmTNF20
. We have now compared the levels of systemic Th1 cytokines, IFN-, IL-12p40, and IL-12p70 (the active form of IL-12), to those in BAL. IFN- and IL-12p40 levels were highly increased in both the serum and the BAL fluids of TNF/LT-–/– mice, this exacerbated secretion of cytokines was attenuated in the presence of the TmTNF form (Figure 3, A to D)
. TmTNF tg mice showed slight differences and lower levels of IFN- when compared to wild-type mice. The levels of IL-12p70 in serum of TNF/LT-–/– mice remained lower than those of wild-type mice while in BAL fluid they were higher (Figure 3, E and F)
. In addition, we have examined the phenotype and frequencies of spleen cell subpopulations 4 weeks after infection (Table 1)
. The phenotypic analysis of spleen cells revealed that TmTNF tg mice had lower frequencies of Mac-1-positive cells and a slight increase in CD4-positive cells compared to wild-type mice. These changes were markedly enhanced in the TNF/LT-–/– mouse that showed an important reduction in Mac-1 cells and an increase in both CD4 and CD8 T cells (Table 1)
.
|
|
–/– Mice
Activation of iNOS in macrophages represents one important bactericidal mechanism to eliminate intracellular bacteria. iNOS is required for protection against BCG infection.6
The absence of iNOS renders mice susceptible to mycobacterial infections.6,7
We have evaluated the iNOS activity from crude extracts of lung and spleen at 4 weeks after BCG inoculation by monitoring the conversion of radioactive L-arginine to L-citruline and NO. We observed that iNOS activity was similar in the lung and spleen of TmTNF tg and wild-type mice suggesting that TmTNF is sufficient to activate this enzyme on an avirulent BCG infection (Figure 4)
. In contrast, TNF/LT-–/– mice showed a statistically significant reduced activity of iNOS mainly in infected lungs but also in spleen (Figure 4)
. These data show that iNOS can be activated by both sTNF and TmTNF and that on BCG infection TmTNF is sufficient for bacillus control and elimination.
|
–/– but Are More Sensitive Than Wild-Type Mice to M. tuberculosis Infection
Susceptibilities of TNF/LT-–/– and TmTNF tg mice to M. tuberculosis (H37Rv, 105 CFU) infection was assessed by bacterial counts in infected organs at 4 weeks of infection. TmTNF tg mice depicted a significant increase of CFU in the lungs but only a tendency in spleen and liver when compared to wild-type mice (Figure 5A)
. Bacterial loads of TNF/LT-–/– mouse infected organs showed a dramatic increase of M. tuberculosis living bacilli compared to organs from TmTNF tg mice. Histological analyses of the lung at 4 weeks of M. tuberculosis infection revealed numerous granulomas containing large well-differentiated macrophages fused onto multinucleated giant cells in wild-type mice. Furthermore, wild-type mice were able to control recruitment of inflammatory cells to the lungs (Figure 5B)
and bacillus growth (Figure 5C)
. In contrast, TmTNF tg mice exhibited important perivascular and peribronchiolar inflammatory cell accumulation, characterized by the presence of lymphocytes and macrophages occupying the alveolar space leading to lung lesions (Figure 5D)
. In TmTNF tg mice, the presence of bacilli was limited and contained within granulomas (Figure 5E)
. Lungs of TNF/LT-–/– mice had dramatic cellular recruitment, numerous caseous necrotic lesions, and edema accumulation in the alveolar space, which was not observed on BCG infection (Figure 5, F and G)
. TNF/LT-–/– mice died between 3 and 7 weeks (median survival time, 31 days) while TmTNF tg mice died between 6 and 13 weeks (median survival time, 56 days) after M. tuberculosis infection (Figure 6A)
. The difference of survival time of TNF/LT-–/– and TmTNF tg mice was statistically significant (P < 0.0001). Comparison of the lung lesions of TmTNF tg and TNF/LT-–/– mice at necropsy depicted a different pathology. Lung lesions of TmTNF tg mice showed marked inflammatory cell infiltration in the presence of low number of bacilli and absence of caseous necrotic lesions that were the predominant lesions in TNF/LT-–/– mice associated with very high number of acid fast bacilli (Figure 6; B to E)
.
|
|
–/– Mice Reveal Altered iNOS Expression after M. tuberculosis Infection
Mycobacterial infection induces cytokine expression that triggers bactericidal mechanisms, such as the iNOS activity, which is expressed by macrophages forming bactericidal granulomas. Therefore, we have analyzed the presence of iNOS in spleen and lung at 4 weeks of infection. On M. tuberculosis infection, the pattern of iNOS-positive cells in the spleen of TmTNF tg mice showed a reduced number of iNOS-expressing cells compared to wild-type mice and very few cells were stained for iNOS in the spleen of TNF/LT-–/– mice (Figure 7; A to C)
. In contrast, the pattern of BCG-infected mice was comparable in wild-type and TmTNF tg mice showing cell clusters forming granulomas (Figure 7, D and E)
, but TNF/LT-–/– mice exhibited a marked reduction in iNOS-positive cells (Figure 7F)
. The iNOS-expressing cells were mainly the multinucleated giant cells, as we could observe in the lung of both BCG-infected wild-type and TmTNF tg mice (Figure 7, G and H)
and also after M. tuberculosis infection. Quantification of iNOS expression in lung and spleen sections from M. tuberculosi-s and BCG-infected mice was performed with an algorithm integrating surface and intensity of immunostaining. We found that TmTNF tg mice infected with M. tuberculosis displayed a marked decrease in iNOS expression in the lungs (Figure 8A)
. They also showed a less marked reduction of iNOS-expressing cells in the spleen (Figure 8B)
. Infection with BCG resulted in comparable iNOS counts in the spleen of both TmTNF tg and wild-type mice but reduction in TNF/LT-–/– mice. Correlation between iNOS staining in tissues and bacillary loads for each individual mouse has been analyzed. We observed that the degree of iNOS staining in the lungs and spleen was directly related to the elimination or the control of bacillary growth (correlation coefficient obtained by scatterplot values from iNOS staining and corresponding CFU were r = –0.7 for M. tuberculosis and r =–0.6 for BCG infection). This suggests a direct association between the capacity to express iNOS in granulomas and the availability to control mycobacterial growth, indicating that the macrophage cell population producing iNOS is critically involved in sensitivity and resistance to mycobacteria.
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
, in the presence or absence of a TmTNF form, and provides evidence that regulation of chemokines, Th1-type cytokines, and mainly iNOS is pivotal in mediating sensitivity and resistance to mycobacterial infections. We show that expression of TmTNF can rescue mice from detrimental pulmonary pathology, characterized by enhanced cellular infiltrates, mycobacterial proliferation, and death of the host observed in BCG-infected TNF/LT-–/– mice. The efficient immune response of TmTNF tg mice against BCG infection was the result of activation and control of chemokines and Th1-type cytokines leading to proper cell recruitment and expansion thus inducing bactericidal mechanisms, eliminating most of bacilli, and preventing their proliferation. Activation of iNOS in cells migrating or differentiating in the lung and spleen of BCG-infected mice correlated with low bacillary loads of TmTNF tg mice whereas the absence of TNF/LT- resulted in profound iNOS deficiency and inability to clear bacteria leading to a dramatic increase in bacillary burden. We have noticed that in wild-type and TmTNF granulomas, the same cells activating iNOS, including multinucleated cells, also produced TNF and represented the reservoir of bacilli. Our data indicate that susceptibility and resistance to BCG infection depends on two main factors: the availability to induce cells that express TmTNF and iNOS, able to destroy intracellular bacilli before they can multiply and, the capacity to orchestrate the production and the regulation of chemokines and Th1-type cytokines in a TNF-dependent manner.
Homing of leukocytes requires the presence of tissue-associated chemokine gradients and the expression of CCRs on migratory cells.32
Mycobacterial infection induces chemokine activation and TNF regulates the early chemokine induction necessary for cell recruitment to the infection site.25,26
Local and systemic amounts of chemokines on BCG infection were measured to address the question of whether absence or presence of TmTNF in TNF/LT-–/– mice could modify chemokine regulation. We have observed that TmTNF tg mice had increased serum levels of MCP-1 when compared to both wild-type and TNF/LT-–/– mice. This suggests that TmTNF can influence MCP-1 expression. Absence of MCP-1 or MCP-1 receptor resulted in a transient increase of bacterial load and a reduction in macrophage recruitment without affecting the survival of mice infected with M. tuberculosis.27,33
Interestingly, we observed that increased pulmonary MCP-1 levels correlated with enhanced cell migration and tissue injury in TNF/LT-–/– mice. TNF/LT-–/– mice also had a substantial increase of RANTES in the lung, which was attenuated in TmTNF tg mice. We observed that induction of RANTES was dependent on TNF at the early stages of the infection. TmTNF form was previously shown to play a dominant role in chemokine steady-state gene expression, particularly in MCP-1 and RANTES.23
A recent study has reported that neutralization of TNF with antibodies resulted in reduced production of RANTES, but did not affect MCP-1 expression during M. tuberculosis infection.34
Moreover, depletion of RANTES was associated with decreased size of M. bovis-induced granuloma.35
Our results suggest that TmTNF can partially regulate RANTES through infection.
TNF has been considered an essential cytokine contributing to the induction of Th1-type immune response required for cell-mediated resistance against mycobacteria. More recent studies are reconsidering this notion in view of new observations in mice unable to use TNF. One study has previously identified TNF as a negative regulatory cytokine for BCG-induced Th1-type immune responses because it was found to limit an otherwise detrimental Th1-type immune response. Animals unable to produce TNF showed overproduction of Th1-type cytokines in the lungs and peripheral blood, and hyperreactive CD4 and CD8 T-cell responses on BCG infection. Depletion of CD4 and CD8 T cells prevented tissue destruction in TNF–/–-infected mice.21
Previous reports have provided evidence for an anti-inflammatory role of TNF involving its capacity to regulate Th1-type cytokines.36,37
TNF/LT-–/– mice did not activate early Th1-type cytokines and then they showed a substantial increase in local and systemic Th1-type cytokines associated with disease severity. This impaired regulation of Th1-type cytokines is likely because of the absence of TNF in TNF/LT-–/– mice. Indeed, BCG-infected LT-–/– mice showed regulated IFN- expression (data not shown), and at 4 weeks of infection, lung lesions were similar to those of wild-type mice.19
Therefore, LT- does not appear to influence this enhanced Th1-type immune response.
Although TmTNF tg mice were able to mount an efficient immune response to BCG infection, they were unable to survive to M. tuberculosis infection but still showed longer survival than TNF/LT-–/– mice. Histopathological examinations at 4 weeks of M. tuberculosis infection showed a relatively modest increase in bacillary burden compared to wild-type mice. Enhanced cellular infiltration and the resulting tissue lesions, however, were probably more relevant factors leading to the fatal issue of the infection than the bacillary overgrowth. Moribund TmTNF tg mice did not develop caseous necrotic lesions observed in the lung of TNF/LT-–/– mice, but massive lymphocytic infiltration was predominant. This indicates that during M. tuberculosis infection, TmTNF was not sufficient to attenuate the hyperactivation of lymphocytes leading to exacerbated cell recruitment to the lung, suggesting that the rapid presence of sTNF at the site of infection is necessary to regulate the cell-mediated immune response against virulent mycobacteria. Our results are in agreement with previous reports showing that local inhibition of TNF in the lungs during M. tuberculosis infection led to the death of mice because of increased production of proinflammatory factors that induced the tissue destruction in the absence of bacillary overgrowth.38
Resistance to BCG and M. tuberculosis infections correlated with the capacity to develop cells producing iNOS in the lung and spleen. Indeed, when animals were able to eliminate intracellular bacilli, iNOS was expressed by multinucleated cells within the granulomas and these cells were also immunoreactive for TNF demonstrating that TNF and iNOS were produced by the same cells and that the absence of TNF prevents proper iNOS expression. Moreover, a strong correlation between iNOS expression in activated macrophages and growth containment of both BCG or M. tuberculosis in infected organs was found. Activation of iNOS generating NO has been shown to be an important bactericidal mechanism for host defense against mycobacterial infections.6,7 Activation of iNOS is instrumental in the regulation of the cell-mediated immune response by the multiple activities exerted by NO.39 NO has been demonstrated to down-modulate T cell responses and the recruitment of inflammatory cells to the sites of infection. We have previously reported that iNOS plays an essential role in down-regulating the immune responses to BCG infection. The absence of iNOS resulted in a dramatic activation of the immune system, up-regulation of TNF, and massive recruitment of lymphoid cells to granulomas, which became caseous necrotic lesions.6 Bekker and colleagues40 have shown that TNF may control intracellular bacilli in the absence of iNOS, thus, TNF may act by both iNOS-dependent (killing of bacilli) and iNOS-independent pathways (growth of bacilli). We have observed that during M. tuberculosis infection TmTNF tg mice were unable to efficiently activate iNOS-expressing cells in the lung leading to exacerbation of local cell immune response and death from pulmonary lesions because of uncontrolled cell recruitment despite modest bacillary growth. Our findings indicate that both iNOS and TNF may be negative regulators of a detrimental and uncontrolled mycobacteria-induced cell-mediated immune response.
The study of TmTNF activities in mycobacterial infections is important because the treatment of human disease targets TNF with antibodies or with TNF receptors neutralizing sTNF and TmTNF. Recently, several reports demonstrated that the treatment of rheumatoid arthritis and Crohn’s disease with TNF inhibitors significantly increases the tuberculosis risk and reactivation of latent tuberculosis.41-45 Similarly, in murine models, the immune system is unable to control latent tuberculosis without TNF.46,47 Our results providing evidence that TmTNF is pivotal in resistance against BCG infection and that it partially protects from M. tuberculosis infection suggest that the development of selective inhibitors neutralizing soluble but sparing TmTNF would be a promising therapeutic strategy.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the Swiss National Foundation for Scientific Research (grant 3200B0-105914 to I.G.), the Roche Foundation, the E. and L. Schmidheiny Foundation, and the Jules Thorn Foundation.
Accepted for publication January 10, 2005.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. C. Cowley, J. D. Sedgwick, and K. L. Elkins Differential Requirements by CD4+ and CD8+ T Cells for Soluble and Membrane TNF in Control of Francisella tularensis Live Vaccine Strain Intramacrophage Growth J. Immunol., December 1, 2007; 179(11): 7709 - 7719. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Zalevsky, T. Secher, S. A. Ezhevsky, L. Janot, P. M. Steed, C. O'Brien, A. Eivazi, J. Kung, D.-H. T. Nguyen, S. K. Doberstein, et al. Dominant-Negative Inhibitors of Soluble TNF Attenuate Experimental Arthritis without Suppressing Innate Immunity to Infection J. Immunol., August 1, 2007; 179(3): 1872 - 1883. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Uchiyama, I. Kawamura, T. Fujimura, M. Kawanishi, K. Tsuchiya, T. Tominaga, T. Kaku, Y. Fukasawa, S. Sakai, T. Nomura, et al. Involvement of Caspase-9 in the Inhibition of Necrosis of RAW 264 Cells Infected with Mycobacterium tuberculosis Infect. Immun., June 1, 2007; 75(6): 2894 - 2902. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Spohn, R. Guler, P. Johansen, I. Keller, M. Jacobs, M. Beck, F. Rohner, M. Bauer, K. Dietmeier, T. M. Kundig, et al. A Virus-Like Particle-Based Vaccine Selectively Targeting Soluble TNF-{alpha} Protects from Arthritis without Inducing Reactivation of Latent Tuberculosis J. Immunol., June 1, 2007; 178(11): 7450 - 7457. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. O'Kane, J. J. Boyle, D. E. Horncastle, P. T. Elkington, and J. S. Friedland Monocyte-Dependent Fibroblast CXCL8 Secretion Occurs in Tuberculosis and Limits Survival of Mycobacteria within Macrophages J. Immunol., March 15, 2007; 178(6): 3767 - 3776. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. K. McCoy, T. N. Martinez, K. A. Ruhn, D. E. Szymkowski, C. G. Smith, B. R. Botterman, K. E. Tansey, and M. G. Tansey Blocking soluble tumor necrosis factor signaling with dominant-negative tumor necrosis factor inhibitor attenuates loss of dopaminergic neurons in models of Parkinson's disease. J. Neurosci., September 13, 2006; 26(37): 9365 - 9375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Musicki, H. Briscoe, S. Tran, W. J. Britton, and B. M. Saunders Differential Requirements for Soluble and Transmembrane Tumor Necrosis Factor in the Immunological Control of Primary and Secondary Listeria monocytogenes Infection. Infect. Immun., June 1, 2006; 74(6): 3180 - 3189. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Torres, L. Janot, V. F.J. Quesniaux, S. I. Grivennikov, I. Maillet, J. D. Sedgwick, B. Ryffel, and F. Erard Membrane Tumor Necrosis Factor Confers Partial Protection to Listeria Infection Am. J. Pathol., December 1, 2005; 167(6): 1677 - 1687. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||
