Tuberculosis is a major cause of death in much of the world. Current estimates are that one-third of the world's population is infected with Mycobacterium tuberculosis. Most infected persons control the infection but in many cases may not eliminate the organism. Reactivation of this clinically latent infection is responsible for a large proportion of active tuberculosis cases. A major risk factor for reactivation of latent tuberculosis is HIV infection, suggesting a role for the CD4+ T cell subset in maintaining the latent persistent infection. In this study, we tested the requirement for CD4+ T cells in preventing reactivation in a murine model of latent tuberculosis. Antibody-mediated depletion of CD4+ T cells resulted in rapid reactivation of a persistent infection, with dramatically increased bacterial numbers in the organs, increased pathology in the lungs, and decreased survival. Although CD4+ T cells are believed to be a major source of interferon (IFN)-γ, expression of the gene for IFN-γ in the lungs of CD4+ T cell–depleted mice was similar to that in control mice. In addition, inducible nitric oxide synthase production and activity was unimpaired after CD4+ T cell depletion, indicating that macrophage activation was present even during CD4+ T cell deficiency. These data indicate that CD4+ T cells are necessary to prevent reactivation but may have roles in addition to IFN-γ production and macrophage activation in controlling a persistent tuberculous infection.

Introduction

Mycobacterium tuberculosis, the etiologic agent of tuberculosis, was responsible for an estimated 8 million cases of tuberculosis and 1.5 million deaths worldwide in 1998 1. In humans, the majority of M. tuberculosis infections are initially controlled, and a clinically latent infection is established, characterized by the persistence of low numbers of possibly dormant bacilli. A small but significant percentage (∼10%) of latent infections reactivate years to decades later to give rise to reactivation tuberculosis 2. Aging 3 and iatrogenic immunosuppression 4 have been associated with reactivation of human latent infections. In recent years, HIV infection has emerged as the biggest risk factor for reactivation of latent tuberculosis 5. The severe CD4+ T cell deficiency in AIDS implicates CD4+ T cells in protection against reactivation tuberculosis.

Cell-mediated immunity, contributed by both CD4+ and CD8+ T lymphocytes, plays an essential role in containing acute M. tuberculosis infection in murine models (for review see reference 6). Studies in mouse models using antibody depletion 7,8,9, adoptive transfer 10, and transgenic mouse strains deficient in either MHC class II 11,12 or CD4 12 have established the absolute requirement for CD4+ T cells in controlling an acute M. tuberculosis challenge. The key role of the CD4+ T cell in tuberculosis is thought to be its ability to produce the cytokine IFN-γ, which is essential in the control of experimental tuberculosis in mice 13,14 and is the first identified human immunologic factor essential to resistance against mycobacterial infection 15. IFN-γ is a critical factor for inducing macrophage synthesis of the enzyme inducible nitric oxide synthase (NOS2) 14,16. Upon activation, macrophages generate nitric oxide and other reactive nitrogen intermediates (RNIs), the best characterized antituberculous effector molecules in the mouse (for review see reference 17). Evidence is mounting that RNIs also play a role in antimycobacterial defense in humans 18. It is also likely that RNI-independent mechanisms induced by IFN-γ participate in protection against tuberculosis 17,19.

Despite the large body of knowledge on the immune response required to control an experimental acute M. tuberculosis infection, little is known about the immunologic mechanisms responsible for maintaining a latent infection. Studies using persistent tuberculosis in mice to model latent tuberculosis in humans have demonstrated that RNIs are required to prevent reactivation of persistent infection 19,20. IFN-γ and TNF-α also participate in maintenance of persistent M. tuberculosis infection in mice (21, 22, and Mohan, V.P., C.A. Scanga, K. Yu, H.M. Scott, K.E. Tanaka, E. Tsang, J.L. Flynn, and J. Chan, manuscript submitted for publication). Although CD4+ T cells clearly are important early in the course of M. tuberculosis infection, the role of these cells beyond the acute phase of infection when a vigorous immune response has been established is unknown. CD4+ T cell–deficient mice succumbed to acute tuberculosis, although the level of IFN-γ was merely delayed compared with control mice; a compensatory increase in CD8+ T cells producing IFN-γ occurred by 4 wk after infection 12. This suggested that subsequent to the induction of an immune response to the infection, other cells might be capable of producing IFN-γ, reducing the requirement for CD4+ T cells.

In this study, we tested whether CD4+ T cells are required to prevent reactivation tuberculosis, using a previously described murine model of persistent tuberculosis 19. Depletion of CD4+ T cells resulted in marked reactivation of the infection. However, the expression of IFN-γ and NOS2 in the lungs of CD4+ T cell–depleted mice was similar to that in control mice, suggesting that the mechanism by which CD4+ T cells maintain a quiescent infection is not simply production of IFN-γ.

Materials And Methods

Mice.

8–10-wk-old female C57BL/6 mice (The Jackson Laboratory and Charles River Laboratories) were housed in microisolator cages under specific pathogen–free biosafety level 3 conditions and monitored for various viruses, bacteria, and parasites. All animal protocols used in this study were approved by the Institutional Animal Care and Use Committees at Albert Einstein College of Medicine and the University of Pittsburgh School of Medicine.

Mycobacteria and Infection.

The virulent Erdman strain (The Trudeau Institute, Saranac Lake, NY) of M. tuberculosis was passed through mice, grown once in culture, and frozen in aliquots. For infections, an aliquot was thawed, diluted in PBS with 0.05% Tween-80, and briefly sonicated in a cup horn sonicator, and 100 μl (containing 5 × 103 viable bacilli) was injected into mice via a lateral tail vein. The low-dose latency model, described previously 19, was used in this study. In brief, mice were infected with M. tuberculosis. Within 1 mo, the numbers of bacilli in the lungs and spleen reach 105–106 and the infection is stably maintained for >10 mo.

In Vivo Depletion of CD4+ T Cells.

6–8 mo after infection, during the period of stable infection, CD4+ T cells were depleted in vivo using 0.5 mg of rat anti-CD4 mAb GK1.5 delivered intraperitoneally weekly (n = 20 mice per experiment). The GK1.5 hybridoma (ATCC) was used to produce ascites (Harlan Bioproducts for Science) and has been used by others for in vivo CD4+ T cell depletion 23. The ascites were subjected to sodium ammonium sulfate precipitation to obtain CD4-specific IgG. Similarly infected control mice (n = 19 mice per experiment) received normal rat IgG (Jackson ImmunoResearch Laboratories). Mice that became moribund during the infection were humanely killed.

CFU Determination.

Organs were homogenized in PBS/0.05% Tween-80 and dilutions were plated on 7H10 agar. Plates were incubated at 37°C in 5% CO2, and colonies were enumerated after 21 d.

Histology and Immunohistochemistry.

Formalin-fixed, paraffin-embedded tissue sections were stained with hematoxylin and eosin for histological analysis and for acid fast bacilli using Kinyoun's method (Difco Labs.) according to the manufacturer's directions. Immunohistochemical staining for NOS2 was performed as described previously 19. In brief, formalin-fixed, paraffin-embedded tissue sections were deparaffinized, subjected to microwave antigen retrieval, and stained using rabbit anti-NOS2 antibody (Transduction Labs.), followed by biotinylated anti–rabbit IgG (Vector Labs.). Visualization of antibody for NOS2 was accomplished using an ABC kit (Vector Labs.) and diaminobenzidine (DAB) or 3-amino-9-ethyl-carbazole substrate (Sigma-Aldrich) followed by a hematoxylin counterstain. Nitrotyrosine was detected in similarly treated tissue sections using a rabbit polyclonal antinitrotyrosine antibody (Upstate Biotechnology) with DAB substrate and a methyl green counterstain. Antibody specificity was confirmed by preincubating the primary antibody with 10 mM 3-nitro-l-tyrosine (Sigma-Aldrich) before incubation with the tissue sections.

FACS® Analysis.

To determine the efficacy of GK1.5 anti-CD4 mAb for CD4+ T cell depletion in vivo, single-cell suspensions were prepared by passing spleens harvested from long term–infected mice treated with GK1.5 or normal rat IgG through mesh bags (Bally Ribbon Mills). RBCs were lysed with Tris/NH4Cl, and the cells were stained with mAbs directed against CD3, CD4, and CD8 (clones 145-2C11, H129.19, and 53-6.7, respectively; BD PharMingen) at 0.2 μg/106 cells. The cells were fixed with 2% paraformaldehyde and subjected to three-color FACS® analysis using CELLQuest™ software on a FACSCalibur™ (Becton Dickinson). Cells were gated on lymphocytes by forward and side scatter parameters.

Intracellular Cytokine Staining.

All antibodies used in FACS® analysis were obtained from BD PharMingen. Single-cell suspensions of lungs were prepared as above. Cells were stimulated overnight with anti-CD3 and anti-CD28 mAb (clones 145-2C11 and 37.51, respectively) at 0.1 and 1.0 μg/ml, respectively. Monensin (3 μM) was added for the final 6 h. Cells were then stained for the surface markers CD4 and CD8 as above, permeabilized with saponin, and stained for intracellular IFN-γ (anti–IFN-γ–PE mAb; clone XMG1.2). Cells were fixed with 1% paraformaldehyde and analyzed by three-color FACS®. Cells were gated on lymphocytes by forward and side scatter parameters.

Ribonuclease Protection Assay.

Total RNA was isolated from flash-frozen lungs using Trizol (GIBCO BRL) according to manufacturer's instructions, with an additional phenol-chloroform extraction before RNA precipitation. Gene expression was assessed using the RiboQuant™ ribonuclease protection assay (RPA) system (BD PharMingen) using customized template sets that included probes for NOS2, IFN-γ, IL-12p40, TNF-α, IL-1α, and IL-1β and the GAPDH and L32 housekeeping genes in one set and IL-4, IL-12p40, IL-10, IL-1α, IL-1β, IFN-γ, GAPDH, and L32 in another. Expression of cell surface marker genes was analyzed with the CD-1 RPA probe (BD PharMingen). Band intensities on the autoradiographs were quantitated by densitometry (Personal Densitometer SI; Molecular Dynamics), and the ratio of band intensity between the gene of interest and a housekeeping gene, either GAPDH or L32, was calculated.

Statistical Analysis.

Where appropriate, values were tested for statistical significance by unpaired Student's t test using InStat (v.2.03; GraphPad Software). CFU values were subjected to log transformation before analysis. P values <0.05 were considered to be significant.

Results

Course of Chronic Persistent Infection after CD4+ T Cell Depletion.

A murine model of latent human tuberculosis 19 was used to assess the contribution of CD4+ T cells to preventing reactivation. C57BL/6 mice were infected intravenously with 5 × 103 CFU M. tuberculosis Erdman strain, and the infection was allowed to progress for 6–8 mo. As detailed previously 19, the bacillary burden in the lungs, spleen, and liver increased for 3–4 wk after infection. As a cellular immune response to the infection is established, the numbers of bacilli stabilize in the organs and remain essentially unchanged for at least 10 mo. During this period, the mice remain clinically healthy. This prolonged period of stable bacillary numbers in an apparently healthy animal may be best characterized as a persistent or chronic infection but can serve as a useful experimental model of human latent tuberculosis.

6 mo after infection, during the period of stable bacterial load, CD4+ T cells were depleted in vivo by weekly administration of rat anti–murine CD4 mAb GK1.5. Splenocytes were analyzed by FACS® to evaluate the efficacy of the depletion regimen 11 d after initiation of the antibody treatment. GK1.5-treated mice exhibited a 93% reduction in the number of splenic CD4+ T cells as compared with similarly infected control mice receiving normal rat IgG at this time point (1.7% CD4+ T cells [mean, two mice] versus 22.7%; n = 1). The number of CD8+ T cells was unchanged in GK1.5-treated mice compared with control mice (13.4 versus 14.0%). The bacillary burden increased steadily in lung, liver, and spleen beginning as early as 2 wk after depletion of CD4+ T cells, while the numbers of bacilli in the organs of animals receiving normal rat IgG remained unchanged (Fig. 1 A). By 9 wk of antibody treatment, the GK1.5-treated mice had ∼60-fold more mycobacteria in the lungs and spleen and 400-fold more in the liver compared with IgG-treated control mice (Fig. 1 A). Although specific mice were not set aside in this experiment to follow mortality, 2 of 20 CD4+ T cell–depleted mice succumbed to tuberculosis (at 50 and 56 d of GK1.5 treatment) throughout the course of this experiment, while all control mice appeared healthy. In a second trial in which mice were infected for 8 mo before anti-CD4 treatment, the kinetics of tissue bacterial burden were virtually identical in GK1.5-treated mice (Fig. 1 B). CD4+ T cell–depleted mice exhibited a mean survival time of 76 ± 25 d with 100% mortality by 109 d of antibody treatment (range 46–109 d; n = 9 mice followed for mortality of 25 total). 4 of 25 control IgG–treated mice died during the antibody treatment period at 27, 35, 54, and 97 d of treatment. Although the bacillary burdens were not determined in the animals that died, analysis of other control mice indicated that bacterial numbers and pathology remained stable throughout the treatment period (Fig. 1 B). It is likely that the four deaths in the control IgG mice were not due to progression of tuberculosis but instead may have been caused by inadvertent penetration of visceral organs during antibody administration. In sum, depletion of CD4+ T cells was detrimental to control of chronic persistent tuberculosis.

Pathology in Persistently Infected Mice after CD4+ T Cell Depletion.

Tissue sections from the CD4+ T cell–depleted and control mice were examined for histopathology. Not surprisingly, substantial pathology was present in lungs of both GK1.5- and control antibody–treated mice, as both groups had been persistently infected with M. tuberculosis for >6 mo. In control mice receiving normal rat IgG, the granulomatous response in the lungs consisted of well delineated lymphoid aggregates containing a minor monocytic component, situated among a predominately lymphohistiocytic infiltrate with poorly defined margins. There was an abundance of foamy macrophages, some of which were intraalveolar (Fig. 2A and Fig. C). Alveolitis and interstitial pneumonitis were also observed. In the CD4+ T cell–depleted mice, a gradual disappearance of the lymphoid aggregates was apparent such that they were barely detectable microscopically by 63 d of GK1.5 treatment (Fig. 2B and Fig. D). Most striking was the prominence of necrosis in the pulmonic granulomata of the CD4+ T cell–depleted mice, involving multiple areas with marked destruction of the lung architecture (Fig. 2 D), particularly at later time points when tissue bacillary burden was high. Eosinophilic plasma cells (Mott cells) were noted in granulomatous areas in the lungs of both the CD4+ T cell–depleted and control mice. These histopathologic findings are consistent with the increased CFU in organs of GK1.5-treated mice, in that high bacterial numbers are frequently associated with necrosis of lung tissue 24. Histologic stains for Pneumocystis carinii were negative (data not shown) and excluded this opportunistic pathogen as contributing to the histopathology observed in the CD4+ T cell–depleted mice.

NOS2 Expression and Activity in CD4+ T Cell–depleted Mice.

The expression of NOS2 protein in lungs, spleen, and liver of CD4+ T cell–depleted and control mice was assessed immunohistochemically. High levels of NOS2 protein localized to foamy macrophages within granulomatous areas were observed at 6 mo after infection, before antibody treatment, as previously described (19; data not shown). There was no discernible difference in tissue NOS2 expression between GK1.5- and normal IgG-treated mice after 2 wk (Fig. 3A and Fig. E), 6 wk (Fig. 3B and Fig. F), or 9 wk (data not shown) of antibody treatment. The loss of CD4+ T cells did not appear to affect the number of NOS2-positive cells (Table), the distribution of those cells, nor the intensity of NOS2 staining in the lung. This observation was supported by RPA analyses of total lung RNA from GK1.5 and control mice (Fig. 4). Expression of the NOS2 gene was similar between CD4+ T cell–depleted and control mice at all time points. More direct evidence of continued production of RNIs in the CD4+ T cell–depleted mice was provided by immunostaining tissue sections from these mice using antinitrotyrosine antibody to detect nitrated proteins. Nitrotyrosine is generated via nitration of tyrosine by peroxynitrite, a product of NO and superoxide anion, and thus reflects NOS2 activity. Nitrotyrosine staining was similar in the lung tissue and granulomas of GK1.5- and normal IgG-treated mice (Fig. 3C and Fig. H). When the antinitrotyrosine antibody was preincubated with nitrotyrosine before being used in this assay, all staining was abolished (Fig. 3D and Fig. G), confirming the staining specificity. These results indicate that CD4+ T cell depletion–induced reactivation was not due to a deficiency in RNI-dependent antimicrobial mechanisms.

IFN-γ Production after CD4+ T Cell Depletion.

IFN-γ is a critical factor required for activating macrophages to produce NOS2 14,16. Therefore, the finding of unimpaired NOS2 production in CD4+ T cell–depleted mice suggested that an alternate source of IFN-γ, independent of CD4+ T cells, existed in GK1.5-treated mice. To address the possible source of IFN-γ in the lungs of the mice, cell populations and cytokine production in the lungs were assessed by FACS® analysis. Staining for CD4+ T cells confirmed that GK1.5 treatment was efficacious in the lungs, with >94% depletion of CD4+ T cells in lungs of GK1.5-treated mice compared with normal rat IgG–treated mice after 4 and 9 wk of antibody treatment (Fig. 5). However, the mean number of cells recovered from lung tissue did not differ significantly between GK1.5- and normal rat IgG–treated mice (Table). There was an increase in the percentage of CD8+ T cells in CD4+ T cell–depleted mice compared with control IgG–treated mice at both 4 wk (mean GK1.5, 20.5%; mean control IgG, 10.8%; P = 0.03) and 9 wk (mean GK1.5, 34.1%; mean control IgG, 26.9%; P = 0.076). Because the numbers of cells recovered from the lungs of mice from both groups were similar (Table), this increased percentage corresponds to an actual increase in total number of CD8+ T cells.

We used intracellular cytokine staining of lung cells to assess production of IFN-γ in the lungs of the CD4+ T cell–depleted, persistently infected mice. Despite effective depletion of the pulmonic CD4+ T cell compartment, there was not a significant decrease in the percentage of total lymphocytes producing IFN-γ after 4 or 9 wk of depletion compared with IgG-treated control mice. Data from representative mice are shown in Fig. 5. Lymphocytes from uninfected mice produce very little IFN-γ after ex vivo stimulation 25. As noted above, there was an increase in numbers of CD8+ T cells in GK1.5-treated mice, and these CD8+ T cells were capable of producing IFN-γ (Fig. 5). RPA analysis of total lung RNA confirmed that the levels of IFN-γ expression in the lungs of GK1.5- and control IgG–treated mice were indeed comparable (Fig. 4). The efficacy of GK1.5 treatment, the increase in CD8+ T cells upon CD4+ T cell depletion, and the comparable IFN-γ expression in the lungs of GK1.5 and control mice was confirmed in a second CD4+ T cell depletion experiment (data not shown). Taken together, these data indicate that in the lungs of persistently infected mice depleted of CD4+ T cells, there was an increase in CD8+ T cells capable of IFN-γ production.

Gene Expression in the Lungs after CD4+ T Cell Depletion in Chronic Persistent Tuberculosis.

RPA analysis of total lung RNA revealed that in the CD4+ T cell–depleted mice, TNF-α expression remained relatively steady, except at the last time point when a significant increase was observed (Fig. 4). This may reflect the more severe pathology in the lungs of these mice relative to the mice receiving normal rat IgG (Fig. 2). Substantial expression of IL-1α and IL–1β was detected over the course of the antibody treatment regimen, but there was little difference between the mice receiving GK1.5 and those receiving control antibody (Fig. 4). A low level of IL-12p40 expression was noted in both groups and did not vary throughout the antibody treatment period (data not shown). Expression of IL-4 (Fig. 4) and IL-10 (data not shown) was low or undetectable at all time points in both groups of mice.

2 wk after initiation of anti-CD4 antibody treatment, there was a decrease in CD4 gene expression in the lungs relative to that in mice receiving control antibody, as assessed by RPA (Fig. 6). This difference was significant (P = 0.002) by 6 wk of antibody treatment and was maintained throughout the remainder of the experiment (Fig. 6). These data confirmed the results obtained by FACS® analysis that GK1.5 was efficacious in depleting CD4+ T cells in the lungs over a long treatment period. The expression of CD3 was not significantly different among the GK1.5- and rat IgG–treated mice at the various time points studied (Fig. 6). There was a significant decrease in the expression of CD19 (a pan-B cell marker) in CD4+ T cell–depleted mice by 6 wk of GK1.5 treatment compared with IgG-treated mice (P = 0.001; Fig. 6), suggesting that fewer B cells were present in those mice.

Discussion

The association between the deficiency of CD4+ T cells observed with AIDS and the incidence of reactivation of latent tuberculosis suggests that CD4+ T cells play an important role in preventing reactivation tuberculosis. In human studies, M. tuberculosis–specific CD4+ T cells secrete IFN-γ 26, and a recent publication suggests that in vitro IFN-γ can induce human macrophages to kill M. tuberculosis coincident with the expression of NOS2 27. CD4+ T cells are thought to be the primary source of IFN-γ and are required to control acute M. tuberculosis infection in mice 7,8,10,12. The importance of IFN-γ in controlling mycobacterial infections has been demonstrated in humans (for review see reference 15) and mice 13,14. NOS2 expression is severely compromised at both the gene 14 and protein (our unpublished observations) levels in M. tuberculosis–infected mice with a disruption in the IFN-γ gene, and these animals succumb quickly to infection 14. In addition, emerging evidence suggests that both IFN-γ 22 and NOS2 19,20 play a role in preventing reactivation of persistent tuberculosis in murine models of latency.

The requirement for CD4+ T cells in preventing reactivation was tested in a murine model of persistent tuberculosis in which CD4+ T cells were depleted by antibody treatment beginning 6–8 mo after infection. Depletion of CD4+ T cells resulted in fatal exacerbation of the quiescent infection with markedly increased bacterial numbers, indicating that CD4+ T cells are required to prevent reactivation. Two surprising findings emerged from this study. First, depletion of CD4+ T cells, arguably the major source of IFN-γ in tuberculous infection, did not result in an appreciable decrease in overall IFN-γ production in the lungs. Second, control of mycobacterial growth was lost in the CD4+ T cell–depleted mice, despite unimpaired NOS2 expression and activity.

CD8+ T cells are capable of producing significant amounts of IFN-γ in response to M. tuberculosis in both murine 11,12,25 and human 28,29 systems. Indeed, in this study, depletion of CD4+ T cells in mice persistently infected with M. tuberculosis resulted in an increase in the number of CD8+ T cells, a subset of which was producing IFN-γ. Therefore, as in acute tuberculosis models using CD4+ T cell–deficient mice 12, a compensatory increase in CD8+ T cells producing IFN-γ in the lungs of GK1.5-treated mice could at least partially account for the undiminished IFN-γ expression seen in mice depleted of CD4+ T cells. Other cell types cannot be excluded from contributing to IFN-γ production. CD48 T cells from Leishmania major–infected mice have been shown to produce IFN-γ in vitro 30, although this was not observed in acute tuberculosis models in CD4+ T cell–deficient mice 12. NK cells have been implicated as important sources of IFN-γ in the early immunologic response against several infectious agents, including L. major 31 and Toxoplasma gondii 32. Finally, macrophages have been reported to produce IFN-γ in response to mycobacteria 33,34. As cell markers other than CD8 and CD4 were not analyzed by FACS®, contribution to the total IFN-γ production by these other cell types cannot be excluded. Whatever the source of the IFN-γ, it is clear that its presence did not prevent reactivation of the infection and thus did not adequately compensate for the lack of CD4+ T cells. Perhaps this cytokine must be delivered directly to the infected macrophage to be effective, and cell types other than CD4+ T cells may not do so as efficiently within the granuloma. Studies are underway to determine the precise location of and the cell type responsible for IFN-γ production in the absence of CD4+ T cells in this model.

Studies using murine models of latent tuberculosis have established the importance of NOS2 in preventing reactivation 19,20. In this study, however, CD4+ T cell depletion resulted in a dramatic increase in the numbers of bacilli despite levels of NOS2 comparable to that of mice receiving control IgG. Nitrated tyrosine residues were detected in lung tissue of both CD4+ T cell–depleted and control mice, and their distribution was colocalized to areas of NOS2 expression. These data provide evidence that NOS2 gene and protein expression and activity were unaffected by the depletion of CD4+ T cells. NOS2 expression is dependent on IFN-γ 14,16, and thus the unimpaired NOS2 expression provides further, indirect evidence of undiminished IFN-γ expression in the lungs of CD4+ T cell–depleted mice. These surprising findings suggest that CD4+ T cells prevent reactivation by a mechanism that is not solely mediated by IFN-γ or NOS2-generated RNIs. Although unlikely, it is possible that there was an undetected decrease in IFN-γ and/or NOS2 production immediately following CD4+ T cell depletion, as the earliest postdepletion time point examined was 2 wk. In acute M. tuberculosis infection of CD4−/− or MHC class II−/− mice, production of wild-type levels of IFN-γ and NOS2 was delayed for only 2 wk 12 and was apparently sufficient to render these mice very susceptible to tuberculosis. However, in a chronic infection, macrophages are already primed to produce NOS2, and a short-lived IFN-γ deficiency would be less likely to have an effect than in the immunologically naive mice used for the acute tuberculosis studies. This study raises the question of how the reactivating bacilli not only survive but actively replicate in the presence of NOS2 and the resultant RNIs. Although NOS2-generated RNIs are a necessary antimycobacterial effector mechanism, it is not sufficient to eliminate M. tuberculosis infection in vivo 19,20,35.

In light of reactivation in the presence of control levels of IFN-γ and NOS2 in the CD4+ T cell–depleted mice, additional CD4+ T cell functions capable of maintaining latency must be considered. First, the requirement of CD4+ T cells for a vigorous granulomatous reaction in mycobaterial infection has been reported 12,36,37. In tuberculosis patients coinfected with HIV, the structural integrity of the tuberculous granuloma appears to correlate with the number of total peripheral CD4+ T cells 38,39. Although the pathology and general loss of granuloma structure observed in the lungs of GK1.5-treated mice during reactivation may be directly related to bacterial numbers, we cannot exclude a role for CD4+ T cells in maintenance of an organized granuloma during latent infection. Second, a key element in controlling M. tuberculosis infection is macrophage activation. CD4+ T cells may have IFN-γ– and NOS2-independent pathways to activate mycobactericidal or mycobacteriostatic effector functions in macrophages. Macrophages can be activated through direct contact via interaction between CD40 on macrophages and CD40L on activated lymphocytes 40, and this can result in NOS2 induction 41. Although this mechanism is apparently necessary for successful immunity to L. major 42,43,44, CD40L−/− mice were not more susceptible to M. tuberculosis infection 45. As negative data in genetically deficient mice can be difficult to interpret, the role of CD40 in tuberculosis remains unclear. Third, it is possible that depletion of CD4+ T cells results in increased production of cytokines capable of deactivating macrophages. Two such cytokines are IL-10 46 and TGF-β 47. Indeed, in HIV-infected individuals, production of IL-10 and TGF-β by PBMCs is augmented (for review see reference 48). IL-10 mRNA was expressed at very low levels that were similar in lungs of CD4+ T cell–depleted and control mice, so this cytokine is not likely to contribute to the reactivation observed in this model. TGF-β produced by M. tuberculosis–infected macrophages has been reported to downregulate macrophage function 49. TGF-β was not examined in this study, and so the possibility remains that this cytokine may increase in the absence of CD4+ T cells, leading to macrophage deactivation and increased bacterial growth. Fourth, CD4+ T cells are thought to be important in maintaining an adequate CD8+ T cell response 50,51,52. We and others have demonstrated previously that CD8+ T cells participate in control of tuberculosis in mice 9,11,25,53,54, and the presence of mycobacterial-specific CD8+ T cells in tuberculosis patients has been recently reported 28,29,55. Mycobacteria-specific CD8+ T cells can produce IFN-γ or act as cytotoxic cells for infected macrophages 25,28,29,56,57. Recent data indicate that CD8+ CTLs that produce granulysin can kill intracellular M. tuberculosis 58. Although CD8+ T cell numbers were increased after CD4+ T cell depletion, and these cells were clearly primed to produce IFN-γ, the actual function of these cells in the lungs in the absence of CD4+ T cells is unclear. Perhaps CD4+ T cells are necessary for complete function or maintenance of CD8+ cytolytic activity, without which control of chronic persistent tuberculosis is compromised.

Results obtained from RPA analysis of cell surface marker expression suggested that there was a substantial reduction in B cells in the lungs of CD4+ T cell–depleted mice. T cells play an important role in modulating B cell differentiation, function, and life span. CD4+ T cells may also regulate production of specific B cell chemoattractants. Plasmacytoid cells were observed in lung sections, but CD19 is downregulated on fully differentiated B cells in humans 59,60, and so this does not contradict the RPA data. The role of antibodies in immunity to M. tuberculosis is controversial 61, and B cells may contribute to the immune response to intracellular infections in an antibody-independent manner, as suggested recently in a Francisella tularensis murine model 62. Studies using B cell–deficient mice are contradictory, suggesting that B cells contribute modestly 63 or not at all 64 to control of M. tuberculosis infection. A careful analysis of the contribution of B cells in latent tuberculosis may further define protective mechanisms against reactivation of latent infections.

In this murine model of latent tuberculosis, inhibition of NOS2 19, neutralization of TNF-α (Mohan, V.P., C.A. Scanga, K. Yu, H.M. Scott, K.E. Tanaka, E. Tsang, J.L. Flynn, and J. Chan, manuscript submitted for publication), or depletion of CD4+ T cells (this study) resulted in reactivation of infection and fatal tuberculosis. However, there were significant differences in the course of reactivation and pathology as well as expression of cytokines, depending on the immunologic manipulation implemented. Inhibition of NOS2 resulted in slowly progressive reactivation primarily in the lungs 19, whereas CD4+ T cell depletion resulted in a rapid and dramatic increase in bacterial numbers in liver and spleen, as well as lung, shortly after initiation of antibody treatment. TNF-α neutralization caused an initial increase in bacterial numbers that stabilized after 3 wk, despite decreased NOS2 expression (Mohan, V.P., C.A. Scanga, K. Yu, H.M. Scott, K.E. Tanaka, E. Tsang, J.L. Flynn, and J. Chan, manuscript submitted for publication). Severe pathology was observed in TNF-α–neutralized mice that was very different than that observed in CD4+ T cell–depleted mice, with marked disorganization of granulomata and increased leukocytic infiltration (Mohan, V.P., C.A. Scanga, K. Yu, H.M. Scott, K.E. Tanaka, E. Tsang, J.L. Flynn, and J. Chan, manuscript submitted for publication). Although interpretations based on interexperimental comparisons must be made with prudence, the results of our previous studies and results presented in this study suggest that individual immunologic components play a variety of roles in preventing reactivation, highlighting the complex nature of the protective response in chronic tuberculosis.

In summary, this study demonstrates the requirement for CD4+ T cells in maintaining a persistent M. tuberculosis infection. However, depletion of CD4+ T cells did not result in decreased tissue expression of IFN-γ or NOS2, the two most characterized CD4+ T cell–related antimycobacterial factors. Thus, the role of the CD4+ T cell in persistent tuberculosis is more complex than merely as a source of IFN-γ for macrophage activation. Further investigation of the mechanism by which the CD4+ T cell functions during latency may guide the development of therapies and vaccines designed to prevent reactivation of tuberculosis.

Acknowledgments

We gratefully acknowledge Dr. Joseph Ahearn for use of his FACSCalibur™ and Dr. Simon Watkins, Director of the University of Pittsburgh Center for Biologic Imaging. We thank Dr. Denise Croix and Natalya Serbina for assistance in analyzing FACS® data and the members of the Flynn and Chan laboratories for helpful discussions.

This work was supported by National Institutes of Health grant AI36990 (to J. Chan and J.L. Flynn).

References

References
World Health Organization. 1999. The World Health Report 1999. Making a Difference. 116.
Stead
W.W.
Pathogenesis of a first episode of chronic pulmonary tuberculosis in manrecrudescence of residuals of the primary infection or exogenous reinfection?
Am. Rev. Resp. Dis.
95
1967
729
745
[PubMed]
Powell
K.E.
,
Farer
L.S.
The rising age of tuberculosis patients
J. Infect. Dis.
142
1980
946
948
[PubMed]
Cisneros
J.R.
,
Murray
K.M.
Corticosteroids in tuberculosis
Ann. Pharmacotherapy.
30
1996
1298
1303
Raviglione
M.C.
,
Snider
D.E.
,
Kochi
A.
Global epidemiology of tuberculosismorbidity and mortality of a global epidemic
JAMA.
273
1995
220
226
[PubMed]
Chan
J.
,
Kaufmann
S.H.E.
,
Immune Mechanisms of Protection
Bloom
B.R.
TuberculosisPathogenesis, Protection and Control
1994
389
415
American Society for Microbiology
Washington, D.C.
Leveton
C.
,
Barnass
S.
,
Champion
B.
,
Lucas
S.
,
De Souza
B.
,
Nicol
M.
,
Banerjee
D.
,
Rook
G.
T-cell mediated protection of mice against virulent Mycobacterium tuberculosis
Infect. Immun.
57
1989
390
395
[PubMed]
Flory
C.
,
Hubbard
R.
,
Collins
F.
Effects of in vivo T lymphocyte subset depletion on mycobacterial infections in mice
J. Leukoc. Biol.
51
1992
225
229
[PubMed]
Muller
I.
,
Cobbold
S.
,
Waldmann
H.
,
Kaufmann
S.H.E.
Impaired resistance to Mycobacterium tuberculosis infection after selective in vivo depletion of L3T4+ and Lyt2+ T cells
Infect. Immunol.
55
1987
2037
2041
[PubMed]
Orme
I.
Characteristics and specificity of acquired immunologic memory to Mycobacterium tuberculosis infection
J. Immunol.
140
1988
3589
3593
[PubMed]
Tascon
R.E.
,
Stavropoulos
E.
,
Lukacs
K.V.
,
Colston
M.J.
Protection against Mycobacterium tuberculosis infection by CD8 T cells requires production of gamma interferon
Infect. Immun.
66
1998
830
834
[PubMed]
Caruso
A.M.
,
Serbina
N.
,
Klein
E.
,
Triebold
K.
,
Bloom
B.R.
,
Flynn
J.L.
Mice deficient in CD4 T cells have only transiently diminished levels of IFN-γ, yet succumb to tuberculosis
J. Immunol.
162
1999
5407
5416
[PubMed]
Cooper
A.M.
,
Dalton
D.K.
,
Stewart
T.A.
,
Griffen
J.P.
,
Russell
D.G.
,
Orme
I.M.
Disseminated tuberculosis in IFN-γ gene-disrupted mice
J. Exp. Med.
178
1993
2243
2248
[PubMed]
Flynn
J.L.
,
Chan
J.
,
Triebold
K.J.
,
Dalton
D.K.
,
Stewart
T.A.
,
Bloom
B.R.
An essential role for interferon-γ in resistance to Mycobacterium tuberculosis infection
J. Exp. Med.
178
1993
2249
2254
[PubMed]
Ottenhof
T.H.
,
Kumararatne
D.
,
Casanova
J.L.
Novel human immunodeficiencies reveal the essential role of type-1 cytokines in immunity to intracellular bacteria
Immunol. Today.
19
1998
491
494
[PubMed]
Dalton
D.K.
,
Pitts-Meek
S.
,
Keshav
S.
,
Figari
I.S.
,
Bradley
A.
,
Stewart
T.A.
Multiple defects of immune cell function in mice with disrupted interferon-gamma genes
Science.
259
1993
1739
1742
[PubMed]
Chan
J.
,
Flynn
J.
,
Nitric oxide in Mycobacterium tuberculosis infection
Fang
F.
Nitric Oxide and Infection
1999
281
310
Plenum Publishers
New York
Nicholson
S.
,
Bonecini-Almeida
M.
,
Silva
J.R.L.
,
Nathan
C.
,
Xie
Q.-W.
,
Mumford
R.
,
Weidner
J.R.
,
Calaycay
J.
,
Geng
J.
,
Boechat
N.
Inducible nitric oxide synthase in pulmonary alveolar macrophages from patients with tuberculosis
J. Exp. Med.
184
1996
2293
2302
[PubMed]
Flynn
J.L.
,
Scanga
C.A.
,
Tanaka
K.E.
,
Chan
J.
Effects of aminoguanidine on latent murine tuberculosis
J. Immunol.
160
1998
1796
1803
[PubMed]
MacMicking
J.D.
,
North
R.J.
,
LaCourse
R.
,
Mudgett
J.S.
,
Shah
S.K.
,
Nathan
C.F.
Identification of nitric oxide synthase as a protective locus against tuberculosis
Proc. Natl. Acad. Sci. USA.
94
1997
5243
5248
[PubMed]
Adams
L.B.
,
Mason
C.M.
,
Kolls
J.K.
,
Scollard
D.
,
Krahenbuhl
J.L.
,
Nelson
S.
Exacerbation of acute and chronic murine tuberculosis by administration of a tumor necrosis factor receptor-expressing adenovirus
J. Infect. Dis.
171
1995
400
405
[PubMed]
Scanga
C.A.
,
Mohan
V.P.
,
Joseph
H.
,
Yu
K.
,
Chan
J.
,
Flynn
J.
Reactivation of latent tuberculosisvariations on the Cornell murine model
Infect. Immun.
67
1999
4531
4538
[PubMed]
Topham
D.J.
,
Tripp
R.A.
,
Doherty
P.C.
CD8+ T cells clear influenza virus by perforin or Fas-dependent processes
J. Immunol.
159
1997
5197
5200
[PubMed]
Rhoades
E.R.
,
Frank
A.A.
,
Orme
I.M.
Progression of chronic pulmonary tuberculosis in mice aerogenically infected with virulent Mycobacterium tuberculosis
Tubercle Lung Dis.
78
1997
57
66
Serbina
N.V.
,
Flynn
J.L.
Early emergence of CD8+ T cells primed for production of type 1 cytokines in the lungs of Mycobacterium tuberculosis-infected mice
Infect. Immun.
67
1999
3980
3988
[PubMed]
Boom
W.H.
,
Wallis
R.S.
,
Chervenak
K.A.
Human Mycobacterium tuberculosis-reactive CD4+ T-cell clonesheterogeneity in antigen recognition, cytokine production, and cytotoxicity for mononuclear phagocytes
Infect. Immun.
59
1991
2737
2743
[PubMed]
Almeida
G.B.
,
Chitale
C.
,
Boutsikakis
I.
,
Geng
J.-Y.
,
Doo
H.
,
He
S.
,
Ho
J.L.
Induction of in vitro human macrophage anti-Mycobacterium tuberculosis activityrequirement for interferon-γ and primed lymphocytes
J. Immunol.
160
1998
4490
4499
[PubMed]
Lalvani
A.
,
Brookes
R.
,
Wilkinson
R.
,
Malin
A.
,
Pathan
A.
,
Andersen
P.
,
Dockrell
H.
,
Pasvol
G.
,
Hill
A.
Human cytolytic and interferon gamma-secreting CD8+ T lymphocytes specific for Mycobacterium tuberculosis
Proc. Natl. Acad. Sci. USA.
95
1998
270
275
[PubMed]
Lewinsohn
D.
,
Alderson
M.
,
Briden
A.
,
Riddell
S.
,
Reed
S.
,
Grabstein
K.
Characterization of human CD8+ T cells reactive with Mycobacterium tuberculosis–infected antigen-presenting cells
J. Exp. Med.
187
1998
1633
1640
[PubMed]
Locksley
R.M.
,
Reiner
S.L.
,
Hatam
F.
,
Littman
D.R.
,
Killeen
N.
Helper T cells without CD4control of leishmaniasis in CD4-deficient mice
Science.
261
1993
1448
1451
[PubMed]
Scharton
T.M.
,
Scott
P.
Natural killer cells are a source of interferon γ that drives differentiation of CD4+ T cell subsets and induces early resistance to Leishmania major in mice
J. Exp. Med.
178
1993
567
577
[PubMed]
Gazzinelli
R.T.
,
Hieny
S.
,
Wynn
T.A.
,
Wolf
S.
,
Sher
A.
Interleukin 12 is required for the T-lymphocyte-independent induction of interferon-γ by an intracellular parasite and induces resistance in T-cell-deficient hosts
Proc. Natl. Acad. Sci. USA.
90
1993
6115
6119
[PubMed]
Fenton
M.J.
,
Vermeulen
M.W.
,
Kim
S.
,
Burdick
M.
,
Strieter
R.M.
,
Kornfeld
H.
Induction of gamma interferon production in human macrophages by Mycobacterium tuberculosis
Infect. Immun.
65
1997
5149
5156
[PubMed]
Wang
J.
,
Wakeham
J.
,
Harkness
R.
,
Xing
Z.
Macrophages are a significant source of type 1 cytokines during mycobacterial infection
J. Clin. Invest.
103
1999
1023
1029
[PubMed]
Chan
J.
,
Tanaka
K.
,
Carroll
D.
,
Flynn
J.L.
,
Bloom
B.R.
Effect of nitric oxide synthase inhibitors on murine infection with Mycobacterium tuberculosis
Infect. Immun.
63
1995
736
740
[PubMed]
Ladel
C.H.
,
Daugelat
S.
,
Kaufmann
S.H.E.
Immune response to Mycobacterium bovis bacille Calmette Guerin infection in major histocompatibility complex class I- and II-deficient knock-out micecontribution of CD4 and CD8 T cells to acquired resistance
Eur. J. Immunol.
25
1995
377
384
[PubMed]
Hansch
H.C.
,
Smith
D.A.
,
Mielke
M.E.
,
Hahn
H.
,
Bancroft
G.J.
,
Ehlers
S.
Mechanisms of granuloma formation in murine Mycobacterium avium infectionthe contribution of CD4+ T cells
Int. Immunol.
8
1996
1299
1310
[PubMed]
Muller
H.
,
Kruger
S.
Immunohistochemical analysis of cell composition and in situ cytokine expression in HIV- and non-HIV-associated tuberculous lymphadenitis
Immunobiology.
191
1994
354
368
[PubMed]
DiPerri
G.
,
Cazzadori
A.
,
Vento
S.
,
Bonora
S.
,
Malena
M.
,
Bontempini
L.
,
Lanzafame
M.
,
Allegranzi
B.
,
Concia
E.
Comparative histopathological study of pulmonary tuberculosis in human immunodeficiency virus-infected and non-infected patients
Tubercle Lung Dis.
77
1996
244
249
Shu
U.
,
Kiniwa
M.
,
Wu
C.Y.
,
Maliszewski
C.
,
Vezzio
N.
,
Kakimi
J.
,
Gately
M.
,
Delespesse
G.
Activated T cells induce Interleukin-12 production by monocytes via CD40-CD40 ligand interaction
Eur. J. Immunol.
25
1995
1125
1128
[PubMed]
Tian
L.
,
Noelle
R.J.
,
Lawrence
D.A.
Activated T cells enhance nitric oxide production by murine splenic macrophages through gp39 and LFI-1
Eur. J. Immunol.
25
1995
306
309
[PubMed]
Campbell
K.A.
,
Ovendale
P.J.
,
Kennedy
M.K.
,
Fanslow
S.G.
,
Reed
S.G.
,
Maliszewski
C.R.
CD40 ligand is required for protective cell-mediated immunity to Leishmania major
Immunity.
4
1996
283
289
[PubMed]
Kamanaka
M.
,
Yu
P.
,
Yasui
T.
,
Yoshida
K.
,
Kawabe
T.
,
Horii
T.
,
Kishimoto
T.
,
Kikutani
H.
Protective role of CD40 in Leishmania major infection at two distinct phases of cell-mediated immunity
Immunity.
4
1996
275
281
[PubMed]
Soong
L.
,
Xu
J.C.
,
Grewal
I.S.
,
Kima
P.
,
Sun
J.
,
Longley
B.J.
,
Ruddle
N.H.
,
McMahon-Pratt
D.
,
Flavell
R.A.
Disruption of CD40-CD40 ligand interactions results in an enhanced susceptibility to Leishmania amazonensis infection
Immunity.
4
1996
263
273
[PubMed]
Campos-Neto
A.
,
Ovendale
P.
,
Bement
T.
,
Koppi
T.A.
,
Fanslow
W.C.
,
Rossi
M.A.
,
Alderson
M.R.
CD40 ligand is not essential for the development of cell-mediated immunity and resistance to Mycobacterium tuberculosis
J. Immunol.
160
1998
2037
2041
[PubMed]
Bogdan
C.
,
Vodovotz
Y.
,
Nathan
C.
Macrophage deactivation by interleukin 10
J. Exp. Med.
174
1991
1549
1555
[PubMed]
Tsunawaki
S.
,
Sporn
M.
,
Ding
A.
,
Nathan
C.
Deactivation of macrophages by transforming growth factor-beta
Nature.
334
1988
260
262
[PubMed]
Fauci
A.S.
Host factors and pathogenesis of HIV-induced disease
Nature.
384
1996
529
534
[PubMed]
Toossi
Z.
,
Gogate
P.
,
Shiratsuchi
H.
,
Young
T.
,
Ellner
J.J.
Enhanced production of TGF-beta by blood monocytes from patients with active tuberculosis and presence of TGF-beta in tuberculosis granulomatous lung lesions
J. Immunol.
154
1995
465
473
[PubMed]
Keene
J.A.
,
Forman
J.
Helper activity is required for the in vivo generation of cytotoxic T lymphocytes
J. Exp. Med.
155
1982
768
782
[PubMed]
von Herrath
M.G.
,
Yokoyama
M.
,
Dockter
J.
,
Oldstone
M.B.
,
Whitton
J.L.
CD4-deficient mice have reduced levels of memory cytotoxic T lymphocytes after immunization and show diminished resistance to subsequent virus challenge
J. Virol.
70
1996
1072
1079
[PubMed]
Bennett
S.R.
,
Carbone
F.R.
,
Karamalis
F.
,
Miller
J.F.
,
Heath
W.R.
Induction of a CD8 cytotoxic T lymphocyte response by cross-priming requires cognate CD4 T cell help
J. Exp. Med.
186
1997
65
70
[PubMed]
Flynn
J.L.
,
Goldstein
M.M.
,
Triebold
K.J.
,
Koller
B.
,
Bloom
B.R.
Major histocompatibility complex class I-restricted T cells are required for resistance to Mycobacterium tuberculosis infection
Proc. Natl. Acad. Sci. USA.
89
1992
12013
12017
[PubMed]
Orme
I.
,
Collins
F.
Adoptive protection of the Mycobacteria tuberculosis-infected lung
Cell. Immunol.
84
1984
113
120
[PubMed]
Stenger
S.
,
Mazzaccaro
R.
,
Uyemura
K.
,
Cho
S.
,
Barnes
P.
,
Rosat
J.
,
Sette
A.
,
Brenner
M.
,
Porcelli
S.
,
Bloom
B.
Differential effects of cytolytic T cell subsets on intracellular infection
Science.
276
1997
1684
1687
[PubMed]
De Libero
G.
,
Flesch
I.
,
Kaufmann
S.H.E.
Mycobacteria-reactive Lyt-2+ T cell lines
Eur. J. Immunol.
18
1988
59
66
[PubMed]
Serbina
N.V.
,
Liu
C.-C.
,
Scanga
C.A.
,
Flynn
J.L.
CD8+ cytotoxic T lymphocytes from lungs of M. tuberculosis infected mice express perforin in vivo and lyse infected macrophages
J. Immunol.
165
2000
353
363
[PubMed]
Stenger
S.
,
Hanson
D.A.
,
Teitelbaum
R.
,
Dewan
P.
,
Niazi
K.R.
,
Froelich
C.J.
,
Ganz
T.
,
Thoma-Uszynski
S.
,
Melian
A.
,
Bogdan
C.
An antimicrobial activity of cytotoxic T cells mediated by granulysin
Science.
282
1998
121
125
[PubMed]
Loken
M.R.
,
Shah
V.O.
,
Dattilio
K.L.
,
Civin
C.I.
Flow cytometric analysis of human bone marrow. II. Normal B cell development
Blood.
70
1987
1316
1324
[PubMed]
Scheuermann
R.H.
,
Racila
E.
CD19 antigen in leukemia and lymphoma diagnosis and immunotherapy
Leuk. Lymphoma.
18
1995
385
397
[PubMed]
Glatman-Freedman
A.
,
Casadevall
A.
Serum therapy for tuberculosis revisitedreappraisal of the role of antibody-mediated immunity against Mycobacterium tuberculosis
Clin. Microbiol. Rev.
11
1998
514
532
[PubMed]
Elkins
K.L.
,
Bosio
C.M.
,
Rhinehart-Jones
T.R.
Importance of B cells, but not specific antibodies, in primary and secondary protective immunity to the intracellular bacterium Francisella tularensis live vaccine strain
Infect. Immun.
67
1999
6002
6007
[PubMed]
Vordermeier
H.M.
,
Venkataprasad
N.
,
Harris
D.P.
,
Ivanyi
J.
Increase of tuberculous infection in the organs of B cell-deficient mice
Clin. Exp. Immunol.
106
1996
312
316
[PubMed]
Johnson
C.M.
,
Cooper
A.M.
,
Frank
A.A.
,
Bonorino
C.B.
,
Wysoki
L.J.
,
Orme
I.M.
Mycobacterium tuberculosis aerogenic challenge infections in B cell-deficient mice
Tubercle Lung Dis.
78
1997
257
261

Abbreviations used in this paper: NOS, nitric oxide synthase; RNIs, reactive nitrogen intermediates; RPA, ribonuclease protection assay.

C.A. Scanga and V.P. Mohan contributed equally to this study.

Address correspondence to John Chan, Dept. of Medicine and Microbiology and Immunology, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, NY 10461. Phone: 718-430-2678; Fax: 718-652-0536; E-mail: jchan@aecom.yu.edu