Selectin Ligand-Independent Priming and Maintenance of T Cell Immunity during Airborne Tuberculosis¹

Immunity to tuberculosis depends on the timely induction of specific, IFN-γ-producing T cells in the lymph nodes and their efficient interaction with mycobacteria-harboring macrophages at the site of mycobacterial implantation in the lung (1). This necessitates the focal accumulation of primed T cells and activated monocytes within a highly structured lesion, the pulmonary granuloma (2, 3).

Constant recruitment of mononuclear cells into the granuloma is required to contain infection. Even after initial healing, the chronic latent state of infection during which mycobacteria persist in a viable, but quiescent state, can be “reactivated” whether local or systemic immunodeficiencies interrupt the continuous supply of freshly activated effector T cells and macrophages (4). Recrudescence is the most common type of disease in industrialized nations and frequently occurs with anti-inflammatory pharmacotherapy. For example, the introduction of novel TNF-targeted biologicals (e.g., infliximab, a neutralizing anti-TNF Ab) for therapy of rheumatoid arthritis or inflammatory bowel disease has been accompanied by a significant incidence of cases of reactivation tuberculosis (5, 6).

Another potential target for anti-inflammatory pharmacotherapy is the interaction of leukocytes with the vascular wall via specialized selectin-selectin ligand interactions (7). E- and P-selectins on vascular endothelium initiate and sustain leukocyte rolling by engaging specific glycoprotein counterreceptors on granulocytes, monocytes, and activated T cells (8, 9). L-selectin mediates adhesive interactions between lymphocytes and glycoprotein counterreceptors on high endothelial venules (HEV)⁴ in secondary lymphoid organs, required for lymphocyte homing, and in sites of chronic inflammation (10, 11). The selectin counterreceptors become functional only after specific α-1,3-fucosylation (12). Studies in gene-deficient mice demonstrated that the α1,3-fucosyltransferase (FucT) VII plays a prominent role, and FucT-IV a supportive role in regulating T cell trafficking to lymph nodes by L-selectin ligand formation on HEV, and that lymphocyte homing was almost completely abrogated in doubly deficient mice (13). In addition, FucT-VII is critically required for Th1 and Tc1 recruitment in a model of contact hypersensitivity in the skin (14), thereby, controlling an essential component of the efferent phase of T cell immune responses.

Most studies addressing the consequences of FucT deficiency have thus far focused on acute inflammatory responses involving granulocytes (12, 15, 16). As a result, the relative importance of FucT-mediated selectin ligand modification in chronic inflammatory processes, which predominantly involve monocytes and T cells, has remained essentially unexplored.

We, therefore, infected mice with a single or double deficiency for FucT-IV and FucT-VII with Mycobacterium tuberculosis by aerosol to answer the following questions: (1) To what extent does decreased lymphocyte homing to regional lymph nodes affect priming of lymphocyte responses to M. tuberculosis? (2) Does the lack of functional P- and E-selectin ligands interfere with efficient lymphocyte recruitment and activation to M. tuberculosis-infected lungs? (3) Is antimycobacterial effector immunity differentially impaired in the absence of FucT-IV, FucT-VII, or both?

C57BL/6 wild-type and FucT-IV-, FucT-VII-, FucT-IV/-VII-deficient (−/−) mice (12, 14) on a C57BL/6 genetic background were raised under specific pathogen-free conditions at the University of Michigan (Ann Arbor, MI). All mice used were between 8 and 16 wk old. In any given experiment, mice were matched for age and sex. For infection experiments, mice were maintained under barrier conditions in the BSL 3 facility at the Research Center Borstel (Borstel, Germany) in individually ventilated cages. All experiments performed were in accordance with the German Animal Protection Law and were approved by the Animal Research Ethics Board of the Ministry of Environment, Nature Protection and Agriculture (Kiel, Germany).

M. tuberculosis (H37Rv) was grown in Middlebrook 7H9 broth (Difco) supplemented with 10% Middlebrook OADC enrichment medium (Invitrogen Life Technologies), 0.5% glycerol, and 0.05% Tween 80. Midlog phase cultures were harvested, aliquoted, and frozen at −80°C. After thawing, viable cell counts were determined by plating serial dilutions of the cultures on Middlebrook 7H10 agar plates supplemented with 10% OADC followed by incubation at 37°C. All experiments were performed in the BSL 3 Laboratories at the Research Center Borstel.

Before infection of experimental animals, stock solutions of M. tuberculosis were diluted in sterile, distilled water and pulmonary infection was performed using an inhalation exposure system (Glas-Col). To infect mice with a low dose of ∼100 CFU/lung, animals were exposed for 40 min to an aerosol generated by nebulizing ∼5.5 ml of a suspension containing 10⁷ live bacteria. Inoculum size was checked 24 h after infection by determining the bacterial load in undiluted homogenates of the entire lung of infected mice.

Bacterial loads in lung, liver, and spleen were evaluated at different time points after infection with M. tuberculosis to follow the course of infection. Organs from sacrificed animals were removed aseptically, weighed, and homogenized. Ten-fold serial dilutions of organ homogenates were plated in duplicates onto Middlebrook 7H10 agar plates containing 10% OADC and incubated at 37°C for 19–21 days. Colonies on plates were enumerated and results are expressed as log₁₀ CFU per organ. One lung lobe, a piece of liver, and spleen per mouse were fixed in 4% PBS-buffered Formalin, set in paraffin blocks, and sectioned (2–3 μm). Histology was performed using standard protocols for H&E staining (17). For immunohistochemical analysis of lymphocytes and lymphocyte proliferation, tissue sections were prepared and stained with anti-mouse Abs to CD45R/B220 (clone RA3-6B2; BD Biosciences), CD3 (clone CD3-12; Serotec), and Ki-67 (polyclonal rabbit anti-mouse antiserum; a gift from J. Gerdes, Research Center Borstel). Stainings were developed as previously described (17). For immunohistochemical detection of chemokine production in lungs from infected mice, tissue sections were incubated overnight at 4°C with primary goat Abs to mouse CXCL13 (R&D Systems) or CCL21 (R&D Systems), followed by incubation with biotin-SP-AffiniPure mouse anti-goat IgG (Jackson ImmunoResearch Laboratories). After developing with alkaline phosphatase coupled to streptavidin (Jackson ImmunoResearch Laboratories) and fuchsin substrate (DakoCytomation), tissue sections were counterstained with Mayer’s hematoxylin, mounted, and submitted to microscopic analysis.

For Ag-specific restimulation and flow cytometric analyses, single-cell suspensions of mediastinal lymph nodes and lungs were prepared from M. tuberculosis-infected mice at different time points. Lymph node cells were isolated by straining through a metal sieve. After depletion of erythrocytes, cells were resuspended in complete IMDM (Invitrogen Life Technologies) supplemented with 10% FCS (Invitrogen Life Technologies), 0.05 mM 2-ME (Sigma-Aldrich), and penicillin and streptomycin (100 U/ml and 100 μg/ml, respectively; Invitrogen Life Technologies), counted, and used for additional experiments. For preparation of single-cell suspensions from lungs, mice were anesthetized and injected i.p. with 150 U heparin (Ratiopharm). Lungs were perfused through the right ventricle with warm PBS. Once lungs appeared white, they were removed and sectioned. Dissected lung tissue was then incubated in collagenase A (0.7 mg/ml; Roche) and DNase (30 μg/ml; Sigma-Aldrich) at 37°C for 2 h. Digested lung tissue was gently disrupted by subsequent passage through a 100-μm pore size nylon cell strainer. Recovered lung cells were counted, diluted in IMDM, and used for additional experiments.

To asses the capacity of T cells to bind to selectins, cells were stained with P-selectin IgM and E-selectin-IgM chimeras as previously described(12).

For flow cytometric analysis of surface markers, cells were washed and incubated with a mixture containing anti-FcγRIII/II mAb (clone 2.4G2) and mouse and rat serum to block nonspecific binding to Fc receptors. Cells were then incubated in consecutive steps for 20 min with optimal concentrations of the following Abs: CD4-allophycocyanin, CD8-allophycocyanin, CD3-PerCP, CD44-FITC, and CD62 ligand (CD62L)-PE (all from BD Biosciences). Fluorescence intensity was analyzed on a FACSCalibur (BD Biosciences) gating on lymphocytes identified by the forward-scatter/side-scatter profile.

For detection of intracellular IFN-γ, an intracellular cytokine staining kit was used (BD Biosciences). Briefly, single-cell suspensions were prepared at 21 and 42 days after infection and 2 × 10⁶ cells were stimulated with plate-bound anti-CD3/CD28 mAb (clone 2C11 and clone 37/51 at 10 μg/ml, respectively) for 4 h in the presence of GolgiPlug (BD Biosciences). Nonspecific Ab binding was blocked by incubation with a mixture containing anti-FcγRIII/II mAb (clone 2.4G2) and mouse and rat serum. Cells were washed and incubated with optimal concentrations of anti-CD4-FITC (BD Biosciences). After staining, cells were fixed and permeabilized with Cytofix/Cytoperm (BD Biosciences), and intracellularly accumulated IFN-γ was stained with a PE-labeled anti-IFN-γ mAb (BD Biosciences). Fluorescence intensity of IFN-γ-producing CD4⁺ T cells was analyzed gating on CD4⁺ lymphocytes.

At 28 days after infection, mice were challenged with a s.c. injection of 10 μg purified protein derivative (PPD; Statens Serum Institute) of M. tuberculosis in 50 μl of PBS in the right and 50 μl of PBS in the left footpad. Swelling in both footpads was measured after 24 h using a Mitutoyo micrometer caliper (Brütsch), and the difference was taken as the amount of Ag-specific DTH. The PPD preparation did not induce swelling in noninfected animals.

For measuring Ag-specific production of IFN-γ by lymphocytes in lymph nodes, single-cell suspensions of mediastinal lymph nodes were prepared in IMDM from mice 20 days after infection with M. tuberculosis. For Ag-specific restimulation, 4 × 10⁵ lymph node cells were incubated with 2 × 10⁵ peritoneal macrophages that were pulsed with 25 μg/ml short-term culture filtrate (ST-CF) from M. tuberculosis (a gift from P. Andersen, Statens Serum Institute, Copenhagen, Denmark) in IMDM. Resident peritoneal macrophages were obtained 1 day before the experiment by peritoneal lavage of uninfected C57BL/6 mice and incubated overnight in 96-well flat-bottom microplates (Nunc) in complete IMDM. After 72 h of restimulation, supernatants were collected and frozen at −80°C until production of IFN-γ was quantified by ELISA. To determine IFN-γ production after Ag-specific restimulation, supernatants were analyzed in 3-fold serial dilutions using a paired sandwich ELISA (BD Biosciences). After incubation with HRP coupled to avidin and developing with tetramethylbenzidine substrate reagent, the absorbance was read on a microplate reader (Sunrise; Tecan). Using a test wavelength of 450 nm and a reference wavelength of 630 nm, samples were compared with appropriate recombinant cytokine standards. The detection limit was 5 pg/ml.

At different time points after aerosol infection with M. tuberculosis, weighed lung samples were homogenized in 5 ml of 4 M guanidinium-isothiocyanate buffer and total RNA was extracted using an RNA-Isolation kit (Roche). cDNA was obtained using murine moloney leukemia virus reverse transcriptase (Invitrogen Life Technologies) and oligo(dT) (12–18 mer; Sigma-Aldrich) as a primer. Quantitative PCR was performed on a Light Cycler (Roche) as previously described (18). Data were analyzed with the “Fit Points” and “Standard Curve Method” using β₂-microglubulin (β2m) as housekeeping gene to calculate the level of gene expression normalized for β2m expression. The following primer sets were used: β2m: sense 5′-TCA CCG GCT TGT ATG CTA TC-3′, antisense 5′-CAG TGT GAG CCA GGA TAT AG-3′; NOS2: sense 5′-AGC TCC TCC CAG GAC CAC AC-3′, antisense 5′-ACG CTG AGT ACC TCA TTG GC-3′; LRG-47: sense 5′-AGC CGC GAA GAC AAT AAC TG-3′, antisense 5′-CAT TTC CGA TAA GGC TTG G-3′; CXCL13: sense 5′-AAC GCT GCT TCT CCT CCT G-3′, antisense 5′-ATG GGC TTC CAG AAT ACC G-3′; CCL19: sense 5′-TGT GGC CTG CCT CAG ATT AT-3′, antisense 5′-AGT CTT CCG CAT CAT TAG CAC-3′; and CCL21: sense 5′-TCC AAG GGC TGC AAG AGA-3′, antisense 5′-TGA AGT TCG TGG GGG ATC T-3′.

Quantifiable data are expressed as the means of individual determinations and SDs. ANOVA was performed using the Dunnett multiple comparison test defining different error probabilities between C57BL/6 and all mutant strains (∗, p ≤ 0.05; ∗∗, p ≤ 0.01). The unpaired Student’s t test was used to compare results obtained in FucT-VII^−/− and FucT-IV^−/−/-VII^−/− mice defining different error probabilities (+, p ≤ 0.05; ++, p ≤ 0.01; +++, p ≤ 0.001).

A prerequisite for Ag-specific T cell activation and the development of a protective T cell response is the presence of a sufficient T cell repertoire in the lymph node draining the challenge site. Studies in FucT-deficient mice demonstrated that FucT-VII plays a prominent role and FucT-IV a supportive role in regulating trafficking of naive T cells to peripheral and mesenterial lymph nodes by L-selectin ligand formation on HEV, and that lymphocyte homing was severely reduced in uninfected doubly deficient mice (13). To examine how pulmonary infection affects lymph node cellularity, wild-type and FucT-deficient mice were challenged with M. tuberculosis by aerosol and the number and activation status of T cells in the draining lymph nodes was assessed by flow cytometry during the course of infection.

Before M. tuberculosis infection, ∼5 × 10⁶ to 6.5 × 10⁶ cells could be found in mediastinal lymph nodes from C57BL/6 and FucT-IV^−/− mice (Fig. 1,a). In contrast, the amount of mediastinal lymph node cells was significantly reduced in FucT-VII^−/− and FucT-IV^−/−/FucT-VII^−/− mice to ∼2 × 10⁶ cells. After infection with M. tuberculosis, the total number of lymphocytes present in the mediastinal lymph nodes from wild-type and FucT-IV^−/− mice increased 4- to 5-fold in the first 6 wk. In contrast, mice with a single deficiency in FucT-VII or a combined deficiency in FucT-IV and FucT-VII had lymphocyte counts only marginally higher than baseline at the onset of infection (Fig. 1,a). In FucT-VII^−/− and, in a more pronounced fashion, in FucT-IV^−/−/FucT-VII^−/− mice, both CD4⁺ and CD8⁺ T cell subsets were severely reduced in numbers compared with wild-type mice (Fig. 1 b).

When T cell subsets were analyzed for the presence of activation markers, such as high levels of CD44 and low levels of CD62L, it became apparent that double-deficient mice had a significantly higher percentage of activated CD4⁺ and CD8⁺ T cell populations in mediastinal nodes than FucT-VII^−/− mice. Whereas 70% of CD4⁺ T cells in the mediastinal lymph nodes from C57BL/6 and FucT-IV^−/− mice displayed a naive phenotype (CD44^low/CD62L^high) on day 42, only 40% of the CD4⁺ T cell population in lymph nodes from FucT-VII^−/− and doubly deficient mice were CD44^low/CD62L^high (data not shown). However, the absolute number of activated CD4⁺ and CD8⁺ T cells was reduced in lymph nodes from FucT-deficient mice (Fig. 1,c). Three weeks after infection, the accumulation of activated CD4⁺ and CD8⁺ T cells in lymph nodes from FucT-IV^−/− and FucT-VII^−/− mice amounted to 50–70% of the numbers found in wild-type mice. However, the total number of activated T cells in lymph nodes of FucT-IV^−/−/FucT-VII^−/− mice was further reduced to 10% of the number present in wild-type mice (Fig. 1,c). Whereas 42 days after infection the amount of activated CD4⁺ and CD8⁺ T cells in lymph nodes from FucT-IV^−/− and FucT-VII^−/− mice was comparable to C57BL/6 mice, the amount of activated T cells remained significantly reduced in lymph nodes from FucT-IV^−/−/FucT-VII^−/− mice (Fig. 1 c).

Together, these data confirm a prominent function of FucT-VII and a supportive role for FucT-IV in the homeostatic regulation of lymphocyte homing to mediastinal lymph nodes. In addition, the increase of lymphocyte counts that occurs either by proliferation or additional recruitment during the course of M. tuberculosis infection is highly dependent on the presence of FucT-VII and, to a lesser extent, on the presence of FucT-IV.

Priming of naive T cells in the draining lymph nodes is a necessary event to expand Ag-specific, protective CD4⁺ effector T cells that can migrate to sites of infection. These cells are the main source of IFN-γ known to be indispensable for macrophage activation and inhibition of bacterial growth.

To estimate the capacity of lymph node cells to mount Ag-specific immune responses, cell suspensions were prepared from mediastinal lymph nodes on day 19 after infection with M. tuberculosis and restimulated in vitro with Ag-pulsed peritoneal macrophages. After incubation with ST-CF-pulsed macrophages, lymph node cells from uninfected animals produce negligible amounts of IFN-γ (data not shown). Although unstimulated lymph node cells did not produce appreciable levels of IFN-γ, Ag-restimulated cells from lymph nodes of infected C57BL/6 and FucT-IV^−/− mice secreted comparably high levels of IFN-γ (Fig. 2,a). In contrast, lymph node cells from FucT-VII^−/− and to a greater extent from FucT-IV^−/−/FucT-VII^−/− mice produced significantly reduced amounts of IFN-γ (Fig. 2 a). Thus, FucT-VII governs priming events in draining lymph nodes that affect Ag-specific responses.

To further quantitate the capacity of FucT-deficient mice to develop differentiated Th1 cells during pulmonary M. tuberculosis infection, IFN-γ production by CD4⁺ T cells in mediastinal lymph nodes was determined, at the single cell level by intracellular staining after polyclonal restimulation with plate-bound anti-CD3/CD28 gating on CD4⁺ cells. At 20 days after infection, within the CD4⁺ T cell compartment ∼70% fewer CD4⁺ T cells from doubly deficient mice produced IFN-γ than did C57BL/6 mice, whereas FucT-IV^−/− mice had only 30% fewer IFN-γ-producing CD4⁺ cells compared with wild-type mice. IFNγ-secreting T cells were also substantially reduced (by 65%) in the sole absence of FucT-VII (Fig. 2 b).

In conclusion, although activated T cells are present in mediastinal lymph nodes from FucT-VII^−/− and FucT-IV^−/−/FucT-VII^−/− mice, they are strongly impaired in secreting IFN-γ in response to polyclonal and Ag-specific stimuli.

Next, we examined the general capacity of infected mice to mount specific cell-mediated immune responses by measuring a DTH reaction to soluble Ag applied in the footpad. In contrast to C57BL/6 and FucT-IV^−/− mice, FucT-VII^−/− and FucT-IV^−/−/FucT-VII^−/− mice displayed a significantly reduced footpad swelling after injection of PPD at 28 days after infection (Fig. 3). Whereas a single FucT-VII deficiency resulted in a reduction of DTH by 25%, footpad swelling was diminished by 60% in doubly deficient mice. Since there was no significantly different DTH reaction between M. tuberculosis-infected FucT-VII^−/− and FucT-IV^−/−/-VII^−/− mice, this observation identifies a prominent role for FucT-VII in DTH to PPD.

Migration of cells into inflamed tissue is thought to require an initial binding of effector and memory lymphocytes to E-and P-selectin. However, histopathological analysis of lungs performed on days 42 after aerosol infection with M. tuberculosis revealed no obvious differences in the kinetics of granuloma formation or the cellular composition of inflammatory infiltrations between the four mouse strains examined (Fig. 4). A more detailed quantitative and qualitative analysis was performed by flow cytometry of lung digests.

Flow cytometric analysis of T cells purified from the lungs of M. tuberculosis-infected FucT-VII^−/− and doubly deficient mice revealed a significantly impaired binding to E- and P-selectin, as observed elsewhere (12, 13) (Fig. 5 a).

Three weeks after infection, the total number of cells (Fig. 5,b), as well as the amount of activated CD4⁺ and CD8⁺ T cells (Fig. 5,c), was slightly, but significantly, decreased in the lungs of FucT-deficient mice. However, 6 wk after infection leukocyte recruitment was similar in the lungs of C57BL/6 and FucT-deficient mice (Fig. 5,b). Flow cytometric analysis revealed that both CD4⁺ and CD8⁺ lymphocyte populations totaled >5 × 10⁶ cells on day 42 after infection, regardless of the mouse strain examined (Fig. 5,b). Similarly, staining for activation markers showed that FucT-deficient mice had the same numbers of CD44^high, CD62L^lowCD4⁺, and CD8⁺ cells in infected lungs as wild-type mice 42 days after infection (Fig. 5 c).

Because development of effector CD4⁺ T cells in draining lymph nodes was disturbed, yet recruitment of activated T cells into inflamed lungs was only slightly delayed in FucT-deficient mice after M. tuberculosis infection, we next wished to determine the IFN-γ-producing capacity of T cells recruited to the lungs during infection at the single cell level. Twenty-one days after M. tuberculosis infection, CD4⁺ cells from the lungs of FucT-IV^−/−/FucT-VII^−/− mice showed significantly decreased IFN-γ production following polyclonal stimulation with anti-CD3/anti-CD28 when compared with CD4⁺ T cells of lungs from C57BL/6 mice (Fig. 6,a). A single FucT-IV deficiency did not result in impaired CD4⁺ T cell-mediated IFN-γ production, but mice singly deficient for FucT-VII had ∼40% fewer IFN-γ-producing T cells in infected lungs than wild-type mice (Fig. 6 a).

IFNγ-mediated activation of resident macrophages and newly recruited monocytes that have engulfed mycobacteria reprograms these cells for mycobacteriocidal activities (19), such as induction of NO synthase 2 (NOS2) or the small 47-kDa GTPase, LRG-47, which is involved in endosome trafficking (20, 21).

In a more comprehensive analysis of gene expression in inflamed lungs, mRNA levels of IFN-γ and of the IFN-γ-induced genes NOS2 and LRG-47 were quantified on days 21 and 42 after aerosol infection with M. tuberculosis. Although expression of all three genes was found to be reduced in FucT-IV^−/−/FucT-VII^−/− mice by ∼70%, IFN-γ, NOS2, and LRG-47 expression in FucT-VII^−/− mice was decreased by ∼40% at day 21 postinfection (Fig. 6,b). In contrast, FucT-IV^−/− mice displayed no impairment in IFN-γ, NOS2, and LRG-47 expression (Fig. 6,b). Six weeks after infection, the expression of IFN-γ and IFN-γ-induced genes in all mouse strains examined was of a comparable magnitude (Fig. 6 b). Accordingly, as revealed by immunohistology, no difference in the expression of NOS2 on the protein level was found in lung sections of either mouse strain 42 days after aerosol infection with M. tuberculosis (data not shown). Together, impairment of IFN-γ and IFN-γ-induced gene expression in lung tissue was only transient and reversible in the absence of FucT-IV and FucT-VII.

To determine whether FucT-VII^−/− and FucT-IV^−/−/FucT-VII^−/− mice were compromised in controlling M. tuberculosis infection, animals were sacrificed at different time points after infection and the bacterial loads were quantified in the lung, liver, and spleen. In line with the expression of IFN-γ and IFN-γ-induced genes in lung tissue from all groups of mice, no significant differences in bacterial loads were found in any organ during the course of infection with 75 CFU of M. tuberculosis (Fig. 7,a). CFU in lungs of all mouse strains reached a plateau level of 10⁶/lung at 3 wk postinfection (Fig. 7,a). Increasing the infection dose to 150 mycobacteria/mouse did not diminish the capacity of FucT-deficient mice to contain the infectious burden, and no significantly different bacterial loads could be detected in any organ at any time point examined (Fig. 7 b)

These results suggest that a mechanism exists which, in the absence of FucT-VII or in the absence of both FucT-IV and FucT-VII, counteracts the severely impaired priming of CD4⁺ T cells in draining lymph nodes.

Recently, mice lacking spleen, lymph nodes, and Peyer’s patches were demonstrated to generate unexpectedly robust primary B and T cell responses to influenza virus infection (22). In these mice, areas of induced BALT have distinct B cell follicles and T cell areas, and specific immune responses and clonal expansion seem to be initiated within induced BALT. To analyze the formation of such neolymphoid follicles during infection with M. tuberculosis and to determine whether they arise in a FucT-dependent way, we performed immunohistochemical analysis of lung sections from all infected groups of mice. Distinct lymphoid follicles, containing B cells and T cells, were readily apparent in wild-type, FucT-IV^−/−, FucT-VII^−/−, and FucT-IV^−/−/FucT-VII^−/− mice (Fig. 8). In these follicle-like structures, B cells were present as well-organized aggregates, whereas adjacent T cells were more diffusely interspersed in the lymphoid cell accumulation. In addition, proliferation of some of these lymphocytes was detected by staining with mAbs specific for the proliferation marker Ki-67, indicating ongoing immune responses within these granuloma-associated neolymphoid structures. Proliferation appeared particularly marked in some follicles from doubly deficient mice.

A prerequisite for the homing of naive and central memory T cells, B lymphocytes, and eventually the formation of secondary lymphoid organs is the expression of the tissue-organizing chemokines CXCL13, CCL21, and CCL19 (23). To determine whether these chemokines may have contributed to the development of neolymphoid follicles in lungs from M. tuberculosis-infected mice, we performed immunohistochemical and gene expression analysis of CXLC13, CCL21, and CCL19 in the lungs of wild-type, FucT-IV^−/−, FucT-VII^−/−, and FucT-IV^−/−/FucT-VII^−/− mice (Fig. 9). Immunohistochemistry revealed that CXCL13 and CCL21 were indeed expressed not only perivascularly throughout the lung, but also in close association with granulomatous lesions in all strains of mice examined at day 42 after M. tuberculosis infection (Fig. 9,a). In contrast to a constitutive expression of all three chemokines in secondary lymphoid organs of uninfected mice (data not shown), mRNAs for CXCL13, CCL21, and CCL19 were not expressed in appreciable levels in the lungs before infection with M. tuberculosis (Fig. 9 b). Pulmonary infection induced gene expression of CXCL13, CCL21, and CCL19 in the lungs of infected mice that was over time more pronounced in tissue taken from FucT-IV/-VII^−/− mice. Together, our data revealed that after M. tuberculosis infection priming, activation, and clonal expansion of effector lymphocytes may proceed in neolymphoid follicles in the lungs of FucT-deficient mice in which lymphocyte priming in draining lymph nodes is impaired.

We conclude from the present study that fucosyltransferase-dependent selectin-selectin ligand interactions are not necessary for the generation and expression of cell-mediated immunity at a level sufficient to contain pulmonary M. tuberculosis infection. Our findings corroborate and extend data obtained in naive and Ag-challenged FucT-deficient mice that FucT-VII- and, to lesser degree, FucT-IV-mediated fucosylation of ligands for L-, E-, and P-selectin substantially contribute to the homing and expansion of lymphocytes in draining lymph nodes, and thereby exert a significant effect on priming T cells for effector functions. However, in stark contrast to selectin ligand-dependent T cell-mediated DTH reactions in the skin, trafficking of effector lymphocytes to the M. tuberculosis-infected lung is completely independent of FucT-IV and FucT-VII function. As a net consequence of decreased priming yet efficient recruitment of T cells, expression of IFN-γ-dependent effector mechanisms in the lungs are merely delayed in FucT-VII-deficient and in doubly deficient mice but are deployed at a sufficient magnitude to control M. tuberculosis replication within well-structured granulomatous lesions. Therefore, despite prominent defects in mediastinal lymph node cellularity before and during infection, FucT-VII and FucT-IV/VII doubly deficient mice are surprisingly well equipped to mount effector immune responses to tuberculosis.

Following pulmonary infection, M. tuberculosis is transported to the draining lymph nodes where priming and clonal expansion of Ag-specific effector T cells occurs (24, 25). Extending published data in naive and skin-sensitized FucT-deficient mice (14), we report in this study that the numbers of CD4⁺ and CD8⁺ T cells in the mediastinal lymph nodes of FucT-VII^−/− and FucT-IV^−/−/FucT-VII^−/− mice remain at low levels throughout the first 6 wk after pulmonary infection with M. tuberculosis. It is therefore evident that even under conditions of chronic inflammatory stress induced by an intracellular infectious agent, no major alternative mechanism of cell recruitment can be mustered that would efficiently compensate for the lymphocyte homing defect imposed by FucT deficiency. We found that, at 3 and 6 wk postinfection, Ag-specific as well as polyclonal stimulation of T lymphocytes obtained from mediastinal lymph nodes resulted in significantly less IFN-γ production in FucT-IV^−/−/FucT-VII^−/− and, to a lesser extent, FucT-VII^−/− mice. It is conceivable that dendritic cell migration from the lung into the regional lymph node is impaired in these FucT-deficient mice and that reduced Ag presentation in the lymph node might therefore account for the observed deficiency in effector cell differentiation. However, in a model of contact hypersensitivity, it was previously observed that the number and function of dendritic cells required for cutaneous Ag processing and presentation is unaltered in all FucT-deficient strains used in our study (14). It, therefore, appears more likely that the prominent numerical deficiency in naive T cells in the lymph nodes in FucT-IV^−/−/FucT-VII^−/− and, to a lesser extent, FucT-VII^−/− mice accounts for the reduced capacity of these mice to expand and differentiate Ag-specific T cells.

After M. tuberculosis infection, recruitment of Ag-specific CD4⁺ T cells to the site of infection in the lung is necessary for protective cell-mediated immune responses, such as granuloma formation and macrophage activation (1, 26). As a surrogate marker for some of the key events involved in the expression of this type of protective immunity, the DTH reaction in the skin is often investigated (27). The absence of E- and P-selectin ligands on Th1 and Tc1 lymphocytes was previously demonstrated to annul trafficking of these T cells to inflamed cutaneous sites, as demonstrated by the loss of cutaneous hypersensitivity against dinitrofluorobenzene (14). Ag-specific cutaneous DTH to an infectious agent, lymphocytic choriomeningitis virus (LCMV), also depends on FucT-VII (28). However, since DTH reactions were only delayed after LCMV infection of FucT-VII^−/− mice (29), showing a retarded extravasation of CD4⁺ and CD8⁺ T cells into the skin, FucT-IV has been inferred to direct a low degree of residual E- and P-selectin ligand expression by activated T cells (14), similar to its function in neutrophils (13). In this study, we report that the DTH response in M. tuberculosis-infected mice is strictly dependent on FucT-VII, with little or no contribution of FucT-IV activity.

CD4⁺ and CD8⁺ T cells obtained from the lungs of M. tuberculosis-infected FucT-VII^−/− and FucT-IV^−/−/FucT-VII^−/− mice failed to significantly bind E- and P-selectin-IgM chimeras. Surprisingly, however, the recruitment of activated T cells into the lungs of infected mice was not impaired in the absence of FucT-VII or in doubly deficient mice. This is at apparent odds with previous observations that E- and P-selectins and their ligands contribute prominently, although not essentially, to lymphocyte recruitment to the lung (30, 31, 32). Our own data obtained in a mycobacterial infection model are, however, reminiscent of recent data in the LCMV model of infection where extravasation of CD4⁺ and CD8⁺ T cells into visceral organs, such as the ovaries or the brain, was not defective in FucT-VII^−/− mice, although the DTH reaction in the skin to LCMV was severely reduced.

FucT-dependent selectins are believed to initiate the earliest step of leukocyte recruitment into inflammatory sites mediating leukocyte rolling along the activated endothelium (33, 34). In contrast, firm attachment and subsequent transendothelial migration are mainly mediated by ICAMs and VCAMs (35). Therefore, recruitment of T cells in the absence of FucTs must be mediated by other selectin-independent mechanisms that initiate leukocyte rolling. It may be envisioned that physical retardation of leukocytes in the small capillaries of the lung is responsible for rolling along the activated endothelium (36, 37). This has been convincingly shown for neutrophils whose initial sequestration in the lung occurs through mediator-induced changes in the biomechanical properties of neutrophils, particularly a stiffening of these cells and a reduction in their ability to deform and pass through the pulmonary capillaries (37). More importantly, LFA-1-mediated retention of activated T cells has been described as a possible selectin ligand-independent mechanism of T cell sequestration in the lungs (38). Most of our observations in FucT-VII^−/− mice are also consistent with the immune defects described in E-selectin and P-selectin doubly deficient mice, which show enhanced susceptibility to cutaneous bacterial infection (39, 40). These mice exhibit a reduced migration of CD4⁺ T cells into skin (34, 41, 42, 43), but a normal leukocyte migration into inflamed liver and CD4⁺ T cell migration into the brain in an experimental autoimmune encephalomyelitis model (40, 44). Few studies, however, also report on impaired leukocyte extravasation into peripheral organs (39, 45). In endothelial E- and P-selectin-deficient mice that have been intratracheally primed and challenged with SRBCs, a similar, only delayed CD4⁺ T cell infiltration into the lungs was observed, suggesting that selectins do not seem to play an important role in T cell homing to the lungs (46). With respect to mycobacterial infections, at least P-selectin has been shown to be dispensable for leukocyte recruitment and granuloma formation (47).

In FucT-IV^−/−/FucT-VII^−/− mice, the kinetics and quality of granuloma formation was unaltered and NOS2 expression within granuloma macrophages was only slightly delayed compared with wild-type mice. Thus, although the absolute levels of IFN-γ elicited by polyclonal stimulation of lung T cells was significantly reduced in FucT-IV^−/−/FucT-VII^−/− mice, this defect was not sufficient to significantly compromise antibacterial protection. As a consequence, M. tuberculosis replication was held at a plateau level in the lungs of infected mice, regardless of the fucosylation status of selectin ligands. Previous studies using the vaccinia and LCMV model had demonstrated that elimination of these viruses from ovaries and brain was normal in FucT-VII^−/− mice. Our study is the first to address the relevance of differentially fucosylated selectin ligands in a model of chronic bacterial infection in the lung.

Our data indicate that priming of Ag-specific T cells in draining lymph nodes may be less critical than previously thought for the induction and maintenance of an effective protective cell-mediated immune response to M. tuberculosis infection in the lung. In typical respiratory immune responses, Ag-specific lymphocytes first appear in the lymph nodes that drain the respiratory tract and subsequently appear in the lungs. However, we found that M. tuberculosis infection also results in the formation of neolymphoid follicles in the lungs, which may have the potential to prime M. tuberculosis-specific T cells that subsequently function to contain mycobacterial growth despite defective priming in conventional lymphoid organs.

Our data are consistent with published reports showing that T cells can be primed directly in the lung (48) and a previous report demonstrating antiviral immunity in mice carrying the paucity of lymph node T cells mutation (49), suggesting that T cell dendritic cell encounters in these mice occur primarily in the marginal zone of the spleen or even in nonlymphatic organs like the lung. In this respect, BALT is occasionally found in the lungs of mice and humans after mycobacterial infection (50), and lymphoid follicles were demonstrated at the periphery of granulomatous lesions (51).

Homing of T lymphocytes to the T zone within secondary lymphoid tissues are regulated by the CCR7 ligands, CCL19 and CCL21, whereas homing of T cells to B cell follicles are mediated by CXCL13, the ligand for CXCR5 (23). We found that all three chemokines are induced in the lungs and expressed, to a significant extent, within granulomatous lesions after infection with M. tuberculosis, indicating that the formation of neolymphoid follicles is a likely consequence of pulmonary tuberculosis infection. In support of this hypothesis, flow cytometric analysis revealed that naive and central memory T cells, expressing CCR7, and follicular helper T cells, expressing CXCR5, are recruited to the site of infection after pulmonary M. tuberculosis infection (data not shown). Therefore, it appears possible that, subsequent to dissemination of M. tuberculosis from the lung via regional lymph nodes, continuous recruitment of naive T cells and priming of T cells also takes place in organized neolymphoid areas within the lungs of FucT-deficient mice in a largely unimpaired fashion. This would also explain why, after an initial inferior capacity of FucT-IV^−/−/FucT-VII^−/− mice to mount IFN-γ responses in the lung, T cell effector functions become fully restored over time.

In summary, our results indicate that FucT-IV and FucT-VII are neither required for inducing nor maintaining T cell effector responses at a level necessary to control pulmonary M. tuberculosis infection. Thus, FucT-IV- and FucT-VII-independent mechanisms are sufficient for efficient expression of protective cell-mediated immune responses during tuberculosis. In terms of anti-inflammatory pharmacotherapy, this finding may indicate that targeting selectin-selectin ligand interactions during chronic inflammatory conditions may avoid some of the side effects noted with the use of TNF-targeted biologicals, notably reactivation tuberculosis.

We thank Alexandra Rausch, Johanna Volz, Susanne Metken, and Manfred Richter for excellent technical assistance; Ilka Monath, Sven Mohr, and Claus Möller for organizing the BSL3 animal facility; and Robert Kelly for taking care of the mice.

The authors have no financial conflict of interest.