Characterization of a B220+ Lymphoid Cell Subpopulation with Immune Modulatory Functions in Nasal-Associated Lymphoid Tissues
http://www.100md.com
免疫学杂志 2005年第1期
Abstract
Complex mechanisms operate on mucosal tissues to regulate immune responsiveness and tolerance. When the lymphocyte subpopulations from murine nasal-associated lymphoid tissues (NALT) were characterized, we observed an accumulation of B220lowCD3lowCD4–CD8–CD19–c-Kit+ cells. TCR transgenic mice and athymic mice were used for monitoring T cell lineage and the presence of extrathymic T cell precursors. The majority of cells from NALT exhibited a T cell precursor phenotype (CD4–CD8–CD19–c-Kit+). Fas-independent apoptosis was their main mechanism of cell death. We also demonstrated that B220lowCD4–CD8–CD19– cells from NALT exhibited the potential to down-regulate the activation of mature T cells. However, the innate immunity receptor TLR2 was also highly expressed by this cell subpopulation. Moreover, nasal stimulation with a TLR2/6 agonist resulted in a partial activation of the double-negative cells. These results suggest that the immune responses in NALT may be in part modulated by a cell subpopulation that maintains a tolerogenic milieu by its proapoptotic status and suppressive activity, which can be reverted through stimulation of a TLR signaling cascade.
Introduction
The bronchial-associated and nasal-associated lymphoid tissues (NALT)2 provide an extensive specific and nonspecific defense system to the respiratory tract. The nasal mucosa, due to its strategic position, is the first contact site with inhaled Ags. In humans, the lymphoid structures associated with the oronasal mucosa are known as the Waldeyer’s ring and consist of a number of tonsils, lymphoid bands, and the adenoids. In the murine nasal passages, tissues equivalent to the Waldeyer’s ring have been found bilaterally at the entrance of the pharyngeal duct, at the level of the ecto- and endoturbinates (1). In addition, there is an abundance of intraepithelial and lamina propria lymphocytes, which are diffusely disseminated in the mucosa of the nasal passages (2, 3, 4). NALT and the nasal mucosa drain directly to the superficial and posterior cervical lymph nodes (cLN). Although an inhaled particulate impacting on the mucus layer of the nasal mucosa can be rapidly cleared by ciliary motion, it could also be selectively delivered to the organized NALT structures, thereby triggering immune responses (5).
One of the primary roles of NALT is to defend against respiratory infections and inhaled toxicants. Previous studies have proven that the nasopharyngeal-associated lymphoreticular tissue is a major mucosal inductive site (5). Thus, new vaccination strategies based on nasal application have been designed to prevent diseases caused by infectious agents with a mucosal portal of entry. In addition to its role in the defense of the upper and lower respiratory tracts, the nasal lymphoid system cooperates with the systemic immune system and affects immune reactions at distant mucosal sites, such as the urogenital tract and the gut (6, 7). However, the introduction of soluble Ags into the nasal mucosa predominantly leads to immunological nonresponsiveness (8, 9). Therefore, there are extremely complex mechanisms operating at the level of the nasal immune system to regulate highly specialized processes, such as immune reactivity and mucosal tolerance.
Despite its central role in mucosal immunity, little is known about the nasal immune system. The phenotype and functions of the different cellular subpopulations are not fully characterized. However, given the intensity of antigenic exposure at this site, it is clear that local T cell activation events require strict control to maintain tissue homeostasis. The generation of an adequate base of knowledge would be critical to understand immune responsiveness under normal or pathologic conditions (e.g., resistance and susceptibility to infection, allergy, anergy, and autoimmunity) as well as for the design of new immune therapeutic and/or prophylactic interventions.
The mucosal immune system has developed separately from the general systemic immune system. As a major consequence, only immune responses initiated at mucosal inductive sites can lead to effective immunity in mucosal tissues. Previous studies have also shown the uniqueness of immune cells from mucosal tissues with respect to those from secondary lymphoid organs (10). The mucosal immune system can be divided into inductive and effector compartments. The NALT is a typical inductive region displaying under physiological conditions strong tolerance (non responsiveness) to exogenous Ags (11, 12, 13). Several mechanisms for peripheral tolerance to self-Ags have been suggested, including T cell anergy, T cell deletion, immunological ignorance, and receptor modulation. Similar mechanisms appear to regulate tolerance to exogenous Ags (14, 15). Some studies suggested that the majority of T cells released from the respiratory epithelium are held in an inactive growth phase or unresponsive state (16, 17). In the small intestine and the genital tract, a population of CD3lowCD45R/B220low cells has been identified that has potential regulatory functions (18, 19, 20). However, there are no data available concerning the NALT. In contrast, the comparison between Peyer’s patches and NALT demonstrated various differences, such as the fact that NALT contains a major population of naive T cells (1, 21). An important question to be addressed is whether differences between NALT, bronchial-associated lymphoid tissues, and Peyer’s patches reflect differences in the mechanisms involved in either the maintenance of local tolerance or the generation of immune responses. The major aim of the present study was to characterize lymphoid cells from the NALT. We observed an accumulation of cells with a T cell precursor phenotype (B220lowCD3lowCD4–CD8–c-Kit+), which expressed the innate immunity receptor TLR2 and seem to exhibit the potential to down-regulate the activation of mature T cells.
Materials and Methods
Mice
Six-week-old female BALB/c, BALB/c nu/nu (H-2d), and MRL lpr/lpr mice were purchased from Harlan-Winkelmann; the heterologous TCR-hemagglutinin (TCR-HA) mice (22) backcrossed onto BALB/c mice were bred at our institute. The animals were maintained in individually ventilated cages in a specific pathogen-free animal facility according to local and European Community guidelines. Transgene expression was determined by DNA analysis by PCR using TCR-V-specific primers and was confirmed by FACS analysis of PBL obtained by tail bleedings (23).
NALT and cell preparation
NALT cells were recovered for flow cytometric analysis and proliferation assays as previously described (5, 21, 24). Briefly, after exsanguination of the mice and isolation of cLN and spleen, the head and fore teeth were cut off. The facial skin, lower jaw, and cheek muscles were removed, and NALT were exposed by carefully peeling away the palate. Individual NALT were removed by microsurgical tweezers under a stereoscopic microscope and placed in ice-cold culture media RPMI 1640/10% FCS. The lymphoid cells were dissociated by teasing with syringe plunger through a 100-μm pore size nylon mesh. RBC were lysed in ammonium chloride lysis (ACK) buffer, and cellular suspensions were filtered through a 40-μm pore size mesh. Cell suspensions from cLN and spleen were obtained following the same process.
Intranasal administration of macrophage-activating lipopeptide of 2 kDa (MALP-2)
BALB/c mice (n = 6) received 20 μl of synthetic MALP-2 (0.5 μg) in PBS or PBS only by the intranasal route (25). Animals were killed 3 or 16 h after treatment. Organs were isolated, and cell suspensions were prepared for further analysis.
Flow cytometric assays
Cellular suspensions were counted using a cell counter and size analyzer Z2 (Beckman Coulter), and lymphocyte concentration was determined using a cell size range of 4.0–8.6 μm. For flow cytometric analysis, 4 x 105 cells/single staining were used. FcRs were blocked at 4°C during 45 min using an Ab specific for CD16/CD32 (BD Pharmingen) diluted in PBS/1% FCS. Cells were then stained at 4°C for 30 min with specific mAbs conjugated with FITC, PE, PerCP, allophycocyanin, or biotin from BD Pharmingen, with the exception of the polyclonal rabbit anti-murine TLR2 Ab that was provided by C. J. Kirschning (University of Munich, Munich, Germany). mAb 6.5 (anti-TCR-HA) was purified from a hybridoma supernatant and used in an FITC-labeled form. Irrelevant labeled Abs were used as isotype controls in all experiments. After washing, cells were examined with a FACSCalibur cytometer and analyzed using CellQuest Pro software (BD Biosciences). Alternatively, a mixture of Abs was used to allow analysis of restricted cellular subpopulations. Specific labeling of dead cells and apoptotic cells was performed using the vital dye 7-aminoactinomycin D (7-AAD; BD Pharmingen) and FITC-labeled annexin V (Molecular Probes), according to the instructions of the manufacturers.
Isolation of cell nuclei and Western blot analysis
Nuclei of cells from NALT or cLN isolated from two mice were purified by ultracentrifugation in sucrose, as previously described (26). All buffers contained 1 mM PMSF, and all steps were conducted on ice. Briefly, cell pellets were resuspended in buffer (0.25 M sucrose, 50 mM Tris-HCl (pH 7.4), and 5 mM MgSO4) and centrifuged at 1,000 x g for 15 min. The pellets containing the crude nuclear preparation were then resuspended in 2.2 M sucrose, 50 mM Tris-HCl (pH 7.4), and 5 mM MgSO4 and layered over a 5-ml cushion of the same buffer before ultracentrifugation at 72,000 x g for 60 min. The pellets containing the purified nuclei and the complexed proteins were resuspended in 200 μl of 10 mM Tris (pH 7.5) buffer supplemented with 0.25 M sucrose, 10 mM NaCl, 3 mM MgCl2, and 40% glycerol and stored at –20°C until electrophoretic analysis. Twenty micrograms of total protein from each sample was separated into a 10% SDS-PAGE gel and electroblotted onto a polyvinylidene difluoride membrane. After 1-h blocking in 5% milk, immunoblotting was performed using rabbit polyclonal Ab (0.2 μg/ml) against poly(ADP-ribose) polymerase (PARP-1; Santa Cruz Biotechnology) at room temperature for 1 h, followed by incubation with protein A-peroxidase at 1 μg/ml (Amersham Biosciences). Staining was visualized by chemiluminescent technique using an ECL kit (Amersham Biosciences) according to the manufacturer’s instructions.
Ag-specific cell proliferation and cytokine measurements
Cells from NALT and spleens from TCR-HA transgenic mice were isolated and prepared as described above. Half the cells from NALT were used for depletion studies in which CD19+ cells were depleted using epoxy beads coated with anti-CD19 Ab (Dynal Biotech), whereas B220+ cells were depleted using B220 Dynabeads ready to use, according to the instructions of the manufacturers. The other half of the cell suspensions was used as the control and received the same treatment, with the exception of bead supplementation. NALT cells were seeded at an estimated concentration of 1.5 x 105 cells/well and cultured in the presence of the specific HA peptide (5 μg/ml) or Con A (5 μg/ml) for 4 days. Peritoneal macrophages isolated from BALB/c mice were irradiated, and 5 x 104 cells were added to each well as an additional source of APCs, to compensate for any potential loss during the depletion process. Cells were grown in RPMI 1640 supplemented with 10% FCS, 100 U/ml penicillin, 50 mg/ml streptomycin, 5 x 10–5 M 2-ME, and 1 mM L-glutamine (Invitrogen Life Technologies) and maintained at 37°C in a humidified 5% CO2 atmosphere. T cell proliferation was measured after pulsing with 1 μCi of [3H]thymidine for the last 16 h. Each condition was tested in quadruplicate. The results were expressed as an index of stimulation corresponding to the ratio of [3H]thymidine uptake in cpm of stimulated cells to the nonstimulated cells. Cells were also stimulated in the same test format, and supernatant fluids were collected after 96 h and stored at –70°C for additional analysis. IL-2, IL-4, IL-5, IL-6, IFN-, and TNF- cytokines were measured using the mouse Th1/Th2 and inflammation cytokine kits (cytometric bead array kit; BD Pharmingen), according to the manufacturer’s instructions. Single samples of the same proliferation assay were tested undiluted in two separate experiments.
Transwell coculture experiments
To evaluate the putative role of soluble factors, lymph node (LN) cells isolated from TCR-HA mice were cultivated separately from NALT cells in a two-chamber system. Transwell inserts with 0.45 μm pore size (Costar) were used to cocultivate irradiated (900 gray) NALT cells in the lower chamber and LN cells in presence of the HA peptide (5 μg/ml) in the upper chamber. A 0.45-μm pore size does not allow direct cell-cell contact, whereas it allows the exchange of peptide and soluble factors. Irradiated LN cells were given to the lower chamber and acted as controls for the irradiated cells from NALT. Additionally, CD4+ depletion of NALT cells was performed using anti-CD4-coated epoxy beads, prepared following the manufacturer’s instructions (Dynal Biotech). Cells were cultured for 72 h in 12-well plates (106 LN cells in the upper chamber and 107 NALT cells or control cells in the lower chamber). The proliferation of T cells from the upper chamber was measured after pulsing with [3H]thymidine for the last 16 h (4 μCi/well).
Results
Two main lymphoid cellular subpopulations are present in murine NALT
We performed four-color flow cytometric analysis of cells from NALT in comparison with those from cLN. A first subpopulation of NALT cells with a large size (FSChigh) was identified, which matched the majority of cells present on cLN in BALB/c mice (Fig. 1A). These cells mainly corresponded to conventional mature B cells (49.0%) and T cells (16.0%), according to cell surface staining for CD19, CD4, and CD8 (Fig. 1B and data not shown). In addition, a second subpopulation of smaller cells (FSClow) was identified in NALT from all tested mouse strains (BALB/c, BALB/c nu/nu, and MRL lpr/lpr; see Fig. 1A). This FSClow cell subpopulation of NALT reached a high proportion of the total cells (41.6% in the BALB/c mice; Fig. 1A). Mature CD19+ or CD4+ or CD8+ cells represented only a small fraction of this FSClow subpopulation from NALT. In contrast, FSClow and FSChigh cells from cLN showed similar phenotypic profiles (Fig. 1B), with a majority of mature B and T cells. In contrast, FSChigh cells from NALT exhibited a complex phenotypic profile, which encompasses the profiles from both conventional cLN cells and FSClow cells from NALT (Fig. 1B).
FIGURE 1. Phenotypic analysis of cellular subpopulations in cLN and NALT. Freshly prepared cellular suspensions obtained from 7- to 9-wk-old BALB/c, BALB/c nu/nu, and MRL lpr/lpr mice were analyzed by flow cytometry. A, Histogram overlays of the cell sizes of cLN (line) and NALT () cells are shown for each mouse strain. The percentages of FSClow (m1) and FSChigh (m2) were calculated considering the cells comprised in the FSC cell range m3. B, Characterization of FSClow and FSChigh cells from cLN and NALT obtained from BALB/c mice. Expression of markers for mature B cells (CD19) and mature T cells (CD4) as well as the pan-B cell (B220) and pan-T cell (CD3) markers are shown () compared with isotype controls (). A representative result from three independent experiments is shown.
The B220lowCD3low subpopulation constitutes the majority of cells from NALT
In addition to their small size, the larger fraction of FSClow cells and some of the FSChigh cells from NALT displayed a very peculiar phenotype, characterized by low surface expression of CD45R/B220 (B220low; Fig. 1B). A thorough characterization demonstrated that these cells corresponded to B220lowCD3lowCD4–CD8–CD19– cells. To strengthen the idea that these CD3lowB220low cells from NALT were indeed of T cell origin, we evaluated whether an epitope-specific TCR could be found in this subpopulation. To this end, we used transgenic mice generated on a BALB/c background, which are characterized by the presence of a high percentage of cells expressing a TCR specific for the peptide 110–120 from the HA of the influenza virus A/PR8/34 (22, 23). The surface expression of HA-specific TCR was characterized on cells isolated from NALT, cLN, and spleen of TCR-HA transgenic mice (Fig. 2). Similarly to what was observed in spleen, 25% of the CD3+CD4+ cells from NALT highly expressed the specific V6.5 chain (Fig. 2A), demonstrating that there is no deviation of TCR expression in NALT. In contrast, only 7% of CD3+CD4– cells from spleen expressed the 6.5 transgenic TCR, whereas a significant proportion (20.7%) of CD3+CD4– cells from NALT expressed it (Fig. 2A). As illustrated in Fig. 2B, CD3+CD4– cells expressing the 6.5 TCR corresponded to CD8+ cells in the case of cLN or spleen, whereas they were mainly B220low T cells in the case of NALT. In both cases, the level of expression of the TCR was lower on CD3+CD4– than on CD3+CD4+ cells (mean fluorescence intensity (MFI), 39.1 vs 67.6 for spleen cells, and 24.0 vs 89.4 for NALT cells; Fig. 2A).
FIGURE 2. T cell lineage analyzed by epitope-specific TCR expression. The expression of the 6.5 TCR on the surface of cells obtained from spleens, cLN, and NALT isolated from TCR-HA transgenic mice was examined by four-color flow cytometry. A, Cells (FSClow and FSChigh together) were gated on the basis of staining with PE-conjugated anti-CD4, and the expression of 6.5 TCR on CD4+ and CD4– cells was evaluated using an FITC-conjugated Ab. The percentage of gated cells staining positively for the 6.5 TCR compared with the isotype control is shown in the histogram for spleen and NALT cells. B, Characterization of the expression of 6.5 TCR on CD4– cells. The 6.5 TCR is mainly expressed on CD8+ T cells from cLN and B220low cells from NALT. Dot plots showed that in both cases CD8+ and B220low T cells had only low expression of 6.5 TCR compared with bright expression on CD4+ cells.
Fas-independent apoptosis is a main cell death mechanism of B220lowCD4–CD8– cells from NALT
Because apoptotic T cells also display a characteristic FSClow phenotype (27), we evaluated whether B220low cells from NALT corresponded to preapoptotic or apoptotic T cells. To this end, apoptosis and cell death were measured by flow cytometry using the annexin V and 7-AAD dyes. As shown in Fig. 3A, NALT cells contained 49 and 34% apoptotic and dead cells, respectively. When mature B cells (CD19+) and dead cells (7-AAD+) were excluded, 26% of the remaining cells were B220+, the majority of which were also annexin V+ (Fig. 3B). In contrast, cells from the cLN contained only 10% apoptotic cells, from which only 0.7% corresponded to B220lowCD19– cells. These results were confirmed by analyzing caspase-dependent cleavage of PARP (Fig. 3C). Extracts from NALT cells revealed, in addition to intact PARP, the presence of a major 89-kDa fragment that is specifically generated during apoptosis (28). In contrast, this fragment was undetectable or remained a minor component when cellular extracts from cLN were tested. Similar results of annexin V binding were obtained with cells isolated from BALB/c mice between 6 and 24 wk of age (data not shown). To approach the mechanisms involved in the induction of apoptosis in NALT, we performed FACS analysis of annexin V binding with cells of NALT from MRL lpr/lpr (Fas-deficient) mice. We found a similar percentage of apoptotic cells with respect to those in BALB/c mice (data not shown). This demonstrated that a Fas-independent apoptotic process occurs in NALT.
FIGURE 3. Quantification of apoptosis and cell death of NALT cells. A flow cytometric analysis of freshly isolated cells from NALT and cLN from 8-wk-old BALB/c mice was performed gating on FSClow and FSChigh cells. A, Dot plots show representative double stainings with annexin V-FITC and 7-AAD of NALT and cLN cells. The percentages of living (double-negative), early apoptotic (annexin V+) and necrotic (double-positive) cells are indicated. B, Apoptotic cells expressing the B220 marker were quantified after exclusion of CD19+ and 7-AAD+ cells. Results are representative of two independent experiments. C, Cleavage of PARP in nuclei preparations purified from NALT and cLN cells were analyzed by Western blot (20 μg of total protein/lane). In vivo PARP proteolytic cleavage associated with apoptosis leads to 89- and a 24-kDa fragments. PARP-1 cleavage by caspases causes its inactivation for the DNA repair process. The molecular masses of the main products are indicated by arrows. Intact PARP has a molecular mass of 113 kDa, and necrosis induces the expression of a 55-kDa fragment. One microgram of recombinant PARP-1 protein (Alexis) and a PARP fraction isolated from nuclei of the murine hepatoma cell (HEPA 1–6) were used as controls. Results are representative of two independent experiments.
CD3lowB220lowCD4–CD8– cells from NALT resemble T cell precursors
Because double-negative cells matched immature T cells, we evaluated the possibility that B220low NALT cells corresponded to a precursor cell type. A staining specific for the lineage precursor marker c-Kit (CD117) was performed. The results showed that a large proportion of B220low cells from NALT expressed the c-Kit marker (Fig. 4A). After exclusion of dead cells (7-AAD+) and mature CD19+ B cells (Fig. 4A), 3.7 and 46% of the remaining cells were found to be c-Kithigh (MFI, 612) and c-Kitlow (MFI, 53.8) with respect to the isotype control (MFI, 13.6), respectively. These results suggested that cells with a precursor c-Kit-positive phenotype were accumulating in NALT. To unravel the origin of the double-negative T cells residing in nasal tissues, we evaluated the link between their presence and the T cell thymic differentiation capacity. Using BALB/c nu/nu mice, in which T cell maturation is impaired as result of a thymic deficiency, we found a similar CD3lowB220low population in NALT (Fig. 4B). Thus, these cells seem neither to be related to the thymic maturation pathway nor derived from mature T cells.
FIGURE 4. Immature, mature, and extrathymic cells from NALT. A, The surface expression of the cell precursor marker c-Kit was analyzed by flow cytometry. A representative dot plot shows the double staining of NALT cells (FSClow and FSChigh together) obtained with B220- and c-Kit-specific Abs. The percentages of cells in the corresponding rectangular regions are indicated. The c-Kit expression on B220+ cells after exclusion of 7-AAD+ and CD19+ cells () is shown in comparison with the isotype control (line). B, Phenotypic analysis of cell subpopulations in NALT and cLN of athymic BALB/c nu/nu mice was performed by flow cytometry gating on both FSClow and FSChigh cells.
CD3lowB220lowCD4–CD8– NALT cells can regulate the activation of mature T cells
A general characteristic of mucosal tissues is their high tolerogenicity. This is particularly important considering the large amount of inhaled Ags. Thus, we investigated which cellular component of NALT might be involved in this regulatory process. CD4+CD25+ T cells can exert immunoregulatory functions; unexpectedly, they were not identified in nasal tissues of naive BALB/c mice. In contrast, 74.3% of the CD4+ T cells from NALT were CD45RB+ (low plus high) vs 65.7% for the CD4+ cells from cLN (Fig. 5A), from which 14.6 and 27.0% were CD45RBlow, respectively. This suggested that NALT contained a higher proportion of naive T cells (CD45RBhigh) and fewer T cells exhibiting an activated, memory, or regulatory phenotype (CD45RBlow). This was also supported by the lack of expression of activation markers, such as CD69 (data not shown). Thus, it seems that potential CD4+CD45RBlow regulatory T cells could constitute only a minor subpopulation within NALT.
FIGURE 5. Immune regulatory function of B220lowCD3+CD4–CD8– cells from NALT. A, Double-stained cells (CD45RB/CD4) isolated from NALT and cLN of 8-wk-old BALB/c mice were analyzed by gating on FSClow and FSChigh cells. Approximately 14.6 and 27.0% of CD4+ cells from NALT and cLN corresponded to CD4+CD45RBlow T cells (as determined by a cut-off above the isotype control and below a MFI of 160), respectively. B, Ag-specific CD4 T cell proliferative responses of total () vs B220-depleted () cell preparations from NALT of TCR-HA mice were measured after in vitro restimulation in the presence of HA peptide 110–120. The stimulation index was determined as the ratio of cpm of cells cultured with peptide (5 μg/ml) or Con A (5 μg/ml) to control cells grown in medium. C, The cytokines secreted into culture supernatants after in vitro restimulation with the HA peptide were measured after 96 h of culture. Results are expressed as mean values; SDs are indicated by vertical lines. D, HA-specific proliferative response of LN cells cocultivated in a Transwell system with either tissue culture medium or irradiated cells (NALT cells, CD4+-depleted NALT cells, or LN cells). Results are expressed as both stimulation indexes and percent inhibition compared with the values of LN control cells cocultivated without additional cells in the lower chamber. Results are representative of two independent experiments. *, p < 0.1; **, p < 0.05.
Additional studies were performed to evaluate the regulatory potential of CD3lowB220lowCD4–CD8– cells from NALT. To this end, we assessed the proliferative capacity of cells isolated from NALT of TCR-HA transgenic mice (22, 23). Significant proliferative responses from NALT cells were observed in the presence of both the specific HA peptide and the mitogenic stimulator Con A (Fig. 5B). However, when NALT cells were depleted from the B220+ subpopulation, we observed a significant increment in HA-specific proliferation (Fig. 5B). No differences were detected in response to Con A (Fig. 5B) or after depletion of CD19+ cells (data not shown). Then we analyzed the cytokines secreted by cells stimulated with the HA peptide. We observed an increment (i.e., mean > control + 2 SD) in the production of TNF-, IFN-, and IL-2, when B220+ cells were depleted, whereas the levels of secreted IL-10, IL-6, IL-5, and IL-4 were not significantly increased (Fig. 5C). These results are in agreement with the dominant Th1 cytokine response pattern of peptide-stimulated spleen cells from TCR-HA animals (Fig. 5C) previously observed (23, 29).
In an attempt to unravel the underlying mechanisms of the observed suppressive activity from NALT cells, Transwell coculture studies were performed. The results showed that the suppression induced by NALT cells is supported by soluble factors able to neutralize up to 70% of the proliferative stimulation (Fig. 5D). Surprisingly, the complete inhibitory effect of NALT cells required the presence of CD4+ T cells, because the removal of these cells resulted in a decreased (40%) suppression (Fig. 5D). Thus, it seems that NALT cells exert their suppressive activity via soluble factors in cooperation with CD4+ T cells.
CD3lowB220lowCD4–CD8– cells from NALT are responsive to danger signals
NALT are not only a tolerogenic site, but are also an important inductive site in the context of appropriate stimulation, such as those delivered by TLR (5, 30). Thus, we evaluated the expression of TLR on the surface of cells from NALT. As shown in Fig. 6A, more cells expressing high levels of TLR2 were observed in NALT than in cLN (63.5 vs 5.8%). For the large majority, these TLR2+ cells from NALT were also B220low positive. This suggests that TLR2 signaling could play an important role in the homeostasis and reactivity of B220low cells as well as in the induction of immune responses in nasal tissues.
FIGURE 6. TLR2 expression and sensitization of B220lowCD4–CD8– from NALT. A, Surface expression of TLR2 on cells isolated from NALT and cLN of 7- to 9-wk-old BALB/c mice were analyzed by flow cytometry. Double-stained TLR2/B220 cells were analyzed by gating on FSClow and FSChigh cells. B, Freshly isolated cells from NALT of mice that received MALP-2 (0.5 μg) or PBS (nonactivated) by the intranasal route 16 h before organ processing were stained with Abs specific for the T cell markers CD4, CTLA-4, and CD62L. Dot plots correspond to FACS analysis gating on both FSClow and FSChigh. The percentages of positive cells in the corresponding quadrants are indicated.
Despite its crucial role in immune activation, TLR2 could also signal for apoptosis (31, 32). Thus, we evaluated activation and apoptosis after stimulation with the TLR-2 ligand MALP-2 (25, 33). The number of apoptotic cells was not increased in NALT 3 or 16 h after intranasal administration of 0.5 μg of MALP-2 (data not shown). In contrast, we observed an increment in the activation status of CD4– T cells, as reflected by the down-regulation of CTLA-4 and CD62L (0.6 vs 6.7% and 7.3 vs 12.8%, respectively; Fig. 6B).
Discussion
Atypical cells have been previously described in association with NALT. Hiroi et al. (34) reported the presence of CD3+CD4–CD8– cells, whereas Wu et al. (5) described a large fraction of small cells. However, the phenotypic and functional characterization of these unconventional cells has remained elusive to date. Our analysis showed that a large fraction of NALT cells (45% of the total cells) corresponded to CD3lowB220lowCD4–CD8– cells. In TCR-HA mice, the transgenic TCR is expressed not only on CD4+CD8– and CD4–CD8+ mature T cells, but also on CD4–CD8– immature T cells (22). By using these TCR transgenic mice, we clearly demonstrated that the double-negative cells observed in NALT belong to the T cell lineage. Interestingly, in accordance to their small size, the majority of the NALT B220lowCD4–CD8– cells were apoptotic. It has been reported that the B cell marker B220 is up-regulated on activated T cell blasts before entering apoptosis (27, 35, 36). Considering that nasal tissues are an overexposed environment, it can be hypothesized that cells from NALT undergo apoptosis after activation by exogenous Ags or superantigen, as previously demonstrated in other experimental systems (27, 35, 37, 38). However, the expression levels of both CD45RB and activation markers suggests that the majority of cells from NALT corresponded to naive T cells (CD45RBhigh). Moreover, a similar cell phenotype was found in NALT from athymic nu/nu mice. Thus, the B220low T cells in NALT do not appear to be related to the presence of mature T cells. Therefore, it seems unlikely that a down-regulation of surface markers as a by-product of apoptosis accounts for the assigned phenotype.
We favor the hypothesis that the expression of the B220 molecule on double-negative cells from NALT is related to a precursor cell phenotype. This is in agreement with the surface expression of the precursor marker c-Kit observed in this study, and it is also supported by the fact that lymphoid progenitor cells have been recently identified in human tonsils (39). B220+CD19– pluripotent precursor cells are also found in bone marrow and fetal liver (40, 41). Although it was accepted that commitment to the T lineage is an intrathymic event, prothymocytes are also found in murine fetal blood (42). In addition, the intestine can occasionally act as a site of positive selection for thymus-derived T cells (43). Concerning NALT, it remains to be elucidated whether and why progenitor or immature T cells migrate or accumulate in the nasal tissues and undergo apoptosis. Studies focused on elucidation of the molecular events leading to initiation and organization of NALT showed that, to date, only deficiency in the regulator of cell differentiation Id2 completely abolished NALT development (24, 44). This deficiency can be overcome by adoptive transfer of fetal liver cells, suggesting a primal origin for NALT initiator cells.
Our hypothesis is that the double-negative B220low cells might be involved in formation of the reticular network maintaining the homeostasis of mature lymphocytes from NALT. We also postulate that this cellular subpopulation might represent an important immunomodulatory system for NALT. Several observations support this hypothesis. First, no or very few classical regulatory T cells, such as CD4+CD25+ and CD4+CD45RBlow, have been found in nasal tissues (45, 46, 47, 48). Second, previous work suggested that B220+CD4–CD8– cells can affect the capacity of activated T cells to proliferate or produce cytokines (18, 36, 49). In this study we demonstrated that B220low double-negative cells in NALT exhibited the capacity to down-regulate mature T cell activation. This finding is in agreement with the fact that apoptosis generates a tolerogenic milieu rather than proinflammatory reactions (50). However, we also showed that CD3lowB220lowCD4–CD8– cells from NALT express high levels of the innate immunity receptor TLR2, and that in vivo stimulation of the TLR2/6 signaling cascade in nasal tissues leads to an activated phenotype of these cells. In conclusion, CD3lowB220lowCD4–CD8– cells constitute the main cellular subpopulation from NALT. It is conceivable that nasal tissues have acquired a regulatory system in which the same cellular pool plays different functions, such as contributing toward homeostasis maintenance, immune suppression (e.g., via production of soluble factors and/or its proapoptotic nature), and immune activation (e.g., via its high expression of innate immunity receptors).
Acknowledgments
We are particularly grateful to Carsten J. Kirschning for the polyclonal rabbit anti-murine TLR2 Ab, to Peter F. Mühlradt for providing us with MALP-2, and to Urte J?ger for her outstanding technical help.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 Address correspondence and reprint requests to Dr. Carlos A. Guzmán, Vaccine Research Group, Division of Microbiology, GBF-German Research Center for Biotechnology, Mascheroder Weg 1, D-38124 Braunschweig, Germany. E-mail address: cag@gbf.de
2 Abbreviations used in this paper: NALT, nasal-associated lymphoid tissue; 7-AAD, 7-aminoactinomycin D; cLN, cervical lymph node; FSC, forward scatter; HA, hemagglutinin; LN, lymph node; MFI, mean fluorescence intensity; PARP, poly(ADP-ribose) polymerase; MALP-2, macrophage-activating lipopeptide of 2 kDa.
Received for publication May 10, 2004. Accepted for publication November 16, 2004.
References
Kuper, C. F., P. J. Koornstra, D. M. Hameleers, J. Biewenga, B. J. Spit, A. M. Duijvestijn, P. J. van Breda Vriesman, T. Sminia. 1992. The role of nasopharyngeal lymphoid tissue. Immunol. Today 13:219.
Wiley, J. A., R. J. Hogan, D. L. Woodland, A. G. Harmsen. 2001. Antigen-specific CD8+ T cells persist in the upper respiratory tract following influenza virus infection. J. Immunol. 167:3293.
Asanuma, H., A. H. Thompson, T. Iwasaki, Y. Sato, Y. Inaba, C. Aizawa, T. Kurata, S. Tamura. 1997. Isolation and characterization of mouse nasal-associated lymphoid tissue. J. Immunol. Methods 202:123.
Liang, B., L. Hyland, S. Hou. 2001. Nasal-associated lymphoid tissue is a site of long-term virus-specific antibody production following respiratory virus infection of mice. J. Virol. 75:5416.
Wu, H. Y., H. H. Nguyen, M. W. Russell. 1997. Nasal lymphoid tissue (NALT) as a mucosal immune inductive site. Scand. J. Immunol. 46:506.(Faiza Rharbaoui, Dunja Br)
Complex mechanisms operate on mucosal tissues to regulate immune responsiveness and tolerance. When the lymphocyte subpopulations from murine nasal-associated lymphoid tissues (NALT) were characterized, we observed an accumulation of B220lowCD3lowCD4–CD8–CD19–c-Kit+ cells. TCR transgenic mice and athymic mice were used for monitoring T cell lineage and the presence of extrathymic T cell precursors. The majority of cells from NALT exhibited a T cell precursor phenotype (CD4–CD8–CD19–c-Kit+). Fas-independent apoptosis was their main mechanism of cell death. We also demonstrated that B220lowCD4–CD8–CD19– cells from NALT exhibited the potential to down-regulate the activation of mature T cells. However, the innate immunity receptor TLR2 was also highly expressed by this cell subpopulation. Moreover, nasal stimulation with a TLR2/6 agonist resulted in a partial activation of the double-negative cells. These results suggest that the immune responses in NALT may be in part modulated by a cell subpopulation that maintains a tolerogenic milieu by its proapoptotic status and suppressive activity, which can be reverted through stimulation of a TLR signaling cascade.
Introduction
The bronchial-associated and nasal-associated lymphoid tissues (NALT)2 provide an extensive specific and nonspecific defense system to the respiratory tract. The nasal mucosa, due to its strategic position, is the first contact site with inhaled Ags. In humans, the lymphoid structures associated with the oronasal mucosa are known as the Waldeyer’s ring and consist of a number of tonsils, lymphoid bands, and the adenoids. In the murine nasal passages, tissues equivalent to the Waldeyer’s ring have been found bilaterally at the entrance of the pharyngeal duct, at the level of the ecto- and endoturbinates (1). In addition, there is an abundance of intraepithelial and lamina propria lymphocytes, which are diffusely disseminated in the mucosa of the nasal passages (2, 3, 4). NALT and the nasal mucosa drain directly to the superficial and posterior cervical lymph nodes (cLN). Although an inhaled particulate impacting on the mucus layer of the nasal mucosa can be rapidly cleared by ciliary motion, it could also be selectively delivered to the organized NALT structures, thereby triggering immune responses (5).
One of the primary roles of NALT is to defend against respiratory infections and inhaled toxicants. Previous studies have proven that the nasopharyngeal-associated lymphoreticular tissue is a major mucosal inductive site (5). Thus, new vaccination strategies based on nasal application have been designed to prevent diseases caused by infectious agents with a mucosal portal of entry. In addition to its role in the defense of the upper and lower respiratory tracts, the nasal lymphoid system cooperates with the systemic immune system and affects immune reactions at distant mucosal sites, such as the urogenital tract and the gut (6, 7). However, the introduction of soluble Ags into the nasal mucosa predominantly leads to immunological nonresponsiveness (8, 9). Therefore, there are extremely complex mechanisms operating at the level of the nasal immune system to regulate highly specialized processes, such as immune reactivity and mucosal tolerance.
Despite its central role in mucosal immunity, little is known about the nasal immune system. The phenotype and functions of the different cellular subpopulations are not fully characterized. However, given the intensity of antigenic exposure at this site, it is clear that local T cell activation events require strict control to maintain tissue homeostasis. The generation of an adequate base of knowledge would be critical to understand immune responsiveness under normal or pathologic conditions (e.g., resistance and susceptibility to infection, allergy, anergy, and autoimmunity) as well as for the design of new immune therapeutic and/or prophylactic interventions.
The mucosal immune system has developed separately from the general systemic immune system. As a major consequence, only immune responses initiated at mucosal inductive sites can lead to effective immunity in mucosal tissues. Previous studies have also shown the uniqueness of immune cells from mucosal tissues with respect to those from secondary lymphoid organs (10). The mucosal immune system can be divided into inductive and effector compartments. The NALT is a typical inductive region displaying under physiological conditions strong tolerance (non responsiveness) to exogenous Ags (11, 12, 13). Several mechanisms for peripheral tolerance to self-Ags have been suggested, including T cell anergy, T cell deletion, immunological ignorance, and receptor modulation. Similar mechanisms appear to regulate tolerance to exogenous Ags (14, 15). Some studies suggested that the majority of T cells released from the respiratory epithelium are held in an inactive growth phase or unresponsive state (16, 17). In the small intestine and the genital tract, a population of CD3lowCD45R/B220low cells has been identified that has potential regulatory functions (18, 19, 20). However, there are no data available concerning the NALT. In contrast, the comparison between Peyer’s patches and NALT demonstrated various differences, such as the fact that NALT contains a major population of naive T cells (1, 21). An important question to be addressed is whether differences between NALT, bronchial-associated lymphoid tissues, and Peyer’s patches reflect differences in the mechanisms involved in either the maintenance of local tolerance or the generation of immune responses. The major aim of the present study was to characterize lymphoid cells from the NALT. We observed an accumulation of cells with a T cell precursor phenotype (B220lowCD3lowCD4–CD8–c-Kit+), which expressed the innate immunity receptor TLR2 and seem to exhibit the potential to down-regulate the activation of mature T cells.
Materials and Methods
Mice
Six-week-old female BALB/c, BALB/c nu/nu (H-2d), and MRL lpr/lpr mice were purchased from Harlan-Winkelmann; the heterologous TCR-hemagglutinin (TCR-HA) mice (22) backcrossed onto BALB/c mice were bred at our institute. The animals were maintained in individually ventilated cages in a specific pathogen-free animal facility according to local and European Community guidelines. Transgene expression was determined by DNA analysis by PCR using TCR-V-specific primers and was confirmed by FACS analysis of PBL obtained by tail bleedings (23).
NALT and cell preparation
NALT cells were recovered for flow cytometric analysis and proliferation assays as previously described (5, 21, 24). Briefly, after exsanguination of the mice and isolation of cLN and spleen, the head and fore teeth were cut off. The facial skin, lower jaw, and cheek muscles were removed, and NALT were exposed by carefully peeling away the palate. Individual NALT were removed by microsurgical tweezers under a stereoscopic microscope and placed in ice-cold culture media RPMI 1640/10% FCS. The lymphoid cells were dissociated by teasing with syringe plunger through a 100-μm pore size nylon mesh. RBC were lysed in ammonium chloride lysis (ACK) buffer, and cellular suspensions were filtered through a 40-μm pore size mesh. Cell suspensions from cLN and spleen were obtained following the same process.
Intranasal administration of macrophage-activating lipopeptide of 2 kDa (MALP-2)
BALB/c mice (n = 6) received 20 μl of synthetic MALP-2 (0.5 μg) in PBS or PBS only by the intranasal route (25). Animals were killed 3 or 16 h after treatment. Organs were isolated, and cell suspensions were prepared for further analysis.
Flow cytometric assays
Cellular suspensions were counted using a cell counter and size analyzer Z2 (Beckman Coulter), and lymphocyte concentration was determined using a cell size range of 4.0–8.6 μm. For flow cytometric analysis, 4 x 105 cells/single staining were used. FcRs were blocked at 4°C during 45 min using an Ab specific for CD16/CD32 (BD Pharmingen) diluted in PBS/1% FCS. Cells were then stained at 4°C for 30 min with specific mAbs conjugated with FITC, PE, PerCP, allophycocyanin, or biotin from BD Pharmingen, with the exception of the polyclonal rabbit anti-murine TLR2 Ab that was provided by C. J. Kirschning (University of Munich, Munich, Germany). mAb 6.5 (anti-TCR-HA) was purified from a hybridoma supernatant and used in an FITC-labeled form. Irrelevant labeled Abs were used as isotype controls in all experiments. After washing, cells were examined with a FACSCalibur cytometer and analyzed using CellQuest Pro software (BD Biosciences). Alternatively, a mixture of Abs was used to allow analysis of restricted cellular subpopulations. Specific labeling of dead cells and apoptotic cells was performed using the vital dye 7-aminoactinomycin D (7-AAD; BD Pharmingen) and FITC-labeled annexin V (Molecular Probes), according to the instructions of the manufacturers.
Isolation of cell nuclei and Western blot analysis
Nuclei of cells from NALT or cLN isolated from two mice were purified by ultracentrifugation in sucrose, as previously described (26). All buffers contained 1 mM PMSF, and all steps were conducted on ice. Briefly, cell pellets were resuspended in buffer (0.25 M sucrose, 50 mM Tris-HCl (pH 7.4), and 5 mM MgSO4) and centrifuged at 1,000 x g for 15 min. The pellets containing the crude nuclear preparation were then resuspended in 2.2 M sucrose, 50 mM Tris-HCl (pH 7.4), and 5 mM MgSO4 and layered over a 5-ml cushion of the same buffer before ultracentrifugation at 72,000 x g for 60 min. The pellets containing the purified nuclei and the complexed proteins were resuspended in 200 μl of 10 mM Tris (pH 7.5) buffer supplemented with 0.25 M sucrose, 10 mM NaCl, 3 mM MgCl2, and 40% glycerol and stored at –20°C until electrophoretic analysis. Twenty micrograms of total protein from each sample was separated into a 10% SDS-PAGE gel and electroblotted onto a polyvinylidene difluoride membrane. After 1-h blocking in 5% milk, immunoblotting was performed using rabbit polyclonal Ab (0.2 μg/ml) against poly(ADP-ribose) polymerase (PARP-1; Santa Cruz Biotechnology) at room temperature for 1 h, followed by incubation with protein A-peroxidase at 1 μg/ml (Amersham Biosciences). Staining was visualized by chemiluminescent technique using an ECL kit (Amersham Biosciences) according to the manufacturer’s instructions.
Ag-specific cell proliferation and cytokine measurements
Cells from NALT and spleens from TCR-HA transgenic mice were isolated and prepared as described above. Half the cells from NALT were used for depletion studies in which CD19+ cells were depleted using epoxy beads coated with anti-CD19 Ab (Dynal Biotech), whereas B220+ cells were depleted using B220 Dynabeads ready to use, according to the instructions of the manufacturers. The other half of the cell suspensions was used as the control and received the same treatment, with the exception of bead supplementation. NALT cells were seeded at an estimated concentration of 1.5 x 105 cells/well and cultured in the presence of the specific HA peptide (5 μg/ml) or Con A (5 μg/ml) for 4 days. Peritoneal macrophages isolated from BALB/c mice were irradiated, and 5 x 104 cells were added to each well as an additional source of APCs, to compensate for any potential loss during the depletion process. Cells were grown in RPMI 1640 supplemented with 10% FCS, 100 U/ml penicillin, 50 mg/ml streptomycin, 5 x 10–5 M 2-ME, and 1 mM L-glutamine (Invitrogen Life Technologies) and maintained at 37°C in a humidified 5% CO2 atmosphere. T cell proliferation was measured after pulsing with 1 μCi of [3H]thymidine for the last 16 h. Each condition was tested in quadruplicate. The results were expressed as an index of stimulation corresponding to the ratio of [3H]thymidine uptake in cpm of stimulated cells to the nonstimulated cells. Cells were also stimulated in the same test format, and supernatant fluids were collected after 96 h and stored at –70°C for additional analysis. IL-2, IL-4, IL-5, IL-6, IFN-, and TNF- cytokines were measured using the mouse Th1/Th2 and inflammation cytokine kits (cytometric bead array kit; BD Pharmingen), according to the manufacturer’s instructions. Single samples of the same proliferation assay were tested undiluted in two separate experiments.
Transwell coculture experiments
To evaluate the putative role of soluble factors, lymph node (LN) cells isolated from TCR-HA mice were cultivated separately from NALT cells in a two-chamber system. Transwell inserts with 0.45 μm pore size (Costar) were used to cocultivate irradiated (900 gray) NALT cells in the lower chamber and LN cells in presence of the HA peptide (5 μg/ml) in the upper chamber. A 0.45-μm pore size does not allow direct cell-cell contact, whereas it allows the exchange of peptide and soluble factors. Irradiated LN cells were given to the lower chamber and acted as controls for the irradiated cells from NALT. Additionally, CD4+ depletion of NALT cells was performed using anti-CD4-coated epoxy beads, prepared following the manufacturer’s instructions (Dynal Biotech). Cells were cultured for 72 h in 12-well plates (106 LN cells in the upper chamber and 107 NALT cells or control cells in the lower chamber). The proliferation of T cells from the upper chamber was measured after pulsing with [3H]thymidine for the last 16 h (4 μCi/well).
Results
Two main lymphoid cellular subpopulations are present in murine NALT
We performed four-color flow cytometric analysis of cells from NALT in comparison with those from cLN. A first subpopulation of NALT cells with a large size (FSChigh) was identified, which matched the majority of cells present on cLN in BALB/c mice (Fig. 1A). These cells mainly corresponded to conventional mature B cells (49.0%) and T cells (16.0%), according to cell surface staining for CD19, CD4, and CD8 (Fig. 1B and data not shown). In addition, a second subpopulation of smaller cells (FSClow) was identified in NALT from all tested mouse strains (BALB/c, BALB/c nu/nu, and MRL lpr/lpr; see Fig. 1A). This FSClow cell subpopulation of NALT reached a high proportion of the total cells (41.6% in the BALB/c mice; Fig. 1A). Mature CD19+ or CD4+ or CD8+ cells represented only a small fraction of this FSClow subpopulation from NALT. In contrast, FSClow and FSChigh cells from cLN showed similar phenotypic profiles (Fig. 1B), with a majority of mature B and T cells. In contrast, FSChigh cells from NALT exhibited a complex phenotypic profile, which encompasses the profiles from both conventional cLN cells and FSClow cells from NALT (Fig. 1B).
FIGURE 1. Phenotypic analysis of cellular subpopulations in cLN and NALT. Freshly prepared cellular suspensions obtained from 7- to 9-wk-old BALB/c, BALB/c nu/nu, and MRL lpr/lpr mice were analyzed by flow cytometry. A, Histogram overlays of the cell sizes of cLN (line) and NALT () cells are shown for each mouse strain. The percentages of FSClow (m1) and FSChigh (m2) were calculated considering the cells comprised in the FSC cell range m3. B, Characterization of FSClow and FSChigh cells from cLN and NALT obtained from BALB/c mice. Expression of markers for mature B cells (CD19) and mature T cells (CD4) as well as the pan-B cell (B220) and pan-T cell (CD3) markers are shown () compared with isotype controls (). A representative result from three independent experiments is shown.
The B220lowCD3low subpopulation constitutes the majority of cells from NALT
In addition to their small size, the larger fraction of FSClow cells and some of the FSChigh cells from NALT displayed a very peculiar phenotype, characterized by low surface expression of CD45R/B220 (B220low; Fig. 1B). A thorough characterization demonstrated that these cells corresponded to B220lowCD3lowCD4–CD8–CD19– cells. To strengthen the idea that these CD3lowB220low cells from NALT were indeed of T cell origin, we evaluated whether an epitope-specific TCR could be found in this subpopulation. To this end, we used transgenic mice generated on a BALB/c background, which are characterized by the presence of a high percentage of cells expressing a TCR specific for the peptide 110–120 from the HA of the influenza virus A/PR8/34 (22, 23). The surface expression of HA-specific TCR was characterized on cells isolated from NALT, cLN, and spleen of TCR-HA transgenic mice (Fig. 2). Similarly to what was observed in spleen, 25% of the CD3+CD4+ cells from NALT highly expressed the specific V6.5 chain (Fig. 2A), demonstrating that there is no deviation of TCR expression in NALT. In contrast, only 7% of CD3+CD4– cells from spleen expressed the 6.5 transgenic TCR, whereas a significant proportion (20.7%) of CD3+CD4– cells from NALT expressed it (Fig. 2A). As illustrated in Fig. 2B, CD3+CD4– cells expressing the 6.5 TCR corresponded to CD8+ cells in the case of cLN or spleen, whereas they were mainly B220low T cells in the case of NALT. In both cases, the level of expression of the TCR was lower on CD3+CD4– than on CD3+CD4+ cells (mean fluorescence intensity (MFI), 39.1 vs 67.6 for spleen cells, and 24.0 vs 89.4 for NALT cells; Fig. 2A).
FIGURE 2. T cell lineage analyzed by epitope-specific TCR expression. The expression of the 6.5 TCR on the surface of cells obtained from spleens, cLN, and NALT isolated from TCR-HA transgenic mice was examined by four-color flow cytometry. A, Cells (FSClow and FSChigh together) were gated on the basis of staining with PE-conjugated anti-CD4, and the expression of 6.5 TCR on CD4+ and CD4– cells was evaluated using an FITC-conjugated Ab. The percentage of gated cells staining positively for the 6.5 TCR compared with the isotype control is shown in the histogram for spleen and NALT cells. B, Characterization of the expression of 6.5 TCR on CD4– cells. The 6.5 TCR is mainly expressed on CD8+ T cells from cLN and B220low cells from NALT. Dot plots showed that in both cases CD8+ and B220low T cells had only low expression of 6.5 TCR compared with bright expression on CD4+ cells.
Fas-independent apoptosis is a main cell death mechanism of B220lowCD4–CD8– cells from NALT
Because apoptotic T cells also display a characteristic FSClow phenotype (27), we evaluated whether B220low cells from NALT corresponded to preapoptotic or apoptotic T cells. To this end, apoptosis and cell death were measured by flow cytometry using the annexin V and 7-AAD dyes. As shown in Fig. 3A, NALT cells contained 49 and 34% apoptotic and dead cells, respectively. When mature B cells (CD19+) and dead cells (7-AAD+) were excluded, 26% of the remaining cells were B220+, the majority of which were also annexin V+ (Fig. 3B). In contrast, cells from the cLN contained only 10% apoptotic cells, from which only 0.7% corresponded to B220lowCD19– cells. These results were confirmed by analyzing caspase-dependent cleavage of PARP (Fig. 3C). Extracts from NALT cells revealed, in addition to intact PARP, the presence of a major 89-kDa fragment that is specifically generated during apoptosis (28). In contrast, this fragment was undetectable or remained a minor component when cellular extracts from cLN were tested. Similar results of annexin V binding were obtained with cells isolated from BALB/c mice between 6 and 24 wk of age (data not shown). To approach the mechanisms involved in the induction of apoptosis in NALT, we performed FACS analysis of annexin V binding with cells of NALT from MRL lpr/lpr (Fas-deficient) mice. We found a similar percentage of apoptotic cells with respect to those in BALB/c mice (data not shown). This demonstrated that a Fas-independent apoptotic process occurs in NALT.
FIGURE 3. Quantification of apoptosis and cell death of NALT cells. A flow cytometric analysis of freshly isolated cells from NALT and cLN from 8-wk-old BALB/c mice was performed gating on FSClow and FSChigh cells. A, Dot plots show representative double stainings with annexin V-FITC and 7-AAD of NALT and cLN cells. The percentages of living (double-negative), early apoptotic (annexin V+) and necrotic (double-positive) cells are indicated. B, Apoptotic cells expressing the B220 marker were quantified after exclusion of CD19+ and 7-AAD+ cells. Results are representative of two independent experiments. C, Cleavage of PARP in nuclei preparations purified from NALT and cLN cells were analyzed by Western blot (20 μg of total protein/lane). In vivo PARP proteolytic cleavage associated with apoptosis leads to 89- and a 24-kDa fragments. PARP-1 cleavage by caspases causes its inactivation for the DNA repair process. The molecular masses of the main products are indicated by arrows. Intact PARP has a molecular mass of 113 kDa, and necrosis induces the expression of a 55-kDa fragment. One microgram of recombinant PARP-1 protein (Alexis) and a PARP fraction isolated from nuclei of the murine hepatoma cell (HEPA 1–6) were used as controls. Results are representative of two independent experiments.
CD3lowB220lowCD4–CD8– cells from NALT resemble T cell precursors
Because double-negative cells matched immature T cells, we evaluated the possibility that B220low NALT cells corresponded to a precursor cell type. A staining specific for the lineage precursor marker c-Kit (CD117) was performed. The results showed that a large proportion of B220low cells from NALT expressed the c-Kit marker (Fig. 4A). After exclusion of dead cells (7-AAD+) and mature CD19+ B cells (Fig. 4A), 3.7 and 46% of the remaining cells were found to be c-Kithigh (MFI, 612) and c-Kitlow (MFI, 53.8) with respect to the isotype control (MFI, 13.6), respectively. These results suggested that cells with a precursor c-Kit-positive phenotype were accumulating in NALT. To unravel the origin of the double-negative T cells residing in nasal tissues, we evaluated the link between their presence and the T cell thymic differentiation capacity. Using BALB/c nu/nu mice, in which T cell maturation is impaired as result of a thymic deficiency, we found a similar CD3lowB220low population in NALT (Fig. 4B). Thus, these cells seem neither to be related to the thymic maturation pathway nor derived from mature T cells.
FIGURE 4. Immature, mature, and extrathymic cells from NALT. A, The surface expression of the cell precursor marker c-Kit was analyzed by flow cytometry. A representative dot plot shows the double staining of NALT cells (FSClow and FSChigh together) obtained with B220- and c-Kit-specific Abs. The percentages of cells in the corresponding rectangular regions are indicated. The c-Kit expression on B220+ cells after exclusion of 7-AAD+ and CD19+ cells () is shown in comparison with the isotype control (line). B, Phenotypic analysis of cell subpopulations in NALT and cLN of athymic BALB/c nu/nu mice was performed by flow cytometry gating on both FSClow and FSChigh cells.
CD3lowB220lowCD4–CD8– NALT cells can regulate the activation of mature T cells
A general characteristic of mucosal tissues is their high tolerogenicity. This is particularly important considering the large amount of inhaled Ags. Thus, we investigated which cellular component of NALT might be involved in this regulatory process. CD4+CD25+ T cells can exert immunoregulatory functions; unexpectedly, they were not identified in nasal tissues of naive BALB/c mice. In contrast, 74.3% of the CD4+ T cells from NALT were CD45RB+ (low plus high) vs 65.7% for the CD4+ cells from cLN (Fig. 5A), from which 14.6 and 27.0% were CD45RBlow, respectively. This suggested that NALT contained a higher proportion of naive T cells (CD45RBhigh) and fewer T cells exhibiting an activated, memory, or regulatory phenotype (CD45RBlow). This was also supported by the lack of expression of activation markers, such as CD69 (data not shown). Thus, it seems that potential CD4+CD45RBlow regulatory T cells could constitute only a minor subpopulation within NALT.
FIGURE 5. Immune regulatory function of B220lowCD3+CD4–CD8– cells from NALT. A, Double-stained cells (CD45RB/CD4) isolated from NALT and cLN of 8-wk-old BALB/c mice were analyzed by gating on FSClow and FSChigh cells. Approximately 14.6 and 27.0% of CD4+ cells from NALT and cLN corresponded to CD4+CD45RBlow T cells (as determined by a cut-off above the isotype control and below a MFI of 160), respectively. B, Ag-specific CD4 T cell proliferative responses of total () vs B220-depleted () cell preparations from NALT of TCR-HA mice were measured after in vitro restimulation in the presence of HA peptide 110–120. The stimulation index was determined as the ratio of cpm of cells cultured with peptide (5 μg/ml) or Con A (5 μg/ml) to control cells grown in medium. C, The cytokines secreted into culture supernatants after in vitro restimulation with the HA peptide were measured after 96 h of culture. Results are expressed as mean values; SDs are indicated by vertical lines. D, HA-specific proliferative response of LN cells cocultivated in a Transwell system with either tissue culture medium or irradiated cells (NALT cells, CD4+-depleted NALT cells, or LN cells). Results are expressed as both stimulation indexes and percent inhibition compared with the values of LN control cells cocultivated without additional cells in the lower chamber. Results are representative of two independent experiments. *, p < 0.1; **, p < 0.05.
Additional studies were performed to evaluate the regulatory potential of CD3lowB220lowCD4–CD8– cells from NALT. To this end, we assessed the proliferative capacity of cells isolated from NALT of TCR-HA transgenic mice (22, 23). Significant proliferative responses from NALT cells were observed in the presence of both the specific HA peptide and the mitogenic stimulator Con A (Fig. 5B). However, when NALT cells were depleted from the B220+ subpopulation, we observed a significant increment in HA-specific proliferation (Fig. 5B). No differences were detected in response to Con A (Fig. 5B) or after depletion of CD19+ cells (data not shown). Then we analyzed the cytokines secreted by cells stimulated with the HA peptide. We observed an increment (i.e., mean > control + 2 SD) in the production of TNF-, IFN-, and IL-2, when B220+ cells were depleted, whereas the levels of secreted IL-10, IL-6, IL-5, and IL-4 were not significantly increased (Fig. 5C). These results are in agreement with the dominant Th1 cytokine response pattern of peptide-stimulated spleen cells from TCR-HA animals (Fig. 5C) previously observed (23, 29).
In an attempt to unravel the underlying mechanisms of the observed suppressive activity from NALT cells, Transwell coculture studies were performed. The results showed that the suppression induced by NALT cells is supported by soluble factors able to neutralize up to 70% of the proliferative stimulation (Fig. 5D). Surprisingly, the complete inhibitory effect of NALT cells required the presence of CD4+ T cells, because the removal of these cells resulted in a decreased (40%) suppression (Fig. 5D). Thus, it seems that NALT cells exert their suppressive activity via soluble factors in cooperation with CD4+ T cells.
CD3lowB220lowCD4–CD8– cells from NALT are responsive to danger signals
NALT are not only a tolerogenic site, but are also an important inductive site in the context of appropriate stimulation, such as those delivered by TLR (5, 30). Thus, we evaluated the expression of TLR on the surface of cells from NALT. As shown in Fig. 6A, more cells expressing high levels of TLR2 were observed in NALT than in cLN (63.5 vs 5.8%). For the large majority, these TLR2+ cells from NALT were also B220low positive. This suggests that TLR2 signaling could play an important role in the homeostasis and reactivity of B220low cells as well as in the induction of immune responses in nasal tissues.
FIGURE 6. TLR2 expression and sensitization of B220lowCD4–CD8– from NALT. A, Surface expression of TLR2 on cells isolated from NALT and cLN of 7- to 9-wk-old BALB/c mice were analyzed by flow cytometry. Double-stained TLR2/B220 cells were analyzed by gating on FSClow and FSChigh cells. B, Freshly isolated cells from NALT of mice that received MALP-2 (0.5 μg) or PBS (nonactivated) by the intranasal route 16 h before organ processing were stained with Abs specific for the T cell markers CD4, CTLA-4, and CD62L. Dot plots correspond to FACS analysis gating on both FSClow and FSChigh. The percentages of positive cells in the corresponding quadrants are indicated.
Despite its crucial role in immune activation, TLR2 could also signal for apoptosis (31, 32). Thus, we evaluated activation and apoptosis after stimulation with the TLR-2 ligand MALP-2 (25, 33). The number of apoptotic cells was not increased in NALT 3 or 16 h after intranasal administration of 0.5 μg of MALP-2 (data not shown). In contrast, we observed an increment in the activation status of CD4– T cells, as reflected by the down-regulation of CTLA-4 and CD62L (0.6 vs 6.7% and 7.3 vs 12.8%, respectively; Fig. 6B).
Discussion
Atypical cells have been previously described in association with NALT. Hiroi et al. (34) reported the presence of CD3+CD4–CD8– cells, whereas Wu et al. (5) described a large fraction of small cells. However, the phenotypic and functional characterization of these unconventional cells has remained elusive to date. Our analysis showed that a large fraction of NALT cells (45% of the total cells) corresponded to CD3lowB220lowCD4–CD8– cells. In TCR-HA mice, the transgenic TCR is expressed not only on CD4+CD8– and CD4–CD8+ mature T cells, but also on CD4–CD8– immature T cells (22). By using these TCR transgenic mice, we clearly demonstrated that the double-negative cells observed in NALT belong to the T cell lineage. Interestingly, in accordance to their small size, the majority of the NALT B220lowCD4–CD8– cells were apoptotic. It has been reported that the B cell marker B220 is up-regulated on activated T cell blasts before entering apoptosis (27, 35, 36). Considering that nasal tissues are an overexposed environment, it can be hypothesized that cells from NALT undergo apoptosis after activation by exogenous Ags or superantigen, as previously demonstrated in other experimental systems (27, 35, 37, 38). However, the expression levels of both CD45RB and activation markers suggests that the majority of cells from NALT corresponded to naive T cells (CD45RBhigh). Moreover, a similar cell phenotype was found in NALT from athymic nu/nu mice. Thus, the B220low T cells in NALT do not appear to be related to the presence of mature T cells. Therefore, it seems unlikely that a down-regulation of surface markers as a by-product of apoptosis accounts for the assigned phenotype.
We favor the hypothesis that the expression of the B220 molecule on double-negative cells from NALT is related to a precursor cell phenotype. This is in agreement with the surface expression of the precursor marker c-Kit observed in this study, and it is also supported by the fact that lymphoid progenitor cells have been recently identified in human tonsils (39). B220+CD19– pluripotent precursor cells are also found in bone marrow and fetal liver (40, 41). Although it was accepted that commitment to the T lineage is an intrathymic event, prothymocytes are also found in murine fetal blood (42). In addition, the intestine can occasionally act as a site of positive selection for thymus-derived T cells (43). Concerning NALT, it remains to be elucidated whether and why progenitor or immature T cells migrate or accumulate in the nasal tissues and undergo apoptosis. Studies focused on elucidation of the molecular events leading to initiation and organization of NALT showed that, to date, only deficiency in the regulator of cell differentiation Id2 completely abolished NALT development (24, 44). This deficiency can be overcome by adoptive transfer of fetal liver cells, suggesting a primal origin for NALT initiator cells.
Our hypothesis is that the double-negative B220low cells might be involved in formation of the reticular network maintaining the homeostasis of mature lymphocytes from NALT. We also postulate that this cellular subpopulation might represent an important immunomodulatory system for NALT. Several observations support this hypothesis. First, no or very few classical regulatory T cells, such as CD4+CD25+ and CD4+CD45RBlow, have been found in nasal tissues (45, 46, 47, 48). Second, previous work suggested that B220+CD4–CD8– cells can affect the capacity of activated T cells to proliferate or produce cytokines (18, 36, 49). In this study we demonstrated that B220low double-negative cells in NALT exhibited the capacity to down-regulate mature T cell activation. This finding is in agreement with the fact that apoptosis generates a tolerogenic milieu rather than proinflammatory reactions (50). However, we also showed that CD3lowB220lowCD4–CD8– cells from NALT express high levels of the innate immunity receptor TLR2, and that in vivo stimulation of the TLR2/6 signaling cascade in nasal tissues leads to an activated phenotype of these cells. In conclusion, CD3lowB220lowCD4–CD8– cells constitute the main cellular subpopulation from NALT. It is conceivable that nasal tissues have acquired a regulatory system in which the same cellular pool plays different functions, such as contributing toward homeostasis maintenance, immune suppression (e.g., via production of soluble factors and/or its proapoptotic nature), and immune activation (e.g., via its high expression of innate immunity receptors).
Acknowledgments
We are particularly grateful to Carsten J. Kirschning for the polyclonal rabbit anti-murine TLR2 Ab, to Peter F. Mühlradt for providing us with MALP-2, and to Urte J?ger for her outstanding technical help.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 Address correspondence and reprint requests to Dr. Carlos A. Guzmán, Vaccine Research Group, Division of Microbiology, GBF-German Research Center for Biotechnology, Mascheroder Weg 1, D-38124 Braunschweig, Germany. E-mail address: cag@gbf.de
2 Abbreviations used in this paper: NALT, nasal-associated lymphoid tissue; 7-AAD, 7-aminoactinomycin D; cLN, cervical lymph node; FSC, forward scatter; HA, hemagglutinin; LN, lymph node; MFI, mean fluorescence intensity; PARP, poly(ADP-ribose) polymerase; MALP-2, macrophage-activating lipopeptide of 2 kDa.
Received for publication May 10, 2004. Accepted for publication November 16, 2004.
References
Kuper, C. F., P. J. Koornstra, D. M. Hameleers, J. Biewenga, B. J. Spit, A. M. Duijvestijn, P. J. van Breda Vriesman, T. Sminia. 1992. The role of nasopharyngeal lymphoid tissue. Immunol. Today 13:219.
Wiley, J. A., R. J. Hogan, D. L. Woodland, A. G. Harmsen. 2001. Antigen-specific CD8+ T cells persist in the upper respiratory tract following influenza virus infection. J. Immunol. 167:3293.
Asanuma, H., A. H. Thompson, T. Iwasaki, Y. Sato, Y. Inaba, C. Aizawa, T. Kurata, S. Tamura. 1997. Isolation and characterization of mouse nasal-associated lymphoid tissue. J. Immunol. Methods 202:123.
Liang, B., L. Hyland, S. Hou. 2001. Nasal-associated lymphoid tissue is a site of long-term virus-specific antibody production following respiratory virus infection of mice. J. Virol. 75:5416.
Wu, H. Y., H. H. Nguyen, M. W. Russell. 1997. Nasal lymphoid tissue (NALT) as a mucosal immune inductive site. Scand. J. Immunol. 46:506.(Faiza Rharbaoui, Dunja Br)