Dendritic cells (DCs) are a heterogeneous population of APCs with critical roles in T cell activation and immune regulation. We report in this study the identification and characterization of a novel subset of DCs resident in skin-draining peripheral lymph nodes of normal mice. This subset of CD11chighCD40highCD8αintermediate (int) DCs expresses the collagen-binding integrin, α1β1, and the E-cadherin-binding integrin, αEβ7. Although α1β1 and αEβ7 are also expressed on CD11chighCD40intCD8αhigh lymphoid DCs, CD11chighCD40highCD8αint DCs demonstrate preferential integrin-mediated adhesion to collagen and fibronectin. This DC subset most likely acquires expression of these integrins in peripheral lymph node, as this subset is not found in the spleen or mesenteric lymph node, and recent DC migrants from the skin lack expression of α1β1 and αEβ7 integrins. Resident CD40high DCs express α1β1 integrin and colocalize with collagen in lymph nodes. When compared with CD11chighCD40highCD8αint DCs lacking expression of these integrins, the α1β1+αEβ7+ DC subset exhibits more efficient formation of Ag-independent conjugates with T cells, and a decreased ability to acquire soluble Ag. Thus, the α1β1 and αEβ7 integrins define a unique population of peripheral lymph node-derived DCs with altered functional properties and adhesive potential that localizes these cells to sites in lymph nodes where Ag presentation to T cells occurs.

Dendritic cells (DCs) 3 are potent APCs that regulate the immune response through the induction of adaptive immunity and the maintenance of peripheral tolerance (1, 2). As a population, DCs are composed of a number of functionally and phenotypically distinct subsets that have been defined by differential expression of several cell surface receptors. Among the most informative markers of DC subsets are integrins, a family of αβ heterodimeric transmembrane glycoproteins that mediate adhesion to the extracellular matrix (ECM) as well as to other cells (3). CD11c, the αX subunit of the αXβ2 integrin, was the first murine DC marker to be identified (4) and has been instrumental in the identification and characterization of DC subsets in the lymph node (LN), spleen, and thymus (5, 6, 7). In addition to CD11c, CD11b, the αM subunit of the αMβ2 (MAC-1) integrin, has become important in differentiating DC subsets (5). Originally identified as a myeloid-specific Ag, CD11b distinguishes myeloid DC lacking CD8α expression from the lymphoid DC subset that expresses CD8α (5). Finally, recent evidence suggests that the α4 integrin subunit (CD49d) is a maturation marker on monocyte-derived human DCs (8).

In addition to their role as antigenic markers, integrins most likely regulate DC function and localization in lymphoid tissue. The LN is a highly structured tissue in which functional and spatial separation is maintained by a reticular network of collagen fibers and other ECM proteins (9). This network of conduits connects the lymphatic vessels draining into the subcapsular sinus of LNs with high endothelial venules (HEVs) and is critical for the movement of soluble Ag and inflammatory cytokines from tissues into draining LNs (10). Although these fibers are ensheathed by fibroblastic reticular cells, ∼10% of the ECM fibers remain exposed to the cellular compartment of the LN and may serve as potential binding sites for DCs and other cell types (11). Within the LN, differences in either the ECM content of the reticular network or the integrin ligands produced by fibroblastic reticular cells that surround these fibers may define discrete functional areas by specifically retaining DCs expressing the appropriate integrins within those areas. The ability of different DC subsets to interact with collagen and other ECM proteins found in these fibers may be critical to our understanding of the function of these DC subsets, as recent studies have shown that the uptake of soluble Ag by skin-derived DCs resident in draining LNs is critical to the initial activation of naive T cells (12). Because this network severely limits the diffusion of soluble Ags in LNs (10), this suggests that DCs capable of adhering to these ECM fibers are uniquely situated to serve as the critical APCs that initiate T cell activation by acquiring soluble Ag and presenting it to T cells. In situ analysis has revealed that DC subsets differentially expressing the αM integrin (CD11b) localize to distinct areas of the LN (13), and the specific localization pattern of each DC subset may be important to its role in the regulation of the immune response (14, 15). However, the mechanism by which DC subsets are situated in distinct regions of LNs, and particularly the role of ECM-binding integrins in this process, is not known.

Integrins may also regulate DC function by controlling the ability of DCs to traffic to the LN from peripheral sites. Blood-borne DCs enter the LN through the HEVs, and both β1 and β2 integrins mediate adhesion of blood-borne DCs to human umbilical vein endothelium (16). By contrast, DCs migrating from solid tissue such as skin arrive at the LN via the afferent lymph. Contact hypersensitivity models have clearly demonstrated that the αLβ2 (LFA-1) integrin and its ligand, ICAM-1, are required for the migration of Langerhans cells to the LN (17, 18). Similar studies have also demonstrated a role for α6 integrins, which mediate adhesion to the ECM protein laminin, in Langerhans cell migration out of the skin (19). In addition to the role of integrins in DC homing to the LN, β2 integrins are critical for the homeostatic accumulation of DC within the lungs (20).

In this study, we identify and characterize a novel subpopulation of DCs in normal mouse peripheral LNs (pLNs) that coordinately expresses high levels of the α1β1 and αEβ7 integrins and exhibits unique adhesive and Ag uptake capabilities.

C57BL/6 (B6) mice were purchased from the National Cancer Institute (Bethesda, MD) and were housed in specific pathogen-free isolation rooms at the University of Minnesota (Minneapolis, MN). OT-I TCR transgenic mice on a B6/PL background were provided by K. Hogquist (University of Minnesota) (21). Mice were used between 8 and 12 wk of age and housed in specific pathogen-free facilities at the University of Minnesota. All experimental protocols involving the use of mice were approved by the Institutional Animal Care and Use Committee at the University of Minnesota.

DCs were isolated from pLN, as described (13). Briefly, superficial inguinal, axillary, superficial cervical, and mandibular LN were isolated from at least two mice and teased apart in a 400 U/ml collagenase D (Roche Molecular Biochemicals, Indianapolis, IN) solution in RPMI 1640 supplemented with 10 mM HEPES, pH 7.2, 2% FCS, penicillin/streptomycin, and l-glutamine. LNs were incubated at 37°C for 20 min before stopping the digestion with 0.1 M EDTA solution (pH 7.2). The LNs were filtered and washed twice in PBS. In experiments in which mesenteric LNs were used, they were treated in the same manner as the pLNs.

DC subsets were identified by staining with anti-CD11c PE or FITC mAb HL3 (BD PharMingen, San Diego, CA), and anti-CD8α FITC and anti-CD40 PE/Cy5 mAb 1C10 (both from eBioscience, San Diego, CA). Integrin expression was determined using Abs against integrin α1 (CD49a) (Ha31/8), α2 (CD49b) (Ha1/29), α4 (CD49d) (R1-2), α5 (CD49e) (5H-10-27 (MFR5)), αL (CD11a) (M17/4), β1 (CD29) (HMβ1-1), αE (CD103) (M290), or αE FITC (M290), and β4 (CD104) (346-11a), all of which were purchased from BD PharMingen. Abs against α3 (CD49c) were purchased from BD Transduction Laboratories (San Diego, CA). Integrin α6 (CD49f) was stained using Hmα6 mAb (kindly provided by H. Yagita, Juntendo University, Tokyo, Japan) (22, 23).

To characterize integrin expression on DC subsets, DCs were isolated, as described above, and resuspended in FACS buffer (HBSS supplemented with 1% BCS and 0.2% sodium azide). Fc receptors were blocked for 15 min at 4°C with either anti-FcR Ab (clone 2.4G2) or mouse IgG and then stained with unconjugated Abs against specific integrin subunits for 30 min at 4°C. Excess Ab was washed off, and the cells were stained with either biotinylated goat anti-hamster or biotinylated goat anti-rat depending on the primary Ab. Unbound secondary Abs were washed off, and any free epitopes were blocked with hamster IgG or rat IgG. Finally, the cells were stained with a mixture of CD8α FITC, CD11c PE, CD40 PE/Cy5, and streptavidin APC (SA-APC). In each experiment, >750,000 cells for each sample were acquired with a FACSCalibur (BD Biosciences, San Jose, CA) and subsequently analyzed using CellQuest software (BD Biosciences). The DCs were identified by gating on the CD11chigh cells. This population of cells was then subdivided based on CD8α and CD40 expression (Fig. 1 A).

FIGURE 1.

Integrin expression on CD11chigh DC subsets in the pLN. LN cells from normal B6 mice were stained with CD8α FITC, CD11c PE, and CD40 PE/Cy5, as described in Materials and Methods. A, CD40 and CD8α expression on CD11chigh cells (based on the gate shown on the left contour plot) from pLN, mesenteric LNs (mLN), and spleens of B6 mice. B, Using the gates shown in A, four-color flow cytometry was used to assess the expression of integrin α and β subunits on CD11chigh DC subsets in pLN. Filled histograms show integrin expression on pLN-derived DCs (CD40highCD8int), myeloid DCs (CD40intCD8αlow), and lymphoid DCs (CD40intCD8αhigh). The negative controls for each DC subset are shown with empty histograms. Data are representative of at least three independent experiments.

FIGURE 1.

Integrin expression on CD11chigh DC subsets in the pLN. LN cells from normal B6 mice were stained with CD8α FITC, CD11c PE, and CD40 PE/Cy5, as described in Materials and Methods. A, CD40 and CD8α expression on CD11chigh cells (based on the gate shown on the left contour plot) from pLN, mesenteric LNs (mLN), and spleens of B6 mice. B, Using the gates shown in A, four-color flow cytometry was used to assess the expression of integrin α and β subunits on CD11chigh DC subsets in pLN. Filled histograms show integrin expression on pLN-derived DCs (CD40highCD8int), myeloid DCs (CD40intCD8αlow), and lymphoid DCs (CD40intCD8αhigh). The negative controls for each DC subset are shown with empty histograms. Data are representative of at least three independent experiments.

Close modal

To determine expression of multiple integrins on DCs, Fc receptors were blocked, as described above, and the cells were subsequently stained with the anti-α1 mAb Ha31/8 for 30 min at 4°C and followed by a biotinylated goat anti-hamster Ab. Unbound secondary Abs were washed off, and any Ag-binding epitopes were blocked with hamster IgG. Finally, the cells were stained with a mixture of Abs that included CD11c PE, CD40 PE/Cy5, αE FITC M290, and SA-APC, and analyzed, as described above.

Flat-bottom 96-well microtiter plates (Costar, Cambridge, MA) were coated overnight at 4°C with the indicated amount of one of the following integrin substrates: type IV collagen (Life Technologies, Rockville, MD), type III collagen (Fibrogen, South San Francisco, CA), laminin (Life Technologies), E-cadherin (R&D Systems, Minneapolis, MN), human plasma fibronectin (Invitrogen Life Technologies, Carlsbad, CA), or murine fibronectin (Invitrogen). Unbound sites were blocked with 0.5% human serum albumin (HSA) for 2 h at 37°C. Three hundred thousand cells were added per well in a final volume of 0.1 ml PBS containing 0.5% HSA. The cells were allowed to settle on the plate for 60 min at 4°C. Following 20 min of stimulation at 37°C, the nonadherent cells were washed away and the adherent cells were removed using a 0.1 mM EDTA solution. For each condition, duplicate samples of cells pooled from six wells were washed in FACS buffer. Samples were stained with CD8α FITC, CD40 PE/Cy5, and CD11c PE (as described above), washed, and then resuspended in exactly 200 μl FACS buffer plus 50 μl PKH26 microbeads (Sigma-Aldrich, St. Louis, MO) for a total sample volume of 250 μl. Each sample was collected on a FACScan (BD Biosciences) for ∼4 min and analyzed using CellQuest software (BD Biosciences). Triplicate samples of preadherent cell populations were prepared in a similar manner, and all of the cells were analyzed by flow cytometry. The number of DCs, based on differential expression of CD8α and CD40 expression on the CD11chigh cells, in each sample was determined, as previously described (24).

Migration of skin-derived DCs to the draining LN was determined, as previously described (7, 19). Green fluorescent Cell Tracker (Molecular Probes, Eugene, OR) was resuspended at a concentration of 10 mM in DMSO. Just before application, the Cell Tracker dye was diluted to a final concentration of 3.3 mM in a 50/50 (v/v) acetone/dibutyl phthalate mixture. Mice were painted with 50 μl on both the dorsal and ventral sides of the ears. After 18 h, LNs were harvested and the DCs were collected and phenotyped, as described above. Briefly, cells were stained with unconjugated anti-α1 or anti-αE integrin Abs, followed by biotinylated secondary mAb. Unbound epitopes on the secondary Abs were blocked, and the cells were stained with CD11c PE, CD40 PE/Cy5, and SA-APC. Cells were analyzed, as described above.

pLN tissue sections were prepared and stained, as previously described (12). Detection of integrin α1 or α2 chain colocalization with collagen was performed using either hamster anti-mouse α1 integrin or hamster anti-mouse α2 integrin Abs, followed by biotinylated goat anti-hamster IgG (Caltag, Burlingame, CA), SA-conjugated HRP (SA-HRP), and tyramide FITC (NEN, Boston, MA) or tyramide rhodamine (NEN). Collagen was detected using a biotinylated goat anti-collagen type III Ab (Southern Biotechnology Associates, Birmingham, AL), followed by SA-HRP and tyramide Cy5 (NEN). Integrin α1 colocalization with CD40 was detected with hamster anti-mouse α1 integrin (as described above) and biotinylated rat anti-mouse CD40 (Caltag), followed by SA-HRP, biotinyl tyramide (NEN), and SA-Cy5 (Caltag). Images were acquired using a Bio-Rad MRC-1000 confocal microscope equipped with a krypton/argon laser (Bio-Rad Life Sciences, Hercules, CA) and CoMOS v. 7.0a (Bio-Rad Life Sciences) software. Separate green, red, or far red (Cy5) images were collected for each section analyzed. Final image processing was performed using Photoshop software (Adobe, San Jose, CA).

pLNs and mesenteric LNs were isolated from OT-I TCR transgenic mice that express the CD90 isoform, Thy-1.1. T cells were isolated by negative selection using MACS magnetic separation technology (Miltenyi Biotec, Auburn, CA). Briefly, LN cells were incubated in a mixture of the following Abs for 20 min at 4°C: goat anti-mouse IgG FITC (0.5 μg/106 cells) (Southern Biotechnology Associates), anti-B220 FITC (0.07 μg/106 cells) (eBioscience), anti-I-Ab FITC (0.07 μg/106 cells) (BD PharMingen), and anti-αE FITC (0.07 μg/106 cells), followed by incubation with anti-FITC microbeads for 20 min at 4°C (10 μl beads/10 × 106 cells). Cells were washed twice and run over MS or LS separation columns.

For conjugate assays, DCs from normal, nontransgenic B6 mice were isolated, as described above, and resuspended at 20 × 106 cells/ml in RPMI 1640 supplemented with 10% FCS (RP10). DCs were pulsed with 1 μM or 100 nM OVA peptide (SIINFEKL), 1 μM SIY peptide (SIYRYYGL), or no peptide for 30 min at 37°C. Peptides were synthesized by Research Genetics (Carlsbad, CA). After pulsing, cells were washed and resuspended in ice-cold RP10. DCs and T cells were then added to flat-bottom 96-well microtiter plates at a concentration of 2 × 106 DCs/well and 0.5 × 106 T cells/well in a total volume of 200 μl/well. Plates were centrifuged for 1 min at 1000 rpm and then incubated for 30 min at 37°C. Following conjugation, plates were placed on ice and Fc receptors were blocked using anti-FcR Ab (clone 2.4G2) for 15 min. Cells were then stained with anti-αE FITC, anti-CD11c PE, anti-CD40 PE/Cy5, and anti-Thy-1.1 APC (eBioscience) Abs for 30 min. Plates were washed with 200 μl FACS buffer, spun down, and fixed with 1% paraformaldehyde for 20 min at room temperature. Cells were washed twice with 200 μl FACS buffer and harvested from the plate. For each condition, duplicate samples of cells were pooled from three wells, and each sample was analyzed with a FACSCalibur (BD Biosciences) using CellQuest software (BD Biosciences). To determine the degree of conjugate formation, we gated on the cells expressing high levels of CD11c and CD40. Then using αE integrin expression, this population was subdivided into two subpopulations: CD11chighCD40high αE positive and CD11chighCD40high αE negative. The percentage of conjugate formation with OT-I T cells was determined as the number of cells within each subpopulation that exhibit FL-4 (Thy-1.1) fluorescence divided by the total number of cells in each subpopulation.

The ears of C57BL/6 mice were injected s.c. with 50 μg of red fluorescent protein DsRed (EαRFP), as described (12). Four hours following injection, DCs were isolated, as described above, from the draining and nondraining LNs. DCs were stained, as described above, with unconjugated anti-α1 integrin Ab, followed by biotinylated secondary mAb. Unbound epitopes on the secondary Abs were blocked, and the cells were then stained with anti-CD11c FITC, anti-CD40 PE/Cy5, and SA-APC. CD11chighCD40high DCs were identified and analyzed, as described above.

Statistical analysis was performed using the Student’s t test.

To minimize effects of ex vivo purification on DC phenotype, we used multicolor flow cytometry to differentiate DC subsets expressing CD11c, CD40, and CD8α. Using these markers, we identified three populations of CD11chigh DCs in pLN: CD40intermediate (int)CD8αlow (24%), CD40intCD8αhigh (25%), and CD40highCD8αint (18%) (Fig. 1 A). Consistent with previous results (7), the CD40intCD8αlow and the CD40intCD8αhigh populations were also found in the mesenteric LNs and spleen. Classically, these two DC subsets have been designated as myeloid (CD11chighCD40intCD8αlow) and lymphoid (CD11chighCD40intCD8αhigh) DCs (25, 26). Although it is clear that these designations no longer reflect the developmental lineage of these cells (27, 28, 29), we will use these terms to describe DCs with the aforementioned phenotypes. In contrast to myeloid and lymphoid DCs, a population of CD11chighCD40highCD8αint DCs was found exclusively within skin-draining pLN. Although DCs from the skin could be of epidermal and dermal origin, CD11c and CD8α have been reported to be differentially expressed by these populations. The expression of high levels of CD11c and intermediate levels of CD8α is consistent with the phenotype of Langerhans cells (7).

Using these gates, we used four-color flow cytometry to examine the expression of several integrin α and β subunits on each subset of CD11chigh DCs in pLN. In addition to CD11c (αX integrin subunit), the expression of αM (CD11b, MAC-1) has been used to differentiate the myeloid and lymphoid DC subsets. In agreement with previous results (5, 30), we found that αM was highly expressed on the myeloid and CD40highCD8αint pLN-derived DCs, but was expressed at low levels on lymphoid DCs (Fig. 1,B). In contrast, the αL integrin subunit (CD11a) was uniformly expressed on each of the CD11chigh DC subsets (Fig. 1 B) (30).

In contrast to the β2 family of integrins, little is known about the expression of the integrins β1 and β7 subunits, or the α subunits that pair with these integrin β-chains, on murine DC subsets. As shown in Fig. 1,B, the integrin β1 subunit was expressed on all three CD11chigh DC subsets. The α4 and α5 subunits, which associate with β1 to form the fibronectin-binding integrins, α4β1 and α5β1, were also expressed at similar and uniform levels on each CD11chigh DC subset. Although α5 dimerizes exclusively with β1, α4 may form heterodimers with either β1 or β7. Although the β7 integrin subunit was expressed on all three DC subsets, there was a subpopulation of CD40highCD8αint pLN-derived DCs that expressed high levels of β7 (Fig. 1 B). Because both β1 and β7 are expressed on all of the CD11chigh subsets, it is possible that DCs express both α4β1 and α4β7 integrin.

In contrast to the fibronectin-binding integrins, several β1 integrin subunit-associating α subunits are differentially expressed on CD11chigh DC subsets. The α6 integrin subunit was expressed predominately by the lymphoid DC subset and to a lesser extent by myeloid DCs. Consistent with previous reports, we did not detect appreciable α6 integrin expression on CD40highCD8αint pLN-derived DCs (19). Although the α6 integrin subunit can form two different laminin-binding integrins, α6β1 and α6β4, the β4 subunit was not expressed on any of the CD11chigh DC subsets (Fig. 1 B). This suggests that α6β1 is the only α6-containing integrin expressed on CD11chigh DCs.

The expression of the α1 integrin subunit was restricted to particular subsets of CD11chigh cells. As shown in Fig. 1,B, α1 was uniformly expressed on lymphoid DCs, but was not expressed on myeloid DCs. Thus, similar to CD8α and αM integrin, α1 integrin is differentially expressed on these two DC subsets. Integrin α1 was also expressed on ∼50% of the CD40highCD8αint pLN-derived DCs, suggesting that there is heterogeneity within this DC subset. In contrast, the α2 and α3 integrins, which can also pair with β1, were not expressed on any of the CD11chigh DC subsets (Figs. 1 B).

The β7 integrin subunit can also pair with the αE integrin subunit to form the αEβ7 integrin, which mediates adhesion of intraepithelial lymphocytes to E-cadherin (31, 32). Like the α1 integrin subunit, the αE integrin subunit was differentially expressed on the CD40int DCs, with higher expression on the lymphoid DCs. In addition, like α1, the αE integrin subunit was expressed on a subpopulation (∼30%) of the CD40highCD8αint pLN-derived DCs. Thus, both the α1 integrin subunit and the αE integrin subunit suggest heterogeneity within this subset of pLN-derived DCs.

We also examined α1 and αE integrin expression on CD11chigh DCs found in spleen. Similar to the results obtained with skin-draining pLN, α1 and αE integrin expression was detected on splenic lymphoid DCs, but not on splenic myeloid DCs (Fig. 2). Similar results were obtained when examining DCs isolated from mesenteric LN (Fig. 2 and data not shown). Thus, unlike what was observed in pLNs, we did not detect a subset of DCs in either the spleen or mesenteric LNs with low or intermediate CD8α expression that also expresses α1 and/or αE integrin.

FIGURE 2.

α1β1 and αEβ7 integrin expression on DC subsets in the spleen and mesenteric LN. Using the gates shown in Fig. 1 A, α1 and αE integrin expression was assessed on myeloid and lymphoid DCs fron spleen and mesenteric LN (mLN). Filled histograms show integrin expression, and negative controls for each DC subset are shown with empty histograms. Data are representative of at least three independent experiments.

FIGURE 2.

α1β1 and αEβ7 integrin expression on DC subsets in the spleen and mesenteric LN. Using the gates shown in Fig. 1 A, α1 and αE integrin expression was assessed on myeloid and lymphoid DCs fron spleen and mesenteric LN (mLN). Filled histograms show integrin expression, and negative controls for each DC subset are shown with empty histograms. Data are representative of at least three independent experiments.

Close modal

Among the three CD11chigh DC subsets in pLN that we analyzed for integrin expression, we observed the most heterogeneity in integrin expression on CD40highCD8αint DCs. Specifically, α1β1 and αEβ7 were expressed on only a subpopulation of CD11chighCD40highCD8αint cells (Fig. 1,B). We performed additional flow cytometric studies to determine whether αEβ7 and α1β1 were expressed on the same subset of CD11chighCD40high pLN-derived DCs. Fig. 3 shows that all of the αE-positive CD40highCD8αint pLN-derived DCs also expressed high levels of the α1 integrin. In contrast, the αE-negative CD40highCD8αint pLN-derived DCs could be further divided into subsets that expressed either no α1 integrin or low levels of α1 integrin. In addition, the forward and side scatter properties of CD40highCD8αint pLN-derived DCs differed based on expression of αE. The CD11chighCD40high pLN-derived DCs that expressed high levels of both αE and α1 were consistently larger and more granular than the cells lacking αE, with increased mean forward scatter and mean side scatter in the αE-positive CD40highCD8αint cells compared with the αE-negative CD40highCD8αint cells. Thus, these results demonstrate that there is a subset of CD40highCD8αint pLN-derived DCs that is defined by high levels of expression of both the α1β1 and αEβ7 integrins, and unique forward and side scatter properties.

FIGURE 3.

Coexpression of αEβ7 and α1β1 integrin on the same subset of pLN-derived CD11chighCD40high DC. pLN cells from B6 mice were stained with anti-CD11c PE and anti-CD40 PE/Cy5. Top, Contour plot of αE integrin expression (x-axis) and α1 integrin expression (y-axis) on CD11chighCD40high DC. Bottom, Forward and side scatter properties of the CD11chighCD40highCD8αintαE DC (left panel) and CD11chighCD40highCD8αintαE+ DC (right panel). The mean forward light scatter value was 633 for the αE-negative subset compared with 718 for the αE+ subset, and the mean side scatter value was 242 for the αE-negative cells compared with 305 for the αE+ cells. Results are representative of at least three independent experiments.

FIGURE 3.

Coexpression of αEβ7 and α1β1 integrin on the same subset of pLN-derived CD11chighCD40high DC. pLN cells from B6 mice were stained with anti-CD11c PE and anti-CD40 PE/Cy5. Top, Contour plot of αE integrin expression (x-axis) and α1 integrin expression (y-axis) on CD11chighCD40high DC. Bottom, Forward and side scatter properties of the CD11chighCD40highCD8αintαE DC (left panel) and CD11chighCD40highCD8αintαE+ DC (right panel). The mean forward light scatter value was 633 for the αE-negative subset compared with 718 for the αE+ subset, and the mean side scatter value was 242 for the αE-negative cells compared with 305 for the αE+ cells. Results are representative of at least three independent experiments.

Close modal

Functional and spatial separation within the LN is maintained by a reticular network of collagen and other ECM fibers ensheathed by fibroblastic reticular cells. We used immunohistochemistry to determine the localization of α1 integrin-expressing cells within the LN and their proximity to collagen filaments (Fig. 4). Cells expressing α1 integrin were located throughout the cortex of the LN, and there was extensive colocalization of anti-α1 integrin Ab staining with anti-type III collagen Ab staining. Both type III collagen and α1 integrin-expressing cells were excluded from the follicles. Because the α1β1 integrin mediates cell adhesion to collagen (33), the extensive overlap of α1 integrin and collagen throughout the LN suggests that fibroblastic reticular cells may be expressing α1β1 integrin. In contrast, cells expressing another collagen-binding integrin, α2β1, localized on the lumenal side of structures that are consistent with HEVs (Fig. 4). The lack of α2 integrin-expressing cells outside of HEVs is consistent with the lack of α2 integrin expression on any of the DC subsets that we examined by flow cytometry (Fig. 1 B).

FIGURE 4.

Cells expressing α1 integrin colocalize with collagen in pLN. Fixed and frozen sections of pLN were stained, as described in Materials and Methods, with Abs specific for type III collagen and α2 integrin (left panels), type III collagen and α1 integrin (middle panels), or CD40 and α1 integrin (right panels). The top set of panels shows pLN sections at a ×10 magnification, and the bottom set of panels shows pLN sections at a ×20 magnification. Note that cells expressing α1 integrins colocalize extensively with collagen throughout the cortex of the pLN. A portion of these α1-expressing cells also expresses high levels of CD40. Cells expressing α2 are localized exclusively to the lumenal side of HEVs. Follicles are marked with F.

FIGURE 4.

Cells expressing α1 integrin colocalize with collagen in pLN. Fixed and frozen sections of pLN were stained, as described in Materials and Methods, with Abs specific for type III collagen and α2 integrin (left panels), type III collagen and α1 integrin (middle panels), or CD40 and α1 integrin (right panels). The top set of panels shows pLN sections at a ×10 magnification, and the bottom set of panels shows pLN sections at a ×20 magnification. Note that cells expressing α1 integrins colocalize extensively with collagen throughout the cortex of the pLN. A portion of these α1-expressing cells also expresses high levels of CD40. Cells expressing α2 are localized exclusively to the lumenal side of HEVs. Follicles are marked with F.

Close modal

To identify DCs within the LN, sections were stained with an anti-CD40 Ab. Although there are many cells in the LN that express moderate levels of CD40 (7), we were only able to detect the cells expressing the highest levels of CD40, as demonstrated by the lack of appreciable anti-CD40 staining in follicles. LN sections stained with both anti-α1 and anti-CD40 clearly demonstrated that some, but not all, of the CD40high cells also expressed α1 integrin (Fig. 4). These data are consistent with expression of α1 on a subset of CD40highCD8αint pLN-derived DCs (Fig. 1 B).

We next used in vitro adhesion assays to assess the adhesion of CD11chigh DC subsets to β1 integrin ligands. Each of the CD11chigh DC subsets adhered to human and mouse fibronectin in a dose-dependent manner (Fig. 5,A and data not shown). We consistently observed slightly higher levels of adhesion of CD40highCD8αint pLN-derived DCs to fibronectin when compared with either the myeloid or lymphoid DC subsets. Ab-blocking studies demonstrated that an anti-β1 integrin Ab completely reduced adhesion of all three DC subsets to background levels (Fig. 5,B). Adhesion of CD40highCD8αint pLN-derived DCs to fibronectin was partially blocked by an anti-α4 or an anti-α5 Ab, suggesting a role for both α4 and α5 integrins in mediating adhesion of these DCs to fibronectin. In contrast, adhesion of myeloid and lymphoid DCs to fibronectin was only blocked by an anti-α4 Ab, while the anti-α5 Ab had a minimal effect, even when used in combination with the anti-α4 Ab (Fig. 5 B). Thus, even though α5β1 integrin is expressed on both myeloid and lymphoid DCs, we could not detect appreciable α5β1-dependent adhesion of these DCs to fibronectin. Adhesion of the three DC subsets to fibronectin was unaffected by an inhibitory anti-αL Ab.

FIGURE 5.

CD11chigh DC subsets adhere to fibronectin in an integrin-dependent manner. A, Adhesion of CD40highCD8αint pLN-derived DCs (circles), myeloid DCs (squares), and lymphoid DCs (triangles) from pLN to increasing amounts of fibronectin was assessed, as described in Materials and Methods. B, Adhesion of DC subsets to 1.5 μg/well of human fibronectin was assessed in the absence of Ab (untreated) or in the presence of Abs specific for the αL integrin (clone M17/4), α4 integrin (clone R1-2), α5 integrin (clone 5H10-27), β1 integrin (clone Ha2/5), or a combination of anti-α4 and anti-α5 integrin Abs. In each experiment, the mean percentage of adhesion of duplicate samples, each consisting of six pooled wells ± SD, is shown. Background adhesion of DC subsets to HSA (<22%) was subtracted. When compared with adhesion to HSA, the adhesion of all DC subsets to fibronectin was statistically different (p < 0.01). ∗, p < 0.01; ∗∗, p < 0.02. Results are representative of at least three independent experiments.

FIGURE 5.

CD11chigh DC subsets adhere to fibronectin in an integrin-dependent manner. A, Adhesion of CD40highCD8αint pLN-derived DCs (circles), myeloid DCs (squares), and lymphoid DCs (triangles) from pLN to increasing amounts of fibronectin was assessed, as described in Materials and Methods. B, Adhesion of DC subsets to 1.5 μg/well of human fibronectin was assessed in the absence of Ab (untreated) or in the presence of Abs specific for the αL integrin (clone M17/4), α4 integrin (clone R1-2), α5 integrin (clone 5H10-27), β1 integrin (clone Ha2/5), or a combination of anti-α4 and anti-α5 integrin Abs. In each experiment, the mean percentage of adhesion of duplicate samples, each consisting of six pooled wells ± SD, is shown. Background adhesion of DC subsets to HSA (<22%) was subtracted. When compared with adhesion to HSA, the adhesion of all DC subsets to fibronectin was statistically different (p < 0.01). ∗, p < 0.01; ∗∗, p < 0.02. Results are representative of at least three independent experiments.

Close modal

Because α1β1 integrin is differentially expressed on CD11chigh DCs, we also analyzed the adhesion of DCs to murine type IV collagen. As shown in Fig. 6,A, neither myeloid nor lymphoid DCs adhered to collagen in a dose-dependent manner, with adhesion ranging between 5 and 13% above background at all of the doses tested. This result was somewhat surprising, because lymphoid DCs express α1β1 integrin (Fig. 1,B). In contrast to the CD40int DCs, CD40highCD8αint pLN-derived DCs adhered to collagen in a dose-dependent manner, with maximal adhesion of 30% over background at the highest dose of collagen tested (Fig. 6,A). This is consistent with the expression of α1β1 integrin on a subset of these DCs. Adhesion of CD40highCD8αint pLN-derived DCs to type IV collagen was mediated, in part, by α1β1 integrin, because blocking Abs against either the α1 or β1 subunit reduced adhesion to collagen compared with untreated cells (Fig. 6 B). In contrast, an anti-αL integrin Ab had no effect on adhesion of this subset of DCs to collagen. CD40highCD8αint pLN-derived DCs also adhered preferentially to a recombinant form of human type III collagen, although overall levels of adhesion to type III collagen were slightly lower than that observed with type IV collagen (data not shown).

FIGURE 6.

CD11chighCD40high pLN-derived DCs adhere to type IV collagen and E-cadherin. A, Adhesion of CD40highCD8αint pLN-derived DCs (circles), myeloid DCs (squares), and lymphoid DCs (triangles) from pLNs of B6 mice to increasing amounts of type IV collagen was assessed, as described in Materials and Methods. ∗, p < 0.01 when compared with adhesion to HSA. B, Adhesion of CD40highCD8αint pLN-derived DCs to 1.5 μg/well of collagen was determined either in the absence of Ab (untreated) or in the presence of 10 μg/ml of Abs specific for αL integrin (clone M174), β1 integrin (clone Ha2/5), or α1 integrin (clone Ha31/8). ∗, p < 0.01 when compared with adhesion to HSA; ∗∗, p < 0.08 when compared with adhesion of untreated CD40highCD8αint pLN-derived DCs; ∗∗∗, p < 0.02 when compared with adhesion of untreated CD40highCD8αint pLN-derived DCs. C, CD11chighCD40high pLN-derived DCs (▪), myeloid DCs (▦), and lymphoid DCs (▪) were isolated from pLN, and adhesion to 0.1 μg/well of E-cadherin was assessed using the in vitro adhesion assay described in Materials and Methods. ∗, p < 0.10 of adhesion of CD40highCD8αint pLN-derived DCs to E-cadherin when compared with adhesion to HSA; ∗∗, p < 0.02 of adhesion of CD40highCD8αint pLN-derived DCs to E-cadherin when compared with adhesion of myeloid DCs to E-cadherin; ∗∗∗, p < 0.07 of adhesion of CD40highCD8αint pLN-derived DCs to E-cadherin when compared with adhesion of lymphoid DCs to E-cadherin. In each experiment, the mean percentage of duplicate samples, each consisting of six pooled wells ± SD, is shown. Background adhesion to HSA (<22%) was subtracted. Results are representative of at least three independent experiments.

FIGURE 6.

CD11chighCD40high pLN-derived DCs adhere to type IV collagen and E-cadherin. A, Adhesion of CD40highCD8αint pLN-derived DCs (circles), myeloid DCs (squares), and lymphoid DCs (triangles) from pLNs of B6 mice to increasing amounts of type IV collagen was assessed, as described in Materials and Methods. ∗, p < 0.01 when compared with adhesion to HSA. B, Adhesion of CD40highCD8αint pLN-derived DCs to 1.5 μg/well of collagen was determined either in the absence of Ab (untreated) or in the presence of 10 μg/ml of Abs specific for αL integrin (clone M174), β1 integrin (clone Ha2/5), or α1 integrin (clone Ha31/8). ∗, p < 0.01 when compared with adhesion to HSA; ∗∗, p < 0.08 when compared with adhesion of untreated CD40highCD8αint pLN-derived DCs; ∗∗∗, p < 0.02 when compared with adhesion of untreated CD40highCD8αint pLN-derived DCs. C, CD11chighCD40high pLN-derived DCs (▪), myeloid DCs (▦), and lymphoid DCs (▪) were isolated from pLN, and adhesion to 0.1 μg/well of E-cadherin was assessed using the in vitro adhesion assay described in Materials and Methods. ∗, p < 0.10 of adhesion of CD40highCD8αint pLN-derived DCs to E-cadherin when compared with adhesion to HSA; ∗∗, p < 0.02 of adhesion of CD40highCD8αint pLN-derived DCs to E-cadherin when compared with adhesion of myeloid DCs to E-cadherin; ∗∗∗, p < 0.07 of adhesion of CD40highCD8αint pLN-derived DCs to E-cadherin when compared with adhesion of lymphoid DCs to E-cadherin. In each experiment, the mean percentage of duplicate samples, each consisting of six pooled wells ± SD, is shown. Background adhesion to HSA (<22%) was subtracted. Results are representative of at least three independent experiments.

Close modal

Because both a subset of CD40highCD8αint pLN-derived DCs and lymphoid DCs express the αE and β7 integrins, we examined adhesion of DCs to purified E-cadherin. Although lymphoid DCs express αEβ7, we did not detect adhesion of these cells to E-cadherin above background (Fig. 6 C). As expected, myeloid DCs, which do not express αEβ7, also did not adhere to E-cadherin. By contrast, the CD40highCD8αint pLN-derived DCs adhered to E-cadherin, with optimal adhesion at a concentration of 0.1 μg/well of E-cadherin.

Given the heterogeneity of α1β1 and αEβ7 expression on CD40highCD8αint pLN-derived DCs in the LN, we used immunohistochemistry to determine integrin expression on Langerhans cells in the epidermis. We were unable to detect any α1- or αE-expressing epidermal DCs (data not shown). To confirm these results, we applied Cell Tracker dye to the ears of mice and assessed integrin expression on CD11chighCD40high DCs in pLN that had recently migrated from the skin. Although a subpopulation of Cell Tracker-negative CD11chighCD40high cells expressed both α1 and αE, CD11chighCD40high cells that were labeled with Cell Tracker were uniformly low in expression of both α1 and αE integrin (Fig. 7).

FIGURE 7.

α1β1 and αEβ7 integrin expression on CD40highCD8αint pLN-derived DCs is acquired in the LN. Each ear of a B6 mouse was painted with 50 μl of a solution containing 3.3 mM Cell Tracker Green solution in a 50:50 (v/v) mixture of acetone and dibutylthalate, as described previously (719 ). Eighteen hours following the application, DCs were isolated from pLNs, as described in Materials and Methods. DCs were stained with anti-CD11c PE, anti-CD40 PE/Cy5, and Abs against either α1 (top panel) or αE (bottom panel). The DCs stained with Cell Tracker were identified based on their fluorescence in the FL-1 channel, and represented 27% of the cells in the top panel and 33% of the cells in the bottom panel. Results are representative of at least three independent experiments.

FIGURE 7.

α1β1 and αEβ7 integrin expression on CD40highCD8αint pLN-derived DCs is acquired in the LN. Each ear of a B6 mouse was painted with 50 μl of a solution containing 3.3 mM Cell Tracker Green solution in a 50:50 (v/v) mixture of acetone and dibutylthalate, as described previously (719 ). Eighteen hours following the application, DCs were isolated from pLNs, as described in Materials and Methods. DCs were stained with anti-CD11c PE, anti-CD40 PE/Cy5, and Abs against either α1 (top panel) or αE (bottom panel). The DCs stained with Cell Tracker were identified based on their fluorescence in the FL-1 channel, and represented 27% of the cells in the top panel and 33% of the cells in the bottom panel. Results are representative of at least three independent experiments.

Close modal

Using αE integrin as a marker to distinguish subsets of CD11chighCD40high pLN-derived DCs, we examined the ability of these DC subsets to form stable conjugates with OT-I TCR transgenic T cells. Purified OT-I T cells were allowed to interact with pLN cells, and multicolor flow cytometry was used to enumerate the percentage of αE-positive and αE-negative CD11chighCD40high DCs in conjugates with T cells. In the absence of relevant peptide Ag, a low, but detectable, percentage of CD11chighCD40high pLN-derived DCs formed conjugates with OT-I T cells (Fig. 8,A). However, we consistently noted that the percentage of αE-positive CD11chighCD40high DCs in conjugates with OT-I T cells was 1.5- to 2-fold higher than the percentage of αE-negative CD11chighCD40high DCs in conjugates. If DCs were first pulsed with peptide Ag (OVA peptide SIINFEKL) before conjugate formation, the percentage of both αE-positive and αE-negative skin-derived DCs in conjugates with T cells dramatically increased. We also observed higher levels of Ag-dependent conjugates between αE-positive pLN-derived DCs and OT-I T cells compared with αE-negative pLN-derived DCs (Fig. 8 A).

FIGURE 8.

Differential conjugate formation and Ag uptake by CD40highCD8αint peripheral LN-derived DCs expressing the α1β1 and αEβ7 integrins. A, Ag-specific transgenic T cells were isolated and purified from LNs of OT-I B6.PL mice. DCs were isolated from the pLN of B6 mice and were pulsed in the absence of peptide or in the presence of the cognate peptide (OVA) or a control peptide (SIY). T cells were incubated with Ag-pulsed DCs, as described in Materials and Methods, and each sample was stained with anti-CD11c PE, anti-CD40 PE/Cy5, anti-αE FITC, and anti-Thy-1.1 APC. DCs were identified by the expression of high levels of CD40 and CD11c, and could be subdivided based on expression of αE. Transgenic OT-I T cells were identified by Thy-1.1 expression. The degree of conjugate formation was determined by the percentage of each subpopulation of DCs that also exhibited FL-4 (Thy-1.1) fluorescence. The mean percentage of duplicate samples, each consisting of three pooled wells ± SD, is shown. Results are representative of at least three independent experiments. ∗, p < 0.07; ∗∗, p < 0.01; ∗∗∗, p < 0.05. B, Four hours following the s.c. injection of the fluorescent Ag, EαRFP, draining and nondraining LN were harvested and stained for α1, CD11c FITC, and CD40 PE/Cy5. DCs were identified by the expression of high levels of CD11c and CD40. The degree of Ag uptake was determined by FL-2 fluorescence, and α1 integrin expression was determined by FL-4 fluorescence.

FIGURE 8.

Differential conjugate formation and Ag uptake by CD40highCD8αint peripheral LN-derived DCs expressing the α1β1 and αEβ7 integrins. A, Ag-specific transgenic T cells were isolated and purified from LNs of OT-I B6.PL mice. DCs were isolated from the pLN of B6 mice and were pulsed in the absence of peptide or in the presence of the cognate peptide (OVA) or a control peptide (SIY). T cells were incubated with Ag-pulsed DCs, as described in Materials and Methods, and each sample was stained with anti-CD11c PE, anti-CD40 PE/Cy5, anti-αE FITC, and anti-Thy-1.1 APC. DCs were identified by the expression of high levels of CD40 and CD11c, and could be subdivided based on expression of αE. Transgenic OT-I T cells were identified by Thy-1.1 expression. The degree of conjugate formation was determined by the percentage of each subpopulation of DCs that also exhibited FL-4 (Thy-1.1) fluorescence. The mean percentage of duplicate samples, each consisting of three pooled wells ± SD, is shown. Results are representative of at least three independent experiments. ∗, p < 0.07; ∗∗, p < 0.01; ∗∗∗, p < 0.05. B, Four hours following the s.c. injection of the fluorescent Ag, EαRFP, draining and nondraining LN were harvested and stained for α1, CD11c FITC, and CD40 PE/Cy5. DCs were identified by the expression of high levels of CD11c and CD40. The degree of Ag uptake was determined by FL-2 fluorescence, and α1 integrin expression was determined by FL-4 fluorescence.

Close modal

Itano et al. (12) have recently used a fluorescent Ag to show that skin-derived DCs resident in lymph nodes can acquire s.c. injected soluble Ag and initiate T cell activation. Using this system, we assessed potential differences in the ability of subsets of pLN-derived DC defined by differential α1 integrin expression to acquire this Ag, which consists of a portion of the I-Edα MHC class II subunit fused to the EαRFP. As previously demonstrated, Ag administered s.c. was present exclusively in the draining LN (Fig. 8,B), and there was minimal Ag uptake by myeloid and lymphoid DCs 4 h after Ag administration (data not shown). Ag uptake by CD11chighCD40high pLN-derived DCs was detected. Interestingly, there was an inverse relationship between α1 integrin expression on CD11chighCD40high pLN-derived DCs and Ag uptake. We observed that higher levels of EαRFP were detected in the α1-negative CD11chighCD40high pLN-derived DC subset when compared with EαRFP uptake by α1-positive CD11chighCD40high pLN-derived DCs (Fig. 8 B). Thus, these studies suggest differences in efficiency of Ag uptake by DC subsets defined by differential expression of the α1 integrin.

In this study, we demonstrate that β1 and β7 integrins are differentially expressed on CD11chigh DC subsets in the pLN of normal mice. Although the expression of fibronectin-binding integrins and adhesion to fibronectin are common to all DC subsets, we show that α1β1 and αEβ7 integrins are important phenotypic and functional DC markers. Specifically, we describe a subpopulation of CD40highCD8αint pLN-derived DCs that coordinately expresses the α1β1 and αEβ7 integrins. When compared with other CD11chighCD40high pLN-derived DCs, this subpopulation has a number of unique properties, including: 1) increased size and granularity; 2) ability to adhere to collagen and E-cadherin in vitro; 3) enhanced ability to form Ag-independent conjugates with T cells; and 4) decreased efficiency of acquisition of a soluble s.c. Ag. CD40high DCs expressing α1β1 integrin colocalize with collagen in LN, but α1 and αE are not expressed on CD11chigh cells in the skin and are not present on CD11chighCD40high cells that have recently emigrated from the skin. Although the αX (CD11c) and αM (CD11b) integrin subunits are routinely used as DC markers (1), analysis of the expression and function of other integrin subunits critical for cell adhesion and migration is lacking. Phenotypic analysis clearly shows that the CD11chighCD40highCD8αint DCs found in normal mouse pLN can be subdivided into two distinct subpopulations defined by differential expression of both α1β1 and αEβ7 integrin. Differential expression of α1β1 integrin is particularly intriguing, as α1β1 interacts with collagen (33), an ECM protein that is a primary component of the reticular network that forms a major structural scaffold in the LN (9). Besides the fibroblastic reticular cells that ensheath these collagen-laden reticular fibers, little is known about the other cells in the LN that interact with these fibers. Our results suggest that, among the DC subsets examined, CD11chighCD40highCD8αint pLN-derived DCs have enhanced capacity to interact with collagen found in the reticular network, because these DCs preferentially adhere to collagen in vitro. This suggests that this DC subset may be uniquely situated to acquire soluble Ag and immunoregulatory cytokines arriving from tissue sites via the afferent lymph, because these low m.w. molecules percolate into LN via the reticular network (10, 12). In addition to localizing DCs within the LN, collagen may directly regulate DC function by promoting secretion of IL-12 and the generation of a Th1 response (34).

Immunohistochemistry analysis demonstrated the colocalization of CD40high DCs expressing α1β1 integrins with collagen in lymph nodes in vivo. However, it is likely that both α1β1-positive and α1β1-negative CD40highCD8αint pLN-derived DCs interact with the collagen-rich reticular network in vivo, because CD40high DCs lacking expression of α1β1 were also found to be colocalized with collagen in lymph nodes in vivo. In addition, we were unable to completely block adhesion of skin-derived DCs to collagen with an inhibitory anti-α1 Ab. Thus, it is possible that other receptors expressed on CD40highCD8αint pLN-derived DCs, particularly α1β1-negative CD40highCD8αint pLN-derived DCs, may also mediate adhesion to collagen. It is unlikely that either α10β1 or α11β1, two recently discovered collagen-binding integrins (35, 36), is responsible for mediating adhesion to collagen because Abs to α1 and β1 reduced adhesion of the CD40highCD8αint pLN-derived subset to similar levels. Although other nonintegrin receptors, such as CD44 and CD26, are also capable of interacting with collagen (37, 38), both CD44 and CD26 are expressed at comparable levels on all CD11chigh DC (data not shown).

The extensive colocalization of α1 integrin staining with collagen in LNs suggests that the fibroblastic reticular cells also express α1 integrin and most likely use this integrin to interact with collagen in the reticular network. The α1 integrin staining pattern is in sharp contrast to that observed with another collagen-binding integrin, α2β1, which is found predominantly with the HEVs in LN. The rest of the LN is devoid of α2 integrin staining, consistent with the lack of α2 integrin staining on all of the CD11chigh DCs examined by flow cytometry.

Expression of the α1 and αE integrin subunits also distinguishes myeloid DCs from lymphoid DCs, as α1 and αE are both expressed at higher levels on the lymphoid DC subset. Thus, the lymphoid DC subset is characterized by higher levels of expression of the α1, αE, and α6 integrin subunits when compared with myeloid DCs. Although lymphoid DCs express α1β1 at levels comparable to CD40highCD8αint pLN-derived DCs, they do not adhere well to collagen in in vitro adhesion assays. It is possible that, like other hemopoietic cells, lymphoid DCs may require an exogenous signal that enhances integrin-mediated adhesion (39, 40, 41). Differential use of β1 integrins by DC subsets is also observed in the adhesion of DC subsets to fibronectin. Although all CD11chigh cells express comparable levels of the fibronectin-binding integrins α4β1 and α5β1, we consistently observed greater adhesion of CD40highCD8αint pLN-derived DCs to fibronectin. This may reflect the lack of α5 integrin-dependent adhesion of CD40int DCs to fibronectin, as an inhibitory anti-α5 mAb partially blocked the adhesion of CD40highCD8αint pLN-derived DCs to fibronectin, but had a minimal effect on the adhesion of myeloid and lymphoid DCs. In general, these results suggest that blood-borne DC subsets express less functionally active β1 integrins than CD40highCD8αint pLN-derived DCs.

We also examined whether α1β1 and αEβ7 integrins were expressed on DCs in the skin or on DCs that migrate from the skin into the LN, because CD11chighCD40highCD8αint DCs found in pLN have a phenotype consistent with Langerhans cells (7). However, we did not detect expression of either integrin on Langerhans cells in skin or on skin-derived DCs that migrate into the LN. These results are supported by the observation that human epidermal Langerhans cells do not adhere to collagen (42, 43) and suggest that α1β1 and αEβ7 integrins, unlike the α6β1 integrin (19), are not involved in the migration of Langerhans cells from the skin into LNs. Thus, our results suggest that a subpopulation of CD11chighCD40highCD8αint DCs acquires expression of α1β1 and αEβ7 subsequent to their migration from the skin into LN. Alternatively, the detection of this CD40high DC subset in pLN may be due to factors in the spleen and mesenteric LN that reduce CD40 expression or factors in pLN that enhance CD40 expression on DCs expressing α1β1 and αEβ7. However, we believe that this is unlikely, given that we were not able to identify a nonlymphoid DC subset that expresses α1β1 and αEβ7 in either the spleen or mesenteric LN. Although the nature of the signals that regulate integrin expression on CD40highCD8αint pLN-derived DCs remains undefined, our results clearly suggest that the local LN microenvironment plays a role in DC differentiation and function.

We identified two primary functional differences between CD11chighCD40high pLN-derived DCs defined by differential expression of α1β1 and αEβ7. The first is the increased Ag-independent adhesion of αE-positive CD40high DCs to T cells in vitro when compared with αE-negative CD40high DCs. The ability of DC subsets to form conjugates has not been extensively examined. Pulsing of CD40high DCs with relevant peptide Ag dramatically enhanced conjugate formation of both subsets of skin-derived DCs with OT-I T cells, consistent with the ability of TCR triggering to enhance integrin-dependent interactions between T cells and APCs (39, 44, 45). The enhanced ability of αE-positive CD40high DCs to interact with T cells, even in the absence of presentation of a relevant cognate Ag, may be important to our understanding of the dramatic fluctuations in the motility of naive T cells in LN, as determined in recent reports using two-photon microscopy (46, 47). The increased Ag-independent adhesion of αE-positive CD40high DCs to T cells, coupled with the presence of this DC subset in normal pLNs, also suggests the possibility that this DC subset may be particularly important in the role of DCs in maintaining tolerance to self Ags in the periphery (48). The second functional difference between these subsets of CD40highCD8αint pLN-derived DCs is the ability to take up soluble Ag in vivo. There was considerably greater Ag uptake in the α1-negative DCs compared with the α1-positive CD40highCD8αint pLN-derived DCs. This suggests that α1-negative CD40highCD8αint pLN-derived DCs are potent scavengers of soluble Ag and retain the endocytic capacity typically characteristic of peripheral DCs. This result is consistent with our finding that recent DC migrants from the skin also lacked expression of α1 integrin.

Together, these data suggest a model in which α1 and αE integrins define two functionally distinct subsets of CD11chighCD40high DCs in pLN. We propose that under homeostatic conditions, DCs resident in the skin lack expression of α1 and αE and migrate at a low, basal level to the LN in a manner independent of the expression of these integrins. After migration into the LN, α1 and αE integrin expression is induced on a subset of these skin-derived DCs, resulting in enhanced adhesion of these DCs to the collagen-rich fibers that are a major structural component of the conduit network. Thus, this DC subset is uniquely situated in the LN to acquire soluble Ag carried from tissue to the draining LN through these conduits. Tight apposition of DCs to these conduits is most likely critical to effective Ag uptake, as the conduit network provides a significant physical barrier to the diffusion of soluble Ags and immunoregulatory cytokines (10). Concomitant with these changes in integrin expression is a reduction in the ability to take up soluble Ag and enhanced ability to interact with T cells independent of Ag. It is interesting to note that the decreased capacity of α1β1+αEβ7+ CD11chighCD40high pLN-derived DCs to acquire Ag in vivo was not associated with decreased ability to form Ag-dependent conjugates with T cells. In fact, α1β1+ skin-derived DCs formed Ag-dependent conjugates at a level comparable to or greater than skin-derived DCs lacking expression of α1β1 integrin. The identification of two unique subsets of CD11chighCD40high pLN-derived DCs is consistent with recent evidence that both immature and mature skin-derived DCs are present in the LN (49) and supports a model in which α1 and αE integrins are markers of a functionally distinct and, possibly, mature subset of CD11chighCD40high skin-derived DCs. Acquisition and presentation of soluble Ag by resident skin-derived DCs have recently been shown to be critical to the initial activation of T cells (12). The novel DC subset identified in this report may be central to this process, because these cells express the appropriate integrin receptors that facilitate interaction with collagen, the primary ECM component found in the reticular network in LNs, and can interact very efficiently with T cells despite a reduction in the ability to acquire soluble Ag. Alternatively, the increased Ag-independent adhesion of this novel DC subset to T cells may be indicative of a function for this DC subpopulation in maintenance of tolerance in the periphery.

We thank Drs. K. Hogquist, M. Jenkins, D. Mayerova, and J. McCarthy for reagents as well as helpful comments and suggestions.

1

This work was supported by National Institutes of Health Grants AI38474 and AI31126 (to Y.S.), the Harry Kay Chair in Biomedical Research (to Y.S), a University of Minnesota Doctoral Dissertation Fellowship (to J.T.P.), the Irvington Institute of Immunological Research (to A.A.I.), and National Institutes of Health Training Grant AI007313 (to K.L.M.).

3

Abbreviations used in this paper: DC, dendritic cell; ECM, extracellular matrix; HEV, high endothelial venule; HSA, human serum albumin; int, intermediate; LN, lymph node; pLN, peripheral LN; RFP, red fluorescent protein; SA, streptavidin.

1
Shortman, K., Y. J. Liu.
2002
. Mouse and human dendritic cell subtypes.
Nat. Rev. Immunol.
2
:
151
.
2
Banchereau, J., R. M. Steinman.
1998
. Dendritic cells and the control of immunity.
Nature
392
:
245
.
3
Shimizu, Y., D. M. Rose, M. H. Ginsberg.
1999
. Integrins and the immune response.
Adv. Immunol.
72
:
325
.
4
Metlay, J. P., M. D. Witmer-Pack, R. Agger, M. T. Crowley, D. Lawless, R. M. Steinman.
1990
. The distinct leukocyte integrins of mouse spleen dendritic cells as identified with new hamster monoclonal antibodies.
J. Exp. Med.
171
:
1753
.
5
Henri, S., D. Vremec, A. Kamath, J. Waithman, S. Williams, C. Benoist, K. Burnham, S. Saeland, E. Handman, K. Shortman.
2001
. The dendritic cell populations of mouse lymph nodes.
J. Immunol.
167
:
741
.
6
Salomon, B., J. L. Cohen, C. Masurier, D. Klatzmann.
1998
. Three populations of mouse lymph node dendritic cells with different origins and dynamics.
J. Immunol.
160
:
708
.
7
Ruedl, C., P. Koebel, M. Bachmann, M. Hess, K. Karjalainen.
2000
. Anatomical origin of dendritic cells determines their life span in peripheral lymph nodes.
J. Immunol.
165
:
4910
.
8
Puig-Kroger, A., F. Sanz-Rodriguez, N. Longo, P. Sanchez-Mateos, L. Botella, J. Teixido, C. Bernabeu, A. L. Corbi.
2000
. Maturation-dependent expression and function of the CD49d integrin on monocyte-derived human dendritic cells.
J. Immunol.
165
:
4338
.
9
Kaldjian, E. P., J. E. Gretz, A. O. Anderson, Y. Shi, S. Shaw.
2001
. Spatial and molecular organization of lymph node T cell cortex: a labyrinthine cavity bounded by an epithelium-like monolayer of fibroblastic reticular cells anchored to basement membrane-like extracellular matrix.
Int. Immunol.
13
:
1243
.
10
Gretz, J. E., C. C. Norbury, A. O. Anderson, A. E. Proudfoot, S. Shaw.
2000
. Lymph-borne chemokines and other low molecular weight molecules reach high endothelial venules via specialized conduits while a functional barrier limits access to the lymphocyte microenvironments in lymph node cortex.
J. Exp. Med.
192
:
1425
.
11
Hayakawa, M., M. Kobayashi, T. Hoshino.
1988
. Direct contact between reticular fibers and migratory cells in the paracortex of mouse lymph nodes: a morphological and quantitative study.
Arch. Histol. Cytol.
51
:
233
.
12
Itano, A. A., S. J. McSorley, R. L. Reinhardt, B. D. Ehst, E. Ingulli, A. Y. Rudensky, M. K. Jenkins.
2003
. Distinct dendritic cell populations sequentially present a subcutaneous antigen to CD4 T cells and stimulate different aspects of cell-mediated immunity.
Immunity
19
:
47
.
13
Ingulli, E., D. R. Ulman, M. M. Lucido, M. K. Jenkins.
2002
. In situ analysis reveals physical interactions between CD11b+ dendritic cells and antigen-specific CD4 T cells after subcutaneous injection of antigen.
J. Immunol.
169
:
2247
.
14
Den Haan, J. M., S. M. Lehar, M. J. Bevan.
2000
. CD8+ but not CD8 dendritic cells cross-prime cytotoxic T cells in vivo.
J. Exp. Med.
192
:
1685
.
15
Pooley, J. L., W. R. Heath, K. Shortman.
2001
. Cutting edge: intravenous soluble antigen is presented to CD4 T cells by CD8 dendritic cells, but cross-presented to CD8 T cells by CD8+ dendritic cells.
J. Immunol.
166
:
5327
.
16
Brown, K. A., P. Bedford, M. Macey, D. A. McCarthy, F. LeRoy, A. J. Vora, A. J. Stagg, D. C. Dumonde, S. C. Knight.
1997
. Human blood dendritic cells: binding to vascular endothelium and expression of adhesion molecules.
Clin. Exp. Immunol.
107
:
601
.
17
Ma, J., J.-H. Wang, Y.-J. Guo, M.-S. Sy, M. Bigby.
1994
. In vivo treatment with anti-ICAM-1 and anti-LFA-1 antibodies inhibits contact sensitization-induced migration of epidermal Langerhans cells to regional lymph nodes.
Cell. Immunol.
158
:
389
.
18
Xu, H., H. Guan, G. Zu, D. Bullard, J. Hanson, M. Slater, C. A. Elmets.
2001
. The role of ICAM-1 molecule in the migration of Langerhans cells in the skin and regional lymph node.
Eur. J. Immunol.
31
:
3085
.
19
Price, A. A., M. Cumberbatch, I. Kimber, A. Ager.
1997
. α6 Integrins are required for Langerhans cell migration from the epidermis.
J. Exp. Med.
186
:
1725
.
20
Schneeberger, E. E., Q. Vu, B. W. LeBlanc, C. M. Doerschuk.
2000
. The accumulation of dendritic cells in the lung is impaired in CD18−/− but not in ICAM-1−/− mutant mice.
J. Immunol.
164
:
2472
.
21
Hogquist, K. A., S. C. Jameson, W. R. Heath, J. L. Howard, M. J. Bevan, F. R. Carbone.
1994
. T cell receptor antagonist peptides induce positive selection.
Cell
76
:
17
.
22
Miyake, S., T. Sakurai, K. Okumura, H. Yagita.
1994
. Identification of collagen and laminin receptor integrins on murine T lymphocytes.
Eur. J. Immunol.
24
:
2000
.
23
Noto, K., K. Kato, K. Okumura, H. Yagita.
1995
. Identification and functional characterization of mouse CD29 with a mAb.
Int. Immunol.
7
:
835
.
24
Kivens, W. J., S. W. Hunt, III, J. L. Mobley, T. Zell, C. L. Dell, B. E. Bierer, Y. Shimizu.
1998
. Identification of a proline-rich sequence in the CD2 cytoplasmic domain critical for regulation of integrin-mediated adhesion and activation of phosphoinositide 3-kinase.
Mol. Cell. Biol.
18
:
5291
.
25
Inaba, K., M. Inaba, M. Deguchi, K. Hagi, R. Yasumizu, S. Ikehara, S. Muramatsu, R. M. Steinman.
1993
. Granulocytes, macrophages, and dendritic cells arise from a common major histocompatibility complex class II-negative progenitor in mouse bone marrow.
Proc. Natl. Acad. Sci. USA
90
:
3038
.
26
Ardavin, C., L. Wu, C. L. Li, K. Shortman.
1993
. Thymic dendritic cells and T cells develop simultaneously within the thymus from a common precursor population.
Nature
362
:
761
.
27
Manz, M. G., D. Traver, T. Miyamoto, I. L. Weissman, K. Akashi.
2001
. Dendritic cell potentials of early lymphoid and myeloid progenitors.
Blood
97
:
3333
.
28
Traver, D., K. Akashi, M. Manz, M. Merad, T. Miyamoto, E. G. Engleman, I. L. Weissman.
2000
. Development of CD8α-positive dendritic cells from a common myeloid progenitor.
Science
290
:
2152
.
29
Wu, L., A. D’Amico, H. Hochrein, M. O’Keeffe, K. Shortman, K. Lucas.
2001
. Development of thymic and splenic dendritic cell populations from different hemopoietic precursors.
Blood
98
:
3376
.
30
Vremec, D., K. Shortman.
1997
. Dendritic cell subtypes in mouse lymphoid organs: cross-correlation of surface markers, changes with incubation, and differences among thymus, spleen, and lymph nodes.
J. Immunol.
159
:
565
.
31
Cepek, K. L., S. K. Shaw, C. M. Parker, G. J. Russell, J. S. Morrow, D. L. Rimm, M. B. Brenner.
1994
. Adhesion between epithelial cells and T lymphocytes mediated by E-cadherin and the αEβ7 integrin.
Nature
372
:
190
.
32
Higgins, J. M. G., D. A. Mandlebrot, S. K. Shaw, G. J. Russell, E. A. Murphy, Y. T. Chen, W. J. Nelson, C. M. Parker, M. B. Brenner.
1998
. Direct and regulated interaction of integrin αEβ7 with E-cadherin.
J. Cell Biol.
140
:
197
.
33
Ignatius, M. J., T. H. Large, M. Houde, J. W. Tawil, A. Barton, F. Esch, S. Carbonetto, L. F. Reichardt.
1990
. Molecular cloning of the rat integrin α1-subunit: a receptor for laminin and collagen.
J. Cell Biol.
111
:
709
.
34
Brand, U., I. Bellinghausen, A. H. Enk, H. Jonuleit, D. Becker, J. Knop, J. Saloga.
1999
. Allergen-specific immune deviation from a TH2 to a TH1 response induced by dendritic cells and collagen type I.
J. Allergy Clin. Immunol.
104
:
1052
.
35
Camper, L., U. Hellman, E. Lundgren-Åkerlund.
1998
. Isolation, cloning, and sequence analysis of the integrin subunit α10, a β1-associated collagen binding integrin expressed on chondrocytes.
J. Biol. Chem.
273
:
20383
.
36
Velling, T., M. Kusche-Gullberg, T. Sejersen, D. Gullberg.
1999
. cDNA cloning and chromosomal localization of human α11 integrin: a collagen-binding, I domain-containing, β1-associated integrin α-chain present in muscle tissues.
J. Biol. Chem.
274
:
25735
.
37
Knutson, J. R., J. Iida, G. B. Fields, J. B. McCarthy.
1996
. CD44/chondroitin sulfate proteoglycan and α2β1 integrin mediate human melanoma cell migration on type IV collagen and invasion of basement membranes.
Mol. Biol. Cell
7
:
383
.
38
Dang, N. H., Y. Torimoto, S. F. Schlossman, C. Morimoto.
1990
. Human CD4 helper T cell activation: functional involvement of two distinct collagen receptors, 1F7 and VLA integrin family.
J. Exp. Med.
172
:
649
.
39
Dustin, M. L., T. A. Springer.
1989
. T-cell receptor cross-linking transiently stimulates adhesiveness through LFA-1.
Nature
341
:
619
.
40
Shimizu, Y., G. A. van Seventer, K. J. Horgan, S. Shaw.
1990
. Regulated expression and binding of three VLA (β1) integrin receptors on T cells.
Nature
345
:
250
.
41
Woods, M. L., Y. Shimizu.
2001
. Signaling networks regulating β1 integrin-mediated adhesion of T lymphocytes to the extracellular matrix.
J. Leukocyte Biol.
69
:
874
.
42
Le Varlet, B., M. J. Staquet, C. Dezutter-Dambuyant, P. Delorme, D. Schmitt.
1992
. In vitro adhesion of human epidermal Langerhans cells to laminin and fibronectin occurs through β1 integrin receptors.
J. Leukocyte Biol.
51
:
415
.
43
Staquet, M.-J., B. Levarlet, C. Dezutter-Dambuyant, D. Schmitt.
1992
. Human epidermal Langerhans cells express β1 integrins that mediate their adhesion to laminin and fibronectin.
J. Invest. Dermatol.
99
:
12S
.
44
Shi, J., T. Cinek, K. E. Truitt, J. B. Imboden.
1997
. Wortmannin, a phosphatidylinositol 3-kinase inhibitor, blocks antigen-mediated, but not CD3 monoclonal antibody-induced, activation of murine CD4+ T cells.
J. Immunol.
158
:
4688
.
45
Freiberg, B. A., H. Kupfer, W. Maslanik, J. Delli, J. Kappler, D. M. Zaller, A. Kupfer.
2002
. Staging and resetting T cell activation in SMACs.
Nat. Immun.
3
:
911
.
46
Miller, M. J., S. H. Wei, I. Parker, M. D. Cahalan.
2002
. Two-photon imaging of lymphocyte motility and antigen response in intact lymph node.
Science
296
:
1869
.
47
Miller, M. J., S. H. Wei, M. D. Cahalan, I. Parker.
2003
. Autonomous T cell trafficking examined in vivo with intravital two-photon microscopy.
Proc. Natl. Acad. Sci. USA
100
:
2604
.
48
Moser, M..
2003
. Dendritic cells in immunity and tolerance: do they display opposite functions?.
Immunity
19
:
5
.
49
Geissmann, F., M. C. Dieu-Nosjean, C. Dezutter, J. Valladeau, S. Kayal, M. Leborgne, N. Brousse, S. Saeland, J. Davoust.
2002
. Accumulation of immature Langerhans cells in human lymph nodes draining chronically inflamed skin.
J. Exp. Med.
196
:
417
.