The Role of CCR7 in TH1 and TH2 Cell Localization and Delivery of B Cell Help in Vivo

See allHide authors and affiliations

Science  10 Dec 1999:
Vol. 286, Issue 5447, pp. 2159-2162
DOI: 10.1126/science.286.5447.2159


Subsets of murine CD4+ T cells localize to different areas of the spleen after adoptive transfer. Naı̈ve and T helper 1 (TH1) cells, which express the chemokine receptor CCR7, are home to the periarteriolar lymphoid sheath, whereas activated TH2 cells, which lack CCR7, form rings at the periphery of the T cell zones near B cell follicles. Retroviral transduction of TH2 cells with CCR7 forces them to localize in a TH1-like pattern and inhibits their participation in B cell help in vivo but not in vitro. Thus, differential expression of chemokine receptors results in unique cellular migration patterns that are important for effective immune responses.

Trafficking of cells within secondary lymphoid tissues is carefully orchestrated to ensure that antigen-specific T cells are able to deliver help to antigen-specific B cells (1, 2). Recent studies have shown that chemokines are important in regulating leukocyte trafficking within secondary lymphoid tissues (3). Treatment of lymphocytes with pertussis toxin, a potent inhibitor of chemokine receptor signaling, prevents them from entering the splenic white pulp (4). The chemokines SLC and ELC, which signal through the chemokine receptor CCR7, and the chemokine BLC, which signals through the chemokine receptor CXCR5, are constitutively expressed in secondary lymphoid tissues and seem to be particularly important in establishing normal lymphoid architecture and trafficking patterns (5, 6). CCR7- and CXCR5-deficient mice have disturbed lymphoid architecture and impaired immune responses, as do mice deficient in SLC and BLC production (6–8).

CD4+ helper T lymphocyte subsets differ in their abilities to provide B cell help (9). TH2 cells efficiently provide B cell help, promoting strong humoral responses with class switching to immunoglobulin G1 (IgG1) and IgE. TH1 cells are inefficient at providing B cell help, although they can induce class switching to IgG2a (10). In humans, TH1 and TH2 cells have also been shown to differ in their repertoires of expressed chemokine receptors (11). Human TH1 cells preferentially express CXCR3 and CCR5, whereas TH2 cells preferentially express CCR3 and CCR4.

To determine if CD4+ T cell subsets home to different microanatomic locations within secondary lymphoid tissues, we adoptively transferred undifferentiated (naı̈ve) and antigen-stimulated, in vitro–differentiated ovalbumin (OVA)-specific TH1 and TH2 cells from DO11.10 transgenic mice into BALB/c recipients and immunized the recipients with OVA (12). Two days later, we defined the localization of the transferred cells by immunostaining frozen sections of spleen and popliteal lymph nodes (LNs) (13). Transferred TH1 and naı̈ve cells were concentrated within the periarteriolar lymphoid sheaths (PALS) (Fig. 1, A to C). In contrast, transferred TH2 cells formed loose rings around the outer PALS in close proximity to the B cell zones. The localization patterns in the spleen were the same with and without antigen immunization. Similar patterns were seen at 1, 4, or 8 days after transfer with the exception that transferred TH2 cells were difficult to detect by 8 days in vivo (Web figure 1). In the popliteal LN (Fig. 1, D to F), TH1 and TH2 cells were both found primarily in the outer cortex in the parafollicular areas. Naı̈ve cells were found in similar locations except that they were recruited in larger numbers and were found throughout the medulla as well. In the absence of local antigen, TH1 and TH2 cells were not detected in the LN, and naı̈ve cells were detected in only small numbers. Pretreatment of the T cells with pertussis toxin before transfer disrupted their localization patterns within the spleen and completely prevented migration into the popliteal LN, suggesting a dependence on chemokine receptor signaling (Web figure 2).

Figure 1

TH1, TH2, and naı̈ve CD4+ T cells display different localization patterns within secondary lymphoid tissues. BALB/c mice received an iv infusion of 2 × 107 T cells, and then were immunized with 100 μg of OVA in incomplete Freund's adjuvant ip (Ato C) or in the footpad (D to F). Two days later the spleen and popliteal LNs from each mouse were collected. Frozen sections were stained with the clonotypic antibody KJ1-26 (blue) to identify the transferred cells and with anti-B220 (brown) to identify B cell follicles. Shown are spleens (A to C) and LNs (D to F) of recipients of TH1 cells (A and D), TH2 cells (B and E), and naı̈ve CD4+ T cells (C and F).

Analysis of activated murine TH1 and TH2 cell chemokine receptor expression by ribonuclease protection assays and Northern (RNA) blotting (14) revealed distinct receptor repertoires in the two populations (Fig. 2, A and B). TH1 and TH2 cells expressed similar levels of CCR1, CCR2, and CCR4 mRNA. TH1 cells preferentially expressed CCR7, CXCR3, and CCR5, whereas TH2 cells expressed more CCR3 and CXCR4. Neither cell type expressed detectable BLR-1. Analysis of CCR7 expression in naı̈ve CD4+ T cells showed even higher levels of CCR7 mRNA than in TH1 cells (15). CCR7 expression was unaffected by addition of interleukin-2 (IL-2) to the cultures over a period of up to 14 days (Web figure 3).

Figure 2

Murine TH1 and TH2 cells differentially express the chemokine receptors CCR3, CCR5, CCR7, and CXCR3. RNA was purified from TH1 and TH2 cells 7 days after restimulation with OVA (323–339) peptide. (A) Ribonuclease protection assay with 10 μg of RNA from TH1 or TH2 cells and the Pharmingen RiboQuant template sets mCR-5 (left) and mCR-6 (right). L32 and GAPDH are housekeeping genes to control for equal loading of the RNA. (B) Northern blot analysis. Samples (40 μg) of RNA from TH1 or TH2 cells were analyzed with a 32P-labeled CCR7 probe. The membrane was stripped and reprobed for CXCR3 and then stripped and probed again for β-actin as a control for loading.

Next we tested whether the differences in receptor expression resulted in functional differences by measuring calcium fluxes in response to treatment with chemokines (16). TH1 cells responded to SLC with calcium fluxes in a dose-dependent manner, whereas TH2 cells failed to respond to SLC at any dose tested up to 300 ng/ml (Fig. 3A). To determine if expression of either CCR7 or CXCR3 was sufficient to confer SLC responsiveness, TH2 cells were stably transduced with chemokine receptor genes or vector controls by using a retroviral system (17) and then tested for SLC responsiveness. TH2 cells transduced with CCR7 responded to SLC in a manner similar to control TH1 cells (Fig. 3B). Neither CXCR3-transduced nor control TH2 cells responded to SLC at the concentrations tested, although transduction with CXCR3 did confer responsiveness to the CXCR3 ligand Mig (18).

Figure 3

(A) TH1 cells, but not TH2 cells, respond to the CCR7 ligand SLC. TH1 cells (•) or TH2 cells (○) were loaded with Fura-2 dye and then analyzed by dual-wavelength fluorimetry for increases in intracellular Ca2+ concentration in response to SLC. Data are plotted as the change in bulk intracellular Ca2+ concentration for a given concentration of SLC. (B) Transduction of TH2 cells with CCR7 but not CXCR3 confers SLC responsiveness. TH1 or TH2 cells were transduced with the retroviral vector alone or the vector containing the CCR7 or CXCR3 cDNA. Transduced cells were then loaded with Fura-2 dye and analyzed by fluorimetry for their responsiveness to SLC. Shown are data for TH1 cells transduced with vector (•), TH2 cells transduced with vector (○), TH2 cells transduced with CCR7 (□), and TH2 cells transduced with CXCR3 (▵).

To determine if CCR7 expression was responsible for the differences in localization patterns between TH1 and TH2 cells, we adoptively transferred TH2 cells transduced with the retrovirus encoding CCR7 or the control retrovirus into wild-type BALB/c mice. As before, the mice were immunized intraperitoneally with OVA, and 2 days later the localization patterns of the transferred cells were analyzed in the recipient spleens. Control transduced TH2 cells showed the typical TH2 cell splenic localization pattern, forming rings around the outer PALS near the B cell zones (Fig. 4A). In contrast, TH2 cells expressing CCR7 behaved in a TH1-like manner, clustering within the central PALS (Fig. 4B). The effect was CCR7-specific. Transduction of TH1 cells with CCR3 or CCR8 and transduction of TH2 cells with CXCR3 had no detectable effect on the microanatomic localization of these cells in the spleen. TH2 cells expressing BLR-1 formed clusters of cells within the B cell follicles (19). TH1 cells transduced with BLR-1 remained primarily within the central PALS, suggesting that the action of CCR7 is dominant.

Figure 4

Forced expression of CCR7 in TH2 cells results in a TH1-like splenic localization pattern. TH2 cells were transduced with the retroviral vector alone (A) or with the vector containing the CCR7 cDNA (B). Transduced cells (2 × 107) were adoptively transferred into BALB/c recipient mice. Two days later, frozen sections from spleens of recipients were stained with KJ1-26 antibody (blue) to identify the transferred cells and with B220 (brown) to identify the B cell follicles.

The proximity of the antigen-stimulated TH2 cells to the B cell zones in the spleen together with their known abilities to provide help for B cells suggested that the loss of CCR7 expression on differentiated TH2 cells might be necessary for efficient delivery of B cell help. To test this hypothesis, we transferred B cells purified from NP-BSA [(4-hydroxy-3-nitrophenyl) acetyl–bovine serum albumin]–primed mice into sublethally irradiated naı̈ve recipients together with 105 OVA-specific CCR7-expressing or control TH2 cells (20). The recipients were immunized intraperitoneally with NP-OVA, and 7 days later sera were analyzed for anti-NP IgG1 antibodies. In parallel, we cultured a portion of the primed B cells together with the CCR7-expressing or control TH2 cells in medium containing either NP-OVA or NP-KLH (keyhole limpet hemocyanin). In vitro, where the T cells and B cells were not physically segregated, both CCR7-expressing and control TH2 cells were effective in providing B cell help when the appropriate antigen, NP-OVA, was present (Fig. 5A). Anti-NP IgG1 was not detected when the cells were cultured with NP-KLH. In contrast, CCR7-expressing TH2 cells were impaired in their ability to help B cells in vivo, presumably because the forced expression of CCR7 directed these cells to the inner PALS away from the splenic B cell zones (Fig. 5B). Similar results were obtained when CCR7-expressing or control TH2 cells were transferred to T cell–deficient (TCRβ−/−) mice rather than irradiated wild-type mice (21).

Figure 5

CCR7 expression impairs the ability of TH2 cells to help B cells in vivo but not in vitro. (A) Evaluation of helper function in vitro. Splenic B cells were purified from mice immunized 14 days earlier with 100 μg of NP-BSA in complete Freund's adjuvant. B cells (1 × 106) were then mixed with 105 TH2 cells that had been transduced with the control retrovirus (TH2 vector) or with the CCR7-encoding retrovirus (TH2 CCR7) cells and incubated in a 96-well dish in the presence of either NP-OVA (10 μg/ml) or NP-KLH (10 μg/ml). After 48 hours, the cells were washed and given fresh medium. Five days later the supernatants were collected and anti-NP IgG1 titers were measured by ELISA. (B) Evaluation of helper function in vivo. Primed B cells (1 × 107), purified as in (A), were adoptively transferred to irradiated BALB/c mice either alone or mixed with 105 TH2 cells that had been transduced either with the control retrovirus (vector) or with the CCR7-encoding retrovirus (CCR7). The mice were then immunized with 100 μg of NP-OVA adsorbed to alum. Sera were collected 7 days later and analyzed by ELISA for anti-NP IgG1 antibodies. Statistical comparisons were made by Student's t test.

CCR7 expression levels, therefore, are critical in determining the location and consequently the function of CD4+ T cell subsets within the spleen. Naı̈ve cells expressing CCR7 are retained in the central PALS (Fig. 1C) near interdigitating dendritic cells that are highly effective at presenting antigen to naı̈ve cells. Antigen-stimulated TH2 cells, which are effective at promoting humoral responses, lose CCR7 expression and migrate to the peripheral T cell zones in close proximity to the B cell zones (Fig. 1B). Indeed, retention of the TH2 cells in the PALS by forced expression of CCR7 interrupts delivery of B cell help (Figs. 4and 5). TH1 cells, which are poor B cell helpers, maintain CCR7 expression and are retained in the PALS (Fig. 1A). The functional significance of the TH1 location remains unknown, but it could be important for regulation of cytolytic CD8+ T cells that traffic through the inner PALS (22). Also, the cellular requirements for inducing an IgG2a response are unclear. The microanatomic separation between interferon-γ (IFN-γ)–secreting TH1 cells and B cells in the spleen would seem to prohibit IFN-γ–dependent class switching to IgG2a in this tissue. Interestingly, in the LN both TH1 and TH2 cells are in close proximity to the B cell zones, suggesting that there the TH1 cells may more readily participate in B cell help (Fig. 1, D to F). In the spleen, the location of the TH1 and TH2 cells could also influence the type of antigen-presenting cells encountered. CD11bbright dendritic cells of monocyte lineage have been shown to reside primarily in the marginal zones and outer PALS, whereas CD11bdull, CD8+ dendritic cells of lymphoid origin are believed to migrate preferentially to the central T cell areas (23). Treatment of mice with granulocyte-macrophage colony-stimulating factor, which increases the number of macrophage-related, CD11bbright dendritic cells, increases TH2 responses. Treatment of mice with Flt3-ligand, which increases the number of lymphoid-related dendritic cells, increases TH1 responses (24). Thus, selected aspects of the spleen microenvironment appear specifically adapted to mediate selected immune responses.

  • * To whom correspondence should be addressed at Howard Hughes Medical Institute, Washington University School of Medicine, 660 South Euclid Avenue, Box 8022, St. Louis, MO 63110, USA. E-mail: chaplin{at}


View Abstract

Navigate This Article