Report

Visualization of Specific B and T Lymphocyte Interactions in the Lymph Node

See allHide authors and affiliations

Science  03 Jul 1998:
Vol. 281, Issue 5373, pp. 96-99
DOI: 10.1126/science.281.5373.96

Abstract

Early events in the humoral immune response were visualized in lymph nodes by simultaneous tracking of antigen-specific CD4 T and B cells after immunization. The T cells were initially activated in the T cell areas when the B cells were still randomly dispersed in the B cell–rich follicles. Both populations then migrated to the edges of the follicles and interacted there, resulting in CD154-dependent B cell proliferation and germinal center formation. These results provide visual documentation of cognate T-B cell interactions and localize them to the follicular border.

Antigen-specific T and B lymphocytes are essential for humoral immune responses to many antigens (1). It is known that the T cell dependence of antibody production is related to the efficient capture and internalization of antigen by surface immunoglobulin (Ig), which allows antigen-specific B cells to efficiently present antigenic peptide–major histocompatibility complex (MHC) molecules to antigen-specific T cells (2). The T cells are thus positioned to provide stimulatory surface molecules and cytokines, thereby helping the B cells produce antibodies (3).

It is widely believed that the earliest T-B cell interactions during the primary response occur as a result of the movement of antigen-specific B cells from the B cell–rich follicles into the T cell–rich areas, where they meet antigen-specific T cells (4). However, T-B cell interactions have not been directly observed because of technical difficulties related to simultaneous in situ detection of the rare naı̈ve T and B cells specific for a given antigen.

We addressed this problem by tracking small populations of antigen receptor transgenic T and B lymphocytes after adoptive transfer into normal recipients. CD4 T cells from DO11.10 T cell antigen receptor (TCR) transgenic (BALB/c × C57BL/6)F1 mice (5) specific for chicken ovalbumin (cOVA) (amino acid residues 323 to 339) complexed to I-Ad (6) and MD4 Ig transgenic (BALB/c × C57BL/6)F1 mice with B cells specific for hen egg lysozyme (HEL) (7) were transferred (8) into nontransgenic (BALB/c-Ighb× C57BL/6)F1 animals. The DO11.10 CD4 T cells were identified in the lymph nodes of the recipients by flow cytometry after staining with the anticlonotypic TCR monoclonal antibody (mAb) KJ1-26 (9); labeled HEL or antibody specific for the IgMa (anti-HEL Ig) allotype was used to identify MD4 B cells (10).

cOVA was chemically coupled to HEL to produce an antigen (cOVA-HEL) containing linked epitopes that could be recognized by both transgenic populations (cOVA-HEL) (11). Immunization with cOVA-HEL caused the DO11.10 CD4 T cells in the draining lymph nodes to undergo clonal expansion that was apparent by day 2, was maximal on days 3 and 4, and declined thereafter (Fig. 1A).

Figure 1

Kinetics of CD4 T and B cell clonal expansion. (A) (BALB/c-Ighb × C57BL/6)F1recipients of DO11.10 T cells and MD4 B cells or MD4 B cells alone were injected with cOVA-HEL (11). The mean number ± SD of CD4+ DO11.10 T cells (closed circles) or IgMa+MD4 B cells (open circles) in recipients of DO11.10 T cells and MD4 B cells or IgMa+ MD4 B cells (triangles) in recipients of MD4 B cells alone was measured in the draining lymph nodes at the indicated time points by multiplying the number of viable cells in the lymph nodes by the percentage of DO11.10 T or MD4 B cells present as assessed by flow cytometry (8). (B and C) (BALB/c-Ighb × C57BL/6)F1 recipients of DO11.10 T cells and MD4 B cells were injected subcutaneously with CFA containing 130 μg of cOVA-HEL (circles), 130 μg of tOVA-HEL (triangles), or 100 μg of cOVA plus 130 μg of tOVA-HEL (squares). The mean number ± SD of MD4 B cells (B) or DO11.10 T cells (C) present in draining lymph nodes was measured at the indicated time points after injection of antigen. (D) (BALB/c-Ighb × C57BL/6)F1 recipients of DO11.10 T cells and MD4 B cells were injected with cOVA-HEL (11) on day 0 and intraperitoneally with 250 μg of anti-CD154 mAb (clone MR1) or control hamster Ig on days 0, 2, and 4. The mean number of cells ± SD in mice injected with cOVA-HEL and control Ig (DO11.10, closed circles; MD4, open circles) or with cOVA-HEL and anti-CD154 mAb (DO11.10, closed inverted triangles; MD4, open inverted triangles) are shown for the indicated time points.

MD4 B cell expansion in the lymph nodes was detected 3 days after immunization, peaked on day 4 or 5, and then declined (Fig. 1A). Maximal clonal expansion of the MD4 B cells was dependent on the presence of the DO11.10 T cells, because a much reduced response was observed in mice that received B cells alone (Fig. 1A). The MD4 B cells responded weakly when recipients of DO11.10 T cells and MD4 B cells were immunized with a mixture of cOVA and HEL that had been conjugated to turkey OVA (tOVA-HEL) (Fig. 1B), which lacks the appropriate peptide recognized by the DO11.10 TCR (Fig. 1C). The poor response of the B cells in this situation could not be explained by a failure of T cell activation because the DO11.10 T cells responded well after immunization with cOVA plus tOVA-HEL (Fig. 1C). The simplest explanation for these results is that MD4 B cells were efficiently activated to proliferate in vivo only when expression of anti-HEL Ig allowed efficient uptake of cOVA and presentation of cOVA peptide to DO11.10 T cells.

This conclusion was supported by the finding that blockage of CD154 (CD40 ligand), a surface molecule expressed by activated CD4 T cells that provides a contact-dependent signal to B cells (12), completely inhibited clonal expansion (Fig. 1D) and antibody production (13) by the MD4 B cells. In contrast, treatment with antibody to CD154 (anti-CD154) inhibited clonal expansion of the DO11.10 T cells by only about 50% (Fig. 1D). These findings suggest that the previously described effects of CD154 blockade on antibody production and germinal center formation (12) can be explained by a failure of B cell clonal expansion.

Immunohistochemical analyses were carried out to determine whether the cognate interactions predicted by the functional experiments could be directly observed in situ. After adoptive transfer, but before immunization, the DO11.10 T cells were dispersed throughout the T cell–rich paracortical regions of the lymph node but were not present in the B cell–rich follicles (Fig. 2A). In contrast, the MD4 B cells were localized mainly in follicles (Fig. 2E). One day after immunization, the DO11.10 T cells had not increased in number but were present in small clusters within the paracortex (Fig. 2B). At this time, the MD4 B cells were difficult to detect at the magnification shown (Fig. 2F). However, inspection at a higher magnification revealed that MD4 B cells were still randomly distributed in the follicles, although the intensity of anti-IgMastaining on the MD4 B cells was greatly reduced (compare Fig. 2, E and F insets).

Figure 2

Anatomic localization of DO11.10 T cells and MD4 B cells during the primary response. (BALB/c-Ighb × C57BL/6)F1 recipients of DO11.10 T cells and MD4 B cells were injected with cOVA-HEL (11). Mice were sacrificed 0 (A and E), 1 (B and F), 2 (C and G), or 3 (D and H) days after immunization and the draining lymph nodes were sectioned and stained (27). (A) and (E), (B) and (F), (C) and (G), and (D) and (H) were from adjacent sections. (A to D) KJ1-26+ DO11.10 T cells are stained blue, and B220+ B cells are stained brown to identify the follicles. (E to H) IgMa+ MD4 B cells are stained blue, and Thy 1.2+ T cells are stained brown to identify the T cell–rich paracortical areas. Bar = 50 μm. (Insets) High-power views of areas within the follicles of (E) and (F). Inset scale bars = 5 μm.

Two days after immunization, the DO11.10 T cells began to increase in number in the paracortex and, as described in (14, 15), moved into the edges of the follicles (Figs. 2C and 3A). IgMa was again easily detected on the MD4 B cells, which had increased slightly in number, and redistributed to the edges of the follicles (Figs. 2G and 3B) where the DO11.10 T cells were located (Fig. 3C). About 70% of the DO11.10 T cells were in physical contact with one or more MD4 B cells (Fig. 3D) in the rim of the follicles after injection of cOVA-HEL. The specificity of these interactions was indicated by the finding that only about 10% of the DO11.10 T cells located in the follicular rim were found in physical contact with MD4 B cells after injection of cOVA plus tOVA-HEL (Fig. 3E), a situation in which cognate interactions would not be expected. Few, if any, interactions between DO11.10 T cells and MD4 B cells were observed in the T cell–rich paracortex at this time (Fig. 3C) or at any time (Fig. 2) after injection of cOVA-HEL.

Figure 3

Colocalization of DO11.10 T cells and MD4 B cells at the follicular border 2 days into the primary response. (BALB/c-Ighb × C57BL/6)F1 recipients of DO11.10 T cells and MD4 B cells were injected with cOVA-HEL (11). The draining lymph nodes were harvested on day 2, sectioned, and stained (27). (A) B220+ B cells are stained brown and KJ1-26+ DO11.10 T cells are stained blue. (B) Thy 1.2+ T cells are stained brown and IgMa+ MD4 B cells are stained blue. (C) KJ1-26+ DO11.10 T cells are stained brown and IgMa+ MD4 B cells are stained blue. (A ) to (C ) are from adjacent areas of the same lymph node. Bar = 50 μm. Higher-power views of the edges of lymph node follicles 2 days after subcutaneous injection of CFA containing 130 μg of cOVA-HEL (D) or 100 μg of cOVA plus 130 μg of tOVA-HEL (E) are shown. KJ1-26+ DO11.10 T cells are stained brown and IgMa+ MD4 B cells are stained blue. Scale bar = 25 μm.

Over the next 2 days after immunization with cOVA-HEL, the DO11.10 T cells continued to accumulate in the paracortex and became evenly distributed in the follicles (Fig. 2D). The MD4 B cells proliferated between days 2 and 3 and were present in small clusters in the follicles on day 3 (Fig. 2H) and in peanut agglutinin-positive germinal centers on day 4 (Fig. 4). Large, darkly staining IgMa+ MD4 B cells, presumably plasma cells (16), accumulated within the medulla when germinal centers appeared (Fig. 4A, arrow).

Figure 4

Germinal center formation and the appearance of B cells in the medulla 4 days into the primary response. (BALB/c-Ighb × C57BL/6)F1 recipients of DO11.10 T cells and MD4 B cells were injected with cOVA-HEL (11). The draining lymph nodes were harvested on day 4, sectioned, and stained (27). (A) Thy 1.2+ T cells are stained brown, and IgMa+ MD4 B cells are stained blue. MD4 B cells expressing high levels of IgMa and present in the medulla are indicated with an arrow. (B) Peanut agglutinin–positive cells are stained blue. Scale bar = 50 μm.

By elevating the frequency of antigen-specific CD4 T and B cells, it was possible to define early events in the primary immune response. We found T cells in small clusters in the T cell areas as early as 1 day after immunization. In previous studies, these clusters were found to result from interactions between T cells and dendritic cells of the recipient (16, 17). At this time, the B cells were still randomly distributed throughout the follicles but had lost much of their Ig, probably as a result of internalization and degradation of surface Ig-HEL complexes (18). The lack of physical interactions between antigen-specific T and B cells at this time is evidence that T cell interactions with dendritic cells precede those with B cells even in a situation in which the number of antigen-specific B cells is unlikely to be limiting. This ordered antigen presentation would ensure that T cells are always activated by a dendritic cell first, thus avoiding initial antigen presentation by B cells, which has been shown to delete or inactivate naı̈ve T cells (19).

Elevation of the frequency of antigen-specific cells also allowed identification of cognate T-B cell interactions and localized them to the edge of the B cell–rich follicles 2 days after immunization. The interactions occurred there because both antigen-specific cell types moved toward each other from their separate starting locations. Our studies extend previous findings that antigen-specific B cells in the spleen accumulate at the border between the follicles and the T cell areas in the presence of antigen (20, 21) by showing that antigen-specific helper T cells are physically interacting with the B cells in this location. Furthermore, our results show that cognate interactions in the lymph node initially take place on the B cell side of this border and not on the T cell side as has been widely believed (4). Our results also shed light on the relevance of the foci of antigen-specific B cells that form in the T cell areas of the spleen in response to some antigens (20) and have been suggested to be the source of germinal center B cells (22). Foci were not observed in the T cell areas of the lymph nodes in our studies; all the initial B cell clonal expansion and subsequent germinal center formation occurred in the follicles. Thus, our study and another (23) indicate that T-B cell interactions in extrafollicular foci are not necessary for antibody production and germinal center formation.

The results presented here suggest a model in which antigen-specific B cells move to the edge of the follicles shortly after immunization and present peptide-MHC complexes to specific CD4 T cells that have recently entered this site after activation by dendritic cells. The B cells then receive lymphokines and CD154 signals from the T cells and proliferate in the follicles. Some of these B cells remain in the follicles and become membrane Ig-expressing germinal center cells. Others become antibody-secreting plasma cells and move to the medulla, probably on their way to the bone marrow (24).

  • * These authors contributed equally to this work.

REFERENCES AND NOTES

View Abstract

Navigate This Article