Skip to main content
NCBI home page
Immunology logoLink to Immunology
. 1999 Jul;97(3):438–446. doi: 10.1046/j.1365-2567.1999.00794.x

Different roles for LFA-1 and VLA-4 integrins in T–B-cell interactions in vivo

M López-Hoyos *,‡, C Revilla *,‡,§, C Conde *,§, E G Del Campo *, A González *,§, J Merino †
PMCID: PMC2326849  PMID: 10447765

Abstract

Adhesion molecules are critical in the cellular interactions involved in specific immune responses. They are used for homing, cell migration, cell–cell contact and, in some cases, for the delivery of costimulatory signals. Since the host-versus-graft (HVG) reaction represents a particular form of T–B-cell interaction, we have explored whether the inhibition of lymphocyte function-associated antigen-1/intracellular adhesion molecule-1 (LFA-1/ICAM-1) interactions and the signalling through very late activation antigen-4 (VLA-4) have any effect on the development of a lupus-like disease in BALB/c mice injected at birth with (BALB/c × C57BL/6)F1 spleen cells. In close association with the development of tolerance to donor allografts, these mice show a polyclonal activation of F1 donor B cells by alloreactive host CD4+ T cells, manifested by the production of autoantibodies (autoAbs) and the development of a mild glomerulonephritis. The dose of the monoclonal antibody (mAb) employed has been adjusted to block completely the molecule on the surface of peripheral lymphocytes without interfering with the induction of neonatal tolerance. Injection of saturating doses (100 μg/2 days) of either anti-LFA-1α or anti-ICAM-1 mAbs, but not anti-VLA-4α or anti-LFA-1β mAbs, blocks the production of anti-ssDNA autoAbs and the thrombocytopenia characteristic of this HVG disease (HVGD). However, anti-VLA-4α treatment is only able to delay the production of autoAbs and the anti-LFA-1β treatment, not to modify the evolution of the HVGD. These results point to the relevance of LFA-1/ICAM-1 interactions, but not of the VLA-4-mediated signal, in the polyclonal B-cell activation occurring during the allogeneic interactions between host T helper type 2 cells and donor B cells in HVGD.

INTRODUCTION

Diverse adhesion molecules are employed by leucocytes to leave the bloodstream and move across the interstitial space in response to inflammatory stimuli. They also maintain the contact of T cells with target cells or antigen-presenting cells (APC), thus being critical for cellular interactions involved in specific immune responses and for the delivery of costimulatory signals.3–5

A group of these adhesion molecules comprises the integrin heterodimers which play a central role in the adhesion of leucocytes to other cells and to the extracellular matrix. Two particular integrins, very late activation antigen-4 (VLA-4; α4β1) and lymphocyte function-associated antigen-1 (LFA-1; αLβ2), have been involved also in the transduction of costimulatory signals in T cells.6 VLA-4 mediates homotypic aggregation of T cells as well as binding of T cells to B cells and target cells.7 This β1 integrin plays an essential role in the homing of activated T cells to inflamed tissues.8 Moreover, signalling through VLA-4 can enhance the proliferation and cytokine production of T cells induced by anti-CD3/T-cell receptor (TCR) antibodies9 and participates in the activation of B cells triggered through the B-cell receptor and the CD40/CD40 ligand (CD40L) interaction.10 Likewise, binding of VLA-4 with its ligands has been considered necessary for the positive selection of thymocytes.11

The β2 integrin LFA-1 binds to three different molecules of the immunoglobulin superfamily: intracellular adhesion molecules -1, -2 and -3 (ICAM-1, ICAM-2 and ICAM-3).2 Three observations emphasize the role of LFA-1/ICAM-1 interaction in T-cell activation: first, an active form of LFA-1α is rapidly and transiently induced after TCR cross-linking;12 second, a LFA-1/ICAM-1 signal is induced after an antigen-specific T-cell activation;13 and third, the treatment with anti-LFA-1 monoclonal antibodies (mAbs) inhibits allogeneic T-cell proliferative responses.14 On the other hand, it has been shown that both, the clonal deletion and maturation of thymocytes, require the LFA-1/ICAM-1 interaction.15 Besides, it has been reported that the disruption of the LFA-1/ICAM-1 interaction with non-cytolytic mAbs induces a prolonged acceptance of heart16 or kidney allografts.17 Similarly, treatments with mAbs against these molecules have therapeutic effects in several experimental models of T-cell-dependent autoimmune diseases, like experimental autoimmune encephalomyelitis.8

In the present study we have explored the possible role of LFA-1 and VLA-4 integrins in the allogeneic T–B-cell interaction responsible for the host-versus-graft disease (HVGD) that occurs after the induction of neonatal tolerance to alloantigens by the injection of semiallogeneic lymphoid cells into parental newborn mice.18 In these chimeras host T helper type 2 (Th2) cells, reactive with allogeneic major histocompatibility complex (MHC) class II antigens, are refractory to the induction of tolerance and polyclonally stimulate donor B cells to produce several autoantibodies (autoAbs). These mice also develop thrombocytopenia, lymphosplenomegaly and a transient glomerulonephritis mediated by immune complex deposition.20 In this model we have described that the treatment of newborn mice with high doses of an anti-LFA-1α mAb abrogates the induction of neonatal tolerance to alloantigens,21 preventing us from evaluating the effects of anti-LFA-1α mAb treatment on the allogeneic T–B-cell interaction responsible for the autoAb production. However, the possible role of LFA-1 molecule in allogeneic interactions has been suggested by the high expression of this molecule in CD3+ cells from tolerant mice.21

In order to analyse the possible role of LFA-1/ICAM-1 and VLA-4 integrins in allogeneic T–B cell interactions in vivo, BALB/c mice neonatally injected with spleen cells from (C57BL/6 × BALB/c)F1 hybrid mice (CB6F1), were treated from birth with saturating doses of non-cytolytic anti-VLA-4α, anti-LFA-1α, anti-LFA-1β, or anti-ICAM-1 mAbs. These mAbs were used at doses that did not interfere with the induction of neonatal tolerance to alloantigens.

MATERIALS AND METHODS

Mice and induction of neonatal tolerance

BALB/c (H-2d) and C57BL/6 (H-2b) mice were purchased from CRIFFA (Barcelona, Spain). (C57BL/6 × BALB/c) F1 hybrid mice (CB6F1) were bred in our animal facilities. Neonatal tolerance to H-2b alloantigens was induced by intraperitoneal (i.p.) injection of 108 spleen cells from CB6F1 hybrid mice into BALB/c newborn mice within the first 24 hr after birth.18

In vivo treatment with mAbs

Groups of BALB/c mice, injected at birth with CB6F1 spleen cells, were injected i.p. with one of the following mAbs: H35-89.9 [rat immunoglobulin G (IgG2b) anti-LFA-1α; provided by Dr M. Pierres, INSERM-CNRS, Marseille-Luminy, France], PS/2 (rat IgG2b anti-mouse VLA-4α), M18/2 (rat IgG2a anti-LFA-1β) or YN1/1.7 (rat IgG2b anti-ICAM-1) (American Type Culture Collection, Rockville, MA). The mAbs were prepared from ascitic fluid obtained in pristane-primed nude mice. The injection of mAbs was started on the first day of life and repeated every 2 days up to the 15th day. Anti-LFA-1α and anti-ICAM-1 mAbs were administered at doses of 200 μg, 100 μg, or 50 μg/2 days, whereas anti-VLA-4α and anti-LFA-1β mAbs were used at doses of 100 μg or 25 μg/2 days. These doses and days of mAb administration were established on the basis of membrane saturation of each molecule, analysed by flow cytometry as described below. Two additional groups of BALB/c mice injected at birth with CB6F1 cells were treated with 100 μg of anti-LFA-1α or anti-VLA-4α every 2 days up to the 6th week of age.

Flow-cytometry studies

Double and triple cell staining was performed on single cell suspensions from spleen and thymus. Cells were stained with appropriate dilutions of purified ascites containing mAbs, either PS/2, H35-89.9, M18/2, or YN1/1.7, together with mAbs conjugated to phycoerythrin (PE) and recognizing either murine B220 or CD3, and with anti-IAb mAb coupled to biotin. Rat IgG2a (R35-95) or rat IgG2b (R35-38) (Pharmingen, San Diego, CA) were used as isotype controls. As second-step reagent, goat F(ab′)2 anti-rat immunoglobulin–fluorescein isothiocyanate (FITC) (Tago, Burlingame, CA) or streptavidin–RED 670 (Gibco BRL, Rockville, MD) were used.

The saturation of VLA-4, LFA-1, or ICAM-1 molecules in the thymus and the spleen of BALB/c mice injected at birth with CB6F1 cells was assessed 48 hr after the end of the treatment. To this end, cell suspensions were stained either with the mAb employed for the treatment followed by FITC-conjugated goat F(ab′)2 anti-rat immunoglobulin or only with the second antibody. As a control, thymocytes and splenocytes from unmanipulated BALB/c mice were stained with those mAbs following the same protocol.

The presence of chimerism was evaluated by staining splenic cell suspensions with a FITC-conjugated anti-H-2Kb mAb (clone AF6-88.5) (Pharmingen, San Diego, CA).

To analyse the effect of the different treatments on the frequencies of lymphocyte subsets, single cell suspensions from spleen and lymph nodes were stained with FITC-conjugated anti-CD3, anti-B220 and anti-CD4 mAbs and with PE-conjugated anti-CD8 mAbs (Pharmingen).

After staining, flow cytometry analyses were performed in a fluorescence-activated cell analyser (FACScan; Becton Dickinson, Palo Alto, CA). Dead cells were excluded by scatter and forward parameters.

Generation of cytotoxic T lymphocytes in mixed lymphocyte cultures

Mixed lymphocyte cultures (MLC) were established as previously described;22 2 × 106 responding spleen cells were mixed with 2 × 106 irradiated spleen cells from CB6F1 mice in 2 ml of Dulbecco’s modified Eagle’s medium supplemented with amino acids, 10 mm HEPES, 2-mercaptoethanol and 5% fetal calf serum (FCS). Cytolytic activity of MLC cells was measured 5 days later by the 51Cr-release assay, with the MBL-2.9 (H-2b) and BW-5147 (H-2k) tumour lines as target cells.22 Cytolytic activity was measured at multiple leucocyte:target cell ratios and was quantified as percentage of specific lysis at a leucocyte:target cell ratio of 100:1.

Analysis of autoimmune manifestations

Anti-ssDNA IgG antibodies were measured in the sera obtained at different ages by enzyme-linked immunosorbent assay (ELISA) and the results were expressed in titration units (TU) referred to a standard curve obtained by serial dilution of a serum pool from 6- to 8-month-old MRL lpr/lpr mice.18 Platelets were counted in an improved Neubauer haemocytometer under phase contrast at 400 × magnification

RESULTS

Expression of LFA-1, ICAM-1 and VLA-4 molecules in cells involved in the allogeneic T–B-cell interaction

The significance of LFA-1, ICAM-1 and VLA-4 in the allogeneic interaction between donor B cells and host CD4+ T cells occurring in BALB/c mice neonatally injected with 108 spleen cells from CB6F1 mice was firstly evaluated by measuring their expression by flow cytometry. Spleen cells were obtained from tolerant mice at 3 weeks of age and stained with mAbs against LFA-1, ICAM-1 and VLA-4. Donor B cells (B220+ Iab+) presented a marked increase in the fluorescence intensity of LFA-1, ICAM-1 and VLA-4 molecules when compared to host B cells (B220+ Iab−) (Fig. 1), which showed similar levels to B cells from BALB/c controls (not shown). A slight increase in the expression of LFA-1 and ICAM-1 in T cells was observed in neonatally injected mice in comparison to unmanipulated controls (not shown), although the comparison between host (CD3+ H-2Kb−) and donor (CD3+ H-2Kb+) cells was impaired by the extremely low frequency of donor T cells. The expression of VLA-4 in T cells from BALB/c mice injected with CB6F1 cells was similar to that of T cells in uninjected controls.

Figure 1.

Figure 1

Open in a new tab

Spleen cells were obtained from tolerant mice at 3 weeks of age and stained with mAbs against LFA-1α, ICAM-1 and VLA-4 together with anti-B220-PE and anti-IAb-biotin mAbs. Expression of VLA-4, LFA-1α and ICAM-1 was analysed by flow cytometry by gating on host B cells (B220+ IAb−, thin line) and donor B cells (B220+ IAb+, thick line). Results are representative of three independent experiments. Shaded areas represent the rat IgG2b isotypic control. Mean of the mean fluorescence intensities (B220+ IAb− cells versus B220+ IAb+ cells) obtained are also indicated in the upper right corner of each histogram.

Effects of the treatment with anti-LFA-1, -ICAM-1 and -VLA-4 mAbs on membrane saturation

Groups of BALB/c mice, injected at birth with CB6F1 spleen cells were treated for 2 or 6 weeks with 50, 100, or 200 μg/2 days of anti-LFA-1α, -LFA-1β, or -ICAM-1 mAbs, or with 25 or 100 μg/2 days of anti-VLA-4α mAb. The saturation of the different molecules with the corresponding mAbs was investigated by flow cytometry in lymph node and spleen cells 2 days after the end of the treatment. More than 95% of lymph node cells (Fig. 2) or spleen cells (not shown) from mice treated with 100 μg/2 days of anti-LFA-1α, -LFA-1β, or -ICAM-1 rat mAbs were positive when the cells were only labelled with anti-rat immunoglobulin-FITC antibody, and no significant increase in the fluorescence signal was seen after double incubation of cells with the corresponding mAb employed for the treatment and the second-stage reagent (Fig. 2). It should be noted that the expression of these molecules in mAb-treated mice (100 μg/2 days) was four- to sixfold lower (for LFA-1) or five- to eightfold lower (for ICAM-1) than in untreated BALB/c mice. No significant differences in the saturation of these molecules were found when mice were treated with 50 or 200 μg/2 days of these mAbs.

Figure 2.

Figure 2

Open in a new tab

The efficacy of the anti-LFA-1α (100 μg/2 days), -ICAM-1 (100 μg/2 days), -LFA-1β (200 μg/2 days) and VLA-4α (100 μg/2 days) mAbs treatments to saturate the corresponding molecules was investigated by flow cytometry in lymph node cells 48 hr after stopping the treatment (at 2 weeks of age). Cell suspensions were incubated with anti-LFA-1α (upper left), anti-ICAM-1 (upper right), anti-LFA-1β (lower left) and VLA-4α (lower right) mAbs and anti-rat immunoglobulin–FITC as second-step reagent. Cells from BALB/c control mice were stained with the second-step reagent only (A: brighter thin line) or with the first and the second antibodies (B: brighter thick line). Cells from CB6F1 cell-injected, treated with the mAbs, were stained with the second antibody only (C: darker thin line) or with the first and the second antibodies (D: darker thick line). These results are representative of three separate experiments.

BALB/c mice neonatally injected with CB6F1 cells were also treated with anti-VLA-4α mAb. In animals treated with 100 μg/2 days of anti-VLA-4α mAb for 2 weeks (Fig. 2) or 6 weeks (not shown) a complete absence of VLA-4+ cells was observed. To analyse whether the absence of VLA-4+ cells in these treated mice was due either to cell depletion or to the modulation of the VLA-4 molecule by the treatment, frequencies of lymphocyte populations in peripheral lymphoid organs were studied by flow cytometry. We observed similar percentages of the different lymphocyte subsets between treated and untreated mice. Percentages of B cells in lymph nodes and spleen were, respectively, 19·3±7·5 and 44·1±6·2 in treated mice versus 25·2±8·0 and 40·8±6·9 in untreated BALB/c mice. Frequencies of CD3+ cells in lymph nodes and spleen were 65·2±12·5 and 42·0±3·8 versus 75·8±5·4 and 30·2±3·4, respectively. Likewise, no differences were observed for CD4+ and CD8+ subpopulations (not shown). Moreover, with lower doses of anti-VLA-4α mAb (25 μg/2 days), only saturation but not modulation of VLA-4 expression was seen (not shown).

High doses of anti-LFA-1α and ICAM-1 mAbs, but not of anti-LFA-1β or anti-VLA-4α, inhibited the induction of neonatal tolerance to alloantigens

To analyse the induction of tolerance to alloantigens in BALB/c mice injected at birth with CB6F1 spleen cells and treated with anti-LFA-1, -ICAM-1, or -VLA-4 mAbs, the presence of cellular chimerism was analysed at 3 and 20 weeks. Tolerized mice treated with 100 μg/2 days of anti-LFA-1α, -LFA-1β, -ICAM-1, or -VLA-4 α mAbs showed a degree of chimerism, characterized by the presence of H-2Kb+ cells in their spleen, similar to untreated BALB/c mice tolerized with CB6F1 cells. Similar results were observed when the treatment with these mAbs was prolonged for 6 weeks (not shown). However, in BALB/c mice injected at birth with CB6F1 spleen cells and treated with 200 μg/2 days of either anti-LFA-1α or -ICAM-1 mAbs the presence of H-2Kb+ cells in spleen and lymph nodes decreased at 3 weeks (Table 1), and was undetectable at 20 weeks of age (not shown).

Frequency of H-2Kb+ donor cells in lymphoid organs of F1 cell-injected mice

graphic file with name imm0097-0438-t1.jpg

Open in a new tab

Similar results were observed when the induction of neonatal tolerance to H-2Kb alloantigens was analysed in vitro (Fig. 3). Spleen cells from BALB/c mice neonatally injected with CB6F1 spleen cells and treated with 200 μg/2 days of either anti-LFA-1α or -ICAM-1 mAbs for the first 2 weeks of age exhibited a cytolytic reactivity in vitro to MBL-2.9 (H-2b) target cells comparable to that observed with spleen cells from unmanipulated BALB/c mice. However, spleen cells from BALB/c mice injected at birth with CB6F1 cells and treated with 100 μg/2 days of either anti-LFA-1α or -ICAM-1 mAbs or with 200 μg/2 days of anti-LFA-1β mAb for 2 weeks did not show any cytolytic activity as observed for untreated BALB/c mice injected with CB6F1 cells (Fig. 3). An absence of cytolytic activity was also observed in BALB/c mice injected with CB6F1 cells treated with 100 μg/2 days of anti-VLA-4α mAb (not shown). As control, all the groups of mice showed cytolytic activity against the tumour cell line BW-5147 (H-2k) (not shown).

Figure 3.

Figure 3

Open in a new tab

BALB/c mice were i.p. injected at birth with 108 spleen cells from CB6F1 hybrid mice (•). Some of these mice were treated every other day from birth up to 2 weeks of age with one of the following antibodies: 100 μg of anti-LFA–1α mAb (□); 200 μg of anti-LFA–1β mAb (✦); 100 μg of anti-ICAM-1 mAb (▾); 200 μg of anti-LFA-1α mAb (▪); or 200 μg of anti-ICAM-1 mAb (▴). Unmanipulated BALB/c mice were used as controls (○).To assess the establishment of tolerance, the in vitro CTL alloreactivity against MBL-2.9 (H-2b) target cells was analysed in all groups of mice at 10 weeks of age by a standard 5 hr 51Cr-release assay. These results are representative of three separate experiments.

The saturation of LFA-1 on thymocytes was proportional to the dose of anti-LFA-1 mAb

The existence of a correlation between the dose of anti-LFA-1α mAb administered and the establishment of neonatal tolerance to alloantigens, despite the presence of a complete saturation of the LFA-1 molecule in the periphery with all the doses employed, prompted us to evaluate whether the LFA-1 molecule was differentially saturated on thymocytes with the different treatment conditions. To this purpose the saturation of the LFA-1 molecule was analysed at 5, 9 and 20 days of age in thymocytes from BALB/c mice treated from birth with 50, 100, or 200 μg/2 days of anti-LFA-1α mAb. At 5 days, the saturation of LFA-1 on thymocytes was complete with 100 and 200 μg of mAb, but not with 50 μg of mAb (Fig. 4). At 9 days, none of these doses reached a complete saturation of LFA-1, although a significant gradient of saturation was observed in relation with the dose employed, being near complete with 200 μg of mAb (Fig. 4). At 20 days thymocytes from mice injected with 50 or 100 μg of anti-LFA-1α mAb showed a complete absence of LFA-1 saturation and only thymocytes from mice receiving 200 μg of anti-LFA-1α mAb were partially saturated (not shown). In a previous report23 we have shown that treatment with 400 μg/2 days of anti-LFA-1α mAb completely saturated this molecule on thymocytes at 15 days of age.

Figure 4.

Figure 4

Open in a new tab

Saturation of LFA-1 on thymocytes from CB6F1 cell-injected mice treated with different doses of anti-LFA-1α mAb was analysed at 5 and 9 days of age by flow cytometry (a). Thymic cell suspensions from these mice were stained with either anti-LFA-1α mAb and anti-rat immunoglobulin–FITC (D: darker thick line) or only the second antibody (C: darker thin line). Thymic cells from BALB/c control mice were stained with the second antibody only (A: brighter thin line) or with the first and the second antibodies (B: brighter thick line). Similar staining was performed for splenic cells (b) at 9 days of age in CB6F1 cell-injected mice treated with anti-LFA-1α mAb. Results are representative of three independent experiments.

Effects of the treatment with anti-VLA-4α, anti-LFA-1α, anti-LFA-1β and anti-ICAM-1 on the autoimmune manifestations in BALB/c mice neonatally injected with CB6F1 cells

The production of IgG anti-ssDNA autoAbs and the development of thrombocytopenia, two hallmarks of the HVGD in mice rendered tolerant to alloantigens,20 were explored in all groups of BALB/c mice injected at birth with CB6F1 spleen cells and subsequently treated or not with the different mAbs employed in the present work (Fig. 5). Titres of IgG anti-ssDNA autoAbs reached their highest levels at 3 and 6 weeks of age in untreated CB6F1 cell-injected mice and, then, decreased progressively. Similarly, the levels of platelets dropped at 3 weeks of age and recovered afterwards.

Figure 5.

Figure 5

Open in a new tab

BALB/c mice were i.p. injected at birth with 108 spleen cells from CB6F1 hybrid mice (•). Some of these mice were treated every other day from birth up to either 2 or 6 weeks of age with one of the following mAbs: 100 μg of anti-LFA-1α mAb (□); 200 μg of anti-LFA-1β mAb (✦); 100 μg of anti-ICAM-1 mAb (▾); 200 μg of anti-ICAM-1 mAb (▴); or 100 μg of anti-VLA-4α mAb (★;). Unmanipulated BALB/c mice were used as controls (○). Serum levels of anti-ssDNA IgG antibodies from mice treated up to either 2 (a) or 6 (b) weeks of age were determined by ELISA methods. Platelet numbers in mice treated during 2 weeks with these mAbs are represented in (c). The duration of treatments is indicated on the x-axis. Results express the mean±SEM of data from eight to 10 mice per group.

The treatment for 2 weeks with 100 μg/2 days of anti-LFA-1α or anti-ICAM-1 mAbs inhibited the production of IgG anti-ssDNA and the thrombocytopenia in tolerant BALB/c mice at 3 weeks of age. However, these mice showed a delayed appearance of these autoAbs and thrombocytopenia after 6 weeks of age (Fig. 5). Furthermore, the number of mice with thrombocytopenia was reduced (five of 11 and five of eight in tolerant mice treated with 100 μg/2 days of either anti-LFA-1α or -ICAM-1 mAbs, respectively, in comparison to 10 of 12 untreated BALB/c mice tolerized with CB6F1 cells). This delayed anti-ssDNA IgG production was prevented if the treatment with anti-LFA-1 or -ICAM-1 mAbs was prolonged up to 6 weeks of age. (Fig. 5). Similar results were observed in tolerant mice treated with 50 μg/2 days of either anti-LFA-1α or anti-ICAM-1 mAbs (not shown). Treatment of tolerant mice with 200 μg/2 days of anti-LFA-1β did not modify either the evolution of anti-ssDNA IgG or thrombocytopenia, as compared to untreated tolerant mice.

Anti-ssDNA autoAbs and thrombocytopenia were absent in non-tolerant BALB/c mice injected with CB6F1 cells and treated for 2 weeks with 200 μg/2 days of either anti-LFA-1α or -ICAM-1 mAbs (Fig. 5).

In contrast to the treatment of tolerant mice with anti-LFA-1α and anti-ICAM-1 mAbs, the treatment with anti-VLA-4α mAb, either with 100 μg/2 days (Fig. 5) or 25 μg/2 days (not shown) for 2 or 6 weeks, did not inhibit the anti-ssDNA autoAb production, although the levels of these autoAb decreased approximately to one half of those found in untreated tolerant mice.

DISCUSSION

The results presented here show that the injection of either anti-LFA-1α or anti-ICAM-1 mAbs, but not of anti-VLA-4α or anti-LFA-1β mAb, blocks the development of the murine HVGD associated to the induction of neonatal tolerance to alloantigens after injection of semiallogeneic spleen cells into parental neonates. This effect is observed using doses of mAbs that saturate the corresponding adhesion molecule in the peripheral lymphoid organs without interfering with the induction of neonatal tolerance to alloantigens.

We have shown previously that the injection of semiallogeneic F1 cells into parental newborn mice induces a state of specific tolerance to the corresponding alloantigens, that is manifested by the absence of cytotoxic T lymphocyte alloreactivity and by the acceptance of allografts.18 However, in this model the induction of tolerance does not seem to be complete since chimeric donor B cells are polyclonally activated by alloreactive host CD4+ T cells belonging to the Th2 functional subset.22 This allogeneic T–B interaction is mediated by the recognition of allogeneic donor MHC class II alloantigens either of the I-A or the I-E locus and it is only elicited when B cells bear semiallogeneic, but not completely allogeneic, MHC class II molecules.26 By contrast, MHC class I or non-MHC alloantigens are not able to trigger this allogeneic T–B-cell interaction.

During the last years much progress has been made in the knowledge of the mechanisms involved in the cellular interactions leading to autoAb production. Besides the recognition of MHC molecules bearing peptides on the APCs by the TCR–CD3 complex on T cells, other costimulatory signals, such as those delivered by interaction of the T-cell membrane molecules CD28, LFA-1, or CD2 with their corresponding ligands on the APC (B7, ICAM-1, LFA-3), have been involved in T–B-cell interaction. Thus, T–B-cell interactions in the absence of costimulatory signals through CD28/B7 lead to cell death or anergy instead of stimulation.28 Similarly, costimulatory signals via CD40/CD40L, which are short in time, have been implicated in the production of antibodies.29 Several adhesion molecules used by leucocytes for homing, such as the pair LFA-1/ICAM-1, are able to potentiate the immune response to antigen–MHC complexes or anti-CD3 mAbs.12–14 In fact, interference with that interaction blocks the T-cell activation by conventional antigens.30

Allorecognition represents a particular form of T-cell activation31 in which the LFA-1/ICAM-1 interaction seems to be also necessary for cytolytic allogeneic reactions, as manifested by the improvement of allograft survival after treatment with saturating doses of anti-LFA-1 or anti-ICAM-1 mAbs in heart and kidney transplantation. Our data show that the binding of LFA-1 with its ligands is also required for the establishment of T-and B-cell allohelper interactions. Indeed, LFA-1 and ICAM-1 molecules are up-regulated on the cells that participate in the allogeneic interaction in neonatally tolerized mice. Moreover, saturation of either LFA-1 or ICAM-1 molecules inhibits the production of autoAbs and the associated thrombocytopenia in these mice. This inhibitory effect on the LFA-1 molecule is specific of the α-subunit, since the saturation of the β-subunit has no effect on the autoimmune syndrome. These results, according to other authors,32 indicate clearly the relevance of the LFA-1/ICAM-1 interaction in the T-cell-dependent polyclonal activation of B cells during the HVGD. The inhibition of the HVGD in tolerant mice treated for 2 weeks with anti-LFA-1α or anti-ICAM-1 mAbs may be secondary to the induction of effective tolerance in the compartment of alloreactive Th2 cells. Nevertheless, the presence of autoimmune manifestations in these mice at 6 weeks of age (4 weeks after the end of the treatment) argues against this possibility. Indeed, when the treatment with anti-LFA-1α or -ICAM-1 mAbs in tolerant mice is prolonged up to 6 weeks of age, the above mentioned autoimmune manifestations never appear.

Some evidence suggests that the LFA-1/ICAM-1 interaction is necessary for the establishment of neonatal tolerance to alloantigens. In fact, in this article and in a previous report21 we have shown that high doses of anti-LFA-1α or -ICAM-1 mAbs are able to interfere with the induction of neonatal tolerance. Clearly, this phenomenon is dependent on the dose of mAb used and may be related to the different availability of these mAbs in the different lymphoid organs. In fact, at low doses of circulating mAbs the existence of the so-called blood–thymus barrier33 impedes the entry of the mAbs into the thymus. However, when the dose of mAb administered increases, the level of circulating mAb can be sufficiently high to allow its passage through the thymic barrier and its efficient saturation of the thymocytes.

An additional, and non-exclusive, possibility to explain the different effects of low and high doses of anti-LFA-1 and anti-ICAM-1 mAbs in the regulation of neonatal tolerance induction and the autoimmune syndrome is that the LFA-1 and ICAM-1 molecules are differently saturated in the APC or effector cells involved in this process. Previous data show that the treatment with anti-interleukin-4 (IL-4) mAbs interferes with induction of neonatal tolerance to alloantigens in mice injected at birth with semiallogeneic F1 cells.34 These results suggest that activation of alloreactive host Th2 cells may be essential for the maintenance of cytolytic unresponsiveness to donor cells, and emphasize the role of APC subsets in the regulation of Th cell differentiation. In this regard, dendritic cells drive the differentiation of naive T cells to either Th1 or Th2 subsets depending on the presence of interleukin-12 (IL-12) or IL-4, respectively.35 Since the production of IL-12 during the perinatal period is very deficient,36 it is possible that dendritic cells from host or donor origin in the newborn mice injected with semiallogeneic F1 splenocytes, may present allopeptides to recipient naive Th0 cells inducing their preferential differentiation to Th2 cells. In agreement with this, it has been demonstrated that the injection of IL-12 interferes with the induction of neonatal tolerance to alloantigens.37 Thus, it is possible that high doses of anti-LFA-1 mAb block the APC–T-cell interaction impairing the induction of tolerance, whereas low doses allow the differentiation to Th2 cells but block the production of autoAbs by interfering with the T–B-cell interaction. Clearly, the evaluation of this possibility merits further investigations.

There are not conclusive results regarding the role of β1 integrins as costimulatory molecules in T–B-cell interactions. The treatment with anti-VLA-4 mAb has a therapeutic effect on murine models of autoimmune diseases, such as experimental allergic encephalomyelitis8 or diabetes.38 Nevertheless, is not clear whether this effect can be attributed to the inhibition of leucocyte homing or to the inhibition of T-cell–APC interaction. In the case of the HVGD model, the expression of VLA-4 is up-regulated in donor B cells, suggesting a role for this integrin in the activation of these cells by host alloreactive Th2 cells. This possibility has been evaluated by the treatment of BALB/c mice tolerized at birth with CB6F1 spleen cells with 100 μg/2 days of anti-VLA-4α mAb. This treatment completely modulates the expression of the VLA-4 molecule, probably by internalization or shedding of the target antigen, and does not induce a depletion of cells bearing VLA-4 since the frequencies of B and T cells in the lymphoid organs of treated mice are essentially normal. Our present results demonstrate that the treatment with anti-VLA-4 does not inhibit the induction of neonatal tolerance to alloantigens and the associated HVGD, clearly indicating that VLA-4 is not an essential signal in the allogeneic T–B-cell interactions that take place in this model.

In conclusion, our results indicate that signalling through LFA-1 and ICAM-1, but not through VLA-4, plays an important role in the polyclonal B-cell activation mediated by the allogeneic interaction between host Th2 cells and donor B cells. The analysis of the precise mechanism by which these molecules function, either mediating cellular adhesion and/or delivering costimulatory signals, requires further investigation. In addition, our results suggest that this type of approach may be useful for the treatment of autoimmune diseases.

Glossary

Abbreviations

APC

antigen-presenting cell

CB6F1

(BALB/c × C57BL/6)F1

HVGD

host-versus-graft disease

LU

lytic units

mAb

monoclonal antibody

MLC

mixed lymphocyte culture

ssDNA

single-stranded DNA

TU

titration units

References

  • 1.Hogg N, Landis RC. Adhesion molecules in cell interactions. Curr Opin Immunol. 1993;5:383. doi: 10.1016/0952-7915(93)90057-y. [DOI] [PubMed] [Google Scholar]
  • 2.Hynes RO. Integrins: versatility, modulation, and signaling in cell adhesion. Cell. 1992;69:11. doi: 10.1016/0092-8674(92)90115-s. [DOI] [PubMed] [Google Scholar]
  • 3.Damle NK, Klussman K, Linsley PS, Aruffo A. Differential costimulatory effects of adhesion molecules B7, ICAM-1, LFA-3, and VCAM-1 on resting and antigen-primed CD4+ T lymphocytes. J Immunol. 1992;148:1985. [PubMed] [Google Scholar]
  • 4.Shimizu Y, Van Seventer GA, Horgan KJ, Shaw S. Costimulation of proliferative responses of resting CD4+ T cells by the interaction of VLA-4 and VLA-5 with fibronectin or VLA-6 with laminin. J Immunol. 1990;145:59. [PubMed] [Google Scholar]
  • 5.Mcintyre BW, Woodside DG, Caruso DA, et al. Regulation of human T lymphocyte coactivation with an a4 integrin antagonist peptide. J Immunol. 1997;158:4180. [PubMed] [Google Scholar]
  • 6.Clark AE, Brugge JS. Integrins and signal transduction pathways: the road taken. Science. 1995;268:233. doi: 10.1126/science.7716514. [DOI] [PubMed] [Google Scholar]
  • 7.Bernarczyk JL, Mcintyre BW. A monoclonal antibody to VLA-4 alpha-chain (Cdw49d) induces homotypic lymphocyte aggregation. J Immunol. 1990;144:777. [PubMed] [Google Scholar]
  • 8.Yednock TA, Cannon C, Fritz LC, Sanchez-Madrid F, Steinman L, Karin N. Prevention of experimental autoimmune encephalomyelitis by antibodies against α4β1 integrin. Nature. 1992;356:63. doi: 10.1038/356063a0. [DOI] [PubMed] [Google Scholar]
  • 9.Udagawa T, Mcintyre BW. A VLA-4 alpha-chain specific monoclonal antibody enhances CD3-induced IL-2/IL-2 receptor-dependent T-cell proliferation. Lymphokine Cytokine Res. 1992;5:193. [PubMed] [Google Scholar]
  • 10.Silvy A, Altevogt P, Mondiere P, Bella C, Defrance T. A role for VLA-4 integrin in the activation of human memory B cells. Eur J Immunol. 1997;27:2757. doi: 10.1002/eji.1830271103. [DOI] [PubMed] [Google Scholar]
  • 11.Salomon DR, Mojcik CF, Chang AC, et al. Constitutive activation of integrin alpha 4 beta 1 defines a unique stage of human thymocyte development. J Exp Med. 1994;179:1573. doi: 10.1084/jem.179.5.1573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Dustin ML, Springer TA. T-cell receptor cross-linking transiently stimulates adhesiveness through LFA-1. Nature. 1989;341:619. doi: 10.1038/341619a0. [DOI] [PubMed] [Google Scholar]
  • 13.Ybarrondo B, O’rourke AM, Brian AA, Mescher MF. Contribution of lymphocyte function-associated- 1/intercellular adhesion molecule-1 binding to the adhesion/ signaling cascade of cytotoxic T lymphocyte activation. J Exp Med. 1994;179:359. doi: 10.1084/jem.179.1.359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Green JM, Zheng XG, Shimizu Y, Thompson CB, Turka LA. T cell receptor stimulation, but not CD28 costimulation, is dependent fon LFA-1-mediated events. Eur J Immunol. 1994;24:265. doi: 10.1002/eji.1830240141. [DOI] [PubMed] [Google Scholar]
  • 15.Carlow DA, Van Oers NS, Teh SJ, Teh HS. Deletion of antigen-specific immature thymocytes by dendritic cells requires LFA-1/ICAM interactions. J Immunol. 1992;148:1595. [PubMed] [Google Scholar]
  • 16.X.U.X.Y. Honjo K, Devore-Carter D, Bucy RP. Immunosuppression by inhibition of cellular adhesion mediated by leukocyte function-associated antigen-1/intercellular adhesion molecule-1 in murine cardiac transplantation. Transplantation. 1997;63:876. doi: 10.1097/00007890-199703270-00014. [DOI] [PubMed] [Google Scholar]
  • 17.Haug CE, Colvin RB, Delmonico FL, et al. A phase Y trial of immunosuppression with anti-ICAM-1 (CD54) mAb in renal allograft recipients. Transplantation. 1993;55:766. doi: 10.1097/00007890-199304000-00016. [DOI] [PubMed] [Google Scholar]
  • 18.Luzuy S, Merino J, Engers HD, Izui S, Lambert PH. Autoimmunity after induction of neonatal tolerance to alloantigens: role of B cell chimerism and F1 donor B cell activation. J Immunol. 1986;136:4420. [PubMed] [Google Scholar]
  • 19.Schurmans S, Merino J, Qin H-Y, et al. Autoimmune syndrome after neonatal induction of tolerance to alloantigens: analysis of the specificity and of the cellular and genetic origin of autoantibodies. Autoimmunity. 1991;9:283. doi: 10.3109/08916939108997130. [DOI] [PubMed] [Google Scholar]
  • 20.Merino J, Qin H-Y, Schurmans S, Gretener D, Grau GE, Lambert PH. Thrombocytopenia associated with the induction of neonatal tolerance to alloantigens: immunopathogenic mechanisms. Eur J Immunol. 1989;19:1693. doi: 10.1002/eji.1830190925. [DOI] [PubMed] [Google Scholar]
  • 21.Schurmans S, Gonzalez AL, Revilla C, Ramos A, Lambert PH, Merino J. Anti-LFA-1 (CD11a) monoclonal antibody interferes with neonatal induction of tolerance to alloantigens. Eur J Immunol. 1994;24:985. doi: 10.1002/eji.1830240431. [DOI] [PubMed] [Google Scholar]
  • 22.Merino J, Schurmans S, Luzuy S, Izui S, Vassalli P, Lambert PH. Autoimmune syndrome after induction of neonatal tolerance to alloantigens: effects of in vivo treatment with anti-T cell subset monoclonal antibodies. J Immunol. 1987;139:1426. [PubMed] [Google Scholar]
  • 23.Revilla C, Gonzalez AL, Conde C, Lopez-Hoyos M, Merino J. Treatment with anti-LFA-1α monoclonal antibody selectively interferes with the maturation of CD4+ thymocytes. Immunology. 1997;90:550. doi: 10.1046/j.1365-2567.1997.00183.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Schurmans S, Brighouse G, Kramar G, et al. Transient T and B cell activation after neonatal induction of tolerance to MHC class II or Mls alloantigens. J Immunol. 1991;146:2152. [PubMed] [Google Scholar]
  • 25.Gonzalez AL, Conde C, Revilla C, Ramos A, Renedo B, Merino J. Autoimmune syndrome after induction of neonatal tolerance to I-E antigens. Eur J Immunol. 1993;23:2353. doi: 10.1002/eji.1830230945. [DOI] [PubMed] [Google Scholar]
  • 26.Ramos A, Gonzalez M, Lopez-Hoyos M, Carrio R, Merino J. Completely allogeneic spleen cells induced cytolytic neonatal tolerance to alloantigens, but failed to establish allo–helper interactions with host T cells. Immunology. 1996;89:413. doi: 10.1046/j.1365-2567.1996.d01-763.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Ramos A, Merino R, Merino J. Differences in non-MHC alloantigens promote tissue rejection but fail to mediate allogeneic co-operation and autoimmunity in mice neonatally injected with semi-allogeneic F1 B cells. Immunology. 1994;82:287. [PMC free article] [PubMed] [Google Scholar]
  • 28.Harding FA, McArthur JG, Gross JA, Raulet DH, Allison JP. CD28-mediated signaling costimulates murine T cells and prevents induction of anergy in T cell clones. Nature. 1992;356:607. doi: 10.1038/356607a0. [DOI] [PubMed] [Google Scholar]
  • 29.Noelle RJ, Roy M, Shepherd DM, Stamenkovic Y, Ledbetter JA, Aruffo A. A 39-kDa protein on activated T helper cells binds CD40 and transduces the signal for cognate activation of B cells. Proc Natl Acad Sci USA. 1992;89:6550. doi: 10.1073/pnas.89.14.6550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Van Seventer GA, Shimizu Y, Horgan KJ, Shaw S. The LFA-1 ligand ICAM-1 provides an important costimulatory signal for T cell receptor-mediated activation of resting T cells. J Immunol. 1990;144:4579. [PubMed] [Google Scholar]
  • 31.Sherman LA, Chattopadhayay S. The molecular basis of allorecognition. Annu Rev Immunol. 1993;11:385. doi: 10.1146/annurev.iy.11.040193.002125. [DOI] [PubMed] [Google Scholar]
  • 32.Dang LH, Rock KL. Stimulation of B lymphocytes through surface Ig receptors induces LFA-1 and ICAM-1-dependent adhesion. J Immunol. 1991;146:3273. [PubMed] [Google Scholar]
  • 33.Raviola E, Karnovsky MJ. Evidence for a blood–thymus barrier using electron-opaque tracers thymic barrier. J Exp Med. 1972;136:466. doi: 10.1084/jem.136.3.466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Schurmans S, Heusser C, Qin HY, Merino J, Brighhouse G, Lambert PH. In vivo effects of anti-IL-4 monoclonal antibody on neonatal induction of tolerance and on a associated autoimmune syndrome. J Immunol. 1990;145:2465. [PubMed] [Google Scholar]
  • 35.Macatonia SE, Hsieh CS, Murphy KM, O’garra A. Dendritic cells and macrophages are required for Th1 development of CD4+ T cells from alpha beta TCR transgenic mice: IL-12 substitution for macrophages to stimulate IFN-gamma production is IFN-gamma-dependent. Int Immunol. 1993;5:1119. doi: 10.1093/intimm/5.9.1119. [DOI] [PubMed] [Google Scholar]
  • 36.Suen Y, Lee SM, Qian J, Van De Ven C, Cairo MS. Dysregulation of lymphokine production in the neonate and its impact on neonatal cell mediated immunity. Vaccine. 1998;16:1369. doi: 10.1016/s0264-410x(98)00094-2. [DOI] [PubMed] [Google Scholar]
  • 37.Donckier V, Flamand V, Desalle F, et al. IL-12 prevents neonatal induction of transplantation tolerance in mice. Eur J Immunol. 1998;28:1426. doi: 10.1002/(SICI)1521-4141(199804)28:04<1426::AID-IMMU1426>3.0.CO;2-P. [DOI] [PubMed] [Google Scholar]
  • 38.Baron JL, Reich E-P, Visintin Y, Janeway CA. The pathogenesis of adoptive murine autoimmune diabetes requires an interaction between α4-integrins and vascular adhesion molecule-1. J Clin Invest. 1994;93:1700. doi: 10.1172/JCI117153. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Immunology are provided here courtesy of British Society for Immunology

RESOURCES