CD40 ligand (CD40L) plays a crucial role in T cell-dependent B cell responses, but whether its abundance is a limiting factor in their development is unclear. This question was addressed in transgenic mice expressing the murine CD40L gene under the control of the IL-2-promoter (CD40Ltg+). The fraction of activated T cells from the CD40Ltg+ mice with detectable levels of surface CD40L was modestly greater (1.1- to 2-fold) than littermate controls and paralleled an ∼1.8-fold increase in CD40L mRNA abundance. In response to trinitrophenol (TNP)-keyhole limpet hemocyanin and tetanus/diphtheria vaccine, CD40Ltg+ mice developed higher titers of high-affinity IgG and IgG1 Ab than wild-type mice. In contrast, the Ab response of CD40Ltg+ and control mice was similar in response to the T-independent Ag TNP-Ficoll. These results suggest that a modest increment in expression of CD40L accelerates the development of T-dependent responses, and that CD40L plays a limiting role in the induction of high-affinity Ab and Ab-class switching.
CD40 ligand (CD40L)3 (CD154) is expressed in mature activated CD4 T cells and other cell types but not in resting T cells (1). The interaction between T and B cells through the engagement of CD40L with CD40 is essential for the development of humoral responses to T-dependent Ags (2). In humans, mutations of the CD40L gene are responsible for the X-linked hyperIgM syndrome, characterized by a lack of germinal centers, isotype switching, affinity maturation, recall responses with repeat immunizations, and some defects in cell-mediated immunity (3, 4). This phenotype has been reproduced in mice, through gene disruption of either CD40 or CD40L (5, 6).
These results indicate that CD40L plays an important role in B cell responses to T cell-dependent Ags but it is not known whether the abundance of CD40L expressed on activated T cells is in excess or is a critical factor limiting T cell-dependent B cell responses. To evaluate this question, transgenic mice were engineered with the murine CD40L gene under the control of the human IL-2 promoter.
Materials and Methods
C57BL/6 and BDF1 mice were purchased from The Jackson Laboratory (Bar Harbor, ME) and maintained at the University of Washington (Seattle, WA) Specific Pathogen-Free animal facility. Mice lacking the CD40L gene were kindly provided by R. Flavell (Yale University, New Haven, CT).
IL-2 promoter-CD40L-HGx transgenic construct
The cDNA for the murine CD40L gene (nucleotides 615 to 1435) was cloned downstream of the human IL-2 promoter/enhancer region from −568 to +47 relative to the transcriptional start site (7, 8) and upstream from a mutated version of the human growth factor (HGx) gene in the pBluescript plasmid (Stratagene, La Jolla, CA). HGx have been shown to increase transgene expression, providing a source of introns and poly(A) signal, but the mutation prevents expression of functional human growth hormone (9).
Generation of transgenic mice
Transgenic mice were generated by microinjection of the construct described above into BDF2 single cell embryos, using previously published protocols (10). Founders were identified by PCR and Southern blotting and crossed sequentially to C57BL/6 mice. Of the 11 founder lines generated, 1 expressing line was selected for further analysis.
Preparation of Ags and immunization of mice
Trinitrophenol-keyhole limpet hemocyanin (TNP14-KLH) and TNP-Ficoll were kindly provided by R. J. Noelle (Dartmouth Medical School, Lebanon, NH). TNP4-BSA and TNP16-BSA were conjugated in our laboratory (11). Mice were immunized by i.p. injection at day 1 with 10% alum in normal saline (NS), TNP14-KLH (50 μg/mouse) in 10% alum in NS, or TNP-Ficoll (50 μg/mouse) in PBS. Sera were collected from the mice before and 7 and 21 days after immunization. Mice were immunized by i.p. injection at day 1 and day 42 with 1 Lf tetanus and 1.3 Lf diphtheria toxoids adsorbed to alum (Connaught Laboratories, Swiftwater, PA). Sera were collected from mice before and at days 7, 21, 47, and 52 after immunization.
TNP4-BSA and TNP16-BSA (10 μg/m; were added to the wells (50 μl each) of polystyrene microtiter trays (MaxiSorp, Nalge Nunc International, Naperville, IL) and incubated at 4°C overnight. Wells were then blocked by incubation for 2 h with 3% BSA in PBS-0.5% Tween-20. After washing the plates six times with PBS-0.5% Tween-20, diluted samples were added to the plates and incubated for 2 h at room temperature (RT) (sera were initially diluted 1:100 and then diluted 2-fold serially). After washing, a 1:2000 dilution of HRP-conjugated anti-mouse IgM (Biosource International, Camarillo, CA), anti-IgG, anti-IgG1, or anti-IgG2a (Southern Biotechnology Associates, Birmingham, AL) was added for 1 h at RT. After washing, reactions were developed with 2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) peroxidase substrate: H2O2 (Kirkegaard & Perry Laboratories, Gaithersburg, MD) for 10–15 min. The A405 was read in a Beckman Coulter (Fullerton, CA) Biomek spectrophotometer. Ab titers were determined by endpoint titrations after subtraction of the values for the preimmune samples. The last dilutions at which the net OD405 exceeded 0.2 for IgM and 0.1 for total IgG and IgG1 were taken as the titer. Total Ab titers were determined in plates coated with TNP16-BSA, and high-affinity Ab titers were determined on plates coated with TNP4-BSA.
Cell suspensions were prepared from various mouse tissues by gently teasing the cells from the organs and macerating the tissue through a wire-mesh screen. Freshly isolated thymocytes, splenocytes and lymph node cells were resuspended in RPMI 1640 (Life Technologies, Grand Island, NY) supplemented with 2 mM l-glutamine, 100 U/ml of penicillin, and 100 μg/ml of streptomycin and 10% FCS (RPMI complete). The RPMI complete medium was supplemented when indicated with 10 μg/ml of PMA and 0.5 μM ionomycin and/or biotinylated MR-1 mAb, specific for murine CD40L, was added (12). The cultures were maintained at 37°C in a humidified atmosphere of 5% CO2 for 3, 5, 8, or 24 h.
All Abs were titrated and used at saturating concentrations for flow cytometric studies. Single cell suspensions were prepared from thymus, spleen, or lymph nodes. RBC were removed from spleen cell preparations by NH4Cl lysis. Cells were washed and then stained for 20 min with various combinations of the following mAbs specific to murine Ags: FITC anti-B220, PE-conjugated anti-CD3, PE-conjugated anti-CD4, FITC-anti-CD8, biotin (BIO)-conjugated anti-CD80, and BIO-conjugated anti-CD86. MR-1 was detected with strepavidin-PE or strepavidin-FITC. Cells were fixed in PBS containing 2% parafomaldehyde. A minimum of 10,000 events per sample was analyzed using a FACScan flow cytometer (Becton Dickinson, San Jose, CA). Analysis of the data was performed using CellQuest software (Becton Dickinson).
Northern blot analysis
Total cellular RNA was isolated from resting and activated splenocytes using Trisol reagent (Biotecx Laboratories, Houston, TX), and 10 μg of total RNA per sample were electrophoreted and transferred as described (13). The 311-bp probe containing part of the third and fourth exons of the CD40L was used to identify CD40L mRNA, and a 1.5-kb probe for elongation factor-1α was used as a control. RNA abundance was quantitated by PhosphoImager analysis after correction for background.
CD40L expression in CD40Ltg+ mice
The human IL-2 promoter fragment used to generate the CD40Ltg+ mice has been shown previously to drive expression of the murine IL-10 gene (14) and LacZ transgenes in activated T cells (7). Consistent with these reports, the fraction of cells in the lymph nodes that expressed CD40L in response to stimulation with PMA/ionomycin was modestly but consistently greater (1.1- to 2-fold, mean 29% positive cells) in four matched pairs of CD40Ltg+ mice than in their littermate controls (mean 18% positive cells) (Fig. 1,A). Two-color analysis showed that nearly all of the cells that expressed CD40L were CD4+ T cells (Fig. 1 B). In contrast, nearly all PMA/ionomycin-stimulated cells from lymph nodes of CD40Ltg+ and control mice expressed CD69 (94 and 90%, respectively). Few cells expressed CD40L in the absence of stimulation, but this number was also greater in CD40Ltg+ than in control mice (mean 2.8 and 1.5%, respectively). The mean proportion of cells in the lymph nodes of control and CD40Ltg+ mice, respectively, which were CD4+ (40%, 46%), CD8+ (20%, 23%), and B220+ (32%, 31%), was similar as was the proportion and numbers of these cell populations in the spleens.
Surface expression of CD40L on splenic T cells was not increased as consistently as on lymph node T cells (data not shown). However, splenocytes from CD40Ltg+ mice stimulated with PMA/ionomycin for 5 h consistently contained more CD40L mRNA, which in average was 1.8-fold more than in splenocytes from controls (Fig. 2). In addition to expressing the species of mRNA detected in controls, CD40Ltg+ cells contained one or more slowly migrating bands, presumably corresponding to mRNA transcripts containing one or more of the HGx exons spliced to the CD40L mRNA. The kinetics of CD40L mRNA expression were similar in CD40Ltg+ and control mice: CD40L mRNA was not detected in unstimulated cells, peaked 3 h after stimulation with PMA + ionomycin, declined between 5 and 8 h, and was no longer detectable by 18 h (not shown). Theses mRNA results indicate that splenic T cells from CD40Ltg+ mice, like those in lymph nodes, expressed more CD40L than did cells from controls.
CD40Ltg+ mice have more efficient Ab responses to T cell-dependent Ags
To determine whether the modest increase in CD40L expression in the tg+ mice affected Ab production, mice were immunized with the T cell-dependent Ag, TNP-KLH. The CD40Ltg+ mice had significantly higher titers of high-affinity anti-TNP IgM Ab at day 7. The CD40Ltg+ mice also had significantly higher titers for total and high-affinity IgG and IgG1 anti-TNP at day 7 (4- and 5- fold more, respectively) and day 21 (3.6- and 1.6- fold more, respectively) than littermate controls (Fig. 3). These results suggested that CD40Ltg+ mice developed a more robust overall response, which reflected in part a more rapid affinity maturation and isotype switch. To determine whether this was true for another T cell-dependent Ag, the response to tetanus toxoid was evaluated. The results paralleled those observed with TNP-KLH: the CD40Ltg+ mice produced more IgG1 and similar amounts of IgM Ab than controls (Fig. 4). In contrast to these T cell-dependent responses, there was no difference in the production of Ab in response to the T cell-independent Ag TNP-Ficoll (Fig. 5).
Expression of CD40L by CD4+ T cells is tightly regulated by transcriptional and posttranslational mechanisms, so that expression is limited to a brief period following activation of the cells (1, 15). The current studies evaluated the consequences of increasing the abundance of CD40L expressed by activated T cells while preserving the physiological pattern of regulation by directing expression under the control of the IL-2 promoter. A 1.1- to 2-fold increase in expression of CD40L on activated T cells allowed CD40Ltg+ mice to produce higher titers of IgG and high-affinity Abs to T cell-dependent but not to T cell-independent Ags. Previous studies in mice have shown that administration of anti-CD40 Abs, recombinant CD40L, or plasmids leading to the constitutive expression of CD40L in transduced cells enhances Ab responses to concomitantly administered Ags (16, 17). However, this pharmacological approach results in B cell activation that is nonphysiological and substantially polyclonal, including the enhancement of Ab responses to T cell-independent polysaccharide Ags (17). In contrast, the current studies directly test the physiological importance of limited and tightly regulated CD40L expression in T cells. The results indicate that the abundance of CD40L expressed by activated T cells normally limits the rate and magnitude of the Ab response to T cell-dependent but not to T cell-independent Ags. The mechanisms by which this occurs are likely to be multiple, since CD40L enhances T cell-dependent Ab responses through direct effects on B cells (18), signals to the T cell itself after engagement of CD40 (19), and influences T cell-dependent responses indirectly by increasing the expression of costimulatory molecules and cytokines, like IL-12, by APC (20, 21).
Our results, in which the abundance but not the pattern of CD40L expression was altered in CD40Ltg+ mice, contrast with those obtained in mice in which this protein was expressed constitutively in T cells (22, 23). Constitutive expression of a CD40L transgene in developing T cells under the proximal lck promoter disrupted thymic architecture and perturbed T cell development (22). Transduction of T cell progenitors from CD40L knockout mice with a retrovirus caused low-level constitutive expression of CD40L and was followed by the development of T lymphoblastic lymphomas in >50% of the mice by 6–9 mo of age. In contrast, the thymus and peripheral T cell compartments were normal in mice expressing the CD40L transgene under the IL-2 promoter. The CD40Ltg+ mice have remained healthy and appear normal at necropsy at >9 mo of age (our unpublished observations). This suggests that disruption of the normal pattern of transient, activation-induced expression accounted for the aberrant development and malignant transformation observed in the other studies.
The finding that CD40L expression by activated T cells is limiting has implications for situations in which the abundance of CD40L may be reduced below levels normally expressed by activated T cells. T cells from murine and human neonates (24, 25, 26), and to some extent naïve T cells from mature mice (27), express lower amounts of CD40L than do T cells from adults in general, and memory/effector T cells in particular (27). Lower level of expression of CD40L by neonatal T cells may underlie, at least in part, their somewhat slower development of T cell-dependent Ab responses and greater bias toward the production of Th2 cytokine-dependent isotypes (28, 29). Decreased expression of CD40L may also be a factor in the greater sensitivity of neonatal and naïve T cells to induction of tolerance (27, 30). Consistent with this, administration of anti-CD40 Abs as a surrogate for CD40L expression by activated T cells abrogates neonatal tolerance induction (24). Similarly, presentation of Ags by mature dendritic cells (31), or the use of adjuvants that may in part bypass the need for CD40L to prime APC (26), may be useful strategies to enhance T cell-dependent responses in cases where the initial expression of CD40L may be limiting.
We thank Dr. Hans D. Ochs and Dr. Alejandro A. Aruffo for helpful scientific discussions; Michael Weaver and Aniko Fekete for technical assistance; Ben Jacobson for transgenic injections; and Kathryn Allen for assistance with flow cytometry.
This work was supported by Grants AI37107 and HD18184 from the National Institutes of Health.
Abbreviations used in this paper: CD40L, CD40 ligand; TNP, trinitrophenol; KLH, keyhole limpet hemocyanin; WT, wildtype.