| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Genetics and Pathology, Uppsala University, Uppsala, Sweden
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RII or CD23), as
shown by its complete absence in mice pretreated with mAbs specific for
CD23 and in CD23-deficient mice. Because the constitutive expression of
murine CD23 is limited to B cells and follicular dendritic cells
(FDCs), one of these cell types is likely to be involved. One of the
suggested modes of action of IgE/CD23 is to increase the ability of B
cells to present Ag to T cells, as demonstrated to take place in vitro.
Another possibility is that FDCs capture the IgE/Ag complexes and
present these directly to B cells. The purpose of the present study was
to determine whether CD23+ B cells or FDCs are responsible
for the IgE/CD23-mediated enhancement of specific Ab responses in vivo.
We show that the enhancement is completely restored in irradiated
CD23-deficient mice reconstituted with CD23+ spleen or bone
marrow cells. In these mice, the B cells are CD23+ and the
FDCs are presumably CD23- because the FDCs are radiation
resistant and are reported not to be replaced by donor cells after this
type of cell transfer. In contrast, enhancement was not restored in
irradiated wild-type mice reconstituted with CD23- cells.
These results indicate that CD23+ B cells, and not FDCs,
are the cells that capture IgE/Ag complexes and induce
enhancement of Ab responses in vivo. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RII or CD23) (1, 2).
Targeting of Ag to CD23 by covalent binding of Ag to CD23-specific mAbs
also induces a strong response, both in vivo and in vitro
(3), further emphasizing the role of CD23 in Ab responses.
In addition, it has been shown in vitro that IgE/Ag complexes are more
efficiently presented to T cells by murine splenic B cells (4, 5) and EBV-transformed human B cells (6, 7, 8) than Ag
alone. Also in these studies, the enhancing effect of IgE is inhibited
with CD23-specific mAbs (4, 5, 6, 7). Three groups have produced
CD23-deficient mice (9, 10, 11). These mice displayed normal
lymphocyte differentiation and could mount normal Ab responses to
thymus-independent Ags and after infection with Nippostrongylus
brasiliensis. The effect on the Ab response to thymus-dependent
Ags was investigated in two of the studies. One group found no
difference between the response of wild-type and CD23-deficient mice
(9), whereas another group found increased IgE levels in
CD23-deficient mice (11). Moreover, the ability of
DNP-specific IgE to enhance the Ab response to OVA-DNP was severely
impaired in CD23-deficient mice (9). In naive mice, the constitutive expression of CD23 is limited to B cells (12) and follicular dendritic cells (FDCs)3 (13). This murine CD23 is probably a human CD23a counterpart, because human CD23a is expressed constitutively on B cells (14). Another human isoform, CD23b, is induced on a wide variety of hematopoietic cell types after IL-4 stimulation (14, 15). Low levels of murine CD23b mRNA are found in B cells and non-B cells after activation with LPS and IL-4 (16). However, only one form of murine CD23, the one analogous to human CD23a, is expressed at detectable protein levels (17). Involvement of the murine CD23b isoform in IgE-mediated enhancement of in vivo Ab responses is unlikely because IgE/Ag complexes induce unperturbed enhancement in mice deficient in IL-4, thereby demonstrating that the constitutively expressed CD23 is sufficient (18). This finding suggests that either CD23+ B cells or FDCs (or both) are the cells that capture the IgE/Ag complexes and initiate enhancement of Ab responses.
During the primary response to a T cell-dependent Ag, activated B cells produce factors required for maturation of the FDC network. The FDCs capture and retain native Ag in complex with Ab and complement at their surface for long periods of time. It is believed that the trapped Ag can be endocytosed by B cells. These B cells become highly efficient APCs that present processed Ag to T cells. The T cells, together with FDCs, then support further proliferation and differentiation of the B cells. During the secondary response to an Ag, the kinetic of the germinal center reaction is much quicker. Abs from the primary response are present and form complexes with Ag. These Ab/Ag complexes are more efficiently trapped by FDCs than free Ag (reviewed in Refs. 19 and 20). An early and enhanced germinal center formation has been demonstrated in mouse regional lymph nodes or spleens following primary injection of Ab/Ag complexes as compared with Ag alone (21, 22). In addition, the immune complex-induced appearance and growth of germinal centers preceded, or paralleled, an observed rapid and enhanced Ab response (21, 22). Most likely the germinal center reaction is involved in IgE/CD23-mediated enhancement of Ab responses. One possibility is that IgE/CD23 increase the ability of B cells to take up and present Ag to T cells. A mutually not exclusive explanation is that FDCs capture the IgE/Ag complexes via CD23 and present them directly to B cells.
The origin of FDCs has not been unequivocally established. A number of
studies suggest that FDCs differentiate from local stromal cells with a
fibroblastic or mesenchymal morphology, whereas others suggest that
they are derived from hematopoietic cells. There is also speculation
that germinal centers may display two types of FDCs of different origin
(reviewed in Refs. 23, 24, 25). Early studies with mouse
chimeras demonstrated that host FDCs are not replaced by donor cells
after transfer of BM cells (26, 27). This is in agreement
with a more recent study demonstrating that organized clusters of FDCs
are absent from the spleen in mice deficient in lymphotoxin 
. When
these mice are reconstituted with CR1/2-deficient bone marrow (BM)
cells, organized CR1+ FDCs are identified
(28), suggesting that FDCs are not transferred by BM
cells, but rather mature from non-BM precursor cells. Further, transfer
of mouse allogeneic spleen cells into SCID mice (also lacking organized
clusters of FDCs) resulted in the generation of FDCs of host origin
(29). However, Kapasi et al. (30) observed
development of both mouse and rat FDCs in SCID mice reconstituted with
BM or fetal liver cells from rat. In the present study we use transfer
experiments to evaluate the contribution of CD23+
B cells and FDCs in IgE/CD23-mediated enhancement of in vivo Ab
responses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Because all H-2b mice have an I-Ab-linked low responsiveness to IgE/BSA-TNP and IgG/BSA-TNP complexes (31, 32), male CD23-deficient founder mice on a H-2b background (9) were mated with female responder CBA/J (H-2k) mice and the offspring were repeatedly backcrossed to the CBA/J strain. The CD23 genotype was analyzed in a PCR using one sense primer and either of two antisense primers: P1 (5'-AAC GCA CGG GTG TTG GGT CGT TTG-3'), P2 (5'-GCT TTC ACA CTT GCA GTC CTC-3'), and P3 (5'-TGA GAC ATT CTG CTC CCA TCC-3'). Gene amplification was done in 50 µl 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 2.5 (P1/P2) or 1.5 (P2/P3) mM MgCl2, 0.2 mM dNTPs, 0.4 µM of primer, and 2.5 U AmpliTaq DNA polymerase (Perkin-Elmer Cetus, Norwalk, CT) for 40 cycles (1 min at 94°C, 1 min at 58°C, 1 min at 74°C). P1/P2 PCR gave a 862-bp fragment, whereas P2/P3 PCR gave a fragment of 1007 bp. By intercross of the fifth generation, mice with homozygous CD23 gene defects and wild-type littermates were obtained. Groups of four to eight male or female mice, 2–5 mo of age, were used and matched for age and sex within each experiment. The mice were bred in the animal facilities at the Department of Genetics and Pathology, Uppsala University.
Abs and Ags
The origin and purification of mAbs from the hybridoma cell
lines IGELb4 (mouse IgE/
anti-TNP) (33) and C4007B4
(7B4, mouse IgG2a anti-TNP) (34) has been described
before (1, 35). BSA, OVA, and TNP were obtained from Sigma
(St. Louis, MO), and TNP was conjugated to BSA in 0.28 M cacodylate
buffer, pH 6.9 (cacodylic acid, sodium salt; Sigma). After 70 min at
room temperature the reaction was stopped by an excess of
glycyl-glycin, 1 mg/ml (Merck, Darmstadt, Germany). The number of
incorporated TNP residues per BSA molecule was 23 as determined
according to Ref. 36 . The proteins were dialyzed against
PBS and stored at 4°C as sterile solutions.
Adoptive transfer and immunization
Single cell suspensions of donor spleens were prepared in PBS by pressing cells through a mesh, whereas BM cells were harvested by flushing the femurs of donor mice with PBS. Recipient mice were irradiated with 600 rad, and 5 x 107 spleen or 1 x 107 BM cells in 0.2 ml of PBS were injected i.v. 24 h after the irradiation. After 1 day (spleen cells) or 42 days (BM cells) the recipient mice received 50 µg TNP-specific IgE or IgG (IGELb4 or 7B4) together with 20 µg BSA-TNP and 20 µg OVA (as a specificity control) in 0.1 ml of PBS. Ab/Ag complexes were formed by incubating Ag and Ab (or Ag and PBS for controls) for 1 h at 37°C immediately before injection.
ELISA
Mice were bled from the tail veins, and isolated sera were analyzed by ELISA as described (2). Hyperimmune antisera against BSA or OVA, affinity purified on BSA- or OVA-Sepharose, were used as standards. Standard curves and calculations were done by the use of a Softmax program (Molecular Devices, Menlo Park, CA). Statistical differences were determined by Student’s t test.
Flow cytometry
Spleen cells were suspended in PBS containing 0.5% FCS. PE-labeled B3B4 (rat anti-mouse CD23) and FITC-labeled 1D3 (rat anti-mouse CD19) from PharMingen (San Diego, CA) were incubated with 106 cells for 30 min at 4°C. After incubation, cells were washed and data of 10,000 cells were collected and analyzed using a FACSort flow cytometer (Becton Dickinson, Mountain View, CA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It has been reported that CD23-deficient mice display impaired
IgE-mediated enhancement of Ab responses (9). To confirm
this finding in our system, IgE/BSA-TNP complexes were injected into
CD23-deficient mice and wild-type controls. Since it has been shown
previously that anti-hapten Abs modulate the Ab response to all
determinants on an Ag (1, 2, 9, 34, 37), the BSA-specific
response was determined. No enhancement was seen in CD23-deficient
mice, whereas wild-type controls had a 50- to 400-fold enhancement of
the BSA-specific Ab response (Fig. 1
,
a and b). We conclude that the IgE-mediated
enhancement is mediated exclusively via CD23 and not via any other IgE
receptor. To investigate whether our CD23-deficient mice were able to
produce Abs in other situations, they were immunized with IgG/BSA-TNP
complexes also known to enhance Ab responses in vivo (34, 38). CD23-deficient mice were indeed capable of mounting an Ab
response to IgG/BSA-TNP and even showed a higher response than
wild-type controls (Fig. 1, c and d). Both IgE-
and IgG-mediated enhancement of Ab responses is strictly Ag specific
(1, 2, 38). To confirm the specificity in the present
study, mice were always immunized with unspecific Ag (OVA) in addition
to the specific Ag (BSA-TNP). Although the specific response was
significantly enhanced, the response to unspecific Ag was never
enhanced (data not shown).
|
To investigate whether the impaired response to IgE/BSA-TNP
complexes in CD23-deficient mice was due to absence of the receptor on
B cells or FDCs, irradiated CD23-deficient mice were reconstituted with
CD23+ BM cells from wild-type controls. In BM
chimeras, the B cells are of donor origin and the FDCs of host origin
because FDCs are radiation resistant (26, 39) and are not
replaced by donor cells after transfer of BM cells to irradiated hosts
(26, 27, 28). The cell surface expression of CD19 and CD23 on
BM cells from donors and on spleen cells from reconstituted
recipients was analyzed by flow cytometry. In a representative
experiment, wild-type donor BM cells consisted of 7.2%
CD19+CD23+ cells and
29% CD19+CD23- cells,
whereas only 0.7%
CD19-CD23+ cells were
detected. CD23-deficient donor BM cells consisted of 56%
CD19+CD23- cells and 0.4%
CD23+ cells (data not shown). Thus the
CD23-deficient donor mice were indeed CD23 negative. Representative
histograms from recipients 42 days after adoptive transfer are shown in
Fig. 2.
CD19+CD23+ cells
constituted 44% of the cells in wild-type mice and 45% of the
cells in CD23-deficient mice reconstituted with
CD23+ cells (Fig. 2, a and
c). Spleen cells from CD23-deficient and wild-type mice
reconstituted with CD23- cells had 
2.3%
CD19+CD23+ cells (Fig. 2, b and d). There were 1.1%
CD19-CD23+ cells in all
the cell suspensions tested by flow cytometry. These findings indicate
that 1) in spleen cell suspensions from naive mice, only
CD19+ cells (i.e., B cells) express CD23, and 2)
CD23-deficient mice are efficiently reconstituted with
CD23+ B cells.
|
a) as well as irradiated
CD23-deficient mice reconstituted with CD23+ BM
cells (Fig. 3c) responded to IgE/Ag complexes with a 100- to
2000-fold enhancement. Irradiated CD23-deficient mice reconstituted
with CD23- BM cells (Fig. 3b) showed
only marginal enhancement of the Ab response. To further analyze the
involvement of FDCs, the converse experiment was performed. Irradiated
wild-type mice were reconstituted with CD23- BM
cells (Fig. 3d), resulting in an animal where the B cells
were CD23- and the FDCs were
CD23+. These mice had very low Ab responses
indistinguishable from those in irradiated CD23-deficient mice
reconstituted with CD23- BM cells (Fig. 3b).
|
Despite the abundance of data indicating that FDCs are not derived
from BM, this is still a controversial area (reviewed in Refs.
23, 24, 25). To avoid any possible differentiation of
transferred BM cells into FDCs, spleen cells instead of BM cells were
used in the following experiments. Single cell suspensions of donor
spleens were prepared by pressing cells through a mesh. This rough
treatment destroys the fragile FDCs (22, 40) but leaves
the B cells unharmed. In addition, it has been reported that host FDCs
are not replaced by donor cells after transfer of spleen cells
(29). Therefore, we assume that the B cells are of donor
origin and the radiation-resistant FDCs of host origin in spleen cell
chimeras. The expression of CD19 and CD23 on spleen cells from donors
and reconstituted recipients was measured by flow cytometry. In a
representative experiment, wild-type donor spleen cells consisted of
34% CD19+CD23+ cells,
6.1% CD19+CD23- cells,
and 1.1% CD19-CD23+
cells. CD23-deficient donor spleen cells consisted of 34%
CD19+CD23- cells and 0.9%
CD23+ cells (data not shown).
Representative histograms from recipients 42 days after adoptive
transfer are shown in Fig. 4. The number
of CD19+CD23+ cells
observed in wild-type and CD23-deficient mice reconstituted with
CD23+ cells ranged from 35 to 41% (Fig. 4, a and c). CD23-deficient and wild-type mice
reconstituted with CD23- cells were consistently
CD23 negative, i.e., 1.9% of the cells were
CD23+ (Fig. 4, b and
d).
|
a) as well as irradiated
CD23-deficient mice reconstituted with CD23+
spleen cells (Fig. 5c) responded to IgE/Ag complexes with a
20- to 500-fold enhancement. As expected, irradiated CD23-deficient
mice reconstituted with CD23- spleen cells
(Fig. 5b) showed no enhancement of the Ab response. In the
converse experiment, irradiated wild-type mice reconstituted with
CD23- spleen cells (Fig. 5d) had a
similar lack of Ab responses as irradiated CD23-deficient mice
reconstituted with CD23- spleen cells (Fig. 5b).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As mentioned, it is unlikely that other cell types in naive mice than B cells and FDCs express CD23. The purpose of the present study was to determine whether CD23 expressed on B cells or FDCs are responsible for the IgE-mediated enhancement of specific Ab responses in vivo. Although some FDCs may be derived from BM (30), donor FDCs are not developed in normal irradiated mice after BM reconstitution using our protocol (26, 27, 28), nor do they survive conventional preparations of spleen cell suspensions (22, 40). This information was used in the construction of chimeric mice expressing CD23 on spleen cells or BM-derived cells (B cells) but not on radiation resistant cells (FDCs), or inversely, mice expressing CD23 on FDCs but not on B cells. Chimeric mice expressing CD23 on B cells, but not on FDCs, produced high Ab titers after immunization with IgE/Ag complexes, whereas mice expressing CD23 on FDCs, but not on B cells, did not. Our interpretation of these observations is that CD23+ B cells are required and sufficient for IgE-mediated enhancement of Ab responses in vivo, and that CD23+ FDCs alone are not sufficient for induction of enhancement.
An interesting possibility is that uptake of IgE/Ag via CD23 on B cells followed by processing and presentation of antigenic peptides to T cells explains IgE-mediated enhancement of Ab responses in vivo. In vitro, mouse or human B cells were shown to present Ag in complex with IgE to T cells much more efficiently than Ag alone (4, 5, 6, 7, 8). This effect was dependent on CD23. An alternative explanation for the in vivo effect of IgE/Ag is that CD23, acting as a coreceptor, initiates positive signaling to the B cell without involving the Ag presentation pathway. Immunization of nude mice with IgE/Ag do not induce Ab responses (2), and therefore direct signaling to the B cell through CD23 would not abrogate the need for T cell help. IgE/BSA-TNP complexes bind to the majority of naive splenic B cells ex vivo (32), suggesting that surface Ig (sIg), specific for the Ag component, is not a prerequisite for mere binding to CD23. Moreover, CD23-dependent endocytosis and presentation of IgE/Ag complexes to T cells is conducted by non-Ag-specific B cells in vitro (6, 7, 8). It has been proposed that activation of these "bystander" B cells by allergen-specific type 2 T-helper cells would cause the deterioration of single allergies into multiple allergies (41). In our system, however, the specificity is strictly maintained and only the Ab response to Ags within the complex is enhanced (present study and Refs. 1 and 2). Therefore, the mechanism behind IgE-enhanced Ab production must encompass a step that prevents unspecific B cells from being activated, even though they may take up IgE/Ag complexes via CD23. Independent ligation of CD23 and sIg cannot sufficiently explain why mice immunized with IgE-anti-TNP/BSA-TNP and an unspecific Ag (hen egg lysozyme (HEL), keyhole limpet hemocyanin (KLH), or OVA) do not produce Abs recognizing HEL, KLH, or OVA (present study and Ref. 2). A possible explanation for our observations is that CD23 and sIg must be co-cross-linked to trigger the stimulatory events leading to increased Ab production in vivo. These events may be connected with Ag presentation and/or with signaling through the B cell receptor complex. Our studies do not allow the distinction between the two, but given the abundance of in vitro studies showing increased presentation of IgE/Ag complexes by B cells, we find it most likely that this is also an in vivo mechanism. Clearly, however, a safety catch preventing all CD23+ B cells from activating T cells must exist and the definition of the molecular mechanism behind this remains an important task.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Susanne Gustavsson at her current address: Microbiology and Tumorbiology Center, Karolinska Institute, P.O. Box 280, SE-171 77 Stockholm, Sweden.
3 Abbreviations used in this paper: FDC, follicular dendritic cells; BM, bone marrow; TNP, 2,4,6-trinitrophenyl.
Received for publication March 31, 1999. Accepted for publication February 1, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RII by antigen-anti-FcRII conjugates. J. Immunol. 152:4388.[Abstract]
RII). J. Immunol. 148:2340.[Abstract]
receptor II (FcRII/CD23): tissue-specific and IL-4-specific regulation of gene expression. Cell 55:611.[Medline]
RII/CD23). Int. Arch. Allergy Immunol. 105:38.[Medline]
RII/CD23. Mol. Immunol. 30:27.[Medline]
and the type I tumor necrosis factor (TNF) receptor in the establishment of follicular dendritic cells from non-bone marrow-derived cells. J. Exp. Med. 186:1997.This article has been cited by other articles:
|
|
 |
|
  F. Hjelm, M. C. I. Karlsson, and B. Heyman A Novel B Cell-Mediated Transport of IgE-Immune Complexes to the Follicle of the Spleen J. Immunol., May 15, 2008; 180(10): 6604 - 6610. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Getahun, F. Hjelm, and B. Heyman IgE Enhances Antibody and T Cell Responses In Vivo via CD23+ B Cells J. Immunol., August 1, 2005; 175(3): 1473 - 1482. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Getahun, J. Dahlstrom, S. Wernersson, and B. Heyman IgG2a-Mediated Enhancement of Antibody and T Cell Responses and Its Relation to Inhibitory and Activating Fc{gamma} Receptors J. Immunol., May 1, 2004; 172(9): 5269 - 5276. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. D. de Stahl, J. Dahlstrom, M. C. Carroll, and B. Heyman A Role for Complement in Feedback Enhancement of Antibody Responses by IgG3 J. Exp. Med., May 5, 2003; 197(9): 1183 - 1190. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Kilmon, R. Ghirlando, M.-P. Strub, R. L. Beavil, H. J. Gould, and D. H. Conrad Regulation of IgE Production Requires Oligomerization of CD23 J. Immunol., September 15, 2001; 167(6): 3139 - 3145. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Payet-Jamroz, S. L. T. Helm, J. Wu, M. Kilmon, M. Fakher, A. Basalp, J. G. Tew, A. K. Szakal, N. Noben-Trauth, and D. H. Conrad Suppression of IgE Responses in CD23-Transgenic Animals Is Due to Expression of CD23 on Nonlymphoid Cells J. Immunol., April 15, 2001; 166(8): 4863 - 4869. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or 50 µg TNP-specific monoclonal
IgE (a and b) or IgG (c
and d) together with 20 µg BSA-TNP and 20 µg OVA
(
). Serum levels of OVA- and BSA-specific IgG were determined by
ELISA 7, 14, and 21 days after immunization. The level of BSA-specific
IgG in normal mouse serum was 0.32 µg/ml. Data are representative of
two independent experiments. Statistical differences: ***,
p 
