| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Divisions of
*
Cancer and Hematology and
Immunology, Walter and Eliza Hall Institute of Medical Research, Royal Melbourne Hospital, Parkville, Victoria, Australia;
School of Microbiology and Immunology, University of New South Wales, Sydney, New South Wales, Australia; and
Howard Hughes Medical Institute, Herman B. Wells Center for Pediatric Research, Section of Pediatric Hematology/Oncology, Departments of Pediatrics and Medicine, Indiana University School of Medicine, Indianapolis, IN 46202
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(SDF-1) (1, 2). Failure to receive such signals results in developmental
blockages at stages specific for the missing signals. Mutations that
reduce but do not eliminate signaling through such receptors result in
correspondingly partial blockages of B cell development (3, 4). Similarly, the selection of emigrant immature B cells into
the population of mature lymphocytes recirculating within the secondary
lymphoid organs is dependent on signals delivered through the BCR
(4, 5, 6, 7, 8). Entry into the recirculating pool, however, is not
sufficient to ensure B cell longevity. A continuous or tonic level of
BCR signaling is required to maintain the peripheral population
(9). This tonic signal is also positively and negatively
modulated by accessory molecules such as CD19 and CD22, because their
absence significantly alters the characteristics of the follicular B
cell population (10, 11). Several numerically minor B lymphocyte populations exist in the periphery, and recruitment into these populations is also regulated by balancing a multitude of external signals. A distinct population of B cells is located in the unique anatomical location of the spleen called the marginal zone (MZ). MZ B cells do not appear to recirculate and differ from follicular B cells in phenotype, differentiation potential, and responsiveness to T cell-independent Ags (12). Analysis of genetically modified mouse strains lacking MZ B cells has led to the suggestion that recruitment into this population is regulated by the strength of BCR signaling and that this can be modulated by coreceptors such as CD19 (13, 14). There is also the B1 population, barely detectable in the spleen but enriched in the peritoneal and pleural cavities of mice. This population develops early from precursors in fetal liver or omentum and also has a distinct phenotype and pattern of reactivity to mitogens (15). B1 cells are thought responsible for the so-called natural Ig in mice, particularly IgM and IgA (16). The formation and maintenance of the B1 population are dependent on signaling through the BCR, because the population is diminished in situations where BCR signaling is impeded, such as in CD19-/- mice, and is expanded where BCR signaling is exacerbated, such as in New Zealand Black and related strains (17, 18).
The capacity of lymphocytes to integrate signals from numerous sources
into clear pathways of differentiation is apparent in the B cell
response to Ag. B cells activated by ligation of the BCR migrate to T
cell areas and interact with Ag-specific T cells. These B cells
proliferate and differentiate along two related pathways. One pathway
results in the formation of clusters or foci of Ab-forming cells (AFC)
that form in the outer periarteriolar lymphoid sheath (PALS). These
cells secrete low-affinity Ab and die by apoptosis after a relatively
short life (19). The second differentiation pathway takes
the B cells into the follicles, where they give rise to germinal
centers (GC). Efficient GC formation and function are dependent on the
B cell receiving information from numerous sources, including Ag,
follicular dendritic cells, the innate immune system via complement and
phagocytic cells, and CD4+ T cells
(20). The correct positioning of T and B cells within the
lymphoid areas during an immune response is mediated by altering their
response to chemokines (21, 22, 23). While GC development and
function require B lymphocyte chemoattractant (BLC;
CXCL13)-dependent organization, the localization of AFC foci produced
during the immune response appears to be due to SDF-1 (CXCL12)
(21).
It is evident that the integration of signals transmitted from the BCR, costimulatory molecules, and chemokine receptors is pivotal to B cell development and function. While such integration probably occurs at multiple levels, the Rho family of guanosine triphosphatases (GTPases), which includes Rho, Rac, Cdc42, and relatives, are attractive candidates as points of confluence for these signals, as they appear to perform analogous roles in other cell types such as fibroblasts (24). Rac proteins in particular have been directly implicated in extracellular signal integration in T lymphocytes (25). Rac2 is a hemopoietic-specific GTPase that is thought to act in concert with its ubiquitously expressed, highly homologous sister protein, Rac1 (26). Unlike other Rho family GTPases, Rac2 is expressed exclusively in myeloid and lymphoid cells, including B lymphocytes (27), suggesting that it is required for unique processes ascribed to these cells. We have recently described crucial roles for Rac2 in the regulation of superoxide production in neutrophils (28) and chemotaxis in neutrophils, mast cells (29), and hemopoietic stem cells (30).
Several studies suggest that Rac proteins are important for B lymphocyte function and for signaling from the BCR and CD19. Analysis of mice deficient in Vav, a guanine exchange factor for Rac and a catalyst of Rac activity, revealed abnormal Ig class switching in response to T-dependent Ags (31) as well as impaired signaling via the BCR and CD19 (31, 32, 33, 34). Furthermore, ligation of CD19 in a human B cell line activates Vav and mitogen-activated protein kinase pathways and induces the association of CD19 and Vav (35), suggesting that Rac may be important for CD19-mediated signal transduction.
The majority of studies to date have focused on the roles of Rac proteins in T lymphocytes, whereas little, if any, data directly address Rac2 function in B lymphocytes. In this study we begin to define the physiological roles of Rac2 in B lymphocyte development and function using a variety of in vitro and in vivo assays that directly compare B lymphocytes containing or lacking Rac2.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Rac2-deficient mice backcrossed nine generations onto the
C57BL/6 strain and wild-type littermate or C57BL/6 mice were used in
these experiments. The generation of Rac2-deficient mice has been
described previously (28). Sex- and age-matched mice,
5–25 wk of age, were used in these experiments. For reciprocal
transplant experiments, four to seven lethally irradiated mice of each
genotype were reconstituted with 5 x 106
bone marrow cells from donors of each genotype, and lymphoid
compartments were analyzed 8 wk postinjection. Reciprocal, labeled
lymphocyte transfer experiments were performed as previously described
(36) with minor alterations. Groups of three recipients of
each genotype were injected i.v. with unfractionated splenocytes
(containing equivalent numbers of B220+ cells for
each genotype, 
30 x 106 for each
independent experiment) from pooled donors of each genotype, after
incubating with 0.5 µM CTG (CellTracker Green CMFDA; Molecular
Probes, Eugene, OR) according to the manufacturer’s instructions. The
frequency of labeled cells in recipient organs was determined 5 h
later by flow cytometric analysis of 100,000 leukocytes. For peripheral
blood, recovery was calculated from the frequency of labeled cells, the
leukocyte count per milliliter, the number of cells injected, and a
presumed blood volume of 2 ml. To examine T-dependent B lymphocyte
responses, mice were injected i.p. with either 100 µg
nitrophenyl-keyhole limpet hemocyanin (NP-KLH) precipitated with
alum and boosted at 6 wk with 4 µg NP-KLH diluted in PBS or with 100
µg NP-KLH in the absence of adjuvant. T-independent type II (TI-II) B
lymphocyte responses were examined by injecting mice i.p. with 10 µg
DNP-dextran diluted in PBS. Analyses of data were performed using
one-way ANOVA and, where appropriate, the Student-Newman-Keuls
test.
Phenotypic characterization of B lymphocyte populations by flow cytometry
B lymphocyte populations were characterized by flow cytometry using FITC-, PE-, or biotin-conjugated mAbs specific for CD45R (B220), IgM, IgD, CD19, CD21, CD23, CD24, and CD5 (provided by Dr. A. Strasser, Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia). Biotinylated Abs were revealed with streptavidin-PE (BD PharMingen, San Diego, CA).
In vitro lymphocyte chemotaxis and F-actin assays
The recombinant mouse chemokines SDF-1 (CXCL12) and BLC
(CXCL13; R&D Systems, Minneapolis, MN) were used for migration and
actin assays. Migration assays were performed as previously described
(37). In brief, 5 x 105 cells
in 100 µl medium were added to the upper chamber of the 5-µm pore
size Transwell insert (Costar, Corning, NY). Various concentrations of
the chemoattractant were added to the appropriate well. The Transwells
were incubated at 37°C in 5% CO2 for 4 h.
Migration of B lymphocytes was determined by flow cytometry using
anti-B220 Abs. Duplicate assays were performed for each sample.
Actin polymerization was quantified as described previously
(37) with minor modifications. Splenocytes (5 x
105) were incubated at 37°C for 5 min before
addition of chemoattractant in varying concentrations for 15 s,
then fixed, stained (80 nM FITC-phalloidin (Sigma-Aldrich, St. Louis,
MO), 0.1 mg/ml L--lysophosphatidylcholine
(Sigma-Aldrich), 3.6% formaldehyde, and PBS) and analyzed by flow
cytometry. Forty thousand events were acquired, and lymphocytes were
identified by forward and side scatter characteristics. Increases in
mean cellular fluorescence of lymphocyte populations relative to
control samples were used to determine percentage increases in
filamentous actin (F-actin).
Immunohistochemistry
Spleens were snap-frozen in OCT compound (Sakura Finetek,
Torrance, CA) and were stored at -70°C. Cryostat sections were
prepared and stained exactly as described previously (38).
Sections were stained with the following reagents in the combinations
indicated in the relevant figures: anti-IgG1-HRP (Southern
Biotechnology Associates, Birmingham, AL), biotinylated peanut
agglutinin, MOMA-1 (specific for MZ metallophilic macrophages, provided
by Dr. I. Caminschi, Walter and Eliza Hall Institute of Medical
Research, Parkville, Victoria, Australia), F4/80 (Serotec,
Oxford, U.K.), RA3-6B2 (anti-B220), and biotinylated KT3
(anti-CD3). Anti-B220 was detected with Abs specific for rat 
L
chain directly coupled to HRP (BD PharMingen). Biotinylated primary
reagents were detected using streptavidin-alkaline phosphatase
(Southern Biotechnology Associates). Alkaline phosphatase was
visualized using the Fast Blue kit (Vector Laboratories, Burlingame,
CA) after endogenous phosphatases were blocked with 2 mM levamisole
(Sigma-Aldrich). HRP was visualized using the 3-amino-9-ethylcarbazole
substrate kit (Vector Laboratories).
ELISA and ELISPOT
Ab titers and the frequency of AFC were determined exactly as previously described (39). Individual Ig isotypes were revealed using specific goat anti-mouse Abs conjugated to HRP (Southern Biotechnology Associates), except IgA, which was detected using biotinylated goat anti-mouse IgA and streptavidin-HRP (Southern Biotechnology Associates). IgG AFC represent a pool of cells secreting IgG1, IgG2a, and IgG2b. For detection of Ag-specific Abs and AFC following immunization, plates were coated with 20 µg/ml NP2BSA or NP20BSA diluted in PBS. Bound IgG1 was detected with isotype-specific goat anti-mouse IgG1 directly coupled to HRP. Secondary Abs were detected using the reaction of peroxidase with ABTS (Sigma-Aldrich).
Proliferation and Ig production
In vitro proliferation and Ig production assays used a pool of two spleens for each assay. B lymphocytes were purified from spleen by negative sorting after removal of RBC. T lymphocytes, macrophages, and plasma cells were labeled with Abs specific for Thy1, Mac1, and 281-Syndecan, respectively. B cell representation among the negatively sorted leukocytes was determined by subsequently staining the sorted population with anti-B220-PE and was routinely >95%. Proliferation assays were conducted at a concentration of 106 cells/ml in 0.2 ml RPMI 1640/2% FCS/2-ME. B lymphocytes were stimulated with LPS (Difco, Detroit, MI) at 2 and 20 µg/ml, F(ab')2 anti-IgM (Jackson ImmunoResearch Laboratories, West Grove, PA) at 1 and 10 µg/ml, and CD40 ligand (CD40L; 1/200 dilution of a baculovirus-infected cell membrane preparation (40)) alone or supplemented with IL-4 and IL-5. Proliferation on day 2 was measured as [3H]thymidine incorporation following a 6-h pulse with 0.5 µCi [3H]thymidine. Ig production assays were conducted at a cell concentration of 2.5 x 105 cells/ml in 400 µl RPMI 1640/2% FCS/2-ME. B lymphocytes were stimulated with LPS, CD40L, and CD40L supplemented with IL-4 and IL-5 as described above. IgM and IgG1 production on day 5 was determined by ELISA. All assays were performed in triplicate. Control wells contained no stimulus.
Calcium signaling
Calcium flux experiments were performed as described previously
(41). In brief, indo-1/AM (Molecular Probes)-loaded
splenocytes were stained with either anti-B220-FITC or CD21-FITC
and CD23-PE and then labeled with biotinylated anti-CD19 and/or
anti-mouse L chain Abs at the concentrations indicated in the
figure. For calcium flux induction, biotinylated Abs were cross-linked
with 20 µg/ml avidin. Changes in intracellular calcium concentrations
were monitored for 4–6 min by calculating the ratio of the
fluorescence emissions at 425 and 530 nm of B220+
cells or follicular B cells
(CD21lowCD23high). Previous
experiments have shown that staining with anti-B220 does not
influence the outcome of BCR cross-linking (41).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Before defining the role of Rac2 in B lymphocyte differentiation,
we first established the basic parameters of the immune system in
Rac2-deficient C57BL/6 mice. The weight and total cellularity of bone
marrow, spleen, lymph nodes, and peritoneal cavity were normal in
rac2-/- animals (Table I
and data not shown). Analysis of bone
marrow from rac2-/- mice, however,
revealed a 30% reduction in overall B lymphocyte numbers. This
deficiency was explained largely by a reduction in the population of
recirculating B lymphocytes in the bone marrow
(B220highsIgM+; Table I and
Fig. 1A). The proportion and
number of precursor
(B220lowsIgM-) and
immature (B220lowsIgM+) B
lymphocytes did not differ between
rac2-/- and control bone marrow,
indicating that the early stages of B lymphocyte development are
normal. Analysis of peripheral blood from adult C57BL/6 Rac2-deficient
mice, however, revealed a significant increase in total leukocyte
number (wild type, 7.6 ± 2.3 x 109
cells/L; rac2-/-, 12.3 ± 3.2
x 109 cells/L; n = 10–11;
p < 0.05) similar to that observed in previous studies
of Rac2-deficient mice on a 129.B6F2 genetic background
(28). B lymphocyte numbers in blood were increased by 40%
(Table I), as were T cells (data not shown).
|
|
). To clarify the nature of the B
cell deficiency in rac2-/- spleen, we
determined the number of cells present in each of several identifiable
B cell subpopulations. Immature
(B220lowHSAhigh) and mature
(B220highHSAlow) B
lymphocyte cell numbers were reduced by 10 and 40%, respectively,
compared with controls, while immature plus MZ B lymphocytes
(IgDlowIgM+) and follicular
(IgDhighIgM+) B lymphocytes
were reduced by 50 and 20%, respectively (Table I and Fig. 1A). Together these data indicate that the B cell deficiency
in the spleens of rac2-/- mice is
primarily due to an insufficiency of mature and/or MZ B cells. This
issue was clarified by flow cytometric analysis of spleen examining the
distribution of CD23- and CD21-expressing cells (Fig. 1B).
This showed a near absence of the MZ B cell population
(CD23lowCD21high), a
finding confirmed by immunohistochemistry (Fig. 1C). The
distribution of MOMA-1-positive metallophilic MZ macrophages was normal
in rac2-/- mice (Fig. 1C), as
were the number and distribution of F4/80+
macrophages (Fig. 1C), suggesting that the absence of MZ B
cells in rac2-/- mice is not due to a
structural defect that prevents cells from accumulating in this area.
The localization of T and B lymphocyte populations in spleen, as
determined by immunohistochemistry, was normal in
rac2-/- mice (data not shown). To
determine whether the absence of MZ B cells was cell-intrinsic,
reciprocal bone marrow transplantation experiments into irradiated
recipients were performed. The percentage and number (Fig. 1B) of MZ B cells observed in wild-type and
rac2-/- recipients 8 wk post-transplant
simply reflected the genotype of the donor, with marked deficiency of
MZ B cells being recapitulated in all recipients of
rac2-/- cells. The absence of MZ B cells
in recipients of rac2-/- cells was also
confirmed by immunohistochemistry (data not shown).
Phenotypic characterization of B lymphocyte populations in lymph node
(central and peripheral) revealed a trend similar to that observed in
the spleen, with a decrease in the proportion of B lymphocytes noted
(wild type, 45 ± 6%; rac2-/-,
36 ± 5%; n = 12–13; p < 0.05),
although the reduction in absolute number of
B220+ lymph node cells was not statistically
significant (Table I). As assessed histologically, Peyer’s
patches were of normal architecture and cellularity (data not
shown). A second striking deficiency in lymphocyte development was
evident within the peritoneal cavity of
rac2-/- mice. Cytological analysis showed
a decrease in the proportion of lymphocytes in the peritoneal cavity
(data not shown). Flow cytometric analysis revealed this decrease to be
specific for the CD5+sIgM+
B-1a lymphocytes, which were decreased in both proportion and total
cell number in rac2-/- mice (Table I and
Fig. 1D).
Thus, analysis of the B cell subpopulations in both bone marrow and the periphery reveals a critical role for Rac2 in the formation of the recirculating lymphocyte population in the bone marrow and of splenic MZ and peritoneal B1a cells.
Decreased serum IgM and IgA, but increased serum IgG1 and IgG2b, in Rac2-deficient animals
To determine whether the aberrant B cell development observed in
rac2-/- mice had a functional corollary,
we measured the levels of several Ig isotypes in the serum of
rac2-/- and control mice (Fig. 2A). The concentrations of
IgG1 and IgG2b were elevated 2-fold in the serum of 8- to 11-wk-old
rac2-/- mice compared with wild-type
littermates, whereas serum IgM was reduced by 95% and IgA by 50%. The
relative levels of serum IgG2c and IgG3 were similar between wild-type
and rac2-/- mice (data not shown). IgE is
normally present at low concentrations in serum, and in this assay was
not detected in either wild-type or
rac2-/- serum, indicating a
concentration of <80 ng/ml (data not shown).
|
B). The reduction in IgM AFC in
rac2-/- mice was also evident in
immunohistochemical stains of spleen sections using Abs to IgM (data
not shown). Given that the total number of leukocytes in bone marrow
and spleen was normal in rac2-/- mice
(Table I), the reduced frequency of IgM AFC in
rac2-/- mice translated into a
significant reduction in total number. Thus, in addition to the B cell developmental defects observed as a consequence of Rac2 ablation, rac2-/- mice show reduced titers of IgM and IgA together with a reduced frequency of IgM AFC.
Differentiation of rac2-/- B cells into AFC is not intrinsically blocked
To clarify whether the observed in vivo deficiency of IgM
AFC resulted from a B cell-intrinsic developmental defect, we next
measured the ability of purified spleen B lymphocytes to differentiate
into AFC in vitro. B lymphocytes were stimulated with two
concentrations of LPS to induce polyclonal activation and IgM
production in the absence of BCR-specific activation signals. As shown
in Fig. 2C, no differences in IgM production were revealed
in response to 2 or 20 µg/ml LPS. We next compared the responses of
rac2-/- and control B cells to a T
cell-derived signal, CD40L. Stimulation with CD40L alone produced low
quantities of IgM in both wild-type and
rac2-/- lymphocytes. Supplementing CD40L
with IL-4 and IL-5 enhanced IgM production 100-fold in both control
and rac2-/- B lymphocytes, although IgM
production by rac2-/- B lymphocytes
remained at 25% of that by control B lymphocytes
(p < 0.05). The production of IgG1 in response
to stimulation with CD40L/IL-4/IL-5 was indistinguishable between
wild-type and rac2-/- lymphocytes,
indicating that rac2-/- lymphocytes
respond normally to IL-4 and IL-5 and are able to class switch to IgG1
in response to a T-dependent stimulus.
Defective Ag receptor-mediated proliferation in Rac2-deficient B lymphocytes
While the reduced frequency of IgM AFC in
rac2-/- mice does not appear to result
from an intrinsic block in B cell differentiation, it may reflect
deficiencies in BCR-mediated signaling. To determine the response of
rac2-/- B lymphocytes to BCR
cross-linking, purified B cells were stimulated with a series of
mitogens, including F(ab')2 anti-IgM. A
significant defect in the proliferation of
rac2-/- B lymphocytes was observed in
response to cross-linking the BCR with 1 µg/ml anti-IgM (Fig. 2D; p < 0.05). This defect was ameliorated
by a higher concentration of anti-IgM (10 µg/ml). No defects were
observed in response to stimulation with LPS, CD40L, or CD40L
supplemented with IL-4 and IL-5. Thus, Rac2 appears to have a specific
role in B cells in setting the threshold of responsiveness to BCR
cross-linking.
Maintenance of calcium flux is impaired in Rac2-deficient B lymphocytes in response to BCR or CD19 ligation
An immediate consequence of BCR ligation is the fluxing of
intracellular calcium in B cells. The magnitude of this flux is related
to the extent of BCR cross-linking and can be influenced by the
coligation of accessory molecules such as CD19 (positively) and CD22
(negatively) with the BCR (42). To determine whether Rac2
deficiency influenced calcium flux in response to BCR ligation, we
measured intracellular Ca2+ using the
calcium-sensitive fluorescent indicator, indo-1/AM. In response to a
low level of BCR cross-linking (0.01 µg/ml anti-),
rac2-/- B cells showed a reduced
Ca2+ flux (Fig. 3A, upper left
panel). However, just as a high concentration of
F(ab')2 anti-IgM restored normal
proliferation in rac2-/- B cells, a
10-fold increase in the concentration of anti- restored
Ca2+ flux in
rac2-/- B cells to that of controls
(Fig. 3A, upper right panel). We next determined
whether coligation of CD19 with BCR amplified
Ca2+ flux in the absence of Rac2. BCR and CD19
were cross-linked with either 0.1 or 0.01 µg/ml anti- and
anti-CD19. While coligation of CD19 with BCR clearly enhanced the
Ca2+ flux in both
rac2-/- and wild-type B cells relative
to BCR cross-linking alone, the maintenance of elevated levels of
Ca2+ was less in
rac2-/- B cells (Fig. 3B).
This was particularly meaningful when high concentrations of
anti- (Fig. 3B, right panels) were used,
because, when used alone, this reagent induced a response of equal
magnitude in both rac2-/- and wild-type
B cells (Fig. 3A, upper right panel).
Cross-linking CD19 in the absence of BCR ligation engendered a small
and late Ca2+ flux in control B cells and no
change in rac2-/- B cells, consistent
with diminished signaling through CD19 in
rac2-/- B cells (Fig. 3A,
lower right panel). To exclude the possibility that the
observed differences in Ca2+ flux were explained
trivially by the absence of MZ B cells, Ca2+ flux
was specifically analyzed in follicular B cells, and the differences
observed between genotypes were identical with those observed for the
total B cell populations (data not shown). Collectively, these results
indicate a role for Rac2 either in the signal transduction pathways
emanating from the BCR and resulting in Ca2+
mobilization or in determining the efficiency of BCR ligation. They
further indicate that Rac2 is required for the full enhancement of
signals engendered by the coligation of CD19 with BCR.
|
Abnormalities in chemotaxis in response to a variety of
chemoattractants have been noted in Rac2-deficient neutrophils
(28), mast cells (29), and hemopoietic stem
cells (30). We therefore assessed the responses of
rac2-/- and control lymphocytes to a
range of chemokines. Chemotaxis of splenocytes was reduced by 50–75%
in response to all concentrations of SDF-1 and BLC tested (Fig. 4, A and B).
Migration of unstimulated splenocytes (control) was also markedly
reduced, indicating that chemokinesis, or random movement, was also
reduced. This latter observation was confirmed by additional studies
showing that migration of rac2-/-
splenocytes in a zero gradient of chemokine (equal concentrations of
chemokine in the upper and lower chambers of a Transwell) was decreased
relative to that in wild-type splenocytes (data not shown). Flow
cytometric analysis of lymphocyte populations before and after
migration showed that B lymphocyte migration was reduced 65–75% in
response to 800 ng/ml SDF-1 and BLC (inset of Fig. 4, A
and B), but only 40% when stimulated with 1500 ng/ml BLC.
Migration of rac2-/- lymph node B
lymphocytes was also significantly reduced in response to CCL21
and macrophage inflammatory protein-3
(data not shown). These
abnormalities were not explained by differences in integrin-mediated
adhesion or chemokine receptor expression. Adhesion of purified B
lymphocytes via 41
and 51 integrins was
similar between genotypes, as measured in static adhesion assays with
fibronectin fragments (data not shown). Furthermore, expression of
CXCR5 mRNA, the receptor for BLC, was similar in splenocytes of both
genotypes. These data suggest that the role of Rac2 in chemokine
responsiveness is in the efficiency of the response, rather than in
whether there is a response at all.
|
C). The frequencies of labeled cells
recovered from peripheral lymph nodes of either genotype were reduced
by 30–50% in two independent experiments when
rac2-/- donor B cells were injected
rather than wild-type B cells (Fig. 4C). These data confirm
the B cell-intrinsic nature of the defect and suggest that the reduced
chemokine responsiveness of rac2-/- B
cells contributes to their altered distribution in vivo. Reduced actin polymerization in B lymphocytes in response to chemokines
Chemotaxis is dependent on remodeling of the actin cytoskeleton,
and this involves polymerization of globular monomers of actin to form
F-actin. Actin polymerization in splenocytes following a 15-s
stimulation (the optimal time point for peak polymerization) with BLC
and SDF-1 was quantified using phalloidin-FITC. A 50–70% reduction
in F-actin generation was recorded in
rac2-/- splenocytes in response to
stimulation with a wide range of concentrations of SDF-1 (Fig. 4D) and BLC (Fig. 4E). Maximal stimulation of
lymphocytes with BLC and SDF-1 did not ameliorate the defect in
actin polymerization relative to wild-type lymphocytes.
Rac2-deficient mice have perturbed responses to TI-II and T-dependent Ags
The deficiencies in B cell development revealed by our earlier experiments raised the possibility that the ability of these cells to respond to Ag in vivo may be compromised. The absence of MZ B cells and the diminished responsiveness to BCR cross-linking, for example, may impact on the immune response of rac2-/- mice to T cell-independent Ags, while diminished responsiveness to chemokines and CD19 coligation may influence the outcome of T cell-dependent immune challenge. To determine whether this was indeed the case, we challenged rac2-/- and control mice with T cell-independent and T cell-dependent Ags.
Rac2-deficient and control mice were immunized with the
hapten-conjugated polysaccharide, DNP-dextran, a TI-II Ag, and the
response was followed by measuring serum Ig able to bind the highly
cross-reactive Ag, TNP. As expected from earlier findings (Fig. 2A), TNP-specific IgM was markedly reduced at baseline (Fig. 5, upper panel). Following
immunization, the production of TNP-specific IgM was reduced 5- to
10-fold at all time points measured, but both the fold rise over
baseline and the rate of increase were the same as those observed for
wild-type mice. The production of TNP-specific IgG3 was normal at all
time points measured, indicating that class switching in response to
the Ag was normal (Fig. 5, lower panel).
|
A). Interestingly, the foci
of IgG1 AFC that form in the outer PALS as a normal part of the immune
response were particularly larger in some
rac2-/- spleens (Fig. 6A).
This apparent overexuberance was confirmed by the finding of
significant increases in the numbers of Ag-specific IgG1 AFC in
rac2-/- spleens (Fig. 6B,
upper panel) 1 wk following immunization and increases in
Ag-specific serum IgG1 2 wk following immunization (Fig. 6C;
p < 0.05). Consistently, 2- and 4-fold elevations in
Ag-specific IgM were observed in rac2-/-
sera 1 and 2 wk following immunization (p <
0.01). Migration of AFC to the bone marrow during the course of the
immune response appears normal in
rac2-/- mice. A similar frequency and
proportion of total (NP20-binding) and
high-affinity (NP2-binding) IgG1 AFC were
apparent in bone marrow both early in the primary response and 1 wk
after boosting late in the response (Fig. 6B, lower
panel). Because bone marrow AFC originate in the GC, this
observation implies that Rac2 is not critical for B cell selection and
differentiation in the GC or for the subsequent migration of these AFC.
Similarly, the production of high-affinity IgG1 was similar between
rac2-/- and wild-type mice, indicating
that affinity maturation in the GC was occurring normally (data not
shown). Immunization with NP-KLH in the absence of adjuvant induced a
substantially lower concentration of Ag-specific serum IgM and IgG1 but
revealed no differences between wild-type and
rac2-/- mice (Fig. 6C).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Rac2 modulates BCR and CD19 signaling thresholds
In vitro, rac2-/- B cells demonstrated both defective proliferation and Ca2+ flux in response to submaximal stimulation of the BCR. Ca2+ flux was also reduced in response to maximal coligation of BCR and CD19 and to CD19 ligation alone. Rather than being absolutely required for BCR or CD19 signaling to occur, Rac2 appears to play a necessary role in grading B cell responses across a range of stimuli intensities and combinations. Consistent with these in vitro observations, Rac2 deficiency in vivo results in abnormal B lymphocyte development in specialized populations of both B1 and B2 lymphocyte lineages. Although the development of precursor and immature B lymphocytes in the bone marrow was normal in rac2-/- mice, mature recirculating B lymphocytes were reduced, as were splenic MZ B cells and peritoneal cavity B1a lymphocytes. For mature recirculating B lymphocyte and B1a lymphocyte populations, recruitment or maintenance of cells is directly dependent upon signals delivered via the BCR and coreceptors such as CD19 (4, 9, 45, 46). In contrast, recent evidence indicates that MZ B cell development is inversely related to the strength of the BCR signaling (13), so the deficiency of MZ B cells in the absence of Rac2 cannot simply be explained by diminished BCR signaling.
In this regard it is pertinent that CD19-deficient mice and Vav-deficient mice share these key phenotypic features of Rac2-deficient mice. The absence of CD19 results in a reduction in MZ B lymphocytes, peritoneal cavity B-1a lymphocytes, and all serum Ig isotypes (11, 14, 17). Vav associates with CD19 (35), and the activity of Vav is augmented by cross-linking of the BCR and CD19 (47). Vav-1-deficient mice lack B-1a cells, and their B cells proliferate poorly in response to cross-linking of the BCR and do not maintain elevated levels of intracellular calcium in response to ligation of the BCR and CD19 (31, 32, 33, 48). These abnormalities are more severe when Vav-2 is also absent (49, 50). Collectively, our data and these observations indicate that Rac2 influences BCR and CD19 signaling under physiological conditions and that Vav is probably its major proximal activator. As yet our data do not distinguish between a direct role in signaling cascades emanating from the BCR or CD19 receptors and an indirect role in influencing the assembly of the BCR:coreceptor signaling complex via regulation of actin cytoskeleton remodeling.
Rac2 regulates B lymphocyte migration in response to chemokines
Another prominent consequence of Rac2 deficiency is a disturbed
distribution of B lymphocytes in vivo. The 40% increase in B
lymphocytes in peripheral blood and the more modest decreases in B
lymphocytes in the spleen and lymph nodes suggest that the migration of
B lymphocytes from the peripheral blood to secondary lymphoid tissue is
diminished. Recirculation of lymphocytes between peripheral blood and
secondary lymphoid tissues is modulated by chemoattractant cues
(51). The reduced chemotaxis of B lymphocytes in response
to a wide range of concentrations of BLC and SDF-1 in
vitro, attributable partly to a decrease in actin polymerization, is
reflected by the reduced homing of transferred B cells to peripheral
lymph nodes. It is likely to explain the B lymphocytosis in peripheral
blood and contribute to the minor deficit in B lymphocytes in spleen,
lymph node, and bone marrow.
Additional abnormalities in the distribution of subpopulations of B lymphocytes were found in rac2-/- mice. Ig-secreting cells migrate under the direction of chemokine gradients from the peripheral lymphoid tissues to gut-associated lymphoid tissues and to the bone marrow, where they become the major source of serum Ig. Recently, it has been demonstrated that the migration of B-1 lymphocytes from the peritoneal cavity to gut-associated lymphoid tissues and spleen is crucial for their differentiation into plasma cells (16, 52). Indeed, we observed a significant reduction in the frequency of IgM-secreting plasma cells in both spleen and bone marrow of rac2-/- mice. As B1a cells and their plasma cell derivatives are major producers of IgM and IgA (16, 53), it is likely that the marked reduction in serum IgM concentration and the modest reduction in IgA concentration seen in rac2-/- mice reflect both the numerical deficiency in B1a peritoneal cells and the defective migration of these cells to become plasma cells in secondary sites. Importantly, the reduced IgM concentration in the serum of rac2-/- mice was not due to an inability to secrete Ab, as IgM production in response to stimulation with LPS and CD40L was normal. We speculate that the observed deficiency in IgM production in vitro in response to CD40L/IL-4/IL-5 can be attributed to the absence of MZ B lymphocytes in the assays, as these cells produce greater quantities of IgM than follicular B lymphocytes in vitro (12).
Rac2 deficiency perturbs T-independent and T-dependent immune responses
Rac2-deficient mice displayed normal IgG3 responses to the TI-II Ag, DNP-dextran, and normal kinetics of rise in IgM. The consistent reduction in TNP-specific IgM throughout the response is most simply explained by a reduced starting frequency of precursor B cells specific for this Ag. Responsiveness to this class of Ag is primarily a property of MZ B cells, due to both their location in the spleen and unique activation properties (12, 54). The pattern of TI-II response observed is in keeping with the marked reduction in MZ B cells in Rac2-deficient mice and indicates that the residual cells are functionally normal.
The foci of AFC that form in the outer PALS of the spleen during the
immune response to T-dependent Ags develop as a consequence of
Ag-specific B cell proliferation and migration. B cells first receive T
cell help in the inner PALS. They subsequently migrate to the outer
PALS, where they proliferate, isotype switch from IgM to IgG, and
differentiate into AFC (55). The migration of B cells in
the spleen is under the direction of gradients of BLC, SDF-1,
CCL21, and macrophage inflammatory protein-3 (36, 43, 44). Given the significant reductions in lymphocyte chemotaxis
in response to several chemokines we observed in vitro, it is
surprising that no gross alterations of the architecture of lymphoid
tissues or major perturbations in T-dependent immune responses were
observed. Subtle abnormalities were evident however: an increase in
NP-specific serum IgM and IgG1 and an increase in the size of
extrafollicular AFC foci. It is possible that these are a consequence
of prolonged B cell stimulation in the PALS due to defective migration
away from Ag-specific T cells along chemotactic gradients of BLC or
SDF-1. No increases in the proliferation of purified B lymphocytes
were observed in vitro in response to T-dependent or T-independent
stimuli, suggesting that the increase in the size of foci in some
Rac2-deficient mice is not due to abnormal proliferation or delayed
apoptosis.
The absence of major perturbations in B cell responses to Ag in
immunization protocols is in sharp contrast to the profound
abnormalities observed in CD19-deficient mice. Similarly, the
relatively normal architecture of lymphoid organs (except for the
absence of MZ cells) is distinct from the severe consequences for
lymphoid tissue organization observed in chemokine- or chemokine
receptor-deficient mice (43, 44). These differences
indicate that residual signaling downstream of these receptors is
sufficient to establish and then maintain correctly functioning
lymphoid tissue. We conclude that while Rac2 is a physiological
regulator of key cellular responses to the BCR, CD19, and chemokine
receptor ligation, redundancy of its activity is apparent. Other
GTPases, such as the highly homologous Rac1, are almost certainly
involved. Evidence for a compensatory increase in Rac1 activity was
demonstrated recently in rac2-/-
hemopoietic stem cells stimulated with SDF-1 (30).
Interestingly, this occurred not at the transcriptional or
translational levels, but in the percentage of Rac1 found in its active
GTP-bound state. Future studies of mice with compound deficiencies of
Rac1 and Rac2 will ultimately determine whether true redundancy exists
between these two Rac proteins. Such studies of mutant mice and their
cells will enable more complete definition of how B cells coordinate
signaling from key cell surface receptors.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Andrew W. Roberts, Walter and Eliza Hall Institute of Medical Research, PO Royal Melbourne Hospital, Parkville, Victoria 3050, Australia. E-mail address: roberts{at}wehi.edu.au
3 Abbreviations used in this paper: BCR, B cell Ag receptor; AFC, Ab-forming cell; CD40L, CD40 ligand; F-actin, filamentous actin; GC, germinal center; GTPase, guanosine triphosphatase; MZ, marginal zone; SDF-1, stromal cell-derived factor 1; NP-KLH, nitrophenyl-keyhole limpet hemocyanin; BLC, B lymphocyte chemoattractant; PALS, periarteriolar lymphoid sheath; TI-II, T-independent type II.
Received for publication November 29, 2001. Accepted for publication January 28, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
chemotaxis and promotes protein kinase C (PKC)-induced internalization of CXCR4. J. Exp. Med. 189:1461.B-inducing kinase (NIK) causes defects in secondary lymphoid tissue chemokine receptor signaling and homing of peritoneal cells to the gut-associated lymphatic tissue system. J. Exp. Med. 191:1477.This article has been cited by other articles:
|
|
 |
|
  F. Guo, J. A. Cancelas, D. Hildeman, D. A. Williams, and Y. Zheng Rac GTPase isoforms Rac1 and Rac2 play a redundant and crucial role in T-cell development Blood, September 1, 2008; 112(5): 1767 - 1775. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
) are shown.
D, Dot plot (n = 6) showing
peritoneal cavity lymphocytes stained with fluorescently conjugated Abs
to CD5 and surface IgM (sIgM).
) mice was determined by
ELISA. B, Frequency of IgM-producing plasma cells in
bone marrow and spleen of rac2-/-
and wild-type mice. C, Production of IgM and IgG1 in
response to T-independent and T-dependent stimuli. Purified splenic B
lymphocytes from wild-type (•) and
rac2-/- (
