| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
*
Microbiology and
Medicine, Boston University School of Medicine, Boston, MA 02118;
Department of Microbiology and Immunology, Temple University School of Medicine, Philadelphia, PA 19140; and
§
Departments of Laboratory Medicine and Immunobiology, Yale University School of Medicine, New Haven, CT 06520
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One such model involves mice that inherit either the lpr or
gld mutation and thus fail to express functional Fas or Fas
ligand (FasL)4
(12, 13). These mice spontaneously develop a systemic
autoimmune disease associated with excessive autoantibody production
and often develop extremely high rheumatoid factor (RF) serum titers
(14, 15). Random sampling of B cell hybridoma cell lines
derived from diseased MRL-lpr or C3H-lpr mice
revealed that 
10–20% of the B cells spontaneously activated by the
autoimmune disease process expressed RF-binding activity. The majority
of these RF were of non-IgM isotypes and restricted in both their
isotype and allotype specificity (16, 17). Subsequent
sequence analysis demonstrated that many of these RF-producing B cells
were derived from the extensive expansion of a limited number of clones
and that members of these clones had undergone considerable somatic
mutation (17, 18). Despite this somatic mutation, the
extent of affinity maturation within a clone was limited and, in
general, the affinity of RF for autologous IgG2a was relatively low
(19). Many of these RF Abs were found to be IgG3 molecules
with the capacity to cryoprecipitate. On adoptive transfer to normal
mice, representative monoclonal RF cryoglobulins caused both skin
vasculitis and glomerulonephritis, pathologies frequently associated
with the lpr disease (20, 21). Thus, RF were
found to be a prevalent prototypic autoantibody in Fas/FasL-deficient
mice with the potential to contribute significantly to the effector
phase of the disease process.
The allotype specificity of MRL-lpr-derived RF B cells made them particularly amenable to experimental analysis. The heavy and light chain genes from AM14, a typical MRL/lpr-derived RF+ hybridoma cell line, were used to construct a B cell receptor (BCR) transgenic (Tg) mouse line designated AM14 (22). The AM14 RF is allotype restricted and recognizes IgG2aj or IgG2aa, but not IgG2ab. This allotype specificity made it possible to study the regulation of a true autoreactive B cell in the presence (on an MRL IgG2aj or BALB/c IgG2aa background) or absence (on a B6 IgG2ab background) of the corresponding autoantigen. In contrast to B cells from other Tg mice, where reactivity of the transgene encoded BCR with either natural (7, 8, 10) or experimental (1, 23) autoantigens led to deletion or anergy, B cells in nonautoimmune prone AM14 mice appeared remarkably unaffected by in vivo exposure to their cognate Ag. AM14 B cells were found to circulate with a normal life span in both IgG2ab and IgG2aa mice. In this state of "clonal ignorance," they were perfectly capable of responding in vivo to challenge with immune complexes (IC) of the appropriate allotype, as long as Ag-specific T cell help was available (9). Despite their clonally ignorant phenotype in wild-type mice, AM14 B cells can become spontaneously activated in IgG2aa lpr mice (24).
A number of factors, not necessarily exclusive, could account for the distinct profiles of AM14 cells in Fas-sufficient and Fas-deficient mice. For example, it is possible that the essentially naive status of AM14 B cells in nonautoimmune mice simply reflects the absence of autoreactive T cell help; more help would be available in lpr mice due to the failure of such T cells to undergo activation-induced cell death (25, 26, 27). Another possibility is that activated AM14 B cells persist in lpr mice under conditions that, in normal mice, would normally lead to elimination by FasL-expressing T cells (28, 29). Alternatively, the level or form of IgG2a present in the sera of the normal mice may not be sufficient to effectively signal through the AM14 receptor, whereas the IgG2a in the sera of autoimmune mice has the capacity, either by virtue of titer or by valence, to reach an activation threshold.
To further address the latter possibility, we undertook an in vitro analysis of the parameters governing the response of AM14 B cells to serum and monoclonal IgG2a Abs. We found that AM14 B cells were effectively stimulated only by the sera of Fas/FasL-deficient mice that expressed the Igha and Ighj allotypes, not by sera from wild-type mice or by sera from Fas/FasL-deficient mice of the inappropriate allotype. Additional studies with mAbs demonstrated that purified autoantibodies reactive with nuclear or cytoplasmic constituents could activate AM14 B cells, probably by complexing with ubiquitous nuclear components released into the media by dying cells. We speculate that similar complexes form in vivo, potentiating the activation of RF B cells. Such a pathway may explain the dominance of RF and anti-nucleosome specificities in lpr mice and in human systemic autoimmune disease.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
BALB/c mice Tg for the AM14 heavy chain (9, 22)
were backcrossed to MRL-+/+ mice and used after the eighth backcross
generation. MRL mice homozygous for the AM14 V
8 light chain were
provided by Dr. J. Erikson (Wistar Institute, Philadelphia, PA). Both
lines were bred and maintained under pathogen-free conditions at the
Boston University School of Medicine Laboratory Animal Sciences Center
and crossed to produce RF+ and RF-
littermates. The RF+ offspring were initially identified by
PCR, and their identity was confirmed by flow cytometry using the 4-G7
monoclonal anti-Id (22).
MRL/MpJ-Faslpr (MRL-lpr),
MRL/MpJ (MRL-+/+), and B6.MRL-Faslpr
(B6-lpr) mice were obtained from The Jackson Laboratory (Bar
Harbor, ME). MRL/MpJ-Fasgld
(MRL-gld) were derived from breeders originally provided by
Dr. C. Sidman (University of Cincinnati, Cincinnati, OH). Double-mutant
MRL-lpr/gld mice were bred and maintained at Boston
University School of Medicine (30).
MRL-lpr.Ighb were originally obtained
from Dr. R. Eisenberg (University of Pennsylvania Medical School,
Philadelphia, PA); the MRL-lpr.Ighb
and light chain Tg MRL-lpr V8 and
MRL-lpr.Ighb V8 were bred at Yale
University School of Medicine (New Haven, CT).
Serum samples and Abs
Blood samples collected by tail bleed or cardiac puncture were allowed to clot at room temperature for 1–2 h at which time the serum was removed and stored in aliquots at -80°C. The nucleosome-specific mAbs PR1-3, PL2-6, PL2-3, LG4-1, MRB-4, LG10-1, LG8-1, PL9-7, PL2-8, PL2-7, and MGC 23 were derived from MRL mice and purified by protein G affinity chromatography (31, 32, 33, 34). The IgG2a anti-Sm mAb 2-12 (Igh-1j) was provided by Dr. S. Clarke (University of North Carolina, Chapel Hill, NC) (35). The IgG2a anti-hapten mAbs 23 (Igh-1j anti-NIP), 15G (Igh-1b anti-NIP), Hy1.2 (Igh-1a anti-TNP), and C4010 (Igh-1b anti-TNP) (9, 16) were purified by protein A affinity chromatography.
Isolation of serum IgG2a
Serum IgG2a was isolated by anti-Ig affinity chromatography. An anti-IgG2a column was made by coupling the mAb TT1 (IgG1 anti-IgG2a) (36) to CNBr-activated Sepharose 4B (Amersham Pharmacia Biotech, Uppsala, Sweden). The control column was coupled with the mAb 44-10 (IgG1 anti-Ars) (37). Aliquots of an MRL-lpr/gld serum pool, diluted to 4% in serum-free RPMI, were passed through either the anti-IgG2a column or the irrelevant IgG1 column. After the flowthrough was collected, the columns were washed extensively with PBS. Bound Ab was eluted with 0.1 M glycine-HCl, pH 2.8, and the eluate was neutralized immediately with 1 M NaOH. After neutralization, the eluate was filtered through a 0.2-µm pore size filter, and 50 µl of the filtered eluate were added to a total well volume of 200 µl in the experiments. Neutralized elution buffer alone was always included in the experiments as a control.
B cell activation
T-depleted spleen cell suspensions were obtained from MRL AM14 Tg (RF+) mice or MRL nontransgenic (RF-) littermates by treatment with the mAbs GK1.5 (anti-CD4), 53.6.72 (anti-CD8) and 13-4 (anti-Thy-1.2) followed by rabbit C' (Pel-freez Biologicals, Rogers, AR). In some experiments, the T-depleted spleen cells were preactivated through CD40 with a CD40 ligand (CD40L)-CD8 fusion protein and anti-CD8 mAb as previously described (28). Serum samples, mAbs, or F(ab')2 goat anti-mouse IgM (15 µg/ml, Jackson ImmunoResearch Laboratories, West Grove, PA) were diluted in RPMI 1640, 10% FCS, filter sterilized, and then added directly to the splenic B cells or added 24 h after the CD40L fusion protein. Cells were incubated at a final concentration of 2 x 106 cells/ml in RPMI 1640, 10% FCS for a total of 36–48 h in 96-well flat-bottom microtiter plates. 3[H]Thymidine (New England Nuclear, Boston, MA) was added for the last 6–16 h at a final concentration of 25 µC/ml, DNA was isolated with a PHD Harvester (Cambridge Technology, Cambridge, MA), and thymidine incorporation was measured with an LKB Wallac 1212 Rackbeta counter (Perkin-Elmer, Wellsley, MA). Data are presented as the mean cpm of triplicate cultures or as a percentage of the anti-IgM response. For B cells pretreated with CD40L, the percentage anti-IgM response was calculated according to the formula [(cpm experimental condition - cpm CD40L alone)/(cpm anti-IgM - cpm CD40L alone)]. For B cells not pretreated with CD40L, the percentage anti-IgM response was calculated according to the formula [(cpm experimental condition - cpm medium alone)/(cpm anti-IgM - cpm medium alone)]. The response of the anti-IgM treated RF+ B cells ranged from 23 to 42% of the response of the anti-IgM treated RF- B cells.
RIA and ELISA for detection of IgG2a, RFs, and ICs
Serum IgG2a titers were determined by competitive RIA as
described previously (36). Circulating RF titers were
determined by direct binding ELISA using a
2a
target Ab (23) and an
anti- detecting Ab as described previously (16). IC
were measured using a modified C1q binding assay (38).
Costar Serocluster U vinyl plates (Costar, Cambridge, MA) were coated
with 10 µg/ml human C1q (Calbiochem-Novabiochem, La Jolla, CA) in
PBS. Plates were blocked with PBS, 5% nonfat milk for 2 h at room
temperature. Experimental samples were titrated in PBS, 0.1% Tween 20,
0.2% BSA (PBS-Tween), added to the plates, and incubated at 37°C for
70 min. The plates were then washed, and biotin-conjugated
F(ab')2 goat anti-mouse IgG (Southern
Biotechnology Associates, Birmingham, AL) diluted 1:5000 was incubated
on the plates overnight at 4°C. Plates were developed with a 1:7500
dilution of streptavidin-HRP (Southern Biotechnology Associates)
followed by ABTS (Sigma, St. Louis, MO). Positive and negative controls
consisted of unmanipulated and heat-aggregated (HA, 63°C for 25 min)
mIgG (Sigma). Nonspecific binding was assessed on duplicate plates not
coated with C1q.
Preparation of nucleosomes
Nucleosomes were isolated as described previously
(39). Briefly, nucleosomes were purified from bovine
thymus (Pel-Freez) by micrococcal nuclease (Worthington Biochemical,
Freehold, NJ) digestion of disrupted nuclei and subsequent separation
of mononucleosomes, dinucleosomes, and higher oligomers on a 5–30%
sucrose gradient. Nucleosome fractions corresponding to mononucleosomes
and dinucleosomes were then 5'-end labeled with
[-33P]ATP (New England Nuclear, Boston, MA)
using T4 polynucleotide kinase (New England Biolabs, Beverly, MA). The
anti-nucleosome mAbs, PL2-3 and PR1-3, and an isotype-matched
anti-NIP mAb, 23, were diluted to 20 µg/ml in PBS and added to a
96-well plate (Costar) for 16 h at 4°C. The plates were then
blocked with 0.2% BSA in PBS for 2 h at room temperature and the
-33P-labeled nucleosomes were then added at
10,000 cpm/well for 90 min at room temperature. After extensive
washing, the extent of nucleosome binding was determined by measuring
ß emission in a liquid scintillation counter.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
AM14 mice express a transgene encoded receptor specific for IgG2aa/j and the majority of B cells in BALB/c Tg+ mice appear as naive RF (RF+) B cells (9). However, on the B6/lpr.Igha, but not B6/lpr.Ighb, background, these RF+ B cells become spontaneously activated (24). We reasoned that this spontaneous activation resulted from the inability of autoreactive T cells to eliminate Fas-deficient B cells and/or unique features of the serum components in these mice. To further explore the potential role of serum components in the activation of such RF+ B cells, it was important to determine whether any concentration or form of the IgG2aj autoantigen could effectively engage the AM14 Ag receptor.
As a first approach, serum pools were collected from 4- to 5-mo-old MRL-+/+ and MRL-lpr mice for use as a source of IgG2a. Based on previous serological data, it was known that on the average the concentration of IgG2a in MRL-+/+ sera is higher than those in sera from AM 14 mice, and that the concentration of IgG2a in the MRL-lpr sera is even greater (9, 40). It was also expected that the IgG2a in the MRL-+/+ sera would be essentially monomeric, whereas a significant proportion of the IgG2a in the MRL-lpr sera would be in the form of IC (14).
The MRL sera were added to cultures of CD40L-preactivated B cells
obtained from either RF+ or RF- mice, and
proliferation was assessed after 2 days of culture. CD40L was used as a
source of pseudo-T cell help and under the conditions of these
experiments routinely elicited only modest proliferation by itself. The
B cells were also cultured with a combination of CD40L and
F(ab')2 anti-IgM, to provide a direct
comparison of the overall proliferative capacity of the RF+
and RF- B cell preparations. As shown in Fig. 1
A, both B cell populations
proliferated moderately in response to CD40L alone and vigorously to
the combination of CD40L and F(ab')2
anti-IgM. The RF+ B cells could also be effectively
stimulated by serum pools from either MRL-lpr or double
mutant MRL-lpr/gld mice, but not by serum pools
from MRL-+/+ mice. In contrast, the RF- B cells were not
stimulated by any of the MRL sera, demonstrating that the effect of the
autoimmune sera on RF+ B cells could not be attributed to
agonist-like autoantibodies (41) or excessive cytokine
titers present in the sera. Because the responses of the
RF+ cells to CD40L alone as well as to the rabbit
F(ab')2 anti-IgM was routinely lower than the
corresponding response of the RF- B cells, the data were
recalculated as the percentage of the F(ab')2
anti-IgM response to best reflect the inherent differences between
the two populations (Fig. 1B). Dilutions of autoimmune serum
pools were further tested on resting and CD40L-preactivated B cells.
Although the extent of activation of the resting cells was
significantly less than that of the CD40L-preactivated cells, the
overall pattern remained the same (Fig. 2); the MRL-lpr and
MRL-lpr/gld serum pools stimulated
RF+ B cells, and the MRL/+ sera did not.
|
|
To stimulate, a serum sample needed to contain IgG2a of the
appropriate allotype. Serum pools from B6-lpr mice
(IgG2ab) and allotype congenic
MRL-lpr.Ighb mice failed to stimulate
RF+ B cells, consistent with the allotype specificity of
the AM14 RF receptor (Fig. 3). Moreover,
when an MRL-lpr serum sample was passed through an affinity
column coupled with monoclonal anti-IgG2a, the stimulatory material
was depleted from the flowthrough and could be recovered in the eluate.
No activity could be eluted from a column coupled with an irrelevant
IgG1 (Fig. 4). Overall, these results
were consistent with the assumption that IgG2a present in the
MRL-lpr serum, but not the MRL-+/+ serum, was able to
stimulate RF+ B cells.
|
|
The stimulatory capacity of the MRL-lpr serum pool in comparison with the MRL-+/+ serum pool was not simply due to a higher concentration of IgG2a in the MRL-lpr sample. When IgG2a titers of various MRL-lpr and MRL-+/+ serum pools were measured by RIA, it was found that wells supplemented with 0.3% MRL-lpr serum routinely contained less total IgG2a than the wells supplemented with 3.0% MRL-+/+ serum, even though RF+ B cells cultured in the presence of 0.3% MRL-lpr cultures were effectively stimulated whereas RF+ B cells cultured with 3.0% MRL-+/+ serum were not.
Sera from individual MRL-lpr mice vary widely with regard to
IgG and specific autoantibody titers (14, 16). To
determine whether the stimulatory activity apparent in the serum pools
from the Fas/FasL-deficient mice reflected a general property of
autoimmune sera, samples from 21 Fas/FasL-deficient
(MRL-lpr/gld, MRL-gld, or
MRL-lpr) mice and 8 individual MRL-+/+ mice, ranging in age
from 2.5 to12 mo, were screened for their ability to stimulate
RF+ B cells. As shown in Fig. 5, sera from younger (11-wk)
MRL-lpr/gld mice were not stimulatory. However, 14 of 18
serum samples from mice 4–6 mo of age, added to the cultures at a 1%
final concentration, elicited responses that ranged from 33 to 117% of
the anti-IgM response. By contrast, most of the MRL-+/+ sera had no
detectable activity; only when MRL-+/+ sera were obtained from mice
that were at least 9 mo old did they show any ability to stimulate, and
even then the levels were low.
|
, and
the data from the entire panel are summarized in Table I. Overall, the results demonstrated a
significant, although not absolute, correlation between stimulatory
activity and C1q binding activity (Fig. 6 and Table I), implicating IC
as the key element in RF activation. However, because the sera
indicated by the open and closed inverted triangles in Fig. 6 have
comparable IgG2a levels and C1q binding activity but differ in their
ability to activate RF+ B cells, the simple presence of IC
cannot be the only factor. The extent to which particular features of
IC, such as size, conformation, valency, or incorporation of complement
components, may be relevant to activation of RF+ B cells is
under investigation.
|
|
To further characterize the nature of the stimulatory Ab, we
decided to examine the stimulatory capacity of specific monoclonal
autoantibodies. Because nucleosomes are a common autoantigen in both
murine and human systemic autoimmune disease (42, 43),
initial studies were conducted with several MRL-lpr-derived
IgG2a anti-nucleosome mAbs (31, 34). Nucleosomes were
purified from bovine thymus as described previously (39).
To demonstrate that the isolated material was in fact nucleosomes that
could be bound by the mAbs, nucleosome fractions 4 and 7 (Fig. 7A) were labeled with
[-33P]ATP. Microtiter plates were then
coated with the IgG2aj anti-nucleosome mAbs
PL2-3 and PR1-3, or with an allotype-matched anti-NIP mAb, 23. The
plates were blocked, and the radiolabeled nucleosomes were shown to
specifically react with PL2-3 and PR1-3 (Fig. 7B). These
anti-nucleosome mAbs, either alone or in combination with purified
nucleosomes, were then added to cultures containing the RF+
and RF- B cells. We expected the anti-nucleosome mAbs
to stimulate only in the cultures supplemented with free nucleosomes
(to allow for IC formation). Surprisingly, we found that the
anti-nucleosome mAbs stimulated the RF+ B cells more
strongly than the F(ab')2 anti-IgM, even in
the cultures without added nucleosomes (Fig. 7C). This was
true over a wide range of concentrations, even when non-CD40L-activated
B cells were used in the assay (data not shown). These same mAbs had no
effect on the RF- B cells. A monoclonal
IgG2aj anti-Sm Ab, 2–12, was also found to
specifically stimulate the RF+ B cells under these
conditions. In contrast, anti-hapten mAbs of the same isotype and
allotype, purified in a comparable manner, failed to stimulate the
RF+ B cells. Presumably, the anti-nucleosome mAb bound
nucleosomes released by dying cells present in the initial ascites
fluids (44) and/or in the B cell cultures per se and
formed IC even without the addition of the purified nucleosome
preparation. This assumption was supported by data from a C1q binding
assay where it was shown that anti-nucleosome mAbs bound to C1q
better than the anti-hapten mAbs (Fig. 8A).
|
|
2a), PL2-6 (2b), and PL2-8 (2b) that expressed comparable
variable domains but differed with regard to heavy chain isotype. The
IgG2a mAbs stimulated RF+, but not
RF-, B cells (Table II), thereby demonstrating that
anti-nucleosome Abs were not inherently mitogenic and that
stimulation by anti-nucleosome mAb was dependent on specific
recognition by the AM14 receptor.
|
If the failure of the anti-hapten Abs to stimulate was simply
due to their monomeric status, then IC of the same Abs would be
expected to stimulate. The anti-hapten mAbs were therefore warmed
to 63°C for 25 min to form heat aggregates. The IC status of the HA
Abs was confirmed by C1q binding (Fig. 8
A). When the HA
preparations were compared with the nonaggregated Abs in an
RF+ B cell proliferation assay, the HA Abs were found to
stimulate effectively, although not as well as the anti-nucleosome
mAbs (Fig. 8B). The conclusion from these experiments is
that AM14 RF+ B cells can be effectively stimulated only by
IgG2a in the form of IC, and not by monomeric IgG2a.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The main parameter assessed in the current study was proliferation; autoimmune sera were found to induce a vigorous proliferative response specifically in the AM14 RF+ B cells. In that effective cross-linking of the BCR has been shown by others to up-regulate MHC class II Ag-processing mechanisms (45), it follows that IC stimulation of RF+ B cells is likely to enhance the ability of these cells to present autoantigen epitopes contained within the complex. If the corresponding autoreactive T cells come in contact with the IC-stimulated B cells, the resulting cognate interactions could lead to autoantibody production, especially in Fas/FasL-deficient mice where activated autoreactive T and B lymphocytes are unaffected by signals that normally lead to FasL-induced apoptosis. Because RF+ B cells can presumably be stimulated by IC containing a wide range of autoantigens, they have the potential to both activate and be stimulated by T cells with diverse autoantigen repertoires. This aspect of epitope spreading might explain why RF+ B cells constitute such a high percentage of the MRL-lpr autoantibody repertoire (16, 17).
In contrast to the stimulatory capacity of the IC described in the
current study, there is a considerable literature documenting the
capacity of IC to very effectively block B cell activation by
cross-linking the B cell surface IgM (sIgM) and FcRIIB receptors
(46, 47). The validity of these studies has been further
supported by recent data demonstrating that FcRII-deficient mice
develop higher Ab titers than FcRII-sufficient mice when challenged
with either T-dependent or T-independent Ags (48). The
results of the current study cannot simply be explained by failure of
FcRIIB receptor to bind Abs of the IgG2a subclass, given that IgG2a
anti-SRBC and anti-hapten Abs have been shown to block B cell
activation very effectively (49, 50). The enhancing effect
of CD40L in this system may result from its ability to override FcR
inhibition. It is also possible that the FcR-binding site on IgG2a
molecules is blocked by the AM14 receptor. Another possibility is that
not all IC are comparable and that the "stimulatory" IC present in
autoimmune sera can effectively cross-link sIgM receptors with one
another but cannot effectively cross-link sIgM and FcRIIB receptors.
"Inhibitory" IC, by comparison, would be able to effectively
cross-link sIgM and FcRIIB receptors.
The size and configuration of circulating IC have been extensively studied by Mannik and colleagues. Although in general, small-latticed IC (containing 1–2 Ab molecules) were shown to persist in the circulation longer than soluble large-latticed IC (containing >2 Ab molecules), relatively high levels of large-latticed C1q-binding IC were detected in the sera of diseased MRL-lpr mice. Such large-latticed IC presumably circulate longer in autoimmune mice where the reticuloendothelial uptake system is saturated, thus slowing the IC clearance mechanisms (38, 51, 52). Because large-latticed IC are preferentially detected in the C1q assay (51), the data showing a correlation between the capacity of sera to bind C1q and stimulate RF+ B cells strongly suggest that in fact large-latticed IC present in the Fas/FasL-deficient sera are responsible for the stimulatory activity. How these serum IC compare to the experimentally assembled IC used for the blocking studies cited above remains to be determined.
It is remarkable that Abs reactive with prototypic systemic lupus erythematosus (SLE)-associated nuclear and cytoplasmic autoantigens were found to activate RF+ B cells particularly well, even without the addition of exogenous autoantigen. The high potency of the anti-nucleosome/nucleosome complexes that evidently form spontaneously also suggested that there is some aspect of the size, structure, or valence of these complexes that made them inherently unique with regard to their ability to stimulate B cells. Nuclear and/or cytoplasmic components released in vivo from dying cells may form complexes of comparable structure that effectively activate RF+ B cells. Moreover, the non-Ig autoantigens in such IC may also serve to activate non-RF autoantigen-specific B cells. The formation of such complexes might explain why the injection of lpr mice with certain monoclonal anti-Sm Abs increases the percentage of lpr mice that go on to make anti-Sm autoantibodies (53). Perhaps these anti-Sm Abs form IC capable of stimulating either RF or Sm-specific B cells and thereby facilitate recognition by Sm-specific autoreactive T cells that then drive subsequent B cell expansion and differentiation.
The strongest responses in the current study resulted from costimulation of resting B cells with IC (or anti-IgM) together with CD40L, included as a substitute for a critical aspect of T cell help. Significant, albeit lower, responses were also induced in the absence of CD40L. Exactly how these in vitro responses reflect in vivo B cell status is not clear. In this regard, non-CD40L-pretreated AM14 B cells from both IgG2aa/j and IgG2ab mice could be activated by the anti-nucleosome Abs (data not shown), indicating that prior activation through the AM14 receptor in vivo was not a factor. Indeed, effective receptor engagement in vivo, in the absence of T cell help, often leads not to B cell expansion but rather to tolerance induction, either deletion, receptor editing, or anergy, as demonstrated in numerous BCR Tg models (1, 2, 3, 4, 5, 6, 7, 8, 10). Ineffective engagement, as in the case of the AM14/IgG2aa mouse (9) or 3-83/soluble H-2Kk mouse (54), may well have little impact. Based on the tolerance models alluded to above, it is reasonable to assume that in the absence of T cells, effective in vivo engagement of B cells by IC may also serve as a tolerogenic signal. Furthermore, it may well be that cytoplasmic and nuclear autoantigens routinely complex with "natural" Ab and that such complexes purge the repertoire of potentially hazardous autoreactive B cells. In this regard, Fas-deficient mice genetically altered to lack circulating IgM, and thus potentially deficient in tolerogenic natural IC, have higher autoantibody titers and shorter life spans than Fas-deficient mice with normal levels of circulating IgM (55).
We have shown that autoimmune sera contain IC that convert a relatively benign form of a genuine autoantigen (monomeric IgG2a) into a form (IgG2a IC) that can very effectively stimulate bona fide autoreactive (AM14 Tg RF+) B cells. IC containing SLE-associated nuclear and cytoplasmic autoantigens were found to be particularly potent B cell activators in this system and are likely to contribute to the "spontaneous" activation of RF+ B cells in AM14 Tg+ IgG2a lpr mice. Appropriate manipulation of such autoantibody/autoantigen complexes may eventually provide a means for therapeutic intervention in patients with systemic autoimmune disorders.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 I.R.R. and E.A.L. contributed equally to this study.
3 Address correspondence and reprint requests to Dr. Ann Marshak-Rothstein, Department of Microbiology, Boston University School of Medicine, 80 East Concord Street, Boston, MA 02118.
4 Abbreviations used in this paper: FasL, Fas ligand; HA, heat-aggregated; IC, immune complexes; RF, rheumatoid factor; CD40L, CD40 ligand; SLE, systemic lupus erythematosus; sIgM, surface IgM; Tg, transgenic.
Received for publication February 14, 2000. Accepted for publication May 9, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RII-deficient mice. Nature 379:346.[Medline]
receptor-mediated inhibition of mouse B cell activation. J. Immunol. 134:2835.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  B. C. Beaudette-Zlatanova, T. Ling, M. J. Shlomchik, A. Marshak-Rothstein, and I. R. Rifkin B Cells and Dendritic Cells from V{kappa}8 Light Chain Transgenic Mice Activate MRL-lpr/gld CD4+ T Cells J. Immunol., July 1, 2006; 177(1): 45 - 52. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Fields, M. H. Metzgar, B. D. Hondowicz, S.-A. Kang, S. T. Alexander, K. D. Hazard, A. C. Hsu, Y.-Z. Du, E. L. Prak, M. Monestier, et al. Exogenous and Endogenous TLR Ligands Activate Anti-Chromatin and Polyreactive B Cells. J. Immunol., June 1, 2006; 176(11): 6491 - 6502. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. M. R. Westwood, P. N. Nelson, and F. C. Hay Rheumatoid factors: what's new? Rheumatology, April 1, 2006; 45(4): 379 - 385. [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. Carter B Cells in Health and Disease Mayo Clin. Proc., March 1, 2006; 81(3): 377 - 384. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. William, C. Euler, E. Leadbetter, A. Marshak-Rothstein, and M. J. Shlomchik Visualizing the Onset and Evolution of an Autoantibody Response in Systemic Autoimmunity J. Immunol., June 1, 2005; 174(11): 6872 - 6878. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Doerre, K. P. Mesires, K. M. Daley, T. McCarty, S. Knoetig, and R. B. Corley Reductions in I{kappa}B{epsilon} and Changes in NF-{kappa}B Activity during B Lymphocyte Differentiation J. Immunol., January 15, 2005; 174(2): 983 - 991. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Latz, A. Visintin, T. Espevik, and D. T. Golenbock Mechanisms of TLR9 activation Innate Immunity, December 1, 2004; 10(6): 406 - 412. [Abstract] [PDF]
|
|
|
|
 |
|
 A. R. Ferguson, M. E. Youd, and R. B. Corley Marginal zone B cells transport and deposit IgM-containing immune complexes onto follicular dendritic cells Int. Immunol., October 1, 2004; 16(10): 1411 - 1422. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Marshak-Rothstein, L. Busconi, C. M. Lau, A. S. Tabor, E. A. Leadbetter, S. Akira, A. M. Krieg, G. B. Lipford, G. A. Viglianti, and I. R. Rifkin Comparison of CpG s-ODNs, chromatin immune complexes, and dsDNA fragment immune complexes in the TLR9-dependent activation of rheumatoid factor B cells Innate Immunity, August 1, 2004; 10(4): 247 - 251. [Abstract] [PDF]
|
|
|
|
|
|
M. P. Cancro and J. F. Kearney B Cell Positive Selection: Road Map to the Primary Repertoire? J. Immunol., July 1, 2004; 173(1): 15 - 19. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. W. Boule, C. Broughton, F. Mackay, S. Akira, A. Marshak-Rothstein, and I. R. Rifkin Toll-like Receptor 9-Dependent and -Independent Dendritic Cell Activation by Chromatin-Immunoglobulin G Complexes J. Exp. Med., June 21, 2004; 199(12): 1631 - 1640. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L E Schiffer, N Hussain, X Wang, W Huang, J Sinha, M Ramanujam, and A Davidson Lowering anti-dsDNA antibodies--what's new? Lupus, December 1, 2002; 11(12): 885 - 894. [Abstract] [PDF]
|
|
|
|
 |
|
 C. H. Nielsen and R. G. Q. Leslie Complement's participation in acquired immunity J. Leukoc. Biol., August 1, 2002; 72(2): 249 - 261. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
-Ig),
or a final concentration of 3, 1, or 0.3% mouse serum, was then added
to each of the wells. Control wells were given an additional 50 µl
medium (Med). The cells were cultured for an additional 40 h;
proliferation was assessed by [3H]thymidine incorporation
during the last 16 h of culture. Data are presented as the mean
cpm of triplicate cultures ± SD (A) or as the
percentage of the anti-IgM response (B).
), MRL-lpr/gld (
),
MRL-gld (
), or MRL-lpr (
) mice of
various ages and tested for their ability to stimulate CD40L-pretreated
RF+ B cells when added to cultures at a final concentration
of 1%.
