| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Division of Medicine, Hadassah University Hospital, Faculty of Medicine of the Hebrew University, Jerusalem, Israel; and
Institute for Genetics, University of Köln, Köln, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
utilization and DNA binding of hybridoma Abs
suggested that immature B cell deletion and light (L) chain editing
were the major mechanisms affecting tolerance. Unexpectedly, these
mechanisms were less effective in targeted mice expressing the
somatically mutated, anti-DNA H chain than in mice expressing the
germline-encoded H chain, possibly due to the greater abundance of high
affinity, anti-DNA immature B cells in the BM. Consequently,
autoreactive B cells that showed features of clonal anergy could be
recovered in the periphery of these mice. Our results suggest that
clonal deletion and receptor editing are interrelated mechanisms that
act in concert to eliminate autoreactive B cells from the immune
system. Clonal anergy may serve as a back-up mechanism for central
tolerance, or it may represent an intermediate step in clonal
deletion. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Transgenic mice expressing a single autoantibody specificity by nearly all B cells have provided excellent in vivo models for the study of tolerance mechanisms. B cells bearing Ig receptors for surface Ags, such as MHC class I, were deleted from the repertoire upon encounter with the cross-linking Ag (20), while those reacting with a soluble protein, like hen egg lysozyme (HEL) were subject to anergy induction, accompanied by down-modulation of their surface IgM receptors (21). The anergic state could be partly reversed by nonspecific mitogens, such as bacterial LPS; under these circumstances, differentiation into Ab-secreting plasma cells was inhibited by the continued binding of the autoantigen to surface Ig receptors (22).
Anti-DNA transgenic mice have also been constructed in an effort to explore functional and aberrant mechanisms of immune tolerance in normal and lupus-prone mice (23, 24, 25). The DNA Ag has been found to exhibit an intermediate behavior between soluble and immobilized Ags, since features of both clonal deletion (26) and anergy (23, 25, 27) could be demonstrated in these mice. Ab affinity was found by Chen et al. (26) to play a major role in the extent of anti-DNA B cell deletion. In contrast, the threshold affinity for deletion of anti-MHC class I autoreactive B cells was very low and could hardly be measured (28). The role of anti-DNA affinity in the induction of B cell anergy has not yet been clarified.
Recently, a novel mechanism of immunologic tolerance that enables
autoreactive B cells to escape clonal deletion by revising their Ag
receptors has been proposed (29, 30, 31) and termed "receptor editing"
(reviewed in 32 . This mechanism allows for further rearrangements
in the L chain (29, 30, 31) or H chain (33) gene loci. Several lines of
evidence support the occurrence of receptor editing, particularly
affecting L chain specificity; these include the up-regulation of RAG-1
and RAG-2 mRNA in the BM and the excess levels of variant B cells
bearing 
L chains in the spleen (28, 29), as well as the marked
shift in the bias of J usage from the V-proximal J1 and J2
(34, 35) to the distal J5 (31).
In the study reported here, we have utilized the recently developed technique of targeted insertion of a rearranged variable region gene into the mouse Ig H chain locus (36), to obtain transgenic "knock in" mice, carrying anti-DNA H chains with different affinity for DNA. Autoreactive H chain-only transgenic mice have been used by several groups of investigators (26, 31, 33, 37) and provide a useful tool for studying populations of B cells that are restricted by a common transgenic H chain but possess a wide variety of endogenous L chains from the entire mouse repertoire. The inserted H chains in this study were derived from the anti-DNA hybridoma D42 (38, 39, 40) that binds both ssDNA and native dsDNA. This hybridoma represents a group of mouse anti-DNA autoantibodies (about 5% of NZB/NZW anti-DNA; 7 whose H chains are encoded by the VH11 (S107) gene segment. Two separate mouse lines were constructed: one carrying the unmutated, germline-encoded VH11 gene and the other carrying the original D42 H chain containing two replacement somatic mutations, a Ser to Asn at position 35 in CDR1 and an Asn to Tyr at position 53 of CDR2. In both cases the VH11 gene is rearranged to a D segment encoding an arginine-rich CDR3 (39, 40) that was found to be essential for DNA binding (40). The difference in DNA affinity between the two H chains, in combination with the D42 L chain was about 12-fold (40). Our results suggest that the major mechanisms of tolerance that prevent B cell autoreactivity in these transgenic mice are clonal deletion and L chain editing; however, the extent of these processes is not correlated with receptor affinity in a simple way. Additionally, anti-DNA B cells that escape deletion and receptor editing do not undergo class switching or somatic mutation and show several features of clonal anergy.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The targeting vector pIVhL2neoR(B6SH)(gl)D42H was constructed by replacing the short homology (SH) region (0.8-kb JH-Cµ intron fragment) of pIVhL2neoR vector (41) with the corresponding fragment derived from C57BL/6 genomic DNA. The glD42H and D42H cassettes, including promoter and rearranged VDJ genes, were excised from the glD42 or D42 expression vector, respectively (40), and cloned into the SalI/ClaI site of pIVhL2neoR(B6SH) to generate the targeting vector. NotI-linearized targeting constructs (80 µg) were transfected by electroporation into 4 x 107 B6III ES cells (42). The transfected cells were selected with G418 (300 µg/ml) and Gancyclovir (2 µM). Double-resistant clones were identified by PCR using a VH11 sense primer and an antisense primer located 3' to the IgH intron enhancer. Putative targeted transfectants, positive for a 1.3-kb PCR fragment, were further analyzed by Southern blotting with a probe spanning the EcoRI-HindIII region, 3' to the IgH intron enhancer. Homologous recombinants were identified by a 6.1-kb band, corresponding to the glD42- or D42-targeted allele, in addition to the 2.3-kb band representing the wt allele. To delete the neor gene, 1 x 107 targeted ES cells were transfected by electroporation with 40 µg of the circular Cre-encoding plasmid pGCre-4 (obtained from Dr. H. Gu, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Twinbrook, MD). Targeted clones lacking the neor gene were identified by having a 5.1-kb band in addition to the 2.3-kb wt band. ES cell clones bearing the rearranged D42VDJ gene or the glD42VDJ gene (with or without the neor gene) were injected into blastocytes of CB.20 mice and transplanted into the uteri of (C57BL/6 x BALB/c)F1 foster mothers. Chimeric mice were mated to C57BL/6 mice.
Flow cytometric analysis
Single cell suspensions from BM, spleen, and peripheral blood were three-color stained with monoclonal or polyclonal Abs and analyzed by FACScan (Becton Dickinson, San Jose, CA) using "Lysis II" program. The following Abs were used: FITC-, PE- or biotin-conjugated MB86 anti-µb (43), FITC-, PE- or biotin-conjugated RS3.1 anti-µa (44), PE-conjugated RA3-6B2 anti-CD45R/B220 (45), FITC-conjugated CFO-1 (anti-Thy1.2) (46), FITC-conjugated S7 anti-CD43 (47), and RF-81, a rabbit anti-D42H Id, primarily directed to VH11 epitopes (R. Fischel and D. Eilat, unpublished observations). Biotin conjugates were revealed by Cy-chrome-, PE- or FITC-conjugated streptavidin (PharMingen, San Diego, CA). RF-81 was revealed by FITC- conjugated F(ab')2 goat anti-rabbit IgG (Southern Biotechnology Associates, Birmingham, AL).
Mouse immunizations
Influenza virus immunization. Wild type and mutant (C57BL/6 x BALB/c)F1 mice (2–4 mo old) were challenged intranasally with 10-6 egg infectious dose (EID50) of live PR8 influenza virus and boosted 3 wk later i.p. with 1200 hemagglutination units (HAU) in 200 µl PBS. Allotype-specific IgG2a serum levels were analyzed 6 days later by ELISA, and fusion of myeloma cells with splenocytes from selected immunized mice was conducted on the following day.
DNA immunization. Wild type (C57BL/6 x BALB/c)F1 mice (3 mo old) and transgenic littermates were immunized i.p. with a Fus1/DNA complex in CFA, then boosted with the same complex in IFA and in buffer 3 and 6 wk later, respectively, according to the protocol described previously (8). Allotype-specific IgM serum levels were analyzed 1 wk after the last immunization.
Ab binding tests
ELISA tests were conducted according to the general procedure described previously (40). Briefly, microtiter plates were incubated at 4°C overnight with 50 µl/well PBS containing 5 µg/ml goat anti-mouse µ-specific Ab (Jackson ImmunoResearch, West Grove, PA), 250 hemagglutination units/ml of live PR8 influenza virus or 100 µg/ml phenol-treated salmon-sperm DNA (Sigma, St. Louis, MO), for testing allotype-specific IgM, anti-influenza, or anti-DNA Abs, respectively. In selected DNA binding experiments, the DNA was treated with S1 nuclease (New England Biolabs, Beverly, MA) and phenol extracted, or heat denatured, before its immobilization. For the detection of allotype-specific IgM Abs and IgM anti-DNA, biotin-conjugated RS3.1 or MB86 mAbs (described above) were employed, while Ig(1a)8.3 mouse anti-mouse IgG2aa (48) and G12–47/30 mouse anti-mouse IgG2ab (G. Seemann and K. Rajewsky, unpublished observations) were used to detect IgG2a allotypes. Biotin-conjugated reagents were further reacted with streptavidin-conjugated alkaline phosphatase (Southern Biotechnology Associates) followed by p-nitrophenyl phosphate developing substrate. The details of the nitrocellulose filter DNA binding assay have been described previously (40).
PCR and sequence analysis of hybridoma Abs
Hybridomas were produced by fusion of BALB/c NSO myeloma cells
with spleen cells from wt or mutant mice following 3 days incubation
with 40 µg/ml LPS (Sigma). To analyze the targeted locus genotype of
IgMa expression clones, PCR was preformed on genomic DNA
using a sense D42 leader primer (5'-GGCGAATTCATGAAGTTGTGGCTGAACTGG-3')
and an antisense JH1 primer
(5'-GCCTATTTTCCATGATTGTCACTGTTCCAC-3'). When the 0.5-kb targeted
band was not present in the PCR product, an additional PCR was
performed using a sense Dsp2 primer
(5'-ACGAATTCAAAGCACAATGCCTGGCT-3') or a universal
degenerate VH primer
(5'-CGATGAATTCAGGT(GC)(AC)A(AG)CT(GT)CTCGAGTC(AT)GG-3')
(both containing an EcoRI site) and an antisense
D42CDR3 primer (5'-ACTCTAGACCTTCCCCGTCGTAACCCTCC-3')
(containing an XbaI site) for the detection of secondary
rearrangements in the targeted locus. DNA from selected PCR bands
(
0.7 kb) was cloned into a pBluescript plasmid and sequenced with a
T7 Sequenase 2.0 Kit (Amersham, Arlington Heights, IL).
To detect somatic mutations on the targeted D42 H chain, PCR was
performed on hybridoma genomic DNA. PCR primers were sense D42 leader
and antisense JH1. The reaction products were run on a
0.7% low melting agarose gel (SeaPlaque GTG, FMC, Vallensbaek Strand,
Denmark), and a 0.5-kb band was cut out from the gel to perform
"in-gel" direct sequencing (49), using D42CDR3 or D42CDR2
(5'-CTCTGTTGTATACCCATTAGCTTTGTTTC-3') primers. To analyze J
usage of hybridoma clones, total RNA was prepared from 3 x
106 hybridoma cells using TRI Reagent (Molecular Research
Center, Cincinnati, OH). RT-PCR was performed using constant-region
primers and a degenerate Vk primer
5'-CCGCTCTAGAGGCTGCAG(GC)TTCAGTGGCAGTGG(AG)TC(AG)AC-3'. The
PCR products were subjected to direct sequencing (49).
LPS activation of splenocytes
Splenocytes (1.5 x 106) were cultured in RPMI 1640 medium (containing 10% FCS) and different concentrations of LPS, in the presence or absence of 25 µg/ml DNase I (Boehringer Mannheim, Indianapolis, IN). Additional DNase I (25 µg/ml) was added daily to the cultures. This concentration had been calibrated for efficient degradation of 32P-labeled DNA in a preliminary experiment. On day 5, cultured cells were collected, and live cells were counted. Analysis of supernatants for IgMb- and anti-DNA IgMb-secreted Abs was performed by ELISA. Ficoll gradients were performed to enrich for live cells, and the percentage of IgMb- and IgMa-expressing cells was determined by FACScan as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Two lines of Ig H chain transgenic mice were constructed by targeted replacement of the JH locus in C57BL/6 ES cells with the rearranged VH gene segment of the D42 anti-DNA Ab, derived from a diseased NZB/NZW F1 female mouse (38). One mouse strain, designated glD42i (neo), carries the unmutated VH11 germline sequence, combined with an arginine-rich Dsp2.3 JH1 CDR3 (39). A variant of this mouse line was generated (glD42i) that lacks the selection marker gene, following Cre-recombinase mediated deletion (50). We have not found significant differences between the glD42i (neo) and glD42i mouse strains in any subsequent studies, and data of the two glD42i variants can be regarded as interchangeable. The second mouse line, designated D42i (neo), carries the original, somatically mutated D42 VH region.
The construction of H chain-only targeted mice was based on two assumptions: the first was that the VH11 H chain would combine with a large variety (but not all; 40 of L chains to give DNA binding B cell receptors, ranging from very low to high affinity. This was based on previous transfection experiments (40), as well as on the large number of Vk gene families, including Vk1 (6, 7, 51, 52), Vk2 (52, 53), Vk4/5 (52, 54), Vk8 (54), Vk10 (7, 52), Vk12/13 (51), Vk20 (51), and Vk21 (7), whose representatives have been reported in the literature to bind DNA by association with the VH11 (S107) H chain. Moreover, since most of these Vk groups comprise several gene members (55, 56) and different members of the same group (e.g., Vk1, Ref. 52; Vk4/5, Ref. 53; and our unpublished transfection experiments) have been shown to combine with VH11 to give DNA specificity, it is reasonable to predict (based on the complexity of each gene family) that about 80% of the presumed 140 mouse Vk gene segments would combine with the transgene(s) to generate some affinity for DNA.
The second assumption was that the dominance of the H chain in determining affinity for DNA (40) would result in higher affinity anti-DNA B cell receptors for most H/L combinations containing the mutated VH11 H chain, as compared with those having the unmutated H chain. This is supported by the finding that the Ser to Asn mutation at position 35 of VHCDR1 is the most consistent somatic mutation found in VH11-encoded anti-DNA autoantibodies (7, 39, 53, 54). Apparently, Asn at this position in the H chain is very instrumental in DNA binding and may be found also in J558 VH genes, such as 3H9 (2) and B/W DNA-16 (3, 8) that consistently appear in high affinity anti-DNA Abs. This assumption was also supported by preliminary transfection experiments of mutated and unmutated D42H with VKD42 (40), Vk4 and Vk8 (Y. Yarkoni, N. Yachimovich and D. Eilat, unpublished results) in which the mutated H chain induced higher anti-DNA affinity in all of the H/L combinations.
The targeting of the rearranged VDJ genes into the DQ52-JH
region of the C57BL/6 mouse Ig H chain locus (36) is illustrated in
Figure 1
A. Recombinant ES
clones were injected into CB20 blastocytes to generate chimeric mice
that were bred to C57BL/6 mice and gave rise to heterozygous transgenic
mice (Fig. 1B). To distinguish between transgenic and
endogenous H chains by allotype specific Abs, the targeted C57BL/6 mice
were further crossed with unmanipulated BALB/c mice, and the
heterozygous transgenic F1 hybrids were selected for all subsequent
studies.
|
). A FACS analysis of peripheral blood
lymphocytes shows (Fig. 2A) that the majority of B cells in
the two independently constructed transgenic mice expressed the D42 Id.
This Id surface expression was restricted to B cells carrying the
transgenic IgMb allotype (Fig. 2B). Staining of
splenocytes with allotype-specific reagents (Fig. 2C) shows,
in agreement with previous results (41), that allelic exclusion at the
individual B cell level was maintained in the mutant mice and was
similar to that observed in wt F1 mice. However, a small
but significant population (3–14%) of transgenic B cells expressed
the endogenous IgMa allele exclusively. This population,
presumably arising as a result of transgene inactivation, was always
1.5- to 2-fold greater in the D42i mice than in glD42i mice.
|
Measurements of serum IgM and IgM anti-DNA in wt and targeted
mice are shown in Figure 3. The
concentrations of IgMb Igs were reduced to about 20% of
normal levels in glD42i and were almost undetectable in D42i mice (Fig. 3A). One explanation for this finding is that the D42 H
chain in combination with endogenous L chains does not provide the
required B cell Ag receptor (BCR) repertoire for the immune
response to environmental Ags. This is unlikely, however, since the
VH11 (S107) gene, unlike certain other anti-DNA
VH genes (8), has been reported to participate in several
different natural and experimental immune responses (57, 58, 59) (see also
Figure 4). A second explanation fulfills
our initial prediction, namely, that the great majority of transgenic B
cells would have H/L combinations with some affinity for DNA and,
therefore, would be subject to regulatory mechanisms of immune
tolerance. The serum IgG2ab levels of wt and transgenic
mice showed a similar pattern to that of IgMb (results not
shown). Interestingly, the IgMa levels in wt and mutant
mice were similar, although only about 10% of transgenic B cells
expressed the endogenous IgMa allotype. This suggests some
type of homeostatic control of serum Abs, as suggested by Cascalho et
al. (60).
|
|
B). The latter anti-DNA activity is
attributed to spontaneously secreted natural Abs, present in normal
mouse sera (9, 10), since the same sera did not bind radioactive DNA in
the more discriminatory nitrocellulose filter solution assay (8, 40, 61) that measures affinities of 106 M-1/bp or
greater (Fig. 3C). It follows that both high and low
affinity anti-DNA B cells representing the whole range of H/L
receptor combinations in the two targeted mouse lines were secreting
neither high affinity (typical of autoimmune mice) nor low affinity
(typical of normal mice) anti-DNA Abs. Active immunization of targeted mice
To test whether or not the transgenic B cells can be activated by
exogenously administered Ags, wt and transgenic mice were immunized
with influenza virus or with an immunogenic complex of Fus1 peptide/DNA
that are known to elicit strong humoral immune responses to influenza
hemagglutinin and DNA, respectively, with the concomitant production of
VH11-encoded Abs (Refs. 8 and 57; and T. Marion,
unpublished observations). As shown in Figure 4A,
immunization of mutant mice with live influenza virus resulted in an
IgG antiinfluenza immune response of both a and b allotype, indicating
that the targeted B cells could be activated by an exogenous Ag. The
titer of allotype b IgG2a was much lower than that of allotype a since
the arginine-rich CDR3 may not be optimal for binding the influenza
hemagglutinin. This may also apply to the somatic mutations in the
anti-DNA H chain, which would make the immune response of D42i mice
weaker than that of glD42i mice (Fig. 4A). Indeed, the
sequence of the targeted VH11 gene from an antiinfluenza
IgG2a hybridoma derived from an immunized D42i mouse (Fig. 4B) shows, in addition to multiple somatic mutations, a
reversal of the mutation in CDR2 from Tyr to the original germline Asn.
This experiment further demonstrates that the targeted H chain was
capable of physiologic class-switching and of the acquisition of
somatic mutations. Immunization of targeted mice with Fus1/DNA complex
in CFA has led to a minimal IgMb immune response in glD42i
(data not shown) and no response in D42i mice (Fig. 4C). In
contrast, B cells expressing the endogenous a allele showed a vigorous
response to peptide/DNA immunization. This difference is striking,
since a large number of targeted B cells was expected to be uniquely
suitable for binding DNA due to the promoting features (CDR3, somatic
mutations) of the H chain. These results suggest that the potentially
reactive B cells are either absent from the mouse peripheral immune
system, or that their activation by immunogenic DNA is inhibited by
regulatory mechanisms of self tolerance.
Analysis of BM and spleen cells from targeted mice
The absence of anti-DNA Abs in the serum of glD42i- and
D42i-targeted mice could result from deletion and/or functional
inactivation of autoreactive B cells. To address these possibilities,
we have measured the size of the different cellular compartments in the
BM of targeted mice, reflecting the developmentally ordered B cell
subpopulations on the basis of cell surface markers (62). The results
are summarized in Table I. They show
about twofold reduction in the absolute numbers of total BM cells and
total lymphocytes in all mutant mice compared with their wt,
nontransgenic littermates. This difference is also seen in the pro-B
and pre-B cell populations; however, these compartments have similar
relative sizes in wt and mutant mice (5% and 50% of total lymphocytes
for pro B and pre B cells, respectively). This may suggest that the
decrease in total cell numbers results from some differences in
developmental patterns at early stages of B cell maturation, due to the
introduction of a rearranged H chain to the mouse germline (41). In
contrast, a twofold reduction was observed in IgM+ immature
B cells from all of the mutant mouse lines (8.8–9.8% of total
lymphocytes, numbers underlined in Table I), compared with wt mice
(17.5% of total lymphocytes). A similar deletion of immature
B cells was also noted when glD42i B cells were compared with B
cells expressing nonautoreactive
anti-(4-hydroxy-3-nitrophenyl)acetyl (NP) Abs in identically
prepared H chain-targeted mice (41). Moreover, a recently
constructed double transgenic VH11/Vk4 anti-DNA
mice had further reduced the relative number of their immature B cells
to one third of that compartment in Vk4-only targeted mice (N.
Yachimovich and D. Eilat, unpublished observations). These results,
also supported by preliminary Annexin staining data that showed
increased apoptosis of glD42i and D42i immature B cells compared with
wt immature B cells, suggest that immature B cells in
VH11-targeted mice are subject to negative selection by
clonal deletion, probably due to their anti-DNA autoreactivity, at
the pre-B to immature B stage or just after expression of the surface
Ig. Similar results were previously obtained by Weigert and coworkers,
who studied a different anti-DNA transgenic H chain, 3H9 (26).
However, in their study, B cell deletion in the BM was observed only
after the affinity of the anti-DNA H chain was increased by
directed mutagenesis. In contrast, we observe a substantial deletion of
B cells, targeted with the germline-encoded VH11 and only a
minor increase in the extent of deletion of presumably higher affinity
B cells, expressing the somatically mutated anti-DNA H chain (Table I).
|
) and splenic B lymphocytes
(not shown), the number of targeted cells was still reduced, compared
with wt cells (this study) and with targeted cells expressing
nonautoreactive H chain specificities such as
(4-hydroxy-3-nitrophenyl)acetyl (NP) (41), but a substantial recovery
of the B cell compartment was observed (10–12.7% mature B cells in
mutant mice, compared with 15.8% in wt mice; Table I). The fraction of
B cells expressing endogenous IgMa H chains was fairly
constant in immature and mature BM B cells (Table I), as well as in
splenic B cells (not shown), and accounted for 3 to 7% in glD42i mice
and for 6 to 14% in D42i mice. These proportions did not change with
increasing age of the mutant mice (data not shown). Production of B cell hybridomas from VH11-targeted mice
Fusions of splenocytes from the two targeted mouse lines were
conducted with NSO myeloma cells, following in vitro activation with
LPS (Table II), to analyze the transgenic
B cells at the single cell level. Hybridomas were first examined for
the expressed H chain isotype (only IgM- and no IgG-expressing cells
were found) and allotype (Table II). In agreement with the FACS
analysis (Fig. 2C), no double expressing hybridomas were
identified. The fraction of clones expressing the endogenous
IgMa allele in glD42i and glD42i (neo) mice was 2 to 5%,
reflecting the proportion of these cells (3–7%) in BM and spleen
(Fig. 2C and Table I). In contrast, the fraction of
IgMa-expressing clones in D42i (neo) mice was, on average,
eightfold higher than that of glD42i mice (Table II), and this was
significantly different from the relative size of these populations in
BM and spleen, suggesting a selective LPS activation before the fusion
experiment (see below). Since there was no evidence for a lack of
allelic exclusion in B cells from the transgenic mouse lines, the
IgMa-expressing cells were likely to have inactivated their
targeted IgMb allele, possibly by secondary rearrangements
(33, 36, 60, 63), with concomitant deletion of the neor
gene, located 5' to the targeted VDJ gene in glD42i (neo) and D42i
(neo) mice. This indeed was found to be the case for the great majority
of IgMa hybridomas, as judged by their sensitivity to the
neomycin analogue, G418 (Table II). In contrast, the
IgMb-expressing clones, with some exceptions discussed
below, were resistant to G418 due to the maintenance of the
neor gene.
|
). While
IgMb-expressing glD42i hybridoma Abs bound DNA with a very
low frequency (4/126 or 3%), which was lower than the "natural"
IgMa anti-DNA activity (6/72 or 8%), a substantial
number of D42 hybridoma Abs (47/165 or 28%) were reactive with DNA in
ELISA. These hybridomas bound S1 nuclease-treated dsDNA and
heat-denatured ssDNA to about the same extent. None of the D42i
hybridoma supernatants bound dsDNA or ssDNA in the nitrocellulose
filter solution assay, reflecting the low affinity nature of these
interactions (40), and no additional somatic mutations were identified
in direct sequencing of several transgenes. These results indicate
that, following LPS stimulation and fusion with myeloma cells,
presumably anergic, low affinity splenic B cells that are present in
D42i, but not in glD42i mice, can be activated to secrete anti-DNA
Abs. Features of B cell anergy in D42i mice
The finding that a significant fraction (30%) of D42i
IgMb hybridomas secreted anti-DNA Abs contrasted with
our inability to measure anti-DNA activity in the sera of these
mice (Fig. 3). This could result from a functional silencing (anergy)
of the relevant B cells through a continuous B cell antigen receptor
(BCR) engagement with DNA; alternatively, the cells could be
nonresponsive to DNA in vivo due to their low affinity for the
autoantigen. The first clue to B cell anergy came from the significant
increase in the ratio of IgMa- to
IgMb-expressing hybridomas of D42i as compared with glD42i
mice, following LPS activation (Table II). This suggested that the
activation of IgMb-expressing cells may be restricted due
to their DNA reactivity. To dissociate the effects of LPS stimulation
and fusion with myeloma cells, in vitro B cell cultures were set up
with increasing concentrations of LPS in the presence or absence of
DNase I. It had been shown previously (11, 21) that a functional
recovery from the anergic state could be achieved by LPS in vitro;
however, differentiation into Ab-secreting plasma cells was inhibited
by the continued binding of autoantigen to surface Ig receptors.
The results on day 5 of LPS stimulation of wt and mutant splenocytes
are shown in Figure 5. The stimulation of
transgenic glD42i B cells by LPS was essentially unaffected by the
presence of DNase I, which had been calibrated to remove all traces of
DNA from the cell cultures. In contrast, LPS stimulation was low in
D42i cells and could reach the levels of glD42i stimulation only if DNA
was removed from the cell cultures. Since the stimulation of wt B cells
was not dependent on the presence of DNase (Fig. 5A), the
ratio of IgMa- to IgMb-expressing cells was
increased in D42i-stimulated cells, in agreement with the hybridoma
data (Table II). The analysis of cell supernatants for DNA binding on
day 5 of LPS stimulation (Fig. 5B) shows that not only
proliferation of anti-DNA IgMb cells but also their
LPS-induced differentiation into Ab-secreting cells was affected by
DNase I, since elevated levels of IgMb anti-DNA Abs
were observed upon removal of DNA by the enzyme. The finding that the
state of anergy is dependent on the continuous engagement of B cells by
DNA in vitro suggests that anergic cells may also be controlled by
permanent Ag encounter in vivo.
|
C, the major
population of IgMb-positive cells in the two transgenic
lines had surface IgM densities that were significantly lower than
those observed for wt IgMb-expressing cells (mean
fluorescence of 190). However, while glD42i mice still contained a
minor population of cells expressing wt IgM densities (mean
fluorescence of 465), these cells were essentially absent in D42i mice.
Instead, a prominent population of low density IgM-expressing B cells
was evident in D42i mice (mean fluorescence of 57). This population may
represent the anergic B cells with anti-DNA autoreactivity. The
fluorescence intensity of IgD did not differ between glD42i and D42i B
cells (not shown). Secondary rearrangements and receptor editing in VH11-targeted mice
The presence in the targeted mice of B cells expressing the
endogenous IgMa allele (Fig. 2C and Table I) and
the G418 sensitivity of most IgMa hybridoma suggested that
the transgene had been inactivated by secondary rearrangement(s) in
these cells. Indeed, Southern blot analysis of IgMa
hybridoma DNA showed that the 6.1-kb band representing the intact
targeted allele (Fig. 1B) was replaced by a different band
in each individual clone (data not shown). Additionally, PCR analysis
of IgMa hybridoma DNA using D42 leader and JH1
primers confirmed that all IgMa-expressing cells had lost
their D42 leader sequence (data not shown). To further study the nature
of these secondary rearrangements, DNA from IgMa-expressing
clones was amplified with Dsp2-D42CDR3 and universal
VH-D42CDR3 primers, and the resulting PCR products were
subjected to DNA sequencing (Fig. 6). The
sequences revealed that, similarly to previous analysis (63), two
heptamer-like motives, located in the leader intron of the
VH11 gene, served as acceptors for upstream D and/or
VH gene segments. Of 27 tested IgMa-expressing
hybridoma clones, 14 had D to VDJ rearrangements (Fig. 6A),
one had VD to VDJ rearrangement (Fig. 6B), and one had a
putative V to VDJ rearrangement (Fig. 6C), although the
absence of D in this case could not be proven. Putative N sequences
were identified in all joints. Secondary rearrangements that led to a
functional replacement of the targeted gene by an upstream
VH gene (33) were apparently very rare in the D42i mice,
since only one IgMb hybridoma was identified that fulfilled
the two criteria of G418 sensitivity and the absence of the D42 leader
sequence. The relevance of H chain secondary rearrangements to receptor
editing is discussed below.
|
L chain locus allows for
secondary rearrangements of upstream Vk gene segments to downstream
J elements, which result in deletion or inactivation (by inversion)
of primary VkJ combinations. Secondary rearrangements of endogenous
L chains in VH11-targeted B cells were indirectly inferred
by a shift in J usage from the V-proximal J1 plus J2 to the
distal J5, as previously suggested by Weigert and his associates
(31, 65, 66, 67). Table III summarizes the
frequency of J gene expression in the different categories of
mutant hybridomas (described in Table II), as compared with hybridomas
derived from wt mouse littermates. The distribution of J usage in wt
mouse hybridomas resembled the reported frequencies in the
literature (34, 35, 66), in that about 80% of Vk gene segments were
rearranged to J1 or J2 and only 17% used J5. In a sharp
contrast, 60 to 70% of non-DNA binding, IgMb-expressing
transgenic B cells (p < 0.002) had their Vk
genes rearranged to J5 (Table III). This striking shift in J
utilization is best explained by editing of autoreactive B cell
receptors through secondary Vk rearrangements, since hybridomas
expressing endogenous IgMa alleles did not differ in J5
usage from wt hybridomas. Additionally, the pattern of J
distribution in wt and targeted mice is suggestive of a stepwise
editing mechanism (no rearrangements to J1 remain in transgenic
-chains), in which successive rounds of L chain editing have taken
place. This could result from the dominance of the anti-DNA H chain
that would retain DNA binding in many H/L combinations. Interestingly,
the group of low affinity IgMb-expressing anti-DNA
hybridomas, described in Table II, had an intermediate distribution of
J usage between wt and mutant hybridomas (Table III), suggesting a
partial failure of the deletion/editing mechanism in D42i mice.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Clonal deletion of anti-DNA B cells
Negative selection by clonal deletion is an effective means for preventing anti-DNA B cell autoimmunity. This mechanism has been previously demonstrated for B cells reactive with membrane-bound Ags (20, 68) as well as for DNA (26, 65). It is remarkable that, within the range of affinities dictated by the two anti-DNA H chain transgenes, there was little difference in the extent of deletion between glD42i and D42i B cells. These results suggest a low threshold affinity for anti-DNA B cell deletion and differ from those obtained with conventional anti-DNA transgenic mice (26, 65); however, they agree with the recent results of Lang et al. (28), in which immature B cells were shown to be exquisitely sensitive to central tolerance and receptor editing by low affinity, membrane-bound Ag. Indeed, the DNA Ag apparently belongs to the class of densely arranged and repetitive Ags (like those of viruses, bacteria, and parasites) that may interact directly with B cells to induce deletion (12, 65).
Contrary to our expectation, deletion of DNA-reactive B cells in the BM
was essentially complete for the lower affinity glD42i transgenic B
cells (Table II) but incomplete for the higher affinity D42i cells. In
the latter case, a significant fraction of the autoreactive B cell
population has found its way to peripheral lymphoid tissue (spleen) and
could be induced to secrete DNA-binding Abs, albeit with low affinity.
This binding may be explained in several different ways: i) the
transgenic B cells are negatively selected in the BM by an autoantigen
that is different from DNA (52), and, in this case, the somatically
mutated H chain may decrease rather than increase the affinity for the
deleting Ag; ii) the L chain repertoire may give rise to H/L receptor
combinations with noncontinuous distribution of DNA binding affinities.
Thus, it is conceivable that L chains producing intermediate affinity
with D42H and low affinity with glD42 H chain would be missing from the
repertoire. In that case, one would expect to find B cells expressing
low affinity Abs (and escaping deletion) in D42i but not in glD42i
mice, as is indeed observed. iii) More likely, clonal deletion is not a
one hit process but takes place in several intermediate steps (69),
possibly including successive rounds of secondary rearrangements
(receptor editing) that give rise to B cell receptors with reduced
affinity. When a multitude of high affinity autoreactive B cells is
present in the BM, as is the case in transgenic mice or in mice prone
to autoimmunity, deletion may become less efficient because of
insufficient time (before B cells exit the BM) or due to the scarcity
of Ag (as may be the case for memory B cell selection in germinal
centers (70)); consequently, low affinity, autoreactive B cells would
leak to peripheral organs. In this view, the anergic state in which
these cells are found in the spleen (Table II) would be regarded either
as a back-up mechanism of functional inactivation for B cells that fail
to undergo editing and deletion, or as an intermediate step in clonal
deletion that fails to be completed (19).
Editing of transgenic B cell receptors
H and L chain editing of autoreactive B cells (29, 30, 31, 32, 33) may be
accomplished in several ways, including i) the inactivation of the
expressed allele and subsequent rearrangement of the second allele, and
ii) replacement of rearranged variable gene segments in cis
with an upstream VH or Vk genes. We regularly observed that
a small fraction of the targeted B cells (Fig. 1 and Tables I and II)
had inactivated their transgenic IgMb allele and expressed
the endogenous IgMa allele. Sequence analysis of several
inactivated IgMb alleles revealed secondary rearrangements
of upstream D or VD segments into the leader intron of the targeted VDJ
gene (Fig. 6). We suspect that these secondary rearrangements may have
occurred spontaneously and may not be due to autoreactivity for the
following reasons: i) most of these rearrangements involved upstream D
segments that are not present in normal, rearranged H chain alleles,
and the nonphysiologic presence of D elements upstream of the VDJ
complex may result in varying degrees of locus instability depending on
the structure of the inserted VDJ gene segment (41); ii) most
rearrangements involved N-sequence additions, typically
introduced by TdT enzyme in pro B cells that are negative for surface
Ig, suggesting that they may have occurred early in B cell development
and irrespective of autoreactivity; and iii) similar inactivating
secondary rearrangements have been demonstrated by targeted H chains
with no apparent autoreactivity, such as T15 (36, 63), and many of
these rearrangements occurred in the early stages of B cell
development, before the expression of surface Ig (63). These arguments
notwithstanding, there remains a possibility that transgene
inactivation was induced or promoted by contact with the autoantigen.
The second route of receptor editing, via V-gene replacement did not
occur with the targeted VH11-encoded H chains, since no
example of V or VD recombination with the heptamer embedded at the 3'
end of most VH genes (33) was found, although both this
heptamer and a putative nonamer with the appropriate spacer length are
present in the VH11 gene. In contrast, however, L chain
editing seemed to represent a major component of B cell tolerance in
VH11-targeted mice (Table III), as judged by the dramatic
shift in J utilization. J usage appears to be a reliable
parameter of L chain editing for the following reasons: i) the
frequencies of individual J expression were found to be relatively
constant (34, 35); and, ii) unlike the biased distribution of Vk gene
segments in receptors of tolerant B cells (that largely depends on Ag
binding preferences), the choice of J is relatively independent of
DNA binding (40). Like clonal deletion, the mechanism of L chain
editing was found to be at least as efficient in glD42i as in D42i
transgenic mice. This finding is again in agreement with the recent
data of Lang et al. (28), using anti-class I transgenic mice; it
also supports the view that clonal deletion and receptor editing are
interrelated processes that occur in the immature B cell population
(26, 71) at roughly the same time and have similar requirements with
respect to Ag specificity and affinity. Indeed, an unknown fraction of
the deleted immature B cells may be accounted for by autoreactive cells
that failed the process of receptor editing, due to nonfunctional
secondary rearrangements.
The close relationship between B cell deletion and receptor editing would also predict that clonal deletion be severely hampered when receptor editing fails to reduce or eliminate autoreactivity by L chain secondary rearrangements. This would further suggest that, if targeted L chains capable of secondary rearrangements or H chain-only mice were to be employed in the anti-lysozyme experiments of Goodnow et al. (21), or in the anti-DNA experiments of Nguyen et al. (64), then more B cell deletion/editing and less B cell anergy would be observed.
Clonal anergy in anti-DNA D42i B cells
A significant fraction of D42i, but not glD42i, splenic
IgMb-expressing B cells could bind DNA with low affinity,
following LPS stimulation and fusion with myeloma cells. These B cells
were shown to be anergic by three independent criteria: their surface
IgM receptors were down modulated, they proliferated poorly and
secreted little Ab upon mitogenic stimulation in the presence of the
autoantigen, and they were poorly responsive to immunization with a
strongly immunogenic peptide/DNA complex in CFA. These low affinity B
cells could be regulated by a distinct, peripheral tolerance mechanism
that would render them anergic; alternatively, as discussed above, they
could be the fugitives of an incomplete process of clonal deletion
and/or receptor editing. Indeed, the J distribution of the anergic
cells (Table III) was suggestive of an intermediate stage of clonal
deletion/receptor editing.
The existence of intermediate stages in B cell clonal deletion that give rise to reversible maturation arrest and functional inactivation has been previously demonstrated by Hartley et al. (69) in anti-HEL transgenic mice. The arrested B cells had decreased surface IgM and were triggered to die after 1 to 3 days exposure to membrane-bound HEL. However, they could be rescued by removal of Ag and were able to secrete Ab upon reactivation with LPS, as was found to be the case with the anergic anti-DNA population in this study. A shortened life span of transgenic anti-HEL B cells was also found by Fulcher et al. (72), who concluded that anergy is a form of delayed deletion and that the distinction between B cell anergy and deletion may be more relative than absolute.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Dan Eilat, Division of Medicine, Hadassah University Hospital, P.O. Box 12000, Jerusalem 01120, Israel. E-mail address:
3 Abbreviations used in this paper: SLE, systemic lupus erythematosus; BM, bone marrow; wt, wild type; LH, long homology; SH, short homology; PE, phycoerythrin; HEL, hen egg lysozyme; H, heavy; L, light; ES, embryonic stem.
Received for publication February 6, 1998. Accepted for publication June 29, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß T cells. J. Immunol. 156:4041.[Abstract]
immunoglobulin light chain locus are used at markedly different frequencies. Proc. Natl. Acad. Sci. USA 81:4756. chain joining region J1 and J2 segments in mouse spleen DNA. Proc. Natl. Acad. Sci. USA 82:6399. locus contains about 140 variable gene segments. Eur. J. Immunol. 26:1613.[Medline]
This article has been cited by other articles:
|
|
 |
|
  L. Yunk, W. Meng, P. L. Cohen, R. A. Eisenberg, and E. T. Luning Prak Antibodies in a Heavy Chain Knock-In Mouse Exhibit Characteristics of Early Heavy Chain Rearrangement J. Immunol., July 1, 2009; 183(1): 452 - 461. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Makdasi, R. Fischel, I. Kat, and D. Eilat Autoreactive Anti-DNA Transgenic B Cells in Lupus-Prone New Zealand Black/New Zealand White Mice Show Near Perfect L Chain Allelic Exclusion J. Immunol., May 15, 2009; 182(10): 6143 - 6148. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kiefer, P. B. Nakajima, J. Oshinsky, S. H. Seeholzer, M. Radic, G. C. Bosma, and M. J. Bosma Antigen Receptor Editing in Anti-DNA Transitional B Cells Deficient for Surface IgM J. Immunol., May 1, 2008; 180(9): 6094 - 6106. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Sekiguchi, L. Yunk, D. Gary, D. Charan, B. Srivastava, D. Allman, M. G. Weigert, and E. T. L. Prak Development and Selection of Edited B Cells in B6.56R Mice. J. Immunol., June 1, 2006; 176(11): 6879 - 6887. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Grimaldi, V. Jeganathan, and B. Diamond Hormonal Regulation of B Cell Development: 17beta-Estradiol Impairs Negative Selection of High-Affinity DNA-Reactive B Cells at More Than One Developmental Checkpoint. J. Immunol., March 1, 2006; 176(5): 2703 - 2710. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Ait-Azzouzene, L. Verkoczy, B. Duong, P. Skog, A. L. Gavin, and D. Nemazee Split Tolerance in Peripheral B Cell Subsets in Mice Expressing a Low Level of Ig{kappa}-Reactive Ligand J. Immunol., January 15, 2006; 176(2): 939 - 948. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Liu and T. Manser Antinuclear Antigen B Cells That Down-Regulate Surface B Cell Receptor during Development to Mature, Follicular Phenotype Do Not Display Features of Anergy In Vitro J. Immunol., April 15, 2005; 174(8): 4505 - 4515. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Novobrantseva, S. Xu, J. E.-L. Tan, M. Maruyama, S. Schwers, R. Pelanda, and K.-P. Lam Stochastic pairing of Ig heavy and light chains frequently generates B cell antigen receptors that are subject to editing in vivo Int. Immunol., April 1, 2005; 17(4): 343 - 350. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Hayashi, T. Nojima, R. Goitsuka, and D. Kitamura Impaired Receptor Editing in the Primary B Cell Repertoire of BASH-Deficient Mice J. Immunol., November 15, 2004; 173(10): 5980 - 5988. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Edry and D. Melamed Receptor Editing in Positive and Negative Selection of B Lymphopoiesis J. Immunol., October 1, 2004; 173(7): 4265 - 4271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hangartner, B. M. Senn, B. Ledermann, U. Kalinke, P. Seiler, E. Bucher, R. M. Zellweger, K. Fink, B. Odermatt, K. Burki, et al. Antiviral immune responses in gene-targeted mice expressing the immunoglobulin heavy chain of virus-neutralizing antibodies PNAS, October 28, 2003; 100(22): 12883 - 12888. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T Winkler Antigenic targets-workshop report Lupus, December 1, 2002; 11(12): 780 - 782. [Abstract] [PDF]
|
|
|
|
|
|
C T Ravirajan and D A Isenberg Transgenic models of tolerance and autoimmunity: with special reference to systemic lupus erythematosus Lupus, December 1, 2002; 11(12): 843 - 849. [Abstract] [PDF]
|
|
|
|
|
|
J. Yan and M. J. Mamula B and T cell tolerance and autoimmunity in autoantibody transgenic mice Int. Immunol., August 1, 2002; 14(8): 963 - 971. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-P. Chu, D. Taylor, H.-G. Yan, B. Diamond, and L. Spatz Persistence of partially functional double-stranded (ds) DNA binding B cells in mice transgenic for the IgM heavy chain of an anti-dsDNA antibody Int. Immunol., January 1, 2002; 14(1): 45 - 54. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Li, L. Fernandez, K. C. O'Connor, T. Imanishi-Kari, and B. D. Stollar The Rearranged VH Domain of a Physiologically Selected Anti-Single-Stranded DNA Antibody as a Precursor for Formation of IgM and IgG Antibodies to Diverse Antigens J. Immunol., October 1, 2001; 167(7): 3746 - 3755. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Litzenburger, H. Bluthmann, P. Morales, D. Pham-Dinh, A. Dautigny, H. Wekerle, and A. Iglesias Development of Myelin Oligodendrocyte Glycoprotein Autoreactive Transgenic B Lymphocytes: Receptor Editing In Vivo After Encounter of a Self-Antigen Distinct from Myelin Oligodendrocyte Glycoprotein J. Immunol., November 1, 2000; 165(9): 5360 - 5366. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wang and M. J. Shlomchik Autoantigen-Specific B Cell Activation in FAS-Deficient Rheumatoid Factor Immunoglobulin Transgenic Mice J. Exp. Med., September 6, 1999; 190(5): 639 - 650. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K.-P. Lam and K. Rajewsky B Cell Antigen Receptor Specificity and Surface Density Together Determine B-1 versus B-2 Cell Development J. Exp. Med., August 16, 1999; 190(4): 471 - 478. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Mehr, M. Shannon, and S. Litwin Models for Antigen Receptor Gene Rearrangement. I. Biased Receptor Editing in B Cells: Implications for Allelic Exclusion J. Immunol., August 15, 1999; 163(4): 1793 - 1798. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Friedmann, N. Yachimovich, G. Mostoslavsky, Y. Pewzner-Jung, A. Ben-Yehuda, K. Rajewsky, and D. Eilat Production of High Affinity Autoantibodies in Autoimmune New Zealand Black/New Zealand White F1 Mice Targeted with an Anti-DNA Heavy Chain J. Immunol., April 15, 1999; 162(8): 4406 - 4416. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Seiler, B. M. Senn, M.-A. Brundler, R. M. Zinkernagel, H. Hengartner, and U. Kalinke In Vivo Selection of Neutralization-Resistant Virus Variants But No Evidence of B Cell Tolerance in Lymphocytic Choriomeningitis Virus Carrier Mice Expressing a Transgenic Virus-Neutralizing Antibody J. Immunol., April 15, 1999; 162(8): 4536 - 4541. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
