| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1,3Gal Epitopes in 1,3-Galactosyltransferase-Deficient Mice1
*
Transplantation Biology Research Center, Surgical Service, Massachusetts General Hospital/Harvard Medical School, Boston, MA 02129; and
BioTransplant, Inc., Charlestown, MA 02129
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-3Gal
1-4GlcNAc (Gal) epitopes
are a major barrier to xenotransplantation. Studies in this report,
which use combined multiparameter flow cytometric sorting and
enzyme-linked immunospot assay, demonstrate that anti-Gal
IgM-producing cells are found exclusively in a small B cell
subpopulation (i.e., CD21-/low IgMhigh
B220low CD5- Mac-1-
493- cells) in the spleens of
1,3-galactosyltransferase-deficient mice. All IgM-producing cells
were detected in a similar splenic subpopulation of
1,3-galactosyltransferase-deficient and wild-type mice. A higher
frequency of B cells with anti-Gal surface IgM receptors was
observed in the peritoneal cavity than in the spleen, but these did not
actively secrete Abs, and showed phenotypic properties of B-1b cells
(CD21-/low IgMhigh CD5-
CD43+ Mac-1+). However, these became
Mac-1- and developed anti-Gal Ab-producing activity
after in vitro culture with LPS. The splenic B cells with anti-Gal
receptors consisted of both Mac-1+ B-1b cells and
Mac-1- B-1b-like cells. The latter comprised most
anti-Gal IgM-producing cells. Our studies indicate that
anti-Gal natural IgM Abs are produced by a B1b-like,
Mac-1- splenic B cell population and not by plasma cells
or B-1a cells. They are consistent with a model whereby B-1b cells lose
Mac-1 expression upon Ag exposure and that these, rather than plasma
cells, become the major IgM Ab-producing cell
population. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1,3-galactosyltransferase
(GalT)4
gene and do not express Gal1-3Gal1-4GlcNAc (Gal)
carbohydrate residues. They produce abundant natural Abs (NAbs)
to the Gal epitope, which are polyclonal, highly specific for Gal, and
not polyreactive (1, 2). Anti-Gal constitutes as much as
1% of circulating total IgG, and 1–4% of total IgM in human sera
(2, 3, 4). Although the physiological significance of
anti-Gal is not entirely clear as yet, it has been implicated in
autoimmune disease and tumor immunity (5, 6, 7). In addition,
anti-Gal Abs are a major barrier to xenotransplantation of pig
organs into humans, because hyperacute rejection is initiated by their
binding to Gal determinants that are abundantly expressed on porcine
endothelial cell glycoproteins and glycolipids (8, 9, 10, 11).
Despite the significance of these Abs, the phenotype and other
properties of the B cell types responding to Gal and actively producing
anti-Gal Abs have not been defined. The repertoire of human NAbs may be driven by selection predominantly by self-Ags or by exposure to microorganisms, or may be the expression of a developmentally determined genetic program. The response of the B cell compartment to both self-Ags and microbial products is thought to derive preferentially from activation of CD5+ (B-1a) B cells, which are the predominant B cell population in the peritoneal cavities of mice (12). Anti-Gal Abs are presumed to be part of this larger pool of NAbs. The majority of human anti-Gal Abs are of the IgM and IgG2 classes, and lower levels of other isotypes also exist (3, 13, 14). Because NAbs against Gal are thought to develop as a result of exposure to environmental bacteria that express this carbohydrate determinant (3, 13), the B-1a lineage has been speculated to be the major population of anti-Gal NAb-producing cells (11). Another candidate for the production of anti-Gal Abs might be the splenic marginal zone (MZ) B cell. MZ B cells have been thought to represent a distinct lineage; they have some characteristics of memory cells, and they primarily recognize and respond to complement-coated polysaccharide Ags. The major population of MZ B cells are IgMhigh IgDlow CD23-/low CD21high large lymphocytes. Like B-1 cells, MZ B cells may be involved in responding primarily to T cell-independent (TI) type 2 Ags and include self-reactive B lymphocytes (15). Also, terminally differentiated plasma cells might be expected to be a major Ab-producing population.
In the present studies, we used combined multiparameter flow cytometric (FCM) sorting and enzyme-linked immunospot (ELISPOT) assay to define the phenotype of the cells actively producing anti-Gal NAbs and all IgM in GalT-deficient (GalT-/-) mice, in which Gal expression is completely eliminated, and naturally occurring anti-Gal Abs are present in sera, similar to humans (16, 17). In addition, by surface staining with fluorochrome-labeled Gal-BSA (synthetic Gal conjugated to BSA), we have identified B cells bearing surface IgM (sIgM) receptors that can recognize Gal epitopes. A distinct expansion of B cells with anti-Gal sIgM was observed in the spleen (SPL) and peritoneal cavity (PerC) in GalT-/- mice, but not in those in GalT+/+ mice after immunization with rabbit RBC that abundantly express Gal (1). Unexpectedly, the phenotype of B cells producing anti-Gal was distinct from plasma cells, Mac1+ B-1 cells, B-1a cells, and also from MZ B cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
GalT-/- mice were bred in our colony (16). Age-matched (10- to 12-wk-old) GalT-/- and wild-type GalT+/+ mice either on a hybrid background (129SV x DBA/2 x C57BL/6) or on the C57BL/6J background (backcrossed seven times to C57BL/6) were used for the experiments. All mice were maintained in a specific pathogen-free microisolator environment. To enhance anti-Gal Ab production, GalT-/- mice were immunized with Gal-bearing xenogeneic cells (i.e., 109 rabbit RBC) (Cocalico Biologicals, Reamstown, PA) 8 days before assay. Rabbit RBC were washed twice and resuspended at 109 cells/ml in PBS before injection.
ELISPOT assay for detecting anti-Gal and total IgM- producing cells
The assay was performed as described previously (18, 19). Briefly, nitrocellulose membranes of a 96-well filtration plate (Millipore, Bedford, MA) were coated with 5 µg/ml Gal-BSA (Alberta Research Council, Alberta, Canada) or with 5 µg/ml goat anti-mouse IgM (Southern Biotech, Birmingham, AL) for detecting anti-Gal or total IgM-producing cells, respectively. Nonspecific binding sites were blocked with 0.4% BSA in IMDM. Serial dilutions of cell suspension in IMDM supplemented with 0.4% BSA, 5 µg/ml insulin, 5 µg/ml transferrin, 5 ng/ml sodium selenite (all from Sigma, St. Louis, MO), 50 µM 2-mercapto-ethanol and 1 µg/ml gentamicin were added to wells in triplicate. After a 24-h culture at 37°C, bound Abs were detected using HRP-conjugated goat anti-mouse IgM Abs (Southern Biotech), followed by color development with 3-amino-9-ethyl carbazole (Sigma). After the membranes were dried, the red spots in each well were counted using a stereo microscope. Tiny spots without halo formation indicating cell-specific Ab output were excluded as false spots caused by cell debris or shed materials.
LPS stimulation
SPL, bone marrow (BM), and PerC cells obtained from either GalT-/- or GalT+/+ mice were cultured at 1.25–5 x 105 cells/ml with an equal amount of irradiated (3 Gy) syngeneic feeder SPL cells in RPMI 1640 supplemented with 10% FCS (Sigma), 50 µM 2-mercapto-ethanol, 1% HEPES buffer, 0.09 mM nonessential amino acid, 1 mM sodium pyruvate, 2 mM glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin, and 10 µg/ml LPS. After in vitro LPS stimulation for 1–7 days, cells were washed and resuspended in medium, then serial dilutions (four 5-fold dilutions, beginning with 1 x 105 cells/well) of each suspension were added to triplicate wells of ELISPOT plates for detecting anti-Gal IgM-producing cells.
FCM analysis and cell sorting
The following mAbs were used: FITC-conjugated anti-CD21
(7G6), anti-B220, PE-conjugated anti-CD5 (Ly-1), anti-CD19,
anti-CD23 (Fc
R), anti-CD43 (Ly-48, S7), anti-Mac-1
(CD11b), anti-CD138 (Syndecan-1), biotin-conjugated anti-mouse
IgM, unconjugated CD21 (all from PharMingen, San Diego, CA), and 493 in
concentrated culture supernatant (hybridoma kindly provided by Dr.
Antonius G. Rolink, Basel Institute for Immunology, Basel, Switzerland;
Ref. 20). Unconjugated mAbs were visualized with
PE-conjugated mouse anti-rat Ig
(PharMingen). To detect B cells
bearing receptors for Gal, FITC-conjugated Gal-BSA (Alberta Research
Council), or control FITC-conjugated BSA (Fisher Scientific, Pittsburgh
PA) were used. FITC-conjugation of Gal-BSA and BSA was performed with
Quick Tag FITC Conjugation Kit (Boehringer Mannheim, Indianapolis, IN)
according to the manufacturer’s instructions. Cells were double or
triple stained with appropriate combinations of FITC-, PE-, and
biotin-conjugated Abs. In some experiments, SPL cells were T
cell-depleted using anti-CD4 and anti-CD8 mAbs and rabbit
complement as described (21) before staining. For Gal-BSA
staining, 1,000,000 cells per 100 µl were incubated with 0.5 µg/100
µl FITC-Gal-BSA or 0.5 µg/100 µl control FITC-BSA in FCM
medium for 2 h at 4°C. The biotinylated mAb was visualized with
PE-streptavidin (for two-color FCM) or Cychrome-streptavidin (for
three-color FCM). Nonspecific Fc
R binding of labeled Abs was blocked
by 2.4G2 (rat anti-mouse FcR mAb). Dead cells were excluded from
the analysis by light-scatter and/or propidium iodide. All analyses
were performed on a FACScan cytometer (Becton Dickinson, Mountain
View, CA).
Based on criteria indicated in the individual figures, cells were sorted under sterile conditions using a MoFlo high-speed cell sorter (Cytomation, Fort Collins, CO). Sorted cells were reanalyzed for purity on a FACScan cytometer, and were immediately resuspended in culture medium and applied to ELISPOT plates precoated with Gal-BSA or anti-mouse IgM to determine the frequency of anti-Gal or total IgM-producing cells (combined FCM sorting and ELISPOT assay).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We first determined the anatomical location of the cells that
actively produce anti-Gal Abs. In
GalT-/- mice that were either untreated
or immunized by i.p. injection of 109
Gal-expressing rabbit RBC (8 days before analysis), the frequency of
anti-Gal IgM- and total IgM-producing cells was quantified by
ELISPOT assay of various tissues; i.e., the SPL, BM, PerC, PBMC,
cervical lymph nodes, and mesenteric lymph nodes. As is shown in Fig. 1
, anti-Gal IgM-producing cells were
localized mainly in the SPL and were undetectable in the PerC of both
untreated and immunized mice. A similar trend was observed for total
IgM-producing cells, regardless of specificity. Rabbit RBC immunization
markedly increased the frequency of anti-Gal IgM-producing cells in
all tested sites other than the PerC (Fig. 1A), but did not
alter that of total IgM-producing cells (Fig. 1B),
reflecting the predominance of the anti-Gal specificity in the
response to rabbit RBC. To assess the possible influence of the route
of immunization on the response of Gal-reactive B cells,
GalT-/- mice were immunized via various
routes; i.e., s.c., i.v., or i.p. inoculation. Despite a marked
increase in the serum levels of anti-Gal in
GalT-/- mice immunized by all three
routes (data not shown), anti-Gal Ab-producing cells were
undetectable in the PerC, whereas SPL contained high frequencies (Fig. 1C). Thus, anti-Gal IgM-producing cells are localized
mainly in the SPL, and these cells are absent in the PerC of
GalT-/- mice.
|
D). These results indicated that
PerC cells have the potential to become anti-Gal Ab secretors upon
in vitro stimulation. The low level detection of Anti-Gal Ab-producing
cells in GalT+/+ mice after LPS culture
probably reflects increased nonspecific background or polyreactive Ab
production in this assay, because similar levels of Ab-producing cells
were detected using control BSA-precoated plates (not shown). Anti-Gal IgM-producing cells are contained exclusively in a small CD21-/low IgMhigh B220low CD5- Mac-1- splenic B cell subpopulation
By combined multiparameter FCM sorting and ELISPOT assay, we
analyzed the cell surface phenotype of B cells making anti-Gal IgM
in the SPL of GalT-/- mice that were
either untreated or immunized with i.p. injection of
109 rabbit RBC (8 days before analyses). SPL
cells were stained with combinations of mAbs directed against IgM,
CD21, CD5, Mac-1, and B220. Ligation of those mAbs per se did not
affect the subsequently performed ELISPOT assay (data not shown). The
cells were sorted on the basis of criteria described in the individual
figures (Figs. 2-4), and the sorted cells
were subjected to ELISPOT assay for the frequency of anti-Gal IgM
and total IgM-producing cells in each cell subpopulation.
|
shows the results of an experiment in which SPL cells were
sorted on the basis of CD21 and IgM expression. Because MZ, follicular
(Fo), and newly formed (NF) B cells, which constitute anatomically
distinct B cell subsets located in the SPL, differ in their expression
of IgM and CD21 (22), the combination of these markers
could be used to separate these three subsets by cell sorting.
CD21high IgMhigh is the
phenotype of MZ B cells, which are uniquely positioned near efficient
Ag-trapping cells of the marginal sinus, where their association with
macrophages facilitates their exposure to blood-borne Ags (22, 23). CD21int IgMint
is the phenotype of Fo B cells. Fo B cells are positioned adjacent to T
cell areas, which may facilitate their response to T cell-dependent Ags
and promote germinal center formation (23).
CD21-/low IgMhigh is
the phenotype of NF B cells, which have recently immigrated from the BM
into the SPL. These B cells penetrate the marginal sinus, reaching the
outer zone of the periarteriolar lymphocytic sheath, and are still
immature at these sites (24). Because B-1 cells also share
similar phenotypic characteristics (25), the
CD21-/low IgMhigh
population consists of a mixture of NF B cells and B-1 cells. As is
shown in Fig. 2B, the CD21-/low
IgMhigh population (fraction b) was greatly
enriched for both anti-Gal and total IgM-producing cells, and the
CD21int-high IgMint-high
(fractions c-e) were markedly depleted for both types of Ab-producing
cells. A similar cell population comprised IgM-producing cells in
wild-type (GalT+/+) mice (data not shown).
The CD21- IgM- plasma
cell-containing cell population sorted from T cell-depleted SPL cells
was devoid of IgM-producing (Fig. 3),
including anti-Gal-producing (not shown) cells, although this
population was enriched for IgG-producing plasma cells (Fig. 3). Thus,
splenic plasma cells produce IgG but not IgM Abs.
|
A). To examine the role
of CD5+ cells in Ab production, we purified
CD21-/low
IgMhighCD5- cells by
three-color FCM sorting. Again, in both untreated and immunized
GalT-/- mice,
CD21-/low IgMhigh cells,
including both CD5+ and
CD5- subsets, were enriched for both
anti-Gal and total IgM-producing cells. However, further enrichment
for both types of Ab-producing cells was observed in the
CD21-/low IgMhigh
CD5- population, indicating that
CD5+ B-1a cells were not the predominant
IgM-producing population. In accordance with these
results, the frequency of anti-Gal IgM-producing
cells was markedly higher in the
CD5-B220low population
than among CD5+ B220low
(B-1a) cells or CD5-
B220high cells (Fig. 4B). The
CD21-/low IgMhigh B cells
expressed lower levels of B220 than Fo or MZ cells (not shown).
|
10–15% of CD21-/low
IgMhigh cells expressed Mac-1; data not shown),
it was possible that
Mac-1+CD5- B-1b cells were
the predominant IgM-producing cell population. To address this
possibility, we purified Mac-1+
B220low, Mac-1-
B220low and Mac-1-
B220high populations by FCM sorting, and the
resulting populations were then analyzed for anti-Gal and total IgM
production. In untreated and immunized
GalT-/- mice, the
Mac-1+ B220low and
Mac-1- B220low populations
were enriched for both anti-Gal and total IgM-producing cells, and
the Mac-1- B220high
population was markedly depleted for both types of Ab-producing cells
(Fig. 5). However, the frequency of
anti-Gal and total IgM-producing cells was markedly higher in the
Mac-1- B220low population
than in the Mac-1+ B220low
(B-1b) population. Thus, although the Mac-1+ B-1b
population might include some IgM-producing cells, IgM-producing cells
are found predominantly in the CD21-/low
IgMhigh B220low
CD5- Mac-1-
subpopulation, regardless of specificity. Similar results for total
IgM-producing cells were obtained in wild-type
(GalT+/+) mice (data not shown).
|
To directly identify B cells bearing sIgM that can bind to Gal,
SPL and PerC cells from GalT-/- mice that
were either untreated or immunized with rabbit RBC were stained with
FITC-labeled synthetic Gal-BSA together with biotin-labeled
anti-sIgM mAb plus PE-streptavidin, and subjected to FCM analysis.
Gal-BSA-binding B cells were detected in the SPL and PerC of
GalT-/- mice, and these populations were
significantly expanded by immunization with Gal-bearing xenogeneic
cells (rabbit RBC; Fig. 6, A
and B). In contrast, even after immunization, no Gal-binding
B cells were detected in the SPL, and only limited numbers of
low-level-Gal-binding B cells, which might be anergic autoreactive B
cells (26), were detected in the PerC of
GalT+/+ mice. The specificity of the
Gal-BSA ligand for the corresponding sIgM on B cells was demonstrated
by showing enrichment of anti-Gal Ab-producing cells among
Gal-binding B cells and a corresponding absence of anti-Gal
Ab-producing cells among non-Gal-binding B cells, using combined FCM
sorting and ELISPOT assay (Fig. 6, C–E), as we
have described previously (19). Previous incubation with
Gal-BSA did not alter the frequency of anti-Gal IgM-producing SPL
and PerC cells (data not shown), ruling out the possibility that
Gal-BSA ligation used for cell sorting might affect the subsequently
performed ELISPOT assay. Of note, the frequency of Gal-binding B cells
in the PerC was significantly higher than that in the SPL of
GalT-/- mice (Fig. 6B), but
anti-Gal Ab-producing cells were undetectable in the PerC (Fig. 1A), even among Gal-binding PerC B cells that were enriched
to 71% by FCM (Fig. 6D).
|
To further characterize the phenotype of B cells with anti-Gal
sIgM, we analyzed the various surface makers expressed on those B cells
by three-color FCM analysis. SPL and PerC B cells were prepared from
rabbit RBC-immunized GalT-/- mice, and
cells bearing sIgM that bound to Gal-BSA (Gal-binding B cells) were
selected by gating (Fig. 7A),
and were evaluated for their cell size and expression of different
surface markers. Both SPL and PerC Gal-binding B cells were
significantly larger in size than non-Gal-binding B cells (Fig. 7B), suggesting that they are in an activated state. As
shown in Fig. 7C, PerC Gal-specific B cells were
CD19+, CD21-/low,
CD23-/low, CD5-,
CD43+, Mac-1+,
CD138-, and 493-; this
phenotype is consistent with the properties of B-1b cells (Table I). The absence of 493 staining confirms
that both the SPL and PerC B cells that recognize Gal are not NF B
cells (20). Although most SPL Gal-specific B cells showed
a similar phenotype, they also included a significant
Mac-1- subpopulation, which comprised most
anti-Gal IgM-producing cells, as demonstrated above (Fig. 5). Thus,
B cells with anti-Gal sIgM are included in the B-1b subset of PerC
B cells, but most of the SPL B cells actively producing anti-Gal
IgM do not express Mac-1.
|
|
D. However, the sorted
PerC Gal-binding B cells (enriched to 77%) developed high levels of
anti-Gal Ab-producing activity after culture with LPS, whereas
non-Gal-binding B cells did not (Fig. 8B). Thus, the PerC
Gal-binding cells can become anti-Gal Ab secretors upon in vitro
stimulation. As expected, the majority of the initially sorted PerC
Gal-binding B cells were Mac-1+. However, this
population lost Mac-1 expression after LPS stimulation (Fig. 8A), when Ab production was detected.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The response of the B cell compartment to environmental Ags (including
TI type 2 Ags) and self-Ags has been thought to be derived
preferentially from activation of CD5+ (B-1a) B
cells, which are the predominant B cells in the mouse PerC (12, 31, 32). A recent report has clearly demonstrated that the
presence of self-Ag promotes PerC B-1a cell accumulation and
serum autoantibody secretion in a murine model system of naturally
generated autoreactive B cells with a germline-encoded
specificity for Thy-1 glycoprotein (33). In contrast, the
results presented herein show that B cells with anti-Gal receptors
have phenotypic properties of CD5- B-1b cells,
which are also primarily located in the PerC, suggesting that
Gal-reactive B cells may be phenotypically distinct and perhaps belong
to a different lineage from autoreactive B cells. Our results are
consistent with a previous human study, in which Gal-expressing porcine
thyroglobulin was used to detect human Gal-reactive B cells, and
CD5+ B cells were not enriched in the porcine
thyroglobulin-reactive B cell population (34). However, in
the present studies, the anti-Gal-specific B cells in the PerC,
which have the phenotype of B-1b cells, were not actively producing
Abs, whereas those detected in the SPL comprised the major anti-Gal
Ab-producing population. These splenic Ab-producing cells showed
similar phenotypic characteristics to B-1b cells, with the exception
that most did not express Mac-1. Because the PerC population developed
anti-Gal Ab-producing cells after in vitro culture with LPS, these
cells seemed to be arrested in their differentiation or anergic in
vivo. The question as to why the Gal-specific precursors did not
develop Ab-producing capacity in the PerC, but could be induced to do
so by in vitro mitogen stimulation, can be answered only speculatively
at the moment. Either the cells need a stimulus, which was provided
only in vitro, or they are inhibited from further differentiation or
from Ab production in the PerC by an unknown mechanism. Previous
reports have suggested that the expansion and activation of PerC B-1
cells may be induced by two steps in autoantibody (anti-RBC Ab)
transgenic mice; enteric bacteria increase the number of PerC B-1
cells, and pathogenic infection induces PerC B-1 cells to produce
anti-RBC Abs (35, 36). In those transgenic mice, oral
administration of LPS could induce the production of autoantibodies
from PerC B-1 cells, resulting in autoimmune symptoms
(35). Another possibility relates to the fact that
PGE2 produced by PerC macrophages reduces B-1 IgM
production without either killing B-1 cells or decreasing their
precursor frequencies (37). Macrophage-derived
PGE2 may provide a needed constraint on this
potentially hyperactive element of the immune response. In a recent
report using a cell transfer model, it was postulated that migration
from the PerC to the systemic circulation is necessary for B cells to
produce Ab. Evidence for this hypothesis was the presence of
transferred donor PerC-derived Ab-secreting cells in the SPL of host B
lymphocyte-defective X-chromosome-linked immune-defective mice
(38). This hypothesis is consistent with our data, and
would suggest a model wherein PerC B-1b cells expand locally and
migrate to the SPL, where they become Ab-producing cells. Supporting
this model, LPS-stimulated PerC Gal-binding B cells, which developed
Ab-producing capacity, phenotypically mimicked the SPL anti-Gal
Ab-producing cells (i.e., they lost Mac-1 expression (Fig. 7)). Several
groups have noted that B-1b cells are mostly
Mac-1+ in the PerC and mostly
Mac-1- in the SPL (39, 40). The
conversion of PerC B-1b cells from Mac-1+ to
Mac-1- upon LPS stimulation suggests a possible
relationship between PerC Mac-1+ and SPL
Mac-1- B-1b cells. An alternative explanation
for the lack of Ab production in the PerC is that PerC B cells with
anti-Gal receptors may be sequestered cells that are prohibited
from responding in vivo but can be differentiated by in vitro mitogen
stimulation. To distinguish these possibilities, additional studies
are needed.
The lineage hypothesis postulates distinct precursors for B-1 cells and conventional B (B-2) cells (41). B-1a cells are differentiated predominantly from precursors in fetal omentum and liver but rarely from B cell precursors in adult BM, which give rise to B-2 and B-1b cells (23, 42). In vivo cell-transfer studies have distinguished three murine B cell lineages: conventional B-2 cells, which develop late and are continuously replenished from progenitors in adult BM; B-1a cells, which develop early and maintain their number by self-replenishment; and B-1b cells, which share many of the properties of CD5+ B-1a cells, including self-replenishment and feedback regulation of development, but can also readily develop from progenitors in adult BM (42). We have also demonstrated that adult marrow poorly reconstitutes the B-1a cell lineage, but fully reconstitutes B-2 and B-1b lineages in severe combined immunodeficient mice (43). In this model, we demonstrated that marrow-derived B cells are efficient anti-pig NAb producers, suggesting that either a B-1b or a B-2 population might be the predominant cell population responsible for NAb production. Furthermore, similar findings were observed in GalT-/- mice that received myeloablative irradiation followed by BM transplantation from GalT-/- mice (i.e., full reconstitution of B-2 and B-1b cells, poor reconstitution of B-1a cells, and normal levels of anti-Gal IgM in their sera (Y.-G.Y. and M.S., unpublished data)). These are all consistent with our present data indicating that B-1b cells can produce anti-Gal IgM. We postulate that there might be a lineage relationship between the B-1b cells and the Mac-1- CD21-/low IgMhigh cell population in SPL that is most enriched for Ab-producing cells.
In contrast to the lineage hypothesis, the induced differentiation hypothesis proposes that both B-1 and B-2 cells are differentiated from a common precursor pool. The acquisition of the B-1a (CD5+) cell phenotype is due to the cross-linking of surface Ig receptors in the absence of cognate T cell help (44, 45). B-1a precursors may be positively selected by TI type 2 Ags and induced to express CD5. Several recent reports support this differentiation hypothesis (32, 46, 47). However, to our knowledge, the induction of the B-1b phenotype (Mac-1+, CD5-) through surface Ig receptor signaling has not been demonstrated.
Phenotypically, B-1a and B-1b cells are essentially identical, being distinguished only by the presence or absence of the CD5 marker. Functionally, no differences between the two populations have yet been clearly identified. However, differing activities of IL-5 and IL-9 on B-1a and B-1b cells have been recently reported. IL-5 transgenic mice have an expanded B-1a population that is associated with high levels of autoantibodies, whereas IL-9 transgenic mice show expansion of the B-1b population without stimulation of autoantibody production, although IgM-production is enhanced in both types of mice (48). In addition, using FCM sorting and single-cell PCR methodology, a recent report demonstrated a different pattern of VH family usage in B-1b cells compared with either B-1a or conventional B cells in mice (i.e., the VH1 (J558) and VH2 (Q52) families were underused and the VH10 (DNA4) and VH3 (3660) families over-represented among B-1b cells), suggesting differences in the repertoires between B-1a and B-1b populations (49). This report also demonstrated that B-1b cells used the VH4 (X24) family at a higher frequency (2.9%) than B-1a (1.8%) or conventional B cells (0%). Interestingly, another recent report using anti-Gal mAb-producing hybridomas derived from GalT-/- mice suggested that VH441, a member of the X24 family of Ig genes, is an important germline progenitor for encoding Ab-binding to the Gal epitopes (50). An unbiased analysis of VH-D-JH sequences from Gal-binding B cells purified by FCM sorting would further clarify this issue.
It is noteworthy that our studies indicate that all IgM-producing cells
express high levels of sIgM and are CD138 (Syndecan-1)-negative, as
shown by the combined FCM sorting and ELISPOT assays. These findings
conflict with the accepted belief that Abs are secreted mainly by
plasma cells, which have significantly down-modulated levels of sIgM,
B220, and other surface markers indicative of B cell maturation. The
CD21- IgM- cell
population sorted from T cell-depleted SPL cells, which should
include plasma cells, was depleted of IgM-producing cells, although
this population was enriched for IgG-producing cells (Fig. 3). It
seems unlikely that the CD21-/low
IgMhigh B220low
CD5- Mac-1- population
represents cells in the transition phase from B cells to plasma cells,
because IgM production is active only among
IgMhigh B cells.
The presence of sIgM, which can bind a synthetic form of Gal, on anti-Gal IgM-producing cells suggests that specific tolerance could be induced toward this epitope or specific depletion might be achieved by targeting these cells by cross-linking Gal epitopes to the corresponding receptors. Consistently, specific tolerance toward Gal can be induced by achieving GalT+/+ to GalT-/- BM mixed chimerism, even in the presence of anti-Gal-producing cells at the time of BM transplantation, suggesting that tolerization among those Ab-producing cells may occur by sIgM-cross-linking by Gal-positive donor cells (19). Preliminary data are consistent with the possibility that mixed chimerism leads to rapid deletion of splenic anti-Gal Ab-producing cells (2 wk after BM transplantation). Such an approach of targeting IgM-producing cells by sIgM cross-linking could also be applied to other Ags, because our data indicate that all IgM-producing cells express high levels of sIgM and show phenotypic properties identical with those of anti-Gal IgM-producing cells, regardless of their specificity. This possibility is also suggested by a previous finding that administration of anti-µ mAb in rats is able to deplete circulating IgM, including natural xenoreactive IgM (with specificity other than Gal; Ref. 51). Thus, our demonstration of a common phenotype for IgM-producing cells suggests a possible approach to using specific tolerogens to overcome Ab-mediated diseases, and to prevent humoral rejection after xenotransplantation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 H.O. was partially supported by Uehara Memorial Foundation.
3 Address correspondence and reprint requests to Dr. Megan Sykes, Bone Marrow Transplantation Section, Transplantation Biology Research Center, Massachusetts General Hospital, MGH East, Building 149-5102, 13th Street, Boston, MA 02129.
4 Abbreviations used in this paper: GalT, 1,3-galactosyltransferase; BM, bone marrow; ELISPOT, enzyme-linked immunospot; FCM, flow cytometry; Fo, follicular; Gal, Gal1-3Gal1-4GlcNAc; MZ, marginal zone; NAb, natural Ab; NF, newly formed; PerC, peritoneal cavity; sIgM, surface IgM; SPL, spleen; TI, T cell-independent.
Received for publication May 12, 2000. Accepted for publication August 21, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-galactosyl IgG. II. The specific recognition of (1,3)-linked galactose residues. J. Exp. Med. 162:573.1,3)galactose epitope. Transplantation 64:501.[Medline]
-galactosyl epitope on human normal and autoimmune thyroid cells. Autoimmunity 10:81.[Medline]
(1,3)Gal antibodies in humans switch immune responses from cellular to humoral. Nat. Med. 4:315.[Medline]
-galactosyl epitopes: a major obstacle for xenotransplantation in humans. Immunol. Today 14:480.[Medline]
1–3Gal epitopes. Proc. Natl. Acad. Sci. USA 90:11391.-galactosyl immunoglobulin G and bacteria of the human flora. Infect. Immun. 56:1730.1–3Gal antibodies after sensitization with xenogeneic tissues. J. Clin. Immunol. 19:116.[Medline]
1,3Gal epitopes implicated in sperm adhesion to the zona pellucida glycoprotein ZP3 are not required for fertilization in the mouse. J. Biol. Chem. 270:21437.1,3-Galactosyltransferase-deficient mice produce naturally occurring cytotoxic anti-Gal antibodies. Transplant. Proc. 28:556.[Medline]
1–3Gal natural antibody-forming B cells by induction of mixed chimerism. J. Exp. Med. 187:1335.1,3Gal-mediated graft rejection. J. Clin. Invest. 104:281.[Medline]
1,3galactosyltransferase. Xenotransplantation 5:191.[Medline]
-gal epitopes on xenografts and allografts: implications for accommodation in xenotransplantation. J. Clin. Invest. 105:301.[Medline]
1,3 pig terminal residues are not produced by CD5+ B-lymphocytes. Transplant. Proc. 28:545.[Medline]
(1, 3)Gal epitope in Gal knockout mice. The 5th Congress of the International Xenotransplantation Association, October 26. Nagoya, Japan. (Abstr.).
This article has been cited by other articles:
|
|
 |
|
  Y. S. Choi and N. Baumgarth Dual role for B-1a cells in immunity to influenza virus infection J. Exp. Med., December 22, 2008; 205(13): 3053 - 3064. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Benatuil, J. Kaye, N. Cretin, J. G. Godwin, A. Cariappa, S. Pillai, and J. Iacomini Ig Knock-In Mice Producing Anti-Carbohydrate Antibodies: Breakthrough of B Cells Producing Low Affinity Anti-Self Antibodies J. Immunol., March 15, 2008; 180(6): 3839 - 3848. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Irei, H. Ohdan, W. Zhou, K. Ishiyama, Y. Tanaka, K. Ide, and T. Asahara The persistent elimination of B cells responding to blood group A carbohydrates by synthetic group A carbohydrates and B-1 cell differentiation blockade: novel concept in preventing antibody-mediated rejection in ABO-incompatible transplantation Blood, December 15, 2007; 110(13): 4567 - 4575. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Li, Y. Yan, Y. Lin, D. M. Bullens, O. Rutgeerts, J. Goebels, C. Segers, L. Boon, A. Kasran, R. De Vos, et al. Rapidly induced, T-cell independent xenoantibody production is mediated by marginal zone B cells and requires help from NK cells Blood, December 1, 2007; 110(12): 3926 - 3935. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Fairfax, L. M. Corcoran, C. Pridans, N. D. Huntington, A. Kallies, S. L. Nutt, and D. M. Tarlinton Different Kinetics of Blimp-1 Induction in B Cell Subsets Revealed by Reporter Gene J. Immunol., April 1, 2007; 178(7): 4104 - 4111. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Shimizu, T. Kawahara, F. Haspot, P. D. Bardwell, M. C. Carroll, and M. Sykes B-cell extrinsic CR1/CR2 promotes natural antibody production and tolerance induction of anti-{alpha}GAL-producing B-1 cells Blood, February 15, 2007; 109(4): 1773 - 1781. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Savitsky and K. Calame B-1 B lymphocytes require Blimp-1 for immunoglobulin secretion J. Exp. Med., October 2, 2006; 203(10): 2305 - 2314. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Sprangers, M. Waer, and A. D. Billiau Xenograft rejection--all that glitters is not Gal Nephrol. Dial. Transplant., June 1, 2006; 21(6): 1486 - 1488. [Full Text] [PDF]
|
|
|
|
|
|
T. Moll, E. Martinez-Soria, M.-L. Santiago-Raber, H. Amano, M. Pihlgren-Bosch, D. Marinkovic, and S. Izui Differential Activation of Anti-Erythrocyte and Anti-DNA Autoreactive B Lymphocytes by the Yaa Mutation J. Immunol., January 15, 2005; 174(2): 702 - 709. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Kawahara, H. Ohdan, G. Zhao, Y.-G. Yang, and M. Sykes Peritoneal Cavity B Cells Are Precursors of Splenic IgM Natural Antibody-Producing Cells J. Immunol., November 15, 2003; 171(10): 5406 - 5414. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Yin, L. Ma, A. Varghese, J. Shen, and A. S.-F. Chong Intact Active Bone Transplantation Synergizes with Anti-CD40 Ligand Therapy to Induce B Cell Tolerance J. Immunol., May 15, 2002; 168(10): 5352 - 5358. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. L. Rothstein Two B-1 or Not To Be One J. Immunol., May 1, 2002; 168(9): 4257 - 4261. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
