| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Laboratory of Immune Regulations and Development, Department of Developmental Biology, and
Flow Cytometry Unit, J. Monod Institute, Centre National de la Recherche Scientifique and Universities Paris 6 and 7, Paris, France;
Sandoz Pharma, Basel, Switzerland; and
§
Division of Basic Sciences, National Jewish Center for Immunology and Respiratory Medicine, Denver, CO 80206
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
80% of Id-bearing B cells, starts at midpregnancy, and is
only observed until term. Such a specific maternal B cell deletion
could contribute to the success of the fetal
allograft. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is now well-documented in mice that the fetal expression of polymorphic MHC class I Ags starts at around day 9 postcoitum and slowly increases during the second half of gestation, never exceeding one-fiftieth of the adult level in any given tissue (10, 11). This low level of expression does not prevent fetal tissues from being able to trigger a cytotoxic response (12). It is also well-established that the fetus can be recognized as foreign by the immune system of the mother. In addition, there are numerous reports of the passage or effraction of fetal cells into the blood or lymphoid organs of the mother, although with variable frequency (13, 14). A reverse passage has also been reported (15).
We have been interested in the consequences of a semiallogeneic gestation on the specific immune repertoire and function of the mother. Such a question could only be addressed in a transgenic (Tg)3 animal in which a major proportion of mature lymphocytes bear a unique specificity, since the expression of Tg AgR molecules shuts off endogenous receptor rearrangements. In non-Tg animals, the proportion of specific T or B lymphocytes is too small to be followed.
We have used a previously described Tg B cell mouse model (16, 17) in which fully rearranged Ig heavy and light chain genes have been introduced into a mouse genome. The Tg Igs (IgM and IgD) are expressed on the B cell surface and present a relatively high affinity for H-2Kk molecules, a lower affinity for H-2Kb molecules, and a very low affinity for H-2Dk molecules (18, 19). They bear the 3.83 Id that is specifically bound by the 54.1 mAb. Such mice have already been used successfully to analyze the mechanisms of tolerance to self Ags at the B cell level (16, 17, 18, 19, 20). We have crossed age-matched 3.83 (anti-H-2Kk) Tg females (on the B10.D2 background, H-2d) with B10 MHC-congenic males of H-2d, H-2k, or H-2f haplotypes. Thus, we were able to analyze the fate of 3.83-positive B lymphocytes during syngeneic (H-2d) or semiallogeneic gestations. In the latter case, two different situations were compared: Semiallogeneic fetuses bearing MHC class I H-2K molecules were either recognized (H-2Kk) or not recognized (H-2Kf as a "third-party" control) by the Tg B cell receptors (BCRs). We show here for the first time that the presence of semiallogeneic fetuses recognized by the Tg BCR significantly alters the fate of specific B lymphocytes (compared with syngeneic or third-party control fetuses); this effect is observed primarily in the spleens of pregnant females starting at midgestation. An average of 70% of Tg B cells are deleted from the spleens of H-2Kk-fetus-bearing females starting at around 10 days of gestation. Such effects are fully reversed as early as 4 days postpartum, even in lactating females in which the hormonal down-regulation of B lymphopoiesis is maintained (21, 22).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3.83 anti-H-2Kk-µ
Tg mice have been
described previously (16). B10.BR (H-2k), B10.D2
(H-2d), and B10.M (H-2f) congenic mice were
obtained from Harlan (Gannat, France). All animal care and handling was
performed according to institutional guidelines.
Cell suspensions
Nucleated cell suspensions were prepared from the blood, spleens, paraaortic lymph nodes (LNs), and bone marrow (BM) of 3.83 Tg pregnant or nonpregnant females. Heparinized blood was obtained from anesthetized animals by intracardiac puncture. BM samples were flushed from the long leg bones (femurs and tibias) with HEPES-buffered 199 tissue-culture medium (Life Technologies, Cergy Pontoise, France). E were eliminated from the blood and spleen cell suspensions by osmotic shock in lysing buffer (0.17 M Tris and 8.3 g/L ammonium chloride, mixed 1 volume/9 volumes, pH 7.3).
Immunofluorescence staining and flow cytometry analyses
The incubation of 4 x 105 nucleated cells per
sample was performed with the following mAbs: 54.1 rat mAb specific for
the 3.83 BCR Id was revealed by an FITC-labeled mouse anti-rat 
light chain Ab (PharMingen, San Diego, CA), biotin-coupled
anti-B220 Abs (PharMingen) (in the presence of a 10-fold excess of
normal mouse serum Ig for single labelings) were followed by
phycoerythrin-labeled streptavidin (Southern Biotechnology Associates,
Birmingham, AL), and goat Abs to mouse µ-chain were coupled to FITC
(PharMingen). Flow cytometry profiles were obtained on an Epics
Elite-ESP flow cytometer that was equipped with a 488-nm argon laser
(Coultronics, Margency, France). The cell populations analyzed were
gated on the basis of forward and side (90° angle) scatter criteria
to avoid potential contamination by dead cells or debris.
ELISA
Microtiter plates (Luxlon, CML, Nemours, France) were coated
with 54.1 purified rat mAb (0.25 µg per well). Duplicate serum
dilutions were added to each well, and bound Abs were revealed by
peroxidase-coupled rat anti-mouse light chain mAbs (Serotec,
Oxford, U.K.). The enzymatic activity was measured after the addition
of the substrate O-phenylenediamine in 0.1 M citrate buffer
containing 0.03% H2O2 at 490 nm using a
microplate reader (MR 5000, Dynatech, Saint-Cloud, France). One single
pool of CosLin D1 supernatant in serial twofold dilutions was used as a
positive control to obtain a standard reference curve in every assay.
This supernatant originated from transfected cells secreting the
3.83 IgM only (16, 17).
Statistical analyses
Mean values were compared using the Student’s t test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A shows spleen B cell
populations in pregnant (day 14.5 of gestation) mice. A significant
decrease in the percentage of 3.83-positive cells (80% in this
experiment) could be seen reproducibly in the spleens of Tg females
that had been mated with B10.BR males. The same observations could be
made when we labeled the cells with anti-IgM Abs. Pregnant females
that had been mated with B10.D2 or B10.M males did not show this
effect, and their B cell percentages were comparable with those of
nonpregnant mice. To see whether this lack of 3.83-positive and
IgM+ cells was due to a modulation of their
receptors or to the disappearance of the cells, we labeled the cells
with the B220 cell marker. It was determined that the percentage of
cells bearing the pan B cell marker B220 was also diminished in a
comparable proportion (Fig. 1A). These results indicated
that a significant proportion of anti-H-2Kk Tg B cells
had disappeared from the spleen in B10.BR-mated pregnant females only.
In contrast, when the paraaortic LNs removed from the same animals were
subjected to the same analyses, no significant difference could be
observed between control and experimental Tg animals (Fig. 1B). The same results were obtained on other peripheral LNs
from the same animals.
|
, only B10.BR-mated females (seven
animals) undergo a significant loss of Tg B cells in their spleens
(70%, p 0.001) (Fig. 2B) and BM (66%,
p 0.001) (Fig. 2C); no loss was observed
in the LNs of these mice (Fig. 2A). This B cell deficit
remains at the same amplitude until day 18.5, as can be inferred from
the small SD values. In contrast, the average percentages of
CD4+ or CD8+ T cells are not significantly
affected in the same spleens (Fig. 2E). Moreover, the
subpopulation of non-Tg B lymphocytes bearing an endogenously
rearranged, Id-negative BCR is not different in the three groups (Fig. 2D). The ratio of IgM to B220 labeling remained close to 1
at all time points, indicating that 
70% of the Tg receptor-bearing
B cells had disappeared from the spleen and BM between days 12.5 and
18.5 of gestation. This Ag-specific deletion was also observed in blood
leukocytes from the same animals (data not shown). The phenomenon is
reversible, since splenic B cell populations are not different from
controls at 4 days postpartum, even in lactating females (data not
shown).
|
, showing a
typical result on day 18.5 of gestation). Again, no such differences
could be detected in the LNs of the same animals.
|
30% of B10.BR pregnant females
(day 10.5). Interestingly, the percentages of 3.83, IgM, and
B220+ cells were diminished in the BM only in
another 30% of animals (Fig. 4,
C vs D); percentages in the spleen did not differ
significantly between control and experimental crosses (Fig. 4, A and B). In these B10.BR-mated females, the
fluorescence labeling intensity of the 3.83-positive or
IgM+ spleen cells was shifted toward lower levels compared
with the control pregnant animals analyzed in the same experiment,
suggesting that the cells had begun to encounter Ag (Fig. 4, A vs B). The remaining one-third of animals did
not differ from controls at this stage.
|
).
|
, lower right plot). In
contrast, B220high cells appeared to be only marginally
affected. Moreover, an important decrease in 3.83 Id-positive B cells
was observed in B10.BR-mated pregnant females only (80% on average,
p 0.001) (cf Fig. 4
D).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A report using Tg mice for a TCR anti-H-2Kb has shown that maternal T lymphocytes do recognize paternal alloantigens during pregnancy (23). This "awareness" leads to a TCR down-modulation on H-2Kb-reactive T cells around midgestation (days 9–11 postcoitum) in the spleens of mice with H-2Kb-bearing fetuses. Moreover, such mice do not reject H-2Kb-bearing P815 tumor cell grafts. These observations suggest that a transient state of maternal tolerance to paternal alloantigens is established during the second half of gestation.
We have studied Tg mice bearing a major population of B lymphocytes with IgM and IgD receptors of a relatively high affinity for H-2Kk molecules (19). Our results are in agreement with the data obtained on T cells (23). We also observed a specific effect on maternal splenic lymphocytes that started around midgestation and was reversible after delivery. However, the T cell phenomenon was reportedly at its peak between days 9 and 11 of gestation and was limited to an AgR down-modulation. The important but incomplete B cell deletion we have observed is seen reproducibly from day 10.5 of gestation in the BM and from day 12.5 in the spleen and blood until day 18.5. It is dependent upon the presence of alloantigen-bearing fetuses in utero, since it is not seen after delivery. To our knowledge, this is the first report on such a phenomenon. In addition, our results suggest that this phenomenon starts at midgestation with a down-modulation of Id-bearing IgRs on the B cell surface. This possibility was observed in the B10.BR-mated females, in which no deletion of spleen Tg B cells had occurred at that time. When total cell numbers instead of frequencies were compared among the different groups, the same basic conclusions could be reached. However, cell frequency data are more accurate for two reasons: 1) they are obtained from more homogeneous cell samples that are gated on lymphoid cells by scatter criteria on the flow cytometer, and 2) although total spleen and LN cell numbers are increased routinely in the pregnant females of all three groups compared with nonpregnant females, the absolute cell numbers present rather substantial variations from one animal to the next, even within the same group. These variations are probably due to the increase in erythropoietic activity in the spleens of pregnant mice. Fowler and Nash have reported that spleen weight and cellularity are increased in parallel to erythropoiesis during gestation in mice, with a peak on day 12 and a positive correlation with litter size (24).
It is interesting to note the similarity in the timing of the onset of
B cell deletion and of the first appearance of MHC class I mRNA and
surface molecules in mouse embryos; both of these events occur around
day 9.5 of gestation (10, 11, 25, 26). The receptor modulation is
observed first and is replaced shortly thereafter by the deletion of
70% of BM and splenic B cells until term. This partial effect is
different from previously described complete (and thus irreversible) Tg
B cell eliminations in an H-2k or an H-2b mouse
(16, 17). It is probably due to the low antigenic load coming from the
semiallogeneic fetuses vs the large numbers of specific B lymphocytes
present in such Tg mice. It is noteworthy that a partial deletion
(although restricted to peripheral lymphoid organs) was also observed
in the same Tg mouse model in the case of a peripheral expression of
low levels of H-2Kb Ag in the liver (18).
The mechanism(s) leading to such a cell disappearance from the spleen
are presently unknown. Trivial side effects have been ruled out by the
B10.M crosses, which allowed us to study 3.83 Tg pregnant mice bearing
semiallogeneic fetuses that were not recognized by the Tg BCR. We have
also confirmed that the B10.BR cross by itself did not have any
particular influence on B cell populations in pregnant mice in two
ways: 1) non-Tg B10.D2 pregnant mice that were crossed with B10.M or
B10.BR males did not have significantly different B cell populations in
their spleens (data not shown), and 2) as shown in Figure 2, E and D, the T cells as well as non-Tg B
lymphocytes present in the same spleens as the
anti-H-2Kk-deleting B cells were not affected. Thus, it
appears that this deleting effect is restricted to Id-positive B cells
and is most likely Ag-specific. We have verified that these cells are
not drawn to or sequestered in other lymphoid organs such as the
peripheral LNs (even the paraaortic LNs draining the uterus), blood, or
BM. The potential homing of these cells to the placenta appears
unlikely: such a possibility has not been reported in the literature,
including reports on TCR anti-H-2Kb Tg pregnant mice
(23). In humans, it has been shown that the leukocytes that are found
in decidual cell suspensions are mostly NK cells, with some macrophages
and T cells but virtually no B cells (27). In mice and rats,
macrophages represent the vast majority of BM-derived cells in the
decidua (28).
It has been shown that soluble Ag such as H-2 (16) or hen egg-white lysozyme (29, 30) are unable to cause cell deletions, except on Ag-activated, high affinity, germinal center B cells (31, 32). As the deletion can be seen in primiparous naive females as soon as the developing fetuses begin expressing H-2Kk Ags, it is assumed, although not demonstrated, that alloantigen-bearing fetal cells are blood-borne to the lymphoid organs, where they encounter specific B cells. The paraaortic LN B cells draining the uterus do not share the same fate, possibly because they are in a different state of differentiation, since they contain very few cells of the IgMhigh, B220low phenotype. An alternative explanation might be that these LN cells do not encounter the fetal alloantigen-bearing cells. These as yet unidentified fetal cells are most likely blood-borne from the placenta and reach the spleen and BM first. BM B cells are known to be highly sensitive to tolerization, which is probably why they are affected first and are already deleting on day 10.5 of gestation. Surprisingly, fetal cells do not appear to be prevented from reaching those sites by the anti-H-2Kk IgM Abs present in plasma. Presumably, these circulating Abs are not of a sufficiently high avidity for the low Ag-expressing cells. They remain as IgMs, since the Tg Ig constructs do not allow a class switch (18, 20). As IgMs do not cross the placental barrier, they remain harmless to the fetuses.
It is well-known that multiparous females frequently display paternal MHC-specific Abs in their serum, apparently without harm to their progeny. Along the same line, earlier work by Mitchison (33) has shown that one can hyperimmunize female mice against paternal Ags without affecting the size or health of subsequent litters. This does not contradict the results that we obtained using naive primiparous females. Moreover, the deletion we observe does not affect all B cells, at least in our present model, and does not occur postpartum. Thus, immunization is most likely to occur at the time of delivery, when fetal cells might be able to enter the maternal circulation in greater numbers.
Interestingly, a recent study by Bonney and Matzinger (14) has reported the presence of fetal cells in the mother (as detected by quantitative PCR) as a relatively rare event. However, fetal cells are responsible for the priming of an anti-HY immune reaction in about one-third of all previously pregnant mice. In our experiments, we observed a specific B cell deletion in all mice studied. The apparent discrepancy between these observations could be due to different mouse strain combinations or to different housing conditions. It could be also the result of a higher level of sensitivity of the immune system of the mother compared with the PCR assay with regard to the detection of fetal cells. We would like to propose an alternate hypothesis that mouse fetal RBCs bearing MHC class I molecules (34, 35) could be responsible, at least in part, for the deletion we have observed. The embryo E are anucleated from day 12.5 of development and would not be detected in the PCR assay.
The results of the ELISAs showed no significant differences between 3.83-positive IgM levels in the various pregnant female groups. This finding could be the result of the absence of a transplacental passage of IgM. The large enough pool of remaining B cells in the peripheral LNs could also be a source of secreted IgM. A further understanding of this observation will require an investigation of the deletional mechanism(s) as well as of the state of activation or responsiveness of the remaining B cells.
What happens in the BM is particularly interesting, since it is the site of B cell development; it has been shown that B lymphopoiesis is down-regulated by gestational estrogenic hormones. As reported by Medina et al. (21, 22), the high estrogen levels that are attained during gestation lead to a remarkable decrease of B lymphopoiesis in the BM, particularly at the small pre-B cell stage. In contrast, the overall hormonal effect on splenic B cell populations is reportedly marginal. Despite the very low B cell numbers left in the BM, we were able to observe a nearly complete deletion of B220low, sIgM+-specific Tg B cells; however, this deletion was found only in mothers bearing H-2Kk fetuses. Thus, the phenomenon we report in the present work appears to be distinct from the one described by Medina et al. for the following reasons: 1) it is restricted to pregnant females bearing H-2Kd/k fetuses, 2) it affects spleen and BM cells to the same extent, and 3) the two phenomena have different kinetics; the specific cell deletion starts around day 10 and stops at parturition. However, both phenomena are superimposed during the second half of gestation.
Our data are consistent with the interpretation that the B cells that are eliminated in the maternal tolerance process are recent BM emigrants; such B cells have an IgMhigh phenotype, are present in the spleen but not in the LNs, and are known to turn over rapidly (36, 37). In this study, the elimination of these cells may be related to their enhanced tolerance sensitivity or to their greater accessibility to fetal Ag. Preliminary evidence from RAG-/-, 3.83 Tg mice suggests that other lymphocytes are not required for the specific maternal B cell deletion (our unpublished observations). Both B cell deletion and TCR down-modulation (23) are potentially important mechanisms involved, among many others, in the success of the fetal allograft. Given the importance of this biologic phenomenon for species survival, it is extremely likely that multiple, potentially redundant mechanisms are at work to ensure the success of viviparity. A refinement of the experimental model would require both the presence of fewer Tg B cells in the lymphoid organs and also their capacity to class-switch. These adjustments would bring the model even closer to the physiologic situation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Colette Kanellopoulos-Langevin, Institut Jacques Monod, Lab. des Regulations Immunitaires et Dev. Tour 43. 2 Place Jussieu, 75251 Paris Cedex, France. E-mail address:
3 Abbreviations used in this paper: Tg, transgenic; LN, lymph node; BM, bone marrow; sIg, surface Ig; BCR, B cell receptor.
Received for publication January 15, 1998. Accepted for publication May 4, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. M. Caucheteux, C. Vernochet, J. Wantyghem, M.-C. Gendron, and C. Kanellopoulos-Langevin Tolerance induction to self-MHC antigens in fetal and neonatal mouse B cells Int. Immunol., January 1, 2008; 20(1): 11 - 20. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Li, L. Ma, J. Shen, and A. S. Chong Peripheral deletion of mature alloreactive B cells induced by costimulation blockade PNAS, July 17, 2007; 104(29): 12093 - 12098. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Caucheteux, M.-C. Gendron, and C. Kanellopoulos-Langevin Pregnancy-induced alterations of B cell maturation and survival are differentially affected by Fas and Bcl-2, independently of BcR expression Int. Immunol., January 1, 2005; 17(1): 55 - 63. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Wolfgang, S. G. Eisele, M. A. Browne, M. L. Schotzko, M. A. Garthwaite, M. Durning, A. Ramezani, R. G. Hawley, J. A. Thomson, and T. G. Golos Rhesus monkey placental transgene expression after lentiviral gene transfer into preimplantation embryos PNAS, September 11, 2001; 98(19): 10728 - 10732. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Ait-Azzouzene, S. Caucheteux, F. Tchang, J. Wantyghem, R. Moutier, A. Langkopf, M.-C. Gendron, and C. Kanellopoulos-Langevin Transgenic Major Histocompatibility Complex Class I Antigen Expressed in Mouse Trophoblast Affects Maternal Immature B Cells Biol Reprod, August 1, 2001; 65(2): 337 - 344. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.P. Adams and D.F. Antczak Ectopic Transplantation of Equine Invasive Trophoblast Biol Reprod, March 1, 2001; 64(3): 753 - 763. [Abstract] [Full Text]
|
|
|
|
 |
|
 J. M. Baker, A. I. Bamford, and D. F. Antczak Modulation of Allospecific CTL Responses During Pregnancy in Equids: An Immunological Barrier to Interspecies Matings? J. Immunol., April 15, 1999; 162(8): 4496 - 4501. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
104 cells were analyzed in each
case): upper rows represent staining with 54.1 rat mAbs
specific for the 3.83 BCR Id; lower rows indicate
staining with anti-B220 Abs. Background staining was <1% in each
case. The total cell numbers in the spleen or LNs were expanded 1.5 to
2.5 times in all pregnant mice compared with nonpregnant
controls.
