SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

The fetal–placental unit is a semi-allograft and immunological recognition of pregnancy, together with the subsequent response of the maternal immune system, is necessary for a successful pregnancy. Dendritic cells (DC) show a biological plasticity that confers them special characteristics regulating both immunity and tolerance. Therapy employing DC proved to diminish the abortion in the DBA/2J-mated CBA/J females; however, the underlying mechanisms remain unknown. Here, we evaluated whether DC therapy influences the presence of immunoregulatory populations of cells at the fetal–maternal interface. To address this hypothesis, we analysed the pregnancy-protective CD8, γδ cell populations as well as transforming growth factor (TGF)-β1 and progesterone-induced blocking factor (PIBF) expression at the fetal–maternal interface from abortion-prone female mice that had previously received adoptive transfer of syngeneic DC. Syngeneic DC therapy induced an increase in the number of CD8 and γδ cells. Additionally, an upregulation of TGF-β1 and PIBF expression could be detected after DC transfer. We suggest that DC therapy differentially upregulates a regulatory/protective population of cells at the fetal–maternal interface. It is reasonable to assure that this mechanism would be responsible for the lower abortion rate.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

Dendritic cells (DC) are the most potent antigen-presenting cells of the immune system and are critically involved in the initiation of primary immune responses, graft rejection, autoimmune diseases and fetal rejection [1–5]. DC can also modulate the nature of the immune response in stimulatory or tolerogenic fashion [6]. DC pulsed in vitro with antigens have been successfully used to induce protective responses against tumour, experimental allergic encephalomyelitis and abortion [2, 5, 7–9]. Indeed, tumour peptide-pulsed mature DC can be used to elicit antigen-specific protective anti-tumour immunity in mice [10]. However, when immature DC are generated from murine bone marrow (BM) and applied for antigen presentation to allogeneic T cells, they induce T-cell anergy [11, 12]. These findings have important implications for the use of DC in immunotherapy in different types of diseases.

The establishment of pregnancy involves maternal recognition of pregnancy and implantation of the blastocyst into the maternal uterus. The phrase maternal recognition of pregnancy was coined by Roger Short in 1969 and can be defined as the physiological process whereby the conceptus signals its presence to the maternal system and prolongs lifespan of the corpus luteum. In most mammals, local mechanisms are triggered to protect the fetus [i.e. T helper (Th)2–3 cytokines, progesterone, progesterone-induced blocking factor (PIBF), indoleamine 2,3-dioxygenase, regulatory T cells, asymmetric antibodies (AAb) and complement regulatory protein] [13–19]. However, the entire range of molecules and mechanisms involved in pregnancy maintenance is not understood. It has been estimated that 50–70% of all conceptions fail and that current pregnancy loss affects 10–15% of couples [20].

Recently, we successfully induced tolerance to fetal antigens using a mouse model for spontaneous abortion by BM-derived DC (BMDC) [5]. In the present study, we investigated the cellular and molecular mechanisms involved in DC-associated pregnancy protection.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

Animals.  Female 6- to 8-week-old CBA/J and male DBA/2J mice were purchased from The Roffo Hospital Laboratory (Buenos Aires, Argentina). Animals were maintained under the guidelines stipulated by the University of Buenos Aires Institutional Animal Care and Use Committee.

To investigate the effect of syngeneic DC adoptive transfer on the abortion rate in DBA/2J-mated CBA/J females, the mice were divided into three groups, one group served as a control, having received no treatment. Another group was inoculated i.v. 100 μl DC-conditioned medium (DCCM). The third group received adoptive transfer of 4 × 106–7 × 106 syngeneic DC. The females were checked for vaginal plugs every morning. The day of vaginal plug detection was considered as day 0.5 of gestation (Gd). The pregnant females were sacrificed on Gd 13.5. Uteri-placental tissues were removed and separately placed in ethanol at 4 °C to be paraffin embedded using a standard protocol. The histological characteristics of uteri-placental tissue samples were investigated on haematoxylin–eosin stained sections.

Syngeneic DC generation.  To obtain BMDC, femora and tibiae were removed from CBA/J female (4- to 6-week-old) mice and mechanically isolated from surrounding tissues. Before removing epiphyses, BM cells were flushed using Iscove-modified Dulbecco medium (IMDM) and cultured in IMDM supplemented with PGE2 (10−8 mol/l) [21] and 30% J558-GM-CSF-conditioned medium [5, 22] at 2 × 106cells/ml in 100-mm Petri dishes. On day 5 of culture, DC were isolated by transferring the non-adherent and loosely adherent cells to new culture plates (leaving behind the adherent macrophages), incubating the plates at 37 °C for at least 2 h, and then repeating the procedure to remove any contaminating macrophages. In all cases, cell concentration and viability were assessed using haemocytometer and trypan blue dye exclusion test respectively. After that, DC were further enriched (>95% purity) by density-gradient centrifugation with dense BSA solution at 1.080 g/ml [23, 24] and used for adoptive transfer.

Abortion rate.  Mice were killed by neck dislocation on day 13.5, the uteri were removed and the total number of implantations and resorption sites (signs of abortion) was recorded. The abortion rate was calculated as percentages of resorption sites calculated from total number of implantations.

Immunohistochemical staining.  The uteri-placental tissues were fixed in 4% paraformaldehyde solution and embedded in paraffin. Tissue sections of 6-μm thick were mounted on polylysine-coated slides, deparaffinized, rehydrated and then heated in 10 mm citrate buffer (pH 6) containing triton X-100 (Sigma-Aldrich, Munich, Germany) 0.1% (v/v). After two washes with phosphate-buffered saline (PBS), slides were then incubated with 0.3% hydrogen peroxide in methanol for 30 min to quench endogenous peroxidase activity. After washing with PBS, tissues were incubated with blocking serum (Vectastain ABC Kit, Wertheim-Bettingen, Germany) at room temperature for 1 h. Then, a primary antibody diluted in blocking serum [CD8, TCR-γδ, transforming growth factor (TGF)-β1 or PIBF 1:50 dilution] was added to the slides and incubated at 4 °C overnight in a humidified chamber. Additional washes were performed and tissue sections were incubated for 30 min with 3 μg/ml biotinylated antibody (anti-rabbit or anti-mouse). Subsequently, slides were washed with PBS and incubated with avidin–biotin complex reagent containing horseradish peroxidase for 30 min. Slides were washed with PBS for 5 min and colour development was achieved using DAB substrate. The tissue sections were counterstained with haematoxylin and mounted using Crystal/Mount TM (Biomeda, Foster City, CA, USA). Negative controls were performed using the same protocol but substituting the primary antibody with normal rabbit or mouse IgG (Vector Laboratories Inc., Burlingame, CA, USA). Photo documentation was performed using an Olympus microscope and a digital image analysis system (Image-Pro®, Silver Spring, MD, USA).

Microscopic evaluation.  Two independent observers evaluated all slides without knowledge of the origin of the sample. For each slide, the signal intensity was scored as follows: −, negative; +−, patchy; +, weak; ++, moderate; +++, high. The results were scored to permit statistical analysis.

Statistics.  Results subjected to statistical analyses were expressed as mean ± SEM. Abortion rate, TGF-β1 and PIBF data were analysed by one-way analysis of variance followed by the Tukey's multiple comparisons test. Differences between experimental groups for CD8 and γδ staining were determined by the non-parametric Mann–Whitney U-test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

Spontaneous abortion could be abrogated by syngeneic DC adoptive transfer

As depicted in Table 1, we reproduced our previous findings that syngeneic DC adoptive transfer significantly decreases the abortion rate in DBA/2J-mated CBA/J females, as analysed on gestation day 13.5. Furthermore, no significant changes on the abortion rate were observed in the mice group that received injection of DCCM compared to the untreated control mice. Fig. 1A shows the total number of implantations, which comprise viable placental units (F + PL) and haemorrhagic units as an indicator for abortion (A). Neither syngeneic DC treatment nor injection of DCCM had a significant effect on the number of implantations (Table 1).

Table 1.   Spontaneous abortion could be abrogated by dendritic cell (DC) adoptive transfer.
GroupNumber of mice per groupNumber of total implantationNumber of total abortion
  1. aMice adoptive transferred with syngeneic DC showed a diminished abortion rate compared with control and inoculated with DC-conditioned medium (DCCM) (*P < 0.05) as analysed by one-way analysis of variance followed by the Tukey's multiple comparisons test.

DBA/2J-mated CBA/J females, without treatment64211
DBA/2J-mated CBA/J females, inoculated with DCCM64410
DBA/2J-mated CBA/J females, inoculated with syngeneic DC7482a
image

Figure 1.  Pregnant uterus appearance, placental histology and immunohistochemical analysis. (A) The picture shows a representative example of pregnant uterus on gestation day 13.5, where F means fetus, PL placenta and A abortion. (B, C) Haematoxylin–eosin staining of a healthy and abortion fetus–placenta unit respectively. (D, E) Immunohistochemical analysis of CD8+ cells showing representative samples of moderate and weak staining. (F) Represents an example of negative control for the immunohistochemical reaction. (G, H) Shows a positive staining γδ cells at the fetus–maternal interface. The images show moderate staining for γδ cells (G) and weak staining (H). (I, J) Transforming growth factor-β1 staining are showed in this figure, where (I) represents moderate and (J) weak stainings. (K, M) Fetus–maternal interface are positive for progesterone-induced blocking factor, but a strong trophoblast staining was also found as can be seen in (M).

Download figure to PowerPoint

To confirm absence of any histological alterations and infection, placental tissue from normal and abortion samples were analysed by microscopy. Fig. 1B clearly demonstrates that normal placental tissue presents no evidence of histological alterations or infection. However, samples from abortion tissue revealed important cell infiltration processes and fibrosis (Fig. 1C).

Protective upregulation of CD8 and γδ cells at the fetal–maternal interface

Recently, it was reported that human CD8+ cells, comprising a small population at the fetal–maternal interface [25, 26], contain a subpopulation with regulatory function similar to that exhibited by murine CD8+ cells [27, 28]. In view of these findings, we investigated the expression of cell surface CD8 on uteri-placental samples. As shown in Fig. 1D, the expression level of cell surface CD8 was high in decidua from DC-adoptive transfer treatment; whereas only a small amount of CD8 was found in decidua in control and DCCM-treated groups (Fig. 1E). The statistic of the results is shown in Fig. 2A. An example of negative control staining of our IC is depicted in Fig. 1F.

image

Figure 2.  CD8-positive and γδ cells are increased after dendritic cell transfers. The figures show the statistical results for the CD8α+ (A) and (B) γδ cells staining. The data are presented as mean ± SEM. *P < 0.05 and **P < 0.01, as analysed by the non-parametric Mann–Whitney U-test.

Download figure to PowerPoint

Gamma-delta cells have been shown to play a role in the suppression of the immune system during pregnancy [29–32]. Similar to resident γδ cells at other mucosal sites [33], the decidual γδ cells are activated but resting. Considering their role in various forms of immune tolerance, we sustained the hypothesis that γδ cells were necessary for the induction of fetal acceptance after DC therapy. Uteri-placental tissues obtained from the groups of females that received adoptive transfer of DC showed high number of positive cells in decidua, as well in the spongy trophoblast zone (Fig. 1G) compared with the control and DCCM-inoculated groups (Figs 1H and 2B).

TGF-β1 is upregulated at the fetal–maternal interface following DC therapy

Considering that DC transfer gave rise to an upregulation of CD8 and γδ cells, we wondered whether it was effective in increasing the protective pregnancy molecule, TGF-β1, as a possible explanation for the therapeutic effect of the DC therapy. We evaluated TGF-β1 expression by IC in uteri-placental samples from animals receiving DC. Intracellular TGF-β1 expression was increased in decidua and placental tissue after syngeneic DC treatment (Figs 1I and 3A). Interestingly, TGF-β1 expression was slightly expressed in decidua and placenta tissue from control groups (Fig. 1J). These results indicate that TGF-β1 is increased following DC treatment; it would probably be an important mediator of DC-associated pregnancy protection.

image

Figure 3.  Transforming growth factor (TGF)-β1 was only increased after non-pulsed dendritic cell therapy. The figures show the statistical results for the TGF-β1 (A) and progesterone-induced blocking factor (PIBF) (B) staining. The data are presented as mean ± SEM. *P < 0.05 and **P < 0.01, as analysed by one-way analysis of variance followed by the Tukey's multiple comparisons test.

Download figure to PowerPoint

DC therapy upregulated the expression of PIBF at the fetal–maternal interface

A possible immunomodulatory mechanism in the maintenance of normal pregnancy is the induction of PIBF. In the presence of progesterone, activated lymphocytes and decidual CD56+ cells synthesize PIBF, which exerts a substantial anti-abortive effect in vivo [34]. Its pregnancy-protective activity is mediated by its effects on the humoral and cellular immune system and by reduction of natural killer (NK)-cell activity [35]. To assess whether the transfer of syngeneic DC was able to increase the expression of PIBF, the number of PIBF+ cells was analysed by IC in uteri-placental tissues. The transfer of DC provoked an increase of the number of PIBF+ cells that was not detected in the rest of the groups (Figs 1K and 3B). We observed fewer PIBF+ cells infiltrating the decidua from controls groups (Fig. 1L) when compared with DC-transferred mice. Surprisingly, spongy and labyrinthine trophoblast also expressed PIBF (Fig. 1M).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

Immature DC are a novel tolerogenic type of antigen-presenting cells, capable of inducing T-cell unresponsiveness in vitro and prolonging allograft survival when administered in vivo [36–38]. We have previously shown that DC transfer treatment abrogated abortion rate in a mouse model. Here, we investigated the beneficial effects of this therapy, focusing on the cellular and molecular microenvironment at the fetal–maternal interface.

How does DC therapy decrease the abortion rate in abortogenic mouse pregnancy? We first considered the possibility that DC transfer releases suppressor/regulatory cells and tested the expression of two pregnancy-protective cell populations at the fetal–maternal interface. As mentioned above, the beneficial role of CD8+ cells during pregnancy has been previously reported [25, 28]. Here, we show that CD8+ cells were increased after DC transfer. The relative increase of CD8+ cells in decidua suggests either an active recruitment or local expansion of this population at the fetal–maternal interface. These findings support the concept that CD8+ cells are critical in regulating the CBA/J × DBA/2J model of spontaneous abortion in mice. Indeed, CD8+ cell depletion studies resulted in an increase of the abortion rate [18, 39]. Furthermore, CD8+ cells have been shown to have a suppressive capacity [40], and recently Tilburgs et al. [26] have identified a significantly higher percentage of CD8+ cells in decidua basalis and parietalis compared to peripheral blood suggesting an important role for this T-cell subset locally at the fetal–maternal interface in humans. Consistent with a role for CD8+ cells in tolerance during pregnancy is the finding, in humans, that placental trophoblasts activate a clonal population of CD8+ cells with regulatory function [25]. These cells are not MHC class I restricted, but require costimulation through a member of the carcinoembryonic Ag family present on early gestation trophoblasts [25]. Besides CD8+ cells, γδ cells also are present at the fetal–maternal interface which have been shown to be important in the acceptance of allogeneic fetus in murine pregnancy [35] and maintenance of early pregnancy in humans [29, 41]. Here, we also showed that γδ cells are increased after DC therapy. It is generally accepted that γδ cells combine unique functions of infection protection and immunoregulation during pregnancy [42, 43]. These cells have been implicated in the downregulation of immune responses in various inflammatory disorders and may acquire immunoregulatory properties at mucosal sites [44]. In the context of immunoregulation it is interesting to note that γδ cells exert their effect via secretion of immunosuppressive cytokines IL-10 and TGF-β1.

We next analysed the expression of TGF-β1 in the uteri-placental unit and found that TGF-β1 was upregulated after DC transfer. These data suggest a beneficial effect in the expression of tolerogenic cytokines, like TGF-β1, during DC therapy. Ingman et al. [45] have demonstrated the important role of TGF-β1 in the outcome of murine pregnancy. Moreover, TGF-β1 has been reported to be one of the molecules involved in mechanisms of regulatory T cells to induce tolerant pathways [46, 47]. It would be tempting to speculate that CD8 and γδ cells act as TGF-β1-producing cells to create a decidual environment that actively tolerates the fetus. Two possible mechanisms by which these cells could induce local uterine tolerance towards the fetus are: (1) the direct pathway, where effector cells at the fetal–maternal interface could be directly inhibited by TGF-β1 [48]. In this pathway CD8 and γδ cells function as regulatory T cells [26, 44]; (2) the indirect pathway, where CD8 and γδ cells could mediate their tolerogenic effect through generation of primed Th0, mainly αβ CD4 cells. Under the influence of TGF-β1, these cells differentiate into TGF-β1-producing Th3 type of cells which in their turn act suppressively on the effector cells. In this pathway CD8 and γδ cells are needed for generation of efferent suppressor cells, but are not suppressor themselves [44, 49].

Accumulating evidence has been indicating the important role of PIBF in pregnancy maintenance. PIBF has several anti-abortive effects in vivo and appears to be the pivotal mediator of progesterone-dependent immunomodulation [41, 50]. In this study, we show for the first time that local PIBF expression was significantly different after DC therapy, confirming with this result that different pregnancy protective molecules are involved in the beneficial effects exerted by DC. PIBF has been shown to act as a pregnancy-protective molecule, enhancing AAb production [32, 51], increasing the production of Th2 cytokines (IL-3, IL-4, IL-10) and decreasing Th1 cytokine IL-12 [52]. Indeed, NK activity in pregnant women is inversely related to the rate of PIBF+ lymphocytes [53]. PIBF keeps NK activity at a low level both by controlling IL-12 production and also by inhibiting perforin release [32, 41, 54].

In summary, we have employed syngeneic DC as a means to induce fetal tolerance, a strategy that is comparable with the one used in DC-based cancer vaccine therapy. Because it exploits strategies that have been applied in clinical settings for several years for the treatment of cancer, this work will facilitate the promotion of DC-based fetal tolerance induction regimens. However, several critical parameters including DC cultures as well as timing, site and frequency of injection need to be evaluated carefully through extensive studies in human subjects to establish a safe and reliable DC-based tolerance induction therapy.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References

The authors are especially grateful to Dr M.E. Roux (Cellular Immunology Lab, Facultad de Farmacia y Bioquímica, UBA) for her expert assistance in photo documentation. This work was supported by CONICET Grant to S.M. (PIP 5503) and Charite Grant to P.C.A. and S.M.B. Gabriela Barrientos thanks DAAD (German Academic Exchange) for the fellowship granted. Sandra M. Blois received a postdoctoral fellowship from Ernst Schering Foundation. Petra C. Arck, Sandra M. Blois, Julia Szekeres-Bartho and Laszlo Szereday are part of the EMBIC Network of Excellence, co-financed by the European Commission throughout the FP6 framework program ‘Life Science, Genomics and Biotechnology for Health’.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References