Decidual stromal cells support tolerance at the human foetal-maternal interface by inducing regulatory M2 macrophages and regulatory T-cells

and T-cells. In this study we report that maternally derived DSC are able to induce homeostatic M2 macrophages and Treg cells. CD14 + monocytes and CD4 + T-cells from healthy non-pregnant women were cultured in the presence or absence of conditioned medium (CM) from DSC isolated from 1 st trimester and term placentas. DSC-CM alone was able to promote the survival of macrophages and to induce a regulatory CD14 bright CD163 + CD209 + CD86 dim phenotype, typical for decidual macrophages and similar to that induced by M-CSF. Interestingly, DSC-CM was also able to overrule the pro-inflammatory effects of GM-CSF by upregulating CD14, CD163 and CD209. Protein-profiling showed that M-CSF was secreted by DSC, and blocking of M-CSF partially reversed the M2 phenotype and reduced viability. DSC-CM also expanded CD25 bright Foxp3 + Treg cells, an expansion that was abolished by a SMAD3-inhibitor, indicating the contribution of TGF-β signaling. In conclusion, our findings collectively emphasize the role of tissue-resident stromal cells in shaping the tolerogenic environment at the foetal-maternal interface.


Introduction
Pregnancy exerts a great challenge for the maternal immune system due to the presence of the semi-allogeneic foetus, and requires adaptation and acquisition of tolerance by the maternal immune system in order to prevent rejection of the foetal-placental unit (Svensson-Arvelund et al., 2013).This tolerance is mainly established locally in the decidua (the uterine endometrium during pregnancy) by cross-talk between maternal leukocytes, stromal cells and invading foetal trophoblast cells (Arck and Hecher, 2013).Failure to induce proper tolerance can lead to pregnancy complications such as preeclampsia (Hsu and Nanan, 2014), preterm birth (Goldenberg et al., 2008) and miscarriages (Christiansen and Nielsen, 2008).
During pregnancy the decidua is populated by a unique composition of leukocytes, where both macrophages and regulatory T (Treg) cells are enriched and associated with immune regulation (Bulmer et al., 2010).Macrophages in the decidua express phenotypical markers such as CD163 and CD209, and show high expression of CD14 (Svensson et al., 2011), while having low expression of the co-stimulatory marker CD86 (Heikkinen et al., 2003).Furthermore, decidual macrophages produce high amounts of anti-inflammatory cytokines and chemokines, such as interleukin (IL)-10 and chemokine (C-C motif) ligand (CCL)18, and in the absence of inflammatory insult they produce low levels of tumour necrosis factor, IL-12 and C-X-C motif chemokine (CXCL) 10, hence being associated with a regulatory, homeostatic phenotype (Svensson-Arvelund and Ernerudh, 2015).Several factors, including macrophage-colony stimulating factor (M-CSF), IL-10 (Svensson et al., 2011) and IL-34 (Lindau et al., 2018) have been shown to influence polarization of macrophages into a regulatory decidual phenotype.CD25 bright Foxp3 + Treg cells are, in contrast to most other T-cell populations, enriched in the decidua during pregnancy compared to blood (Heikkinen et al., 2004;Mjösberg et al., 2010;Tilburgs et al., 2008) and are crucial for a successful pregnancy (Rowe et al., 2012).For induction and expansion of Treg cells, the transforming growth factor (TGF)-β family of cytokines holds a central position (Li and Flavell, 2008).
Decidual stromal cells (DSCs), the largest cell population in the decidua, develop from endometrial stromal cells under the influence a of post-ovulatory rise in oestrogen and progesterone levels.This process, termed decidualization, prepares the uterus for implantation of the embryo.DSCs promote pregnancy by production of hormones, such as prolactin, but also by influencing the local immunological microenvironment through production of cytokines and chemokines (Oreshkova et al., 2012;Vinketova et al., 2016).Furthermore, DSCs have been shown to limit natural killer cell and dendritic cell activation (Croxatto et al., 2014;Hu et al., 2015;James-Allan et al., 2020;Yang et al., 2019).However, it is not known if DSCs contribute to the regulation of decidual macrophages and Treg cells.In this article we report the secreted protein profile of tissue-derived DSCs and demonstrate that soluble factors, in particular through M-CSF and TGF-β signalling, promote the expansion of regulatory macrophages and Treg cells, thereby contributing to a tolerant micro-environment at the foetal-maternal interface.
First trimester decidual tissues (7th -12th weeks of gestation) were derived from four women undergoing elective termination of pregnancy for non-medical reasons.The isolation procedure was modified from (Saleh et al., 2011) Briefly, decidual tissues were washed in ice-cold PBS, cut into small pieces using a scalpel, and stored overnight at 4 • C in phenol red-free Dulbeccos modified eagle medium (DMEM), supplemented with 0.05 mg/mL gentamicin (both from Invitrogen).The tissues were digested in Hank's balanced salt solution (HBSS) containing 25 mM HEPES, 12.5 mg/mL DNase (Sigma-Aldrich) and 2 mg/mL Collagenase I (Worthington) at 37 • C with gentle agitation, in two subsequent digestion steps of 45 min each.The digested cells were filtered through a 70 μm cell strainer (Falcon) and washed twice with HBSS.Red blood cells were lysed by incubating cells in a buffer containing 155 mM NH4Cl (Sigma-Aldrich), 10 mM KHCO 3 (Merck) and 0.1 mM EDTA (Merck), pH 7.3, for 5 min.Cells were washed twice with HBSS and filtered through a 40 μm cell strainer (Falcon).Isolated cells were seeded in phenol red-free DMEM/F12, supplemented with 10 % heat-inactivated foetal calf serum (FCS) (PAA Laboratories GmbH, Pasching, Austria), 1% ITS + Premix (BD Biosciences), and 50 μg/mL Gentamicin and cultivated at 37 • C, 5% CO 2 and 95 % humidity.At 90 % confluency, DSCs were washed with PBS and passaged using 0.25 % Trypsin-EDTA (Gibco).The first trimester DSC have previously been phenotypically characterised, and the purity is high (>99 %), in both 1 st and term DSC cultures, since other cells are prohibited from growth because of the expanding and resistant DSCs (Saleh et al., 2011).For functional studies, we used passage 1-4 for 1 st trim DSCs and for protein screening passage 4 was used.

Collection of conditioned medium from decidual stromal cells
For preparation of CM from DSCs, 1 × 10 6 cells were seeded per well of a 24-well plate in 1 mL cDMEM and grown for 20− 24 h at 37 • C, 5% CO 2 .This procedure was used in order to standardise the number of cells and duration of culture, thereby minimising inter-individual variations in growth rate.The DSC-CM was collected, centrifuged at 500 x g for min at 7 • C, and stored in aliquots at − 70 • C.

Isolation of CD14 + and CD4 + cells
Buffy coats or blood was obtained from healthy, non-pregnant women (18-45 years old) and stored in citrate phosphate dextrosecoated bags or heparin-coated tubes.Peripheral blood mononuclear cells were obtained using Lymphoprep (Axis-Shield, Oslo, Norway) according to the manufacturer's protocol.CD14 + or CD4 + cells were isolated with positive selection using magnetic activated cell sorting (MACS) according to the manufacturers protocol (Miltenyi Biotec, Bergisch Gladbach, Germany) and the purity was <97 % for CD14+ cells and <99 % for CD4+ cells.
To analyse the effect of DSC-CM on T-cells, freshly isolated CD4 + Tcells (0.25 × 10 6 cells/mL) from healthy control women were cultured in a 96-well plate for 5 d at 37 • C, 5% CO 2 , in cDMEM or DSC-CM supplemented with or without vehicle control (DMSO) or 7.5μM specific inhibitor of SMAD3 (SIS3) (Tocris Bioscience, Bristol, UK).

Flow cytometry
Phenotyping of macrophages (CD163, CD14, CD209, CD86 CD206 and HLA-DR) and T-cells (CD4, CD25, Foxp3) was performed after 6 and 5 days of culture, respectively.Viability was assessed by 7AAD and Annexin V (both from BD Biosciences).Antibodies used in phenotyping are listed in Table 1.For macrophage markers, isotype controls were used to set gates.Additionally, relative marker expressions were calculated as the ratio of the MFI of Ab-stained cells and isotype control Abstained cells.Gating of T-cells were contour-based, and no isotype controls were used.The staining procedures and gating strategies (Fig. S1) are described in detail in Supplementary data.FACSCanto II and FACSDiva software version 8 (BD Biosciences) were used for phenotyping and data acquisition and Kaluza software version 1.2 (Beckman Coulter) was used for data analysis.

Protein profiling of conditioned medium from decidual stromal cells
Conditioned media from 1 st trimester and term DSC (n = 4) were analysed with proximity extension assay (PEA) using the Olink Inflammation panel, which includes 92 proteins (https://www.olink.com/products/inflammation) at the Clinical Biomarkers facility, Science for Life Laboratory, Uppsala University, Uppsala, Sweden.Briefly, oligonucleotide-conjugated Abs, directed towards two different epitopes of the proteins, were incubated with 1 μL sample.When pair-wise binding of the antibodies occurs, the proximity of the oligonucleotides initiates proximity-dependent DNA polymerisation, creating new PCR targets.Next, real-time PCR (Fluidigm® BioMark™HD System) is used to amplify and quantify the PCR targets.The data is shown as arbitrary Normalised Protein eXpression (NPX) values, which are in a Log 2 scale, used for relative quantification.In addition, multiple bead technology was used to measure additional cytokines and chemokines, see Supplementary data.

Statistics
One-way ANOVA for paired data with Sidak's post-hoc test was used for flow cytometric data.Student's t-test was used to compare protein levels by PEA, while Mann-Whitney U test was used for the multiplex bead assay.Two-tailed tests were used, and p-value <0.05 was considered significant.All data were analysed using GraphPad Prism version 8.2.0 (La Jolla, CA, USA).

Conditioned medium from term decidual stromal cells over-rule the pro-inflammatory effects of GM-CSF and induce regulatory M2 macrophages
To investigate the potential role of DSCs in polarisation of macrophages, we utilised a previously established in vitro assay (Svensson-Arvelund et al., 2015;Svensson et al., 2011) in which CD14 + blood monocytes cultured with GM-CSF gain a pro-inflammatory M1-like phenotype.Notably, addition of conditioned medium from term DSC was able to over-rule this pro-inflammatory phenotype as evident by significantly increased expression of decidual macrophages markers CD163, CD14 and CD209, and decreased expression of the co-stimulatory molecule CD86, as shown by proportion of cells (Fig. 1A-D), and by MFI (Fig. S2).For CD206 and HLA-DR, DSC-CM did not significantly alter the expression (data for CD206 and HLA-DR are shown in Fig. S3).Unexpectedly, we also observed an increase in viability of the macrophages cultured in the presence of DSC-CM (Fig. 1E).These results demonstrate that DSCs not only possess the ability to over-rule the pro-inflammatory effects of GM-CSF and promote expansion of macrophages with a decidual macrophage phenotype, but also possess the ability to support macrophage survival.

Conditioned medium from term and 1 st trim decidual stromal cells alone promotes polarisation of macrophages into a decidual phenotype and support their survival
Inspired by the viability-promoting effects, we investigated if DSC-CM alone, without addition of any growth factor, had the ability to support growth and polarisation of macrophages.Blood monocytes from healthy non-pregnant women were therefore cultured with either 100 % term DSC-CM or 100 % 1 st trim DSC-CM, and compared with monocytes cultured with either GM-CSF, representing M1-like macrophages, or M-CSF, representing M2-like macrophages similar to those in decidua (Svensson et al., 2011).Indeed, CM alone, both from term (Fig. 1F-H) and from 1 st trimester (Fig. 1K-M), was able to polarise monocytes into a phenotype that was similar to that of M-CSF-polarised macrophages, and significantly different compared with GM-CSF-polarised macrophages with high expression of CD163 and CD14.The proportion of cells expressing CD209 was not significantly affected, while the expression of CD209 per cell (MFI) increased significantly in presence of both term and 1 st trimester CM (Fig. S2) as compared with GM-CSF.In accordance, CD86 showed a lower expression (Fig. 1I and N) compared with macrophages cultured in GM-CSF, and also compared with M-CSF.Moreover, also in this setting, macrophages cultured in 100 % DSC-CM exhibited a high viability, which indeed was higher compared with M-CSF and GM-CSF cultured macrophages (Fig. 1J and O).Collectively, DSC-CM alone, both from term and 1 st trimester, is able to promote survival of macrophages and polarise them into a phenotype that is similar to that of decidual and M-CSF-polarised macrophages, and strikingly different from that of GM-CSF.

M-CSF is partially responsible for the M2-polarizing and the viabilitypromoting properties of decidual stromal cells
To determine factors responsible for the polarising and viability promoting effects, we blocked M-CSF since it is known to polarise macrophages into a decidual phenotype (Svensson et al., 2011;Svensson-Arvelund et al., 2015), and since it secreted by DSC (see section 3.5 below).Neutralising antibodies towards M-CSF were found to significantly decrease the expression of CD163 (proportion and MFI) on macrophages, both when using DSC-CM together with GM-CSF as a basal growth factor (Fig. 2A-B), as well as when using DSC-CM alone (Fig. 2D-E).For the other macrophage markers (CD14, CD209, CD86, CD206, HLA-DR), there were few and inconsistent changes (Fig. S4).M-CSF neutralising antibodies also reduced the viability in the setting with DSC-CM and GM-CSF (Fig. 2C), although this was not seen when using DSC-CM alone (Fig. 2F).Thus, M-CSF is in part responsible for the polarising and survival promoting effects of DSC-CM.However, both polarisation and viability properties were only partially reversed by M-CSF blocking; hence other factors are also in play.

Conditioned medium from decidual stromal cells induces regulatory T-cells, an induction that is dependent on TGF-β family signalling
We also investigated if CM from DSCs could induce Treg cells, another central immune regulatory cell at the foetal-maternal interface.Therefore, CD4 + cells were isolated from healthy non-pregnant women and cultured in the presence of DSC-CM.After 5 days of culture, the CD4 + T-cells treated with 1 st trimester and term DSC-CM (Fig. 3) acquired an increased proportion of CD25 bright Foxp3 + Treg cells compared with CD4 + T-cells cultured in control medium.The increase in T-reg cells could be abolished by blocking the TGF-β signalling molecule SMAD3, using the inhibitor SIS3 (Fig. 3).

Protein profiling of conditioned medium from decidual stromal cells
In order to map the secreted protein profile of DSCs, 24-h supernatants (conditioned medium) were collected from 1 st trimester and term DSCs cultured ex vivo.We used PEA for measurement of 92 and multiplex bead assay (Fig. S5) for measurement of 26 inflammatory and antiinflammatory proteins.The proteins that were detected in at least 75 % of samples in the PEA are presented in Fig. 4A (term) and 4 B (1 st trimester) and multiplex bead assay findings are presented in Fig. S5, and the remaining proteins are listed in Supplementary Table 1.In general, 1 st trimester and term DSC-CM showed similar findings.Of note, M-CSF and TGF-β were among the abundantly secreted proteins, while GM-CSF and G-CSF were detected only at low levels.Besides M-CSF and TGF-β, DSCs secreted several other regulatory proteins like CCL2, HGF and LIF, while also several pro-inflammatory cytokines, like CXCL8 and IL-6, as well as tissue-remodelling proteinases including MMP-1 and MMP-10, were secreted, showing the multipotent capacity of DSCs.The multiple bead assay added some growth factors that were not included in the PEA (GM-CSF, G-CSF, IL-3), and it also confirmed some findings in the PEA, like secretion of CCL2, CXCL1, CXCL8, IL-6 and M-CSF (Fig. S5).

Discussion
In this study we provide evidence that tissue-resident DSCs actively contribute to shaping the foetal-maternal interface by inducing regulatory M2 macrophages and Treg cells, the two main regulatory and tolerance promoting maternal leukocyte populations in the decidua.M-CSF and TGF-β signalling were in part responsible for these effects, and protein profiling revealed secretion of these, as well as of several other cytokines, chemokines, growth factors and tissue remodelling proteins.
A main finding of this study is the prominent ability of DSC-CM to induce macrophages with a high expression of CD163, CD14, CD209 and low expression of CD86, a phenotype associated with decidual macrophages (Heikkinen et al., 2003;Svensson et al., 2011).We showed this ability both in the setting where DSC-CM over-ruled the pro-inflammatory properties of GM-CSF, and in the setting of macrophages cultured in DSC-CM only, which was possible because of the viability-promoting effects of DSC-CM.In addition to CD163, CD14, CD209 and CD86, we also used CD206 and HLA-DR, since they have been reported to be expressed on decidual macrophages (Heikkinen et al., 2003;Svensson-Arvelund and Ernerudh, 2015).However, we find that the pro-inflammatory GM-CSF, induces a high expression of CD206, in line with our previous findings (Svensson et al., 2011), and data on HLA-DR expression was inconclusive (commented in more detail in Supplementary data file), hence CD206 and HLA-DR markers did not contribute to the interpretation of findings.
It was not surprising that M-CSF was secreted by DSCs and that it was partially responsible for inducing the decidual phenotype of macrophages.We have previously shown that M-CSF and IL-10 are key cytokines in polarisation of macrophages to a phenotype that resembles that of decidual macrophages, a phenotype that is distinctly different from macrophages induced by the Th2 associated cytokines IL-4 and IL-13 (Svensson et al., 2011;Svensson-Arvelund et al., 2015).Accordingly, we find that M-CSF is abundantly secreted by DSCs, while IL-4 and IL-13 are not.Since IL-34 binds to the same receptor as M-CSF (Lin et al., 2008), IL-34 is a potential candidate for macrophage polarisation.In line with this, we previously showed that IL-34 promotes survival of and polarises macrophages to a phenotype similar to that of M-CSF.However, although IL-34 is expressed by 1 st trimester DSCs, we were unable to detect secreted IL-34 in supernatants (Lindau et al., 2018).Taken together, M-CSF emerges as a major factor for DSC-induced macrophage polarisation, although other factors may also be involved.Another novel finding was the clear survival-promoting effects exerted by DSCs.Interestingly, M-CSF was involved also here, further emphasising the central homeostatic and tolerance promoting role of M-CSF at the foetal-maternal interface.
DSCs also induced expansion of CD25 bright Foxp3 + Treg cells, cells that are enriched in the decidua during a normal pregnancy (Heikkinen et al., 2004;Mjösberg et al., 2010;Tilburgs et al., 2008).Mesenchymal stem cells (MSCs) are also able to induce Treg cells (Mareschi et al., 2016), but given their localisation in the placental matrix, we believe that the DSCs, in close proximity to maternal decidual leukocytes should have a more prominent role.We previously showed that trophoblast cells, including extravillous trophoblasts, also induce Treg cells, which was mediated by TGF-β, IL-10 and TRAIL (Svensson-Arvelund et al.,  2015).Blocking experiments of DSCs showed a prominent role of signalling via activation of SMAD3, which after translocation to the nucleus (Moustakas et al., 2001) binds to the foxp3 enhancer element CNS1 (Schlenner et al., 2012).The high abundance of TGF-β in the protein screening, corroborating previous finding of TGF-β secretion by DSCs (Erkers et al., 2013), in combination with the lack of both TRAIL and IL-10, support the notion that TGF-β signalling is the main inducer of Treg cells by DSC-CM.However, SMAD3 is not only activated by TGF-β (Chen and Ten Dijke, 2016), thus other TGF-β family members could also contribute.
Interestingly, the polarising and survival-promoting effects on macrophages, as well as the Treg cell-inducing effects were executed by DSCs from both 1 st trimester and from term placentas, and the secreted protein profiles were also similar.Since DSCs were obtained from gestational weeks 7-38, DSCs seem to have a tolerance inducing role during most of the pregnancy.
Given the distinct role of DSCs in instructing macrophages and Treg cells, a dysfunction of these interactions could affect pregnancy outcome.In preeclampsia, a lower relative frequency of CD163 + macrophages was found in the decidua (Schonkeren et al., 2011), as well as an imbalance of polarising factors with increased levels of GM-CSF (Huang et al., 2010).The importance of macrophage polarisation is further evident in M-CSF knockout mice, which experience increased rates of spontaneous abortions (Pollard et al., 1991).Similarly, mice lacking the foxp3 enhancer element CNS1 showed deficiencies in Treg cells numbers and functions, and consequently, they displayed increased resorption rates, and abnormal spiral artery remodelling (Samstein et al., 2012).
The prominent immune-modulating effects of DSCs are currently under investigation as a novel treatment strategy in inflammatory conditions, in particular graft-versus-host disease (GvHD) (Karlsson et al., 2012;Ringden et al., 2018).These maternal stromal cells possess several advantages over the more commonly used bone marrow-derived mesenchymal stem cells (MSCs) since they are easily obtainable in high amounts, are half the size of MSCs, do not differentiate to unwanted phenotypes, and have a stronger expression of adhesion molecules for homing to inflamed tissues.Our present findings, adding a prominent role of DSCs to inhibit an inflammatory M1-like environment and promoting an M2-like anti-inflammatory environment, further strengthen the potential of DSCs as a treatment strategy in inflammatory conditions including multiple sclerosis (Shapouri-Moghaddam et al., 2018).
In conclusion, we show that DSCs promote the viability and polarisation of monocytes into macrophages of a decidual phenotype, as well as induce the expansion of Treg cells and actively produce factors required for tissue homeostasis.The data support a major role for tissueresident DSCs in shaping the micro-environment at the foetal-maternal interface and strengthen the potential use of DSCs for treatment of inflammatory conditions.

Funding
The study was supported by the Swedish Research Council grant 2018-02776, the Medical Research Council of Southeast Sweden, ALF grants (Jan Ernerudh) and by Linköping University.HK was supported by grants from the Swedish Research Council(2019-01311), the Cancer Society in Stockholm (141193), the Swedish Cancer Foundation (CAN 2014/793), the Swedish Childhood Cancer Foundation (PR2015-0059) and Karolinska Institutet.  ) from passage 4 were cultured at standardized conditions for 20-24 h and protein levels were measured in the supernatants (conditioned medium).Proteins that were detected in >75 % of samples are reported, the remaining proteins are shown in Suppl.Table 1.Comparisons between conditioned medium from (A) term decidual stromal cells (n = 4) and (B) 1 st trimester decidual stromal cells (n = 4) showed no significant differences (Student's t-test).Data are expressed as normalised protein eXpression (NPX) values, which are in a Log 2 NPX scale.Graphs show mean and SD.DSC-CM, conditioned medium from decidual stromal cells.

Fig. 2 .
Fig. 2. Neutralising M-CSF in cultures with conditioned medium from term decidual stromal cells reduces the viability and the expression of the M2 marker CD163.Macrophages were generated by culturing blood monocytes from healthy, non-pregnant women, for 6 days in the presence of 5 ng/mL GM-CSF and 50 % conditioned medium from term decidual stromal cells, with either isotype control antibodies or αM-CSF antibodies (A-C).In (D-F), macrophages were generated by 100 % conditioned medium from term decidual stromal cells, with the addition of isotype control antibodies or with αM-CSF antibodies.Expression of CD163 and viability markers (Annexin V and 7AAD) were evaluated by flow cytometry.(A-C n = 10, D-E n = 7, F n = 5).Data were analysed by one-way ANOVA with Sidak's multiple comparison test.Graphs show mean and SD.* p < 0.05, ** p < 0.01, *** p < 0.001.CM, conditioned medium, ctrl, control, GM, GM-CSF, MFI, median fluorescence intensity, MΦ, Macrophages.

Fig. 3 .
Fig. 3. Conditioned medium from 1 st trimester and term decidual stromal cells induce expansion of regulatory T cells, which is reversed by the SMAD3 blocker SIS3.(A) representative dot plots.(B and C) CD4 + T cells were isolated from healthy, nonpregnant women and cultured for 5 days in either (B) 1 st trimester decidual stromal cell conditioned medium, with or without vehicle control or 7.5μM SIS3 (n = 6); or in (C) conditioned medium from term decidual stromal cells with or without vehicle control or 7.5μM SIS3 (n = 6).Expression of Foxp3 and CD25 were analysed by flow cytometry.Data were analysed with one-way ANOVA with Sidak's multiple comparison test.Graphs show the mean and SD.* p < 0.05, ** p < 0.01.CM, conditioned medium, Ctrl, control, SIS3, specific inhibitor of SMAD3, VC, vehicle control.

Fig. 4 .
Fig. 4. Secreted protein profile of decidual stromal cells by proximity extension assay.Decidual stromal cells (1 × 10 6) from passage 4 were cultured at standardized conditions for 20-24 h and protein levels were measured in the supernatants (conditioned medium).Proteins that were detected in >75 % of samples are reported, the remaining proteins are shown in Suppl.Table1.Comparisons between conditioned medium from (A) term decidual stromal cells (n = 4) and (B) 1 st trimester decidual stromal cells (n = 4) showed no significant differences (Student's t-test).Data are expressed as normalised protein eXpression (NPX) values, which are in a Log 2 NPX scale.Graphs show mean and SD.DSC-CM, conditioned medium from decidual stromal cells.

Table 1
Antibodies used for phenotyping of macrophages and CD4 + T-cells.