The effects of human pregnancy-specific β1-glycoprotein preparation on Th17 polarization of CD4⁺ cells and their cytokine profile

Isolated CD4⁺ T cells were cultivated for 72 h in the presence of a native PSG, TCR–activator, IL-1β, and IL-6, with subsequent assessment of proliferation using flow cytometry, detection of RORɣτ (RAR (retinoic acid receptor)-related orphan receptor gamma) and Ki-67 expression, and evaluation of different cytokine concentrations in culture supernatants (Fig. 1). We used concentrations of PSG (1, 10, and 100 μg/mL) that correspond to the first, first-second, and second-third trimesters of pregnancy, respectively [6, 25].

When estimating the effect of PSG on the proliferation of CD4⁺ T cells that accompanied their Th17 polarization, it was established that the addition of PSG into CD4⁺ cultures at concentrations of 10 and 100 μg/mL dose-dependently changed the percentages of proliferating and non-proliferating cells (Table 1). According to the differential gating method, PSG concentrations of 10 μg/mL and 100 μg/mL decreased the percentage of proliferating and increased the percentage of non-proliferating helper T cells. PSG did not affect the percentage of dead and apoptotic cells. Thus, PSG, in our study, primarily suppressed the proliferation of CD4⁺ T cells in Th17-polarizing conditions (Table 1).

The antiproliferative effect of PSG was previously known [8, 26], but there was no evidence of the effect of this placental protein hormone on the Th17 pro-inflammatory subpopulation. In our previous study, the general inhibitory effect of PSG on CD4⁺ T cell proliferation was shown [27], but we now confirmed this effect for RORγτ⁺ Th17. This effect is probably due to a single mechanism of latent TGF-β1 activation for all lymphocytes [28, 29].

When studying the effect of PSG on Th17 differentiation, it was found that PSG at a concentration corresponding to the last trimester of pregnancy (100 μg/mL) significantly decreased the number of CD4⁺ lymphocytes expressing the RORγτ transcription factor, the main intranuclear marker of Th17 cells [30] (Fig. 2).

The differentiation processes of Th17 cells are accompanied by the active proliferation of these cells [31]. Therefore, we evaluated the level of intracellular expression of Ki-67, which appears in actively proliferating cells [32, 33].

PSG (10 and 100 μg/mL) suppressed the expression of Ki-67 in Th17-polarized CD4⁺ cells (Fig. 2). This effect correlates with the data from the differential gating method (10 μg/mL: r = 0.69, P = 0.011; 100 μg/mL: r = 0.65, P = 0.018).

Next, we analyzed the effect of PSG on RORγτ expression, depending on the proliferative status of the cells. For this purpose, the percentages of RORγτ⁺Ki-67⁺ (proliferating) and RORγτ⁺Ki-67⁻ (non-proliferating) cells were determined (Fig. 2). PSG at concentrations of 10 and 100 μg/mL decreased the amount of proliferating Th17 (RORγτ⁺Ki-67⁺) cells in the CD4⁺ cell culture. Correspondingly, the percentage of non-proliferating Th17 cells (RORγτ⁺Ki-67⁻) in the culture increased (Fig. 2).

Thus, PSG regulates both the proliferation and differentiation of CD4⁺ cells under Th17 polarizing conditions, rendering mainly an inhibitory effect.

When analyzing the cytokine profile of Th17-polarized CD4⁺ cell cultures, we found that PSG at concentrations of 10 and 100 μg/mL decreased the production of IL-10, IFN-γ, MIP-1β, and TNF-α (Table 2).

In some cases, an inhibitory effect of only a high concentration of PSG on the production of cytokines was found. PSG at concentration of 100 μg/mL suppressed the production of IL-4, IL-5, IL-8, IL-12, IL-13, IL-17, G-CSF, and GM-CSF (Table 2). Among the above-mentioned cytokines, the pregnancy protein had the most pronounced inhibitory effect on the secretion of IL-17 (2.64-fold) and GM-CSF (2.74-fold).

Reduction in IL-17 output in 100 μg/mL PSG CD4⁺ culture correlates with a decrease in the proliferating RORγτ⁺Ki67⁺ cell percentage and inversely correlates with an increase in the non-proliferating RORγτ⁺Ki67⁻ cell percentage and thus may be directly associated with the antiproliferative effect of PSG.

In addition to IL-17, their major effector cytokine, Th17 cells can produce IL-21, IL-22, TNFα, IFNγ, and GM-CSF. This impressive arsenal helps cells cope with extracellular and intracellular bacteria; ensure protective immunity to Mycobacterium tuberculosis, Chlamydia trachomatis, fungi, and viruses; protect mucosal homeostasis; and enhance the neutrophil response. Conversely, it is involved in inflammation and several autoimmune diseases [34]. IL-17 has been shown to activate the nuclear factor (NF)-kB downstream signaling pathway, which results in the expression of pro-inflammatory cytokine genes, such as TNF-α, IL-1, IL-6, G-CSF, and GM-CSF; the chemokines CXCL1, CXCL5, IL-8, CCL2, and CCL7; the matrix metalloproteinases MMP1, MMP3, MMP9, and MMP13; and the antimicrobial peptides defensins and S100 proteins [35]. Since IL-17 is a decisive factor triggering inflammatory reactions, it is quite logical that in our study, a decrease in its concentration in 100 μg/ml PSG cultures correlated with a decrease in TNF-α (r = 0.78, P = 0.003) and IFN-γ (r = 0.73, P = 0.007). TNF-α, like other IL-17-derived pro-inflammatory cytokines, is highly undesirable during pregnancy, as it can causе pregnancy complications, such as recurrent miscarriage, premature rupture of fetal membranes, preeclampsia and intrauterine fetal growth retardation [36].

In our study, PSG at a concentration of 100 μg/mL inhibited the production of G-CSF and GM-CSF by CD4⁺ cells under Th17-polarizing conditions (Table 2). These hematopoietic colony-stimulating factors are necessary for the onset and progression of pregnancy, but above all, locally. Other pregnancy proteins may stimulate their synthesis in vivo. In particular, it is known that the expression of GM-CSF is triggered by chorionic gonadotropin [37], the concentration and therefore the dose-dependent effect of which are significantly higher in the first trimester of pregnancy than PSG.

In addition to inhibiting the “classic” pro-inflammatory cytokines and chemokines, PSG, in our study, also decreased the production of anti-inflammatory cytokines such as Il-4, IL-10, and IL-13. The decrease in IL-10 output in the 10 μg/mL PSG culture was inversely correlated with the percentage of RORγτ⁺ cells. This result is logical since the main producers of IL-10 are Treg, Th2, and the effector Th1 [34, 38]. The presence of these cells in the cultures is very likely since we Th17 polarized not only naive but all CD4⁺ T cells. The cultures initially contained a mixture of different T helper subsets, including Th1, T reg, and Th17, which are known to have plasticity and can transdifferentiate from one subpopulation to another, depending on the cytokine environment [39, 40]. Most likely, the main portion of helper T cells polarized into RORγτ⁺ Th17 due to the cytokine background (Fig. 2); however, some of the CD4⁺ T cells “remained true” to their original phenotype.

Interestingly, although the concentration of IL-4 in cultures of Th17-polarized CD4⁺ cells was initially low in the control (Me = 11.7 pg/ml), in the 100 μg/mL PSG culture, there was a significant decrease in the production of this cytokine (Me = 10.46 pg/ml), which correlated (r = 0.67, P = 0.014) with a reduction in the percentage of Ki67⁺ and RORγτ⁺Ki67⁺ cells. Perhaps, due to the plasticity of the T helper phenotypes, a small percentage of transitional cell subsets co-expressing, along with IL-17, a variety of other cytokines, including IFN-γ, IL-10, and IL-4, is formed in the pro-inflammatory cytokine environment, as occurs with autoimmune diseases [40].

Regarding the suppression of cytokine production, our previous studies have already shown a simultaneous decrease in the production of both pro-inflammatory (IFN-γ) and anti-inflammatory (IL-4) cytokines [27] by CD4⁺ T cells under the influence of PSG. This effect can probably be associated with the “universal” antiproliferative effect of this pregnancy hormone since, in this study, a correlation of the percentage of Ki67⁺ (proliferating) cells was detected with the decrease in concentrations of several cytokines, including IL-4, IL-17, GM-CSF, IFN-γ, and TNF-α, and the chemokine MIP-1β [28, 29, 41].

Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation (Diacoll 1077, Dia-m, Russia, ρ = 1.077 g/cm³). The maximum time between collection of a blood sample and separation in a density gradient was 30 min.

PSG was purified from the blood of healthy pregnant women according to the technique developed by Mikhail Rayev [42] and described in detail elsewhere [27]. PSG preparation is a mixture of the following proteins: PSG1, PSG3, PSG7, PSG9, and some of their isoforms and precursors.

CD4⁺ cells were obtained from PBMCs by positive immunomagnetic separation (MACS® MicroBeads and MS Columns, Miltenyi Biotec, Germany). The purity of the isolated T helper population was checked using a two-color BD Simultest™ IMK-Lymphocyte kit (Becton, Dickinson and Company, BD Biosciences, USA) and a CytoFLEX S Flow Cytometer (Beckman Coulter, USA). The percentage of B-cells (CD3⁻CD19⁺) did not exceed 0.02%, monocytes (CD45⁺CD14⁺) – 0.2%, CTLs (CD3⁺CD8⁺) – 1%, and NK-cells (CD3⁻CD16/56⁺) – 0.05%. The portion of CD3⁺CD4⁺ cells varied within 97–99%.

Isolated CD4⁺ cells were cultured in 96-well plates (1 × 10⁶ cells/mL, 200 μL) in complete medium (CM), RPMI-1640 (Sigma-Aldrich, USA) supplemented with 10% FBS (Sigma, USA), 10 mM HEPES, 2 mM L-glutamine (both from ICN Pharmaceuticals, USA), and 30 μg/mL gentamycin (KRKA, Slovenia), and in a humidified CO₂ incubator at 37 °C and 5% CO₂ for 72 h.

Cultures without PSG served as a control. The viability of the cells after incubation evaluated using 0.4% trypan blue (Invitrogen, USA) was 95–98%. PSG did not affect either the cell number or the viability of cells.

To polarize CD4⁺ cells into Th17 cells, we used TCR-activator (T Cell Activation/Expansion Kit human, Miltenyi Biotec, Germany) and the IL-1β and IL-6 cytokines at concentrations of 10 ng/mL (Miltenyi Biotec, Germany) [43].

Th17 polarization resulted in a significant increase in the percentage of proliferating CD4⁺ cells (Me (Q1-Q3), from 11.32 (9.32–15.63) to 42.33 (34.03–50.14)), accompanied by a simultaneous decrease in the percentage of non-proliferating cells (from 83.56 (77.85–97.53) to 47.84 (39.47–53.57)), while the level of apoptosis cells did not change. In general, this result indicates an adequate activation of CD4⁺ cells in the present experimental model. The obtained data are consistent with similar experiments, where we studied the proliferation of TCR-activated helper T cells in the presence of IL-2 [44].

The independent effect of Th17 polarization consisted of a significant (more than 60-fold) increase in the number of cells expressing Ki-67 (Additional file 1, Table S1).

The effect of Th17 polarization on the cytokine/chemokine profile of CD4⁺ cells was expressed as a reliable increase in IL-2, IL-4, IL-5, IL-7, IL-10, IL-17, G-CSF, GM-CSF, IFN-γ, MIP-1β, and TNF-α output. The most pronounced increase was observed in the TNF-α and IL-17 concentrations, thus confirming the predominance of IL-17-producing cells in culture and, in general, the successful formation of a pro-inflammatory background (Additional file 2, Fig. S1).

The main transcription factor of Th17 cells is RORγτ [26]. Therefore, after 72 h of cultivation, we determined the frequency of Th17 cells as a percentage of RORγτ⁺ CD4⁺ cells. To evaluate the percentage of proliferating Th17 cells, we assessed the level of intracellular Ki-67 protein.

Sample preparation for intracellular/intranuclear staining was performed with a FOXP3 Fix/Perm Buffer Set (Biolegend, USA). Stained samples were analyzed by running a two-color flow cytometry assay with a CytoFLEX S (Beckman Coulter, USA). The antibodies used were anti-RORg(t)-PE, human and mouse, and anti-Ki-67-PerCP-Vio700™, human and mouse (both Miltenyi Biotec, Germany).

The frequencies of RORγτ⁺, Ki-67⁺, ROR-γτ⁺Ki-67⁺, and RORγτ⁺Ki-67⁻ cells were assessed. The data are presented as percentages of RORγτ⁺, Ki-67⁺, RORγτ⁺Ki-67⁺, and RORγτ⁺Ki-67⁻ cells from all events in the gate of living СD4⁺ lymphocytes established according to FSC and SSC properties. The threshold between positive and negative cells was determined using the fluorescence minus one (FMO) controls. Flow cytometry data were analyzed using CytExpert 2.0 software (Beckman Coulter, USA).

An author-modified differential gating method [45] was used to determine the proliferative status of cells. In our study, in contrast to the above-referenced method, we did not calculate the absolute but rather the relative number of proliferating cells. That is, we calculated the percentage of cells in each gate (proliferating, non-proliferating, and apoptotic) from all cells in the three gates [27, 46].

Data were acquired on a CytoFLEX S Flow Cytometer and analyzed in CytExpert 2.0 software (Beckman Coulter, USA).

After 72 h of incubation in the culture supernatants of Th17-polarized CD4⁺ cells, the IL-2, IL-4, IL-5, IL-7, IL-8, IL-10, IL-17, G-CSF, GM-CSF, IFN-γ, MCP-1, MIP-1β, and TNF-α concentrations were determined in the supernatants of CD4⁺ cell cultures using a Bio-Plex® Multiplex Immunoassay System (Bio-Rad, USA).

Procedures were performed according to the “Bio-Plex Pro™ Human Cytokine 27-plex Assay” and “Bio-Plex Pro™ Human Th17 Cytokine Panel 15-Plex” protocols. The results were recorded using a Bio-Plex automatic microplate photometer and Bio-Plex Manager software (Bio-Plex® 200 Systems, Bio-Rad, USA). The cytokine concentrations were determined from a calibration curve for each cytokine (dynamic range from 2 to 32,000 pg/mL) according to the manufacturer’s recommendations.

The statistical data analysis was performed with GraphPad Prism 6 using one-way ANOVA (proliferation, differential gating) and the paired Friedman test with Dunn’s multiple comparisons test (RORɣτ and Ki-67, cytokines). Data are presented as arithmetic means and standard deviations (M ± SD) and medians with first and third quartile values (Me (Q1 − Q3)), respectively. Differences were considered significant at P < 0.05. In some cases, the Spearman correlation coefficient was calculated.