Regulatory T Cell Apoptosis during Preeclampsia May Be Prevented by Gal-2

Abstract

There are several open questions to be answered regarding the pathophysiology of the development of preeclampsia (PE). Numerous factors are involved in its genesis, such as defective placentation, vascular impairment, and an altered immune response. The activation of the adaptive and innate immune system represents an immunologic, particularity during PE. Proinflammatory cytokines are predominantly produced, whereas immune regulatory and immune suppressive factors are diminished in PE. In the present study, we focused on the recruitment of regulatory T cells (Tregs) which are key players in processes mediating immune tolerance. To identify Tregs in the decidua, an immunohistochemical staining of FoxP3 of 32 PE and 34 control placentas was performed. A clearly reduced number of FoxP3-positive cells in the decidua of preeclamptic women could be shown in our analysis (p = 0.036). Furthermore, CCL22, a well-known Treg chemoattractant, was immunohistochemically evaluated. Interestingly, CCL22 expression was increased at the maternal-fetal interface in PE-affected pregnancies (p_{syncytiotrophoblast} = 0.035, p_decidua = 0.004). Therefore, the hypothesis that Tregs undergo apoptosis at the materno-fetal interface during PE was generated, and verified by FoxP3/TUNEL (TdT-mediated dUTP-biotin nick end labeling) staining. Galectin-2 (Gal-2), a member of the family of carbohydrate-binding proteins, which is known to be downregulated during PE, seems to play a pivotal role in T cell apoptosis. By performing a cell culture experiment with isolated Tregs, we could identify Gal-2 as a factor that seems to prevent the apoptosis of Tregs. Our findings point to a cascade of apoptosis of Tregs at the materno-fetal interface during PE. Gal-2 might be a potential therapeutic target in PE to regulate immune tolerance.

Keywords: Galectin-2; apoptosis; preeclampsia; regulatory T cells.

Figure 1

Immunohistochemical staining results of FoxP3: (A) boxplot of the average number of FoxP3-positive Tregs per visual field in control and PE placentas (control n = 34, PE n = 30), mean ± SD; p-values were calculated with Mann-Whitney-U-Test, * p = 0.046; (B) representative picture of control placenta; (C) representative picture of PE placenta. Detected Tregs are marked with arrows. The circle in A symbolizes an outlier value with its respective number for identification. A respective negative and positive control is shown in the Supplementary Data (Figure S1).

Figure 2

Immunohistochemichal staining results of CCL22 in the syncytium: (A) boxplot of the mean IRS of syncytial staining of CCL22 in control and PE placentas (control n = 34, PE n= 30), mean ± SD; p-values were calculated with Mann-Whitney-U-Test, * p = 0.013; (B) boxplot of the mean intensity of syncytial staining of CCL22 in control and PE placentas (control n = 34, PE n = 30), mean ± SD; p-values were calculated with Mann-Whitney-U-Test, * p < 0.001; (C); representative picture of control placenta (IRS = 3); (D) representative picture of PE placenta (IRS = 12). The circles in A and B symbolize outlier values with their respective number for identification. The star in A symbolizes an extreme outlier value with its respective number for identification. A respective negative and positive control is shown in the Supplementary Data (Figure S2).

Figure 3

Immunohistochemical staining results of CCL22 in the EVT (extravillous trophoblast): (A) boxplot of the mean IRS of decidual staining of CCL22 in control and PE placentas (control n = 34, PE n = 30), mean ± SD; p-values were calculated with Mann-Whitney-U-Test; * p = 0.006 (B) boxplot of the mean intensity of decidual staining of CCL22 in control and PE placentas (control n = 34, PE n = 30), mean ± SD; p-values were calculated with Mann-Whitney-U-Test; (C) representative picture of control placenta (IRS = 3), (D) representative picture of PE placenta (IRS = 6). The circles in (A) and (B) symbolize outlier values. Respective negative and positive control is shown in the Supplementary Data (Figure S2).

Figure 4

Immunofluorescence staining of CCL22 (red) and CK7 (green) in the decidua of PE and control placentas. Nuclear staining with DAPI is shown in blue in each case. Representative pictures of control placenta (A–D) and PE placenta (E–H), single staining of CCL22 (A,E) and CK7 (B,F), double staining of CCL22 and CK7 (C,G), and merge including nuclear staining (D,H) are shown. Respective negative control picture is shown in the Supplementary Data (Figure S3).

Figure 5

Immunofluorescence staining of apoptotic Tregs. Graphical representation of all Tregs (FoxP3⁺) and all apoptotic Tregs (FoxP3⁺/TUNEL⁺) counted per sample (A), mean ± SD; * p-values were calculated with Mann-Whitney-U-Test; mean percentage of apoptotic Tregs counted in controls and PE (B), mean ± SD; * p-values were calculated with Mann-Whitney-U-Test; Representative pictures of control placenta (C–F) and PE placenta (G–J), single staining of FoxP3-positive Tregs (red) (C,G), single TUNEL staining (green) (D,H), double staining of apopototic Tregs (E,I) (yellow), merge (F,J), and nuclear staining with DAPI (C,D,F,G,H,J). Non-apoptotic Tregs are marked with arrows; apoptotic Tregs are marked with arrowheads. A respective negative control is shown in the Supplementary Data (Figure S3).

Figure 6

Concentration of standardized active Caspase 3 after exclusion of the second donor, in the groups untreated Treg (Treg, 0.20 ± 0.231), Treg with induction of apoptosis through FasL (Treg + FasL, 1.00 ± 0.024), and the group with the Gal-2 treatment (Treg + FasL + Gal-2, 0.57 ± 0.283). The one-way-ANOVA-Kruskal-Wallis-Test showed significant differences among the three groups (p = 0.002). Further analysis revealed this significant difference between the group Treg and Treg + FasL (p < 0.001), as well as between Treg + FasL and the group Treg+FasL+Gal2 (p = 0.018).