Anti-phospholipid induced murine fetal loss: novel protective effect of a peptide targeting the β2 glycoprotein I phospholipid-binding site. Implications for human fetal loss

Abstract

β2 glycoprotein I (β2GPI)-dependent anti-phospholipid antibodies (aPL) induce thrombosis and affect pregnancy. The CMV-derived synthetic peptide TIFI mimics the PL-binding site of β2GPI and inhibits β2GPI cell-binding in vitro and aPL-mediated thrombosis in vivo. Here we investigated the effect of TIFI on aPL-induced fetal loss in mice. TIFI inhibitory effect on in vitro aPL binding to human trophoblasts was evaluated by indirect immunofluorescence and ELISA. TIFI effect on aPL-induced fetal loss was investigated in pregnant C57BL/6 mice treated with aPL or normal IgG (NHS). Placenta/fetus weight and histology and RNA expression were analyzed. TIFI, but not the control peptide VITT, displayed a dose-dependent inhibition of aPL binding to trophoblasts in vitro. Injection of low doses of aPL at day 0 of pregnancy caused growth retardation and increased fetal loss rate, both significantly reduced by TIFI but not VITT. Consistent with observations in humans, histological analysis showed no evidence of inflammation in this model, as confirmed by the absence of an inflammatory signature in gene expression analysis, which in turn revealed a TIFI-dependent modulation of molecules involved in differentiation and development processes. These findings support the non-inflammatory pathogenic role of aPL and suggest innovative therapeutic approaches to aPL-dependent fetal loss.

Figure 1

Human anti- β2GPI antibody binding to trophoblast cells (A to C) Trophoblast monolayers were exposed to IS3 (25 μg/ml; panels B and C) or irrelevant moAb (panel A) in the absence (panel B) or in the presence of TIFI (20 μg/ml) (panel C) and exogenous β2GPI (5 μg/ml) (panel B and C). The moAb binding was revealed by FITC-labeled goat anti human IgG and evaluated by fluorescence microscopy. (D) Trophoblast cell cultures were incubated with aPL (50 μg/ml), β2GPI (5 μg/ml) and serial concentrations of TIFI or VITT. The aPL binding was revealed by alkaline phosphatase-labeled goat anti-human IgG and expressed as mean optical density (ODx10⁻³) units ± SEM. *p<0.01 vs VITT-exposed cells by Student t test.

Figure 2. Dose-response of aPL-induced fetal damage

Effect of increasing amounts of aPL on fetal loss (panel A) and fetus weight (panel B) at day 15 of pregnancy. Results are reported as mean ± SEM of the total number of fetuses from 3 mice/group. *p<0.05 and **p<0.01 vs PBS-treated mice by Fisher exact test (panel A) and Student t test (panel B).

Figure 3. TIFI effect on aPL-induced fetal damage

TIFI or VITT (40 μg/mouse on day 0, 5 and 10) effect on fetal loss (panel A), weight of embryonic sacs (panel B), fetuses (panel C) and placentas (panel D). Mice were treated with: PBS (C; 8 mice), NHS (11 animals), aPL (8 animals), aPL+TIFI (11 animals) or aPL+VITT (4 animals). Numbers under columns are total number of events evaluated. In panel A results are reported as percentage of resorbed fetuses and statistical significance is calculated by Fisher exact test. In panels B to D results are reported as mg, mean ± SEM, and statistical significance is calculated by ANOVA with Tukey post-test. *p<0.05 and **p<0.01 vs PBS-treated mice.

Figure 4. Histological analysis of the placentas

Hematoxylin-eosin staining (OM: 10x) of placenta sections at day 15 of pregnancy from animals treated with aPL (panel A), aPL+TIFI (panel B), aPL+VITT (panel C), or NHS (panel D). In all groups no signs of overt inflammation are evident in the labyrinth (L) and the junctional zone (jz), and a few leukocytes and focal areas of mild necrosis (**) are detected in the maternal site of placenta (decidua basalis, db).

Figure 5. Gene expression analysis

Venn diagram representing number of genes differentially expressed comparing aPL vs NHS, aPL vs aPL+TIFI and aPL+TIFI vs NHS treated mice (panel A). Over-represented biological processes related to differentially expressed genes identified by GO categories enrichment in aPL vs NHS (filled columns) and aPL vs aPL+TIFI (open columns) comparisons (panel B).