LMWH prevents thromboinflammation in the placenta via HBEGF-AKT signaling

Abstract

Low molecular weight heparins (LMWH) are used to prevent or treat thromboembolic events during pregnancy. Although studies suggest an overall protective effect of LMWH in preeclampsia (PE), their use in PE remains controversial. LMWH may convey beneficial effects in PE independent of their anticoagulant activity, possibly by inhibiting inflammation. Here, we evaluated whether LMWH inhibit placental thromboinflammation and trophoblast NLRP3 inflammasome activation. Using an established procoagulant extracellular vesicle-induced and platelet-dependent PE-like mouse model, we show that LMWH reduces pregnancy loss and trophoblast inflammasome activation, restores altered trophoblast differentiation, and improves trophoblast proliferation in vivo and in vitro. Moreover, LMWH inhibits platelet-independent trophoblast NLRP3 (NLR family pyrin domain containing 3) inflammasome activation. Mechanistically, LMWH activates via heparin-binding epidermal growth factor (HBEGF) signaling the PI3-kinase-AKT pathway in trophoblasts, thus preventing inflammasome activation. In human PE placental explants, inflammasome activation and PI3-kinase-AKT signaling events were reduced with LMWH treatment compared with those without LMWH treatment. Thus, LMWH inhibits sterile inflammation via the HBEGF signaling pathway in trophoblasts and ameliorates PE-associated complications. These findings suggest that drugs targeting the inflammasome may be evaluated in PE and identify a signaling mechanism through which LMWH ameliorates PE, thus providing a rationale for the use of LMWH in PE.

Graphical abstract

Figure 1.

LMWH prevents EV-mediated placental dysfunction. (A-C) Representative images (A; top, embryos; bottom, placenta) and bar graphs (B-C) showing reduced fetal death (B) and improved embryonic height (C) after LMWH treatment in EV-injected mice. Pregnancy outcome assessed in C57BL/6 mice at day 12.5 p.c. after IV injection of mouse endothelial cell–derived procoagulant EVs at day 10.5 p.c. and 11.5 p.c. (D-E) Hematoxylin and eosin (H&E) staining of murine placenta (representative images [D]; bar graph summarizing results [E]) showing LMWH prevents EV-induced enhanced maternal vascularization (blood lacunae; enucleated erythrocytes, arrows) and reduced fetal vascularization (nucleated erythrocytes, arrow heads) after EV injections. Vascularized areas were calculated using ImageJ. Analyses performed at day 12.5 p.c. Size bar represent 20 μm. (F-G) Ki-67 staining on murine placenta (representative images [F]; bar graph summarizing results [G]) showing LMWH restores EV-induced impaired placental proliferation. Size bar represent 20 μm. (H) Representative Immunostaining images from mice placenta showing increased CD62P (P-selectin, red; DAPI [4′,6-diamidino-2-phenylindole]–stained nucleus, blue) signals in placentas from EV-injected dams, suggesting increased activated platelets within the placenta. LMWH-treated mice (EV+LMWH) showed reduced CD62P signaling, suggesting LMWH prevents platelet activation. Size bar represent 20 μm. (I) Bar graphs summarizing results from qRT-PCR for mouse trophoblast differentiation gene (PL-II, Tpbpa, Esx1, and Gcm-1) showing that LMWH prevents EV-induced abnormal placental differentiation. Control mice, C, were injected with the supernatant obtained after the last PBS wash during EV isolation. For panels B-C,E,G,I, data shown represent mean ± standard error of the mean (SEM); n = 5 to 8 placentas from 3 dams (∗P < .05, nonsignificant [ns]; analysis of variance [ANOVA]; Šídák multiple comparisons test). C, control; EV, extracellular vesicle (EV)-injected mice; EV+LMWH, EV-injected mice with low-molecular weight heaprin (LMWH) injections; qRT-PCR, quantitative reverse transcriptase-polymerase chain reaction; PL-II, placental lactogen-II; Tpbpa, trophoblast specific protein alpha; Esx1, Extraembryonic, Spermatogenesis, Homeobox 1 and Gcm-1, Glial Cells Missing Transcription Factor 1.

Figure 2.

LMWH prevents NLRP3 inflammasome activation. (A-B) Inflammasome activation in murine placentas after LMWH and EV injections. Representative immunoblots (A; each lane represents individual mice placentae) and bar graphs (B) showing reduced EV and platelet-induced expression of NLRP3, cleaved IL-1β, and cleaved caspase-1 in murine placenta analyzed at day 12.5 p.c. Control mice, C, were injected with the supernatant obtained after the last PBS wash during EV isolation. (C-E) Inflammasome activation in murine and human trophoblast cells. Representative immunoblots (C) and bar graphs (D, E) showing reduced EV and platelet (PRP) induced expression of NLRP3, cleaved IL-1β, and cleaved caspase-1 due to LMWH in mouse (trophoblast cells [TS] cells C-D) and human (JEG-3, C,E) trophoblasts. Control cells, C, were exposed to supernatant obtained after the last PBS wash during EV isolation. EV represents cells exposed to EV and platelets (PRP); EV+LMWH, cells pretreated with LMWH and exposed to EVs and platelets. (F-H) Representative immunoblots (F) and bar graphs (G-H) showing that LMWH prevented LPS+ATP–induced expression of NLRP3, cleaved IL-1β, and cleaved caspase-1 in mouse (TS cells [F-G]) and human (JEG-3 [F,H]) trophoblasts. Control cells, C, were exposed to PBS. LPS+ATP cells were pretreated with 200 ng/mL LPS (24 hours) and then exposed to 1-mM ATP (1 hour). ATP+LMWH cells pretreated with LMWH (30 minutes), followed by treatment with LPS (24 hour) and then exposed to ATP (1 hour). Inactive (proform) and active (cleaved form) forms of IL-1β or caspase-1 are indicated in panels A,C,F. Only the active form (cleaved form) was quantified in panels B,D-E,G-H. For panels B,D-E,G-H, data shown represent mean ± SEM; n = 5 to 8 placentas from 3 dams for panel B or 3 independent repeat experiments for panels D-E,G-H. ∗P < .05; ns; ANOVA, Šídák multiple comparisons test. EV, EV-injected mice; EV+LMWH, EV-injected mice with LMWH injections.

Figure 3.

LMWH prevents platelet and EV-induced inflammasome activation via HBEGF-AKT signaling. (A-B) LMWH activates HBEGF-AKT signaling in placenta. Immunoblots of placenta from EV-injected pregnant mice (representative immunoblots [A]; bar graphs summarizing results [B]) showing restored HBEGF, pEGFR, pAKT, and pERK1/2 levels upon LMWH treatment. Placentas were assessed from C57BL/6 mice at day 12.5 p.c. after IV injection of mouse endothelial cell–derived procoagulant EVs at days 10.5 p.c. and 11.5 p.c. Control mice (C) were injected with the supernatant obtained after the last PBS wash during EV isolation. (C-D) Inhibition of HBEGF (HBEGFi, CRM197) or AKT (AKTi, LY294002) in vitro inhibits (representative immunoblots [C]; bar graphs summarizing results [D]) LMWH-mediated protective effects on platelet and EV-induced expression of NLRP3, cleaved IL-1β, and cleaved caspase-1. Cells were pretreated respective inhibitors before LMWH treatment. Control cells, C, were exposed to supernatant obtained after the last PBS wash during EV isolation. EV represents cells exposed to EV and platelets (PRP); EV+LMWH, cells pretreated with LMWH and exposed to EVs and platelets; EV+LMWH+HBEGFi, cells pretreated with HBEGFi and LMWH and exposed to EVs and platelets; EV+LMWH+AKTi, cells pretreated with AKTi and LMWH and exposed to EVs and platelets. Inactive (proform) and active (cleaved form) of IL-1β or caspase-1 are indicated in panels A,C. Only the active form (cleaved form) was quantified in panels B,D. Data shown represent mean ± SEM from 6 to 8 placentas from 3 dams for panels A-B or 3 independent repeat experiments for panels C-D. ∗P < .05, ns; ANOVA; Šídák multiple comparisons test. EV, EV-injected mice; EV+LMWH, EV-injected mice with LMWH injections.

Figure 4.

LMWH prevents inflammasome activation in human PE placental explants. (A-B) Human explants from term placentas obtained from patients with PE exposed to LMWH ex vivo (PE+LMWH) and healthy controls (Cs; non-PE) showed an inhibition of NLRP3, cleaved IL-1β and cleaved caspase-1 and an increase in HBEGF-AKT signaling (representative immunoblots [A]; bar graphs summarizing results [B]). Control explants, C, were obtained from healthy term placenta and exposed to PBS. Data shown represent mean ± SEM from 3 different patients. ∗P < .05, ns, ANOVA, Šídák multiple comparisons test.

Figure 5.

Proposed mechanism by which LMWH promotes HBEGF-AKT signaling and prevents ensuing placental thromboinflammation, proliferation, and differentiation. On 1 hand, LMWH prevents platelet activation. On the other, LMWH activates HBEGF signaling, thereby activating AKT signaling and preventing inflammasome activation. Cl., cleaved.