Lipid nanoparticles and siRNA targeting plasminogen provide lasting inhibition of fibrinolysis in mouse and dog models of hemophilia A

Abstract

Antifibrinolytic drugs are used extensively for on-demand treatment of severe acute bleeding. Controlling fibrinolysis may also be an effective strategy to prevent or lessen chronic recurring bleeding in bleeding disorders such as hemophilia A (HA), but current antifibrinolytics have unfavorable pharmacokinetic profiles. Here, we developed a long-lasting antifibrinolytic using small interfering RNA (siRNA) targeting plasminogen packaged in clinically used lipid nanoparticles (LNPs) and tested it to determine whether reducing plasmin activity in animal models of HA could decrease bleeding frequency and severity. Treatment with the siRNA-carrying LNPs reduced circulating plasminogen and suppressed fibrinolysis in wild-type and HA mice and dogs. In HA mice, hemostatic efficacy depended on the injury model; plasminogen knockdown improved hemostasis after a saphenous vein injury but not tail vein transection injury, suggesting that saphenous vein injury is a murine bleeding model sensitive to the contribution of fibrinolysis. In dogs with HA, LNPs carrying siRNA targeting plasminogen were as effective at stabilizing clots as tranexamic acid, a clinical antifibrinolytic, and in a pilot study of two dogs with HA, the incidence of spontaneous or excess bleeding was reduced during 4 months of prolonged knockdown. Collectively, these data demonstrate that long-acting antifibrinolytic therapy can be achieved and that it provides hemostatic benefit in animal models of HA.

Fig. 1.. Plasminogen is depleted for weeks after a single administration of siPlg in mice and dogs.

(A) Mice were injected with a single dose of siLuc (gray) or siPlg (teal blue). Plg mRNA in liver tissue was measured using qPCR, normalized against the housekeeping gene, Ppia. (B) Representative Western blot against plasminogen, where each lane contains the plasma from an individual mouse in either treatment group. The triangular marker indicates the expected molecular weight of plasminogen (92 kDa). (C) Plasminogen protein in plasma measured in a portion of the mice enrolled at each time point using densitometry, normalized to a loading control, and graphed relative to untreated mice. (D) Plasminogen protein in plasma quantified after administration of siPlg to dogs at 0.027 (purple, n = 3), 0.054 (green, n = 2), or 0.54 (teal blue, n = 1) mg siRNA/kg body weight. (E) Representative Western blot against plasminogen, where each lane contains plasma collected from a single WT dog at a different time point before or after siPlg administration at 0.054 mg/kg. Data are presented as mean ± SEM and were analyzed by two-tailed unpaired Student’s t test (A) or by two-way ANOVA (C); *P < 0.05.

Fig. 2.. Long-term knockdown of plasminogen does not lead to pathologies associated with complete plasminogen deficiency.

(A) Plasma plasminogen (left y axis) quantified during 9 months of repeat administration of siPlg. The weight of mice (right y axis) was tracked over this period during administration of siLuc (gray) and siPlg (orange). (B) Periodontal bone loss measured by the distance between the cementoenamel junction and the alveolar bone crest (orange lines) in samples stained with methylene blue from mice administered siLuc (n = 6) or siPlg (n = 7) for 9 months compared to untreated (Unt WT) age-matched WT controls (purple line). Scale bar, 2 mm. (C to F) Inflammatory markers IL-6 (C and D) and TNF-α (E and F) were measured using qPCR in liver tissue or enzyme-linked immunosorbent assay in plasma from siLuc-treated (n = 8) or siPlg-treated (n = 10) mice compared to untreated (Unt) controls (purple line). (G) Immunohistochemistry against fibrin(ogen) in microscopy images of liver tissue of siPlg-treated mice showed inflammatory infiltrates but no fibrinous lesions. Scale bar, 100 μm. (H) Immunohistochemistry against fibrin(ogen) in clots formed ex vivo in whole blood from mice treated with siLuc (left) or siPlg (right). Scale bar, 25 μm. For all graphs, values represent mean ± SEM, ns indicates difference not statistically significant (P > 0.05), analyzed by two-tailed unpaired Student’s t test.

Fig. 3.. Plasminogen knockdown enhances hemostasis after SVP but not TVT in a mouse model of HA.

(A) Schematic showing SVP and TVT injury models. (B) Plasminogen in HA mice at the time of TVT bleed. (C and D) Blood loss in microliters of blood per gram of body weight (C) and bleeding time (D) after TVT in WT, untreated HA, and treated HA mice (n = 5 to 7). (E and F) Blood loss in microliters of blood per gram of body weight (E) and bleeding time (F) to end of observation period at 40 min (gray dashed line) after SVP (n = 7 to 10). Values represent mean ± SEM; *P < 0.05, **P < 0.01, and ns indicates not significant (P > 0.05), analyzed by one-way ANOVA.

Fig. 4.. Plasminogen knockdown stabilizes clots ex vivo in two dogs with HA.

Samples were collected from two HA dogs at baseline (green), within 4 hours of TXA (10 mg/kg iv, orange), within 3 weeks of siPlg administration without TXA (teal), or after administration of both siPlg and TXA (purple). (A) Representative TEG curve tracings using plasma from two HA dogs (top and bottom graphs) with added tPA. (B) Clot lysis time within 180-min monitoring period, values represent mean, and error bars represent ± SEM. (C) Thrombin generation in plasma collected weekly from two HA dogs or pooled plasma from WT dogs (WT, gray). (D) Correlation between plasminogen protein in plasma and thrombin generation peak (open and closed markers distinguish the two siPlg-treated dogs). (E) Plasmin generation in plasma collected weekly from two HA dogs or pooled plasma from WT dogs. (F) Correlation between plasma plasminogen and plasmin generation peak.

Fig. 5.. Plasminogen knockdown decreases bleeding events in two dogs with HA over a 4-month period.

HA dogs were administered five doses of siPlg over 15 weeks. Vertical dashed lines indicate days of an administration. (A) Plasminogen in plasma quantified using densitometry of Western blots. Values represent mean ± SEM. (B) Annualized bleeding rate extrapolated from 10 months before treatment and the 4-month period of siPlg treatment in two HA dogs (circle and diamond each indicate an individual dog).