Self-anticoagulant sponge for whole blood auto-transfusion and its mechanism of coagulation factor inactivation

Abstract

Clinical use of intraoperative auto-transfusion requires the removal of platelets and plasma proteins due to pump-based suction and water-soluble anticoagulant administration, which causes dilutional coagulopathy. Herein, we develop a carboxylated and sulfonated heparin-mimetic polymer-modified sponge with spontaneous blood adsorption and instantaneous anticoagulation. We find that intrinsic coagulation factors, especially XI, are inactivated by adsorption to the sponge surface, while inactivation of thrombin in the sponge-treated plasma effectively inhibits the common coagulation pathway. We show whole blood auto-transfusion in trauma-induced hemorrhage, benefiting from the multiple inhibitory effects of the sponge on coagulation enzymes and calcium depletion. We demonstrate that the transfusion of collected blood favors faster recovery of hemostasis compared to traditional heparinized blood in a rabbit model. Our work not only develops a safe and convenient approach for whole blood auto-transfusion, but also provides the mechanism of action of self-anticoagulant heparin-mimetic polymer-modified surfaces.

Fig. 1. Design and characterization of the sponges for instantaneous blood auto-transfusion.

a Schematic diagram of the preparation of the HMP-modified sponges. The inset SEM images show the rough surface morphology after modification. Scale bars, 10 μm. b Pore-size distributions of MS, MS@D and MS@D-HMP. The experiments were performed independently in duplicate, with similar results obtained. c Permeating behaviors of water (20 μL droplet) on the surface of MS, MS@D and MS@D-HMP. Scale bars, 2 mm. The experiments were performed independently in duplicate, with similar results obtained. d Quantitative evaluation of water (left) and blood (right) sorption into MS@D-HMP. The measured weight of water (left) and blood (right) absorbed into MS@D-HMP per unit contact area (m_s) versus the square root of sorption time (t^1/2). The experiments were performed independently in duplicate, with similar results obtained. e Zeta potentials of MS, MS@D and MS@D-HMP (n = 6 independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). f Compressive stress-strain curves of MS@D-HMP after loading-unloading cycles. The experiments were performed independently in duplicate, with similar results obtained. Source data are provided as a Source Data file.

Fig. 2. Coagulation inhibition behaviors of the anticoagulant sponges in vitro.

a Concentration-dependent prolongation of clotting times (aPTT, PT and TT) for the citrate-anticoagulated PPP after incubation with MS@D-HMP for 30 min (n = 3 biologically independent samples, mean ± SD). b aPTT for the citrate-anticoagulated PPP (100 μL) after incubation with MS@D-HMP (10 mg) for different time intervals (n = 3 biologically independent samples, mean ± SD). c Concentration-dependent prolongation of PRTs for the citrate-anticoagulated PPP after incubation with MS@D-HMP (n = 3 biologically independent samples, mean ± SD). PPP (300 μL) was re-calcified with CaCl₂ solution (final Ca²⁺ concentration at 12.5 mM), and then incubated with MS@D-HMP. The treated PPP was taken out immediately (within 5 s), and monitored for clotting. d Photographs of recalcification of fresh plasma (upper) and MS@D-HMP-treated plasma (lower) in the glass vial. Scale bars, 10 mm. e Schematic diagrams and photographs of the whole blood (without anticoagulant) after incubation with MS@D-HMP in vitro (created with BioRender.com). Scale bars, 10 mm. f TEG traces for pristine whole blood and blood after incubation with MS@D-HMP. g Generated concentrations of TAT in the citrate-anticoagulated whole blood after incubation with MS, MS@D, and MS@D-HMP for 30 min. Positive control (recalcified blood, final Ca²⁺ concentration at 10 mM) and negative control (citrate-anticoagulated blood) are also shown (n = 3 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). Source data are provided as a Source Data file.

Fig. 3. Mechanisms of coagulation factor inactivation of the sponges.

a Activities of FVIII, FIX, FXI and FXII for the normal citrate-anticoagulated PPP after incubation with MS@D-HMP at different concentrations for 30 min. The activities were determined based on aPTT by mixing incubated plasma with corresponding factor deficient plasma (n = 5 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). b Abundance of coagulation-related proteins in the normal plasma and MS@D-HMP-treated plasma. Citrate-anticoagulated PPP (300 μL) was incubated with MS@D-HMP (30 mg) for 30 min. Values are expressed in relative percentages based on total proteins. c Abundance of contact system-related proteins in normal PPP and MS@D-HMP-incubated plasma. Values are expressed in relative percentages based on total proteins. d Schematic of two-stage successive incubations combined with PBS treatment. In stage 1, tightly- and loosely-bound proteins were adsorbed on the sponge in 1st incubation with fresh plasma. Such sponge without washing was allowed for 2nd incubation with another fresh plasma. Then, the sponge undergoing stage 1 was washed by PBS to remove the loosely-bound proteins, and used in stage 2 with similar procedures as those of stage 1 (created with BioRender.com). e Equilibrium binding capacity and the corresponding inactivation efficiency of FXI by MS@D-HMP. FXI-deficient plasma (100 μL) supplemented with different amounts of FXI proteins was incubated with MS@D-HMP (6 mg) for 30 min (n = 3 biologically independent samples, mean ± SD). f aPTT values for the PPP after successive incubations with MS@D-HMP in stages 1 and 2 (n = 8 biologically independent samples, mean ± SD. Two-way ANOVA with Bonferroni post-hoc tests). g Abundance of coagulation-related proteins in fresh plasma, plasma after 1st incubation with MS@D-HMP in stages 1, and plasma after 1st incubation with MS@D-HMP in stages 2. Values are expressed in relative percentages based on total proteins. h Abundance of coagulation-related proteins in the tightly-bound protein layer on the surface of MS@D-HMP in stages 1 and 2. Values are expressed in relative percentages based on total proteins. Source data are provided as a Source Data file.

Fig. 4. Evaluation of inhibited coagulation pathways in the sponges-treated plasma.

a Schematic of exploring the procoagulant activity of FXIa in the MS@D-HMP-incubated plasma via fluorometric-based FIXa activity assay. b Quantitative monitoring of FIXa in the normal hirudin-anticoagulated plasma after incubation with MS@D-HMP for 10 min. Actin (aPTT reagent) was used as positive control and citrate was used to block the activation of FIX by FXIa (n = 4 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). c Generated concentrations of bradykinin in the citrate-anticoagulated plasma after incubation with MS@D-HMP for 20 min. Plasma without treatment and plasma treated with glass power (1 mg/100 μL plasma) for 20 min were used as negative and positive controls, respectively (n = 6 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). Evaluation of FXIIa activity in normal plasma (d) and FXII-deficient plasma (e) after incubation with MS@D-HMP for 15 min using hydrolysis of chromogenic substrate S-2302 at an absorbance λ = 405 nm. Positive control (actin, aPTT reagent) was shown in (d) (n = 6 biologically independent samples for (d), mean ± SD. One-way ANOVA with Bonferroni post-hoc tests for (d). n = 3 biologically independent samples for (e), mean ± SD. Unpaired, two-tailed student’s t-test for (e)). f Quantitative monitoring of FIXa in the FXII-deficient plasma after incubation with MS@D-HMP for 10 min (n = 3 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). g Activities of FVIII, FIX and FXI for the FXII-deficient plasma after incubation with MS@D-HMP for 30 min (n = 3 biologically independent samples, mean ± SD. Unpaired, two-tailed student’s t-test). h Schematic of initiated and inhibited coagulation cascade by traditional negatively charged surface and MS@D-HMP, respectively (created with BioRender.com). Source data are provided as a Source Data file.

Fig. 5. Biocompatibility evaluation of the anticoagulant sponges in vitro.

a Blood cell count for pristine blood and blood after incubation with MS@D-HMP (n = 6 biologically independent samples, mean ± SD. Unpaired, two-tailed student’s t-test). b Differential white blood cell count (DIFF) scatter (forward (FS), side (SS)) charts of pristine blood and blood after incubation with MS@D-HMP. Monocyte (pink), lymphocyte (green), neutrophil (cyan), eosinophil (red). c Hemolysis ratio of the blood after incubation with MS@D-HMP (n = 3 biologically independent samples, mean ± SD. Unpaired, two-tailed student’s t-test). d Quantitative evaluation of adhered platelets on the sponges by LDH assay (n = 8 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). e Generation of PF4 in the blood after incubation with sponges (n = 3 biologically independent samples, mean ± SD. Kruskal–Wallis test with Dunn’s multiple comparisons). f Platelet activation in PRP based on the expression of platelet activation marker CD62p. Thrombin receptor activating peptide (TRAP, 0.1 mM) and control PRP were used as positive and negative control, respectively (n = 5 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). g Monocyte activation in whole blood based on the expression of monocyte activation marker CD11b. Lipopolysaccharide (10 ng/mL) and pristine blood were used as positive and negative control, respectively (n = 4 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). h Concentration of D-dimer in the citrate-anticoagulated plasma after incubation with sponge (n = 4 biologically independent samples for PPP and MS@D-HMP, n = 3 biologically independent samples for Positive, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). Generation of C3a (i) and C5a (j) in the blood after incubation with sponges. Cobra-venom factor (CVF, 25 μg/mL) and PBS were used as positive and negative controls, respectively (n = 6 biologically independent samples for Blood, MS, MS@D and MS@D-HMP, n = 4 biologically independent samples for Positive, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests for (i), Kruskal–Wallis test with Dunn’s multiple comparisons for (j)). Source data are provided as a Source Data file.

Fig. 6. Blood collection and anticoagulation using MS@D-HMP in rabbits.

a Schematic of whole blood collection and instantaneous anticoagulation using heparin or MS@D-HMP in rabbits (created with BioRender.com). b Images of WBCTs for heparin-treated group and MS@D-HMP-treated group. The collected blood (150 μL/well) was incubated in a 96-well polystyrene plate, and the un-clotted blood was washed away by PBS (150 μL) after different time intervals. The collected blood without anticoagulant (before) was shown as comparison. Scale bars, 2 mm. c Relative hemoglobin absorbance RHA(t) plot for the collected whole blood in heparin-treated group and MS@D-HMP-treated group at different time intervals. The free RBCs in the un-clotted blood was resuspended in 1 mL deionized water, and hemoglobin concentration was measured at an absorbance λ = 540 nm (n = 3 biologically independent samples, mean ± SD. Two-way ANOVA with Geisser–Greenhouse correction and Bonferroni post-hoc tests). d aPTT, PT and TT values for heparin-treated group and MS@D-HMP-treated group (n = 3 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). e Activity of intrinsic coagulation factors measured by titration experiments in the dFVIII, dFIX, dFXI and dFXII PPP for heparin-treated group and MS@D-HMP-treated group (n = 3 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). f Change in the serum calcium concentrations after treatment with heparin or MS@D-HMP (n = 3 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests). Hemolysis ratio (g), blood cell levels (h) and protein levels (i) of the collected blood in heparin-treated group and MS@D-HMP-treated group (n = 3 biologically independent samples, mean ± SD. One-way ANOVA with Bonferroni post-hoc tests for (g). Paired, two-tailed student’s t-test for (i) and RBC and Platelet in (h). Two-tailed Wilcoxon matched-pairs signed-tank test for WBC in (h)). Source data are provided as a Source Data file.

Fig. 7. Efficacy and safety of MS@D-HMP in a rabbit femoral artery hemorrhage model.

a Schematic of the whole blood auto-transfusion and experimental time line showing the sequence of events (created with BioRender.com). aPTT (b), TT (c), FDP (d) and D-dimer (e) for heparin-treated group and MS@D-HMP-treated group (n = 3 biologically independent samples, mean ± SD. Two-way ANOVA with Geisser–Greenhouse correction and Bonferroni post-hoc tests). f Blood cell count for heparin-treated group and MS@D-HMP-treated group (n = 3 biologically independent samples, mean ± SD). Analysis of liver function parameters (g, h), renal function parameters (i), serum complement and immunoglobulins levels (j), blood lipids levels (k) and serum electrolyte levels (l) of the rabbits in heparin-treated group and MS@D-HMP-treated group: g total protein (TP; g per litre); albumin (ALB; g per litre); globulin (GLB; g per litre); and albumin globulin ratio (A/G); h alkaline phosphatase (ALP; units per litre); alanine aminotransferase (ALT; units per litre); aspartate aminotransferase (AST; units per litre); gamma-glutamyl transferase (GGT; units per litre); direct bilirubin (DBIL; μmol per litre); total bilirubin (TBIL; μmol per litre); i urea (UREA; mmol per litre); creatinine (CREA; μmol per litre); and uric acid (UA; μmol per litre); j complement component 4 (C4; g l⁻¹); complement component 3 (C3; g l⁻¹); immunoglobulin A (IGA; g l⁻¹); immunoglobulin G (IGG; g l⁻¹) and immunoglobulin M (IGM; g l⁻¹). k cholesterol (CHOL; mmol per litre); triglyceride (TG; mmol per litre); high-density lipoprotein (HDL; mmol per litre); low-density lipoprotein (LDL; mmol per litre); l potassium, sodium, chlorine, phosphorus and calcium (quantity in mmol per litre) (n = 3 biologically independent samples, mean ± SD). Source data are provided as a Source Data file.