Human Kunitz-type protease inhibitor engineered for enhanced matrix retention extends longevity of fibrin biomaterials

Abstract

Aprotinin is a broad-spectrum serine protease inhibitor used in the clinic as an anti-fibrinolytic agent in fibrin-based tissue sealants. However, upon re-exposure, some patients suffer from hypersensitivity immune reactions likely related to the bovine origin of aprotinin. Here, we aimed to develop a human-derived substitute to aprotinin. Based on sequence homology analyses, we identified the Kunitz-type protease inhibitor (KPI) domain of human amyloid-β A4 precursor protein as being a potential candidate. While KPI has a lower intrinsic anti-fibrinolytic activity than aprotinin, we reasoned that its efficacy is additionally limited by its fast release from fibrin material, just as aprotinin's is. Thus, we engineered KPI variants for controlled retention in fibrin biomaterials, using either covalent binding through incorporation of a substrate for the coagulation transglutaminase Factor XIIIa or through engineering of extracellular matrix protein super-affinity domains for sequestration into fibrin. We showed that both engineered KPI variants significantly slowed plasmin-mediated fibrinolysis in vitro, outperforming aprotinin. In vivo, our best engineered KPI variant (incorporating the transglutaminase substrate) extended fibrin matrix longevity by 50%, at a dose at which aprotinin did not show efficacy, thus qualifying it as a competitive substitute of aprotinin in fibrin sealants.

Keywords: Aprotinin; Fibrin biomaterial; Human Kunitz-type protease inhibitor; Plasmin; Protein engineering.

Figure 1. Sequence and structure comparison of the protease inhibitors bovine-derived aprotinin and human-derived KPI

(A) Alignment of aprotinin and KPI sequences using CLUSTAL Omega (1.2.3) multiple sequence alignment program (European Bioinformatics Institute) demonstrates high degree of homology. Identical amino-acids (*) are highlighted in gray, and conserved substitutions (:) and semi-conserved substitutions (.) are denoted. (B) Ribbon-view structures of aprotinin (PDB database: 6PTI [18]), KPI (PDB database: 1AAP [14]), and their structural alignment (USCF Chimera software, University of California, USA) emphasize the structural similarity of the two protease inhibitors.

Figure 2. Design and characterization of wild-type and engineered KPI variants

(A) KPI was fused to C-terminal fibrin-targeting domains α₂PI_1-8 or PlGF-2_123-144A. (B) KPI variant size and purity were assessed by SDS- PAGE. KPI, KPI-α₂PI_1-8 and KPI-PlGF-2_123-144A migrated at sizes ranging from 12 to 15 kDa. (C) Determination of Michaelis constant (K_m) of plasmin and N-succinyl-Ala-Phe-Lys-7-amido-4-methylcoumarin acetate salt plasmin substrate (n=6, mean ± SEM). (D) Bioactivity of KPI variants characterized via a plasmin inhibition assay (n=4, mean ± SEM). (E) Inhibition constant K_i were calculated from IC₅₀ values obtained by data interpolation using a sigmoidal curve fitting (n=4, mean ± SEM). All KPI variants have similar inhibitory activities on plasmin, with K_is ranging from 29 to 46 nM. Aprotinin K_i was determined experimentally and corresponded to published data [15].

Figure 3. Engineered KPI-α₂PI_1-8 and KPI-PlGF-2_123-144A slow the rate of fibrinolysis in vitro

Fluorescent fibrin gels containing 15 μM of KPI variants or aprotinin were incubated in presence of 2.5 nM of plasmin. (A) Representative images of fluorescent fibrin gel degradation over time. (B) Gel degradation was monitored by quantification of fluorescence decay. Both KPI-α₂PI_1-8 and KPI-PlGF-2_123-144A inhibit plasmin and prolong the longevity of fibrin matrices compared to matrices containing aprotinin or no inhibitor. KPI-α₂PI_1-8 offered the best protection against fibrinolysis (n=4 gels, mean±SEM, ANOVA with Bonferroni post-hoc test).

Figure 4. Optimization of fibrinogen and inhibitor concentrations to enhance fibrin longevity in vitro

Three concentrations of fibrin and 5 concentrations of inhibitors were tested to compare the degradation time of fluorescently-labeled fibrin matrices in the presence of 2.5 nM plasmin. Gel degradation was quantified daily by fluorescence measurement and the time of complete degradation was defined by the absence of fluorescence. (A) Fibrin matrix longevity during plasmin exposure, as a function of fibrin and inhibitor concentrations. Tables indicate days at which gels were fully degraded. (B) Comparison of the inhibitors, at various concentrations, on the degradation time of 12 mg/mL fibrin matrices. (C) Influence of fibrin concentration on the degradation time of matrices supplemented with KPI-α₂PI_1-8.

Figure 5. KPI-α₂PI_1-8 extends fibrin matrix longevity in vivo when implanted subcutaneously

Fluorescent fibrin matrices containing 40 μM of KPI variants or aprotinin were implanted subcutaneously in mice. (A) In vivo gel degradation was quantified through fluorescence measurements every 6 days. KPI-α₂PI_1-8 significantly increased fibrin longevity compared to the other inhibitors (n≥7 for KPI-α₂PI_1-8 and ‘no inhibitor’ groups, n≥4 in remaining groups, mean ± SEM, Kruskal-Wallis with Dunns post-hoc test). (B) Apparent gel degradation rate at day 18 as a function of the different inhibitors (n≥7 for KPI-α₂PI_1-8 and ‘no inhibitor’ groups, n≥4 in other groups, mean ± SEM, Kruskal-Wallis with Dunns post-hoc test). (C) Representative images of fluorescence decay measurements of subcutaneous fibrin implants loaded with the different inhibitors over time. (Blue circles: region-of-interests of 1 cm in diameter). (D) Immunohistochemistry of subcutaneously implanted fibrin matrices containing the different KPI inhibitor variants at 15 days post-implantation (DAPI/cell nuclei in blue, fibrinogen in red, human KPI in green). Both KPI-α₂PI_1-8 and KPI-PlGF-2_123-144A remain sequestered into fibrin matrices compared to wild-type KPI. Scale bar = 100 μm.