Rigidification of the autolysis loop enhances Na(+) binding to thrombin

Abstract

Binding of Na(+) to thrombin ensures high activity toward physiological substrates and optimizes the procoagulant and prothrombotic roles of the enzyme in vivo. Under physiological conditions of pH and temperature, the binding affinity of Na(+) is weak due to large heat capacity and enthalpy changes associated with binding, and the K(d)=80 mM ensures only 64% saturation of the site at the concentration of Na(+) in the blood (140 mM). Residues controlling Na(+) binding and activation have been identified. Yet, attempts to improve the interaction of Na(+) with thrombin and possibly increase catalytic activity under physiological conditions have so far been unsuccessful. Here we report how replacement of the flexible autolysis loop of human thrombin with the homologous rigid domain of the murine enzyme results in a drastic (up to 10-fold) increase in Na(+) affinity and a significant improvement in the catalytic activity of the enzyme. Rigidification of the autolysis loop abolishes the heat capacity change associated with Na(+) binding observed in the wild-type and also increases the stability of thrombin. These findings have general relevance to protein engineering studies of clotting proteases and trypsin-like enzymes.

Figure 1. Role of the autolysis loop on thrombin stability

The value of [urea]_1/2 necessary to unfold 50% of the enzyme in ChCl is plotted vs the analogous value in NaCl. The stabilizing effect of Na⁺ on wild-type and chimera is evident from the plot, and so is the much higher stability of murine thrombin compared to the human enzyme and its lack of response to Na⁺. The values for chymotrypsin and trypsin are also shown for comparison. Analysis of denaturation curves according to eq 1 in the text gives the following best-fit values for [urea]_1/2: (NaCl) 3.11 M (human thrombin), 4.25 M (chimera), 4.58 (murine thrombin), 4.71 M (chymotrypsin), 4.62 M (trypsin); (ChCl) 2.56 M (human thrombin), 2.95 M (chimera), 4.61 (murine thrombin), 4.70 M (chymotrypsin), 4.63 M (trypsin). Errors are 1–2% of the values.

Figure 2. Binding of DAPA monitored by fluorescence quenching

Experimental conditions are: 50 nM wild-type (open symbols) and 20 nM chimera (closed symbols), 10 mM Bis-Tris Propane, 0.1% PEG8000, pH 7.4 at 37 °C in the presence of 145 mM NaCl (circles) or ChCl (squares). Upon complex formation the total fluorescence decreases up to 30–35%. Best-fit parameters values for the affinity of DAPA are: (wild-type) K_d=79±5 nM (NaCl), K_d,=350±10 nM (ChCl); (chimera) K_d=44±2 nM (NaCl), K_d=89±2 nM (ChCl).

Figure 3. Kinetic traces of Na⁺ binding to thrombin wild-type (A) and chimera (B)

In both cases, Na⁺ binding obeys a two-step mechanism, with a fast-phase completed within the dead time (< 0.5 ms) of the spectrometer, followed by a single-exponential slow phase. Experimental conditions are: 50 nM enzyme, 50 mM Tris, 0.1% PEG8000, pH 8.0 at 15 °C. (C) Values of k_obs for the slow phase of fluorescence increase for thrombin wild-type (open circles) and chimera (closed circles). Continuous line were drawn according to eq 3 in the text with best-fit parameters values: (wild-type) k_r=91±2 s⁻¹, k_−r=83±2 s⁻¹, K_A=160±20 M⁻¹; (chimera) k_r=46±1 s⁻¹, k_−r=28±5 s⁻¹, K_A=350±30 M⁻¹.

Figure 4

(A) Na⁺ binding to thrombin wild-type (open circles) and chimera (close circles) monitored by fluorescence spectroscopy. Experimental conditions are: 10 mM Bis-Tris Propane, 0.1% PEG8000, pH 7.4 at 37 °C, I=600 mM. The equilibrium dissociation constant K_d for Na⁺ binding is 109±5 mM for wild-type and 20±1 mM for the chimera. (B) van't Hoff plot of Na⁺ binding to thrombin wild-type (open circles) and chimera (close circles). Shown are the values of K_d obtained from fluorescence titration over the temperature range 2–37 °C, under experimental conditions of 80 nM enzyme, 10 mM Bis-Tris Propane, 0.1% PEG8000, pH 7.4, I=600 mM kept constant with ChCl. Excitation and emission wavelengths were 280 and 341 nm, respectively. Continuous lines were drawn according to eq 2 in the text with best-fit parameters values: (wild-type) ΔH₀=−23.6±0.5 kcal/mol, ΔS₀=−71±4 cal/mol/K, ΔC_p=−860±60 cal/mol/K; (chimera) ΔH₀=−17.4±0.5 kcal/mol, ΔS₀=−49±2 cal/mol/K, ΔC_p=0. (C) Effect of ionic strength on the binding of Na⁺ to thrombin wild-type (open circles) and chimera (close circles). Experimental conditions are: 50 mM Tris, 0.1% PEG8000, pH 7.4 at 37 °C. The ionic strength was changed by addition of ChCl. Continuous lines were drawn according to the equation −log K_d=A₀+Γlog[NaCl] (93), with best-fit parameter values: (wild-type) A₀=0.96±0.02, Γ=−0.19±0.06; (chimera) A₀=1.69±0.04, Γ=−0.015±0.005.

Figure 5. Effect of Na⁺ on the hydrolysis of physiological substrates

Shown is the dependence of k_cat/K_m for the cleavage of fibrinogen (circles) and protein C (squares) in the presence of 5 mM Ca²⁺ and 100 nM thrombomodulin for thrombin wild-type (open circles) and chimera (closed circles). Experimental conditions are: 20 mM Tris, 0.1% PEG8000, pH 7.4 at 37 °C. The ionic strength was kept constant at 145 mM with ChCl. Continuous lines were drawn according to the linkage equation $k_{\mathit{cat}}∕K_m=s=\frac{s_0+s_1{K}_{\text{app}}\left[{\text{Na}}^{+}\right]}{1+K_{\text{app}}\left[{\text{Na}}^{+}\right]}$, where s₀ and s₁ are the values of k_cat/K_m at [Na⁺]=0 M and under saturating conditions, with best-fit parameter values: (wild-type, fibrinogen): s₀=1.5±0.1 μM⁻¹s⁻¹, s₁=21±2 μM⁻¹s⁻¹, K_app=12±1 M⁻¹; (chimera, fibrinogen) s₀=4.5±0.2 μM⁻¹s⁻¹, s₁=10±1 μM⁻¹s⁻¹, K_app=50±2 M⁻¹; (wild-type, protein C) s₀=0.32±0.02 μM⁻¹s⁻¹, s₁=0.20±0.01 μM⁻¹s⁻¹, K_app=12±1 M^−1;; (chimera, protein C) s₀=0.29±0.01 μM⁻¹s⁻¹, s₁=0.21±0.03 μM⁻¹s⁻¹, K_app=50±10 M⁻¹.

Figure 6. Crystal structure of the chimera

(A) Superposition of human wild-type thrombin (1SG8, cyan) and chimera (3R3G, gold) (rmsd=0.422 Å). Residues of the catalytic triad (H57, D102, S195) and Trp residues reporting the E*-E-E:Na⁺ transitions (W215, W141, W148) (90) are rendered as sticks. Relevant domains of the enzyme are noted. The bound Na⁺ is rendered as a sphere. (B) The coodination shell of Na⁺ in the chimera is similar to that of wild-type and involves four water molecules and the backbone O atoms of R221a and K224. H-bonding distances are noted. The electron density 2F₀-F_c map (green) is countered at 1 σ. (C) The autolysis loop of the chimera is stabilized by the hydrophobic cluster composed of W148, I149c and I150, the H-bond between N149b and N149d and a salt bridge between R145 and E18 that anchors the N-terminus of the heavy chain. The electron density 2F₀-F_c map (green) is countered at 0.7 σ.

Figure 7. Flexibility of the autolysis loop probed by limited proteolysis

Thrombin wild-type (A) or chimera (B) were incubated with chymotrypsin using an ezyme:substrate ratio of 1:150 (w/w) under experimental conditions of 20 mM Tris, 200 mM ChCl, pH 7.4. Samples were taken at 0, 5, 10, 20, 40, 60 and 120 min and loaded into a 4–12% gradient gel under reducing conditions. HC is the heavy chain, F1 and F2 are the cleavage products generated by chymotrypsin, encompassing the sequence 36–148 (17 kDa) and 149–259 (10 kDa), respectively. F3* is a secondary product of the reaction that is further proteolyzed by chymotrypsin as a function of time.

SCHEME 1