Introduction: Soluble plasma fibronectin (Fn) with its inactive compact structure requires unfolding to assemble into active fibrils. Fibril formation of Fn is cell-mediated (Mao and Schwarzbauer, 2005) and depending on interactions of Fn with integrin receptors, through binding to αIIbβ3, α5β1, or αvβ3. Less is known about the contribution of biomechanical forces on the fibrillogenesis of Fn. Tension forces generated by cells via cytoskeleton could modulate Fn fibrillogenesis. The aim of this study was to investigate conformational changes of Fn, as induced by (1) platelet integrins, (2) cytoskeletal forces and/or (3) shear rates simulating venous or arterial flow conditions.
Methods: Human plasma Fn (100 μg/ml) was added to plates pre-coated with 100 μg/ml of Fn or collagen in the presence or absence of washed platelets (2.5 x 107/ml). Subsequently, the solutions were exposed to shear using a cone-plate rheometer (Haake Rheostress 1). For microscopic analysis (LSM 510, Carl Zeiss), Alexa flour 488-conjugated Fn was used. In parallel experiments, a N-terminal 70kDa fragment of Fn (70 μg/ml) was incubated with soluble Fn at room temperature for 20 min before exposure to shear. To examine the role of distinct platelet integrins on fibril formation of Fn, washed platelets were incubated with monoclonal antibodies LM609, P1D6, 10E5, or abciximab (10 μg/ml, each) for 30 min at room temperature to block αvβ3, α5β1, αIIbβ3, or both αIIbβ3 and αvβ3, prior to the addition of Fn (100 μg/ml) and subsequent exposure to shear. In parallel experiments, washed platelets were pre-incubated with actin-modifying reagents, jasplakinolide (1 μM) or cytochalasin D (10 μM). In all experiments, flow conditions were simulated by shear rates, stepwise increasing from 50 s-1 to 5000 s-1 within 5 min and subsequently decreasing from 5000 s-1 to 50 s-1 within 5 min. Viscosities (mPa's) of shear-exposed solutions were recorded over 10 min. To study the structure of Fn fibrils, solutions were examined by laser scanning microscopy after exposure to shear. To quantify the amount of fibril formation, deoxycholate solubility assays and densitometric analysis of Western blots were performed. Control experiments were conducted under static conditions.
Results: Microscopic analyses showed that exposing Fn solutions to shear resulted in fibril formation. Fn fibril diameter varied from 0.5 to 5 μm. Observed fibrils were linked with each other and varied in length (from 50 to 300 μm). Addition of washed platelets to Fn solution resulted in a higher intertwined matrix of fibrils. Treatment of Fn with the N-terminal 70 kDa fragment of Fn, which is known to inhibit Fn matrix assembly, blocked fibril formation of Fn. Western blotting and densitometric analyses revealed that, in the absence of washed platelets, fibril formation (calculated as the ratio of insoluble to soluble Fn) on plates coated with collagen were 2-fold higher than on Fn-immobilized plates (n = 4, p < 0.05). Addition of washed platelets to Fn solution (100 μg/ml) resulted in increases of 20- or 7-fold in fibril formation of Fn, generated by shear on Fn- and collagen-immobilized plates, respectively (p < 0.05, n = 3). In contrast, 10E5 or abxicimab blocking αIIbβ3, or both αIIbβ3 and αvβ3 caused a reduction by 82% or 74% in fibril formation of Fn (p < 0.05, n = 3 each), in comparison to samples without antibodies. Blocking α5β1 or αvβ3by P1D6 or LM609 only caused a reduction by 17% or 56% (p > 0.05, n = 3 each). Incubation of platelets with jasplakinolide, which stabilizes actin, caused an increase in fibril formation by 41%, as compared to samples with untreated platelets (p > 0.05, n = 3). In contrast, disruption of actin by cytochalasin D resulted in a decrease by 86% (p < 0.05, n = 3). Under static conditions, no fibril formation was detected.
Conclusions: Our results indicate that fibrillogenesis of Fn is modulated by shear conditions in a surface-dependent manner. Furthermore, formation of fibrils is induced by platelet integrins and actin cytoskeleton. Hereby, αIIbβ3 plays a predominant role, while α5β1 has a minor part among the three examined platelet integrins in terms of Fn fibril formation.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.