The Effect of IL-6-Primed Platelets on ADAMTS13- Mediated Clearance of Platelet-Bearing ULVWF and Its Mechanism


 Inflammation is an essential contributing factor in the development of thrombosis. Using a microfluidic flow chamber, we investigated how the proinflammatory cytokine interleukin 6 (IL-6) affects the cleavage of platelet-bearing ultra-large VWF (ULVWF) by plasma ADAMTS13. We found that IL-6-treated platelets perfused at arteriolar shear stress significantly enhanced the ULVWF-platelet complex formation on activated endothelial cells and suppressed their clearance by ADAMTS13 under flow conditions. We also detected the phosphorylation of the serine/threonine kinase Akt and extracellular signal-regulated protein kinases 1 and 2 (ERK1/2) in platelets treated with IL-6. Treatment of IL-6-primed platelets with either the phosphoinositol-3 kinase (PI3K) inhibitor LY294002 or the mitogen-activated protein kinase kinase (MEK) inhibitor U0126 reduced the ULVWF-platelet complex formation and restored the clearance of the complex by plasma ADAMTS13, compared to IL-6-primed platelets. Furthermore, IL-6 enhanced the phosphorylation of the intracellular adaptor molecule 14-3-3ζ, which regulates VWF binding to the glycoprotein (GP) Ib-IX complex. The 14-3-3 antagonist R18 significantly increased ADAMTS-13 cleavage of ULVWF strings with adherent IL-6-treated platelets. These findings indicate that IL-6 related intracellular signals of platelet is involved in regulating ULVWF-platelet binding and ULVWF cleavage by ADAMTS13.


Introduction
VWF enables platelets to adhere to the subendothelial matrix that can resist the pulling force of blood ow, which can be as high as or even higher than shear stress found in arterioles [1,2]. While platelets initiate hemostasis when they bind the subendothelial matrix, they also bind the strings of VWF multimers freshly released from activated endothelial cells and anchored to the endothelial surface.
ADAMTS13 effectively cleaves and thus remove these platelet-bearing VWF strings [3], especially the hyperactive ULVWF from the endothelial surface because platelets on VWF exert a stretching force on VWF to expose the sessile bond in the A2 domain for ADAMTS-13. The importance of this enzymatic activity of ADAMTS13 is well-demonstrated in patients with thrombotic thrombocytopenic purpura, whose ADAMTS13 activity is either suppressed by autoantibodies to the metalloprotease or de cient due to genetic mutations of the ADAMTS13 gene. If the cleavage is delayed, more platelets could be tethered to the VWF strings before cleavage, thus releasing platelet-VWF aggregates that could cause systemic arterial thrombosis.
The clearance of platelet-bearing ULVWF strings can also be affected by in ammation. We have previously shown that IL-6, a universal in ammatory cytokine, signi cantly suppressed the ADAMTS13 activity for cleaving platelet-bearing ULVWF strings [4]. It was speculated that IL-6 binds and interferes with ADAMTS13 by steric hindrance on the VWF-ADAMTS13 interaction. However, the fact that IL-6 regulates platelet function [5] also raises the possibility that platelets primed for activation by IL-6 affects the cleavage of ULVWF bearing those platelets. Platelets can also regulate VWF cleavage by providing the surface on which VWF and ADAMTS13 interact to form a proteolytic complex [6] or by VWF undergoing conformational changes upon binding to GP Ibα to facilitate or promote the cleavage.
In addition to their hemostatic activity, platelets also play an active role in in ammation because they contain a plethora of bioactive molecules in their granules [7] and release in ammatory mediators upon activation. Platelets express cytokine receptors that participate in in ammation and can be activated by cytokine bound to these receptors [8]. For example, 14-3-3, which is a GP Ibα-associated adaptor [9], is activated by in ammatory cytokines [10,11]. In ammation and thrombosis are therefore considered the tightly interrelated processes that cross-regulate each other [12,13]. In this regard, understanding the effects of in ammatory cytokines on VWF-bound platelets and their clearance from the vessel wall by ADAMTS13 can provide new insight into the pathophysiology of thrombosis and in ammation and identify new and target therapies.

Ethics statement
Healthy adult volunteers were recruited for blood collection with written informed consent. The study was approved by the Institutional Review Board of Yonsei University, College of Medicine and performed in accordance with the Declaration of Helsinki.

Platelet rich plasma and suspension preparation
To prepare platelet-rich plasma (PRP), we modi ed the previously published procedures [24]. Brie y, blood samples from healthy donors were drawn into acid citrate dextrose (ACD) tubes. PRP was obtained by

Fabrication of a micro uidic chip
We fabricated a single straight-channel micro uidic chip using polydimethylsiloxane (PDMS) to monitor ULVWF-mediated platelet binding to the surface of activated endothelial cells. The workloads and time required for procedure were also considerably reduced by replacing ow chambers with micro uidic chips [25]. Brie y, a mold of a micro uidic chip was formed by photolithography, with the micro uidic channel measured 800 µm wide, 125 µm deep, and 28 mm long on a wafer. A 10:1 w/w mixture of PDMS base and curing agent (Sylgard 184, Dow Corning, Midland, MI, USA) was poured over the mold and cured in an oven at 65°C for 2 hours. The solid PDMS was separated from the mold, and two openings were cut at both ends of the micro uidic channel. The PDMS chip was adhered to a glass slide by a corona treater to complete the micro uidic chip and then sterilized in an autoclave.

Endothelial cell culture in micro uidic chip
The channel of the micro uidic chip was lled with 1% polyethyleneimine solution and incubated at room temperature for 10 minutes. The chip was washed with distilled water and treated with 0.1% glutaraldehyde for 30 minutes. After washing three times, the chip was coated with 0.3 mg/mL collagen (Sigma) for 20 minutes at 37°C. After coating, the channel was lled with endothelial cell culture medium (EGM-2 bullet kit, Lonza, Walkersville, MD, USA) for 30 minutes at 37°C. At this point, Human umbilical vein endothelial cells (HUVECs, Gibco, Grand Island, NY, USA) culture asks were detached by trypsin, centrifuged at 180g for 7 minutes, resuspended in cell culture media at a density of 5×10 6 cells/mL and then seeded in the microchip. HUVECs were cultured under static conditions for 24 hrs before experiments.

Perfusion of platelet suspensions in a micro uidic chip
Platelets are often used to visualize the ULVWF anchored on endothelial surface in a chamber slide setting. This in vitro technique is useful in studying the mechanism of thrombotic disease. HUVECs were grown in the micro uidic chip until full con uence and then activated by 150 µM histamine (Sigma) for 30 minutes at 37°C. The histamine-treated micro uidic chip was placed on an optical microscope stage. Platelet suspensions were treated with 100 ng/mL IL-6 (Sigma) or vehicle for 30 minutes at 37°C, washed once, and then resuspended in Tyrode's albumin buffer. In subsets of samples, platelets were also incubated with 25 µM LY294002 (Sigma), 10 µM U0126 (Cell Signaling Technology, Danvers, MA, USA), or 10 µM R18 (Tocris, Ellisville, MO, USA) for 10 minutes before perfusion. Microtubing connected to the injection pump was linked to the outlet of the micro uidic chip, and a platelet-inhaled microtubing was connected to the inlet. Pretreated platelets were perfused over con uent HUVECs through the chamber with continuous video recording. The ow rate was set at 80 µL/min, which generated 6.4 dynes/cm2 shear stress with 1 cp viscocity. The number of linear structures in which the ULVWF and the platelets were combined was counted. We also monitored with image recording the removal pattern of plateletbearing ULVWF strings after perfusing plasma for 4 minutes in a single eld of view (original magni cation 200×). In our previous study, the metalloprotease activity of ADAMTS13 in plasma was blocked by EDTA, which inhibited string cleavage [25].

Statistical analysis
Statistical analyses were performed using one-way analysis of variance (ANOVA) and Student's t-test to evaluate differences between groups. p<0.05 was considered statistically signi cant.

Results And Discussion
Formation of IL-6-treated platelet-ULVWF strings and clearance by ADAMTS13 IL-6, a representative in ammatory mediator, promotes hemostasis through several pathways including platelet aggregation [26,27]. IL-6-treated platelets are thrombosis-prone in that they are primed for activation by several platelet agonists at subthreshold concentrations [28][29][30]. We modi ed the owchamber experiments to a miniaturized micro uidic system to perform the experiments with a small volume of blood [25]. We examined IL-6-treated platelets for their in uence on the formation of platelet-bearing ULVWF strings on activated ECs under ow conditions, as compared to mock-treated platelets. As shown in Figure 1A and B, more platelet-bearing ULVWF strings were detected when IL-6-treated platelets were perfused than perfusing untreated platelets (mean ± standard deviation [SD]: 21.6±10.6 versus 16.2±6.5; 1.30±0.16-fold increase; P < 0.01).
(A) HUVECs cultured in a micro uidic channel were stimulated with histamine, and then perfused with platelets untreated (left panel; control) or treated with IL-6 (right panel). ULVWF-platelet formed the beadson-string appearance (arrow heads) the endothelial surface (original magni cation 200x). (B) Data summary from multiple experiments shows fold increase in the number of ULVWF strings, as compared to control (data presented as SEM, n=5, Student's t test, *P < 0.01).
(A) ULVWF-platelet strings formed during the perfusion of IL-6-treated or untreated platelets were cleaved by ADAMTS13 in plasma and washed away from the endothelial surface (▶ ULVWF-platelet strings initially formed, ▶ Strings showing decrease in length, ▷ Strings eliminated from endothelial surface).
The platelet IL-6 signaling pathway in clearance of platelet-bearing ULVWF by ADAMTS13 The results demonstrated that IL-6 primed platelets affect the formation and clearance of ULVWF-platelet strings. To explain this phenomenon, we hypothesized that there exists a pathway between the point of the IL-6 signal and VWF-GPIbα binding. It has been shown that the PI3K/Akt and MEK/ERK signal pathways play a role in VWF-GPIb-IX-dependent platelet aggregation [31][32][33]. Here, we investigated whether IL-6 activates the PI3K/Akt or MEK/ERK pathway in platelets [34,35] by examining the phosphorylation of Akt or ERK1/2. The IL-6 treatment induced a 1.54-fold (95% CI 1.48-1.60) and 1.63fold (95% CI 1.25-2.01) increases in the phosphorylation of Akt and ERK, respectively, as compared to the mock treatment (Fig. 3).
Western blot analysis was performed to evaluate Akt and ERK1/2 phosphorylation after the treatment of platelets with 100 ng/mL IL-6 for 30 minutes. β-actin was used as a loading control. Histograms show quantitative analysis of pAkt normalized to total AKT and pERK1/2 normalized to total ERK1/2. Graphs represent the mean±SEM of three separate experiments performed in triplicate, *P < 0.05, **P < 0.01 compared to the untreated group (C).
IL-6-treated or resting platelets were treated with or without LY294002 or U0126 and then perfused through activated ECs in the channel. Summarized data presented the related ratio of formed strings. Mean±SEM, n=3, *P < 0.05 compared to the control group, #P < 0.05 compared to the IL-6-treated group.
IL-6-treated or untreated platelets were perfused into the channel after treatment with or without LY294002 (A, C) or U0126 (B, D). ULVWF-platelet strings were cleaved by ADAMTS13 contained in plasma and washed away from the endothelial surface. ▶ ULVWF-platelet strings initially formed, ▶ Strings showing decrease in length, ▷ Strings eliminated from the endothelial surface. Summarized data show the clearance rate of ULVWF-platelet strings by ADAMTS13. Mean±SEM, n=3, * P < 0.05 versus the IL-6 only-treated group.
In contrast, inhibitors of JAK or STAT3 did not signi cant affect the cleavage of ULVWF strings bearing IL-6-treated platelets (data not shown). Together, these results suggest that IL activated PI3K/Akt and MEK/ERK signaling pathways regulated the formation and clearance of VWF-platelet strings formed on activated ECs.
Western blot analysis was performed to evaluate phosphorylation and protein levels of 14-3-3ζ after treatment of platelets with 100 ng/mL IL-6 for 30 minutes. β-Actin was used as a loading control.
IL-6-treated or untreated platelets were perfused into the channel after treatment with or without R18.
Many ULVWF-platelet strings were present on the endothelial surface. Summarized data were mean±SEM, n=4, * P < 0.05 compared to control.
(A) IL-6-treated or untreated platelets were perfused into the channel after treatment with or without R18. ULVWF-platelet strings were cleaved by ADAMTS13 contained in plasma and washed away from the endothelial surface. ▶ ULVWF-platelet strings initially formed, ▶ Strings showing decrease in length, ▷ Strings eliminated from the endothelial surface. (B) Summarized data showed the clearance rate of ULVWF-platelet strings by ADAMTS13. Mean±SEM, n=4, * P < 0.05 compared to the control group, # P < 0.05 compared to IL-6 alone-treated group.

Conclusions
The VWF-platelet complex anchored on the endothelial surface acts as the seed for intravascular thrombus. Its formation and clearance are affected by IL-6, a major in ammatory cytokine, but the exact mechanism is unclear yet. IL-6 primed platelets enhance the ULVWF-platelet complex formation. Also, IL-6 primed platelets are less e ciently cleared by plasma ADAMTS13. PI3K/Akt and MEK/ERK pathways seem to mediate the effect of IL-6 in regulating the formation and the clearance of the ULVWF-platelet complex. IL-6 increases the phosphorylation of 14-3-3ζ, and the inhibition of 14-3-3ζ signi cantly enhances the clearance of the ULVWF-platelet (IL-6-treated) complex.

Declarations
Declaration of interest None Funding Source Disclosure Figure 1 <p><strong>Effect of IL-6-induced platelets on the formation of ULVWF strings under ow conditions. </strong> </p><p> (A) HUVECs cultured in a micro uidic channel were stimulated with histamine, and then perfused with platelets untreated (left panel; control) or treated with IL-6 (right panel). ULVWFplatelet formed the beads-on-string appearance (arrow heads) the endothelial surface (original magni cation 200x). (B) Data summary from multiple experiments shows fold increase in the number of ULVWF strings, as compared to control (data presented as SEM, n=5, Student's t test, *P &lt; 0.01).</p> <p><strong>Activation of the Akt and ERK signaling pathway in IL-6-treated platelets.</strong> </p><p> Western blot analysis was performed to evaluate Akt and ERK1/2 phosphorylation after the treatment of platelets with 100 ng/mL IL-6 for 30 minutes. β-actin was used as a loading control. Histograms show quantitative analysis of pAkt normalized to total AKT and pERK1/2 normalized to total ERK1/2. Graphs represent the mean±SEM of three separate experiments performed in triplicate, *P &lt; 0.05, **P &lt; 0.01 compared to the untreated group (C).</p> Figure 4 <p><strong>Effect of PI3K and MEK inhibitors on platelet string formation under ow conditions. </strong></p><p> IL-6-treated or resting platelets were treated with or without LY294002 or U0126 and then perfused through activated ECs in the channel. Summarized data presented the related ratio of formed strings. Mean±SEM, n=3, *P &lt; 0.05 compared to the control group, #P &lt; 0.05 compared to the IL-6-treated group.</p> Figure 5 <p><strong>The PI3K or MAPK/ERK pathways regulated the cleavage of ULVWF-platelet strings under ow conditions. </strong></p><p> IL-6-treated or untreated platelets were perfused into the channel after treatment with or without LY294002 (A, C) or U0126 (B, D). ULVWF-platelet strings were cleaved by ADAMTS13 contained in plasma and washed away from the endothelial surface. ▶ ULVWF-platelet strings initially formed, ▶ Strings showing decrease in length, ▷ Strings eliminated from the endothelial surface. Summarized data show the clearance rate of ULVWF-platelet strings by ADAMTS13. Mean±SEM, n=3, <sup>*</sup><em>P</em> &lt; 0.05 versus the IL-6 only-treated group.</p> Figure 6 <p><strong>Phosphorylation of 14-3-3ζ in IL-6-treated platelets. </strong></p><p> Western blot analysis was performed to evaluate phosphorylation and protein levels of 14-3-3ζ after treatment of platelets with 100 ng/mL IL-6 for 30 minutes. b-Actin was used as a loading control. Histograms show quantitative analysis of p14-3-3ζ normalized to b-actin, 14-3-3ζ normalized to b-actin, and p14-3-3ζ normalized to total 14-3-3ζ. Graphs represent the mean±SEM of three separate experiments performed in triplicate, <sup>*</sup><em>P</em> &lt; 0.05 compared to the untreated group (C).</p> Figure 7 <p><strong>Effect of a 14-3-3 inhibitor on platelet string formation under ow conditions. </strong></p> <p> IL-6-treated or untreated platelets were perfused into the channel after treatment with or without R18.