Clots Are Potent Triggers of Inflammatory Cell Gene Expression Indications for Timely Fibrinolysis

Basic Sciences

Robert A. Campbell, Adriana Vieira-de-Abreu, Jesse W. Rowley, Zechariah G. Franks, Bhanu Kanth Manne, Matthew T. Rondina, Larry W. Kraiss, Jennifer J. Majersik, Guy A. Zimmerman, Andrew S. Weyrich

Objective—Blood vessel wall damage often results in the formation of a fibrin clot that traps inflammatory cells, including monocytes. The effect of clot formation and subsequent lysis on the expression of monocyte-derived genes involved in the development and progression of ischemic stroke and other vascular diseases, however, is unknown. Determine whether clot formation and lysis regulates the expression of human monocyte-derived genes that modulate vascular diseases. Approach and Results—We performed next-generation RNA sequencing on monocytes extracted from whole blood clots

Downloaded from and using a purified plasma clot system. Numerous mRNAs were differentially expressed by monocytes embedded in clots compared with unclotted controls, and IL-8 (interleukin 8) and MCP-1 (monocyte chemoattractant protein-1) were among the upregulated transcripts in both models. Clotted plasma also increased expression of IL-8 and MCP-1, which far exceeded responses observed in lipopolysaccharide-stimulated monocytes. Upregulation of IL-8 and MCP-1 occurred in a thrombin- independent but fibrin-dependent manner. Fibrinolysis initiated shortly after plasma clot formation (ie, 1–2 hours) reduced

http://atvb.ahajournals.org/ the synthesis of IL-8 and MCP-1, whereas delayed fibrinolysis was far less effective. Consistent with these in vitro models, monocytes embedded in unresolved thrombi from patients undergoing thrombectomy stained positively for IL-8 and MCP-1. Conclusions—These findings demonstrate that clots are potent inducers of monocyte gene expression and that timely fibrinolysis attenuates inflammatory responses, specifically IL-8 and MCP-1. Dampening of inflammatory gene expression by timely clot lysis may contribute to the clinically proven efficacy of fibrinolytic drug treatment within hours of stroke onset. Visual Overview—An online visual overview is available for this article. (Arterioscler Thromb Vasc Biol. 2017;37: 1819-1827. DOI: 10.1161/ATVBAHA.117.309794.)

by guest on July 10, 2018 Key Words: fibrin◼ fibrinogen◼ inflammation◼ monocytes ◼ thrombin

oagulation contributes to the pathogenesis of many dis- suggests that clots are ineffective triggers of gene expression Cease processes, including sepsis, myocardial infarction, by monocytes, but this assumption has not been rigorously and ischemic stroke.1,2 On vessel injury, blood is exposed tested. Accordingly, we determined whether intact clots alter to tissue factor, resulting in the formation of thrombin that gene expression patterns in monocytes using human models stems blood loss by activating coagulation proteases and of whole blood and plasma clot formation. Our data dem- platelets.3–5 Thrombin also induces the conversion of fibrin- onstrate that clots induce marked changes in the monocyte ogen to fibrin, resulting in the formation of clots that seal transcriptome, including increased expression of IL-8 (inter- wounds.6,7 leukin-8) and MCP-1 (monocyte chemoattractant protein-1). When clots form, they trap numerous blood cells, includ- Remarkably, these shifts were more intense than changes in ing monocytes. Monocytes play an important role in bridg- gene expression induced by LPS. ing hemostasis and inflammation because they generate After fibrin formation occurs under physiological condi- tissue factor and inflammatory cytokines in response to tions, tPA (tissue-type plasminogen activator) is generated external stimuli.8–10 Perhaps the most well-known stimulus of and immediately begins breaking down the fibrin matrix.18 monocyte gene expression is lipopolysaccharide (LPS).11,12 Iatrogenic fibrinolysis is used as a therapeutic intervention However, LPS is not present in most clinical situations where in pathological thrombosis.19,20 In the setting of stroke, tPA is clots form, and individual clotting factors, such as tissue fac- administered at high levels to quickly dissolve the clot and tor, thrombin,13,14 and fibrin,15–17 have variable and generally improve blood flow to ischemic brain. In either setting, the mild stimulatory effects on monocyte gene expression. This effect of fibrinolysis on cytokine expression has not been

Received on: November 9, 2016; final version accepted on: July 21, 2017. From the Program in Molecular Medicine (R.A.C., J.W.R., Z.G.F., B.K.M., M.T.R., L.W.K., A.S.W.) and Departments of Internal Medicine (R.A.C., A.V.-d.-A., J.W.R., M.T.R., G.A.Z., A.S.W.), Surgery (L.W.K.), and Neurology (J.J.M.), University of Utah, Salt Lake City. The online-only Data Supplement is available with this article at http://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.117.309794/-/DC1. Correspondence to Robert Campbell, PhD, Eccles Institute of Human Genetics, Bldg 533, Rm 4280, University of Utah, Salt Lake City, Utah 84112. E-mail [email protected] © 2017 American Heart Association, Inc. Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.117.309794

1819 1820 Arterioscler Thromb Vasc Biol October 2017

I in the online-only Data Supplement). RNA-seq revealed that Nonstandard Abbreviations and Acronyms monocytes expressed over 10 000 transcripts and nearly 1000 FXIII factor XIII were differentially expressed (ie, >4-fold) in clot-retrieved IL-8 interleukin-8 monocytes compared with baseline monocytes (Figure 1A). LPS lipopolysaccharide Among the differentially expressed transcripts were IL-8 and MCP-1 monocyte chemoattractant protein-1 MCP-1, which were increased by >6- and 48-fold, respec- NF-κB nuclear factor kappa B tively, in clot-retrieved monocytes (Figure 1A and 1B). TLR4 toll-like receptor-4 Increased mRNA expression was confirmed by real-time tPA tissue-type plasminogen activator polymerase chain reaction (Figure II in the online-only Data Supplement) and was associated with elevated levels of IL-8 and MCP-1 protein, which accumulated in a time-dependent examined. Our results demonstrate that fibrinolysis signifi- fashion (Figure 1C). In addition to IL-8 and MCP-1, 13 other cantly downregulates IL-8 and MCP-1 synthesis by mono- proteins involved in inflammation were increased after whole cytes, depending on when lysis was initiated. These findings blood clot formation, and mRNA for these genes correlated have significant implications in ischemic stroke because with their protein expression levels (Figure III in the online- timely administration of tPA after stroke dictates outcomes. only Data Supplement). These cytokines were also expressed in thrombi extracted from patients (Figure IV in the online- Materials and Methods only Data Supplement), indicating that in vivo clot formation Downloaded from Materials and Methods is available in the online-only Data is also associated with the generation of inflammatory proteins. Supplement. Plasma Clot Formation Induces Inflammatory Results Gene Expression in Monocytes Whole Blood Clots Trigger Inflammatory Next we determined whether clotted plasma triggers gene http://atvb.ahajournals.org/ Gene Expression in Monocytes expression in a manner similar to whole blood. For these To test whether clots trigger inflammatory gene expression, studies, recombinant tissue factor was added to recalci- whole blood was clotted with recombinant tissue factor in the fied, cell-free plasma and allowed to clot before addition of presence of calcium for 2 hours, and gene expression patterns purified autologous monocytes (Figure V in theonline-only were examined in purified monocytes. Isolated monocytes Data Supplement). Changes in monocyte gene expression were mostly CD14+C16− (>90%), and this population did in response to clot formation were then assessed. RNA-seq change significantly after whole blood clot formation (Figure analyses revealed that mRNAs were differentially expressed by guest on July 10, 2018

Figure 1. Whole blood clot formation markedly alters gene expression patterns in monocytes. A, RNA sequencing was performed in baseline or clot-retrieved monocytes, and relative expression of individual genes were plotted against each other. Red and green indicate increased and decreased gene expression, respectively, between clotted compared with baseline (4-fold difference or greater). IL-8 (interleukin 8) and MCP-1 (monocyte chemoattractant protein-1) mRNA expression compared with other genes is highlighted by arrows. B, Dis- tribution of RNA-seq reads across MCP-1 and IL-8 transcripts (with intron/exon structures depicted below each plot) in monocytes from unclotted (baseline) blood and clotted whole blood. C, Protein for IL-8 and MCP-1 protein was measured in the plasma of clotted whole blood at specific times postclot formation. The bars in this graph indicate mean±SEM of 3 independent experiments (n=3). The asterisk indicates P<0.05 compared with freshly isolated monocytes that were processed immediately. BL indicates baseline; and NT, not treated. Campbell et al Clots Trigger Inflammatory Cell Gene Expression 1821

in monocytes embedded within clots compared with freshly Monocytes embedded in plasma clots displayed increased isolated monocytes (baseline), unstimulated monocytes expression for IL-8 and MCP-1 mRNA compared with (untreated) that were left in suspension culture for equivalent all control groups as measured by RNA-seq (Figure 2D). time periods, or thrombin-stimulated monocytes (Figure 2A Quantitative real-time polymerase chain reaction confirmed and B. Similarity matrices demonstrated that RNA patterns increases in IL-8 and MCP-1 (Figure 2E) and demonstrated for freshly isolated, untreated and thrombin-stimulated, a time-dependent change in mRNA expression (Figure VII in and clot-retrieved monocytes clustered in distinct nodes the online-only Data Supplement). Consistent with changes in (Figure 2A and B). More than 1100 transcripts were differ- mRNA, high levels of IL-8 and MCP-1 protein were detected in entially expressed by >4-fold between the clotted and control plasma clot samples (Figure 2D). In contrast, protein for IL-8 groups (Figure 2C). In addition, patterns of gene expression and MCP-1 was not observed in untreated or thrombin-stim- in monocytes isolated from whole blood and plasma clots ulated samples resuspended in M199 (Figure 2F). Monocytes were moderately correlated with one another (Figure VI in the embedded within clots also produced more IL-8 and MCP-1 online-only Data Supplement; R=0.364 and P=0.0008), dem- protein than LPS-stimulated monocytes (Figure 2F), and onstrating that the reduced plasma model clot system mimics monocytes stained similarly for IL-8 and MCP-1 as compared responses observed in a whole blood clotting model. with thrombi isolated from human patients (Figure VIII in the Downloaded from http://atvb.ahajournals.org/ by guest on July 10, 2018

Figure 2. Monocytes embedded in plasma fibrin clots undergo significant changes in gene expression.A–C , Monocytes were embedded in plasma clots for 18 hours (CLOT), and gene expression patterns were compared with control samples. The control samples included freshly isolated monocytes (BL), monocytes left in suspension M199 culture for 18 hours (NT), or monocytes in M199 stimulated with 0.1 U/mL of thrombin (Thr). Heatmaps of gene expression patterns and global sample distance clustering of monocytes that were retrieved from clots compared with control samples. Upregulated and downregulated genes (P<0.05 and >4-fold difference) in clot-retrieved monocytes compared with monocytes left in suspension culture. Expression of IL-8 (interleukin 8) and MCP-1 (monocyte chemoattractant protein-1) mRNA in monocytes retrieved from clots versus control samples, as measured by RNA sequence analysis (D) or real-time polymerase chain reaction (PCR; E). F, Synthesis of IL-8 and MCP-1 protein by monocytes embedded in plasma clots versus thrombin or LPS-stimulated monocytes. G, Global protein synthesis, as assessed by incorporation of S35-labeled methionine/cysteine into proteins, in clot-retrieved monocytes compared to monocytes left in suspension, monocytes treated with thrombin, or monocytes stimulated with LPS. The bars represent the mean±SEM >3 independent experiments for each group using 3 different donors. The asterisk and hastag indicates P<0.05 compared to BL, NT, and Thr samples (D, E). The asterisk indicates P<0.05 compared to Thr and LPS (F). The asterisk indicates P<0.05 compared to NT and Thr (G). BL indicates baseline; and NT, not treated. 1822 Arterioscler Thromb Vasc Biol October 2017

Figure 3. Synthesis of IL-8 (interleukin 8) and MCP-1 (monocyte chemoattractant protein-1) occurs at the transcriptional level. A, Time- dependent secretion of IL-8 and MCP-1 protein. B and C, IL-8 and MCP-1 protein levels (18 hours postclot formation) in the presence of cycloheximide, an inhibitor of global protein synthesis (B), or actinomycin D, a transcriptional inhibitor (C). The bars in the panels represent the mean±SEM of 3 independent experiments for each group. The asterisk and hashtag indicates P<0.05 compared with 2, 4, and 8 hour time points (A), cycloheximide treatment (B), or actinomycin treatment (C).

online-only Data Supplement). Global protein synthesis was IX in the online-only Data Supplement). We also found also markedly higher in clotted samples when compared with that increasing concentrations of purified thrombin had no not treated and LPS-treated samples (Figure 2G and Figure appreciable effect on IL-8 or MCP-1 synthesis (Figure 4A).

Downloaded from IX in the online-only Data Supplement). However, pretreatment of plasma with heparin, which blunts thrombin generation and, thereby, clot formation, significantly MCP-1 and IL-8 Expression Are Regulated reduced IL-8 and MCP-1 synthesis (Figure 4B). The specific at the Transcriptional Level thrombin inhibitor lepirudin also reduced cytokine synthesis Protein for IL-8 and MCP-1 was markedly increased after in a dose-dependent manner (Figure 4C). These studies indi-

http://atvb.ahajournals.org/ 18 hours but not before then (Figure 3A). Accumulation of cated that thrombin generation, but not thrombin itself, is criti- IL-8 and MCP-1 protein was completely blocked in the pres- cal for clot-dependent cytokine production. ence of cycloheximide, a global inhibitor of protein synthesis Because thrombin converts fibrinogen to fibrin, we deter- (Figure 3B; Figure X in the online-only Data Supplement). mined whether mediators released into the clotted milieu or Actinomycin D also prevented monocytes from generating IL-8 fibrin formation are required for cytokine synthesis. The addi- and MCP-1 protein (Figure 3C). To further examine transcription of fibrin-free serum did not induce IL-8 or MCP-1 synthe- tional regulation of IL-8 and MCP-1 induced by plasma clots, sis (Figure 5A), suggesting that fibrin formation is the primary we treated monocytes with a NF-κB (nuclear factor kappa B) instigator of these synthetic events. To examine this postulate by guest on July 10, 2018 inhibitor. Inhibition of NF-κB blocked IL-8 and MCP-1 syn- in more detail, we added Gly-Pro-Arg-Pro to thrombin-treated thesis (Figure XI in the online-only Data Supplement). Taken plasma to see if inhibition of fibrin polymerization prevents together, these results demonstrate that the synthetic events MCP-1 and IL-8 synthesis. As shown in Figure 5B, the addi- were primarily controlled at the transcriptional level. tion of Gly-Pro-Arg-Pro blocked IL-8 and MCP-1 synthesis. The concentration of Gly-Pro-Arg-Pro to inhibit fibrin forma- Fibrin Formation Is Required tion had no effect on LPS-stimulated monocytes (Figure XII in for Cytokine Synthesis the online-only Data Supplement). We also found that the addi- Results displayed in Figure 2E and 2F demonstrate that treat- tion of tissue factor and calcium to fibrinogen-deficient plasma ment of purified monocytes (ie, not in plasma) with low did not induce a robust cytokine response as compared with the concentrations of thrombin (0.1 U/mL) did not have appre- presence of fibrinogen (Figure 5C). In addition, tissue factor, ciable effects on IL-8 and MCP-1 mRNA or protein levels, alone, had little effect on cytokine production (Figure XIII in the and thrombin did not increase global protein synthesis (Figure online-only Data Supplement). Reintroduction of fibrinogen to

Figure 4. The role of thrombin in the induction of inflammatory responses in monocytes.A , Monocytes were embedded in plasma clots triggered by tissue factor and calcium or stimulated in the absence of plasma with increasing concentrations of thrombin. B, Monocytes were embedded in plasma clots in the presence or absence of heparin. C, Monocytes were embedded in plasma clots that were formed in the presence of the direct thrombin inhibitor lepirudin. In all treatment conditions, cell-free supernatants were harvested after 18 hours, and IL-8 (interleukin-8) and MCP-1 (monocyte chemoattractant protein-1) levels were assessed. The bars in the panels represent the mean±SEM of 3 independent experiments for each group. The asterisk and hashtag indicates P>0.05 compared with all thrombin concentrations (A), heparin-treated samples (B), and 10 or 100 U/mL of lepirudin (C). Campbell et al Clots Trigger Inflammatory Cell Gene Expression 1823 Downloaded from http://atvb.ahajournals.org/

Figure 5. Fibrin formation induces the inflammatory response in monocytes.A , Purified, cell-free plasma clots were triggered with recom- by guest on July 10, 2018 binant tissue factor and calcium. After 2 hours, serum was removed from the plasma clots by centrifuging the sample at 12 000g for 20 minutes. Monocytes were then incubated with plasma clots or in residual serum. B, Monocytes were embedded into plasma fibrin clots formed in the presence of Gly-Pro-Arg-Pro (GPRP), a fibrin polymerization inhibitor or vehicle (water).C , Monocytes were embedded in plasma clots that were generated from normal plasma or fibrinogen-deficient plasma.D , Monocytes were embedded in fibrinogen- deficient plasma that was reconstituted with increasing concentrations of fibrinogen.E , Plasma fibrin clots were formed in the presence of heparin to prevent thrombin generation and in the presence or absence of batroxobin (1 U/mL), which generates fibrin in a thrombin- independent manner. As a control, monocytes were resuspended in M199 and stimulated with batroxobin. In all treatment conditions, cell-free supernatants were harvested after 18 hours, and IL-8 (interleukin-8) and MCP-1 (monocyte chemoattractant protein-1) levels were assessed. The bars in the panels represent the mean±SEM of 3 independent experiments for each group. The asterisk and hashtag indicates P>0.05 in clot samples versus their comparative controls.

fibrinogen-deficient plasma, however, triggered IL-8 and MCP-1 fibrinogen and thrombin did not recapitulate the plasma system synthesis in a concentration-dependent manner (Figure 5D). (Figure XV in the online-only Data Supplement). However, the When thrombin generation was blocked by the presence of hep- addition of serum to purified fibrinogen and thrombin induced arin, but fibrin formation was allowed to proceed by the addition robust IL-8 and MCP-1 secretion. Taken together, these results of batroxobin—a snake venom capable of cleaving fibrino- indicate that fibrin formation in the setting of plasma induces gen—monocytes synthesized IL-8 and MCP-1 (Figure 5E). inflammatory gene expression in monocytes. Fibrin stabilization through factor (F) XIII may play a critical role inducing monocyte-derived IL-8 and MCP-1. To address Fibrinolysis Blunts Clot-Induced Cytokine the role of FXIII, we formed plasma clots in the presence and Synthesis in a Time-Dependent Manner absence of a specific FXIII inhibitor, T101. Plasma clots formed Next we determined whether fibrinolysis (breakdown of the fibrin in the presence of T101 had reduced fibrin cross-linking but had clot) affects clot-induced cytokine synthesis. Monocyte mRNA no effect on cytokine production (Figure XIV in the online-only expression patterns were examined in plasma clots that were Data Supplement). In addition, plasma clots were generated in lysed with tPA 30 minutes after clot formation and the addition of the presence and absence of FXIII using FXIII-deficient plasma. monocytes. As shown in Figure 6A, mRNA expression patterns In the absence of FXIII, we observed decreased fibrin cross- were markedly different in clot-retrieved monocytes compared linking but no change in IL-8 and MCP-1 synthesis (Figure with untreated monocytes, and these changes were blunted in XIV in the online-only Data Supplement). We next examined the presence of tPA. Furthermore, patterns of gene expression in whether purified fibrinogen and thrombin alone were sufficient monocytes isolated from plasma clots and plasma clots plus tPA to induce IL-8 and MCP-1 synthesis. Interestingly, purified demonstrated a negative correlation with one another (R=0.807, 1824 Arterioscler Thromb Vasc Biol October 2017 Downloaded from http://atvb.ahajournals.org/ by guest on July 10, 2018

Figure 6. Timely fibrinolysis of clots blunts the inflammatory response in monocytes.A , Monocytes were left in suspension culture (NT), embedded in plasma fibrin clots, or embedded in plasma fibrin clots for 30 minutes and then lysed with tPA (tissue-type plasminogen activator). After 18 hours, monocytes were harvested and prepared for RNA sequencing. A, Heatmaps of mRNA expression patterns (4-fold change or greater) in NT, Clot, and Clot+tPA samples. B, Distribution of RNA sequencing reads for IL-8 (interleukin-8) and MCP-1 (monocyte chemoattractant protein-1) in NT, Clot, and Clot+tPA samples. C, IL-8 and MCP-1 protein levels in NT, Clot, and Clot+tPA samples. E, Monocytes were embedded in plasma clots and subsequently treated with tPA at different times postclot formation. IL-8 and MCP-1 protein was assessed after 18 hours (D). Whole blood was clotted in the presence (added 1 hour postclot formation) or absence of tPA, and after 18 hours, IL-8 and MCP-1 protein levels were determined. The asterisk indicates P>0.05 in clot samples versus their comparative controls. The bars in C–E represent the mean±SEM of 3 independent experiments. NT indicates not treated.

P<0.001; Figure XVI in the online-only Data Supplement), dem- Discussion onstrating that the fibrinolysis significantly blunted inflamma- Monocytes play prominent roles in driving coagulation and tory gene responses. Monocytes isolated from lysed samples had inflammation, in part, through de novo synthesis of throm- reduced expression of IL-8 and MCP-1 mRNA when compared boinflammatory proteins on stimulation. Here, we determined with clot-retrieved monocytes (Figure 6B), and protein for these for the first time intact clots trigger robust, global expression cytokines was similarly reduced (Figure 6C). Decreases in IL-8 of inflammatory genes in monocytes, including IL-8 and and MCP-1 were time-dependent and not durable when tPA- MCP-1. Thrombi extracted from patients demonstrated sim- treatment was delayed by >2 hours after addition of monocytes ilar patterns of cytokine expression, suggesting that in vivo (Figure 6D). tPA treatment similarly reduced IL-8 and MCP-1 clot formation can also induce the generation of IL-8 and protein levels when it was administered to whole blood 1 hour MCP-1 (Figure IV in the online-only Data Supplement). The after clots were formed (Figure 6C). ability of the clot to activate multiple inflammatory cytokines Campbell et al Clots Trigger Inflammatory Cell Gene Expression 1825

(IL-8, MCP-1, and others; Figure XVII in the online-only signaling, had no effect on macrophage cytokine synthesis.24 Data Supplement) and pathways underscores the complexity Taken together, these findings indicate only a small role for of the interaction between coagulation reactions and inflam- thrombin in regulating these inflammatory responses. mation. Furthermore, these data suggest that dysregulation Although thrombin did not directly modulate gene expres- of clot formation may significantly impact downstream pro- sion in purified, plasma-free monocytes, our data demonstrate cesses, such as wound healing, and could contribute to ongo- that thrombin is still indirectly necessary for cytokine responses ing coagulation responses because genes, such as tissue factor by acting on fibrinogen. Our data suggest that thrombin exerts display, increased expression in clot-retrieved monocytes its effect by inducing fibrin formation, which directly triggers (Figure XVIII in the online-only Data Supplement). gene expression pathways in monocytes. Indeed, pretreat- RNA-seq analysis of monocytes derived from whole ment of plasma with the peptide Gly-Pro-Arg-Pro, which blood clots demonstrated robust differential gene expression inhibits fibrin polymerization by blocking A:a interactions, compared with monocytes isolated immediately after whole significantly reduced cytokine synthesis.25,26 Likewise, fibrin- blood was drawn. To further verify specific changes in the ogen-deficient plasma failed to support cytokine production, monocyte transcriptome because of plasma clot formation, but reintroduction of fibrinogen to plasma restored clot for- we performed additional RNA-seq analysis from mono- mation and cytokine synthesis. Batroxobin,27 which cleaves cytes embedded in a purified plasma clot system. RNA-seq fibrinogen into fibrin independent of thrombin, also restored analysis revealed that monocytes embedded in plasma clots clot formation and supported IL-8 and MCP-1 production in mounted a robust gene expression response, and differen- heparinized plasma. These data demonstrate that fibrin forma- Downloaded from tially expressed mRNAs resembled observed changes in the tion elicits gene expression responses in monocytes. However, more complex whole blood clot (Figure VI in the online- the absence of fibrinogen did not completely abolish cytokine only Data Supplement; R=0.364; P value =0.0008). Plasma formation (Figure 5C), and purified fibrin generated by throm- clots induced robust synthesis of IL-8 and MCP-1 mRNA bin was unable to induce IL-8 and MCP-1 synthesis (Figure and protein and a marked increase in global protein synthe- XV in the online-only Data Supplement). These findings sug- http://atvb.ahajournals.org/ sis. Consistent with an increase in global protein synthesis, gest that additional factors circulating in the plasma or gener- ribosomes attached to endoplasmic reticulum were more ated during clot formation are necessary in addition to fibrin frequently observed in clot-embedded monocytes compared to induce robust cytokine response in monocytes. While the with unstimulated monocytes (Figure IX in the online-only current study focused on the role of fibrin in the setting of Data Supplement). The upregulation of mRNAs followed by plasma, additional studies are warranted to determine other subsequent synthesis of their corresponding proteins (Figure critical factors important in regulating this response. XVII in the online-only Data Supplement) in clot-retrieved The exact mechanism by which fibrinogen/fibrin drive monocytes suggests that the synthesis of numerous genes is IL-8, MCP-1, and other inflammatory gene expression is not by guest on July 10, 2018 regulated at the transcriptional level. In this regard, transcrip- known. Previous studies have demonstrated that fibrin(ogen) tional inhibitors completely abrogated the synthesis of IL-8 is capable of inducing synthesis of IL-8, MCP-1, IL-6, IL-1β, and MCP-1, demonstrating that clots induce the transcription and other cytokines in immune cells and vascular endothe- and then translation of both mRNAs. lial cell in vitro and in vivo.24,28–31 The mechanism behind The data indicate that generation of fibrin is essential for fibrinogen- or fibrin-driven cytokine responses was thought plasma clot–induced changes in gene expression. We first to be through TLR4 (toll-like receptor-4) on macrophages.32,33 demonstrated using heparin, a clinically useful anticoagu- However, inhibition of TLR434,35 had no effect on cytokine lant, a significant reduction in cytokine protein expression. To production (Figure XIX in the online-only Data Supplement), focus on the role of thrombin, we next used the direct throm- suggesting that ligation of TLR4 by fibrinogen is not the bin inhibitor lepirudin,21 which also blocked the production mechanism for induction of cytokine synthesis in our experi- of cytokines. We did not observe IL-8 and MCP-1 synthesis ments.29 In addition, these data indicate that low levels of LPS when increasing concentrations of thrombin were introduced in clots are not responsible for increased IL-8 and MCP-1 to purified monocytes in the absence of plasma, despite the synthesis. The generation of FXIIIa and its role in stabilizing fact that the higher doses used exceeded previously reported fibrin through cross-linking may also play a role in driving peak levels of thrombin generation.22,23 We also found that the cytokine response in monocytes. Previous studies have

thrombin did not increase global protein synthesis by mono- demonstrated that FXIIIa is capable of binding to αvβ3 on cytes, and mRNA patterns in thrombin-stimulated monocytes monocytes and upregulating proliferation and migration while clustered with mRNA expression in unstimulated monocytes. preventing apoptosis.36 Using pharmacological inhibition and These results are similar to those produced by Nieuwenhuizen FXIII-deficient plasma, we observed minimal changes in IL-8 et al13 who demonstrated that high concentrations of thrombin and MCP-1 synthesis (Figure XIV in the online-only Data did not induce cytokine synthesis but contrast a report show- Supplement). Another possibility is CD11b/CD18, which ing that thrombin regulates the expression of inflammatory binds fibrinogen/fibrin with high affinity through the c-termi- mRNAs.14 The reasons for these differences are not obvious nal region of fibrinogen’s gamma chain.37 It has been shown but may be because of the techniques (RNA-sequencing ver- that this site becomes exposed on conversion of fibrinogen to sus microarray analysis) and the experimental milieus used. fibrin,38,39 and previous publications have shown that interac- Others have used genetics models to examine the role of tions of fibrin with CD11b/CD18 regulate NF-κB-dependent thrombin in regulating inflammation. For example, deletion of cytokine production. Interestingly, pharmacological inhibition protease activator receptor 1, the major receptor for thrombin of NF-κB blunts clot-induced synthesis of IL-8 and MCP-1 1826 Arterioscler Thromb Vasc Biol October 2017

(Figure XI in the online-only Data Supplement). However, an thank Ms Diana Lim for her excellent preparation of the figures. We inhibitory antibody against CD11b blocked fibrinogen bind- thank Drs Tim LaPine and Tom Martins at ARUP for their help with ing but had little effect on cytokine production (Figure XX the multiplex analysis. In addition, we thank the High-Throughput Genomics and Bioinformatics Analysis and the Electron Microscopy in the online-only Data Supplement). Additional studies using cores at the University of Utah. mice deficient in CD11b/CD18 are necessary to fully examine the role CD11b/CD18 in plasma clot–mediated IL-8 and Sources of Funding MCP-1 synthesis because studies using CD11b/C18 knockout This work was funded by the following: HL066277 and mice demonstrated that they are protected from cerebral isch- HL112311 (A.S. Weyrich), AG040631 and HL092161 (M.T. emia reperfusion injury,40 suggesting that interactions of fibrin Rondina), R37HL044525 (G.A. Zimmerman), Utah Stroke Net with CD11b/CD18 may regulate inflammatory processes dur- Fellowship 4U10NS086606, and the American Heart Association ing stroke. 11POST7290019 (R.A. Campbell). While the ability of fibrin to interact with vessels and cells is important in stemming blood loss, fibrin deposition Disclosures and its prompt removal are necessary in wound repair.41 The None. removal of fibrin from a thrombus is regulated by the generation of plasmin from its proform, plasminogen. Plasminogen References binding to fibrin in conjunction with tPA results in plasmin 1. Foley JH, Conway EM. Cross talk pathways between coagulation generation and the subsequent degradation of the fibrin and inflammation. Circ Res. 2016;118:1392–1408. doi: 10.1161/ Downloaded from clot18; therefore, a delicate balance is needed under physi- CIRCRESAHA.116.306853. 2. Lindsberg PJ, Grau AJ. Inflammation and infections as risk factors ological conditions between procoagulant and fibrinolytic for ischemic stroke. Stroke. 2003;34:2518–2532. doi: 10.1161/01. processes to prevent bleeding or thrombosis and to ensure STR.0000089015.51603.CC. proper wound repair. Previous studies in plasminogen- 3. Monroe DM, Hoffman M, Roberts HR. Transmission of a procoagu- deficient mice have demonstrated exacerbated inflamma- lant signal from tissue factor-bearing cell to platelets. Blood Coagul Fibrinolysis. 1996;7:459–464. http://atvb.ahajournals.org/ tory diseases associated with arthritis, nerve damage, and 4. Campbell RA, Overmyer KA, Bagnell CR, Wolberg AS. Cellular procoag- osteoporosis.42–45 Genetic or pharmacological depletion of ulant activity dictates clot structure and stability as a function of distance fibrinogen rescued the associated phenotype, suggesting from the cell surface. Arterioscler Thromb Vasc Biol. 2008;28:2247–2254. that fibrin and its subsequent lysis modulates the disease doi: 10.1161/ATVBAHA.108.176008. 5. Campbell RA, Overmyer KA, Selzman CH, Sheridan BC, Wolberg AS. 42,44,45 46 process. Additionally, Cole et al demonstrated that Contributions of extravascular and intravascular cells to fibrin network plasminogen-deficient animals have elevated levels of IL-6. formation, structure, and stability. Blood. 2009;114:4886–4896. doi: These studies, together with our findings, demonstrate that 10.1182/blood-2009-06-228940. 6. Wolberg AS, Campbell RA. Thrombin generation, fibrin clot formation the balance between fibrin formation and lysis critically

by guest on July 10, 2018 and hemostasis. Transfus Apher Sci. 2008;38:15–23. doi: 10.1016/j. regulates IL-8, MCP-1, and other inflammatory cytokine transci.2007.12.005. production by monocytes. 7. Weisel JW. Structure of fibrin: impact on clot stability. J Thromb Haemost. As many as 795 000 individuals experience a new or recur- 2007;5(suppl 1):116–124. doi: 10.1111/j.1538-7836.2007.02504.x. 8. Hogg N. Human monocytes are associated with the formation of fibrin. 47 rent stroke in the United States each year. Intravenous fibri- J Exp Med. 1983;157:473–485. nolytic therapy involving the use of tPA for acute stroke is 9. von Brühl ML, Stark K, Steinhart A, et al. Monocytes, neutrophils, and widely agreed on to be beneficial if administered within 3 to platelets cooperate to initiate and propagate venous thrombosis in mice in 4.5 hours of symptom onset.48 This demonstrates that timely vivo. J Exp Med. 2012;209:819–835. doi: 10.1084/jem.20112322. 10. Tezono K, Sarker KP, Kikuchi H, Nasu M, Kitajima I, Maruyama I. administration of tPA is necessary for maximal benefit to the Bioactivity of the vascular endothelial growth factor trapped in fibrin clots: patients. While fibrinolytics allow for reperfusion in the isch- production of IL-6 and IL-8 in monocytes by fibrin clots. Haemostasis. emic area, our data suggest an added benefit of timely fibri- 2001;31:71–79. doi: 48047. 11. Guha M, Mackman N. LPS induction of gene expression in human mono- nolysis is modulation of cytokine production in the ischemic cytes. Cell Signal. 2001;13:85–94. area. Specifically, addition of tPA to clots within 2 hours sig- 12. Sharif O, Bolshakov VN, Raines S, Newham P, Perkins ND. Transcriptional nificantly reduces the synthesis of IL-8 and MCP-1. Increased profiling of the LPS induced NF-kappaB response in macrophages. BMC inflammation is a hallmark finding in stroke, as evidenced by Immunol. 2007;8:1. doi: 10.1186/1471-2172-8-1. 13. Nieuwenhuizen L, Falkenburg WJ, Schutgens RE, Roosendaal G, van increased levels of IL-8, MCP-1, IL-1β, and other cytokines Veghel K, Biesma DH, Lafeber FP. Stimulation of naïve monocytes and in the blood—which associate with poor outcomes.49–51 While PBMCs with coagulation proteases results in thrombin-mediated and these studies suggest certain cytokines are important in regu- PAR-1-dependent cytokine release and cell proliferation in PBMCs only. lating outcomes in stroke, they have not examined the role Scand J Immunol. 2013;77:339–349. doi: 10.1111/sji.12033. 14. López ML, Bruges G, Crespo G, Salazar V, Deglesne PA, Schneider H, of fibrinolytic therapy in regulating cytokine expression after Cabrera-Fuentes H, Schmitz ML, Preissner KT. Thrombin selectively stroke. Our findings demonstrate for the first time and strongly induces transcription of genes in human monocytes involved in inflam- suggest that the positive benefits of timely tPA administra- mation and wound healing. Thromb Haemost. 2014;112:992–1001. doi: 10.1160/TH14-01-0034. tion in stroke is due, in part, to an attenuated inflammatory 15. Lee ME, Kweon SM, Ha KS, Nham SU. Fibrin stimulates microfilament response. reorganization and IL-1beta production in human monocytic THP-1 cells. Mol Cells. 2001;11:13–20. 16. Flick MJ, Du X, Degen JL. Fibrin(ogen)-alpha M beta 2 interactions Acknowledgments regulate leukocyte function and innate immunity in vivo. Exp Biol Med We thank Dr Tammy Smith for input on experimental design and data (Maywood). 2004;229:1105–1110. collection in reference to Figure 1 and Figures III and IX in the online- 17. Flick MJ, Du X, Witte DP, Jirousková M, Soloviev DA, Busuttil SJ, Plow only Data Supplement, as well as critically reviewing the article. We EF, Degen JL. Leukocyte engagement of fibrin(ogen) via the integrin Campbell et al Clots Trigger Inflammatory Cell Gene Expression 1827

receptor alphaMbeta2/Mac-1 is critical for host inflammatory response in 35. Rondina MT, Schwertz H, Harris ES, Kraemer BF, Campbell RA, vivo. J Clin Invest. 2004;113:1596–1606. doi: 10.1172/JCI20741. Mackman N, Grissom CK, Weyrich AS, Zimmerman GA. The sep- 18. Cesarman-Maus G, Hajjar KA. Molecular mechanisms of fibrinolysis.Br tic milieu triggers expression of spliced tissue factor mRNA in human J Haematol. 2005;129:307–321. doi: 10.1111/j.1365-2141.2005.05444.x. platelets. J Thromb Haemost. 2011;9:748–758. doi: 10.1111/j.1538- 19. Hacke W, Kaste M, Fieschi C, Toni D, Lesaffre E, von Kummer R, 7836.2011.04208.x. Boysen G, Bluhmki E, Höxter G, Mahagne MH. Intravenous thromboly- 36. Dardik R, Krapp T, Rosenthal E, Loscalzo J, Inbal A. Effect of FXIII on sis with recombinant tissue plasminogen activator for acute hemispheric monocyte and fibroblast function. Cell Physiol Biochem. 2007;19:113– stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA. 120. doi: 10.1159/000099199. 1995;274:1017–1025. 37. Ugarova TP, Solovjov DA, Zhang L, Loukinov DI, Yee VC, Medved 20. Adams HP Jr, del Zoppo G, Alberts MJ, et al; American Heart Association; LV, Plow EF. Identification of a novel recognition sequence for integ- American Stroke Association Stroke Council; Clinical Cardiology Council; rin alphaM beta2 within the gamma-chain of fibrinogen. J Biol Chem. Cardiovascular Radiology and Intervention Council; Atherosclerotic 1998;273:22519–22527. Peripheral Vascular Disease and Quality of Care Outcomes in Research 38. Lishko VK, Kudryk B, Yakubenko VP, Yee VC, Ugarova TP. Regulated Interdisciplinary Working Groups. Guidelines for the early management of unmasking of the cryptic binding site for integrin alpha M beta 2 in the adults with ischemic stroke: a guideline from the American Heart Association/ gamma C-domain of fibrinogen. Biochemistry. 2002;41:12942–12951. American Stroke Association Stroke Council, Clinical Cardiology Council, 39. Sitrin RG, Pan PM, Srikanth S, Todd RF III. Fibrinogen activates Cardiovascular Radiology and Intervention Council, and the Atherosclerotic NF-kappa B transcription factors in mononuclear phagocytes. J Immunol. Peripheral Vascular Disease and Quality of Care Outcomes in Research 1998;161:1462–1470. Interdisciplinary Working Groups: the American Academy of Neurology 40. Soriano SG, Coxon A, Wang YF, Frosch MP, Lipton SA, Hickey PR, affirms the value of this guideline as an educational tool for neurologists. Mayadas TN. Mice deficient in Mac-1 (CD11b/CD18) are less susceptible Stroke. 2007;38:1655–1711. doi: 10.1161/STROKEAHA.107.181486. to cerebral ischemia/reperfusion injury. Stroke. 1999;30:134–139. 21. Gosselin RC, Dager WE, King JH, Janatpour K, Mahackian K, Larkin 41. Clark RA. Fibrin and wound healing. Ann N Y Acad Sci. 2001;936:355–367. EC, Owings JT. Effect of direct thrombin inhibitors, bivalirudin, lepirudin, 42. Bugge TH, Kombrinck KW, Flick MJ, Daugherty CC, Danton MJ, Degen Downloaded from and argatroban, on prothrombin time and INR values. Am J Clin Pathol. JL. Loss of fibrinogen rescues mice from the pleiotropic effects of plas- 2004;121:593–599. doi: 10.1309/D79K-4YG7-8NTN-YY38. minogen deficiency. Cell. 1996;87:709–719. 22. Machlus KR, Colby EA, Wu JR, Koch GG, Key NS, Wolberg AS. Effects 43. Hultman K, Cortes-Canteli M, Bounoutas A, Richards AT, Strickland S, of tissue factor, thrombomodulin and elevated clotting factor levels on Norris EH. Plasmin deficiency leads to fibrin accumulation and a com- thrombin generation in the calibrated automated thrombogram. Thromb promised inflammatory response in the mouse brain. J Thromb Haemost. Haemost. 2009;102:936–944. doi: 10.1160/TH09-03-0180. 2014;12:701–712. doi: 10.1111/jth.12553. 23. Allen GA, Wolberg AS, Oliver JA, Hoffman M, Roberts HR, Monroe DM. 44. Kao WW, Kao CW, Kaufman AH, Kombrinck KW, Converse RL, Good http://atvb.ahajournals.org/ Impact of procoagulant concentration on rate, peak and total thrombin WV, Bugge TH, Degen JL. Healing of corneal epithelial defects in plas- generation in a model system. J Thromb Haemost. 2004;2:402–413. minogen- and fibrinogen-deficient mice. Invest Ophthalmol Vis Sci. 24. Szaba FM, Smiley ST. Roles for thrombin and fibrin(ogen) in cyto- 1998;39:502–508. kine/chemokine production and macrophage adhesion in vivo. Blood. 45. Akassoglou K, Kombrinck KW, Degen JL, Strickland S. Tissue plas- 2002;99:1053–1059. minogen activator-mediated fibrinolysis protects against axonal 25. Laudano AP, Doolittle RF. Synthetic peptide derivatives that bind to fibrin- degeneration and demyelination after sciatic nerve injury. J Cell Biol. ogen and prevent the polymerization of fibrin monomers. Proc Natl Acad 2000;149:1157–1166. Sci U S A. 1978;75:3085–3089. 46. Cole HA, Ohba T, Nyman JS, Hirotaka H, Cates JM, Flick MJ, Degen 26. Laudano AP, Doolittle RF. Studies on synthetic peptides that bind to fibrin- JL, Schoenecker JG. Fibrin accumulation secondary to loss of plasmin-

by guest on July 10, 2018 ogen and prevent fibrin polymerization. Structural requirements, number of mediated fibrinolysis drives inflammatory osteoporosis in mice. Arthritis binding sites, and species differences. Biochemistry. 1980;19:1013–1019. Rheumatol. 2014;66:2222–2233. doi: 10.1002/art.38639. 27. Stocker K, Barlow GH. The coagulant enzyme from Bothrops atrox 47. Mozaffarian D, Benjamin EJ, Go AS, et al; American Heart Association venom (batroxobin). Methods Enzymol. 1976;45:214–223. Statistics Committee and Stroke Statistics Subcommittee. Heart dis- 28. Qi J, Kreutzer DL. Fibrin activation of vascular endothelial cells. Induction ease and stroke statistics–2015 update: a report from the American of IL-8 expression. J Immunol. 1995;155:867–876. Heart Association. Circulation. 2015;131:e29–e322. doi: 10.1161/ 29. Kuhns DB, Priel DA, Gallin JI. Induction of human monocyte interleukin CIR.0000000000000152. (IL)-8 by fibrinogen through the toll-like receptor pathway. Inflammation. 48. Jauch EC, Saver JL, Adams HP Jr, Bruno A, Connors JJ, Demaerschalk 2007;30:178–188. doi: 10.1007/s10753-007-9035-1. BM, Khatri P, McMullan PW Jr, Qureshi AI, Rosenfield K, Scott PA, 30. Lee ME, Rhee KJ, Nham SU. Fragment E derived from both fibrin and Summers DR, Wang DZ, Wintermark M, Yonas H; American Heart fibrinogen stimulates interleukin-6 production in rat peritoneal macro- Association Stroke Council; Council on Cardiovascular Nursing; phages. Mol Cells. 1999;9:7–13. Council on Peripheral Vascular Disease; Council on Clinical Cardiology. 31. Perez RL, Roman J. Fibrin enhances the expression of IL-1 beta by human Guidelines for the early management of patients with acute ischemic peripheral blood mononuclear cells. Implications in pulmonary inflamma- stroke: a guideline for healthcare professionals from the American Heart tion. J Immunol. 1995;154:1879–1887. Association/American Stroke Association. Stroke. 2013;44:870–947. doi: 32. Smiley ST, King JA, Hancock WW. Fibrinogen stimulates macro- 10.1161/STR.0b013e318284056a. phage chemokine secretion through toll-like receptor 4. J Immunol. 49. Kostulas N, Kivisäkk P, Huang Y, Matusevicius D, Kostulas V, Link H. 2001;167:2887–2894. Ischemic stroke is associated with a systemic increase of blood mono- 33. Al-ofi E, Coffelt SB, Anumba DO. Fibrinogen, an endogenous ligand nuclear cells expressing interleukin-8 mRNA. Stroke. 1998;29:462–466. of Toll-like receptor 4, activates monocytes in pre-eclamptic patients. 50. Whiteley W, Jackson C, Lewis S, Lowe G, Rumley A, Sandercock P, J Reprod Immunol. 2014;103:23–28. doi: 10.1016/j.jri.2014.02.004. Wardlaw J, Dennis M, Sudlow C. Inflammatory markers and poor outcome 34. Belcher JD, Chen C, Nguyen J, Milbauer L, Abdulla F, Alayash AI, after stroke: a prospective cohort study and systematic review of interleu- Smith A, Nath KA, Hebbel RP, Vercellotti GM. Heme triggers TLR4 kin-6. PLoS Med. 2009;6:e1000145. doi: 10.1371/journal.pmed.1000145. signaling leading to endothelial cell activation and vaso-occlusion in 51. Amantea D, Nappi G, Bernardi G, Bagetta G, Corasaniti MT. Post- murine sickle cell disease. Blood. 2014;123:377–390. doi: 10.1182/ ischemic brain damage: pathophysiology and role of inflammatory media- blood-2013-04-495887. tors. FEBS J. 2009;276:13–26. doi: 10.1111/j.1742-4658.2008.06766.x.

Highlights • Clot formation induces a robust inflammatory response in monocytes specifically through fibrin-dependent mechanisms. • Timely fibrinolysis blunts monocyte inflammatory responses, revealing a potential novel mechanism of fibrinolytic therapy in stroke. Downloaded from Clots Are Potent Triggers of Inflammatory Cell Gene Expression: Indications for Timely Fibrinolysis Robert A. Campbell, Adriana Vieira-de-Abreu, Jesse W. Rowley, Zechariah G. Franks, Bhanu Kanth Manne, Matthew T. Rondina, Larry W. Kraiss, Jennifer J. Majersik, Guy A. Zimmerman and Andrew S. Weyrich http://atvb.ahajournals.org/

Arterioscler Thromb Vasc Biol. 2017;37:1819-1827; originally published online August 3, 2017; doi: 10.1161/ATVBAHA.117.309794 Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2017 American Heart Association, Inc. All rights reserved. Print ISSN: 1079-5642. Online ISSN: 1524-4636 by guest on July 10, 2018

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://atvb.ahajournals.org/content/37/10/1819

Data Supplement (unedited) at: http://atvb.ahajournals.org/content/suppl/2017/08/01/ATVBAHA.117.309794.DC1

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at: http://www.lww.com/reprints

Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular Biology is online at: http://atvb.ahajournals.org//subscriptions/

Supplemental Figure I. Whole blood clot formation does not alter CD14+CD16- and

CD14+CD16+ populations. Whole blood (10 mLs) was left untreated or allowed to clot

(20 mM CaCl2 final and 1:20000 Innovin, final) for two hours. Monocytes were isolated using CD14+ microbeads from miltenyi biotec as described in the material and methods.

Roughly 2 million monocytes were isolated from untreated blood and clotted blood (data not shown). Monocytes were then stained for CD14 and CD16. CD14 and CD16 positivity was determined by flow cytometry compared to IgG controls. CD14++CD16-,

CD14++CD16+, and CD14+CD16++ populations were unchanged between the two groups

(N=3).

Supplemental Figure II. IL-8 and MCP-1 mRNA increase compared to untreated whole blood. Monocytes were isolated from untreated whole blood (NT) or whole blood

o clots (WB Clot) after 2 hours at 37 C at 5% CO2 and mRNA isolated. Real-time PCR to measure IL-8 and MCP-1 levels was then performed (n>3, *p<0.05).

Supplemental Figure III. Clot induced changes in the expression of mRNAs and their corresponding proteins correlate with one another. (A, B) Whole blood clots were formed as described in the Supplemental Methods. Proteins were assessed by a dot blot cytokine protein array in freshly-isolated whole blood (BL- baseline) and compared to a whole blood clot after 18 hours. IL-8 (red) and MCP-1 (blue) are indicated on the actual dot blot and the relative levels of proteins that were increased in response to clotting are shown in panel B. (C) Proteins with increased expression (Panel B) were correlated to their corresponding mRNA expression levels, as measured by RNA-sequencing.

Supplemental Figure IV. IL-8 and MCP-1 are expressed in clots extracted from patients with renal disease. Thrombi were harvested from the iliac artery of a patient who developed thrombosis after undergoing renal transplantation. The clots were stained for CD14 (green), IL- 8 (red), or MCP-1 (red). IgG using isotype controls at similar concentrations are shown to assess background stain. Nucleated cells were stained with DAPI (blue). Differential contrast is shown in the far-right panels. The images are representative of three independent samples taken from patients with varying degrees of renal disease.

Supplemental Figure V. Tissue factor driven thrombin generation induces greater levels of thrombin and clot formation compared to the contact pathway. Plasma was isolated as described in the material and methods. Plasma was recalcified (20 mM final) in the presence (TF pathway) or absence (contact pathway) of recombinant tissue factor (1:20000 final) and thrombin generation measured using a fluoregenic substrate as previously described (A). Clot formation was measured in parallel using a thermomax plate reader that monitored changes in optical density at 405 nm (B).

Fibrinolysis was also measured with the addition tissue factor plasminogen (tPA) in parallel clot formation assays (C). All assays were monitored for 2 hours at 37oC.

Tracings are from representative experiments (n=3).

Supplemental Figure VI. Gene expression changes observed in whole blood and plasma fibrin clots correlate with one another. Inflammatory mRNA expression fold changes in monocytes from whole blood clots versus BL whole blood compared to fold changes in monocytes from plasma clot versus NT. Inflammatory genes selected according to the inflammation Gene Ontology (GO) (GO term: inflammation) Genes that did not significantly change (p<0.05) in response to clotting in either group were excluded from the analysis. A Pearson correlation revealed a positive correlation

(p<0.05) between whole blood and plasma clots.

Supplemental Figure VII. IL-8 and MCP-1 mRNA increase over time in plasma clots. Monocytes were embedded into plasma clots or stimulated with 0.1 U/mL thrombin and lysed in Trizol. IL-8 and MCP-1 mRNA was measured at various time point (n>3, *p<0.05).

Supplemental Figure VIII. IL-8 and MCP-1 is expressed by monocytes embedded into plasma clots. Monocytes embedded into plasma clots after 18 hours were processed similar to human thrombi isolation for cytokine staining (see material and methods). Cryosectioned plasma clots were stained for IL-8 (red), MCP-1 (red) and

CD14 (green). DAPI was used to stain the nucleus. IgGs for IL-8 and MCP-1 are the same. Light transmission images were taken to examine plasma clot morphology (DIC, far right). IL-8 and MCP-1 staining co-localized with CD14, a specific monocyte marker.

The images are representative of three independent experiments.

Supplemental Figure IX. Monocytes embedded in plasma fibrin clots have increased frequency of ribosomes tracking along endoplasmic reticulum. (A)

Electron micrographs depicting monocytes at baseline compared to monocytes embedded in plasma fibrin clots. Monocytes embedded in plasma fibrin clots have increased cytoplasm to nuclear ratio as well as increased frequency of ribosomes attached to endoplasmic reticulum (white arrow). These images are representative of

3 independent experiments.

Supplemental Figure X. Production of IL-8 and MCP-1 is transcriptionally and translationally controlled. Monocytes were treated with vehicle, actinomycin D (ActD;

5µg/mL), or cycloheximide (5µg/mL) for one hour before being embedded into plasma clots. Supernatant were harvested at the indicated time ponts and IL-8 and MCP-1 were measured (N>3, *p<0.05).

Supplemental Figure XI. Production of IL-8 and MCP-1 is NFkB-dependent.

Monocytes were treated with vehicle, or the NFkB inhibitor (BAY 11-7082, 5 µM final) for one hour before being embedded into plasma clots. Supernatants were harvested at the indicated time ponts and IL-8 and MCP-1 protein was assessed (N>3, *p<0.05).

Supplemental Figure XII. The fibrin inhibitor GPRP does not affect LPS-induced cytokine synthesis. Monocytes were treated with 40 mg/mL GPRP or vehicle control

(water) for 1 hour before addition of LPS (100 ng/mL). Supernatants were collected at 18 hours and cytokine synthesis was measured by ELISA (N>3).

Supplemental Figure XIII. Tissue factor does not directly induce MCP-1 synthesis.

Monocytes were left untreated, stimulated with recombinant tissue factor (1:20000 dilution, final) or LPS (100 ng/mL, final). After 18 hours, supernatants were harvested and MCP-1 was measured (n = 3, *p<0.05).

Supplemental Figure XIV. FXIII does not influence IL-8 or MCP-1 production from monocytes embedded in plasma clots. Plasma clots using FXIII deficient plasma were formed in the presence or absence of recombinant FXIIIa (10µg/mL, final). Plasma clots were also formed in the presence of 5µM T101, a specific FXIII inhibitor. After 60 minutes, reactions were stopped by the addition of 50 mM DTT, 12.5 mM EDTA, and 8 M urea. Samples were then incubated at 60°C for 1 hour with occasional agitation. Samples were reduced, boiled, and probed with a rabbit anti-human fibrinogen polyclonal antibodies to determine the presence of g-g dimers. Monocytes were embedded into clots formed in the presence or absence of FXIII or in the presence of T101 or vehicle and cytokine production was measured after 18 hours.

Supplemental Figure XV. Monocytes embedded in fibrinogen clots combined with serum synthesize IL-8 and MCP-1. Monocytes were embedded in plasma clots as described in the materials and methods or clots formed from purified fibrinogen (2 mg/mL, final) and thrombin (1 U/mL, final). To some purified fibrin clots, serum from autologous donors was added to recapitulate plasma fibrin clots (N=5, p<0.05).

Supplemental Figure XVI. Fibrinolysis blunts clot-induced changes in inflammatory gene expression. Using genes selected according to the inflammation

GO term, a correlation analysis was performed between monocyte-derived mRNAs that displayed altered expression (p<0.05) when embedded in plasma clots (compared to

NT) versus their expression level when the plasma clots were lysed with tPA (compared to plasma clot) (lysis = 30 minutes post-clot formation). A Pearson correlation revealed a negative correlation (p<0.05) in mRNA expression levels with and without lysis.

Supplemental Figure XVII. Plasma fibrin clots induce robust cytokine protein expression. Baseline plasma and supernatants from monocytes stimulated by plasma clots after 18 hours were harvested for multiplex cytokine analysis that was performed at ARUP Laboratories as described in the Supplemental Methods. N=3 for all experiments. Significance indicated by * with p<0.05.

80 *

Factor mRNA Tissue GADPH) to Normalized 0

NT Baseline to Compared Change (Fold

Baseline Plasma Clot

Supplemental Figure XVIII. Plasma fibrin clots induce tissue factor expression in monocytes. Tissue factor (TF) mRNA expression was assessed in monocytes left untreated (NT) or embedded in plasma clot for 2 hours compared to baseline monocytes. TF express levels were normalized to GADPH. Significance indicated by * with p<0.05 compared to baseline and NT (N=3).

Supplemental Figure XIX. Plasma clots do not induce inflammatory gene expression via toll-like receptor-4 (TLR4). Monocytes were pretreated with the TLR4 inhibitor CLI-095 (1 µM) and then embedded in plasma clots or stimulated with LPS

(100 ng/mL). IL-8 (left panel) and MCP-1 (right panel) protein levels were assessed after 18 hours. The bars in this figure represent the mean±SEM of 3 independent experiments. The asterisk indicates p<0.05 in untreated clot or LPS-treated samples versus the TLR4 inhibitor.

Supplemental Figure XX. Inhibition of monocyte CD11b does reduce IL-8 or MCP-

1 synthesis in response to plasma clot stimulation. Monocytes were pretreated with

5 µg/mL anti-CD11b antibody (red) or isotype control (blue) before the addition of 10 µg total Alexa-fluor 555 labeled fibrinogen. Fibrinogen binding was assessed by flow cytometry (A). Monocytes were left alone or pretreated with 5 µg/mL anti-CD11b antibody or isotype control before cells were added to plasma clots (B). After 18 hours, supernatants were harvested and IL-8 and MCP-1 were measured (n>3).

Supplemental Methods

Whole blood collection and plasma isolation

The University of Utah Institutional Review Board approved this study and all subjects provided informed consent. Human peripheral venous blood from healthy, medication- free adult subjects was drawn into acid-citrate-dextrose (pH 5.1, 1.4 ml ACD/8.6 ml blood) through standard venipuncture and used immediately upon collection. ACD had no effect on plasma pH level (Supplemental Methods Figure I). Plasma was harvested by sequential centrifugation of the whole blood at 150 x g for 20 minutes

(min) followed by another spin at 1500 x g for 20 minutes and then once more at 13,000 x g for 2 min to remove remaining cell contaminants.

Whole Blood Clot Formation

For RNA-seq analysis, 9.5 mL of whole blood was recalcified (20 mM final) by adding whole blood to 500 uL of 400 mM CaCl2 and gently inverting the tube. Recombinant tissue factor (1:20000, final Dade Innovin, Siemens, Tarrytown, NY) was then added

o and the whole blood was allowed to clot for 2 hours at 37 C at 5%CO2. The addition of

CaCl2 did not result in any visual hemolysis after plasma isolation. After two hours, the whole blood clot was dissociated with mechanical shear to remove the monocytes from the clot. Monocytes were isolated with ficoll-paque and positively selected for using

CD14 magnetic beads (see below). Purified monocytes were counted and then lysed in

Trizol for RNA isolation. Monocytes isolated from whole blood clots were compared to baseline monocytes (i.e. cells immediately isolated after whole blood was drawn) for

RNA seq-analysis. For cytokine studies, whole blood was allowed to clot for 18 hours

1 before the blood was centrifuged at 13000 x g for 10 minutes. After the initial centrifugation, the supernatant was centrifuged again at 13000 x g for two minutes to remove residual cells. In indicated experiments, tissue-type plasminogen activator (tPA,

American Diagnostica, Stamford, CT, USA) was added (1 μg/mL, final) at the initiation of clot formation.

Monocyte Isolation

Human monocytes were isolated by drawing human peripheral venous blood (500 ml) from healthy, medication-free donors. Blood was centrifuged at 150 x g for 20 min at

20oC to separate platelet-rich plasma (PRP) from red and white blood cells (RBC/WBC).

The RBC/ WBC mixture was resuspended with 0.9% sterile saline back to the original volume and layered over an equal volume of Ficoll-Paque Plus (GE Healthcare

Biosciences, Piscataway, NJ, USA). The layered cells were then centrifuged for 30 min at 400 x g at 20o C. After 30 min, the mononuclear leukocyte layer was removed and washed with Hank’s Balance Salt Solution (Sigma-Aldrich, St. Louis, MO, USA) with 1% human serum albumin (HBSS/A) (University of Utah Hospital, Salt Lake City, UT, USA) and centrifuged for 10 min at 400 x g at 20o C. The cell pellet was then resuspended and CD14 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) were added for

15 min at 4oC. The cells were then washed with HBSS/A to remove any free CD14 microbeads and then resuspended in HBSS/A. The monocytes were then isolated by running the cell solution through an autoMACs cell separator (Miltenyi Biotec) using the

PosselD2 program. Cells were then washed with HBSS/A and resuspended in M199

2 (BioWhitaker, Walkersville, MD, USA) and counted. The purity of monocytes was >

90%.

Purified Plasma Clot Formation

Whole blood was drawn into ACD (pH 5.1) in a ratio of 1:7 (ACD:Blood). Whole blood was then centrifuged for 20 minutes at 150 x g to isolate platelet-rich plasma (PRP).

PRP was then centrifuged for 20 minutes at 1500 x g to generate platelet-poor plasma

(PPP). Finally, PPP was centrifuged for 2 minutes at 12000 x g to generate cell-free plasma (CFP). CFP (450 µL) was added to a single well of a 24-well tissue culture treated plate and re-calcified with the addition of 25 µL of 400 mM CaCl2. Recombinant tissue factor (Dade Innovin) was added at the same time (25 µL, 1:20000 dilution final) to trigger tissue factor dependent fibrin clot formation. Clot formation was allowed to

o occur for 30 minutes at 37 C in humidified incubator with 5% CO2. Monocytes isolated using CD14 positive selection and resuspended with M199 without supplemental factors were added to the CFP clot at a final concentration of 2 x 106 total cells in a volume of

1000 µL. The final volume of clot and cells was 1500 µL. The cells and clot were

o allowed to incubate in a humidified incubator for 18 hours at 37 C at 5% CO2. In experiments with heparin, plasma from the same individual was drawn into heparinized tubes and centrifuged in a similar manner. Heparinized plasma had only Innovin added back. Plasma was sometimes treated with cycloheximide (5 µg/mL,final)(Sigma-

Aldrich), actinomycin D (5 µg/mL, final)(Sigma-Aldrich), lepirudin (University of Utah

Pharmacy), or GPRP diluted in water (Pentapharm, 40 mg/mL, final). In some assays, tPA (America Diagnostica, 1 µg/ml, final) was added at the indicated time. Fibrinogen

3 deficient plasma was from Affinity Biologicals and was reconstituted with fibrinogen from

Enzyme Research Laboratories at the indicated concentrations. FXIII deficient plasma was from Affinity Biologicals and was reconstituted with FXIIIa from Hematologic

Technologies (10 µg/mL). The FXIII inhibitor, 1,3,4,5-Tetramethyl-2-[(2- oxopropyl)thio]imidazolium chloride (T101) was from Zedira (San Diego, CA, USA) and was added to plasma clots (5 µM, final) before the addition of tissue factor and calcium.

Serum was made by adding thrombin (Sigma-Aldrich, 1.0 U/mL, final) to plasma and allowing the clot to form for an hour. After an hour, the clot was centrifuged for 10 minutes at 13000 x g to remove the clot. Thrombin-independent plasma clots were formed using Batroxobin (Pentapharm, 1 U/mL, final). In some experiments, monocytes were treated with the TLR4 inhibitor, CLI-095 (Invivogen, 1 µM, final), for 1 hour before addition to the plasma fibrin clot. In some experiments, a CD11b monoclonal antibody

(M1/70 eBiosciences) or IgG (5 µg/mL, final) was pre-incubated with monocytes before addition to plasma clots. The inhibitory nature of the antibody was tested using Alexa- fluor 555 labeled fibrinogen (eBiosciences) and measuring fibrinogen binding by flow cytometry. In some experiments, monocytes (2 x 106, final concentration) resuspended in M199 without supplemental factors1-5 and in a total volume of 1500 µL were stimulated with thrombin (0.1 U/mL, final) or lipopolysaccharide (100 ng/mL, final

Sigma-Aldrich) and IL-8 and MCP-1 measured by ELISA. For purified fibrin clot formation, monocytes (2 x 106, final concentration) resuspended in M199 and in a total volume of 1000 µL were added to 500 µL of purified fibrin clots (2 mg/mL, final Enzyme

Research Laboratories) generated by the addition of 1.0 U/mL thrombin (final).

4 FXIII crosslinking experiments

Plasma clots in the absence or presence of FXIII or in the absence or presence of T101 were formed as described above. After 60 minutes, crosslinking and clotting were stopped using a solution containing 50 mM DTT, 12.5 mM EDTA, 8 M urea. Samples were then incubated at 60°C for 1 hour. Samples were reduced, boiled, separated on

10% gels, and transferred to nitrocellulose membranes. Membranes were probed with rabbit anti-human fibrinogen polyclonal antibodies (Dako).

Radiolabeled Protein Synthesis

Monocytes were resuspended in DMEM without methionine or cysteine and allowed to rest for 30 minutes. After 30 minutes, EasyTag™ EXPRESS35S Protein Labeling Mix

(0.06 mCI total) (Perkin Elmer) was added to each reaction and allowed to incubate overnight. The next day the monocytes were washed three times in complete media and then lysed in radioimmunoprecipitation assay buffer (RIPA) (1X PBS with 1% NP-

40, 0.5% sodium deoxycholate and 0.1% sodium dodecyl sulfate). Lysates were cleared and then Trichloroacetic acid (TCA) precipitated using 20% TCA on ice for 30 minutes. The precipitated proteins were loaded onto a Whatman grade GF/C glass microfiber filter (VWR) and washed five times with 10% TCA and five times with 95% ethanol. Filter papers were then read using a liquid scintillation counter.

RNA-Seq Analysis

Monocytes were purified using immune-magnetic positive selection as described previously6, 7. For global screening of RNAs (RNA-seq), monocytes were purified from

5 individual donors. For the RNA-seq and real-time PCR analysis of mRNA expression, washed monocytes were lysed in Trizol and DNAse treated total RNA was isolated, as previously described (Rowley et al). An agilent bio-analyzer was used to Quality

Control (QC) and quantitate RNA. RNA Integrity Number (RIN) scores were similar between all samples. RNA-seq libraries were prepared with TruSeq V2 with oligo-dT selection (Illumina, San Diego CA). For whole blood clots, 36 base pair (bp) paired end reads were sequenced on an Illumina GAIIx. For all other samples, 50 bp single end reads were sequenced on an Illumina Hiseq 2000. Reads were aligned (Novoalign) to the reference genome GRCh37/hg19 and a psuedotranscriptome containing splice junctions. The USeq analysis package was used to assign reads to composite transcripts (one per gene) and quantitate FPKMs as previously described8-11. Data are deposited in Sequence Read Archive (SRA) through NCBI (SRA number

PRJNA397431).

Patient Enrollment and Thrombus Retrieval

This was a prospective cohort study of patients referred to a university tertiary-care referral center with acute thrombosis. The local institutional review board approved this study and all patients provided informed consent (IRB# 42054). Inclusion criteria included patients with acute thrombosis undergoing surgical thrombectomy. Exclusion criteria included inability to provide informed consent or obtain a thrombus specimen. All thrombus specimens were removed by a vascular surgeon, placed immediately into sterile specimen cups, and transported to the laboratory for processing. Demographic data, medical history, and clinical laboratory results were recorded for each patient.

Immunohistochemistry

Thrombi from patients were fixed with 4% paraformaldehyde for 30 minutes at room temperature before being placed in a 7.5% sucrose/4% paraformaldehyde solution for four hours. After four hours, the thrombi were placed in a 15% sucrose/4% paraformaldehyde solution overnight. The thrombi were then embedded into OCT mounting solution and froze at -80oC. The thrombi were then sectioned in 5-20 µm thick sections and stained. Clots from the in vitro clot system were embedded with OCT mounting solution and then frozen at -80oC. The clots were then sectioned in 5-20 µm thick sections and stained. Sections were permeablized using 0.1 X Triton followed by blocking with 10% goat serum. Staining was performed using a Rabbit anti-MCP-1

(Abcam 9669), Rabbit anti-IL8 (Abcam 7747), Mouse anti-CD14 (Abcam 63319), Mouse

IgG (Santa Cruz 2025), and Rabbit IgG (Abcam 171870) in 10% goat serum overnight.

The next day a goat anti-rabbit Alexa Fluor 546 and goat anti-mouse Alexa Fluor 488 were added and the slides were then imaged.

Imaging

High-resolution confocal reflection microscopy was performed with an Olympus IX81,

FV300, and a FV1000 (Olympus) confocal-scanning microscope equipped with a 60 ×

/1.42 NA oil objective for viewing platelets. An Olympus FVS-PSU/IX2-UCB camera and scanning unit and Olympus Fluoview FV 300 and FV1000 Version 5.0 image acquisition software was used for recording. The images were further analyzed with the use of

Adobe Photoshop CS Version 8.0, and ImageJ Version 1.50b (National Institutes of

Health).

7 For the ultrastructural analyses, monocytes embedded in clots were fixed in 2.5% glutaraldehyde in PBS buffer. The samples were subsequently washed and postfixed with 2% osmium tetroxide, rewashed, dehydrated by a graded series of acetone concentrations (50%, 70%, 90%, 100%; 2 × 10 minutes each), and embedded in Epon.

Thin sections were counterstained (ie, uranyl acetate and lead citrate), viewed with a

JEOL JEM-1011 electron microscope (JEOL), and digital images were captured with a side-mounted Advantage HR CCD camera (Advanced Microscopy Techniques).

Quantification of Chemokine and Cytokine Protein Expression

To assess the response of monocytes embedded in plasma clots, we used a multiplexed sandwich capture assay for the quantification of 13 chemokines and cytokines developed at the ARUP Institute for Experimental and Clinical Pathology at the University of Utah as previously described12. The chemokines or cytokines assayed included: CD40 ligand, interferon-γ, IL-10, IL-12, IL-13, IL-1β, IL-2, IL-2 receptor, IL-4,

IL-5, IL-6, IL-8, and tumor necrosis factor-α (TNF-α). To confirm and extend the chemokine results obtained using the multiplex sandwich capture assay, quantification of IL-8 and MCP-1 in supernatants and cell lysates was performed by ELISA (R&D

Systems, Inc.) as per the manufacturer's instructions. Whole blood cytokine analysis was also examined in cell-free supernatants using a Human Cytokine Array Kit (R&D

Systems, Inc.)

8 Statistical methods

For RNA-seq analysis, Deseq2 was used to identify differentially expressed transcripts.

To focus on robustly expressed transcripts, only transcripts with per group average > 3

FPKM were included in the analysis. For the plasma clot samples, a q value < .05 and fold change of 4-fold were used as thresholds for differential expression. For the whole blood clot model, assignment of differentially expressed transcripts was based on fold change of 4-fold. For relevant studies, we calculated the mean ± SEM and performed

ANOVAs to identify differences among multiple experimental groups. If significant differences existed, a Student Newman-Keuls posthoc procedure was used to determine the location of the difference between groups. When single comparisons were performed, a Student’s t-test was employed. Statistical significance was set at P

< .05.

Supplemental Methods Figure I. Plasma isolated from whole blood drawn into

ACD has a physiologic pH. Whole blood was drawn into ACD (pH 5.1; ratio 1:7 ACD to blood) and centrifuged for 20 minutes at 150 x g to isolate platelet rich plasma. Cell free plasma was then generated by sequential centrifugation for 20 minutes at 1500 x g followed by an additional 2 minutes at 12,000 x g. The pH of isolated cell free plasma was determined using pH strips. Shown are pH strips (1, 2, 3) from three independent experiments with the reference guide on the right. The color of these three experiment test strips approximated physiologic pH (i.e. pH 7.4, black arrow).

References

1. Dixon DA, Tolley ND, Bemis-Standoli K, Martinez ML, Weyrich AS, Morrow JD, Prescott SM, Zimmerman GA. Expression of cox-2 in platelet-monocyte interactions occurs via combinatorial regulation involving adhesion and cytokine signaling. J Clin Invest. 2006;116:2727-2738 2. Galt SW, Lindemann S, Medd D, Allen LL, Kraiss LW, Harris ES, Prescott SM, McIntyre TM, Weyrich AS, Zimmerman GA. Differential regulation of matrix metalloproteinase-9 by monocytes adherent to collagen and platelets. Circ Res. 2001;89:509-516 3. Hottz ED, Medeiros-de-Moraes IM, Vieira-de-Abreu A, de Assis EF, Vals-de- Souza R, Castro-Faria-Neto HC, Weyrich AS, Zimmerman GA, Bozza FA, Bozza PT. Platelet activation and apoptosis modulate monocyte inflammatory responses in dengue. J Immunol. 2014;193:1864-1872 4. Weyrich AS, Elstad MR, McEver RP, McIntyre TM, Moore KL, Morrissey JH, Prescott SM, Zimmerman GA. Activated platelets signal chemokine synthesis by human monocytes. J Clin Invest. 1996;97:1525-1534 5. Weyrich AS, McIntyre TM, McEver RP, Prescott SM, Zimmerman GA. Monocyte tethering by p-selectin regulates monocyte chemotactic protein-1 and tumor necrosis factor-alpha secretion. Signal integration and nf-kappa b translocation. J Clin Invest. 1995;95:2297-2303 6. Major HD, Campbell RA, Silver RM, Branch DW, Weyrich AS. Synthesis of sflt-1 by platelet-monocyte aggregates contributes to the pathogenesis of preeclampsia. Am J Obstet Gynecol. 2014;210:547 e541-547 7. Franks Z, Campbell RA, Vieira de Abreu A, Holloway JT, Marvin JE, Kraemer BF, Zimmerman GA, Weyrich AS, Rondina MT. Methicillin-resistant staphylococcus aureus-induced thrombo-inflammatory response is reduced with timely antibiotic administration. Thromb Haemost. 2013;109:684-695 8. Rondina MT, Tatsumi K, Bastarache JA, Mackman N. Microvesicle tissue factor activity and interleukin-8 levels are associated with mortality in patients with influenza a/h1n1 infection. Crit Care Med. 2016 9. Rowley JW, Chappaz S, Corduan A, Chong MM, Campbell R, Khoury A, Manne BK, Wurtzel JG, Michael JV, Goldfinger LE, Mumaw MM, Nieman MT, Kile BT, Provost P, Weyrich AS. Dicer1-mediated mirna processing shapes the mrna profile and function of murine platelets. Blood. 2016;127:1743-1751 10. Kahr WH, Hinckley J, Li L, Schwertz H, Christensen H, Rowley JW, Pluthero FG, Urban D, Fabbro S, Nixon B, Gadzinski R, Storck M, Wang K, Ryu GY, Jobe SM, Schutte BC, Moseley J, Loughran NB, Parkinson J, Weyrich AS, Di Paola J. Mutations in nbeal2, encoding a beach protein, cause gray platelet syndrome. Nature genetics. 2011;43:738-740 11. Rowley JW, Oler AJ, Tolley ND, Hunter BN, Low EN, Nix DA, Yost CC, Zimmerman GA, Weyrich AS. Genome-wide rna-seq analysis of human and mouse platelet transcriptomes. Blood. 2011;118:e101-111

11 12. Thornton NL, Cody MJ, Yost CC. Toll-like receptor 1/2 stimulation induces elevated interleukin-8 secretion in polymorphonuclear leukocytes isolated from preterm and term newborn infants. Neonatology. 2012;101:140-146