sign in sign up
<<
Home , DAB1, Protein C, Reeler, Reelin, VLDL receptor

Activated C ligation of ApoER2 (LRP8) causes Dab1-dependent signaling in U937 cells

Xia V. Yanga, Yajnavalka Banerjeea, Jose´ A. Ferna´ ndeza, Hiroshi Deguchia, Xiao Xua, Laurent O. Mosniera, Rolf. T. Urbanusb, Phillip G. de Grootb, Tara C. White-Adamsc, Owen J. T. McCartyc, and John H. Griffina,1

aDepartment of Molecular and Experimental , The Scripps Research Institute, La Jolla, CA 92037; bLaboratory of and Haematology, University Medical Centre Utrecht, Utrecht, The Netherlands; and cDepartments of Biomedical Engineering and Cell and , Oregon Health & Science University, Portland, OR 97239

Edited by Charles T. Esmon, Oklahoma Medical Research Foundation, Oklahoma City, OK, and approved November 18, 2008 (received for review August 2, 2008) Binding of activated (APC) to cells triggers multiple ischemic , NMDA excitotoxicity injury, and severe beneficial cytoprotective activities that suppress , inflam- models (11–14) . mation, and endothelial barrier breakdown. One paradigm for Recently, two kinds of findings imply that the APC-EPCR- APC’s signaling emphasizes its binding to endothelial cell protein PAR1 paradigm is not always applicable and that alternative cell C (EPCR) and subsequent activated receptor surface receptors might contribute to initiation of (PAR)-1 activation. Here we used human monocytic-like U937 cells by APC. First, for a few in vitro assays of APC’s effects on certain to evaluate E receptor 2 (ApoER2)-dependent sig- cells, it appeared that APC-initiated cell signaling did not naling by APC and found that APC initiated rapid require PAR1 (15, 16) or EPCR (17). Second, in a report of of Tyr-220 in the adaptor protein disabled-1 (Dab1) and of Ser-473 APC- interactions, an important role was posited for in Akt. APC also induced phosphorylation of Ser-9 in glycogen APC binding to receptor-2 [ApoER2; aka LDL synthase 3␤ (GSK3␤), which was blocked by the PI3K receptor (LDLR)-related protein 8; LRP8] (18) and to glycop- inhibitor LY294002. Receptor-associated protein (RAP), a general rotein Ib␣ (19). ApoER2 is a member of the LDLR family that antagonist for binding of ligands to LDL receptor family members, includes, among others, the LDLR, the LDLR related protein inhibited APC-induced phosphorylation of Dab1 and GSK3␤, (LRP), ApoER2, and the very low density receptor whereas anti-EPCR or anti-PAR1 blocking antibodies did not. (VLDLR) (20–23). Various LDLR family members provide Knocking down ApoER2 by using siRNA-ablated APC induced Dab1 major endocytotic activities for cells whereas other family mem- phosphorylation, suggesting that RAP-sensitive APC-induced sig- bers also play prominent roles for signal transduction. LDLR naling requires ApoER2. In surface plasmon resonance equilibrium family members can bind a spectrum of ligands on various cells, binding studies, APC bound with high affinity to soluble (s) ApoER2 not limited to , and ligation of many members of this (apparent Kd, Ϸ30 nM) but not to soluble very low density lipopro- receptor family suffices to initiate cell signaling. tein receptor. RAP blocked APC binding to sApoER2 but not to The physiologic neuronal signaling pathway involves sEPCR. RAP blocked binding of U937 cells to immobilized APC. RAP Reelin ligation of both ApoER2 and VLDLR and subsequent also blocked APC’s ability to inhibit endotoxin-induced tissue phosphorylation of the adaptor protein disabled-1 (Dab1) in- factor pro-coagulant activity of U937 cells. Thus, we propose that volving the Src family Src and Fyn, followed by phos- ligation of ApoER2 by APC signals via Dab1 phosphorylation and phorylation of phosphatidylinositol 3-phosphate kinase (PI3K), subsequent activation of PI3K and Akt and inactivation of GSK3␤, Akt, and glycogen synthase kinase 3␤ (GSK3␤) (20–23). Re- thereby contributing to APC’s beneficial effects on cells. ceptor-associated protein (RAP) (24) is an antagonist for ApoER2 ligand binding that blocks Reelin’s effects on cells. endothelial protein C receptor ͉ reelin ͉ Akt ͉ PI3K ͉ monocyte Remarkably, murine genetic deficiencies of Reelin, ApoER2 plus VLDLR, Dab1, or Src plus Fyn produce mice with indis- ecombinant activated protein C (APC) reduces mortality in tinguishable (, scrambler, yotari) (22, 23, 25–29) Rsevere sepsis (1). APC, a well known plasma protein, can involving impaired motor coordination, compromised gait, trem- exert both and cytoprotective activities by dis- ors, and , plus abnormal patterns of neuronal positioning tinctly different mechanisms (2–4). Because two potent antico- in the . Moreover, reeler adult mice exhibit learning defects agulant plasma , and pathway and have with compromised long-term potentiation inhibitor, failed to reduce mortality in large phase III trials of functioning (30, 31). severe sepsis (5, 6), one may infer that one or more of APC’s Monocytic cell lines, such as the human leukemic monoblast cytoprotective actions contribute to mortality reduction by APC U937 cell line, have been used to study APC’s ability to modulate . APC’s cytoprotective actions include inter alii anti-apo- monocyte apoptosis, phagocytosis, , and tissue ptotic and anti-inflammatory activities, alterations of ex- factor generation (4, 8, 32–34). To evaluate the potential rele- pression, pro-angiogenic activity, and endothelial barrier stabi- vance of ApoER2 for APC cell signaling mechanisms, APC- lization (4, 7, 8). initiated signaling in U937 cells was assayed based on the In vitro studies of the effects of APC on various cells support a paradigm for APC’s cell signaling based on two key receptors, Author contributions: X.V.Y., H.D., and J.H.G. designed research; X.V.Y., Y.B., J.A.F., H.D., endothelial protein C receptor (EPCR) (9) and the G protein- X.X., L.O.M., T.C.W.-A., and O.J.T.M. performed research; R.T.U. and P.G.d.G. contributed coupled receptor protease activated receptor-1 (PAR1) (10), new reagents/analytic tools; X.V.Y., Y.B., J.A.F., H.D., X.X., L.O.M., T.C.W.-A., and O.J.T.M. which mediate downstream cytoprotective effects (4). Central to analyzed data; and X.V.Y. and J.H.G. wrote the paper. this paradigm is the binding of APC’s ␥-carboxyglutamic acid- The authors declare no conflict of interest. rich N-terminal domain to the EPCR with subsequent cleavage This article is a PNAS Direct Submission. of PAR1 by APC’s protease domain, followed by PAR1-induced 1To whom correspondence should be addressed. E-mail: jgriffi[email protected]. direct effects of APC on cells. Strong support for the in vivo This article contains supporting information online at www.pnas.org/cgi/content/full/ relevance of this mechanism for triggering cell signaling by 0807594106/DCSupplemental. pharmacologic doses of APC comes from studies of murine © 2008 by The National Academy of Sciences of the USA

274–279 ͉ PNAS ͉ January 6, 2009 ͉ vol. 106 ͉ no. 1 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807594106 Downloaded by guest on September 25, 2021 hypothesis that APC ligation of ApoER2 would signal similarly to the Reelin signaling pathway wherein ApoER2 ligation by Reelin promotes signaling via phosphorylation of Dab1, which binds to an NPxY sequence of the intracellular C-terminal region of ApoER2 (21–23,26). Here we present data showing that APC binds to purified soluble (s) ApoER2, that U937 cells bind to immobilized APC, and that APC initiates RAP-sensitive phos- phorylation of Dab1 with consequent PI3K-dependent signaling via of Akt and GSK3␤. Our results indicate that APC initiates a Reelin-like signaling pathway and that ApoER2 should join EPCR and PAR1 as candidates for cell receptors or co-receptors that mediate APC’s cell signaling activities. Results U937 Cells Contain Two Forms of ApoER2. Because in various tissues determines the variable expression of ApoER2 in different tissues and different species (18, 35–38), we characterized the forms of ApoER2 mRNA in U937 cells [supporting information (SI) Fig. S1]. Two isoforms of ApoER2 were identified in U937 cells, both with deletions of 5 and 18 and one with an additional deletion of 15 that codes for the extracellular O-linked sugar region (Fig. S1C).

APC Causes Phosphorylation of Dab1, Akt, and GSK3␤. Addition of APC to U937 cells caused phosphorylation of tyrosine residues in Dab1 within 10 min (Fig. 1A). Based on immunoblotting by using antibodies specific for pTyr220, APC caused phosphory- lation of Tyr-220 in Dab1 (data not shown). Phosphorylation of Tyr residues in Dab1 was not inhibited by a (RCR252) that blocks binding of APC to EPCR or by two anti-PAR1 monoclonal antibodies (ATAP2 and WEDE15) that block PAR1 signaling by APC. However, Dab1 phosphorylation was blocked by RAP, which blocks binding of ligands to LDLR family members such as ApoER2 (Fig. 1B). Knocking down Fig. 1. APC stimulates RAP-sensitive phosphorylation of Dab-1 and GSK-3␤. ApoER2 by using siRNA ablated APC-induced Dab1 phosphor- (A) Cells were incubated with or without APC (2 ␮g/mL) for 5 or 10 min at 37 °C. ylation (Fig. 2 A and B), indicating that ApoER2 is required for Dab1 was immunoprecipitated from cell lysates and was separated on SDS/ Dab1 phosphorylation under these conditions. Similar to the PAGE and blotted with anti-phosphorylated Tyr (␣-pY) or anti-Dab1 (␣-Dab1) Reelin signaling pathway that causes activation of Akt by phos- as described in Materials and Methods.(B) U937 cells were pre-treated with phorylation of Ser-473 (39, 40) downstream of Dab1 phosphor- control, anti-EPCR mAb RCR252 (15 ␮g/mL), RAP (2 ␮M), or anti-PAR1 anti- ylation, APC caused rapid phosphorylation of Ser-473 in Akt bodies (ATAP2 plus WEDE15, 15 ␮g/mL each) for 15 min, then incubated with ␮ (Fig. 1C). Further similarly (39, 40), APC caused downstream or without APC (2 g/mL) for 10 min, and Dab1 tyrosine phosphorylation (pY) was examined as in A.(C) Cells were incubated with or without APC (2 ␮g/mL) phosphorylation of GSK3␤ on Ser-9 (Fig. 1D). The ratio of ␤ for 5 min at room temperature, and then Akt was immunoprecipitated from phosphorylated to total GSK3 (Fig. 1D) indicates that the cell lysates and analyzed by Western blot using anti-Akt-pSer473 or anti-Akt MEDICAL SCIENCES relative amount of phosphorylated increased with time as described in Materials and Methods.(D) Cells were incubated with APC (2 (Fig. 1E). Because PI3K mediates Dab1-dependent activation of ␮g/mL) for various times and cell lysates were directly subjected to Western Akt (39, 40), we used the PI3K inhibitor Ly294002 and showed blotting using anti-GSK3␤-pSer9 or anti-GSK3␤ as described in Materials and that it blocked the APC-induced phosphorylation of GSK3␤ Methods.(E) Band intensities (see D) were quantified using Un-Scan-IT (Silk (Fig. 1F). When cells were preincubated with various receptor- Scientific), and the ratio of GSK3␤-pSer9 to total GSK3␤ antigen was calcu- blocking agents before APC addition and phosphorylation of lated for each time. (F) Cells were pre-treated with LY294002 (50 ␮M) or Ser-9 in GSK3␤ at 5 min was determined, we found that only control buffer for 15 min at room temperature and then incubated with or without APC (2 ␮g/mL) for 5 min. Endogenous GSK3␤-pSer9 and GSK3␤ were RAP blocked this reaction (Fig. 1G), showing that a RAP- directly analyzed by Western blot. (G) Cells were pre-treated with anti-EPCR sensitive receptor was required for APC-induced signaling. monoclonal RCR252 (15 ␮g/mL), RAP (2 ␮M), or anti-PAR1 antibodies (ATAP2 Remarkably, antibodies that block APC binding to EPCR or and WEDE15, 15 ␮g/mL each) for 15 min, then incubated with or without APC PAR1-dependent signaling had no significant effect on this (2 ␮g/mL) for 5 min; GSK3␤-pSer9 and GSK3␤ were then analyzed by Western action of APC (Fig. 1G). blot. Western blots shown are representative of at least three independent experiments. RAP Inhibits Binding of APC to sApoER2 But Not to sEPCR. Previously, by using a non-equilibrium binding method, we showed that APC binding of the Reelin signaling pathway receptor sVLDLR to binds to sApoER2 immobilized on a microtiter plate (19). Here APC was detected by SPR (see Fig. 3A Inset) or by solid-phase we used the equilibrium binding method of surface plasmon binding assays (Fig. S2). Thus, RAP blocks high-affinity binding resonance (SPR) analysis to compare parameters for APC of APC to sApoER2, consistent with the hypothesis that the binding to sApoER2 and to sEPCR. sApoER2 bound with RAP-sensitive, APC-induced signaling in U937 cells is caused by apparently higher affinity to APC (34 nM) than sEPCR (195 APC ligation of ApoER2 and not by binding of APC to VLDLR. nM), and RAP completely blocked binding of sApoER2 to APC but not binding of sEPCR (Fig. 3). The kinetics of APC binding RAP and Anti-EPCR Antibodies Inhibit Binding of U937 Cells to Immo- to sApoER2 and sEPCR showed remarkable differences, with bilized APC. To determine if RAP blocks binding of U937 cells to much faster on-rates and off-rates for the latter receptor. No immobilized APC, cells were preincubated with RAP or an

Yang et al. PNAS ͉ January 6, 2009 ͉ vol. 106 ͉ no. 1 ͉ 275 Downloaded by guest on September 25, 2021 Fig. 2. Knocking down ApoER2 expression by using siRNA ablates APC- induced Dab1 phosphorylation. (A) U937 Cells were transfected with ApoER2 siRNA or control siRNA for 48 h before incubation with or without APC (2 ␮g/mL) for 10 min at 37 °C. Dab1 was immunoprecipitated from cell lysates (IP:Dab1) containing equal amount of total protein and was separated on SDS/PAGE and immunoblotted with anti-phosphorylated Tyr (WB:␣-pY) or anti-Dab1 (WB:␣-Dab1) as described in Materials and Methods.(B) Knock- down of ApoER2 gene and protein expression was confirmed by semi- quantitative RT-PCR and Western blot analyses. Total RNA was isolated from U937 cells treated with ApoER2 siRNA or control siRNA and reverse- transcribed to cDNA as described in Fig. S1. ␤- gene and protein expres- sion was analyzed in the same samples to ensure specific knock-down of ApoER2. For protein expression analysis, ApoER2 was immunoprecipitated from whole-cell lysates containing equal amounts of total protein using rabbit polyclonal antibody raised against the C terminus of ApoER2 (Santa Cruz Biotechnology) and subjected to Western blot analysis using a mono- clonal antibody recognizing the N terminus of ApoER2 (Abcam). The same whole-cell lysates were also subjected to Western blotting using anti-␤-actin antibody (Sigma). Data shown are representative of results from three inde- pendent experiments.

anti-EPCR antibody (RCR252) that blocks APC binding or control IgG, and then cells were incubated in chambers con- Fig. 3. Binding of sApoER2 and sEPCR to APC and effect of RAP on APC taining APC or fibronectin that had been immobilized on the binding. WT APC with a V5 tag at its carboxyl terminus was captured on a CM5 chip using an immobilized anti-V5-antibody, and SPR was used to monitor surface. Following vigorous washing, the binding of cells to the binding of soluble receptor domains to APC and to determine the effects of surface was quantified by using phase-contrast microscopy. Cells RAP on receptor binding. (A) Binding of sApoER2 (117, 138, 159, 191, 255, and were bound similarly to APC or fibronectin but not to BSA on 308 nM) to APC was monitored by SPR. RAP (600 nM; dotted line) completely the surface (Fig. 4A). RAP and the blocking anti-EPCR blocked binding of sApoER2 (300 nM) to APC. No binding of sVLDLR (705 nM) RCR252 antibody ablated binding of cells to immobilized APC to APC was observed (Inset). (B) Binding of sEPCR (0.08, 0.17, 0.35, 0.7, and 1.4 but not to fibronectin (Fig. 4 A and B). Our data show that both ␮M) to APC was monitored. No change in binding of sEPCR (1.4 ␮M) to APC was ␮ a RAP-sensitive receptor and EPCR were required for static observed in the presence of RAP (2 M), as indicated by the dotted line that was indistinguishable from the solid line for sEPCR alone (1.4 ␮M). Rate adhesion of U937 cells to immobilized APC after vigorous constants k and k for binding of each soluble receptor to APC are shown washing. on off (Inset), along with the calculated apparent dissociation constant KDapp.

RAP and Anti-EPCR Antibodies Inhibit APC’s Ability to Block LPS- Induced Pro-Coagulant Activity of U937 Cells. To see if ApoER2 domain of sApoER2 (apparent Kd, Ϸ30 nM) but not to a closely might contribute to APC’s previously demonstrated ability to related LDL receptor family member, sVLDLR. Several iso- down-regulate tissue factor activity that appears following U937 forms of ApoER2 are abundant in brain and testis (18) and also cell stimulation with LPS (33), we pre-treated cells with various present in endothelial cells (41) and (42), as well as the blocking antibodies or RAP and then treated cells with LPS and U937 monocytic cell line. We hypothesized that APC ligation of APC. When pro-coagulant activity (PCA) of reaction mixtures ApoER2 on U937 cells promotes Reelin-like signaling that is was measured at6hafterLPSstimulation, APC diminished well described in neurons. Reelin ligation of ApoER2 causes LPS-induced PCA, as expected (Fig. 5). RAP alone or the phosphorylation of the adaptor protein Dab1 by Src family blocking anti-EPCR RCR252 antibody alone totally ablated kinases (Src and Fyn), followed by activation of PI3K and Akt; APC’s effects (Fig. 5). However, two antibodies (ATAP2 and and, in a key downstream step, GSK3␤ is phosphorylated and WEDE15) that block APC activation of PAR1 had no effect on thereby inhibited by activated Akt (22, 23, 43–45). APC’s ability to blunt LPS induction of tissue factor activity (Fig. Here we show that addition of APC to U937 cells initiated 5). The observed ability of the blocking anti-EPCR antibody to rapid tyrosine phosphorylation of Dab1 and of Ser-473 in Akt. prevent APC’s down-regulation of tissue factor activity is similar APC also induced phosphorylation of Ser-9 in GSK3␤, which to the findings of Shu et al. (33). Thus, both an RAP-sensitive was blocked by the PI3K inhibitor LY294002. RAP, an antag- receptor and EPCR were required for APC to down-regulate onist for ApoER2 ligand binding as well as a general antagonist tissue factor PCA on LPS-stimulated cells. for ligand binding to LDLR family members, inhibited APC- induced phosphorylation of Dab1 and GSK3␤, whereas anti- Discussion EPCR or anti-PAR1 blocking antibodies did not. Moreover, APC has multiple signaling effects that may involve multiple binding of APC to sApoER2 in purified systems, like APC- receptors and cell signaling pathways. In equilibrium binding induced signaling in U937 cells, was blocked by RAP, whereas studies, APC bound with high affinity to the extracellular RAP had no effect on binding of APC to sEPCR. Direct

276 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807594106 Yang et al. Downloaded by guest on September 25, 2021 Fig. 5. RAP and anti-EPCR antibody (RCR252) blocked APC’s ability to down-regulate LPS-induced PCA. Cells were incubated with buffer (ᮀ), LPS alone (25 ng/mL; o), or LPS (25 ng/mL) plus APC (1 ␮g/mL; ■) as described in Materials and Methods, and PCA was determined. Data are expressed as percentages of control PCA (cell PCA in the absence of LPS exposure or treatment with RAP or antibodies). Values represent mean Ϯ SD for data from four independent experiments.

EPCR, such a situation is not inconsistent with data that suggest that EPCR binds to only APC’s N-terminal Gla-domain (46, 47) whereas sApoER2 binds to Gla-domain-less APC (19). One possibility is that the distinct APC binding characteristics for Gla-domain-dependent EPCR binding (i.e., fast association) combined with Gla-domain-independent ApoER2 binding (i.e., slow dissociation) allow for the temporal retention of APC on the cell membrane beyond what can be accomplished by either Fig. 4. ApoER2 and EPCR mediate U937 cell binding to immobilized APC. U937 cells that were bound to Surface-coated APC or Surface-coated receptor alone. It is also possible that EPCR binds APC to cells fibronectin (FN) were visualized (A) and quantified (B) using phase-contrast in a manner that permits APC-induced signaling via other microscopy. Pre-treatment of cells with vehicle or RAP or an anti-EPCR currently unidentified receptors. antibody (RCR252) that blocks APC binding to EPCR was performed as Depending on cell type, APC induces signaling via the MAPK described in Materials and Methods. Data are shown as mean Ϯ SEM. In pathway with phosphorylation of ERK-1/2 that is reported to be controls using vehicle with BSA, only minimal numbers of U937 cells were PAR-1-dependent and EPCR-dependent (48–51) or PAR1- observed on BSA-coated wells (small gray bar; 2.8 Ϯ 0.4 cells/mm2; n ϭ 5; *, dependent but EPCR-independent (17). Multiple members of P Ͻ 0.01). the MAPK pathway can strongly influence a cell’s fate for better or for worse, and there are multiple modes of regulating members of the MAPK pathway (52–57). In U937 cells, data evidence of a required role for ApoER2 for APC’s signaling

here show that APC causes phosphorylation of Ser-473 in Akt, MEDICAL SCIENCES came from siRNA studies showing that specific knock-down of a key reaction for Akt activation (58), and that APC can induce ApoER2 ablated APC-induced Dab1 phosphorylation. Notably, signaling via ApoER2 and the PI3K-Akt pathway without any neither EPCR nor PAR1 appeared necessary for APC’s signal- apparent requirement for EPCR or PAR1. Akt is a major node ing effects via Dab1. Thus, the major hypothesis supported by in the cell’s signaling communications network with multiple our data is that APC ligation of ApoER2 causes Dab1- major downstream targets that may modulate cell survival, dependent signaling via the PI3K/Akt pathway in U937 cells. proliferation, metabolism, cell cycle regulation, and angiogenesis Acting via an EPCR-dependent process, APC prevents elab- (58, 59). Two major downstream Akt targets are GSK3␤ and oration of tissue factor activity on the surface of activated U937 endothelial NO synthase, and APC induces phosphorylation of cells (33). Remarkably, RAP effectively blocked APC’s ability to GSK3␤ and of endothelial NO synthase in a PI3K-dependent inhibit endotoxin induction of tissue factor PCA of U937 cells, manner (49). Thus, APC can cause alterations of at least two implicating a requirement for a RAP-sensitive receptor such as major downstream targets of the PI3K-Akt pathway, with some ApoER2. Anti-EPCR antibodies that block binding of APC to broad implications (58, 59). Studies indicate that the PI3K/Akt EPCR on U937 cells prevented APC’s effects on tissue factor pathway activation can down-regulate LPS-induced PCA or induction as reported (33), implying that binding of APC to both tissue factor expression in monocytes (60, 61). These reports are EPCR and ApoER2 on the cell surface is necessary for effective consistent with the implication that ligation of ApoER2 by APC APC signaling needed to blunt tissue factor induction. Similarly, with subsequent activation of the PI3K/Akt pathway may help either RAP or anti-EPCR antibodies blocked binding of U937 explain the mechanism for APC’s ability to down-regulate cells to immobilized APC. As the assay for cell binding to LPS-induced PCA. It will be interesting to clarify the potential immobilized APC involved vigorous washing steps, these data do effects of APC signaling via ApoER2 and the PI3K-Akt pathway not reflect a simple equilibrium binding reaction, and no defin- and how this might be integrated with EPCR/PAR1 pathway itive conclusion can be drawn regarding receptor-receptor in- signaling. teractions or regarding the energetic contributions of either In addition to reducing death by sepsis in humans and murine individual receptor binding to APC. Although no data directly models (1, 14), APC has remarkable in vivo neuroprotective indicate that APC simultaneously binds to both ApoER2 and activities in murine brain injury model studies (11–13, 62, 63).

Yang et al. PNAS ͉ January 6, 2009 ͉ vol. 106 ͉ no. 1 ͉ 277 Downloaded by guest on September 25, 2021 Identification of an ApoER2-dependent, Reelin-like signaling (Sigma). Appropriate secondary horseradish peroxidase-linked antibodies pathway for APC on U937 cells that activates the PI3K-Akt were used and developed with SuperSignal West Pico Chemiluminescent pathway raises questions about potential roles for this new APC Substrate (Pierce) and exposed to CL-XPosure film (Pierce). signaling mechanism for neurons and other brain cells. Inter- RNA Interference. ApoER2 expression in U937 cells was knocked down using a estingly, a number of studies implicate the Akt survival pathway pool of three target-specific 20- to 25-nucleotide siRNAs (Santa Cruz Biotech- as relevant for survival of neurons after ischemic stroke (re- nology) following the manufacturer’s instructions. A non-targeting 20- to viewed in ref. 64). Also relevant to such future investigations of 25-nucleotide siRNA consisting of a scrambled sequence from the same com- ApoER2-dependent effects of APC on neurons is the direct pany was used as a negative control. interaction of ApoER2 with the NMDA receptor (65, 66) which could provide an additional mechanism for APC’s actions on SPR Analysis. Binding was assessed by SPR using a BIAcore 3000 biosensor neurons. system. An anti-V5 tag antibody (V5-AB; ICL) was covalently immobilized on a Engagement of receptors and signaling pathways that are CM5 sensor chip (BIAcore) and saturated with WT APC with a V5 tag at its Ϫ stimulated by APC (e.g., G protein-coupled receptors such as COOH terminal. The kinetics of APC binding to sApoER2 or sEPCR were determined as described in SI Text. PAR1 or PAR3 and the MAPK pathway or ApoER2 and a Reelin-like signaling pathway that includes Dab1 phosphoryla- U937 Assays. Polystyrene 24-well plates were incubated with tion and the PI3K-Akt pathway) generates multiple signals that APC or fibronectin (100 ␮g/mL) for 60 min at room temperature followed by are interconnected components of complex systems whose out- washing and blocking with 2% -free BSA for 60 min, then followed put must be integrated for cellular outputs and survival. The net by washing with PBS solution. U937 cells (1 ϫ 106/mL in Hepes buffered saline physiologic or pharmacological effects of APC’s signaling must solution HBSS buffer containing 1% BSA) were incubated in APC-coated or be ultimately determined by the context of each cell, each tissue, fibronectin-coated wells for 60 min at 37 °C. Then, wells were washed with each blood vessel, and each animal. As shown here, the addition HBSS and fixed with paraformaldehyde for 10 min and the number of bound U937 cells was quantified via phase-contrast microscopy. In selected experi- of a novel ApoER2-dependent signaling pathway to the list of ments, U937 cells were pre-treated for 10 min with RAP (40 ␮g/mL), a rat potential mechanisms for APC’s signaling activities should con- anti-EPCR mAb RCR-252 that blocks APC binding (50 ␮g/mL), or control rat tribute significantly to efforts to decipher mechanisms that IgG1 mAb (50 ␮g/mL) for 10 min. explain how APC can signal in various cells and tissues to prevent apoptosis, induce selective alterations of , reduce Analysis of Tissue Factor PCA. Reaction mixtures containing U937 cells sus- inflammation, stabilize endothelial barriers, provide neuropro- pended in RPMI were preincubated with RAP (2 ␮M), non-immune mouse IgG tection following ischemic stroke, minimize damage in chronic (10 ␮g/mL), anti-EPCR RCR252 antibody (10 ␮g/mL), or two anti-PAR-1 anti- ␮ ␮ neurodegenerative disease animal models, and reduce death bodies (ATAP2 and WEDE15, each at 10 g/mL) for 30 min, and then 1 g/mL APC was added and incubated for 20 min. Then the reaction mixtures con- from severe sepsis. taining U937 cells were stimulated with LPS (Sigma; 055:B5) for 6 h. PCA in the reaction mixture was determined by a single-stage clotting Materials and Methods assay as described (33, 67). Briefly, 50 ␮L of U937 suspension was mixed with Recombinant Proteins. Human GST-RAP, sApoER2, sEPCR, WT APC, and APC-V5 50 ␮L of 25 mM CaCl2. Clotting was initiated by the addition of 50 ␮L of normal were prepared as described in SI Text. human pooled plasma, and the clotting time at 37 °C was recorded using an Amelung KC4 coagulometer (Sigma). The PCA standard curve was calibrated Analysis of Dab1, Akt, and GSK-3␤ Phosphorylation. U937 cells (ATCC) treated using Innovin (Dade). with or without APC were harvested in radio-immunoprecipitation assay buffer with a mixture of protease and phosphatase inhibitor cocktails (Pierce). ACKNOWLEDGMENTS. We thank Ms. Phuong M. Nguyen, Ms. Sarah K. Coit, Dab1 or Akt was immunoprecipitated from 100 ␮g lysate with goat polyclonal and Ms. Tal Eshel for technical assistance. We thank Dr. D. Strickland (Univer- anti-Dab1 antibody (Santa Cruz Biotechnology) or rabbit polyclonal anti-Akt sity of Maryland, Baltimore, MD) for the generous gift of sVLDLR and anti- antibody (Cell Signaling). Equal amounts of immunoprecipitated samples or sVLDLR antibodies, Dr. S. Gonias (University of California, San Diego, CA) for the human GST-RAP construct, and Dr. L. Brass (University of Pennsylvania, cell lysates were subjected to Western blot analysis using polyclonal or mono- Philadelphia, PA) for anti-PAR-1 antibodies. This work was supported in part clonal antibodies directed against phosphorylated Tyr (pY) (Invitrogen), Dab-1 by National Institutes of Health Grants HL31950 and HL52246 (to J.H.G) and (Santa Cruz Biotechnology), Akt-pSer473 (Cell Signaling), Akt (Cell Signaling), HL087618 (to L.O.M.), and American Heart Association Grant 0665512Z (to GSK3␤-pSer9 (Cell Signaling), GSK3␤ (Santa Cruz Biotechnology), or ␤-actin O.J.T.M.).

1. Bernard GR, et al. (2001) Efficacy and safety of recombinant human activated protein 15. Yuda H, et al. (2004) Activated protein C inhibits bronchial hyperresponsiveness and C for severe sepsis. N Engl J Med 344:699–709. Th2 expression in mice. Blood 103:2196–2204. 2. Esmon CT, et al. (1999) Inflammation, sepsis, and . Haematologica 84:254– 16. Finigan JH, et al. (2005) Activated protein C mediates novel endothelial barrier 259. enhancement: role of sphingosine 1-phosphate receptor transactivation. J Biol Chem 3. Dahlback B, Villoutreix BO (2005) The anticoagulant protein C pathway. FEBS Lett 280:17286–17293. 579:3310–3316. 17. O’Brien LA, et al. (2007) Activated protein C decreases tumor factor related 4. Mosnier LO, Zlokovic BV, Griffin JH (2007) The cytoprotective protein C pathway. Blood apoptosis-inducing ligand by an EPCR- independent mechanism involving Egr-1/Erk- 109:3161–3172. 1/2 activation. Arterioscler Thromb Vasc Biol 27:2634–2641. 5. Warren BL, et al. (2001) Caring for the critically ill patient: high-dose antithrombin III 18. Kim DH, et al. (1996) Human apolipoprotein E receptor 2. J Biol Chem 271:8373–8380. in severe sepsis: a randomized controlled trial. JAMA 286:1869–1878. 19. White TC, et al. (2008) Protein C supports platelet binding and activation under flow: 6. Abraham E, et al. (2003) Efficacy and safety of tifacogin (recombinant tissue factor role of Ib and apolipoprotein E receptor 2. J Thromb Haemost 6:995–1002. pathway inhibitor) in severe sepsis: a randomized controlled trial. Jama 290:238–247. 20. Strickland DK, Gonias SL, Argraves WS (2002) Diverse roles for the LDL receptor family. 7. Mosnier LO, Gale AJ, Yegneswaran S, Griffin JH (2004) Activated protein C variants with Trends Endocrinol Metab 13:66–74. normal cytoprotective but reduced anticoagulant activity. Blood 104:1740–1744. 21. Schneider WJ, Nimpf J (2003) LDL receptor relatives at the crossroad of and 8. Toltl LJ, Shin LY, Liaw PC (2007) Activated protein C in sepsis and beyond: update 2006. signaling. Cell Mol Life Sci 60:892–903. Front Biosci 12:1963–1972. 22. Herz J, Chen Y (2006) Reelin, lipoprotein receptors and . Nat Rev 9. Esmon CT, et al. (1999) Endothelial protein C receptor. Thromb Haemost 82:251–258. Neurosci 7:850–859. 10. Coughlin SR (2000) signalling and protease-activated receptors. Nature 23. Stolt PC, Bock HH (2006) Modulation of lipoprotein receptor functions by intracellular 407:258–264. adaptor proteins. Cell Signal 18:1560–1571. 11. Cheng T, et al. (2003) Activated protein C blocks -mediated apoptosis in ischemic 24. Lee D, et al. (2007) The structure of receptor-associated protein (RAP). Protein Sci human brain and is neuroprotective. Nat Med 9:338–342. 16:1628–1640. 12. Guo H, et al. (2004) Activated protein C prevents neuronal apoptosis via protease 25. D’Arcangelo G, et al. (1995) A protein related to proteins deleted activated receptors 1 and 3. 41:563–572. in the mouse mutant reeler. Nature 374:719–723. 13. Cheng T, et al. (2006) Activated protein C inhibits tissue -induced 26. Ware ML, et al. (1997) Aberrant splicing of a mouse disabled homolog, mdab1, in the brain hemorrhage. Nat Med 12:1278–1285. scrambler mouse. Neuron 19:239–249. 14. Kerschen EJ, et al. (2007) Endotoxemia and sepsis mortality reduction by non- 27. Trommsdorff M, et al. (1999) Reeler/Disabled-like disruption of neuronal migration in anticoagulant activated protein C. J Exp Med 204:2439–2448. knockout mice lacking the VLDL receptor and ApoE receptor 2. Cell 97:689–701.

278 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807594106 Yang et al. Downloaded by guest on September 25, 2021 28. Howell BW, et al. (2000) Dab1 tyrosine phosphorylation sites relay positional signals 47. Oganesyan V, et al. (2002) The crystal structure of the endothelial protein C receptor during mouse brain development. Curr Biol 10:877–885. and a bound . J Biol Chem 277:24851–24854. 29. Kuo G, Arnaud L, Kronstad-O’Brien P, Cooper JA (2005) Absence of Fyn and Src causes 48. Riewald M, et al. (2002) Activation of endothelial cell protease activated receptor 1 by a reeler-like . J Neurosci 25:8578–8586. the protein C pathway. Science 296:1880–1882. 30. Weeber EJ, et al. (2002) Reelin and ApoE receptors cooperate to enhance hippocampal 49. Uchiba M, et al. (2004) Activated protein C induces endothelial cell proliferation by synaptic plasticity and learning. J Biol Chem 277:39944–39952. mitogen-activated protein kinase activation in vitro and angiogenesis in vivo. Circ Res 31. Beffert U, et al. (2006) Functional dissection of Reelin signaling by site-directed 95:34–41. disruption of disabled-1 adaptor binding to apolipoprotein e receptor 2: distinct roles 50. Xue M, et al. (2005) Endothelial protein C receptor and protease-activated receptor-1 in development and synaptic plasticity. J Neurosci 26:2041–2052. mediate induction of a wound-healing phenotype in human keratinocytes by acti- 32. Joyce DE, Grinnell BW (2002) Recombinant human activated protein C attenuates the vated protein C. J Invest Dermatol 125:1279–1285. inflammatory response in endothelium and monocytes by modulating nuclear factor- 51. Xue M, Campbell D, Jackson CJ (2007) Protein C is an autocrine growth factor for kappaB. Crit Care Med 30:S288–S293. human skin keratinocytes. J Biol Chem 282:13610–13616. 33. Shu F, et al. (2000) Activated protein C suppresses tissue factor expression on U937 cells 52. Kolch W (2005) Coordinating ERK/MAPK signalling through scaffolds and inhibitors. in the endothelial protein C receptor-dependent manner. FEBS Lett 477:208–212. Nat Rev Mol Cell Biol 6:827–837. 34. Stephenson DA, Toltl LJ, Beaudin S, Liaw PC (2006) Modulation of monocyte function 53. Frebel K, Wiese S (2006) Signaling molecules essential for neuronal survival and by activated protein C, a natural anticoagulant. J Immunol 177:2115–2122. differentiation. Biochem Soc Trans 34:1287–1290. 35. Kim DH, et al. (1997) Exon/intron organization, localization, alternative 54. Mor A, Philips MR (2006) Compartmentalized Ras/MAPK signaling. Annu Rev Immunol splicing, and transcription units of the human apolipoprotein E receptor 2 gene. J Biol 24:771–800. Chem 272:8498–8504. 55. Simo S, et al. (2007) Reelin induces the detachment of postnatal 36. Kim HJ, et al. (1998) Evolution of the apolipoprotein E receptor 2 gene by exon loss. cells and the expression of the Egr-1 through Erk1/2 activation. Cereb Cortex 17:294– J Biochem 124:451–456. 303. 37. Brandes C, et al. (1997) Avian and murine LR8B and human apolipoprotein E receptor 56. Raman M, Chen W, Cobb MH (2007) Differential regulation and properties of MAPKs. 2: differentially spliced products from corresponding . Genomics 42:185–191. Oncogene 26:3100–3112. 38. Brandes C, et al. (2001) Alternative splicing in the ligand binding domain of mouse 57. Shaul YD, Seger R (2007) The MEK/ERK cascade: from signaling specificity to diverse ApoE receptor-2 produces receptor variants binding Reelin but not alpha 2-macro- functions. Biochim Biophys Acta 1773:1213–1226. globulin. J Biol Chem 276:22160–22169. 58. Manning BD, Cantley LC (2007) AKT/PKB signaling: navigating downstream. Cell 39. Bock HH, et al. (2003) Phosphatidylinositol 3-kinase interacts with the adaptor protein 129:1261–1274. Dab1 in response to Reelin signaling and is required for normal cortical lamination. 59. Engelman JA, Luo J, Cantley LC (2006) The evolution of phosphatidylinositol 3-kinases J Biol Chem 278:38772–38779. as regulators of growth and metabolism. Nat Rev Genet 7:606–619. 40. Jossin Y, Goffinet AM (2007) Reelin signals through phosphatidylinositol 3-kinase and 60. Schabbauer G, et al. (2004) PI3K-Akt pathway suppresses coagulation and inflamma- Akt to control cortical development and through mTor to regulate dendritic growth. tion in endotoxemic mice. Arterioscler Thromb Vasc Biol 24:1963–1969. Mol Cell Biol 27:7113–7124. 61. Luyendyk JP, et al. (2008) Genetic analysis of the role of the PI3K-Akt pathway in 41. Korschineck, I et al. (2001) Identification of a novel exon in apolipoprotein E receptor -induced cytokine and tissue factor gene expression in monocytes/ 2 leading to alternatively spliced mRNAs found in cells of the vascular wall but not in . J Immunol 180:4218–4226. neuronal tissue. J Biol Chem 276:13192–13197. 62. Liu D, et al. (2004) Tissue plasminogen activator neurovascular toxicity is controlled by 42. Riddell DR, et al. (1999) Identification and characterization of LRP8 (apoER2) in human activated protein C. Nat Med 10:1379–1383. blood platelets. J Lipid Res 40:1925–1930. 63. Han MH, et al. (2008) Proteomic analysis of active multiple sclerosis lesions reveals 43. Rice DS, Curran T (2001) Role of the Reelin signaling pathway in therapeutic targets. Nature 451:1076–1081. development. Annu Rev Neurosci 24:1005–1039. 64. Zhao H, Sapolsky R, Steinberg G (2006) Phosphoinositide-3-kinase/Akt survival signal 44. Grimes CA, Jope RS (2001) The multifaceted roles of glycogen synthase kinase 3␤ in pathways are implicated in neuronal survival after stroke. Mol Neurobiol 34:249–269. cellular signaling. Progr Neurobiol 65:391–426. 65. Hoe HS, et al. (2006) Apolipoprotein E receptor 2 interactions with the N-methyl-D- 45. Cohen P, Goedert M (2004) GSK3 inhibitors: development and therapeutic potential. aspartate receptor. J Biol Chem 281:3425–3431. Nat Rev Drug Discov 3:479–487. 66. Rebeck GW, et al. (2006) The generation and function of soluble apoE receptors in the 46. Regan LM, Mollica JS, Rezaie AR, Esmon CT (1997) The interaction between the CNS. Mol Neurodegen 1:15. endothelial cell protein C receptor and protein C is dictated by the gamma- 67. Deguchi H, et al. (1997) Dilazep, an antiplatelet agent, inhibits tissue factor expression carboxyglutamic acid domain of protein C. J Biol Chem 272:26279–26284. in endothelial cells and monocytes. Blood 90:2345–2356. MEDICAL SCIENCES

Yang et al. PNAS ͉ January 6, 2009 ͉ vol. 106 ͉ no. 1 ͉ 279 Downloaded by guest on September 25, 2021