The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

Contents lists available at ScienceDirect

The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel

2-Arachidonoylglycerol modulates human endothelial cell/leukocyte interactions by controlling selectin expression through CB1 and CB2 receptors Valeria Gasperi a,1 , Daniela Evangelista a,1 , Valerio Chiurchiù b,c , Fulvio Florenzano d , Isabella Savini a , Sergio Oddi b,c , Luciana Avigliano a , Maria Valeria Catani a,∗,2 , Mauro Maccarrone c,e,∗∗,2 a

Department of Experimental Medicine & Surgery, Tor Vergata University of Rome, Via Montpellier 1, 00133 Rome, Italy Faculty of Veterinary Medicine, University of Teramo, Piazza A. Moro 45, 64100 Teramo, Italy c European Center for Brain Research (CERC)/IRCCS Santa Lucia Foundation, Via del Fosso di Fiorano 64-65, 00143 Rome, Italy d European Brain Research Institute (EBRI), Via del Fosso di Fiorano 64-65, 00143 Rome, Italy e Center of Integrated Research, Campus Bio-Medico University of Rome, Via Alvaro del Portillo 21, 00128 Rome, Italy b

a r t i c l e

i n f o

Article history: Received 15 October 2013 Received in revised form 10 March 2014 Accepted 28 March 2014 Available online 8 April 2014 Keywords: Endocannabinoids Endothelium/leucocyte interaction TNF-␣ Kinases Selectins

a b s t r a c t Accumulated evidence points to a key role for endocannabinoids in cell migration, and here we sought to characterize the role of these substances in early events that modulate communication between endothelial cells and leukocytes. We found that 2-arachidonoylglycerol (2-AG) was able to initiate and complete the leukocyte adhesion cascade, by modulating the expression of selectins. A short exposure of primary human umbilical vein endothelial cells (HUVECs) to 2-AG was sufficient to prime them towards an activated state: within 1 h of treatment, endothelial cells showed time-dependent plasma membrane expression of P- and E-selectins, which both trigger the initial steps (i.e., capture and rolling) of leukocyte adhesion. The effect of 2-AG was mediated by CB1 and CB2 receptors and was long lasting, because endothelial cells incubated with 2-AG for 1 h released the pro-inflammatory cytokine tumour necrosis factor-␣ (TNF-␣) for up to 24 h. Consistently, TNF-␣-containing medium was able to promote leukocyte recruitment: human Jurkat T cells grown in conditioned medium derived from 2-AG-treated HUVECs showed enhanced L-selectin and P-selectin glycoprotein ligand-1 (PSGL1) expression, as well as increased efficiency of adhesion and trans-migration. In conclusion, our in vitro data indicate that 2-AG, by acting on endothelial cells, might indirectly promote leukocyte recruitment, thus representing a potential therapeutic target for treatment of diseases where impaired endothelium/leukocyte interactions take place. © 2014 Elsevier Ltd. All rights reserved.

Abbreviations: ABHD6, ␣/␤-hydrolase domain 6; APC, allophycocyanin; 2-AG, 2-Arachidonoylglycerol; DAPI, 4 ,6-diamidino-2-phenylindole; CB1 , type-1 cannabinoid receptor; CB2 , type-2 cannabinoid receptor; eCBs, endocannabinoids; ECL, enhanced chemiluminescence; ELISA, enzyme-linked immunosorbent assay; ERK, extracellular signal-regulated kinase; FAAH, fatty acid amide hydrolase; LPS, lipopolysaccharide; MAPK, mitogen-activated protein kinase; MAGL, monoacylglycerol lipase; PI3 K/PKB or Akt, phosphatidylinositol-3-kinase/protein kinase B; PBS, phosphate buffered saline; PE, phycoerythrin; PUFA, polyunsaturated fatty acid; PSGL-1, P-selectin glycoprotein ligand-1; RT-PCR, reverse transcriptase-polymerase chain reaction; sL-selectin, soluble L-selectin; TNF-␣, tumour necrosis factor-␣; TNFR, tumour necrosis factor receptor. ∗ Corresponding author at: Department of Experimental Medicine & Surgery, Tor Vergata University of Rome, Via Montpellier 1, 00133 Rome, Italy. Tel.: +39 06 72596463; fax: +39 06 72596465. ∗∗ Corresponding author at: Center of Integrated Research, Campus Bio-Medico University of Rome, Via Alvaro del Portillo 21, 00128 Rome, Italy. Tel.: +39 06 2254 19169; fax: +39 06 2254 1456. E-mail addresses: [email protected] (M.V. Catani), [email protected] (M. Maccarrone). 1 Equally first authors. 2 Equally senior authors. http://dx.doi.org/10.1016/j.biocel.2014.03.028 1357-2725/© 2014 Elsevier Ltd. All rights reserved.

80

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

1. Introduction Cell migration is crucial in several pathophysiological processes, such as immune response, inflammation, angiogenesis and tumour metastasis (Ley et al., 2007; Polacheck et al., 2013). In particular, adhesiveness and migration of leukocytes on endothelial cells involve sequential steps, all orchestrated by specific adhesion molecules that include selectins, chemokines and integrins (Ley et al., 2007). Among these compounds, P-, E- and L-selectins trigger the initial recruitment of leukocytes to the activated endothelium, and their subsequent rolling (Langer and Chavakis, 2009; Ley, 2003). Following specific stimuli, P-selectin quickly translocates from Weibel-Palade bodies to the cell surface of endothelial cells (Ley et al., 1995). E-selectin, also expressed in endothelial cells, is transcriptionally activated by pro-inflammatory cytokines, such as tumour necrosis factor-␣ (TNF-␣) and interleukin-1 (Bevilacqua et al., 1987; Ley, 2003). L-selectin, constitutively expressed by leukocytes, is responsible for lymphocyte binding to endothelial venules in lymphonodes, and for invasion of neutrophils in outbreaks of infection and inflammation (Arbones et al., 1994). Following cellular activation, leukocytes increase expression of L-selectin, which is then shed by proteolytic cleavage into a soluble form that inhibits leukocyte adhesion, and hence modulates the speed of leukocyte rolling (Hafezi-Moghadam et al., 2001; Smalley and Ley, 2005). Selectins bind to specific ligands through weak interactions, which enable leukocytes to roll on endothelium; among them, the P-selectin glycoprotein ligand-1 (PSGL-1), constitutively expressed by leukocytes, binds all three members of the selectin family (Moore, 1998). It is widely accepted that endocannabinoids (eCBs) play an important role in cell migration. In particular, the marked expression of type-2 cannabinoid (CB2 ) receptor in tissues (tonsils, spleen) and cells (mast cells, B cells, macrophages, NK cells) of the immune system suggests its major implication in immune responses (Graham et al., 2010). Understanding the role of eCBs as immunomodulators is a rather complex task, because of the tight dose-dependence of cellular effects and of the heterogeneity of these endogenous compounds (Croxford and Yamamura, 2005; Gokoh et al., 2005; Kishimoto et al., 2006; Rajesh et al., 2007). The major endogenous ligand of CB1 and CB2 receptors, 2-arachidonoylglycerol (2-AG), stimulates polymerization of actin filaments (Malorni et al., 2004), thus promoting motility and chemotaxis of polymorphonuclear leukocytes, as well as leukocyte adhesion to fibronectin, overall facilitating infiltration of immune cells into inflamed tissues (Gokoh et al., 2005; Kishimoto et al., 2006). It should also be emphasized that available data are mainly based on studies carried out on single cells and/or at late stages of the adhesion cascade (Kishimoto et al., 2004; Kobayashi et al., 2001). However, it is well known that in real life different cell types need to cooperate with each other in a dynamic sequence of events to allow adhesion (Ley et al., 2007). With the aim of shedding light on cell/cell interactions, here we focused on the early events of endothelium/leukocyte cross-talks. We show for the first time that 2-AG upregulates the expression of selectins. Indeed, after a short exposure to 2-AG, human umbilical vein endothelial cells (HUVECs) are activated and signal to Jurkat T lymphocytes, thus allowing an efficient endothelium/leukocyte cross-talk.

were from Sigma Chemical Co. (St. Louis, MO). ACEA, HU210 and JWH015 were from Tocris Bioscience (Bristol, UK). WWL70 and SR144528 were purchased from Cayman Chemical Co (Ann Arbor, MI). All compounds were endotoxin-free. Anti-Akt, anti-phospho-AktSer473 , anti-PSGL-1, anti-tubulin, anti-ERK, anti-phospho-ERKTyr204 , anti-L-selectin antibodies, as well as secondary antibodies conjugated to horseradish peroxidase and the enhanced chemiluminescence (ECL) kit, were from Santa Cruz Biotechnology (Santa Cruz, CA). Allophycocyanin (APC)conjugated CD62P antibody was purchased from Miltenyi Biotec (Cologne, Germany), and phycoerythrin (PE)-conjugated CD62E antibody was from Biolegend (San Diego, CA). Anti-CB1 and antiCB2 primary antibodies were from Cayman Chemical Co. Alexa-488 secondary antibody was from Invitrogen (Carlsbad, CA). Calceinacetoxymethyl ester (calcein-AM) was from Life Technologies (Milan, Italy). 2.2. Cell cultures Human Jurkat T cells (ATCC, Manassas, VA) were grown in DMEM:F12 (1:1) medium, supplemented with 10% heatinactivated foetal bovine serum (Invitrogen). HUVECs (Lonza Group Ltd, Basel, Switzerland) were grown in EGM-2 Bullet kit medium (BioWhittaker, Radnor, PA), as reported (Maccarrone et al., 2002). In order to remove serum, which is known to contain trace amounts of eCBs (Marazzi et al., 2011), cells were accurately washed three times in phosphate buffered saline (PBS), before setting each experiment. Subconfluent cells were incubated in serum-free medium containing compounds to be tested, for the indicated periods of time. LPS, the endotoxin of Gram-negative bacteria, was used as positive control (Wright et al., 1990). In conditioned medium (CM) experiments, HUVECs were treated with different compounds for 1 hour; after the incubation period, cells were washed three times with PBS, fresh serum-free medium was added, and the incubation was carried out for two additional hours. Finally, the culture medium was collected, and used to resuspend Jurkat T cells, that were further incubated for the indicated periods of time. 2.3. FACS analysis Surface and intracellular P- and E-selectins, as well as CB1 and CB2 receptor expression, were quantified by flow cytometry in a FACSCanto instrument (Beckton Dikison, NJ). Briefly, 2-AGtreated HUVECs were collected and stained with APC-conjugated CD62P and PE-conjugated CD62E specific antibodies. For intracellular staining, cells were fixed with 4% paraformaldehyde for 15 min, and then stained intracellularly with the above-mentioned selectin-specific antibodies in 0.5% saponin, at room temperature. For CB receptor expression, HUVECs were stained with specific antiCB1 and anti-CB2 primary antibodies for 15 min at 4 ◦ C and then stained with Alexa-488 secondary antibody. Isotype-matched and secondary antibodies were employed to assess background staining and specificity. For each analysis 100,000 events were acquired and viable cells were analyzed using the Flowjo software (TreeStar, Ashland, OR). 2.4. Western blotting

2. Materials and methods 2.1. Reagents Chemicals were of the purest analytical grade. 2-AG, lipopolysaccharide (LPS), AM281, URB597, JZL184 and OMDM1

Cells were lysed in 50 mM Tris–HCl (pH 7.4), containing protease and phosphatase inhibitors. Proteins (30 ␮g/lane) from whole lysates or membrane fractions were subjected to SDS-PAGE, electroblotted onto PVDF membranes, incubated with specific antibodies and detected with ECL, as reported (Catani et al., 2010).

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

81

2.5. Quantitation of TNF-˛ and L-selectin shedding

2.7. Adhesion and trans-migration assays

HUVECs and Jurkat T cells were centrifuged at 200 × g for 10 min. Media were collected to measure TNF-␣ and soluble L-selectin (sL-selectin) content, through the Human TNF-␣ Screening Set kit (Thermo Fisher Scientific Inc. Rockford, IL), and the human L-selectin/CD62L ELISA kit (R&D Systems Inc, Minneapolis, MN), according to the manufacturer’s instructions. The absorbance values at 450 nm of unknown samples were always within the linearity range of calibration curves drawn with increasing concentrations of recombinant TNF-␣ (0–1000 pg/mL) or human L-selectin (0–5000 pg/mL), supplied by the kit manufacturers. For neutralization assays, 1 mL CM from 2-AG-stimulated HUVECs was incubated with 1 ␮g anti-TNF-␣ antibody (Abcam, Cambridge, UK), with 1 ␮g anti-TNFR2 antibody (R&D Systems), with 0.5 U human recombinant fatty acid amide hydrolase (FAAH) (Cayman Chemical Co) or with 0.05 U proteinase k (Sigma), for 30 min at 37 ◦ C. Then, Jurkat T cells were incubated in CM for 6 h, and sL-selectin release was quantified as described above.

HUVECs were treated with different compounds for 1 h at 37 ◦ C, then culture medium was replaced by compound-free medium and cells were further incubated for 5 h. Calcein-AM-labelled Jurkat cells were overlaid on HUVECs for 2 h at 37 ◦ C, the monolayer was carefully washed with phosphate-buffered saline to remove unbound lymphocytes, and fluorescence was measured through a Victor 3 1420-040 multilabel counter (Perkin Elmer Inc.Waltham, MA), with excitation at 485 nm and emission at 535 nm. HUVECs were also grown to confluence on 3 ␮m Transwell-filter inserts (24-multiwell plates). After incubation with LPS or 2-AG for 1 h at 37 ◦ C, the medium was discarded and 1.5 × 106 Jurkat T cells were added to endothelial monolayers. After 6 h, T cells migrated from the upper to the lower compartment were counted by using a phase-contrast microscope (100× magnification) on a total of eight fields.

2.6. Live imaging HUVECs were treated in 8-well chamber slides (Ibidi, Milan, Italy) and divided in 4 groups: three groups were incubated with vehicle, LPS or 2-AG for 1 h at 37 ◦ C, whereas the fourth group was pre-treated with AM281 (0.1 ␮M, 30 min) before adding 2AG. HUVEC nuclei were labelled by incubating them with 1 ng/mL Hoechst 33342 during the treatments. Jurkat T cells (1 × 105 cells/test) were labelled with 1 ␮M calcein-AM for 1 h at 37 ◦ C, washed twice, resuspended in medium and added to HUVEC monolayers. Interactions between Jurkat T lymphocytes and endothelial cells were visualized and recorded by using a time-lapse system mounted on an inverted fluorescence microscope (TiE; Nikon, Japan), equipped with an incubation chamber (Okolab), a cooled CCD camera (Clara; Andor), and a Niss Elements imaging software (Nikon). Video recordings were performed with a 40× objective, for at least 20 min of incubation, by using the MultiTrack Niss Elements module that allowed image acquisition of all experimental groups in the same recording session. Three channels were acquired: transmitted light through a differential interference contrast (DIC; Nomarski) filter, to evaluate cellular morphological features; blue channel (Hoechst 33342), for nuclei acquisition of endothelial cells; and green channel (calcein), for Jurkat cell body visualization. Then, video images were analyzed by using the Imaris Suite 7.6® software (Bitplane A.G., Zurich, Switzerland). Image analysis was performed under visual control, in order to determine the threshold that subtracts background noise and takes into account cellular structures and kinetic tracks. During processing, the images were compared with the original raw data, to make sure that no structures were introduced that were not seen in the original data series, nor that structures present in the original data series were removed. Video images were first segmented for the recognition of relevant objects and their separation from the background in every frame. Morphological and kinetic parameters analyzed were circularity (roundness of a shape; the closer to 1 is the circularity, the more round is the cell), track displacement length (distance between first and last cell positions), track length (total distance travelled by the cell), track straightness (migration directionality; the closer to 1 is the straightness, the more linear is the track), and mean speed (mean of all the instantaneous velocity for each recording). All indexes were evaluated by using the Spot module (Bitplane A.G.), except the circularity that was analyzed with ImageJ software (http://rsb.info.nih.gov/).

Fig. 1. Dose-dependent selectin expression in 2-AG-activated HUVECs. Surface (grey bars) and intracellular (white bars) P- (A) and E- (B) selectin expression. Cells were left untreated or were treated with increasing (0.1–10 ␮M) concentrations of 2-AG for 30 min, before staining with CD62P or CD62E antibodies; for intracellular selectin expression, cells were treated as above, except that they were fixed and permeabilized before staining. The intracellular selectin content was calculated by subtracting the mean fluorescent intensity of surface expression from the total fluorescence signal in permeabilized samples. Results are expressed as percentage of untreated cells, set to 100%. Values are the means ± S.D. of three independent experiments, each performed in triplicate. *p < 0.01 vs surface expression levels of untreated cells; **p < 0.05 vs surface expression levels of untreated cells; # p < 0.05 vs intracellular expression levels of untreated cells.

82

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

Fig. 2. Time-dependent selectin expression in 2-AG-activated HUVECs. Surface P- (A) and E- (D) selectin expression. Cells were left untreated (grey peak) or were treated with 1 ␮M 2-AG (white peak) for 1 h, before staining with CD62P or CD62E antibodies. For intracellular P- (B) and E- (E) selectin expression, cells were treated as above, except that they were fixed and permeabilized before staining. The intracellular selectin content was calculated by subtracting the mean fluorescent intensity of surface expression from the total fluorescence signal in permeabilized samples. Summary histograms of FACS analysis are shown in panels C and F: cells treated with 2-AG were checked for surface (grey bars) and intracellular (white bars) P- (C) and E- (F) selectin content. Results are expressed as percentage of untreated cells, set to 100%. Values are the means ± S.D. of three independent experiments, each performed in triplicate. *p < 0.01 vs surface expression levels of untreated cells; **p < 0.05 vs surface expression levels of untreated cells; # p < 0.05 vs intracellular expression levels of untreated cells.

2.8. Statistical analysis Statistical analysis was performed by the unpaired Student’s t test or one-way ANOVA (followed by Bonferroni post hoc analysis), using the InStat 3 program (GraphPAD Software for Science, San Diego, CA). Differences were considered significant at p < 0.05. 3. Results 3.1. 2-AG triggers plasma membrane expression of P- and E-selectins in HUVECs We firstly assessed the effect of increasing concentrations (0.1–10 ␮M) of 2-AG on surface exposure of selectins, that are early modulators of endothelium/leukocyte cross-talks. By FACS analysis, we found a dose-dependent increase in plasma membrane levels of P- and E-selectins, paralleled by a decrease in their intracellular content (Fig. 1). On the basis of these results, all subsequent experiments were performed at 1 ␮M 2-AG. As shown by time-dependent experiments, P-selectin was rapidly (within 1 h) mobilized to the cell surface (Fig. 2A and B) and, at longer incubation times, it was internalized again (Fig. 2C). Likewise, 2-AG triggered the expression of surface E-selectin. Already after 30 min – 1 hour, the protein was exposed on the surface of HUVECs (Fig. 2D and F), and in parallel its intracellular content decreased (Fig. 2E and F). Then, surface E-selectin levels decreased after 3 h (Fig. 2F). Incidentally, such findings are in agreements with literature data showing how expression kinetics of P- and Eselectins on the surface of vascular endothelium strongly depend on either the inflammatory stimulus or the endothelial cell type

(primary vs immortalized cells and/or human vs mouse), used as paradigm (Sanders et al., 1992; Somers et al., 2000; Khew-Goodall et al., 1996; Wang et al., 2010). 3.2. 2-AG elicits TNF-˛ secretion from HUVECs We next assessed TNF-␣ (a cytokine involved in immunemediated inflammatory responses) secretion in the medium of HUVECs incubated with 1 ␮M 2-AG for 1 h. Much alike LPS (used as a positive control), 2-AG significantly enhanced TNF-␣ release in the culture medium of HUVECs (Fig. 3A). The effect of 2-AG was long lasting, because HUVECs incubated with 2-AG for 1 h released TNF-␣ for up to 24 h (data not shown). To evaluate the possible site of action of 2-AG, we inhibited cellular uptake of this compound by using OMDM1, a selective inhibitor of eCB transmembrane transport (Ortar et al., 2003). Treatment with OMDM1 led to a significant increase in TNF-␣ secretion (Fig. 3A), thus indicating that the stimulatory effect of 2-AG depended on activation of an extracellular binding site. The same potentiating effect occurred upon pre-incubation with the selective inhibitor of monoacilglycerol lipase (MAGL), JZL184 (Long et al., 2009), of FAAH, URB597 (Fegley et al., 2005), and of ␣/␤-hydrolase domain 6 (ABHD6), WWL70 (Li et al., 2007), the main enzymes responsible for 2-AG hydrolysis in HUVECs (Fig. 3A and Supplemental Fig. 1). Although arachidonic acid per se was able to induce TNF-␣ release (Fig. 3A), nonetheless the potentiating effects observed with each inhibitor of 2-AG-degrading enzyme indicated that the effect of eCB was independent, at least in part, from products of its hydrolysis. We next investigated the involvement of CB receptors by pharmacological modulation. The CB1 receptor antagonist AM281

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

83

Overall, expression and pharmacological data suggest that the effect of 2-AG likely occurred via CB1 and CB2 receptors. 3.3. 2-AG-activated HUVECs promote Jurkat T cell activation

Fig. 3. (A) TNF-␣ secretion from 2-AG-activated HUVECs. Cells were left untreated (ctrl) or were treated with 100 ng/ml LPS (used as positive control) or with 1 ␮M 2AG, used alone or after pre-incubation with 0.1 ␮M AM281 (CB1 antagonist), 0.1 ␮M SR144528 (CB2 antagonist), 10 ␮M WWL70 (ABHD6 inhibitor), 0.1 ␮M URB597 (FAAH inhibitor), 1 ␮M JZL184 (MAGL inhibitor), or 0.1 ␮M OMDM1 (eCB transmembrane transport inhibitor), each used at. Cells were also incubated with ␮M ACEA and HU210 (CB1 agonists), JWH015 (CB2 agonist), all used at 0.1 ␮M, as well as with 0.1 and 1 ␮M arachidonic acid (AA). After incubation, TNF-␣ levels in culture medium were assessed by ELISA. Results are expressed as percentage of controls, set to 100% (absolute value for controls: 8.70 ± 0.05 pg/mL). Values are the means ± S.D. of three independent experiments, each performed in quintuplicate. *p < 0.0001 vs untreated cells; p < 0.0001 vs 2-AG-treated cells; **p < 0.05 vs 2-AG-treated cells. (B) Expression of CB1 and CB2 receptors in HUVECs. Cells (1 × 105 ) were stained at cell surface with anti-CB1 and anti-CB2 specific antibodies, as well as with isotype-matched antibody. Summary histograms of FACS analysis are shown. Results are expressed as mean fluorescent intensity (MFI) ± S.D. of three independent experiments.

(Pertwee et al., 2010) attenuated the effect of 2-AG (Fig. 3A), that instead was mimicked by HU210 and, to a lesser extent, by ACEA (Fig. 3A), two specific CB1 receptor agonists (Pertwee et al., 2010); similar results were obtained with JWH015 and SR144528, selective agonist and antagonist of CB2 receptor, respectively (Fig. 3A). The involvement of CB2 receptor was somehow unexpected, as our previous studies suggested that HUVECs were devoid of a functional CB2 subtype receptor (Maccarrone et al., 2000; Bari et al., 2006). Therefore, to gain further insight, we investigated CB expression by FACS analysis, which showed that HUVECs express both receptor subtypes (Fig. 3B). In this context, it should be noted that the apparent discrepancy may be explained taking into account that previous studies were based on pharmacological analyses (i.e., use of selective synthetic agonists and antagonists, as well as compounds able to interfere with CB1 signalling), and usage of AEA, that it is now thought to be a high affinity, CB1 -selective partial agonist; conversely, here we tested the effect of 2-AG that is recognized as a moderate affinity, CB1 /CB2 full agonist (Di Marzo and De Petrocellis, 2012).

We further extended our study by investigating whether 2-AG was able to modulate cell/cell cross-talk. To this end, HUVECs were exposed to 1 ␮M 2-AG for 1 h, the culture medium was replaced with compound-free fresh medium and cells were incubated for additional 2 h. The resulting CM is known to contain one or more soluble factors, released by endothelial cells and able to activate lymphocytes, thus representing an in vitro system that could mimic the in vivo situation (Kokura et al., 2000; Ley et al., 2007; von Dadelszen et al., 1999). By monitoring the expression of L-selectin, we found that activation of Jurkat T cells occurred after 3 hours of incubation with CM (Fig. 4A). Exposure to CM also promoted the release of sL-selectin from Jurkat T lymphocytes after 6 h (Fig. 4B); at 24 h protein levels remained high (Fig. 4A), but L-selectin was not shed any more (Fig. 4B). PSGL1 expression also increased in Jurkat T lymphocytes treated with CM, in a time-dependent manner: plasma membrane PSGL1 levels increased within 1 h of incubation, reached a maximum at 3 h, and then decreased at 24 h (Fig. 4C). Incidentally, the inducing effect of 2-AG on sL-selectin release appeared to be more evident by using human lymphocytes obtained from healthy blood donors. Indeed, peripheral lymphocytes receiving CM from 2-AG-treated HUVECs increased sL-selectin release, which reached a peak at 6 h (Supplemental Fig. 2). To identify the soluble factor(s) released from HUVECs that could be active on Jurkat T cells, we pre-incubated CM with proteinase K or with anti-TNF-␣ antibody; we also tested human recombinant FAAH, in order to explore the possibility that 2-AG itself, through a paracrine loop, could be released from activated endothelial cells, thus acting directly on lymphocytes, as already reported (Maccarrone et al., 2001, 2002). As shown in Fig. 4D, L-selectin release from Jurkat T cells was dependent on TNF-␣ secretion, since proteinase K and anti-TNF-␣ antibody were both able to revert the effect of 2-AG. Depending on the stimulus and the cell type, TNF-␣ might act via activation of TNFR1, which is engaged to triggers death signalling pathway, or/and TNFR2, that is important in T cell signalling and responses to infection (Faustman and Davis, 2013); therefore, we also tested the effect of an antibody which specifically recognized TNFR2 but not TNFR1. As reported in Fig. 4D, anti-TNFR2 antibody gave similar results as TNF-␣ depleting antibody, thus suggesting the involvement of this receptor in sL-selectin release. Interestingly, 2-AG did not appear to act per se on lymphocytes, because the presence of FAAH in the CM did not significantly affect sL-selectin release (Fig. 4D). Finally, we checked the possible engagement of specific kinases (like ERK and Akt), known to be involved in integrin activation, L-selectin ectodomain shedding and actin-myosin assembly dur´ 2010; ing cell migration (Cheresh et al., 1999; Killock and Ivetic, Wang et al., 2007). Although with a different time-dependence, both enzymes were activated in Jurkat T cells exposed to CM (Fig. 4E): ERK phosphorylation reached a peak after 5 min of incubation and slowly decreased within 60 min, while Akt phosphorylation increased linearly with time, reaching a maximum at 30–60 min. 3.4. 2-AG-dependent HUVEC/Jurkat T cell interactions enable completion of the leukocyte adhesion cascade We next wondered whether 2-AG-mediated effect on selectins was sufficient to prime and complete the leukocyte adhesion cascade. To this end, Jurkat T cells were seeded over a monolayer of HUVECs subjected to different treatments, and tracked by time-lapse microscopy, where we evaluated motility (Fig. 5, and

84

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

Fig. 4. Effects of CM on L-selectin and PSGL1 expression in Jurkat T cells. HUVECs were left untreated or were treated with 1 ␮M 2-AG for 1 h; then, medium was replaced with compound-free fresh medium and incubation was prolonged for additional 2 h. The resulting CM was used to resuspend Jurkat T cells, further incubated for 3, 6 and 24 h. Expression (A) and release (B) of L-selectin were assessed by Western blotting and ELISA, respectively. The radiograph is representative of three independent experiments; mean densitometric values are represented in boldface (S.D. ≤ 5%), and are reported as fold over control (set to 1), after normalization to tubulin content. In (B), data are reported as percentage of controls, set to 100% (absolute value for controls: 0.20 ± 0.03 pg/mL); values are the means ± S.D. of three independent experiments, each performed in quintuplicate. (C) Jurkat T cells, treated with CM for different periods of time, were checked for PSGL1 expression by Western blotting. The radiograph is representative of three independent experiments; mean densitometric values are represented in boldface (S.D. ≤ 5%), and are reported as fold over control (set to 1), after normalization to tubulin content. (D) CM, pre-incubated with anti-TNF-␣ antibody (␣-TNF␣), anti-TNFR2 antibody (␣-TNFR2), human recombinant FAAH or proteinase K (Pr.K), was used to resuspend Jurkat cells, and then sL-selectin levels were assessed by ELISA. (E) Phosphorylated ERK and Akt were detected in Jurkat T cells incubated with CM for 1 h (upper panels); the amount of total kinases is shown on the same filter (lower panels). The radiograph is representative of three independent experiments; mean densitometric values are represented in boldface (S.D. ≤ 5%), and are reported as fold over control (set to 1), after normalization to total kinase content. Data are reported as percentage of controls, set to 100%. Values are the means ± S.D. of three independent experiments, each performed in quintuplicate.*p < 0.05 vs untreated cells; **p < 0.01 vs untreated cells; # p < 0.05 vs untreated cells, but not vs 2-AG-treated cells; p < 0.01 vs 2-AG-treated cells.

supplementary videos 1–5) and morphological indexes (Fig. 6) able to give information on leukocyte rolling and crawling activities. Jurkat cells overlaid on 2-AG-treated HUVECs (2-AG group) showed increased trajectory length (∼130%; Fig. 5B) and track displacement length (∼155%; Fig. 5C), indicating that they covered a longer net distance than lymphocytes that migrated on vehicle-treated HUVECs (control group). Track straightness (Fig. 5D) and speed (Fig. 5E) also significantly increased (∼130% and 150%, respectively) in 2-AG group, indicating that Jurkat cells moved faster and in a more directional manner with respect to controls (see also Fig. 5A, and supplementary videos). Interestingly, AM281 and SR144528 led all motility parameters back to baseline values (Fig. 5), indicating that engagement of CB receptors on HUVECs was required to sustain the activity of 2-AG on Jurkat lymphocytes. The latter cells were also found to have distinct morphological profiles in response to 2-AG (Fig. 6). As assessed by the circularity parameter, cells migrating on 2-AG-activated HUVECs showed reduced cell roundness (∼80%), indicating that they acquired a more amoeboid phenotype with respect to controls. Again, 2-AG effects were reverted by pre-treating endothelial cells with the CB receptor antagonists AM281 and SR144528 (Fig. 6B). Taken together, live imaging measurements documented an overall increase in rolling-like activity of Jurkat cells upon exposure to 2-AG, via a CB-dependent mechanism. Incidentally, kinetic and morphological effects triggered by LPS, used as a positive control of migration

(Silber et al., 1994), appeared to be quite different from those exerted by 2-AG (Figs. 5 and 6). Endothelium/leukocyte interactions promoted by 2-AG allowed completeness of migration, indeed activated HUVECs stimulated firm adhesion of Jurkat T cells (Fig. 7A). The use of specific agonists (ACEA and JWH015) and antagonists (AM281 and SR144528) confirmed, once again, the involvement of both CB1 and CB2 receptors (Fig. 7A). In addition, activation of both Akt and MAPK pathways seemed to be necessary for an optimal effect of 2-AG on adhesion, as demonstrated by pre-treatment with pharmacological inhibitors of these kinases. When used alone, the ERK inhibitor PD98059 (Dudley et al., 1995) and the PI3K/Akt inhibitor wortmannin (Arcaro and Wymann, 1993) slightly inhibited adhesion of Jurkat T cells, whereas pre-treatment with either compound completely blocked the effect of 2-AG (Fig. 7A). Finally, stimulation of HUVECs with 2-AG also promoted transmigration of Jurkat T lymphocytes. Indeed, addition of 2-AG to the upper chamber of the plate (containing the HUVEC monolayer) increased the amount of Jurkat T cells migrating from the upper to the lower chamber, compared to unstimulated cells (Fig. 7B). 4. Discussion This investigation focused on early events in the interactions between leukocytes and endothelial cells. The most interesting

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

85

Fig. 5. Time-lapse microscopy analysis of in vitro endothelium/lymphocyte interactions. (A) Rendering elaboration of the track measurements of Jurkat T cells plated on HUVECs. Experimental groups were: untreated cells (Ctrl) or cells treated with LPS or 2-AG, the last being used alone or after pre-treatment with AM281 or SR144528 (SR2). White arrows: track displacement length. Coloured lines: track length. Histograms of the motion parameters analyzed were track length (B), track displacement length (C), track straightness (D) and speed (E). All measures are expressed as mean ± S.D. (n = 200). *p < 0.05 vs untreated and LPS-treated cells; **p < 0.01 vs untreated and LPS-treated cells; # p < 0.05 vs 2-AG-treated cells.

outcome of our study is that 2-AG is able to induce plasma membrane exposure of P- and E-selectins in primary endothelial cells, through a CB1 - and CB2 -dependent mechanism; then, activated cells release soluble factors (TNF-␣ included) that activate Jurkat T lymphocytes, thus allowing a coordinated action between the two cell types. Incidentally, the TNF-␣ concentrations that we measured here have been shown to be associated in vivo to inflammatory diseases, like ankylosing spondylitis, acute and chronic coronary artery disease, and cardiac autonomic neuropathy in type 2 diabetic patients (Bal et al., 2007; Hassanzadeh et al., 2006; Jung et al., 2012). Another novel finding that emerges from our study is that 2-AG may preferentially act on the first steps of migration, whereby Jurkat lymphocytes overlaid on 2-AG-treated HUVECs increase all the motility indexes, and migrate very fast according to a straightforward trajectory. Live imaging experiments also showed that, although both 2-AG and LPS are able to trigger leukocyte migration on endothelial cells, their effect on the adhesion cascade appears different, an observation that seems to deserve further analysis in an independent study. It should be noted that phenotype and responsiveness of endothelial cells are specifically related to the vascular bed (artery vs vein) and the age of subjects (newborn vs adult) from which they derived (Lawson and Weinstein, 2002); indeed, arterial [including human aortic (HAECs) and coronary artery (HCAECs) endothelial cells] and venous (such as HUVECs) endothelial cells show different gene expression programmes, so that opposite responses to

inflammatory and mechanical stimuli may occur (Deng et al., 2006; Bátkai et al., 2007; Buzby et al., 1994). Such a difference in experimental models employed might, therefore, explain the apparent discrepancy between our results and literature data reporting a suppressive effect of eCBs (2-AG or AEA) on pro-inflammatory molecules that promote adhesion of T cells to endothelium (Bátkai et al., 2007; Opitz et al., 2007; Rajesh et al., 2007). In addition, taking into account that eCB effects are strictly dose- and timedependent, the interplay between circulating factors (including eCBs) and properties of the local vascular wall may be critical for vascular disease susceptibility (Deng et al., 2006). One might speculate that eCBs are protective against vascular inflammation in arterial cells (more susceptible to atherosclerosis), but they exert an opposite effect in venous cells, thus regulating inflammationrelated processes, including angiogenesis. Noticeably, AEA has been reported to promote HUVEC proliferation in response to proangiogenic factors, while genetic and pharmacologic inactivation of CB1 receptor inhibits it (Pisanti et al., 2011; Guabiraba et al., 2013). In this context, it should be noted that TNF-␣ plays a crucial role in angiogenesis, by synergistically acting with angiogenic mediators and by priming endothelial cells for angiogenic sprouting (Sainson et al., 2008; Montrucchio et al., 1994). By regulating TNF-␣ release from endothelial cells, 2-AG can be viewed as a proangiogenic factor, promoting intercellular interactions between endothelial and immune cells. Interestingly, by mass spectrometry analysis, we found that 2-AG levels were halved after stimulation

86

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

Fig. 6. Morphological features of in vitro endothelium/lymphocyte interactions. (A) Representative fluorescence microscopy images of Jurkat T cells plated on HUVECs, left untreated (Ctrl) or treated with LPS or 2-AG, the last being used alone or after pre-treatment with AM281 or SR144528 (SR2). Blue: HUVECs nuclei stained with Hoechst 33342. Green: Jurkat T cells labelled with calcein-AM. (B) Histogram of circularity of Jurkat T cells treated as above. All measures are expressed as mean ± S.D. (n = 200). *p < 0.05 vs untreated and LPS-treated cells; **p < 0.05 vs 2-AG-treated cells.

with TNF-␣ (data not shown), thus suggesting that an autocrine loop may be established in order to tightly modulate the degree of inflammation. Besides the recognized role of selectins in inflammation and haemostasis, evidence has been accumulated for their contribution to other pathophysiological processes, including neuroinflammatory conditions (e.g., Multiple Sclerosis) and cancer metastasis (Bahbouhi et al., 2009; Coupland et al., 2012). Impairment of the blood-brain barrier facilitates leukocyte infiltration, and immune cell trafficking into the meningeal compartment is supported by selectins and PSGL1, although the role of these adhesion molecules in sustaining leukocyte extravasation remains controversial (Alvarez et al., 2011; Engelhardt, 2008). Selectin-dependent interactions may also contribute to the formation of a permissive microenvironment for metastasis: several cancer cells express on their surface carbohydrate determinants, which are specific selectin ligands, and the degree of expression correlates with metastasis and poor prognosis for cancer patients (Läubli and Borsig, 2010; Witz, 2008). Hence, 2-AG can also be viewed as a factor promoting interactions between endothelial cells and cancer cells. It should be recalled that tissue levels of eCBs are influenced by dietary constituents and that, in particular, 2-AG levels are affected by the ratio of plasma ␻-6/␻-3 PUFAs. Indeed, dietary ␻-3 PUFAs

dampen inflammatory responses by reducing biosynthesis, and thus plasma levels, of 2-AG (Banni et al., 2011; Batetta et al., 2009). On the other hand, blood and tissue 2-AG levels (normally in the nanomolar–low micromolar range) can increase locally (Kase et al., 2008; Obata et al., 2003), due to: i) activated platelets that release 2-AG (Maccarrone et al., 2001); ii) 2-AG secretion from endothelial cells and macrophages (Chicca et al., 2012; Gauthier et al., 2005); and iii) LPS-induced inhibition of eCB hydrolysis in human lymphocytes (Sugiura et al., 1998; Wagner et al., 1997). Taken together, it is tempting to speculate that a nutritional approach, able to influence eCB tissue levels by providing a suitable ratio of ␻-6/␻-3 PUFAs, may be useful for modulating energy homeostasis, inflammation and immune response. From a more general point of view, 2-AG (and possibly other eCBs) may be considered as local mediator, able to regulate cell/cell cross-talks, as already reported for platelet/endothelium (Maccarrone et al., 2002), fat/muscle (Eckardt et al., 2009), neuron/astrocyte (Navarrete and Araque, 2008), and sperm/oviduct communications (Gervasi et al., 2009), as well as for the attachment of a blastocyst to the uterine luminal epithelium during embryo implantation (Maccarrone, 2009). Therefore, by stimulating endothelial cells to release soluble factors able to recruit immune-competent cells, 2-AG may represent an interesting therapeutic target.

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

Fig. 7. Late events of the leukocyte adhesion cascade primed by 2-AG. (A) Adhesion of calcein-labelled Jurkat T cells on HUVEC monolayers left untreated (ctrl) or treated with 2-AG (alone or after pre-treatment with AM281 or SR144528), or with specific CB agonists (ACEA or JWH015). Adhesion assays were also performed by preincubating cells with 0.1 ␮M wortmannin (Akt inhibitor; Wt) or 20 ␮M PD98059 (ERK inhibitor; PD), alone or in combination. (B) Trans-migration assay of Jurkat T cells across HUVEC monolayers, left untreated (ctrl) or treated with LPS or 2AG. Results are expressed as percentage of controls, set to 100%. Values are the means ± S.D. of three independent experiments, each performed in quintuplicate. *p < 0.001 vs untreated cells; # p< 0.001 vs 2-AG-treated cells; ## p < 0.05 vs 2-AGtreated cells.

Acknowledgements This investigation was supported by Ministero dell’Istruzione, dell’Università e della Ricerca (PRIN 2009 to L.A., PRIN 2010-2011 to M.M.), and by Fondazione Italiana Sclerosi Multipla (FISM grant 2010 to M.M.) and Fondazione TERCAS (grant 2009-2012 to M.M.). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.biocel.2014.03.028. References Alvarez JI, Cayrol R, Prat A. Disruption of central nervous system barriers in multiple sclerosis. Biochim Biophys Acta 2011;1812:252–64. Arbones ML, Ord DC, Ley K, Ratech H, Maynard-Curry C, Otten G, Capon DJ, Tedder TF. Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin-deficient mice. Immunity 1994;1:247–60. Arcaro A, Wymann MP. Wortmannin is a potent phosphatidylinositol 3-kinase inhibitor: the role of phosphatidylinositol 3,4,5-trisphosphate in neutrophil responses. Biochem J 1993;296:297–301. Bahbouhi B, Berthelot L, Pettré S, Michel L, Wiertlewski S, Weksler B, Romero IA, Miller F, Couraud PO, Brouard S, Laplaud DA, Soulillou JP. Peripheral blood CD4+ T lymphocytes from multiple sclerosis patients are characterized by higher PSGL-1

87

expression and transmigration capacity across a human blood-brain barrierderived endothelial cell line. J Leukoc Biol 2009;86:1049–63. Bal A, Unlu E, Bahar G, Aydog E, Eksioglu E, Yorgancioglu R. Comparison of serum IL-1 beta, sIL-2R, IL-6, and TNF-alpha levels with disease activity parameters in ankylosing spondylitis. Clin Rheumatol 2007;26:211–5. Banni S, Carta G, Murru E, Cordeddu L, Giordano E, Sirigu AR, Berge K, Vik H, Maki KC, Di Marzo V, Griinari M. Krill oil significantly decreases 2-arachidonoylglycerol plasma levels in obese subjects. Nutr Metab 2011;8:7–12. Bari M, Spagnuolo P, Fezza F, Oddi S, Pasquariello N, Finazzi-Agro‘ A, Maccarrone M. Effect of lipid rafts on CB2 receptor signaling and 2-arachidonoyl-glycerol metabolism in human immune cells. J Immunol 2006;177:4971–80. Batetta B, Griinari M, Carta G, Murru E, Ligresti A, Cordeddu L, Giordano E, Sanna F, Bisogno T, Uda S, Collu M, Bruheim I, Di Marzo V, Banni S. Endocannabinoids may mediate the ability of (n-3) fatty acids to reduce ectopic fat and inflammatory mediators in obese zucker rats. J Nutr 2009;139:1495–501. Bátkai S, Rajesh M, Mukhopadhyay P, Haskó G, Liaudet L, Cravatt BF, Csiszár A, Ungvári Z, Pacher P. Decreased age-related cardiac dysfunction, myocardial nitrative stress, inflammatory gene expression, and apoptosis in mice lacking fatty acid amide hydrolase. Am J Physiol Heart Circ Physiol 2007;293: H909–18. Bevilacqua MP, Pober JS, Mendrick DL, Cotran RS, Gimbrone MA Jr. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc Natl Acad Sci USA 1987;84:9238–42. Buzby JS, Knoppel EM, Cairo MS. Coordinate regulation of Steel factor, its receptor (Kit), and cytoadhesion molecule (ICAM-1 and ELAM-1) mRNA expression in human vascular endothelial cells of differing origins. Exp Hematol 1994;22:122–9. Catani MV, Gasperi V, Catanzaro G, Baldassarri S, Bertoni A, Sinigaglia F, Avigliano L, Maccarrone M. Human platelets express authentic CB1 and CB2 receptors. Curr Neurovasc Res 2010;7:311–8. Cheresh DA, Leng J, Klemke RL. Regulation of cell contraction and membrane ruffling by distinct signals in migratory cells. J Cell Biol 1999;146:1107–16. Chicca A, Marazzi J, Nicolussi S, Gertsch J. Evidence for bidirectional endocannabinoid transport across cell membranes. J Biol Chem 2012;287:34660–82. Coupland LA, Chong BH, Parish CR. Platelets and P-selectin control tumour cell metastasis in an organ-specific manner and independently of NK cells. Cancer Res 2012;72:4662–71. Croxford JL, Yamamura T. Cannabinoids and the immune system: potential for the treatment of inflammatory diseases. J Neuroimmunol 2005;166:3–18. Deng DX, Tsalenko A, Vailaya A, Ben-Dor A, Kundu R, Estay I, Tabibiazar R, Kincaid R, Yakhini Z, Bruhn L, Quertermous T. Differences in vascular bed disease susceptibility reflect differences in gene expression response to atherogenic stimuli. Circ Res 2006;98:200–8. Di Marzo V, De Petrocellis L. Why do cannabinoid receptors have more than one endogenous ligand. Philos Trans R Soc Lond B Biol Sci 2012;367:3216–28. Dudley DT, Pang L, Decker SJ, Bridges AJ, Saltiel AR. A synthetic inhibitor of the mitogen-activated protein kinase cascade. Proc Natl Acad Sci U S A 1995;92:7686–9. Eckardt K, Sell H, Taube A, Koenen M, Platzbecker B, Cramer A, Horrighs A, Lehtonen M, Tennagels N, Eckel J. Cannabinoid type 1receptors in human skeletal muscle cells participate in the negative crosstalk between fat and muscle. Diabetologia 2009;52:664–74. Engelhardt B. Immune cell entry into the central nervous system: involvement of adhesion molecules and chemokines. J Neurol Sci 2008;274:23–6. Faustman DL, Davis M. TNF Receptor 2 and Disease: Autoimmunity and Regenerative Medicine. Front Immunol 2013;4:478, eCollection 2013. Fegley D, Gaetani S, Duranti A, Tontini A, Mor M, Tarzia G, Piomelli D. Characterization of the fatty acid amide hydrolase inhibitor cyclohexyl carbamic acid 3’-carbamoyl-biphenyl-3-yl ester (URB597): effects on anandamide and oleoylethanolamide deactivation. J Pharmacol Exp Ther 2005;313: 352–8. Gauthier KM, Baewer DV, Hittner S, Hillard CJ, Nithipatikom K, Reddy DS, Falck JR, Campbell WB. Endothelium-derived 2-arachidonylglycerol: an intermediate in vasodilatory eicosanoid release in bovine coronary arteries. Am J Physiol Heart Circ Physiol 2005;288:H1344–51. Gervasi MG, Rapanelli M, Ribeiro ML, Farina M, Billi S, Franchi AM, Perez Martinez S. The endocannabinoid system in bull sperm and bovine oviductal epithelium: role of anandamide in sperm–oviduct interaction. Reproduction 2009;137:403–14. Gokoh M, Kishimoto S, Oka S, Mori M, Waku K, Ishima Y, Sugiura T. 2arachidonoylglycerol, an endogenous cannabinoid receptor ligand, induces rapid actin polymerization in HL-60 cells differentiated into macrophage-like cells. Biochem J 2005;386:583–9. Graham ES, Angel CE, Schwarcz LE, Dunbar PR, Glass M. Detailed characterisation of CB2 receptor protein expression in peripheral blood immune cells from healthy human volunteers using flow cytometry. Int J Immunopathol Pharmacol 2010;23:25–34. Guabiraba R, Russo RC, Coelho AM, Ferreira MA, Lopes GA, Gomes AK, Andrade SP, Barcelos LS, Teixeira MM. Blockade of cannabinoid receptors reduces inflammation, leukocyte accumulation and neovascularization in a model of sponge-induced inflammatory angiogenesis. Inflamm Res 2013;62:811–21. Hafezi-Moghadam A, Thomas KL, Prorock AJ, Huo Y, Ley K. L-selectin shedding regulates leukocyte recruitment. J Exp Med 2001;193:863–72. Hassanzadeh M, Faridhosseini R, Mahini M, Faridhosseini F, Ranjbar A. Serum levels of TNF-, IL-6, and selenium in patients with acute and chronic coronary artery disease. Iran J Immunol 2006;3:142–5.

88

V. Gasperi et al. / The International Journal of Biochemistry & Cell Biology 51 (2014) 79–88

Jung CH, Kim BY, Kim CH, Kang SK, Jung SH, Mok JO. Association of serum adipocytokine levels with cardiac autonomic neuropathy in type 2 diabetic patients. Cardiovasc Diabetol 2012;11:24–32. Kase Y, Obata T, Okamoto Y, Iwai K, Saito K, Yokoyama K, Takinami M, Tanifuji Y. Removal of 2-arachidonylglycerol by direct hemoperfusion therapy with polymyxin B immobilized fibers benefits patients with septic shock. Ther Apher Dial 2008;12:374–80. Killock DJ, Ivetic´ A. The cytoplasmic domains of TNFalpha-converting enzyme (TACE/ADAM17) and L-selectin are regulated differently by p38 MAPK and PKC to promote ectodomain shedding. Biochem J 2010;428:293–304. Kishimoto S, Kobayaschi Y, Oka S, Gokoh M, Waku K, Sugiura T. 2arachidonylglycerol, an endogenous cannabinoid receptor ligand, induces accelerated production of chemokines in HL-60 cells. J Biochem 2004;135:517–24. Kishimoto S, Oka S, Gokoh M, Sugiura T. Chemotaxis of human peripheral blood eosinophils to 2-arachidonoylglycerol: comparison with other eosinophil chemoattractants. Int Arch Allergy Immunol 2006;140:3–7. Kobayashi Y, Arai S, Waku K, Sugiura T. Activation by 2-arachidonoylglycerol, an endogenous cannabinoid receptor ligand, of p42/44 mitogen-activated protein kinase in HL-60 cells. J Biochem 2001;129:665–9. Kokura S, Wolf RE, Yoshikawa T, Granger DN, Aw TY. T-lymphocyte-derived tumor necrosis factor exacerbates anoxia-reoxygenation-induced neutrophilendothelial cell adhesion. Circ Res 2000;86:205–13. Khew-Goodall Y, Butcher CM, Litwin MS, Newlands S, Korpelainen EI, Noack LM, Berndt MC, Lopez AF, Gamble JR, Vadas MA. Chronic expression of P-selectin on endothelial cells stimulated by the T-cell cytokine, interleukin-3. Blood 1996;87:1432–8. Langer HF, Chavakis T. Leukocyte-endothelial interactions in inflammation. J Cell Mol Med 2009;13:1211–20. Läubli H, Borsig L. Selectins promote tumor metastasis. Semin Cancer Biol 2010;20:169–77. Lawson ND, Weinstein BM. Arteries and veins: making a difference with zebrafish. Nat Rev Genet 2002;3:674–82. Ley K. The role of selectins in inflammation and disease. Trends Mol Med 2003;9:263–8. Ley K, Bullard DC, Arbones ML, Bosse R, Vestweber D, Tedder TF, Beaudet AL. Sequential contribution of L- and P-selectin to leukocyte rolling in vivo. J Exp Med 1995;181:669–75. Ley K, Laudanna C, Cybulsky MI, Nourshargh S. Getting to the site of inflammation: the leukocyte adhesion cascade updated. Nat Rev Immunol 2007;7:678–89. Li W, Blankman JL, Cravatt BF. A functional proteomic strategy to discover inhibitors for uncharacterized hydrolases. J Am Chem Soc 2007;129:9594–5. Long JZ, Li W, Booker L, Burston JJ, Kinsey SG, Schlosburg JE, Pavón FJ, Serrano AM, Selley DE, Parsons LH, Lichtman AH, Cravatt BF. Selective blockade of 2arachidonoylglycerol hydrolysis produces cannabinoid behavioral effects. Nat Chem Biol 2009;5:37–44. Maccarrone M. Endocannabinoids: friends and foes of reproduction. Prog Lipid Res 2009;48:344–54. Maccarrone M, Bari M, Lorenzon T, Bisogno T, Di Marzo V, Finazzi-Agrò A. Anandamide uptake by human endothelial cells and its regulation by nitric oxide. J Biol Chem 2000;275:13484–92. Maccarrone M, Bari M, Battista N, Finazzi-Agrò A. Estrogen stimulates arachidonoylethanolamide release from human endothelial cells and platelet activation. Blood 2002;100:4040–8. Maccarrone M, Bari M, Menichelli A, Giuliani E, Del Principe D, Finazzi-Agrò A. Human platelets bind and degrade 2-arachidonoylglycerol, which activates these cells through a cannabinoid receptor. Eur J Biochem 2001;268: 819–25. Malorni W, Bari M, Straface E, Battista N, Matarrese P, Finazzi-Agrò A, Del Principe D, Maccarrone M. Morphological evidence that 2-arachidonoyl-glycerol is a true agonist of human platelets. Thromb Haemost 2004;92:1159–61. Marazzi J1, Kleyer J, Paredes JM, Gertsch J. Endocannabinoid content in fetal bovine sera – unexpected effects on mononuclear cells and osteoclastogenesis. J Immunol Methods 2011;373:219–28. Montrucchio G, Lupia E, Battaglia E, Passerini G, Bussolino F, Emanuelli G, Camussi G. Tumor necrosis factor alpha-induced angiogenesis depends on in situ plateletactivating factor biosynthesis. J Exp Med 1994;180:377–82.

Moore KL. Structure and function of P-selectin glycoprotein ligand-1. Leuk Lymphoma 1998;29:1–15. Navarrete M, Araque A. Endocannabinoids mediate neuron-astrocyte communication. Neuron 2008;57:883893. Obata T, Sakurai Y, Kase Y, Tanifuji Y, Horiguchi T. Simultaneous determination of endocannabinoids (arachidonylethanolamide and 2-arachidonylglycerol) and isoprostane (8-epiprostaglandin F2alpha) by gas chromatography-mass spectrometry-selected ion monitoring for medical samples. J Chromatogr B Analyt Technol Biomed Life Sci 2003;792:131–40. Opitz CA, Rimmerman N, Zhang Y, Mead LE, Yoder MC, Ingram DA, Walker JM, Rehman J. Production of the endocannabinoids anandamide and 2-arachidonoylglycerol by endothelial progenitor cells. FEBS Lett 2007;581:4927–31. Ortar G, Ligresti A, De Petrocellis L, Morera E, Di Marzo V. Novel selective and metabolically stable inhibitors of anandamide cellular uptake. Biochem Pharmacol 2003;65:1473–81. Pertwee RG, Howlett AC, Abood ME, Alexander SP, Di Marzo V, Elphick MR, Greasley PJ, Hansen HS, Kunos G, Mackie K, Mechoulam R, Ross RA. International Union of Basic and Clinical Pharmacology. LXXIX. Cannabinoid receptors and their ligands: beyond CB1 and CB2 . Pharmacol Rev 2010;62:588–631. Pisanti S, Picardi P, Prota L, Proto MC, Laezza C, McGuire PG, Morbidelli L, Gazzerro P, Ziche M, Das A, Bifulco M. Genetic and pharmacologic inactivation of cannabinoid CB1 receptor inhibits angiogenesis. Blood 2011;117:5541–50. Polacheck WJ, Zervantonakis IK, Kamm RD. Tumor cell migration in complex microenvironments. Cell Mol Life Sci 2013;70:1335–56. Rajesh M, Mukhopadhyay P, Bátkai S, Haskó G, Liaudet L, Huffman JW, Csiszar A, Ungvari Z, Mackie K, Chatterjee S, Pacher P. CB2-receptor stimulation attenuates TNF-alpha-induced human endothelial cell activation, transendothelial migration of monocytes, and monocyte-endothelial adhesion. Am J Physiol Heart Circ Physiol 2007;293:2210–8. Sainson RC, Johnston DA, Chu HC, Holderfield MT, Nakatsu MN, Crampton SP, Davis J, Conn E, Hughes CC. TNF primes endothelial cells for angiogenic sprouting by inducing a tip cell phenotype. Blood 2008;111:4997–5007. Sanders WE, Wilson RW, Ballantyne CM, Beaudet AL. Molecular cloning and analysis of in vivo expression of murine P-selectin. Blood 1992;80:795–800. Silber A, Newman W, Reimann KA, Hendricks E, Walsh D, Ringler DJ. Kinetic expression of endothelial adhesion molecules and relationship to leukocyte recruitment in two cutaneous models of inflammation. Lab Invest 1994;70:163–75. Smalley DM, Ley K. L-selectin: mechanisms and physiological significance of ectodomain cleavage. J Cell Mol Med 2005;9:255–66. Somers WS, Tang J, Shaw GD, Camphausen RT. Insights into the molecular basis of leukocyte tethering and rolling revealed by structures of P- and E-selectin bound to SLe(X) and PSGL-1. Cell 2000;103:467–79. Sugiura T, Kodaka T, Nakane S, Kishimoto S, Kondo S, Waku K. Detection of an endogenous cannabimimetic molecule, 2-arachidonoylglycerol, and cannabinoid CB1 receptor mRNA in human vascular cells: is 2-arachidonoylglycerol a possible vasomodulator. Biochem Biophys Res Commun 1998;243:838–43. von Dadelszen P, Hurst G, Redman CW. Supernatants from co-cultured endothelial cells and syncytiotrophoblast microvillous membranes activate peripheral blood leukocytes in vitro. Hum Reprod 1999;14:919–24. Wagner JA, Varga K, Ellis EF, Rzigalinski BA, Martin BR, Kunos G. Activation of peripheral CB1 cannabinoid receptors in haemorrhagic shock. Nature 1997;390:518–21. Wang HB, Wang JT, Zhang L, Geng ZH, Xu WL, Xu T, Huo Y, Zhu X, Plow EF, Chen M, Geng JG. P-selectin primes leukocyte integrin activation during inflammation. Nat Immunol 2007;8:882–92. Wang GF, Wu SY, Xu W, Jin H, Zhu ZG, Li ZH, Tian YX, Zhang JJ, Rao JJ, Wu SG. Geniposide inhibits high glucose-induced cell adhesion through the NF-kappaB signaling pathway in human umbilical vein endothelial cells. Acta Pharmacol Sin 2010;31:953–6. Witz IP. The selectin–selectin ligand axis in tumor progression. Cancer Metastasis Rev 2008;27:19–30. Wright SD, Ramos RA, Tobias PS, Ulevitch RJ, Mathison JC. CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding protein. Science 1990;252:1321–2.