Endothelial Expression of Guidance Cues in Vessel Wall Homeostasis: Dysregulation under pro-atherosclerotic conditions (original) (raw)

Arterioscler Thromb Vasc Biol. Author manuscript; available in PMC 2014 May 1.

Published in final edited form as:

PMCID: PMC3647028

NIHMSID: NIHMS457812

Janine M. van Gils,1,* Bhama Ramkhelawon,1,* Luciana Fernandes,2 Merran C. Stewart,1 Liang Guo,1 Tara Seibert,3 Gustavo B. Menezes,2 Denise C. Cara,2 Camille Chow,4 T. Bernard Kinane,4 Edward A. Fisher,1 Mercedes Balcells,5,6,4 Jacqueline Alvarez-Leite,2 and Kathryn J. Moore1

Janine M. van Gils

1Marc and Ruti Bell Vascular Biology and Disease Program, Leon H. Charney Division of Cardiology, Dept of Medicine, New York University School of Medicine, New York, NY, USA

Bhama Ramkhelawon

1Marc and Ruti Bell Vascular Biology and Disease Program, Leon H. Charney Division of Cardiology, Dept of Medicine, New York University School of Medicine, New York, NY, USA

Luciana Fernandes

2Department of Biochemistry and Immunology, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil

Merran C. Stewart

1Marc and Ruti Bell Vascular Biology and Disease Program, Leon H. Charney Division of Cardiology, Dept of Medicine, New York University School of Medicine, New York, NY, USA

Liang Guo

1Marc and Ruti Bell Vascular Biology and Disease Program, Leon H. Charney Division of Cardiology, Dept of Medicine, New York University School of Medicine, New York, NY, USA

Tara Seibert

3University of Ottawa Heart Institute, Ottawa, Canada

Gustavo B. Menezes

2Department of Biochemistry and Immunology, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil

Denise C. Cara

2Department of Biochemistry and Immunology, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil

Camille Chow

4Program of Developmental Immunology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

T. Bernard Kinane

4Program of Developmental Immunology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

Edward A. Fisher

1Marc and Ruti Bell Vascular Biology and Disease Program, Leon H. Charney Division of Cardiology, Dept of Medicine, New York University School of Medicine, New York, NY, USA

Mercedes Balcells

4Program of Developmental Immunology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

5Massachusetts Institute of Technology, Division of Health Sciences and Technology, Cambridge, MA, USA

6Institut Quimic de Sarria, IQS School of Engineering, Ramon Llull University, Barcelona, Spain

Jacqueline Alvarez-Leite

2Department of Biochemistry and Immunology, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil

Kathryn J. Moore

1Marc and Ruti Bell Vascular Biology and Disease Program, Leon H. Charney Division of Cardiology, Dept of Medicine, New York University School of Medicine, New York, NY, USA

1Marc and Ruti Bell Vascular Biology and Disease Program, Leon H. Charney Division of Cardiology, Dept of Medicine, New York University School of Medicine, New York, NY, USA

2Department of Biochemistry and Immunology, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil

3University of Ottawa Heart Institute, Ottawa, Canada

4Program of Developmental Immunology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

5Massachusetts Institute of Technology, Division of Health Sciences and Technology, Cambridge, MA, USA

6Institut Quimic de Sarria, IQS School of Engineering, Ramon Llull University, Barcelona, Spain

Corresponding author: Kathryn J Moore, New York University School of Medicine, 522 First Avenue, Smilow 705, New York, NY 10016, Tel: 212-263-9259 gro.cmuyn@eroom.nyrhtaK

*equal contribution

Abstract

Objective

Emerging evidence suggests that neuronal guidance cues, typically expressed during development, are involved in both physiological and pathological immune responses. We hypothesized that endothelial expression of such guidance cues may regulate leukocyte trafficking into the vascular wall during atherogenesis.

Approach/Results

We demonstrate that members of the Netrin, Semaphorin and Ephrin family of guidance molecules are differentially regulated under conditions that promote or protect from atherosclerosis. Netrin-1 and Semaphorin3A are expressed by coronary artery endothelial cells and potently inhibit chemokine-directed migration of human monocytes. Endothelial expression of these negative guidance cues is down-regulated by pro-atherogenic factors, including oscillatory shear stress and pro-inflammatory cytokines associated with monocyte entry into the vessel wall. Furthermore, we show using intravital microscopy that inhibition of Netrin-1 or Semaphorin3A using blocking peptides increases leukocyte adhesion to the endothelium. Unlike Netrin-1 and Semaphorin3A, the guidance cue EphrinB2 is up-regulated under pro-atherosclerotic flow conditions and functions as a chemoattractant, increasing leukocyte migration in the absence of additional chemokines.

Conclusions

The concurrent regulation of negative and positive guidance cues may facilitate leukocyte infiltration of the endothelium through a balance between chemoattraction and chemorepulsion. These data indicate a previously unappreciated role for axonal guidance cues in maintaining the endothelial barrier and regulating leukocyte trafficking during atherogenesis.

Keywords: axonal guidance, migration, atherosclerosis, endothelial-leukocyte interaction

Introduction

Atherosclerosis is a chronic inflammatory disease of the large arteries in which monocyte recruitment into the vessel wall plays an essential role in both the initiation and progression of disease. Atherosclerotic plaques form predominantly at sites of disturbed laminar flow, notably, arterial branch points and bifurcations, and are initiated by the subendothelial accumulation of apolipoprotein B-containing lipoproteins1, 2. The key early inflammatory response to these atherogenic lipoproteins is activation of overlying endothelial cells in a manner that leads to recruitment of blood-borne monocytes3. Research in the last decades has provided a rich description of the molecules involved in the recruitment of leukocytes into the artery wall, and highlighted the importance of the coordinated action of chemokines, integrins and other adhesion molecules in directing this pathologic process4. The chemoattractant and adhesive forces that recruit leukocytes to the vessel wall have been well studied in this context5. However, our understanding of other aspects of immune cell migration remains incomplete, particularly the mechanisms by which these cells are excluded from the artery in the absence of inflammation. In addition to the well known positive migration cues that promote atherosclerotic plaque formation, it is reasonable to assume that chemorepulsive forces, or negative guidance cues, also exist in order to inhibit leukocyte-endothelial interactions under homeostatic conditions that may become dysregulated during disease.

The integration of chemoattractive and chemorepulsive signals is essential for controlling neuronal migration during development, and this paradigm is evolutionarily conserved from Caenorhabditis elegans to mammals6. Netrins, Slits, Semaphorins and Ephrins compose the four major families of conserved neuronal guidance cues that control neuronal migration and vascular patterning through a complex interplay of signals. However, accumulating evidence points to additional functions for these guidance molecules in regulating cell migration outside of the nervous system, including roles in the physiological and pathological regulation of the immune response7–9. Such studies have shown that members of the Netrin, Slit, Semaphorin and Ephrin families of guidance cues can regulate immune cell activation or differentiation, and have both chemoattractive and chemorepulsive effects on leukocyte migration9–12. For example, recent work from our group identified Netrin-1 as a leukocyte guidance cue expressed by the endothelium, where its expression was modulated during acute inflammation due to Staphylococcus aureus infection11. In this model, Netrin-1 was found to be expressed on the luminal surface of lung endothelial cells, where it acted to block the migration of monocytes to such bacterial factors as the N-formylated peptide fMLP (N-formyl-methionine-leucine-phenylalanine), suggesting that it may play role in endothelial barrier function. At the onset of S. aureus infection or upon treatment with TNFα in vitro, endothelial expression of Netrin-1 was rapidly downregulated consistent with a lowering of this barrier to leukocyte infiltration of the tissues11.

Given the emerging roles for neuronal guidance cues in regulating inflammation, we hypothesized that these molecules may also have roles in early atherosclerosis, when, as alluded to above, endothelial dysfunction due to injury or lipoprotein retention precedes the development of atherosclerotic plaques. To perform a systematic evaluation of the 4 major classes of evolutionarily conserved neuronal guidance molecules, we used a microarray approach using RNA samples harvested from atherosclerosis prone and resistant aortic sites in a mouse model of early atherogenesis. We identified key members of the Netrin, Semaphorin and Ephrin families that are regulated in the initial stages of plaque formation and confirmed these changes in arterial endothelial cells exposed to atherogenic factors in vitro. Notably, we identified guidance cues that both block (Netrin-1 and Semaphorin3A) or promote (EphrinB2) monocyte migration, and showed that their expression by endothelium is concurrently regulated by pro-atherosclerotic conditions in a manner that would facilitate leukocyte recruitment. Furthermore, using intravital microscopy, we demonstrated that blocking peptides targeting these molecules alters leukocyte adhesion to the endothelium in vivo as would have been predicted by the array and in vitro results. Together, these data suggest an emerging paradigm for the coordination of leukocyte-endothelial interactions in vessel wall homeostasis by neuroimmune guidance cues.

RESULTS

Neuronal Guidance Molecules are Differentially Regulated by Flow

It is well established that atherosclerotic lesions develop in areas of branching or high curvature that are associated with oscillatory or turbulent flow, so-called ‘athero-prone regions’13. We therefore investigated the expression of neuronal guidance molecules in two different regions of the vessel wall (i) the inner curvature (athero-prone region) and (ii) the outer curvature (athero-protected or homeostatic region) of the aortic arch (Figure 1A). We placed _Ldlr_−/− mice on a Western diet for two weeks in order to model conditions for the early initiation of atherosclerosis, and isolated the inner and outer curvature from their aortic arches. Using custom mRNA arrays that covered all four families of neuronal guidance molecules, we compared the expression profile between these two vascular sites. In the aortic arch, we found that 31 of the 36 neuronal guidance molecules we profiled were expressed (Figure 1B–C). Comparing the expression of neuronal guidance molecules from the inner curvature relative to the outer curvature, we found that 10 were more than 25% downregulated in the inner compared to the outer curvature, and only 2 were more than 25% upregulated (Figure 1C). From these, we selected candidate molecules from the netrin, semaphorin, slit and ephrin families and confirmed the array findings by independent qRT-PCR (Figure 1D). Expression of netrin-1 (Ntn1), which we previously showed can limit the migration of monocytes, granulocytes and lymphocytes11, was reduced by 48% in the inner curvature (athero-prone) compared to the outer curvature. Like Ntn1, another reported repulsive cue of the semaphorin family, semaphorin3A (Sema3a), was downregulated in the inner-curvature compared to the outer-curvature (−53%) consistent with roles in endothelial barrier function for these two guidance molecules. From the slit family, Slit1 was not detectable, while Slit2 was expressed at very low levels. However, Slit3 was confirmed to be expressed at similar levels between the inner and outer curvature. From the ephrin family, the ephrinB2 gene (Efnb2) was selected for validation, as ephrinB2 has been demonstrated to be expressed in vascular endothelial cells and may be sufficient to initiate monocyte-endothelial interactions12, 14. However, we did not detect differential expression of Efnb2 mRNA between the inner curvature and outer curvature of the aorta (Figure 1D).

An external file that holds a picture, illustration, etc. Object name is nihms457812f1.jpg

Differential Expression of Neuronal Guidance Molecules in the inner vs. outer curvature of the aortic arch

(A) _Ldlr_−/− mice were fed a Western diet for 2 weeks after which their aortas were harvested and the athero-protected outer- or athero-prone inner-curvature isolated. CD31 staining confirmed disruption of endothelial integrity suggestive of early atherosclerosis in the inner compared to outer-curvature. (B–C) mRNA expression profiling of these aortic regions for netrin, slit, ephrin and semaphorin family members using custom qRT-PCR arrays (n=3 mice); (B) Heat-map; (C) Relative fold change in mRNA expression in the inner compared to outer curvature. (D) Validation of candidate guidance cue expression in the inner and outer curvature by qRT-PCR. Data are mean ± s.e.m. *P<0.05.

To directly investigate whether it was the endothelial cells of the aorta that contributed to the observed differences in Ntn1, Sema3a and Efnb2 gene expression, and to extend those differences to the protein level we performed immunostaining of longitudinal sections of the aortic arch of the _Ldlr_−/− mice fed a Western diet for two weeks. At this time point, changes in endothelial cell morphology are observed in the disturbed flow regions of the lesser curvature of the aorta, prior to the subendothelial accumulation of monocytes15. Netrin-1 immunostaining was detected in endothelial cells (positive for CD31) of the outer curvature, but was highly downregulated on endothelial cells of the inner curvature of the aortic arch (Figure 2). Similarly, semaphorin3A was expressed by endothelial cells in the athero-protected outer curvature, while there was little to no semaphorin3A expression by endothelial cells of the inner aortic curvature. In contrast, endothelial cells of the inner aortic curvature expressed ephrinB2; however staining for ephrinB2 was not detected in the athero-protected outer curvature.

An external file that holds a picture, illustration, etc. Object name is nihms457812f2.jpg

Immunofluorescent staining of Netrin-1, Semaphorin3A and EphrinB2 in athero-protected and athero-prone regions of the aorta

Longitudinal sections of the aortic arch of the _Ldlr_−/− mice fed a Western diet for two weeks were stained for CD31 (green), netrin-1, semaphorin3A, or ephrinB2 (red) and DAPI (blue) in the athero-prone inner curvature (left) or athero-protected outer curvature (right) of the aortic arch. Areas of co-localization (yellow) are shown in the merged image. Images are representative of 5 mice. (L= lumen)

Endothelial Expression of Netrin-1, Semaphorin3A and EphrinB2 are Regulated by Pro-atherosclerotic Conditions in Vitro

Given the difference in netrin-1, semaphorin3A and ephrinB2 mRNA and protein levels in athero-protected vs athero-prone regions of the aorta, we postulated that hemodynamic pro-atherosclerotic forces regulate the expression of these guidance cues on endothelial cells. We therefore seeded human coronary artery endothelial cells (HCAECs) onto fibronectin-coated gas-permeable laboratory tubes and exposed them to arterial (athero-protective), oscillatory (athero-prone) flow, or no flow (static) for 6 hours. As a positive control, we measured endothelial nitric oxide (Nos3), an atheroprotective factor known to be expressed under arterial laminar flow conditions, and found that its mRNA abundance was upregulated 3-fold compared to static control, but was not significantly changed by oscillatory flow conditions compared to static control (Figure 3A). Similarly, Ntn1 mRNA was almost 5-fold upregulated by arterial flow, but unchanged in oscillatory flow conditions when compared to static control (Figure 3A). A different pattern emerged for Sema3a; expression of this gene was unchanged between static and arterial flow conditions, however, it was highly downregulated by oscillatory flow (~50%, Figure 3B). Notably, Efnb2 gene expression was also regulated by oscillatory flow, but unlike Sema3a it was significantly upregulated under these pro-atherosclerotic flow conditions (1.7 fold, Figure 3B). Slit3 mRNA and protein were not affected by either athero-protective or athero-prone flow conditions, and, therefore, was not investigated further (Supplemental Figure 1A).

An external file that holds a picture, illustration, etc. Object name is nihms457812f3.jpg

Regulation of Netrin-1, Semaphorin3A and EphrinB2 expression by arterial flow conditions

(A) Ntn1, Nos3 (eNOS) and (B) Sema3a, Efnb2 and Slit3 mRNA in HCAECs subjected to arterial (17±1 dynes/cm2 continuous flow in one direction) or oscillatory (17±1 dynes/cm2 biphasic flow in opposing directions) for 6 hours. Data are expressed as the fold change in mRNA in each condition compared with static flow. Data are the mean ± s.d. of quadruplicate samples in a single experiment and are representative of an experimental n of 4. *P<0.05

The observed changes in the expression of Ntn1, Sema3a and Efnb2 under differential flow conditions led us to consider whether other pro-atherogenic factors could modulate the expression of these molecules. To examine this we stimulated HCAECs with chemokines that have been implicated in the recruitment of monocytes during early atherosclerosis, including CCL2 (MCP-1), CX3CL1 (Fractalkine; FKN), or IL-85. Analysis of HCAEC mRNA expression revealed that MCP-1, FKN, and IL-8 all significantly reduced Ntn1 after 6 hours of treatment (Figure 4A). Similarly, MCP-1 and FKN, but not IL-8 reduced Semaphorin3A expression under these conditions (Figure 4A). In contrast, EphrinB2, which we found to be induced by oscillatory flow conditions, was not affected by atherosclerotic cytokines/chemokines (Figure 4A). Similar effects on Ntn1 and Sema3a mRNA were seen with other classical pro-atherogenic activators of the endothelium, such as IL-1β and TNFα, as well as with bacterial lipopolysaccharide (LPS) (Figure 4B). Notably, these pro-inflammatory stimuli increased Efnb2 mRNA in HCAEC (Figure 4B). In order to ascertain whether these changes were reflected at the protein level, we treated HCAECs with MCP-1 or TNFα for 24 hours and performed Western blotting for netrin-1, semaphorin3A and ephrinB2. Expression of netrin-1 and semaphorin3A by HCAECs was decreased after treatment with MCP-1 and TNFα, whereas ephrinB2 was increased under similar conditions (Figure 4C).

An external file that holds a picture, illustration, etc. Object name is nihms457812f4.jpg

Pro-atherosclerotic chemokines regulate Netrin-1, Semaphorin3A and EphrinB2 expression in HCAECs

(A–B) Ntn1, Sema3a, and Efnb2 mRNA in HCAECs treated with pro-atherosclerotic chemokines/cytokines (A) [MCP-1 (5 nM), Fractalkine (5 nM), and IL-8 (5 nM)] or (B) LPS (1μg/ml), IL-1b (20ng/ml) or TNFa (10ng/ml). Data are the mean ± s.d. of triplicate samples in a single experiment and are representative of an experimental n of 3. *P<0.05. (C) Western blot analysis of netrin-1, sema3A, ephrinB2 or GAPDH (internal control) in HCAECs stimulated with MCP-1 or TNFα for 24 h.

Netrin-1, Semaphorin3A and EphrinB2 Regulate Leukocyte Migration

Because we observed lower expression of netrin-1 and semaphorin3A in regions of the aorta subject to pro-atherosclerotic/oscillatory flow and in endothelial cells exposed to oscillatory flow conditions or pro-atherosclerotic cytokines, we hypothesized that endothelial expression of netrin-1 and semaphorin3A may act as a barrier to prevent monocyte migration into the arterial intima under basal conditions. To test this hypothesis, we investigated the effects of netrin-1 and semaphorin3A on monocyte migration to two chemokines implicated in monocyte recruitment during early atherogenesis; Fractalkine (CX3CL1) and MCP-1 (CCL2). Cultured THP-1 monocytes or freshly isolated human peripheral blood mononuclear cells (PBMCs) were used in modified Boyden chamber assays to examine monocyte migration to Fractalkine or MCP-1 by adding recombinant netrin-1 to the lower chamber in the absence or presence of chemoattractants. As we previously reported11, recombinant netrin-1 did not alter monocyte migration in the absence of chemokine (data not shown). However, addition of recombinant netrin-1 inhibited PBMC migration to Fractalkine and MCP-1 by up to 50% (Figure 5A, B). This inhibitory effect of netrin-1 was reversed by pre-incubation of PBMCs with an antibody that binds to the extracellular domain of Unc5b, a receptor for netrin-1 expressed by monocytes11 (Figure 5C). Similarly, when semaphorin3A was added in combination with Fractalkine or MCP-1, it was found to be a potent inhibitor of PBMC chemotaxis to both Fractalkine (Figure 5D) and MCP-1 (Figure 5E). In dose-response experiments, semaphorin3A-dependent effects on migration produces a bell-shaped curve (also typically seen with chemokines), with 60–90% inhibition observed at 250 ng/ml and reduced effects observed at the highest doses (1000 ng/ml) (Figure 5D–E).

An external file that holds a picture, illustration, etc. Object name is nihms457812f5.jpg

Netrin-1, Semaphorin3A and EphrinB2 alter leukocyte migration

Migration of PBMCs to (A) Fractalkine (FKN 1 nM) or (B) MCP-1 (10 nM) with or without recombinant Netrin-1 at the concentrations indicated. (C) Pre-incubation of PBMCs with an Unc5b-blocking antibody (5 μg/ml) blocks the inhibitory effect of Netrin-1 (250 ng/ml) on migration to MCP-1 (10 nM). (DE) Migration of THP-1 monocytes to Fractalkine (1 nM) (D) or MCP-1 (10 nM) (E) with or without recombinant semaphorin3A at the concentrations indicated. (F) Pre-incubation of THP-1 monocytes with a blocking Neuropilin-1 antibody (5 μg/ml), but not a control antibody (5 μg/ml), reverses the inhibitory effect of Semaphorin3A on migration to 10nM MCP-1. (G) Migration of PBMCs to ephrinB2 added at the indicated concentrations in the absence of a chemotactic stimulus. Data are the mean ± s.d. of triplicate samples in a single experiment and are representative of an experimental n=3. *P<0.05.

It has been previously demonstrated that class III secreted semaphorins can signal through Neuropilins and Plexins to mediate their effects on target cells16, 17. We therefore investigated whether Neuropilin-1 was required for semaphorin3A-mediated inhibition of monocyte chemotaxis. THP-1 monocytes were pre-incubated with a blocking antibody to Neuropilin-1 or an IgG isotype control antibody. Pre-incubation with the control antibody did not restore monocyte chemotaxis to MCP-1 (Figure 5F). However, pre-incubation with the Neuropilin-1 blocking antibody restored the ability of THP-1 monocytes to respond to MCP-1 (Figure 5F).

In contrast to netrin-1 and semaphorin3A, ephrinB2 expression was observed to be upregulated in endothelial cells exposed to oscillatory flow conditions. Thus, we hypothesized that the upregulation of ephrinB2 may act to recruit monocytes into the arterial intima. To test this hypothesis, we examined the ability of ephrinB2 to act as a monocyte chemoattractant. Indeed we found that ephrinB2 was a potent monocyte chemoattractant when added to the lower chamber of Boyden migration assays (Figure 5G). Migration was induced in a dose-dependent manner, with maximal chemotaxis observed at 250 ng/mL of ephrinB2. When ephrinB2 was added with FKN or MCP-1 to the lower well, chemotaxis was similar to that seen with ephrinB2 alone (data not shown), indicating that the effects on chemotaxis were not additive.

Netrin-1 and Semaphorin3A Inhibit Leukocyte Adhesion

We next investigated whether netrin-1 and semaphorin3A could alter leukocyte adhesion. Treatment with netrin-1 or semaphorin3A, both at 250 ng/ml, reduced the adhesion of RAW264.7 cells to surfaces coated with BSA by 50–60% (Fig. 6A). As netrin-1 and semaphorin3A inhibit both leukocyte migration and adhesion in vitro, we next determined whether enhancing or blocking these two negative guidance molecules would alter leukocyte-endothelial dynamics. We find that netrin-1 and semaphorin3A actively inhibit THP-1 monocyte binding to HCAEC treated with TNFα (Fig 6B, C) or LPS (Supplemental Fig. 1B). Conversely, blocking peptides to netrin-1 and Semaphorin3A increased THP-1 adhesion to quiescent HCAEC (Fig. 6D). Hence, the expression of netrin and semaphorin3A by the endothelium would be expected to actively prevent leukocyte arrest and tissue entry.

An external file that holds a picture, illustration, etc. Object name is nihms457812f6.jpg

Netrin-1 and Semaphorin3A inhibits leukocyte adhesion

(A) Adhesion of RAW264.7 cells, treated without or with netrin-1 (250 ng/ml) and sema3A (250 ng/ml), to BSA-coated tissue culture plates. (B) Adhesion of THP-1 monocytes to untreated or TNFα (10ng/ml) treated HCAECs in the presence/absence of recombinant netrin-1 or sema3A (250 ng/ml). Data are the mean ± sem of 4 experiments. (C) Representative images of labeled THP-1 monocyte adhesion to HCAECs under the conditions shown in B. (D) Adhesion of THP-1 monocytes to HCAECs treated with either netrin-1- or semaphorin3A-blocking peptide or control peptides. Data are the mean ± sem of 6 experiments. *P<0.05.

To examine this we used intravital microscopy, a technique that allows the observation of leukocyte-endothelial cell interactions in vivo18. Similar to areas of disturbed blood flow in the conduit artery, the venous system is exposed to low shear stress ranging from 1 to 6 dynes/cm2. Thus, intravital microscopy of the cremaster vein has proven to be a useful in vivo model for assessing leukocyte-endothelial cell interactions relevant to atherosclerosis19–21. Intravenous infusion of netrin-1- or semaphorin3A-blocking peptides to C57BL/6J mice did not alter leukocyte rolling, but significantly increased leukocyte adhesion to the endothelium compared to control peptides (Figure 7A, B). The number of leukocytes adherent to the endothelium increased approximately 2-fold following the addition of netrin-1- or semaphorin3A-blocking peptides (Figure 7B), further supporting our hypothesis that these guidance cues act to impair this earliest interaction with endothelial cells in vivo.

An external file that holds a picture, illustration, etc. Object name is nihms457812f7.jpg

Blocking Netrin-1 or Semaphorin3A increases leukocyte adhesion to endothelium in vivo.

Leukocyte rolling and adhesion before and after administration of a (A–B) netrin-1- or (C–D) semaphorin3A-blocking peptide or control peptide measured by intravital microscopy in C57BL/6 mice. Data are the mean ± s.e.m. n=5–7 mice/group, *P<0.05. (E) Schematic diagram of the regulated expression of neuroimmune guidance cues under athero-protective and athero-prone conditions. Under conditions of laminar flow (left) netrin-1 and semaphorin3A are expressed on endothelial cells and inhibit leukocyte recruitment, maintaining an athero-protective state. In contrast, oscillatory flow characteristic of atheroprone regions of the vasculature downregulates expression of netrin-1 and semaphorin3A, while increasing levels of ephrinB2, thereby facilitating leukocyte recruitment.

DISCUSSION

The pioneering studies from the Gimbrone laboratory showed that the properties of athero-prone and susceptible regions of the aorta varied in a number of significant ways that could be attributed to the different flow characteristics at specific sites13. For example, endothelial cells in the lesser (inner) curvature of the aortic arch or at branch points, where the laminar flow tended to be lower, were relatively activated. These observations were subsequently extended by a number of investigators to show that before there was evidence of monocyte infiltration, a host of molecules, such as P-selectin, VCAM-1, ICAM-1, and eNOS, were already adversely affected in the susceptible areas3, 13. Based on the results of the present studies, we can now add certain neuronal guidance molecules as early disease stage, athero-regulatory, factors that also operate at the level of the endothelium.

Based on the microarray gene expression results (Figure 1), we focused on netrin-1 and semaphorin3A, because of their decreased expression in the atheroprone, lesser curvature in the aortae of pre-lesional, but diet-stimulated, _Ldlr_−/− mice. We also included analyses of eprinB2 and slit3 as members of their respective classes of neuronal guidance molecules. In our previously published studies we reported that netrin-1 inhibits monocyte migration in vitro in response to bacterial fMLP11. In that same study we found that netrin-1 was expressed in the vascular endothelium of the lung, as well as in human umbilical vein endothelial cells, and that its expression was rapidly downregulated during acute inflammation. Based on these findings, we hypothesized that the higher expression of netrin-1 in the endothelium of the athero-resistant greater curvature reflected an important and novel mechanism by which leukocyte interactions with endothelial cells are limited in homeostatic conditions. Furthermore, we considered that there could also be regulation of leukocyte entry into susceptible arterial regions by other neuronal guidance molecules, given some overlap in properties among the family members.

Importantly, we found that the expression of netrin-1 and semaphorin3A was increased under conditions that recapitulated the steady flow found in healthy arterial segments relative to the oscillatory pattern found at arterial regions prone to development of atherosclerotic plaques, as would be expected if these molecules were athero-protective at the endothelial cell level. Endothelial expression of netrin-1 and semaphorin3A was also decreased by treatment with proinflammatory cytokines or chemokines implicated in leukocyte recruitment to athero-prone regions. We had previously demonstrated that most leukocyte subclasses, including monocytes/macrophages and neutrophils, express the chemorepulsive receptor for netrin1, Unc5b11, 22. This suggests that when circulating monocytes encounter an endothelial cell expressing relatively more netrin-1, a barrier mechanism would be triggered either because secreted netrin-1 inhibited monocyte chemotaxis by creating a diffusible netrin-1 gradient across endothelial cell layers (similar to that created by endothelial cell-secreted MCP-1), or through presentation of netrin-1 on the surface of endothelial cells. In this regard, there is evidence of netrin-1 binding to α6β4 and α3β1 integrins on pancreatic epithelial cells23, and this is one mechanism by which netrin-1 could be presented on arterial endothelial cells.

Consistent with either scenario, our results show that 1) netrin-1 inhibits leukocyte migration to the pro-atherogenic chemokines MCP-1 and Fractalkine in vitro, in a manner that involves the Unc5b receptor; 2) netrin-1 inhibits monocyte binding to endothelial cells; and 3) peptides that block netrin-1 increase leukocyte adhesion to endothelium in vitro. In addition to supporting an atheroprotective role for netrin-1 at the endothelial level, the current study also suggested a similar role for semaphorin3A. For example, semaphorin3A inhibited monocyte migration to MCP-1 in vitro in a process dependent on its receptor Neuropilin, and a semaphorin3A blocking peptide also reduced leukocyte adhesion in vitro. Notably, using intravital microscopy of the cremaster vein, we show that administration of netrin-1 and semaphorin3A blocking peptides similarly increase leukocyte adhesion in vivo.

Turning to ephrinB2, in contrast to netrin-1 and semaphorin 3A, it was upregulated in HCAECs in vitro by pro-atherogenic (oscillatory) flow (Figure 3) and at the protein level it was more highly expressed in the inner curvature endothelium (Figure 2). Another contrast was that it acted to stimulate leukocyte recruitment. These findings are supported by Goettsch et al. (2004), who found that ephrinB2 expression was down-regulated by arterial, but not venous, laminar shear stress in HUVECs24, suggesting that this downregulation may keep endothelial cells in a nonactivated state. Other supporting data for a role of ephrinB2 in vivo are contained in a report that showed that monocytes express EphB2, one of the possible receptors for ephrinB2, and that the expression of EphB2 in monocytes is increased upon their adhesion14. In agreement with the present results, in that report, there was also greater expression of ephrinB2 in the endothelium of the lesser curvature, which correlated with an increased abundance of macrophages in the underlying neointima14.

While we have shown that there is concurrent regulation of netrin-1, semaphorin3A, and ephrinB2 in the endothelium, resulting in concerted athero-protective effects, whether this represents a common coordinating regulatory mechanisms or a more complex process, remains to be determined. In contrast to Ntn1 and Sema3a, we did not detect a difference in EfnB2 mRNA expression in the athero-prone and resistant sites of the aorta, which implies that a combination of transcriptional and post-transcriptional mechanisms may regulate the expression of these factors. Similarly, whereas TNFα treatment of HCAECs markedly downregulated Ntn1 mRNA, as well as netrin-1 and semaphorin3A protein levels, this cytokine increased Sema3A mRNA in vitro. There are scant data about the regulation of endothelial expression for any of the factors we studied. There is some evidence for Netrin-1 that NF-κB and HIF-1α may be involved in its transcriptional upregulation25, 26, but on the surface, this would be counter to the present results in that 1) treatment with LPS or TNFα (activators of NF-κB) or MCP-1 (a classic target of NF-κB) was associated with decr eased netrin-1 expression (Figure 4), and 2) in areas of disrupted laminar flow (as in the lesser curvature of the aortic arch), HIF-1α is induced independent of hypoxia. Thus, while the phenomenon of intersecting protective changes in the 3 guidance molecules is clear, a dissection of the underlying regulatory mechanisms awaits detailed analyses beyond the scope of the present report.

Our data suggest that netrin-1 and semaphorin3A may reduce leukocyte migration and recruitment to atherosclerotic plaque by inhibiting firm adhesion to the vessel wall. In vivo, leukocyte recruitment to inflamed sites is mediated by a combination of adhesion and signaling molecules. Selectins, expressed at inflamed sites, slow leukocytes and initiate rolling; next, chemokine signaling to leukocytes activates integrins to mediate firm adhesion and subsequent extravasation. Our observations that netrin-1 and semaphorin3A inhibit the initiation of firm adhesion are consistent with known roles for neuronal guidance molecules. For example, Semaphorin3A inhibits the activation and adhesive function of β1 and β3 integrins in endothelial cells27 and similar effects might be expected to regulate activity of β2 integrins on leukocytes, which are primarily responsible for firm adhesion of leukocytes to the vessel wall. Netrin-1 has also been shown to bind to integrins, (α3 and α6) and to activate integrin ligand binding through focal adhesion kinase (FAK) and small GTPases.23

In summary, our data support a model (Figure 7) in which the inhibitory guidance molecules netrin-1 and semaphorin3A are more highly expressed by endothelial cells in athero-resistant aortic regions, where they help to maintain tissue homeostasis by preventing leukocyte influx. In athero-prone regions, however, they become down-regulated by disturbed laminar flow and local inflammation, which lowers the endothelial barrier to leukocyte entry. Consistent with this model, systemic delivery of netrin-1 to _Ldlr_−/− mice using adenovirus-associated virus reduced atherosclerosis28, presumably by increasing its expression on endothelial cells. Adding to the increased predilection for leukocyte recruitment in the prone regions is the upregulation of another guidance molecule, eprinB2, which increases monocyte migration in vitro. In addition to these guidance cues which we selected for in depth study, our microarray profiling data suggest that additional family members, including Ntng2, Sema3g, Sema4d, and Sema7a may also be relevant to atherosclerotic processes. During development, the various families of guidance molecules have been shown to have overlapping effects and to also co-regulate each other positively or negatively. Further exploration of the expression of neuronal guidance molecules and their receptors under resting and inflammatory conditions, and their functional interactions, will likely identify new regulatory mechanisms and therapeutic targets in atherosclerosis and other inflammatory disorders.

SIGNIFICANCE

We identified key members of the Netrin, Semaphorin and Ephrin families that are regulated on the endothelium in the initial stages of atherosclerotic plaque formation. We identified guidance cues that both block (Netrin-1 and Semaphorin3A) or promote (EphrinB2) monocyte migration, and showed that their expression by endothelium is concurrently regulated by pro-atherosclerotic conditions in a manner that would facilitate leukocyte recruitment. Furthermore, blocking peptides targeting these molecules alter leukocyte adhesion to the endothelium in vivo as would have been predicted by the array and in vitro results. Together, these data suggest an emerging paradigm for the coordination of leukocyte-endothelial interactions in vessel wall homeostasis by neuroimmune guidance cues.

Supplementary Material

1

2

Acknowledgments

b) Sources of Funding: Support for this work came from the American Heart Association (Grant-in-aid 0655840T to K.J.M; 09POST2080250 to J.M.vG), Fondation de la recherche Medicale to BR, National Institutes of Health (RC1HL100815 to K.J.M; R01 HL084312 to EAF; R01/GM049039 MB), Canadian Institute of Health Research (CGS-MSFSS to T.S.), Conselho Nacional de Desenvolvimento Científico e Tecnológico (01558/2007-6 to L.R.F and J.I.A-L), Ministerio de Innovación Plan Nacional BFU2009-09804 (MB), Fundació Empreses IQS (MB), and POSIMAT (MB).

Footnotes

References

1. Tabas I, Williams KJ, Boren J. Subendothelial lipoprotein retention as the initiating process in atherosclerosis: Update and therapeutic implications. Circulation. 2007;116:1832–1844. [PubMed] [Google Scholar]

2. Williams KJ, Tabas I. The response-to-retention hypothesis of early atherogenesis. Arterioscler Thromb Vasc Biol. 1995;15:551–561. [PMC free article] [PubMed] [Google Scholar]

3. Mestas J, Ley K. Monocyte-endothelial cell interactions in the development of atherosclerosis. Trends Cardiovasc Med. 2008;18:228–232. [PMC free article] [PubMed] [Google Scholar]

5. Galkina E, Ley K. Leukocyte influx in atherosclerosis. Curr Drug Targets. 2007;8:1239–1248. [PubMed] [Google Scholar]

6. Kolodkin AL, Tessier-Lavigne M. Mechanisms and molecules of neuronal wiring: A primer. Cold Spring Harb Perspect Biol. 2011:3. [PMC free article] [PubMed] [Google Scholar]

7. Suzuki K, Kumanogoh A, Kikutani H. Semaphorins and their receptors in immune cell interactions. Nat Immunol. 2008;9:17–23. [PubMed] [Google Scholar]

8. Cirulli V, Yebra M. Netrins: Beyond the brain. Nat Rev Mol Cell Biol. 2007;8:296–306. [PubMed] [Google Scholar]

9. Guan H, Zu G, Xie Y, Tang H, Johnson M, Xu X, Kevil C, Xiong WC, Elmets C, Rao Y, Wu JY, Xu H. Neuronal repellent slit2 inhibits dendritic cell migration and the development of immune responses. J Immunol. 2003;171:6519–6526. [PubMed] [Google Scholar]

10. Mirakaj V, Brown S, Laucher S, Steinl C, Klein G, Kohler D, Skutella T, Meisel C, Brommer B, Rosenberger P, Schwab JM. Repulsive guidance molecule-a (rgm-a) inhibits leukocyte migration and mitigates inflammation. Proc Natl Acad Sci U S A. 2011;108:6555–6560. [PMC free article] [PubMed] [Google Scholar]

11. Ly NP, Komatsuzaki K, Fraser IP, Tseng AA, Prodhan P, Moore KJ, Kinane TB. Netrin-1 inhibits leukocyte migration in vitro and in vivo. Proc Natl Acad Sci U S A. 2005;102:14729–14734. [PMC free article] [PubMed] [Google Scholar]

12. Korff T, Braun J, Pfaff D, Augustin HG, Hecker M. Role of ephrinb2 expression in endothelial cells during arteriogenesis: Impact on smooth muscle cell migration and monocyte recruitment. Blood. 2008;112:73–81. [PubMed] [Google Scholar]

13. Gimbrone MA, Jr, Topper JN, Nagel T, Anderson KR, Garcia-Cardena G. Endothelial dysfunction, hemodynamic forces, and atherogenesis. Ann N Y Acad Sci. 2000;902:230–239. discussion 239–240. [PubMed] [Google Scholar]

14. Braun J, Hoffmann SC, Feldner A, Ludwig T, Henning R, Hecker M, Korff T. Endothelial cell ephrinb2-dependent activation of monocytes in arteriosclerosis. Arterioscler Thromb Vasc Biol. 2011;31:297–305. [PubMed] [Google Scholar]

15. Mullick AE, Soldau K, Kiosses WB, Bell TA, 3rd, Tobias PS, Curtiss LK. Increased endothelial expression of toll-like receptor 2 at sites of disturbed blood flow exacerbates early atherogenic events. J Exp Med. 2008;205:373–383. [PMC free article] [PubMed] [Google Scholar]

16. Horie T, Ono K, Horiguchi M, Nishi H, Nakamura T, Nagao K, Kinoshita M, Kuwabara Y, Marusawa H, Iwanaga Y, Hasegawa K, Yokode M, Kimura T, Kita T. Microrna-33 encoded by an intron of sterol regulatory element-binding protein 2 (srebp2) regulates hdl in vivo. Proc Natl Acad Sci U S A. 2010;107:17321–17326. [PMC free article] [PubMed] [Google Scholar]

17. Lwigale PY, Bronner-Fraser M. Semaphorin3a/neuropilin-1 signaling acts as a molecular switch regulating neural crest migration during cornea development. Developmental biology. 2009;336:257–265. [PMC free article] [PubMed] [Google Scholar]

18. Gavins FN, Chatterjee BE. Intravital microscopy for the study of mouse microcirculation in anti-inflammatory drug research: Focus on the mesentery and cremaster preparations. Journal of pharmacological and toxicological methods. 2004;49:1–14. [PubMed] [Google Scholar]

19. Jin SY, Tohyama J, Bauer RC, Cao NN, Rader DJ, Zheng XL. Genetic ablation of adamts13 gene dramatically accelerates the formation of early atherosclerosis in a murine model. Arterioscler Thromb Vasc Biol. 2012;32:1817–1823. [PMC free article] [PubMed] [Google Scholar]

20. Eriksson EE. Intravital microscopy on atherosclerosis in apolipoprotein e-deficient mice establishes microvessels as major entry pathways for leukocytes to advanced lesions. Circulation. 2011;124:2129–2138. [PubMed] [Google Scholar]

21. Missiou A, Kostlin N, Varo N, Rudolf P, Aichele P, Ernst S, Munkel C, Walter C, Stachon P, Sommer B, Pfeifer D, Zirlik K, MacFarlane L, Wolf D, Tsitsikov E, Bode C, Libby P, Zirlik A. Tumor necrosis factor receptor-associated factor 1 (traf1) deficiency attenuates atherosclerosis in mice by impairing monocyte recruitment to the vessel wall. Circulation. 2010;121:2033–2044. [PMC free article] [PubMed] [Google Scholar]

22. van Gils JM, Derby MC, Fernandes LR, Ramkhelawon B, Ray TD, Rayner KJ, Parathath S, Distel E, Feig JL, Alvarez-Leite JI, Rayner AJ, McDonald TO, O’Brien KD, Stuart LM, Fisher EA, Lacy-Hulbert A, Moore KJ. The neuroimmune guidance cue netrin-1 promotes atherosclerosis by inhibiting the emigration of macrophages from plaques. Nature immunology. 2012;13:136–143. [PMC free article] [PubMed] [Google Scholar]

23. Nikolopoulos SN, Giancotti FG. Netrin-integrin signaling in epithelial morphogenesis, axon guidance and vascular patterning. Cell Cycle. 2005;4:e131–135. [PubMed] [Google Scholar]

24. Goettsch W, Augustin HG, Morawietz H. Down-regulation of endothelial ephrinb2 expression by laminar shear stress. Endothelium : journal of endothelial cell research. 2004;11:259–265. [PubMed] [Google Scholar]

25. Paradisi A, Maisse C, Bernet A, Coissieux MM, Maccarrone M, Scoazec JY, Mehlen P. Nf-kappab regulates netrin-1 expression and affects the conditional tumor suppressive activity of the netrin-1 receptors. Gastroenterology. 2008;135:1248–1257. [PubMed] [Google Scholar]

26. Rosenberger P, Schwab JM, Mirakaj V, Masekowsky E, Mager A, Morote-Garcia JC, Unertl K, Eltzschig HK. Hypoxia-inducible factor-dependent induction of netrin-1 dampens inflammation caused by hypoxia. Nat Immunol. 2009;10:195–202. [PubMed] [Google Scholar]

27. Serini G, Valdembri D, Zanivan S, Morterra G, Burkhardt C, Caccavari F, Zammataro L, Primo L, Tamagnone L, Logan M, Tessier-Lavigne M, Taniguchi M, Puschel AW, Bussolino F. Class 3 semaphorins control vascular morphogenesis by inhibiting integrin function. Nature. 2003;424:391–397. [PubMed] [Google Scholar]

28. Khan JA, Cao M, Kang BY, Liu Y, Mehta JL, Hermonat PL. Systemic human netrin-1 gene delivery by adeno-associated virus type 8 alters leukocyte accumulation and atherogenesis in vivo. Gene Ther. 2011;18:437–444. [PubMed] [Google Scholar]