Monocyte/Macrophages are implicated in initiation of angiogenesis, tissue/organ perfusion and atherosclerosis biology. We recently showed that chemokine receptor CX3CR1 is an essential regulator of monocyte/macrophage derived smooth muscle cell differentiation in the vessel wall after injury. Here we hypothesised the contribution of CX3CR1- CX3CL1 interaction to in vivo neovascularization and studied the functional consequences of genetic and pharmacologic targeting of CX3CR1 in formation, maturation and maintenance of microvascular integrity. Cells functionally deficient in CX3CR1 lacked matrix tunnelling and tubulation capacity in a 3D Matrigel assay. These morphogenic and cytokinetic responses were driven by CX3CL1-CX3CR1 interaction and totally abrogated by a Rho antagonist. To evaluate the role of CX3CR1 system in vivo, Matrigel plugs were implanted in competent CX3CR1+/gfp and functionally deficient CX3CR1gfp/gfp mice. Leaky microvessels (MV) were formed in the Matrigel implanted in CX3CR1gfp/gfp but not in CX3CR1+/gfp mice. In experimental plaque neovascularization immature MV phenotype was observed in CX3CR1gfp/gfp mice, lacking CX3CR1 positive smooth muscle-like cells, extracellular collagen and basement membrane (BM) laminin compared to competent CX3CR1+/gfp mice. This was associated with increased extravasation of platelets into the intima of CX3CR1gfp/gfp but not functionally competent CX3CR1 mice. Pharmacologic targeting using CX3CR1 receptor antagonist in wild type mice resulted in formation of plaque MV with poor BM coverage and a leaky phenotype. Our data indicate a hitherto unrecognised role for functional CX3CR1 in Matrigel and experimental plaque neovascularization in vivo, which may buttress MV collectively in favour of a more stable non-leaky phenotype.
Citation: Kumar AHS, Martin K, Turner EC, Buneker CK, Dorgham K, Deterre P, et al. (2013) Role of CX3CR1 Receptor in Monocyte/Macrophage Driven Neovascularization. PLoS ONE 8(2): e57230. https://doi.org/10.1371/journal.pone.0057230
Editor: Leonard Eisenberg, New York Medical College, United States of America
Received: August 19, 2012; Accepted: January 18, 2013; Published: February 21, 2013
Copyright: © 2013 Kumar et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by grants from Science Foundation Ireland (PI and RFP-NMC) and from ANR (France) (n°ANR-09-PIRI-0003). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The monocyte–macrophage system (MPS) exhibits numerous functions with respect to vascular remodeling which include classical phagocytosis, regulation of antigen presentation and atherosclerotic plaque formation , , , . MPS has also been implicated in paracrine-mediated pro-angiogenic and vascular structure changes orchestrated in response to cues as diverse as atherosclerotic plaque progression, tissue injury and ischemia, inflammation and tumour infiltration , , , , , , . For instance, macrophages have been reported to tunnel through stromal architecture of numerous tissues including the heart and the retina connecting putative microvascular networks , . However, the mechanism and molecular regulators underlying this process are largely unknown and it remains unclear whether this is a generalized feature of MPS infiltration of tissues or if it represents a more specialized function of cellular components of MPS that might be a target for pharmacologic modification.
We have previously identified a CX3CR1 monocyte subpopulation capable of phagocytic functions typical of classical MPS as well as mural vascular smooth muscle cell-like functions that occur subsequent to CX3CL1-CX3CR1 interaction in the injured vessel wall , . Interference with CX3CL1-CX3CR1 interaction was also observed to decrease neointima formation and atherosclerosis in murine animal models , , , thus implicating the CX3CL1-CX3CR1 axis in atherosclerotic plaque development. Moreover polymorphisms in CX3CR1 receptors have been associated with variability in prevalence of human atherosclerosis and coronary artery disease . Given that monocyte/macrophages are proposed to participate in angiogenesis , , , ,  we hypothesized that CX3CL1-CX3CR1 signaling may also be implicated in mural cell fate determination during plaque microvessel formation both in terms of cell recruitment to the perivascular space but also in cytoskeletal re-organization of recruited mural cells that integrate into nascent and maturing microvessels. In the current study we selected the chemokine receptor CX3CR1 as a transgenic marker allowing baseline and loss of function status to be examined in mice that had green fluorescence protein knocked in at one or both CX3CR1 alleles , . This facilitated simultaneous in vivo tracking of a major monocyte/macrophage subset into evolving neovascularization networks in solid matrix and in medium sized vessel wall microvasculature. These transgenic models also facilitated interrogation of loss of CX3CR1 function effects on CX3CR1 cell recruitment to the perivascular space, mural cell integration into evolving microvasculature, microvessel maturation and extracellular matrix production and microvessel haemorrhage and permeability. In vitro studies evaluated the signaling pathway downstream of CX3CR1 activation and its effects on cytoskeletal re-organization and solid matrix tunneling of CX3CR1 positive cells.
We show here for the first time that interaction between CX3CL1 and CX3CR1 initiates signaling through a Rho dependent pathway that promotes cell tubulation and tunneling through extracellular matrix. Moreover, functional competence of CX3CR1 to interact with its cognate ligand CX3CL1 is essential for microvessel maturation and expansion, and monocyte/macrophage-derived mural cell integration into evolving vasculature during neovascularization in vivo. Finally, CX3CR1 deficiency or therapeutic targeting with a peptide-antagonist reduces experimental plaque and Matrigel neovascularization but favours a leaky, haemorrhagic microvessel phenotype.
Materials and Methods
All in vivo procedures were approved by University College Cork Animal Experimentation Ethics Committee. Mice (male) employed in experiments were 8–12 weeks of age. Initial C57BL/6J breeding colonies were from Charles River UK. Transgenic CX3CR1gfp (C57BL6/J background) mice ,, in which either one (CX3CR1gfp/+) or both (CX3CR1gfp/gfp) copies of the CX3CR1 gene were interrupted by enhanced gene fluorescent protein (eGFP) and were originally from the European Mutant Mice Archive (EMMA).
CX3CR1 peptide antagonist (F1)
The CX3CR1 peptide antagonist (F1) was engineered from phage library of CX3CL1 mutants consisting of modified N terminus (QHHGVT sequence of the native CX3CL1 replaced with ILDNGVS). F1 peptide was purified from E. Coli inclusion bodies using standard procedures and its specificity to CX3CR1 receptor and pharmacodynamics has previously been described . Based on the static adhesion assay and prior pharmacodynamics data , 50 µg of F1 peptide was dissolved in 100 µl of saline and administered intraperitoneally every third day from the 2nd to 4th week post carotid artery ligation.
Static adhesion of mice bone marrow cells to CX3CL1 coated plate
48-well plates were coated for 1 hour with a 1∶200 dilution of a mouse antipolyHis tag antibody (HIS-1, Sigma Aldrich, Arklow, Ireland) in bicarbonate buffer, pH 9.6 with 0.1% BSA. Subsequently, selected wells were coated with 100 nM rmCX3CL1 (R&D System, Abingdon, UK) for 1 hr. Since recombinant CX3CL1 is expressed with a C-terminal 6-histidine tag to facilitate purification using nickel agarose, an anti-His tag antibody was coated onto the plate to assist CX3CL1 capture and orientation . CX3CR1 cells isolated from bone marrow of CX3CR1gfp/+ mice were pre-treated with or without CX3CR1 peptide antagonist (F1; 2.5 and 5 µM) and were allowed to adhere to wells and then washed with PBS (3×). Attached cells were then labelled with Hoechst dye and counted using UV microscopy. The number of cells per field (10× objective) was expressed as mean ± SEM.
Mouse model of experimental plaque neovascularization
A murine plaque angiogenesis model previously employed by Sasaki et al  was used. Mice were anesthetised by intraperitoneal administration of ketamine (90 mg/kg) and xylazine (10 mg/kg) and the left carotid artery at its bifurcation was exposed by standard microsurgical procedure. An 8/0 suture was passed under the artery at its bifurcation and tied to create a complete occlusion of the artery following which the subcutaneous tissue and skin was closed with 6/0 sutures by standard surgical process. The sham mice underwent the surgical steps described above except for the ligation procedure. In treatment groups, two weeks post carotid ligation, the mice were treated with F1 (50 µg total dose) or saline i.p. every third day for 2 weeks. 24 hrs before tissue harvest, the mice were injected with fluorescent microspheres (2–2.5 µm diameter) via tail vein following which the mice were sacrificed by over dose of Sodium Pentobarbital (i.p.).The carotid artery was perfusion fixed using chilled PBS and 4% paraformaldehyde. Perfusion fixed carotid artery was dissected out and cleaned of surrounding fat and cellular debris before being embedded in OCT for cryosectioning.
In vivo Matrigel neovascularization assay
500 µl of Matrigel (Gibco®; Bio-Sciences, Dublin, Ireland) fortified with 1 ng/ml FGF was subcutaneously implanted into the right and left flank region of mice. At 4 weeks post Matrigel implantation the mice were injected i.v. with 100 µl of 0.05% Evans blue or saline via tail vein and the mice were sacrificed 4 hrs post injection by overdose of sodium pentobarbital (i.p.). In a separate set of studies 200 µg of Rhodamine Griffonia (Bandeiraea) Simplicifolia Lectin I(GSL I, BSL I) (Vector Labs, CA, USA) was administered to mice by tail vein 24 hrs before culling. The Evans blue content of Matrigel was determined by dissolving the Matrigel in formic acid and then by reading the absorbance at 620 nm. While the heme content was quantified by pyridine hemochrome spectrophotometric heme assay . Separate Matrigel samples embedded in OCT were processed for immunofluorescence and immunocytochemistry.
Morphometric and immunofluorescence analysis of injured carotid artery and Matrigel
5 µM thick frozen sections from carotid artery were cut and stained with hematoxylin-Z and eosin and Trichrome stains (CellPath Ltd, Powys, UK). Tissue images were acquired using Nikon CCD camera attached to a Nikon microscope. Morphometric analysis was performed using NIH ImageJ software for data quantification.
For immunofluorescence detection of smooth muscle specific protein and GFP expression carotid artery/Matrigel sections were stained for calponin (rabbit anti calponin: 1∶200; Epitomics, CA, USA), CX3CR1 (rabbit anti CX3CR1:1∶50; ProSci Incorporated, CA, USA), F4/80 (Rat anti mouse F4/80: 1∶100; eBioscience; Hatfield, UK), Laminin (mouse anti laminin: 1∶500; Novus Biologicals, Cambridge, UK) and/or CD42b (rat anti CD42b: 1∶200; emfret Analytics GmbH & co.KG, Eibelstadt, Germany). Goat anti-rabbit, anti-mouse or anti rat Alexa Fluor 546 or 488 (1∶250 or 500; Abcam, Cambridge, UK) was used as secondary antibody and DAPI was used for nuclear staining. Rabbit, mouse or rat IgG (1∶50 or 200; Abcam, Cambridge, UK) were used as isotype controls. No additional labelling was required for GFP. Several thin optical slices (thickness<0.7 µm) of high-resolution sequential confocal scans (Nikon eC1 plus, TE2000E) of carotid artery sections were acquired to identify single GFP+ cells co-expressing smooth muscle marker and to avoid false positive images due to overlapping cells. The acquired images were coded and analysed by two observers blinded to the codes. The 3D reconstruction and analysis of the images was performed using IMARIS software (Bitplane Scientific Software, Zurich, Switzerland).
In vitro cell tubulation and tunneling assay in 3D matrix
CX3CR1 positive cells were FACS sorted from whole bone marrow of CX3CR1gfp/+ and CX3CR1gfp/gfp mice and were sandwiched between DQ-red BSA (Invitrogen; Bio-Sciences, Dublin, Ireland) fortified Matrigel layers with CX3CL1 gradient in a 15 μ-angiogenesis slide (Ibidi, Munich, Germany). CX3CL1 gradient was created by serially layering two layers of Matrigel consisting of 50 and 100 nM rmCX3CL1 respectively. The cells were imaged using high-resolution sequential confocal scans (Nikon eC1 plus, TE2000E) at 3 hrs and every day up to day 7 post seeding. A separate series of experiments were conducted in the absence and presence of Rho inhibitor (10 µM; Y27632; MERCK, Darmstadt, Germany). The acquired confocal images were 3D reconstructed and analysed using IMARIS software.
Extracellular matrix detection
CX3CR1 positive cells were FACS sorted from whole bone marrow of CX3CR1gfp/+ and CX3CR1gfp/gfp mice and were grown on 6 well plates at a seeding density of 7×105 cells per well for 5 days with or without 100 nM rmCX3CL1 stimulation. In a parallel set of experiments the cell viability was assessed using FACS at day 5 using 7AAD staining. On day 5 the cells were removed by treating with 500 µl of 20 mM Ammonium Hydroxide +0.1% Triton X-100 incubated for 5 mins at 37°C. Following six washes with 1 ml TBS buffer (Tris HCL 50 mM, NaCl 100 mM, pH 8.0), ECM on the plate was removed in the presence of 1% SDS for Western blot analysis. Equal total protein was loaded for Western blot analysis and β-actin was probed as a loading control. The Laminin bands were quantified by digital densitometry normalised to density of respective β-actin bands.
RhoA activation assay
Active RhoA expression during the in vitro cell tubulation and tunneling process was assessed using a commercial RhoA activation assay kit (Millipore Ireland BV, Cork, Ireland). Briefly, CX3CR1 positive cells FACS sorted from whole bone marrow of CX3CR1gfp/+ mice were plated in 6 well plates at a density of 7×105 cells per well and were stimulated with 100 nM rm CX3CL1 for 30 mins. The cells were then lysed (Mg2+ lysis/wash buffer) and the lysates were incubated with GST-tagged Rhotekin Rho binding domain for affinity precipitation of GTP-Rho. The bound GTP-Rho protein was pulled-down and resolved on 12% SDS-PAGE. The gel was transferred onto nitrocellulose membrane and detected by immunoblotting using 1∶250 anti RhoA primary antibody (clone 55) and the appropriate peroxidase-conjugated secondary antibody and luminol based chemiluminescent HRP substrate (Millipore Ireland BV, Cork, Ireland). Whole cell lysate samples from all groups were also analysed by SDS-PAGE as total RhoA controls. The active-RhoA bands were quantified by digital densitometry normalised to density of respective total RhoA bands.
All data are represented as mean ± SEM and analysed by unpaired Student t-test or analysis of variance (ANOVA) using Graphpad Prism Version 4 (GraphPad Software, Inc., California, USA). Animal experiments with tissue section analysis were compared based on intra-experiment variation involving 3–4 mice/group that were not littermates.
Bone marrow derived CX3CR1 positive cells contribute to formation of microvessels
In this study, myeloid phenotype of CX3CR1 expressing cells was established by FACS and tissue immunofluorescence (Figure S1 A, B.) indicating that the CX3CR1 transgenic models selected would be suitable for assessing structure function relationships between monocyte/macrophages and evolving neovascularization networks. We first assessed whether CX3CR1 positive cells contribute to microvasculature in a well-established in vivo model of Matrigel plug neovascularization. Following implantation of Matrigel (500 µl) in functional CX3CR1gfp/+ and functionally incompetent CX3CR1gfp/gfp mice microvascular association (perivascular) and integration(vascular) of CX3CR1 positive cells was assessed by immunofluorescence imaging, high resolution confocal microscopy with Z stacking and volume rendering. Functioning microvessels (MV) were identified by intravital staining of rhodamine conjugated lectin injected via tail vein. CX3CR1 positive cells were noted in the perivascular region of MV and integrated within the walls of MV (Figure 1 A–C). However Matrigels implanted in CX3CR1gfp/+ mice had significantly increased CX3CR1 positive cells integrated into MV wall (4.09±0.34 cells/Matrigel MV/HPF) compared to Matrigels implanted in CX3CR1gfp/gfp mice (1.33±0.05 cells/Matrigel MV/HPF) (Figure 1 B, p<0.05) and predominantly a perivascular MV association was observed for CX3CR1 positive cells in Matrigel of CX3CR1gfp/gfp mice (Figure 1B). Furthermore, we observed increased incidence of leaky vessels in the absence of functional CX3CR1, which was evident from increase in Matrigel heme content (Figure 1 D, E, p<0.05), Evans blue dye leak (Figure 1 F, G, p<0.05), and extravasated platelet antigenicity in the extravascular space (Figure 1 H, I, p<0.05) in Matrigel plugs isolated from CX3CR1gfp/gfp mice compared to functional CX3CR1gfp/+ mice.
500 µl of Matrigel subcutaneously implanted in CX3CR1+/gfp and CX3CR1gfp/gfp mice for 4 weeks. A, Representative sections of Matrigel isolated from CX3CR1+/gfp and CX3CR1gfp/gfp mice, intravitally stained with lectin (Rhodamine Griffonia Simplicifolia; Red) to show functional MV and DAPI (Nucleus; Blue) show presence of CX3CR1 (GFP) positive MV. B, Deficiency of CX3CR1 restricted CX3CR1 cells to mostly the perivascular region of functional MV within Matrigel. Moreover, the number of integrated vascular associated CX3CR1 positive MV were significantly higher in Matrigels isolated from CX3CR1+/gfp mice compared to Matrigels isolated from CX3CR1gfp/gfp mice. Data is expressed as mean ± SEM of 25 Matrigel sections/mice (n = 3 independently performed experiments). C, Three dimensional reconstruction of the confocal images using IMARIS indicate CX3CR1 positive cells integrated into periluminal vascular wall in the Matrigel implanted in CX3CR1+/gfp but not in CX3CR1gfp/gfp mice where cells remained in perivascular region (Scale bar: 10 µm). Microvessels formed in Matrigels isolated from CX3CR1gfp/gfp mice were leaky, which was evident by increased heme content (D & E), Evan's blue dye leak (F & G), and extravasated platelets (CD42b; Red) (H & I) in the matrix compared to Matrigels implanted in CX3CR1+/gfp mice. Data is represented as mean ± SEM of 8 Matrigels/group and 10 Matrigel sections/mice (n = 4 independently performed experiments); * denotes p<0.05.
Functional CX3CR1 positive cells contribute to formation of plaque microvessels
Having established the perivascular and intravascular association of CX3CR1 cells with angiogenic MV in Matrigel, we investigated whether such an association exists with intimal MV using a recently described carotid artery ligation model of plaque neovascularization in mice . We first established consistent development of carotid artery plaque that was rich in intimal microvessels (MV) with active perfusion as evidenced by presence of red blood cells(RBCs) within MV lumina at 4 weeks post arterial ligation (Figure 2 A). Sham operated animals did not develop plaque neovascularization (data not shown). Significantly greater numbers of intimal MV containing RBCs were noted in CX3CR1gfp/+ compared to functionally deficient CX3CR1gfp/gfp mice (Figure 2 B, p<0.05). Indeed similar to Matrigel neovascularization experiments there was a significantly greater (p<0.05) vascular contribution of CX3CR1 positive cells to formation of MV (Figure 2 C, D, E, Figure S2 D, E) in the carotid artery plaque intima in CX3CR1gfp/+ animals and CX3CR1 positive cells were observed in perivascular association with MV within plaque in significantly greater numbers (p<0.05) compared to CX3CR1gfp/gfp mice (Figure 2 C, D; Figure S2 D, E). Moreover ∼30% of all MV co- expressed the myeloid marker F4/80 (Figure S2 A, B) and 75% of all GFP positive cells co-expressed F4/80 (Figure S1 B). The total number of GFP positive cells in neointima was similar in CX3CR1gfp/+ and CX3CR1gfp/gfp mice indicating additional chemokine signals to CX3CL1 (Figure S2 C) may be involved in recruitment of monocytes/macrophages to the neointima. Functional deficiency of CX3CR1 significantly impaired CX3CR1 cell contribution to formation of MV in neointima and in these animals signet-ring like structures were observed. In addition, the total number of signet-ring like cells per unit area were significantly reduced in CX3CR1gfp/gfp mice compared to competent CX3CR1gfp/+ mice (Figure 2 F, G, p<0.05). Of the total MV in the plaque 14.4±1.6% MV were positive for CX3CR1 cells associated with MV wall, while 6.8±0.9% MV had perivascularly associated CX3CR1 cells (Figure 2 D). However this was significantly (p<0.05) impaired by the functional deficiency of CX3CR1 (5.7±0.6% and 1.6±0.1%; vascular wall and perivascular cells respectively). Furthermore 11.1±2.3% MV were positive for CX3CR1 cells co-expressing smooth muscle markers (Calponin or α-smooth muscle actin) and this was significantly impaired in the CX3CR1gfp/gfp mice (1.8±1.3% MV) (Figure 2 E, Figure S2 D, E, p<0.01). Therefore with respect to intimal microvessel formation in this model CX3CR1 positive cells formed either signet ring-like structures, were integrated into MV wall or were perivascularly associated (Figure 2 H) with these MV phenotypes being more mature and more prevalent in the CX3CR1gfp/+ functionally competent animals.
Plaque angiogenesis was created by ligation of carotid artery at its bifurcation for 4 weeks in CX3CR1+/gfp and CX3CR1gfp/gfp mice. Perfusion fixed carotid arteries were isolated and OCT embedded. 5 µm cross sections were stained with smooth muscle marker (Calponin; Red) and confocal images were acquired. A, Representative bright field cross section image of carotid artery from CX3CR1+/gfp mice stained with DAPI (Nucleus; Blue). Red blood cells (RBCs) (arrow heads) were observed in the microvessels indicating these microvessels were functional. B, Mice competent for CX3CR1 function had a higher proportion of MV containing RBCs C, Representative cross section image of carotid artery from CX3CR1+/gfp and CX3CR1gfp/gfp mice were stained with calponin (Red) and DAPI (nucleus; blue). CX3CR1 positive cells (GFP positive; Green) integrated into microvascular wall and were also present in perivascular region and co-expressed smooth muscle marker (Calponin; Red) (Scale bar: 10 µm). In the CX3CR1 functionally deficient (CX3CR1gfp/gfp) mice the number of vascular and perivascular cells (GFP positive; Green) (D) and co-expressing smooth muscle marker (E) were significantly reduced. Data is expressed as mean ± SEM of 20 carotid artery cross sections/mice (n = 4 independently performed experiments). F, Microvessels in the plaque stained with CX3CR1 (Green) and DAPI (nucleus; blue) were 3D reconstructed using IMARIS software to depict signet ring structures and their tubular architecture. G, The number of signet-ring cells was significantly reduced in CX3CR1gfp/gfp mice. H, Schematic representing the three major phenotypes of CX3CR1 cells associated with microvascular structures in the plaque. * denotes p<0.05.
Functional deficiency or pharmacological inhibition of CX3CR1 results in decreased MV formation and leaky MV phenotype in experimental plaque neovascularization
In an experimental model of plaque neovascularization, functional deficiency of CX3CR1 increased intraplaque haemorrhage (Figure S3 A, B, p<0.01) and enhanced extravasation of platelets in the plaque interstitial space in CX3CR1gfp/gfp mice compared to CX3CR1 competent animals (Figure 3 A, B, p<0.05), thus indicating a leaky MV phenotype. The leaky MV phenotype in these mice was associated with reduced extracellular matrix and MV basement membrane staining (Figure 3 C, D, p<0.05) in addition to impaired association of CX3CR1 cells derived smooth muscle-like cells.
Representative cross sectional images of carotid artery plaques from CX3CR1+/gfp and CX3CR1gfp/gfp mice stained with CD42b (Platelets; Red) (A) or laminin (Basement membrane; Red) (C) and DAPI (Nucleus; Blue). B, Significantly increased staining for platelet CD42b was observed in the neointimal interstitial space in CX3CR1gfp/gfp mice compared to competent CX3CR1+/gfp mice (Scale bar: 50 µm). D, A significantly greater number of CX3CR1 positive microvessels were covered by basement membrane laminin in CX3CR1+/gfp mice compared to CX3CR1gfp/gfp mice (insets in panel C show GFP and laminin co-staining present in CX3CR1+/gfp but not CX3CR1gfp/gfp mice). Data is represented as mean ± SEM of 15 plaque sections/mice (n = 4 independently performed experiments); * denotes p<0.05.
To determine whether the effects of genetic functional deletion of CX3CR1 receptor on plaque neovascularization could be recapitulated by pharmacological inhibition of receptor activity, we treated wild type C57BL/J6 mice with a CX3CL1 analogue (F1 peptide) inhibitor of CX3CR1 receptor. The in vivo dose of F1 peptide was optimised based on the capacity of F1 peptide to inhibit binding of CX3CR1 positive cells to CX3CL1 coated plates in a static adhesion assay (Figure S4) and on previously reported pharmacodynamic data . 50 µg of F1 peptide was repeatedly administered every third day for 2 weeks by intraperitoneal injection starting from 2 weeks post carotid artery ligation (Figure 4 A). As observed in CX3CR1gfp/gfp mice, F1 treated wild type mice exhibited significantly reduced staining for MV associated basement membrane laminin within the plaque (Figure 4 B, C, p<0.01), had reduced extracellular matrix collagen staining (Figure S5 A, B, p<0.01) and had leaky MV phenotype indicated by increased extravasation of platelets (Figure 4 D, E, p<0.01) and intravenously administered microspheres (Figure 4 F, G, p<0.01) in the intimal interstitial space.
A, Drug (F1) study protocol: Carotid artery was ligated in the C57BL6J mice (wt mice) and animals were treated with a selective CX3CR1 inhibitor (F1) (from second week post carotid ligation for another two weeks). B, Representative cross sectional images of carotid artery from C57BL6J mice treated with saline or a selective CX3CR1 inhibitor (F1) and stained for laminin (B; Basement membrane; Red), or CD42b (D; Platelets; Red), CX3CR1 (GFP; Green) and DAPI (Nucleus; Blue). B & C, Significantly greater number of CX3CR1 positive microvessels were covered by basement membrane laminin in saline treated C57BL6J mice compared to F1 treated mice. D & E, Increased platelet CD42b staining was observed in the neointimal interstitial space in the F1 treated mice compared to saline treated mice (Scale bar: 50 µm). In addition the leaky microvessel phenotype in mice treated with F1 was confirmed by presence of intravenously administered (tail vein) 2–2.5 µm diameter microspheres (red spheres) in the neointimal lesion (F & G). IgG control staining for isotype-matched antibodies shown. Data is represented as mean ± SEM of 15 plaque sections/mice (n = 4 independently performed experiments); * denotes p<0.01.
Functional deficiency of CX3CR1 impairs 3D Matrigel monocyte/macrophage morphogenesis (tubulation) and tunneling capacity in vitro
To evaluate the morphogenic and cytokinetic effects of CX3CR1- CX3CL1 interaction cells sandwiched between 2 Matrigel layers were examined in 3D for tube formation in presence or absence of CX3CL1 stimulation. CX3CR1gfp/+ cells formed tube-like structures with clear lumen, an effect augmented by CX3CL1 stimulation and in contrast, CX3CR1 deficient cells showed impaired tubulation in presence and absence of CX3CL1 (Figure 5 A, B, p<0.01). Overall CX3CR1 positive cells isolated from both CX3CR1gfp/+ and CX3CR1gfp/gfp mice lacked the ability to form complete tube like structures in the absence of CX3CL1 gradient and the number of incompletely tubulated cells was significantly higher under CX3CR1 deficient conditions (Figure 5 B, p<0.01). Matrigel tunnelling capacity (Figure 5 C, D) and extracellular matrix (laminin) synthesis (Figure 5 E, F, p<0.01) by CX3CR1- CX3CL1 stimulated cells isolated from CX3CR1gfp/gfp mice was also significantly reduced compared to CX3CR1 positive cells isolated from CX3CR1gfp/+ mice. Importantly equal numbers of cells were viable (44.8–47.2%) in all groups at day 5 in culture. Moreover CX3CR1 positive cells from CX3CR1gfp/+ mice developed larger tubular structures of greater length and with greater external and internal luminal diameter compared to their counterparts from CX3CR1gfp/gfp mice (Figure S6 A, B, C, p<0.01). As cellular vacuolation and tubulation have been directly associated with RhoA driven cytoskeletal reorganization , , , we examined active RhoA expression in CX3CL1 stimulated CX3CR1 positive cells isolated from CX3CR1gfp/+ and CX3CR1gfp/gfp mice. Initially we observed increased activation of RhoA at 30 minutes post CX3CL1 stimulation of CX3CR1 positive cells isolated from CX3CR1gfp/+ but not CX3CR1gfp/gfp mice (Figure 6 A, B, p<0.01). Subsequent inhibition of Rho with an antagonist Y27632 severely impaired RhoA activation (Figure 6 A, B, p<0.01) and the capacity of CX3CR1 positive cells to form complete tubular structures in 3D Matrigel sandwich (Figure 6 C, D, p<0.01).
CX3CR1 cells isolated from bone marrow of CX3CR1+/gfp or CX3CR1gfp/gfp mice were sandwiched between two DQ-red fortified (20 µg/ml; substrate for proteases) Matrigel layers with or without 10 ng/ml CX3CL1 gradient and were cultured for 5 days. A & B, CX3CR1 cells remodelled into tube-like structures (tubulation) especially under CX3CL1 gradient (Scale bar: 10 µm). However these tubular structures were incompletely formed by CX3CR1 deficient cells. Furthermore, the CX3CR1 deficient cells lacked tunneling capacity (dotted lines-protease activation red) in the Matrigel (C & D) and produced significantly less laminin following CX3CL1 stimulation (E & F). Data is represented as mean ± SEM of 4 independently performed experiments;* denotes p<0.01.
CX3CR1 cells isolated from CX3CR1+/gfp and CX3CR1gfp/gfp mice were treated with 100 nM CX3CL1 and 30 minutes post probed for expression of activated RhoA. A, Representative immuno blot showing active RhoA protein expression. B, Active RhoA expression within 30 mins post CX3CL1 stimulation was significantly higher in CX3CR1 cells isolated from CX3CR1+/gfp compared to CX3CR1gfp/gfp mice. C, CX3CR1 cells isolated from bone marrow of CX3CR1+/gfp mice were sandwiched between two Matrigel layers with 10 ng/ml CX3CL1 gradient and were cultured for 5 days in absence or presence of Rho inhibitor (Y27632; 10 µM, Scale bar: 10 µm). C & D, Inhibition of Rho using Y27632 (10 µM) prevented the formation of tube-like structures by CX3CR1 cells isolated from CX3CR1+/gfp mice. Data is represented as mean ± SEM of 4 independently performed experiments;* denotes p<0.01.
The major and new finding of this study is that functional competence of the monocyte/macrophage chemokine receptor CX3CR1 is essential for nascent microvessel formation, structural integrity and maturation in two models of neovascularization. Moreover, monocyte/macrophage tubulation, extracellular matrix tunneling and differentiation into smooth muscle-like cells within the microvasculature in vivo is driven in part by CX3CR1 interaction with its cognate ligand. CX3CR1-CX3CL1 association activates RhoA signalling, a central pathway in neovascularization/angiogenesis initiation and maintenance . Abrogation of CX3CL1-CX3CR1 interaction reduces mural cell investment of nascent proangiogenic tubes in both models of neovascularization and reduces extracellular matrix protein production of laminin and collagen in a plaque neovascularization model (Figure S5). Finally loss of CX3CR1 function or therapeutic targeting of CX3CR1 interaction using a mutant peptide CX3CL1-mimetic leads to development of smaller, poorly developed, leaky and haemorrhagic microvessels in Matrigel neovascularization and experimental plaque neovascularization models respectively.
Monocyte/macrophage regulation of neovascularization has been extensively investigated over the past decade and is known to involve both paracrine  and cell-cell interactions . Moreover, pioneering studies by Moldovan and colleagues have demonstrated that monocyte-macrophages drill tunnels in extracellular matrix to facilitate initial ingrowth of new capillaries , , , . Thus monocyte/macrophages can provide paracrine and scaffold support for initiation, expansion and maturation of microvascular networks. However little is known about which myeloid cell population or surface receptors are implicated in signalling and morphogenesis and maturation of neovascular networks.
In both models of neovascularization used in our study it is likely that CX3CL1  as well as other chemokines ,  initiated recruitment of CX3CR1 positive cells into areas of microvessel growth. Although CX3CR1 is the only known receptor for CX3CL1, a recent study reported CX3CL1 induced angiogenesis, cell tubulation and signalling being partially insensitive to pertussis toxin pre-treatment . Interestingly we also observed CX3CL1 induced tubulation events in CX3CR1 deficient cells (Figure 5B) and signet ring like structures in carotid artery plaques from CX3CR1 deficient mice (Figure 2C), thus indicating possibility of CX3CR1 independent effects of CX3CL1. Nevertheless, we show here that for the most part competence of CX3CR1 cells to interact with CX3CL1 ligand likely determines the spectrum of cell differentiation events that subsequently occurs post chemokine-receptor interaction , , . Previously, we have identified smooth muscle cell differentiation as an important effect of CX3CL1-CX3CR1 interaction post arterial wire injury . In the current study we extend these findings to a spectrum of additional downstream effects starting with augmented RhoA signalling, a pathway central to neovascularization/angiogenesis initiation with key functions regulating actin cytoskeleton, cell polarity and vacuolation . We show in vitro that CX3CR1- CX3CL1 interaction increases RhoA activation and that monocyte/macrophage vacuolation downstream of CX3CL1 activation is completely abrogated by a Rho antagonist-Y27632 suggesting this cytoskeletal effect of CX3CR1 receptor activation may be driven at least in part by a Rho dependent mechanism. Since Rho activation initiates multiple other downstream pathways including smooth muscle cell differentiation  and extracellular matrix production , , , ,  it is not surprising that CX3CR1– CX3CL1 interaction also augments laminin production and extracellular matrix tunnelling and migration activity of these activated monocyte/macrophages.
Consistent with these in vitro findings we show that CX3CR1 positive cells also exhibit a distinct tubulation phenotype change in vivo in the form of signet ring cell (a single DAPI nucleus and GFP positive ring cytoplasm) formation (Figure 2F). Monocyte/macrophage tubulation (signet-ring formation) was impaired in CX3CR1 deficient mice and together with previous work by Moldovan and colleagues ,  it is conceivable that impairment of Rho dependent cytoskeletal changes may have impaired initiation and maturation of nascent microvessels in these animals. Although no RBCs were observed within these tubular single cell structures we could readily demonstrate RBCs within two other MV-associated CX3CR1 phenotypes namely perivascular and mural (vascular) cell types. Consistent with our previous evidence for CX3CR1 differentiation into SM-like cells we show that almost all mural CX3CR1 positive cells within MVs co-express smooth muscle markers , . More importantly functional deficiency of CX3CR1 markedly reduced in vivo mural and perivascular association of CX3CR1 cells within plaque MVs and mural association within Matrigel MV. Functional deficiency of CX3CR1 also reduced MV associated extracellular matrix production with more than 70% reduction in MV-associated laminin and collagen staining in CX3CR1 deficient mice. The functional consequences of aberrant CX3CR1 phenotype included increased microvascular leakage of blood, platelets, and intravascularly administered solutes (Evans Blue dye). It is tempting to speculate whether failure of CX3CR1 deficient cells to tubulate, or initiate differentiation into mural smooth muscle-like cells or impairment of extracellular matrix production may be significant contributors to increased microvessel porosity to blood and solutes in vivo thus providing a mechanism underlying the leaky MV phenotype. It is also conceivable that reduced mural CX3CR1 positive cells may have effects on microvascular endothelial permeability .
Our current findings contribute to emerging literature on the direct involvement of monocyte/macrophages to smooth muscle cell differentiation , ,  and extra cellular matrix , ,  production within the vessel wall and supports the hypothesis of CX3CR1 positive cells contributing to formation of stable MV within neovascular networks and within experimental plaque. The monocyte/macrophage contribution to stable MV may thus occur via a number of mechanisms including inter alia differentiation of myeloid cells into smooth muscle cells, intra-mural and perivascular cell integration, synthesis of basement membrane, and paracrine secretion of other extracellular matrix support factors by these cells. Additionally CX3CR1 positive cells contribute to formation of stable MV in a plaque-independent setting, as indicated by our in vivo Matrigel induced neovascularization data.
While the implantation of a subcutaneous Matrigel plug is accepted as an excellent model of in vivo neovascularization, controversy remains on the appropriateness of current plaque neovascularization models used in mice. This limitation was recently addressed by Sasaki et al  wherein ligation of common carotid artery in a murine model was shown to develop a lesion with reproducible intimal neovascularization . Thus we adapted this model for our current CX3CR1 studies and provide clear evidence of functional MV formation in this model using high resolution 3D confocal imaging.
Given that microvessel integrity and function is increasingly investigated as a potential contributor to plaque growth and stability we examined whether therapeutic targeting of CX3CR1 would alter MV structure and function. Remarkably, treatment of wild type mice with a CX3CL1 peptide-mimetic (F1) that specifically antagonized CX3CR1 recapitulated the leaky microvessel phenotype seen in CX3CR1 deficient mice. These F1 treated animals also exhibited marked reduction in basement membrane-laminin MV coverage, increased platelet extravasation and excess leakage of intravascularly administered microspheres compared to control treated animals. These data confirm that CX3CR1 deficiency or specific small molecule targeting of CX3CR1 reduces neovascularization (Figure S7, S8) and alters microvascular structure in vivo. Our data also illustrate that therapeutic interference in microvascular structure-function may potentially be a double edged sword with unintended consequences such as microvascular leakage. This is especially relevant given the recent increase in cardiovascular events observed in patients receiving anti-angiogenic therapy for cancer (VEGF inhibitor-Avastin) . Thus, although enthusiasm is growing for targeting plaque neovascularization in humans our data suggests caution is required as complex interactions between evolving microvasculature and plaque growth and stability may not be easily amenable to single molecule therapeutic approaches.
Myeloid phenotype of CX3CR1-GFP cells. A, Flow cytometric analysis of the CD45/CX3CR1-GFP dual positive cell population of mouse bone marrow cells showing ∼85% or greater coexpression of GFP cells with cfms, CD11b or F4/80, with lower levels of coexpression with Gr-1. Data is representative of 3 independently performed experiments, the values inset are mean ± SEM of 3 independently performed experiments. B, CX3CR1-GFP and F4/80 co-staining in plaque (upper panels) and Matrigel angiogenesis (lower panels) models showing majority of CX3CR1-GFP cells positive for myeloid marker (F4/80). Data is expressed as mean ± SEM of 10 carotid artery/Matrigel cross sections/mice (n = 3 independently performed experiments).
Contribution of myeloid cells to formation of microvessels in neointima. A, Representative cross section image of carotid artery from wild type mice stained with myeloid (F4/80) and smooth muscle (calponin) marker. Nuclei are stained with DAPI (Blue). B, ∼30% of microvessels in the neointima were positive for myeloid marker and ∼15% of microvessels co-expressed myeloid and smooth muscle marker. C, The total number of GFP positive cells in the neointima was similar in CX3CR1+/gfp and CX3CR1gfp/gfp mice. Data is expressed as mean ± SEM of 20 carotid artery cross sections/mice (n = 3 animals). D, Representative cross section image of carotid artery from CX3CR1+/gfp and CX3CR1gfp/gfp mice were stained with α-smooth muscle actin (α-SMA; Red) and DAPI (nucleus; blue). CX3CR1 positive cells (GFP positive; Green) integrated into microvascular wall and were also present in perivascular region and co-expressed smooth muscle marker (α-SMA; Red) (Scale bar: 10 µm). E, In the CX3CR1 functionally deficient (CX3CR1gfp/gfp) mice the number microvascular CX3CR1 positive cells co-expressing smooth muscle marker were significantly reduced. Data is expressed as mean ± SEM of 20 carotid artery cross sections/mice (n = 4 independently performed experiments). * denotes p<0.01.
Neointimal lesions from CX3CR1gfp/gfp mice show increased intra plaque haemorrhage. A, Representative cross section image of carotid artery from CX3CR1+/gfp and CX3CR1gfp/gfp mice stained with Hematoxylin and Eosin, showing increased intraplaque haemorrhage (Dark red staining). B, Data is expressed as mean ± SEM of 20 carotid artery cross sections/mice (n = 4 independently performed experiments); * denotes p<0.01.
Dose dependent inhibition of CX3CR1 cells to CX3CL1 coated plates by selective CX3CR1 inhibitor (F1) in a static cell adhesion assay. Data is represented as mean ± SEM of 4 independently performed experiments; * denotes p<0.01.
A, Pharmacological inhibition of CX3CR1 receptors using CX3CR1 inhibitor (F1) reduces collagen staining (light blue on trichrome) in the neointima. B, Data is expressed as mean ± SEM of 20 carotid artery cross sections/mice (n = 4 independently performed experiments); * denotes p<0.01.
Cell length (A), cell diameter (B) and lumen diameter (C) of CX3CR1 cells isolated from CX3CR1+/gfp or CX3CR1gfp/gfp mice cultured for 5 days in Matrigel sandwich with 10 ng/ml CX3CL1 gradient. Data is represented as mean ± SEM of 4 independently performed experiments;* denotes p<0.01.
Neointimal lesions from CX3CR1gfp/gfp mice are significantly reduced compared to CX3CR1+/gfp mice. Data is expressed as mean ± SEM of 20 carotid artery cross sections/mice (n = 8 independently performed experiments); * denotes p<0.01.
Pharmacological inhibition of CX3CR1 receptors using CX3CR1 inhibitor (F1) significantly reduces the neointimal area. Data is expressed as mean ± SEM of 15 carotid artery cross sections/mice (n = 4 independently performed experiments); * denotes p<0.01.
Conceived and designed the experiments: AK NC. Performed the experiments: AK KM ET CB. Analyzed the data: AK KM. Contributed reagents/materials/analysis tools: KD PD. Wrote the paper: AK NC.
- 1. Angelillo-Scherrer A (2012) Leukocyte-derived microparticles in vascular homeostasis. Circ Res 110: 356–369.
- 2. Swirski FK, Weissleder R, Pittet MJ (2009) Heterogeneous in vivo behavior of monocyte subsets in atherosclerosis. Arterioscler Thromb Vasc Biol 29: 1424–1432.
- 3. Weber C, Zernecke A, Libby P (2008) The multifaceted contributions of leukocyte subsets to atherosclerosis: lessons from mouse models. Nat Rev Immunol 8: 802–815.
- 4. Anghelina M, Krishnan P, Moldovan L, Moldovan NI (2006) Monocytes/macrophages cooperate with progenitor cells during neovascularization and tissue repair: conversion of cell columns into fibrovascular bundles. Am J Pathol 168: 529–541.
- 5. Finn AV, Jain RK (2010) Coronary plaque neovascularization and hemorrhage: a potential target for plaque stabilization? JACC Cardiovasc Imaging 3: 41–44.
- 6. Jain RK, Finn AV, Kolodgie FD, Gold HK, Virmani R (2007) Antiangiogenic therapy for normalization of atherosclerotic plaque vasculature: a potential strategy for plaque stabilization. Nat Clin Pract Cardiovasc Med 4: 491–502.
- 7. Carmeliet P, Jain RK (2000) Angiogenesis in cancer and other diseases. Nature 407: 249–257.
- 8. Lucerna M, Zernecke A, de Nooijer R, de Jager SC, Bot I, et al. (2007) Vascular endothelial growth factor-A induces plaque expansion in ApoE knock-out mice by promoting de novo leukocyte recruitment. Blood 109: 122–129.
- 9. Lobov IB, Rao S, Carroll TJ, Vallance JE, Ito M, et al. (2005) WNT7b mediates macrophage-induced programmed cell death in patterning of the vasculature. Nature 437: 417–421.
- 10. Kinnaird T, Stabile E, Burnett MS, Shou M, Lee CW, et al. (2004) Local delivery of marrow-derived stromal cells augments collateral perfusion through paracrine mechanisms. Circulation 109: 1543–1549.
- 11. Kumar AH, Caplice NM (2010) Clinical potential of adult vascular progenitor cells. Arterioscler Thromb Vasc Biol 30: 1080–1087.
- 12. Moldovan NI (2002) Role of monocytes and macrophages in adult angiogenesis: a light at the tunnel's end. J Hematother Stem Cell Res 11: 179–194.
- 13. Moldovan NI, Goldschmidt-Clermont PJ, Parker-Thornburg J, Shapiro SD, Kolattukudy PE (2000) Contribution of monocytes/macrophages to compensatory neovascularization: the drilling of metalloelastase-positive tunnels in ischemic myocardium. Circ Res 87: 378–384.
- 14. Tigges U, Hyer EG, Scharf J, Stallcup WB (2008) FGF2-dependent neovascularization of subcutaneous Matrigel plugs is initiated by bone marrow-derived pericytes and macrophages. Development 135: 523–532.
- 15. Goichberg P, Bai Y, D'Amario D, Ferreira-Martins J, Fiorini C, et al. (2011) The ephrin A1-EphA2 system promotes cardiac stem cell migration after infarction. Circ Res 108: 1071–1083.
- 16. Metharom P, Kumar AH, Weiss S, Caplice NM (2010) A specific subset of mouse bone marrow cells has smooth muscle cell differentiation capacity-brief report. Arterioscler Thromb Vasc Biol 30: 533–535.
- 17. Kumar AH, Metharom P, Schmeckpeper J, Weiss S, Martin K, et al. (2010) Bone marrow-derived CX3CR1 progenitors contribute to neointimal smooth muscle cells via fractalkine CX3CR1 interaction. FASEB J 24: 81–92.
- 18. Lesnik P, Haskell CA, Charo IF (2003) Decreased atherosclerosis in CX3CR1-/- mice reveals a role for fractalkine in atherogenesis. J Clin Invest 111: 333–340.
- 19. Liu P, Patil S, Rojas M, Fong AM, Smyth SS, et al. (2006) CX3CR1 deficiency confers protection from intimal hyperplasia after arterial injury. Arterioscler Thromb Vasc Biol 26: 2056–2062.
- 20. Pamukcu B, Lip GY, Devitt A, Griffiths H, Shantsila E (2010) The role of monocytes in atherosclerotic coronary artery disease. Ann Med 42: 394–403.
- 21. Lee SJ, Namkoong S, Kim YM, Kim CK, Lee H, et al. (2006) Fractalkine stimulates angiogenesis by activating the Raf-1/MEK/ERK- and PI3K/Akt/eNOS-dependent signal pathways. Am J Physiol Heart Circ Physiol 291: H2836–2846.
- 22. Combadiere C, Feumi C, Raoul W, Keller N, Rodero M, et al. (2007) CX3CR1-dependent subretinal microglia cell accumulation is associated with cardinal features of age-related macular degeneration. J Clin Invest 117: 2920–2928.
- 23. Jung S, Aliberti J, Graemmel P, Sunshine MJ, Kreutzberg GW, et al. (2000) Analysis of fractalkine receptor CX(3)CR1 function by targeted deletion and green fluorescent protein reporter gene insertion. Mol Cell Biol 20: 4106–4114.
- 24. Dorgham K, Ghadiri A, Hermand P, Rodero M, Poupel L, et al. (2009) An engineered CX3CR1 antagonist endowed with anti-inflammatory activity. J Leukoc Biol 86: 903–911.
- 25. Haskell CA, Cleary MD, Charo IF (1999) Molecular uncoupling of fractalkine-mediated cell adhesion and signal transduction. Rapid flow arrest of CX3CR1-expressing cells is independent of G-protein activation. J Biol Chem 274: 10053–10058.
- 26. Sasaki T, Kuzuya M, Nakamura K, Cheng XW, Shibata T, et al. (2006) A simple method of plaque rupture induction in apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol 26: 1304–1309.
- 27. Sinclair PR, Gorman N, Jacobs JM (2001) Measurement of heme concentration. Curr Protoc Toxicol Chapter 8: Unit 8 3.
- 28. Ryu J, Lee CW, Hong KH, Shin JA, Lim SH, et al. (2008) Activation of fractalkine/CX3CR1 by vascular endothelial cells induces angiogenesis through VEGF-A/KDR and reverses hindlimb ischaemia. Cardiovasc Res 78: 333–340.
- 29. Bryan BA, D'Amore PA (2007) What tangled webs they weave: Rho-GTPase control of angiogenesis. Cell Mol Life Sci 64: 2053–2065.
- 30. Bell SE, Mavila A, Salazar R, Bayless KJ, Kanagala S, et al. (2001) Differential gene expression during capillary morphogenesis in 3D collagen matrices: regulated expression of genes involved in basement membrane matrix assembly, cell cycle progression, cellular differentiation and G-protein signaling. J Cell Sci 114: 2755–2773.
- 31. Nucera S, Biziato D, De Palma M (2011) The interplay between macrophages and angiogenesis in development, tissue injury and regeneration. Int J Dev Biol 55: 495–503.
- 32. Fantin A, Vieira JM, Gestri G, Denti L, Schwarz Q, et al. (2010) Tissue macrophages act as cellular chaperones for vascular anastomosis downstream of VEGF-mediated endothelial tip cell induction. Blood 116: 829–840.
- 33. Anghelina M, Krishnan P, Moldovan L, Moldovan NI (2004) Monocytes and macrophages form branched cell columns in matrigel: implications for a role in neovascularization. Stem Cells Dev 13: 665–676.
- 34. White GE, Tan TC, John AE, Whatling C, McPheat WL, et al. (2010) Fractalkine has anti-apoptotic and proliferative effects on human vascular smooth muscle cells via epidermal growth factor receptor signalling. Cardiovasc Res 85: 825–835.
- 35. Landsman L, Bar-On L, Zernecke A, Kim KW, Krauthgamer R, et al. (2009) CX3CR1 is required for monocyte homeostasis and atherogenesis by promoting cell survival. Blood 113: 963–972.
- 36. Martin K, Weiss S, Metharom P, Schmeckpeper J, Hynes B, et al. (2009) Thrombin stimulates smooth muscle cell differentiation from peripheral blood mononuclear cells via protease-activated receptor-1, RhoA, and myocardin. Circ Res 105: 214–218.
- 37. Schnoor M, Cullen P, Lorkowski J, Stolle K, Robenek H, et al. (2008) Production of type VI collagen by human macrophages: a new dimension in macrophage functional heterogeneity. J Immunol 180: 5707–5719.
- 38. Khalil N, Bereznay O, Sporn M, Greenberg AH (1989) Macrophage production of transforming growth factor beta and fibroblast collagen synthesis in chronic pulmonary inflammation. J Exp Med 170: 727–737.
- 39. Murray PJ, Wynn TA (2011) Protective and pathogenic functions of macrophage subsets. Nat Rev Immunol 11: 723–737.
- 40. Sprague L, Muccioli M, Pate M, Meles E, McGinty J, et al. (2011) The interplay between surfaces and soluble factors define the immunologic and angiogenic properties of myeloid dendritic cells. BMC Immunol 12: 35.
- 41. Pedraza C, Geberhiwot T, Ingerpuu S, Assefa D, Wondimu Z, et al. (2000) Monocytic cells synthesize, adhere to, and migrate on laminin-8 (alpha 4 beta 1 gamma 1). J Immunol 165: 5831–5838.
- 42. Schulz B, Pruessmeyer J, Maretzky T, Ludwig A, Blobel CP, et al. (2008) ADAM10 regulates endothelial permeability and T-Cell transmigration by proteolysis of vascular endothelial cadherin. Circ Res 102: 1192–1201.
- 43. Newby AC (2005) Dual role of matrix metalloproteinases (matrixins) in intimal thickening and atherosclerotic plaque rupture. Physiol Rev 85: 1–31.
- 44. Rader DJ, Daugherty A (2008) Translating molecular discoveries into new therapies for atherosclerosis. Nature 451: 904–913.
- 45. Doyle B, Caplice N (2007) Plaque neovascularization and antiangiogenic therapy for atherosclerosis. J Am Coll Cardiol 49: 2073–2080.
- 46. Sasaki T, Nakamura K, Kuzuya M (2007) Plaque rupture model in mice. Methods Mol Med 139: 67–75.
- 47. Daher IN, Yeh ET (2008) Vascular complications of selected cancer therapies. Nat Clin Pract Cardiovasc Med 5: 797–805.