|Home | About | Journals | Submit | Contact Us | Français|
Calcification of the vessel wall is a regulated process with many similarities to osteogenesis. Progenitor cells may play a role in this process. Previously, we identified a novel gene, Vascular Calcification Associated Factor (VCAF), which was shown to be important in pericyte osteogenic differentiation. The aim of this study was to determine the localization and expression pattern of VCAF in human cells and tissues. Immunohistochemical analysis of seven atherosclerotic arteries confirmed VCAF protein expression within calcified lesions. In addition, individual VCAF-positive cells were detected within the intima and adventitia in areas where sporadic 3G5-positive pericytes were localized. Furthermore, VCAF-positive cells were identified in newly formed microvessels in association with CD34-positive/CD146-positive/c-kit-positive cells as well as in intact CD31-positive endothelium in internal mammary arteries. Western blot analysis confirmed the presence of VCAF (18 kD) in protein lysates extracted from human smooth muscle cells, endothelial cells, macrophages, and osteoblasts. In fracture callus samples from three patients, VCAF was detected in osteoblasts and microvessels. This study demonstrates the presence of VCAF in neovessels and raises the possibility that VCAF could be a new marker for vascular progenitor cells involved in a number of differentiation pathways. These data may have implications for the prevention or treatment of vascular disease.
Treatment aimed at preventing vascular calcification in cardiovascular diseases, such as atherosclerosis, diabetes, end-stage renal disease, and vein graft failure, is limited by the lack of a full understanding of the cells or factors involved in this pathology.
In an earlier report, using subtraction hybridization, we identified a novel gene that was up-regulated during the osteogenic differentiation of bovine pericytes in vitro  We termed this gene Vascular Calcification Associated Factor (VCAF), as reverse transcriptase-polymerase chain reaction analysis and in situ hybridization demonstrated VCAF mRNA expression in human arterial calcified lesions, but not in non-calcified lesions . We also reported that knockdown of VCAF, using an adenoviral-mediated antisense-VCAF gene transfer strategy, caused an increase in pericyte nodule formation and size and accelerated pericyte mineralization in vitro . A subsequent study, using dexamethasone, caused accelerated osteogenic differentiation of pericytes in vitro, with a concomitant down-regulation of VCAF, matrix Gla protein (MGP), and osteopontin (OPN) . These data suggest that a host defence mechanism may up-regulate VCAF in order to inhibit calcification in vascular cells, thereby providing protection against osteogenic differentiation and/or mineralization in vivo. Despite these exciting new findings, the specific cell types expressing VCAF protein in human tissue were not known. Therefore, the purpose of the present study was to establish the localization of VCAF in normal and diseased tissue and determine the identity of the cells expressing VCAF.
Growing evidence suggests that the vessel wall contains a resident population of progenitor cells  which contribute to atherosclerosis [4-8], some of which are thought to be either endothelial or smooth muscle-derived, depending on the severity and duration of the vascular damage . Pericytes, postulated progenitor cells [10-13], have been localized in the adventitia, media, and intima of different sized arteries using a 3G5 antibody [14, 15]. Watson et al identified and characterized a subset of aortic vascular smooth muscle cells (VSMCs) , which have been termed calcifying vascular cells, because they can be induced to deposit a mineralized matrix in vitro. In addition, Tintut et al have shown the multilineage potentiality of calcifying vascular cells in vitro and suggested a phenotypic similarity to pericytes in this regard . Taken together, these studies suggest that multipotent, mesenchymal progenitor cells reside in the vessel wall and are recruited to an osteogenic lineage during active phases of vascular calcification [4, 5, 17, 18]. What is still unclear, is the origin of these progenitor cells and where they are located in the vessel wall. Strategies to identify the nature of these cells involves the use of specific markers of mesenchymal stem cells (MSCs), although specificity is dependent on the micro-environment . This study provides evidence for the association of VCAF protein with progenitor cell markers, raising the possibility that VCAF could act as a novel early marker for vascular progenitor cells which, if proved correct, might have implications for cellular, genetic, and tissue engineering approaches to vascular disease.
Ethical approval was granted for this study and procedures were in accordance with institutional guidelines. Atheromatous femoral arterial (FA) specimens (n = 18) from seven lower limb amputations; internal mammary artery (IMA) specimens (n = 3) from three patients; and six bone fracture callus = specimens from three patients were used. Vessels were fixed in phosphate-buffered formalin and characterized using von Kossa and alizarin red staining as previously described . FA segments exhibited complex lesions with calcification. Fracture callus biopsy samples were obtained from surgery for internal fixation or for malposition as part of routine treatment, and processed as previously described .
Sections (6 μm) were mounted onto positively charged slides (SuperFrost Plus; DAKO), dewaxed, and rehydrated. Antigens were retrieved by microwaving in 1mM citric acid (pH 6.0). Endogenous peroxidase activity was blocked with 3% H2O2 for 5 min and non-specific binding was blocked by incubation with either 10% goat or rabbit serum (DAKO) in 1% bovine serum albumen (BSA)/phosphate-buffered saline (PBS). Primary antibodies were diluted in 1% BSA/PBS, incubated at 22 °C for 1 h, and detected using biotin-conjugated secondary antibodies and the ABC system (DAKO, Ely, UK) followed by 3′, 3′-diaminobenzidine (Sigma, Poole, UK) staining. Sections were counterstained with Mayer’s haematoxylin (RA Lamb, Eastbourne, UK) and mounted (Vectastain: Vector, Peterborough, UK). Primary antibodies used were purified rabbit anti-VCAF  (1: 200; Sigma-Genosys), goat anti-c-kit (1: 50; sc-168; Santa Cruz, Heidelberg, Germany) or mouse monoclonal antibodies: OPN (1: 50; NCL-O-PONTIN; Novacastra, Newcastle, UK), CD68 (1: 100; M0814; DAKO), CD146 (1: 150; Mesoblast Ltd, Australia), CD34, CD31 (1: 30; R7170, M0823; DAKO), and 3G5 (neat hybridoma supernatant, ATCC, Teddington, UK). For c-kit, CD34, CD146, and VCAF staining, frozen sections (6 μm) were fixed in 50% acetone/50% methanol for 5 min and immunostained as above. For 3G5 staining, frozen sections were incubated with the antibody before fixing with 4% formaldehyde/PBS and immunostained as outlined above. Non-immune IgGs were used as controls.
SMCs were explanted from femoral arteries and characterized as α-smooth muscle-actin-positive and vonWillebrand factor-negative. The cells were grown in Dulbecco’s modified Eagle’s medium/10% fetal calf serum (FCS). Other human cells used were: coronary artery SMCs (Promocell, Heidelberg, Germany), aortic endothelial cells (UK Cascade Biologics, Mansfield, UK), dermal fibroblasts (a gift from Dr Sarah Herrick, University of Manchester), and MG63 osteoblasts, which were cultured according to the manufacturer’s instructions. Human macrophages were differentiated from monocytes (THP-1; ECACC, Poole, UK) in media (RMP1/10%FCS; Cambrex, Wokingham, UK) containing 200 μM phorbol-12-myristate-13 acetate (Sigma) for 72 h.
Protein was isolated from 80% confluent cells by incubating in lysis buffer (200 mM Tris-HCl,150 mM NaCl, 5mM NaF, 1% nonidet P-40, 10% glycerol, 1 mM EDTA, 0.1% sodium dodecyl sulphate (SDS), 2 mM Na3VO4) containing protease inhibitors (1 mM phenylmethylsulphonylfluoride, 1 μg/ml pepstatin A, 5 mM leupeptin, 2 mM vanadate, 1 μM NaPP) for 20 min at 4 °C and centrifuged (20 000 g for 10 min at 4 °C). Protein concentrations were determined using the bicinchoninic acid protein assay reagent (Pierce, Perbio Science, Cramlington, UK). Protein (20 μg) was resolved by SDS-polyacrylamide gel electrophoresis (PAGE) under reducing conditions and electrotransferred to nitrocellulose membranes (Bio-Rad, UK). The membranes were incubated with rabbit anti-VCAF antibody (1: 200) in 4% skimmed milk for 1 h at room temperature, washed with PBS containing 0.01% Tween, and incubated with swine anti-rabbit horseradish peroxidase-conjugated secondary antibody (1: 1000; DAKO). Signal was detected using ECL-Plus detection reagent (GE Healthcare, Amersham, UK).
We identified VCAF-positive cells in close association with calcified areas of human atherosclerotic plaques and in osteoblasts within human healing bone fractures. VCAF-positive cells were also detected in microvessels which stain positive for c-kit, CD34, and CD146 in FA sections. VCAF protein was detected in VSMCs, osteoblasts, macrophages, and endothelial cells in vitro.
Figure 1A(i) shows a representative FA section, demonstrating alizarin red-stained areas of calcification within the plaque (arrow) and the media (arrowhead). VCAF-positive cells were detected at the base of the calcified plaque (Figure 1Aii; square box in (i) enlarged in ii-vi) in close spatial association with OPN (Figure 1Aiv) and MGP (Figure 1Av) immunoreactivity in consecutive sections. CD68-positive staining identified the presence of macrophages in this region (Figure 1Avi). Another representative sample, exhibiting a complex plaque with discrete microfoci of calcification is shown in Figure 1Bi (arrows); the rectangular box is enlarged in Figure 1Bii-iv. VCAF protein was detected at the base of the plaque (Figure 1Bii) and is associated with CD68 macrophage immunoreactivity in a sequential section (Figure 1Biii). All IgG controls were negative (Figure 1Aiii and 1Biv).
A representative inferior mesenteric artery (IMA) is shown in Figure 1Ci with an intact endothelium (rectangular box enlarged in Figure 1Cii-iv). VCAF was detected in the endothelial layer of three specimens analysed (Figure 1Cii), with positive staining for CD31 in corresponding consecutive sections (Figure 1Ciii). All IgG controls were negative (Figure 1Civ). In addition, VCAF staining was detected in small areas of intact endothelium in the FA specimens which had evidence of localized calcified plaque on the opposite vessel wall. In contrast, there was no staining of either CD31 or VCAF on the luminal side of the plaque (data not shown).
In four patient samples analysed, VCAF staining was detected in sporadic, individual cells in the intima and adventitia; Figure 2i and ii show a representative section. The solitary VCAF-positive cells display a peculiar morphology, either stellate-shaped (Figure 2i) or elongated (Figure 2ii). We used a 3G5 antibody that recognizes a cell surface ganglioside present on pericytes, but not on fibroblasts, SMCs or endothelial cells  to localize pericytes in calcified atherosclerotic arteries. The VCAF immunoreactivity was strikingly similar to the staining detected for 3G5-positive pericytes in large artery walls [14, 15] (Figure 2).
We demonstrated strong VCAF immunoreactivity around microvessels, located both in the adventitia (Figure 3Ai, box A, high power in Figure 3Aii; and Figure 3Ai box B, high power in Figure 3Bi) and intima (Figure 3Ci, box enlarged in Cii) of a grossly occluded femoral artery with slight stippling of calcification. A number of markers were used to characterize these cells: c-kit, a haematopoietic stem cell (HSC), and an MSC-derived cell marker ; CD34, an HSC and endothelial progenitor cell (EPC) marker [22, 23]; and CD146, reported as a primitive pluripotential cell marker , a circulating and mature endothelial cell marker, and also expressed by SMCs [23, 24]. Interestingly, all markers were detected in the VCAF-positive microvessels (Figure 3A), in both large (Figure 3A, box A) and small microvessels (Figure 3A, box B). In addition, VCAF was detected in a number of microvessels within the neointima of occluded arteries (Figure 3Ci and ii, box enlarged in ii-iv) in association with c-kit (Figure 3Ciii). IgG controls were negative (Figure 3Civ).
On sections of human fracture callus, we localized VCAF to osteoblasts embedded in the matrix of newly formed bone (Figure 4i and iii, arrows). In contrast, in older, mature bone on the same section, VCAF protein was below the level of detection (Figure 4iv). In addition, VCAF staining was associated with microvessels in the periosteum, similar to the staining seen in the microvessels of FA segments (Figure 4v, box enlarged in vi).
Using western blot analysis, we investigated the expression pattern of VCAF in a range of lysates from human cells. A band was detected at the apparent molecular weight of VCAF (18 kD) in macrophages, aortic endothelial cells, VSMCs, and MG63 osteoblasts (Figure 5i and ii). A higher molecular weight band with an apparent molecular weight of 22 kD was detected in macrophages and osteoblasts, which could be indicative of either post-translational modification of VCAF or cross-reaction with another protein. This expression pattern correlates well with the results obtained from the immunolocalization in tissue sections. VCAF expression was also detected in human fibroblasts, but at much lower levels (Figure 5ii) and was also evident as an 18 and 22 kD band in these cells.
This study confirms VCAF protein localization in calcified arterial segments, in association with recognized osteogenic differentiation markers, OPN and MGP. These findings support and extend our previously published data, where VCAF mRNA was detected deep within the plaque in the same area as OPN mRNA . We also identify the presence of VCAF-positive cells in the same location as macrophages, which are known to contribute to vascular calcification by producing proatherogenic factors . Western blot analysis confirms expression of VCAF by macrophages. Previously, we reported that down-regulation of VCAF appeared to accelerate mineralization of cells in culture and consequently, we proposed an inhibitory role for VCAF in the prevention of vascular calcification [1, 2]. We now extend these studies and demonstrate that VCAF is also associated with neovascularization and is expressed by endothelial cells, highlighting the potential importance of VCAF protein in vessel wall homoeostasis.
The functional integrity of the endothelium is essential for vessel wall patency and although atherosclerosis affects all vascular beds, some areas of the vasculature are more prone to atherosclerosis than others. For example, the coronary artery is widely affected, whereas the IMA is relatively protected from formation of atherosclerotic lesions . In this study, it is noteworthy that VCAF is expressed in the intact endothelial layer of an IMA and can be detected in the small areas of intact endothelium of femoral arteries, where there is less evidence of plaque development. We also confirmed that VCAF is expressed by endothelial cells using western blot analysis. Although the IMA does not display typical atherosclerotic lesions, the endothelium may display some dysfunction as it is obtained from a patient undergoing bypass surgery. However, despite this limitation, these results could indicate a novel antiatherogenic role of VCAF, related to protection against endothelial damage/loss associated with the development of atherosclerotic plaques.
This study identifies 3G5 positivity in individual, sporadic cells in atherosclerotic lesions, providing evidence for the involvement of pericytes in atherosclerotic plaque development. It is also of interest, that VCAF is detected in similar, solitary cells in areas where macrophages are present. It is tempting to speculate that these VCAF-positive cells respond to inflammatory signals, are recruited to the damaged tissue, and undergo a multistep differentiation programme with commitment to either a myofibroblast or an osteoblast lineage during plaque development.
Asahara et al were the first to report the existence of circulating progenitor cells with the capacity to differentiate into endothelial cells and incorporate into sites of angiogenesis in vivo . Subsequent work by the same group  and others  proposed that newly developed microvessels could transport EPCs and immune cells into the deep areas of plaques, and thus contribute to pathological neovascularization. Recent evidence from animal model and human studies shows that EPCs have a pivotal role in endothelial maintenance and repair and neovascularization [7, 30].
Our data support the evidence for the involvement of progenitor cells in microvessel growth within atherosclerotic lesions. CD146 is a surface marker of MSCs , which are known to differentiate into different cell types, including endothelial cells , SMCs [32-34], and osteoblasts , and we detect strong CD146 immunoreactivity in association with microvessels in the adventitia. In addition, we detect pronounced staining of the primitive pluripotent cell marker, c-kit, around microvessels, which are also positive for CD34, a transmembrane glycoprotein constitutively expressed on both HSCs and EPCs and maintained on more mature endothelial cells . VCAF protein and all three progenitor cell markers are detected in the same microvessels within the adventitia of atherosclerotic plaques in consecutive sections, lending credence to the possibility that VCAF may act as an early marker of such a progenitor cell.
Marrow-derived and periosteal-derived progenitor cells have been shown to produce bone and cartilage in numerous in vivo and in vitro studies . It has been suggested that osteoprogenitor cells are not necessarily confined to bone marrow or circulating progenitor cells, but may reside in the microvasculature [10, 11]. Fracture repair requires new bone formation, which entails the transformation of undifferentiated osteochondral progenitor cells to mature osteoblasts and chondrocytes. Furthermore, it is essential that both angiogenesis and mineralization occur in order to repair bone fracture. Therefore the presence of angiogenic or mineral-regulating mediators is essential for the regulation of vascular and mineral deposition in skeletal tissues. We demonstrate that VCAF is present in osteoblasts of newly forming bone in human fracture callus, but not in older bone. Western blot analysis of osteoblasts in vitro confirms VCAF expression in these cells and furthermore, shows the presence of two bands. Western blot analysis also confirms the expression of VCAF in cultured SMCs and to a lesser extent in fibroblasts, the latter also showing the presence of 18 kD and 22 kD bands, as was found in macrophages and osteoblasts. The upper band might represent a cross-reacting protein in these cells. However, since sequence analysis of the VCAF protein shows putative phosphorylation and glycosylation sites , another possibility is that the upper band detected in MG63 osteoblasts, macrophages, and fibroblasts represents post-translational modification of the VCAF protein. Many proteins are regulated by a post-translational modification, for example, OPN acts as an inhibitor or inducer of calcification, depending on its phosphorylation status . VCAF regulatory domains will be confirmed in future structural analysis studies.
In fracture callus samples, VCAF staining was strong in microvessels in the periosteum, consistent with the VCAF staining detected in microvessels in the vessel wall. We support the idea of a hierarchy of progenitor cells, both in the systemic circulation and the vasculature, which can give rise to a number of cell lineages under different environmental conditions and propose that VCAF may be a marker of an early uncommitted progenitor cell. Indeed, it might seem paradoxical that VCAF-positive cells differentiate into osteoblasts and deposit a mineralized matrix in the atherogenic environment and yet also contribute to new vessel formation in plaque lesions. However, it is evident that angiogenesis and calcification co-exist , and pericytes are known to be independently involved in both processes . Therefore, understanding the molecular identity of progenitor cells and the signals involved in directing cell fate are key elements, which offer powerful opportunities for the future development of cardiovascular cell and gene therapy.
In conclusion, we have (i) identified a novel protein (VCAF) which is expressed in human calcified lesions in association with OPN and MGP expression, and CD68-positive and 3G5-positive cells; (ii) confirmed the localization of VCAF-positive staining in intact endothelium and in individual sporadic cells within plaque lesions and within human microvessels; (iii) demonstrated the presence of VCAF protein in osteoblasts in newly formed bone and in microvessels in the periosteum in fracture callus, but not in osteoblasts in older mature bone within the same section; (iv) established an association of VCAF protein with early progenitor cell markers in neovessels within advanced atherosclerotic lesions; and (v) demonstrated VCAF expression in human VSMCs, endothelial cells, macrophages, and osteoblasts using western blot analysis.
We show that VCAF is expressed in early progenitor cells and may, under different micro-environmental conditions, play a role in a number of differentiation pathways. Our study equally raises several pertinent questions, such as: (i) What is the identity of mediators and cellular mechanisms that promote migration of VCAF-positive cells to sites of damage? and (ii) What are the triggering factors involved in VCAF expression which might influence the proliferation and differentiation of VCAF-positive cells? These topics will be examined in future studies.
Funding from the British Heart Foundation (project grant number PG/05/065 awarded to MYA and AEC) is gratefully acknowledged. We thank Mr J Vincent Smyth and M. Daniel Keenan from the Division of Medicine and Surgery, for providing us with the patient tissue.
No conflicts of interest were declared.