Several studies have suggested CRP as biomarker for cardiovascular and cerebrovascular diseases [13, 14]; however, this protein also seems to be a mediator of atherosclerosis [26, 27]. Angiogenesis is a recognised mechanism involved in the development of complicated atherosclerotic plaques and previous studies have provided controversial data regarding the possible angiogenic or anti-angiogenic effect of CRP. In this work, we have examined the hypothesis that CRP might play a role in modulating angiogenesis, and as such, its presence in vascular regions of developing arterial plaques may be implicated in their progression to unstable, haemorrhagic lesions prone to rupture. Our data shows that CRP, at concentrations commonly found in the circulation of patients with active carotid disease, is highly pro-angiogenic both in vitro, and in vivo. Furthermore, CRP activates cell signalling and increases expression of key genes associated with angiogenesis.
In order to show that the effects of CRP were due to CRP itself and not other components, we have attempted to address the controversy surrounding the influence of ¨contaminants¨ present in commercial preparations of CRP. Notably, Pepys et al. demonstrated that pure natural human CRP free from endotoxins and azide, was not pro-inflammatory to macrophages in vitro . More specifically, inhibition of HUVEC cell proliferation, eNOS gene expression and increased apoptosis measured by activation of caspase-3/9, have been attributed to the presence of azide in CRP preparations. Similarly, contaminating LPS was shown to increase IL-8, ICAM-1 and MCP-1 gene expression [39, 40]. It is worthy of note, however, that the effects of azide were demonstrated at concentrations (0.0025%) equal to those produced by administration of 50 μg/ml CRP. In our study, concentrations of CRP and therefore azide used in the majority of experiments were much lower and in fact, had no significant effects on cell growth and gene expression as demonstrated in the TaqMan microarrays. Even so, this highlights the importance of using satisfactory controls when conducting this type of assays. In our studies, all of the controls, used for comparison with CRP tests, contained equivalent amounts of azide so that direct comparisons were able to be made. Finally, in our study, we show that the treatment of CRP with detoxi-gel columns did not negate the pro-angiogenic effects of CRP. Interestingly, assays performed using purified LPS at concentrations that were similar to or exceeded those reported in high purity human recombinant CRP from our commercial supplier [43, 44] demonstrated no significant effects on BAEC or HCAEC angiogenesis or IL-8 gene activation. This is in agreement with the recent findings of Dasu et al. who showed that purified CRP was able to activate IL-8, IL-6, IL-1β, PAI-1 and eNOS in Toll-like receptor 4 knockout HAEC, indicating that the effects were not due to LPS contamination .
In this study, we showed that CRP induces angiogenesis in vitro in two separate sources of primary cultured vascular EC as well as in vivo as judged by an increase in capillary formation in the CAM assay, aortic ring assay and 3D-Matrigel. The angiogenic potency of CRPdt was quite strong, being 50–75% of that demonstrated by our positive control FGF-2 (25 ng/ml; comparisons not included). One other study showed previously that CRP induced proliferation of rabbit thoracic EC with a concomitant increase in expression of p-ERK1/2, similar to our own findings . In their study, they controlled for LPS by measuring levels with the Limulus assay which were recorded as < 0.125 EU/ml. The authors, however, found that increased proliferation occurred only at significantly high CRP concentrations of (> 20 μg/ml), and furthermore, they did not consider the effects of azide. In agreement with our findings, Bello et al (2008), showed that CRP increases VEGF-A expression via PI3-kinase and ERK1/2 pathway and thus could play a role in the angiogenesis process . Orozlan et al. found no effects of CRP on HUVEC proliferation although they did not supply data indicating the concentration of CRP used . In our previous studies where we have characterised various types of vascular cells, we have found HUVEC to be the least responsive to pro-angiogenic stimuli including VEGF, FGF-2 and oligosaccharides of hyaluronan; therefore, these results are perhaps not surprising . The fact that CRP did not stimulate HCAEC proliferation suggests that there may be variability in cellular responses dependent on cell type, and EC have been shown in many studies to be heterogeneous in this regard. Indeed, in contrast to our results, one recent study showed induction of apoptosis in human umbilical EC (HUVEC) incubated with CRP, analyzed by TUNEL and caspase-3 activity assay and inhibition . However, the above study used 10 μg/ml of CRP in their assay and we observed a pro-angiogenic effect of CRP at 1 and 5 μg/ml. Moreover, the apoptotic effect of CRP was demonstrated in HUVEC and cellular response may be dependent on EC origin. In addition, in the same study the authors showed that CRP treatment of monocluclear cells induced production of MMP-9 which is involved in extraclellular matrix degradation, cell migration and release of angiogenic factors necessary to elicit angiogesis [48, 49] Verma et al. showed that purified CRP attenuated NO release in human saphenous vein EC, increased apoptosis and inhibited capillary-like tube formation in matrigel at 5–25 μg/ml ; however, they did not control for the effects of azide. The same authors also demonstrated that CRP inhibited EC progenitor differentiation, survival and function through a process involving a reduction in NO expression appeared to be carried out without the use of suitable azide and endotoxin controls .
Our data showing a potent angiogenic effect of CRP/CRPdt is strongly backed up by the results of our real-time TaqMan PCR microarrays. Using our specifically designed targeted microfluidity cards, we showed up-regulation of key genes involved in promotion of vascularisation. We found 6 genes up-regulated by CRPdt on HCAEC between 12–72 h. VEGF receptor-2 (KDR), the main receptor mediating both signal transduction and the biologic responses, including angiogenesis, triggered by VEGF in endothelial cells , was significantly induced by CRPdt. Up-regulation of this receptor together with growth factors such as PDGF- B involved in cell proliferation and angiogenesis , could be one mechanism through which CRP initiates its angiogenic effects. Notch1 was increased after 12 h of CRP treatment. Notch1 is required for formation of correct sprouting and branching patterns during VEGF-stimulated angiogenesis in vivo . Recently, cyclic strain was shown to up-regulate Notch1 in human vascular EC and this process was responsible for significantly increased tube-formation in matrigel suggesting a role in development of atherosclerosis . Notch3 is other member of the Notch family that seems to be critical for vascular cell survival and is required for the arterial identity and maturation of vascular cells [55, 56]. One particularly novel finding was the up-regulation of CYR61. CYR61 is an extracellular matrix-associated protein expressed within developing vasculature, which promotes angiogenesis both in vitro and in vivo . CYR61 binds directly to the integrin αvβ3 present on activated EC and mediates chemotaxis and tube formation. Finally, ID1, originally identify as a dominant-negative antagonist of the basis helix-loop-helix (HLH) transcription factors has been recently involved in VEGF-induced angiogenesis in human endothelial cells . Future work should be addressed to determine the expression, localization and relevance of these proteins in angiogenic regions of developing complicated atherosclerotic lesions.
This discussion would not be complete without a brief mention of the recently characterised modified CRP (mCRP). Evidence has emerged that native pentametric CRP can change its structural conformation following separation into monomers. Following re-arrangement, formation of the mCRP sub-unit has increased binding affinity for plasma membranes, and has been shown to be preferentially expressed in tissues . Recent evidence suggest that mCRP may be a significantly weaker stimulator of pro-inflammatory molecules in vascular EC (e.g. IL-8, PAI-1 and prostaglandin F1-α), and hence atherogenic effects . However, it is important to remember that native CRP can be found at high concentration in many hospital patients and most of them do not develop acute cardiovascular events. This suggests that cardiovascular effects of CRP if indeed important maybe due to its modified form.