| COMMENTARIES |
Will Genomics Provide New Insights into the Diversity of Plaque Macrophages?
Andrew C. Newby, Graciela B. Sala-Newby, Anita C. Thomas, and Jason L. Johnson,
Bristol Heart Institute,
University of Bristol, Bristol BS2 8HW, UK
Please address correspondence to:
Prof. Andrew C Newby
British Heart Foundation Chair of Vascular Cell Biology
Bristol Heart Institute
Bristol Royal Infirmary
Bristol BS2 8HW, UK
Tel: +441179283583
Fax: +441179283581
E-mail: a.newby@bris.ac.uk
Dual Role of Macrophages in Fibrous Cap Formation and Atherosclerotic Plaque Rupture
Macrophages are found in atherosclerotic plaques at all stages [1]. Once low density lipoprotein (LDL) enters and becomes trapped in the vascular intima, monocytes are recruited from the circulation, differentiate to macrophages, and accumulate cholesteryl esters to become foam cells [2,3]. Early fatty streaks consist of foamy macrophages and only a few foamy smooth muscle cells (SMCs) and therefore appear highly inflammatory. Nevertheless fatty streaks do not cause symptoms, are covered by an intact endothelium, and appear completely reversible [4]. With continuing hypercholesterolemia small and then larger extracellular lipid pools develop probably owing to foam cell death, although deposition of red cell membranes contributes in late lesions that have been invaded by vasa vasora [5]. Fibroatheromas with a significant fibrous cap develop only later thanks to recruitment and proliferation of SMCs [6]. The ability of foamy and nonfoamy macrophages to produce growth factors, inflammatory cytokines, and extracellular proteases most likely aids migration and proliferation of resident SMCs and stimulates their ability to lay down extracellular matrix [7,8]. Circulating precursors or other cells (including macrophages) may perhaps also differentiate into SMC-like cells [6]. According to this view, macrophages play key roles in orchestrating fibrous cap formation, as they do in other kinds of fibrosis related to inflammation. Nevertheless, foam cell macrophages are abundant at the shoulder regions of thin cap fibro-atheromas (TCFAs), the lesions that most frequently give rise to plaque ruptures [5,9]. Presence of macrophages in these areas of plaque correlates with the breakdown of interstitial collagen and fewer SMCs. There are also high levels of extracellular proteases including several metalloproteinases (MMPs) and cathepsins [10]. These macrophages are therefore thought to be responsible for degrading the extracellular matrix (ECM) and, by provoking apoptosis of SMCs or dying themselves [11], contribute to cellular rarefaction of the plaque cap, which is commonly observed in advanced lesions. Hence macrophages have been implicated as the main culprits in fibrous cap disruption. However, if plaque rupture and coronary thrombosis are not fatal the thrombus becomes organized, a process of fibrosis in which appropriately stimulated macrophages also play an important part [12]. In summary, macrophages seem to promote recruitment of SMC-like cells and the laying-down of ECM during fibrous cap formation and thrombus reorganization, but destroy ECM and promote death of connective tissue cells during plaque rupture. In our view one of the key goals of future research should be to investigate whether and how macrophages can take on these apparently contradictory behaviors.
Genomics of Foam-Cell and Non-Foamy Macrophages
Defining the transcriptome (and ultimately the proteome) of macrophages at different stages of atherosclerosis appears to be a powerful approach to address the problem set out above. However it is beset with technical difficulties. Genomic studies comparing atherosclerotic plaque and normal tissues detect many marker genes for activated macrophages [13] but this may reflect the different proportion of cells present rather than changes within any particular cell type [14]. Laser micro-dissection allows the isolation of plaque macrophages for genomic studies [13,15,16] but dissected samples are too small for metabolic studies to investigate the functional significance of any changes. In a recent study [17], we offered an alternative approach. Polyurethane sponges implanted under the skin of rabbits provoke a vigorous foreign body reaction with abundant non-foamy macrophages. If rabbits are fed a cholesterol-rich diet, foam-cell macrophages are obtained. Large quantities of highly pure foamy and non-foamy macrophages are easily isolated from the sponges and can then be used for genomic, proteomic (unpublished data) and functional assays. We identified the scavenger receptor, Lox-1, and the metalloproteinase, MMP-12, among the genes up-regulated in foam-cell macrophages, whereas arginase-I was significant among the down-regulated genes. The genomics were confirmed by mRNA and protein studies on bulk preparations of foamy and non-foamy cells. Moreover the results were validated in rabbit and human plaques using immunohistochemistry. Finally we were able to investigate the metabolic consequence of arginase-I down-regulation. Since arginase-I and nitric oxide synthase compete for the same substrate, we predicted increased ability of foam cells to generate NO and confirmed this by measurements of nitrite in the conditioned medium. Arginase-I down-regulation is also implicated in plaque instability [18], perhaps through promoting apoptosis. Similarly MMP-12 up-regulation has been implicated in ECM degradation and plaque instability in rabbits and mice [19-21]. Hence the changes we have observed appear to reflect the destructive function of macrophage occurring in later plaques. Consistent with this we found MMP-12 up-regulation and arginase-I down-regulation only in macrophage foam-cells within the deep layers of advanced rabbit aortic and human carotid plaques. Moreover the genomic changes could not be replicated simply by loading macrophages with oxidized LDL.
Limitations and Prospect
Our current studies represent only a proof of principle, and perhaps a staging post towards more powerful methodologies. For example, rabbits provide an abundant source of cells but genomics are limited by the lack of commercial gene chips. Unfortunately the method is not directly translatable to mice (unpublished observations), although some effective variant technique may be feasible. Likewise the consequences of Lox-1 and MMP-12 up-regulation and arginase-I down-regulation on macrophage function remain to be fully elucidated. The added knowledge to be gained from proteomics remains to be garnered. However, almost one hundred years on from Anitschkow’s initial observations on cholesterol-fed rabbits [22], we feel confident that there are still useful lessons to be learned from applying modern post-genomic technologies to the model and our variant of it.
Acknowledgements
The authors’ work is supported by the British Heart Foundation and the European Vascular Genomics Network. ACT is a CJ Martin Fellow of the Australian Medical Research Council.
References