Sources of CRP in Atherosclerotic Lesions

To the Editor-in-Chief:

We read with great interest the recent article by Sun and colleagues¹ in which they show, in two rabbit models, that C-reactive protein (CRP) is present in the atherosclerotic lesion and indeed correlates with lesion area. Furthermore, they show in these two models a significant correlation between total cholesterol and plasma CRP, a finding that has not been observed in human patients. They provide no explanation for this correlation. Is it possible that the hypercholesterolemia in these models induces inflammation in the liver resulting in the elevated levels of CRP, which then enters the atheroma?

More importantly, however, in a sample size of eight, they demonstrate the presence of CRP in the atherosclerotic lesion, where it did not co-localize with macrophages or smooth muscle cells, and CRP was also found in the surface area of the lesion (see Sun et al,¹ Figures 6 and 7). Using real-time reverse transcriptase-polymerase chain reaction (RT-PCR), they could not demonstrate CRP mRNA in the atherosclerotic lesion and concluded that the CRP was derived from the liver. This finding challenges a great body of data suggesting that CRP is synthesized in the atheroma.

The first observation in this regard was made by Yasojima and colleagues² who showed 10-fold more mRNA for CRP in atherosclerotic lesions versus normal vessels (n = 10 arterial samples). This finding has now been confirmed by numerous groups. Kobayashi and colleagues³ used anti-sense riboprobe to show, in 39 directed coronary atherectomy samples, that CRP is present in the coronary atheroma. Furthermore, Jabs and colleagues⁴ have shown by real-time RT-PCR that CRP is expressed in coronary artery venous bypass grafts (n = 11 of 15), and Vainas and colleagues⁵ have documented CRP mRNA in abdominal aortic aneurysmal tissue (n = 4 of 16). In addition to real-time RT-PCR, some of these investigators have used other sensitive techniques to document CRP. Using Western blotting, immunohistochemistry, and real-time RT-PCR, Sattler and colleagues⁶ have recently documented CRP protein and mRNA in plaques obtained from patients undergoing carotid endarterectomy (n = 41). Furthermore, the CRP staining was present in plaque shoulders, microvessels, or borders, mainly in foam cells and endothelial cells.

The major cells of the atherosclerotic lesion include endothelial cells, monocytes/macrophages, T lymphocytes, and smooth muscle cells. Calabro and colleagues⁷ have shown by RT-PCR and enzyme-linked immunosorbent assay that vascular smooth muscle cells in culture express and secrete CRP. In a recent article in The American Journal of Pathology, we made the novel observation that both coronary and aortic endothelial cells express CRP mRNA and protein using a comprehensive approach of examining mRNA by RT-PCR and in situ hybridization and protein by Western blotting. Most importantly, we showed a 100-fold increase in secretion of CRP after incubation of aortic endothelial cells with macrophage-conditioned media.⁸

It is interesting that in the article by Sun and colleagues,¹ no mention is made about the possibility that CRP is derived from aortic endothelial cells, despite surface expression of CRP in the human atherosclerotic tissue in their study. Also, using an activated monocytic cell line, U937 cells, they fail to detect mRNA for human CRP. We have preliminary data that human monocyte-derived macrophages express CRP mRNA (unpublished data), concurring with similar findings in alveolar macrophages.⁹ The authors also fail to mention other sources for CRP. In fact, CRP has now been reported in neurons of Alzheimer’s brain, in renal tubular epithelial cells, and also in alveolar macrophages.^8,9 Based on our experimental experience, as well as that of others, we argue for potential paracrine and autocrine effects in microdomains in the intima resulting in potentially very high CRP levels. Although it is the general consensus that the majority of the CRP in the atheroma derives from the liver, it is not unreasonable that both vascular cells and adipose tissue contribute to some degree. Future studies will help determine their relative contributions. CRP produced in endothelial, vascular smooth muscle cells, and macrophages via autocrine and paracrine loops could exert biological effects on these cells contributing to atherothrombosis and plaque instability.¹⁰

In conclusion, the majority of published studies document CRP expression in atheroma and further support the notion that CRP is elaborated by cells of the vessel wall including endothelial cells, smooth muscle cells, and human monocytes/macrophages. We believe it is premature for Sun and colleagues¹ to arrive at the conclusion that CRP is derived exclusively from the liver. In fact, Ouchi and colleagues¹¹ have shown that the presence of CRP mRNA in adipose tissue has an inverse relationship with adiponectin. In addition, CRP secreted from adipose tissue has recently been documented.¹² Thus, we believe this is an area that clearly requires further scientific inquiry to establish the major sources of CRP in the atheroma. We hypothesize that CRP in atheroma may derive from the liver, the atherosclerosic lesion, and adipose tissue. In this regard it is important to mention that CRP levels have been reported to be significantly higher in the coronary sinus versus the peripheral circulation,¹³ arguing that at least some of the CRP is elaborated in the atheroma itself.