Diabetes is an insidious disease that significantly contributes to the development of atherosclerotic cardiovascular disease (CVD). One of the major factors contributing to atherosclerosis is the accumulation of lipid-laden macrophages in the vascular wall (1). Removal of cholesterol from macrophages is essential for the regression of atherosclerotic lesions. This can be achieved by cholesterol-lowering drugs (i.e., statins) or by enhancing the reverse cholesterol transport pathway (i.e., raising HDL). Unfortunately, in the setting of diabetes, cholesterol-lowering drugs such as statins can be less effective (2), and preclinical models have revealed a major defect in lesion regression even when cholesterol levels are normalized (3–5). This suggests that a major defect in cholesterol metabolism occurs in the setting of diabetes.
The Hedrick and Oram laboratories originally discovered that diabetes causes a downregulation of the ATP binding cassette transporters ABCG1 (6) and ABCA1 (7) in macrophages, promoting foam cell formation. ABCA1 and ABCG1 provide the dominant pathway of cholesterol removal from a number of cells in the body. Yvan-Charvet et al. (8) made the seminal discovery that codeletion of ABCA1/G1 results in a dramatic increase in atherosclerosis. Extensive research on the antiatherogenic role of these transporters has revealed their importance in removing cholesterol from plaque macrophages (9). However, recent discoveries have shown that ABCA1 and ABCG1 regulate the proliferation of hematopoietic stem and progenitor cells to control the abundance of blood monocytes (10). Along with promoting cholesterol efflux from macrophages, suppressing monocyte levels can directly inhibit the progression of atherosclerotic lesions, promote lesion regression, and provide beneficial outcomes after a myocardial infarction (11). Nagareddy et al. (4) found that diabetes induces the production of monocytes via neutrophil-derived S100A8/A9 binding to the receptor for advanced glycation end products (RAGE) on hematopoietic progenitors, which ultimately contributes to impaired atherosclerotic lesion regression. Subsequently, a decrease in the expression of Abca1/g1 in the hematopoietic progenitors of diabetic mice was also noted, and inhibition of microRNA (miR)-33 restored the expression of Abca1/g1 (3). This was accompanied by a reduction in blood monocytes that promoted lesion regression (3). However, the mechanism(s) contributing to the downregulation of these key cholesterol efflux genes in the setting of diabetes was unknown.
In this issue, Daffu et al. (12) report on an important breakthrough using a comprehensive set of studies to link the RAGE signaling axis with the suppression of macrophage ABCA1/G1 in the settings of diabetes (Fig. 1). This resulted in a significant impairment of macrophages to promote cholesterol efflux to HDL and apolipoprotein A1 (apoA1). A plethora of studies have documented the atherogenic role of RAGE, particularly in the setting of diabetes (13). More recently, the importance of the RAGE pathway has been confirmed in human atherosclerotic disease and has been linked to plaque vulnerability (14). RAGE is classed as a pattern recognition receptor having multiple endogenous ligands (13). It is expressed in a variety of cells, and deletion in either hematopoietic or nonhematopoietic cells affords protection against diabetes-accelerated atherosclerosis (15).
In vivo macrophage reverse cholesterol transport (RCT) studies reveal diabetes significantly impairs this process, which was restored when RAGE was deleted (12). Supporting the downregulation of ABCA1/G1, lower HDL levels are observed in diabetic mice and the deletion of RAGE returns HDL levels to normal. The finding of reduced HDL levels in diabetes and restoration with RAGE deletion is not a universal observation as studies by Soro-Paavonen et al. (16) did not observe such an effect in diabetic Apoe−/− mice. Further, HDL can undergo modifications in diabetes resulting in impaired function (17), suggesting a major failure of this important pathway at both the plasma and cellular levels (Fig. 1). While the experiments by Daffu et al. (12) reveal that RAGE plays a role in suppressing ABCA1/G1 expression under normoglycemic conditions, nondiabetic Rage−/− mice display normal cholesterol levels, suggesting that this is perhaps specific to macrophages, which do not significantly contribute to the circulating pool of HDL (18).
Delving deeper into the mechanism using luciferase reporter assays, carboxymethyllysine-advanced glycation end product (CML-AGE) was found to suppress the transcription of ABCG1 through its peroxisome proliferator–activated receptor (PPAR)γ–responsive promoter element, at least in part via the mitogen-activated protein kinase (MAPK) pathway (Fig. 1). CML-AGE is only one of many RAGE ligands shown to be upregulated in the diabetes setting such as damage-associated molecular pattern molecules (4,5,16), which are not investigated by Daffu et al. However, it is possible that this is a universal mechanism downstream of RAGE as S100B has been shown to downregulate Abca1/g1 (19). Interestingly, RAGE signaling does not prevent ABCA1/G1 expression by suppressing the induction of these genes via liver X receptors (LXRs). This could be significant as Mauldin et al. (6) have previously shown that LXR agonists can induce the expression of ABCA1/G1 in the setting of diabetes to reduce foam cell formation, and LXRs can also oppose RAGE-S100B–induced suppression of Abca1/g1 (19), suggesting a utility of macrophage-selective LXR agonists to restore RCT in diabetes (Fig. 1). The relevance of these findings to atherosclerosis was explored using diabetic Ldlr−/− mice compared with Ldlr−/−Rage−/− mice. Laser-capture microdissection of CD68+ macrophages from atherosclerotic plaques of Ldlr−/−Rage−/− mice displayed higher levels of Abca1, Abcg1, and Pparγ mRNA transcripts versus RAGE-expressing Ldlr−/− mice independent of glycemic and lipid control (12). However, a recent comparison revealed only Abca1 transcripts were suppressed in plaque macrophages from diabetic mice (3). Moreover, the total body knockouts and the reduction of atherosclerosis observed here are likely not only due to changes in macrophage ABCA1/G1 expression, as there are certainly important nonmacrophage atherogenic roles for RAGE (15). It would have also been of interest to conduct these studies with inducers of Abca1/g1, such as LXR or PPARγ agonists.
Another point of interest would be to determine the type of macrophages within the atherosclerotic lesion in the current study, as previous reports have shown a major skewing toward the M1-inflammatory phenotype in the setting of diabetes (3–5). Treating diabetic mice with the ABC-inducing anti–miR-33 (20) does suggest that restoring the expression of these genes equates to less M1- and more M2-like macrophages (Fig. 1) (3). Interestingly, restoration of ABCA1/G1 has no effect on the ability of macrophages to be retained or egress from the lesion and points to a major role in the influx of monocytes to the atherosclerotic plaque that should be considered when developing targets in the setting of diabetes (3). Nonetheless, Daffu et al. (12) rightly conclude that the RAGE signaling axis may fill an important therapeutic gap in the treatment of diabetic macrovascular complications, particularly by this previously unexplored mechanism linking cholesterol metabolism.
See accompanying article, p. 4046.
Funding. K.A.J.-D. and A.J.M. are supported by a fellowship from the Australian National Health and Medical Research Council. A.J.M. is also supported by a Future Fellowship (100440) from the National Heart Foundation of Australia.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.