Background
Chronic kidney disease (CKD) has reached epidemic proportions, affecting 10–13% of the global population [
1‐
3]. Cardiovascular disease (CVD) is responsible for most of the deaths in the CKD population. While CVD mortality is due to different causes, atherosclerotic CAD is consistently higher in CKD than in the general population [
4,
5] and potentiation of CVD risk is seen even for modest impairment in kidney function, which constitutes the largest fraction of the CKD population [
6‐
10]. The mechanisms by which CKD imparts the heightened CVD risk remain unclear, constraining development of risk-reduction interventions in this population. Lipid-lowering agents, primarily statins, are the therapeutic cornerstone to decrease CVD risk. However, even in the general population this intervention provides incomplete risk reduction, a shortcoming that becomes more evident among subjects with CKD.
The hallmark of atherosclerotic plaque is macrophage accumulation of cholesterol and transformation into foam cells, consequence of the unrelenting uptake of atherogenic lipoproteins, and of the failure to mobilize excess cholesterol to extracellular cholesterol acceptors, primarily HDL [
11‐
13]. Epidemiologic studies have established that reduced HDL cholesterol (HDL-C) is associated with increased CVD; however, increasing HDL levels by drug therapies have not resulted in clinical benefits [
14‐
16]. These observations have given rise to the concept of dysfunctional HDL to include impaired efflux capacity and pro-inflammatory properties. Importantly, recent human studies indicate that reduced HDL cholesterol efflux is an independent CVD risk factor [
17‐
20], as capacity of HDL to stimulate cholesterol efflux from cultured macrophages predicted subclinical atherosclerosis and coronary artery disease in non-CKD populations. Consequently, conditions that impair cholesterol efflux or promote cellular uptake can accelerate atherosclerosis and CVD even in the absence of hypercholesterolemia, and may also foster resistance to treatment with lipid-lowering agents.
We, and others, have shown that patients with end stage renal disease (ESRD) requiring dialysis have dysfunctional HDL with reduced efflux and increased inflammation [
21‐
30]. Less is known about the extent to which earlier stages of CKD affect HDL functionality. Although moderate CKD decreases HDL’s anti-inflammatory response, the impact on efflux capacity is less understood [
22,
27,
29,
31‐
34]. It is also unclear whether, and which of the HDL functionalities can be modified. Increasing kidney function by renal transplantation can improve HDL anti-inflammatory/anti-oxidant function but not cholesterol handling [
29]. There are no studies evaluating the utility of targeting macrophage cholesterol handling and inflammation that might constitute additional therapeutic targets beyond kidney transplantation or standard lipid lowering. Interestingly, we previously showed that uninephrectomy-induced amplification of murine atherosclerosis is linked to impaired cholesterol efflux reflecting repression of macrophage ATP-binding cassette transporter A1 (ABCA1) [
35]. Liver X receptors (LXR) directly activate ABCA1, and experimental and clinical studies support potentially beneficial effects of LXR agonism to decrease foam cell formation and atherosclerosis in non-CKD settings [
36,
37]. Therefore, we sought to evaluate both lipid handling and inflammatory response to lipoproteins of patients with moderate to severe (stage 3 and 4) CKD and assess the potential benefits of activating cellular cholesterol transporters with LXR agonism.
Discussion
Our data indicate that HDL isolated from patients with moderate to severe CKD has impaired capacity to elicit cholesterol efflux from lipid-loaded macrophages and potentiates inflammation, compared with HDL of individuals with normal kidney function. Both cholesterol handling and regulation of vascular inflammation represent potentially atheroprotective functions of HDL that may be influenced by pharmacologic intervention. We show that upregulation of ABCA1 transporters with LXR agonism significantly increases cellular cholesterol efflux both to normal HDL and to the dysfunctional HDL from CKD patients. However, and unexpectedly, LXR activation also amplifies the already heightened cellular inflammatory cytokine response to HDLCKD by mechanisms linked to activation of TLRs and the ERK1/2 pathway.
These results have important implications. First, our data show that even moderate CKD modifies HDL function and promotes cellular cholesterol accumulation. Interestingly, the lipid accumulation does not reflect greater uptake of LDL. Although macrophage uptake of unmodified LDL is small, CKD is well recognized to cause LDL oxidation which would augment cellular cholesterol loading via scavenger receptor uptake [
45]. However, we did not see increased uptake and instead observed significantly impaired cholesterol efflux, results that echo previous findings by us and others in patients with ESRD on dialysis [
21,
23,
24,
27,
29]. The results also complement findings by Shroff et al. who reported a graded change in cholesterol efflux capacity in children with CKD stage 2–5 [
29]. By contrast, a recent study reported no difference in cholesterol efflux capacity of HDL from normal subjects and individuals with various degrees of kidney impairment (pre-dialysis, maintenance hemodialysis and transplant recipients). However, the HDL concentration used in the assay was only 20% of that commonly applied in efflux assays and may have been insufficient to effectively elicit cholesterol movement from lipid-laden cells [
46]. Taken together, our findings support the hypothesis that even moderate kidney impairment causes HDL dysfunction, including reduced capacity for cholesterol efflux. These observations have particular importance since moderate to severe degree of CKD comprise > 50% of all patients with kidney disease and because efflux capacity has been shown to be a marker for atherosclerosis and risk of cardiovascular events in individuals without kidney damage, although not patients requiring dialysis [
17,
18,
47]. Implications of these findings are further heightened by the fact that standard lipid-lowering therapies reduce the CVD risk by only 20%. Thus, there is the potential of sizable additional residual CV risk reduction in the high risk CKD population using therapies targeted at this derangement [
48].
Our data suggest that age might have some impact on HDL’s efflux capacity, observations that complement the previously described effects of aging on HDL function [
49]. Specifically, adjustment for age eliminated the statistically significant difference in efflux capacity between the CKD and control groups. Interestingly, there were no differences in efflux capacity between CKD and controls in individuals who were less than 50 years old. It is interesting that cellular activation of LXR also increased cholesterol efflux to both HDL
Cont and HDL
CKD, the statistically significant difference persisting after adjustment for age. These data suggest the possibility that, even in the face of dysfunctional HDL, improving cellular transporter activity can increase cholesterol export, which would predict a reduction in foam cells in the arterial wall. Our finding indicate that this may be beneficial to CKD patients regardless of age. Such therapeutic possibility is supported by experimental observations that synthetic LXR agonists can decrease atherosclerosis in animal models through mechanisms that include macrophage efflux and increased reverse cholesterol transport [
37,
50].
In addition to modulating cholesterol transport, LXR activators have been shown to have anti-inflammatory effects that may be critically important in their antiatherogenic effects [
36]. Therefore, we also assessed the effects of LXR agonism on the cellular inflammatory response using the same HDL samples applied to the efflux assay. Our results show macrophage expression of IL-1β, IL-6, and TNF-α is greater in response to HDL
CKD than HDL
Cont. Furthermore, LXR-activated cells exposed to HDL
CKD showed unexpected potentiation in the inflammatory response (Fig.
3). These results suggest that while HDL
CKD is abnormal in terms of both efflux and inflammation, the abnormalities are not synchronized and therapeutic modalities may have variable effects on distinct HDL actions.
Interestingly, a recent study reported that treatment of LDLR
−/− mice with the LXR activator T0901317, substantially reduced the extent of atherosclerotic lesions, independent of the macrophage ABCA1/G1 cholesterol efflux pathways [
36]. Instead, the beneficial effects of LXR activation reflected the anti-inflammatory actions of LXR thought to be mediated through transrepression mechanisms involving the small ubiquitin-like modifier (SUMO)-ylation of LXR, that targets promoters of NF-κB target genes and results in repression of inflammatory genes downstream of NF-κB [
51]. Our study shows that compared to macrophages exposed to normal HDL, cells exposed to HDL
CKD have dramatically increased activity of NF-κB (Fig.
4), central role of NF-κB in regulation of inflammatory processes when renal function is reduced [
52]. Notably, however, activation of LXR did not affect expression of NF-κB, suggesting that increased cytokine response in this context does not involve the NF-κB-dependent transrepression pathway. Instead, LXR agonism was associated with direct effects on TLR2 and TLR4. While TLR2 and TLR4 can activate the same signaling pathway, such as JNK, the ERK1/2 pathway is primarily activated by TLR2 [
53]. Our data show that both HDL
Cont and HDL
CKD activate TLR4, whereas only HDL
CKD activates both TLR4 and TLR2. Moreover, TLR2 activation was associated with phosphorylation of ERK1/2 in response to HDL
CKD. Thus, whereas both TLR4 and TLR2 can stimulate a cytokine response to HDL
Cont and HDL
CKD, the exaggerated response to HDL
CKD is more closely linked to TLR2 than TLR4. Our findings that TLRs are involved in HDL
CKD-augmented inflammatory response complements previous reports that TLR2 has a critical role in impaired endothelial protection and anti-inflammatory capacity of HDL isolated from adults and children with stage 2–4 CKD [
33]. It is interesting that in human macrophages pretreatment with LXR agonists for 24 h significantly reduced the inflammatory response induced by LPS whereas longer exposures paradoxically enhanced the inflammatory response [
50]. Further, the increased cytokine production involved TLR4 [
54,
55], as well as cytokine production through pathways driven by TLR2 [
55]. The LXR augmentation of TLR-induced cytokine secretion has been proposed as the molecular pathway responsible for synovial inflammation in rheumatoid arthritis [
55]. Similarly, it is possible that in the CKD setting, LXR agonism produces undesirable effects by activating TLRs and further potentiating cytokine production.