D-4F reduces EO6 immunoreactivity, SREBP-1c mRNA levels, and renal inflammation in LDL receptor-null mice fed a Western diet.

LDL receptor-null (LDLR(-/-)) mice on a Western diet (WD) develop endothelial dysfunction and atherosclerosis, which are improved by the apolipoprotein A-I (apoA-I) mimetic peptide D-4F. Focusing on the kidney, LDLR(-/-)mice were fed a WD with D-4F or the inactive control peptide scrambled D-4F (ScD-4F) added to their drinking water. The control mice (ScD-4F) developed glomerular changes, increased immunostaining for MCP-1/CCL2 chemokine, increased macrophage CD68 and F4/80 antigens, and increased oxidized phospholipids recognized by the EO6 monoclonal antibody in both glomerular and tublo-interstitial areas. All of these parameters were significantly reduced by D-4F treatment, approaching levels found in wild-type C57BL/6J or LDLR(-/-) mice fed a chow diet. Sterol-regulatory element binding protein-1c (SREBP-1c) mRNA levels and triglyceride levels were elevated in the kidneys of the control mice (ScD-4F) fed the WD compared with C57BL/6J and LDLR(-/-) mice on chow (P < 0.001 and P < 0.001, respectively) and compared with D-4F-treated mice on the WD (P < 0.01). There was no significant difference in plasma lipids, lipoproteins, glucose, blood pressure, or renal apoB levels between D-4F- and ScD-4F-treated mice. We conclude that D-4F reduced renal oxidized phospholipids, resulting in lower expression of SREBP-1c, which, in turn, resulted in lower triglyceride content and reduced renal inflammation.

Feeding a Western diet (WD) to LDL receptor-null (LDLR 2/2 ) mice has been shown to induce atherosclerosis, endothelial dysfunction, and inflammation of brain arterioles associated with cognitive dysfunction, all of which have been reported to be significantly ameliorated by treatment with the oral apolipoprotein A-I (apoA-I) mimetic peptide D-4F without a change in plasma lipid levels or blood pressure (1)(2)(3).
In addition to hyperlipidemia, feeding a WD to LDLR 2/2 mice results in insulin resistance and elevated plasma glucose levels (4,5). A major problem facing Western societies is an increase in chronic renal disease that appears to be associated with dyslipidemia and diabetes in addition to hypertension (6). Dyslipidemia has been emphasized as a factor that can exacerbate hypertension-induced renal damage as well as a factor that by itself may have adverse effects on the kidney (6). Diet-induced obesity in C57BL/6J mice has been reported to cause lipid accumulation in kidneys and glomerulosclerosis via a sterol-regulatory element binding protein 1c (SREBP-1c)-dependent pathway (7).
Oxidized phospholipids have been identified as potent mediators of inflammation (8,9). Berliner and colleagues have reported that oxidized phospholipids induce cytokine production in endothelial cells through induction of SREBP (10) via an endothelial nitric oxide synthase (eNOS)mediated mechanism (11).
Pritchard and colleagues reported that the apoA-I mimetic peptide L-4F (identical to D-4F except that the peptide is synthesized from L-amino acids) restored the balance between nitric oxide and superoxide anion production by eNOS in LDL-treated endothelial cells in culture (12). Injection of L-4F in LDLR 2/2 mice on a WD dramatically improved eNOS-mediated vasodilation of small facial arteries (13). Pritchard and colleagues also demonstrated that oral D-4F similarly improved eNOS-mediated vasodilation in this mouse model and reduced the thickness of small artery walls if apoA-I was present (2). Tight-skin mice have many of the myocardial and vascular features seen in scleroderma patients. Pritchard and colleagues reported that these mice also had impaired eNOS-mediated vasodilation that was significantly improved by treatment with D-4F. The tight-skin mice had elevated levels of triglycerides, which were normalized with D-4F treatment. The hearts from the tight-skin mice contained significantly higher levels of autoantibodies against oxidized phospholipids, which were reduced by D-4F treatment (14).
We previously reported that the apoA-I mimetic peptide D-4F reduced the levels of oxidized phospholipids (but not nonoxidized phospholipids) following viral infection of type II pneumocytes in vitro. This reduction in oxidized phospholipids was associated with a dramatic reduction in inflammatory parameters (15).
Given the widespread vascular inflammation reported in LDLR 2/2 mice on a WD (1-3), we hypothesized that feeding a WD to LDLR 2/2 mice would induce the formation of oxidized phospholipids in the kidney and would result in renal inflammation. We also hypothesized that D-4F treatment would reduce oxidized phospholipids in the kidney, which, in turn, would result in decreased renal inflammation. On the basis of the work of Jiang and colleagues (7) and Berliner and colleagues (10,11), we also hypothesized that the WD would induce increased SREBP-1c mRNA levels and increased kidney triglyceride levels that would be ameliorated by D-4F treatment.
As shown here, feeding a WD to female LDLR 2/2 mice for only 7 weeks induced hyperlipidemia and elevated Fig. 1. The Western diet (WD) induced an increase in the glomerulus-to-capsule diameter ratio in the kidneys of LDL receptornull (LDLR 2/2 ) mice treated with the inactive control peptide scrambled D-4F (ScD-4F) but not in mice treated with the active peptide D-4F. In the D-4F kidneys, the glomerulus-to-capsule ratio was similar to that observed in the wild-type (WT) C57BL/6J and LDLR 2/2 mice fed a regular chow diet. Morphometry and staining were performed as described in Materials and Methods. The data shown are mean 6 SEM (n 5 15 mice per group). Note the increased number of nuclei in the glomerular tuft in the ScD-4F glomerulus. The glomerulus shown for the D-4F mice is histologically similar to the control wild-type C57BL/6J glomeruli (not shown). Magnification 3400. plasma glucose levels, which were associated with increased kidney SREBP-1c mRNA levels, increased kidney triglyceride levels, increased kidney oxidized phospholipid levels, and renal inflammation. Treatment with oral D-4F significantly reduced the formation of oxidized phospholipids in mice fed the WD and largely prevented the increase in SREBP-1c mRNA expression elicited by the WD, prevented triglyceride accumulation in the kidneys, and significantly reduced renal inflammation without altering plasma lipids, lipoproteins, glucose, or blood pressure.

Materials
D-4F and scrambled D-4F (ScD-4F; a control peptide with the same D-amino acids but arranged in a sequence that does not promote lipid binding, thus rendering the peptide inactive) were synthesized as described (1,3). All other reagents were from sources previously reported (1,3).

Mice
Female 7-week-old wild-type C57BL/6J and LDLR 2/2 mice on a C57BL/6J background (Jackson Laboratories, Bar Harbor, ME) were maintained on a chow diet (Ralston Purina; 15% kcal from fat). Some of the LDLR 2/2 mice were fed a WD (Teklad/ Harlan, Madison, WI; diet No. 88137, 42% kcal from fat, 0.15% cholesterol, w/w) for 7 weeks. The mice receiving the WD were given either D-4F or ScD-4F at 300 mg/ml in their drinking water. The mice consumed approximately 2.5 ml of water per day per mouse, and there was no significant difference in water or food consumption between groups. The kidneys from some of the mice used in our studies of mouse brains (3) were collected as described below and are included in this report. The University of California, Los Angeles Animal Research Committee approved all studies.

Morphometry and associated statistical methods
All glomerular measurements were performed on 20-30 randomly selected cross sectioned glomeruli per mouse. Using an Olympus BH-2 microscope (403 objective), the glomeruli were photographed by means of SPOT Image software. All measurements were performed on a single focal plane. Photography, cell counting, quantitative measurements, and determination of semiquantitative parameters were performed on blinded slides. The following histological parameters were used to determine the glomerular changes (7, 21-24): a) glomerular tuft/Bowman's capsule diameter ratio (three measurements of the glomerular tuft and of glomerulus plus Bowman's capsule diameters were taken for each and averaged, and the ratios were calculated); b) hypercellularity (determined by counting the number of nuclei in the glomerular tuft); c) integrity of Bowman's capsule visceral layer was determined; the coefficient of variation for three observers was 13.5 6 1%; d) morphology of glomerular capillaries was assessed semiquantitatively (0 5 normal, 1 5 dilated); the coefficient of variation for three observers was 9.5 6 1%; and e) PAS staining for mesangial area. Glomerular immunostaining for MCP-1 and for periglomerular F4/80 accumulation was quantified on 20-30 photographed glomerular cross sections per mouse after importing the images into Image Pro Plus software (Media Cybernetics, Silver Spring, MD). The number of glomeruli immunolabeled for CD68, EO6, and LF3 were counted in 20-30 high-power fields in each mouse and calculated as percent immunopositive glomeruli. Tubulo-interstial immunostaining for MCP-1, EO6, and LF3 was quantified using Image Pro Plus software on 20-30 photographed high-power fields per mouse. For calculating the percent area of the tublo-interstitium, three random areas of similar size (100 3 100 mm squares) devoid of glomeruli were measured and averaged from each high-power (4003) microscope field. Twenty fields from each mouse were counted. The formula used for determining the percent area stained in tubulo-interstitial areas was: (labeled area/total area selected) 3 100. Statistical calculations for morphometry were performed using InStat software (Graph Pad, San Diego, CA).

Quantitative RT-PCR analysis
Total RNA was isolated from kidneys using the RNeasy Kit (Qiagen) according to the manufacturer's protocol. First-strand cDNA was synthesized using random primers and Multiscribe RT (Applied Biosystems). Real-time RT-PCR was then performed on 2 ml cDNA using iQ SYBR Green Supermix (Bio-Rad) on the MyiQ Single-color Real-time Detection System (Bio-Rad). The primers used for specific cDNA synthesis were mouse GAPDH: 5 ¶-AAC TTT GGC ATT GTG GAA GG-3 ¶ and 5 ¶-GGA TGC AGG GAT GAT GTT CT-3 ¶ and mouse SREBP-1c: 5 ¶-GGA GCC ATG GAT TGC ACA TT-3 ¶ and 5 ¶-GGC CCG GGA AGT CAC TGT-3 ¶ (25). The PCR conditions were: 95jC 3 min; 40 cycles of 95jC 30 s, 60jC 50 s, and 72jC 30 s. Values were normalized to GAPDH and calculated using the comparative C T method.

Other procedures
Plasma blood urea nitrogen levels were determined using a colorimetric assay (Bioquart, San Diego, CA). Plasma glucose levels were measured using the ACE: clinical chemistry system (Alfa Wassermann, Inc., West Caldwell, NJ). Plasma lipoprotein and lipid levels were determined as described previously (26). Urine was spot collected and microalbuminuria determined using the competitive binding ELISA assay Albuwell M (Exocell, Philadelphia, PA). Blood pressure was determined as described previously (3).

D-4F prevents WD-induced histological changes in the glomerulus
Measurement of the glomerular tuft-to-Bowman's capsule diameter ratio showed a significant (P , 0.001) increase in the WD-fed mice treated with the control inactive peptide ScD-4F compared with the wild-type C57BL/6J or the LDLR 2/2 mice on chow. Addition of the active peptide D-4F to the drinking water prevented the increase in the glomerular tuft/capsule ratio observed in the ScD-4F mice. The D-4F-treated mice had values similar to those of the chow-fed LDLR 2/2 mice (Fig. 1). To determine whether hypercellularity caused the differences in the glomerulus-to-capsule ratio, we counted the number of nuclei in the glomerulus and found a significant increase (P , 0.0001) in the number of cells per glomerulus in the mice on the WD receiving ScD-4F compared with those receiving D-4F (Fig. 2).
Mesangial cells proliferate in response to atherogenic lipoproteins (27)(28)(29). Using the PAS-hematoxylin stain, we determined that a significantly larger area (P , 0.0001) of the glomerulus was occupied by mesangial cells in the ScD-4F mice as compared with the D-4F-treated mice (Fig. 3), suggesting that the hypercellularity shown in Fig. 2 is at least in part due to an increased number of mesangial cells.
Another glomerular change that is often seen in animal models of glomerular disease is the integrity of the visceral layer of Bowman's capsule. A significantly lower percent of the ScD-4F glomeruli had a normal, clear capsular space (P , 0.001), and conversely, a higher percent of ScD-4F glomeruli showed focal adhesions and a disrupted visceral layer accompanied by spilling of glomerular cell content into the capsular space than was the case in the kidneys of the D-4F group (Fig. 4).  Examination of the capillary contour in the glomerular tuft of the SD-4F mice showed fewer normal and more dilated capillaries than in the D-4F mice, where there were significantly more normal, clear capillaries (P , 0.001) and very few dilated capillaries in the glomerular tuft (Fig. 5).

D-4F decreases WD-induced kidney MCP-1/CCL2
Immunostaining for the chemokine MCP-1/CCL2 showed relatively few immunopositive areas in both the glomerulus (Fig. 6A, B) and in the tubulo-interstitium (Fig. 6C, D) of the D-4F-treated kidneys compared with both the glo- meruli (P , 0.0001) and tubulo-interstitial areas (P , 0.0183) of the control ScD-4F kidneys. Owing to the disruption in the visceral membrane of Bowman's capsule, some of the glomerular content spilled into the capsular space of the control ScD-4F mice and immunostained for MCP-1/CCL2 (Fig. 6B). Except for glomeruli showing disruption in the visceral membrane of Bowman's capsule, immunostaining for MCP-1/CCL2 was closely associated with cellular structures in the glomeruli and in the tubulointerstitial areas.

D-4F reduces WD-induced macrophage accumulation in kidney cortex
Owing to the heterogeneity of macrophage antigen expression in mouse kidney, there have been conflicting reports regarding macrophage distribution in the kidney (30). Therefore, we immunostained for two groups of macrophage antigens, CD68 and F4/80. As Fig. 7 shows, CD68 macrophage immunolabeling was present in both the tubulo-interstitial areas and in the glomeruli. A very small percent of glomeruli were positive for the CD68 antigen in the chow-fed mice, with a significant increase in the WD-fed mice (Fig. 7A, B). Although the mice on the WD that were treated with D-4F showed a significant increase in the percent of glomeruli immunopositive for the CD68 antigen when compared with the chow-fed LDLR 2/2 mice, the CD68 values in the D-4F kidneys were significantly lower than those in the control ScD-4F kidneys (P , 0.001) (Fig. 7B).
Quantification of tubulo-interstitial CD68 immunostaining revealed a marked accumulation of CD68 macrophages in the kidneys of the control ScD-4F mice, with significantly less immunolabeling in the D-4F-treated mice, with levels of immunostaining in the D-4F-treated mice comparable to those of the chow-fed mice (Fig. 7C).
In contrast to CD68, and confirming the report of Masaki et al. (30), there were no F4/80-labeled macrophages in the glomeruli. F4/80-labeled macrophages were found Oxidized phospholipids and SREBP-1c in kidney adjacent to Bowman's capsule, in close proximity to and tightly connected with the parietal wall of Bowman's capsule (31) (Fig. 7D). Quantification of the F4/80-immunolabeled periglomerular areas showed that approximately 5% of the area stained positive for F4/80 in the chow-fed mice, with a significant increase in F4/80 immunostaining in mice fed the WD (Fig. 7E). There was a .4-fold increase in F4/80 immunolabeling in the mice fed the WD and treated with the inactive peptide ScD-4F (P , 0.001) compared with an approximately 1.6-fold increase in the mice on the WD that received D-4F (P , 0.001) (Fig. 7E). The combined CD68 and F4/80 data strongly suggest that the WD induced an inflammatory process characterized by increased macrophage presence in the kidneys, which was largely ameliorated by D-4F treatment but not by treatment with ScD-4F.

D-4F reduces WD-induced renal oxidized phospholipids without altering apoB
The presence of oxidized phospholipid epitopes recognized by the monoclonal antibody EO6 was demonstrated in both the glomeruli and in tubulo-interstitial areas (Fig. 8). Less than 5% of glomeruli of C57BL/6J mice on chow were EO6 immunopositive. In contrast, ap- proximately 40% of the glomeruli in LDLR 2/2 mice on WD treated with the control inactive peptide ScD-4F were EO6 immunopositive, and there was a highly significant difference between these mice and the other three groups. The differences between C57BL/6J mice on chow, LDLR 2/2 mice on chow, and LDLR 2/2 mice on the WD with D-4F treatment were not significant (Fig. 8C). EO6 immunostaining of the tubulo-interstitial areas revealed a significant increase in the LDLR 2/2 mice on chow compared with the wild-type C57BL/6J mice on chow. EO6 immunostaining of the tubulo-interstitial areas of LDLR 2/2 mice on the WD treated with the inactive control ScD-4F peptide was significantly higher than in LDLR 2/2 mice on chow or LDLR 2/2 mice on the WD with D-4F treatment. There was no significant difference in EO6 immunostaining of the tubulo-interstitial areas of LDLR 2/2 mice on chow compared with LDLR 2/2 mice on the WD with D-4F treatment (Fig. 8D).
Although there was a highly significant increase in EO6 immunostaining in the kidneys of LDLR 2/2 mice on the WD treated with the inactive control ScD-4F peptide compared with LDLR 2/2 mice on the WD treated with D-4F ( Fig. 8B-D), there was no significant difference in apoB immunostaining of the glomeruli or the tubulo-interstitial areas (Fig. 8E-G). There was a significant increase in apoB immunostaining between the mice on the WD compared with the mice on the chow diet, but there was no significant difference in apoB immunostaining in mice on the WD treated with the inactive control ScD-4F peptide compared with those treated with the active D-4F peptide (Fig. 8E-G). Similarly, there was a highly significant difference in plasma lipid and lipoprotein levels between chow-fed mice and mice fed the WD, but there was no significant difference in plasma lipid and lipoprotein levels in mice fed the WD and treated with ScD-4F compared with those treated with D-4F ( Table 1).
D-4F treatment prevents the dramatic increase in SREBP-1c mRNA levels induced by feeding LDLR 2/2 mice a WD On the basis of the work of Jiang and colleagues (7) and Berliner and colleagues (10,11), we hypothesized that the WD would increase SREBP-1c mRNA levels and kidney triglyceride levels. We further hypothesized that these changes would be ameliorated by D-4F treatment. Using quantitative PCR, we found that kidney SREBP-1c mRNA levels were low in C57BL/6J wild-type mice on a chow diet. There was a trend toward higher levels in LDLR 2/2 mice on chow, but this increase did not reach statistical significance (Fig. 9). As predicted by the work of Jiang et al. (7), there was a dramatic increase in kidney SREBP-1c mRNA levels in the mice fed the WD and treated with the control inactive ScD-4F peptide. Berliner and colleagues (10,11) demonstrated that oxidized phospholipids induce SREBP. Mice fed the WD and treated with the active peptide D-4F had renal levels of oxidized phospholipids not significantly different from LDLR 2/2 mice on chow (Fig. 8). The mice fed the WD and treated with D-4F also had kidney levels of SREBP-1c mRNA that were not significantly different from LDLR 2/2 mice on chow (Fig. 9). Consistent with the increased SREBP-1c mRNA levels in the control mice on the WD, kidney triglyceride levels were significantly increased in the mice on the WD that were treated with the control inactive ScD-4F peptide but were not increased in mice on the WD that were treated with the active D-4F peptide ( Table 1). As shown in Table 1, there was a significant increase in kidney weight in mice on the WD, but there was no significant difference in kidney weight between mice treated with the control inactive ScD-4F peptide and those treated with the active D-4F peptide. Kidney total cholesterol levels in mice on a WD treated with the control inactive ScD-4F peptide were significantly higher compared with those of C57BL/6J wild-type mice on chow, but there was no significant difference compared with LDLR 2/2 mice on chow or on the WD with D-4F treatment. (Table 1).

DISCUSSION
After feeding C57BL/6J mice a 60 kcal% saturated (lard) fat diet for 12 weeks, Jiang et al. (7) observed a significant increase in kidney triglyceride levels with a trend toward increased kidney cholesterol levels that was not statistically significant. These changes were associated with a significant increase in the expression of SREBP and significant increases in mRNA levels for genes known to be regulated by SREBP-1c (e.g., acetyl-CoA carboxylase and fatty acid synthase) as well as significant increases in mRNA levels for genes implicated in glomerulosclerosis [e.g., plasminogen activator inhibitor-1 (PAI-1), vascular endothelial growth factor (VEGF), type IV collagen, and fibronectin]. These mice also demonstrated increased PAS staining, which is a hallmark of mesangial expansion. In mice lacking SREBP-1c, the diet failed to induce an increase in renal triglyceride levels, and the induction of PAI-1, VEGF, type IV collagen, and fibronectin was markedly attenuated. The authors concluded that diet-induced obesity in the C57BL/6J mice caused increased renal lipid accumulation and glomerulosclerosis via an SREBP-1cdependent pathway (7).
The plasma lipid changes in the wild-type C57BL/6J mice on the high-fat diet in the studies of Jiang et al. (7) were much less than was the case in our studies with LDLR 2/2 mice on a WD. Plasma total cholesterol levels were 223 6 16 mg/dl in Jiang's study compared with 1,444 6 55 mg/dl in LDLR 2/2 mice on the WD and treated with ScD-4F (Table 1). Similarly, plasma triglycerides in Jiang's study were 105 6 7 mg/dl versus 388 6 52 mg/dl in LDLR 2/2 mice on WD and treated with ScD-4F (Table 1). Plasma glucose levels were also different but less so, with values of 216 6 16 mg/dl in Jiang's study compared with values of 305 6 15 mg/dl in LDLR 2/2 mice on the WD and treated with ScD-4F (Table 1). Interestingly, Jiang and colleagues (7) found proteinuria and we did not. This may have been due to the fact that the mice in Jiang's study were slightly older at the start of the study (8 versus 7 weeks) and were maintained on the high-fat diet for almost twice as long as was the case in our study (12 versus 7 weeks). Additionally, the composition of the WD is significantly different compared with the diet in Jiang's study, and we studied the effects of the WD in LDLR 2/2 mice on a C57BL/6J background, whereas Jiang and colleagues studied wild-type C57BL/6J mice (7).
Similar to the findings of Jiang et al. (7), on feeding LDLR 2/2 mice the WD, we found that they developed Fig. 9. D-4F treatment prevents the dramatic increase in sterolregulatory element binding protein-1c (SREBP-1c) mRNA levels induced by feeding LDLR 2/2 mice a WD. SREBP-1c mRNA levels were determined by quantitative PCR analysis as described in Materials and Methods in C57BL/6J WT mice on chow, in LDLR 2/2 mice on chow, and in LDLR 2/2 mice on the WD treated with the active peptide (WD1D-4F) or the inactive control ScD-4F peptide (WD1ScD-4F). Data are expressed as mean 6 SEM (n 5 4 mice per group). The results are representative of two out of two separate experiments. glomerular changes, including evidence of mesangial expansion (Figs. 1-5). Additionally, we found evidence of chronic inflammation, such as increased expression of the potent monocyte chemoattractant MCP-1 (Fig. 6), and evidence of increased renal macrophage content (Fig. 7). The increase in MCP-1 and in macrophages in the kidney is similar to the inflammatory changes that were previously seen in the aorta (1) and in the brain (3) after feeding a WD to LDLR 2/2 mice. These inflammatory changes are precisely what would be expected from an accumulation of oxidized phospholipids (9). Indeed, we found a marked increase in renal oxidized phospholipids (Fig. 8A-E). We also found a significant increase in renal triglycerides (Table 1), and renal SREBP-1c mRNA levels ( Fig. 9). In contrast to these changes in the LDLR 2/2 mice on the WD treated with the control inactive ScD-4F peptide, treating the mice with the active D-4F peptide resulted in a significant reduction in renal oxidized phospholipids ( Fig. 8A-E), triglycerides (Table 1), and renal SREBP-1c mRNA levels (Fig. 9). The reduction in renal oxidized phospholipids, renal triglyceride levels, and renal SREBP-1c mRNA levels in the D-4F-treated mice occurred without a change in plasma lipid, lipoprotein, or glucose levels, without a change in renal apoB levels, and without a change in blood pressure (Table 1). Moreover, the reduction in renal oxidized phospholipids, renal triglyceride levels, and renal SREBP-1c mRNA levels with D-4F treatment was accompanied by a significant reduction in mesangial expansion and other glomerular changes (Figs. [1][2][3][4][5], and a significant reduction in renal MCP-1 (Fig. 6) and renal macrophage content (Fig. 7). Yeh et al. (10) reported that oxidized phospholipids that are recognized by EO6 (16)(17)(18) dramatically induced the expression of SREBP, which was required for the oxidized phospholipids to induce cytokine expression in endothelial cells. We think the most likely explanation for the findings reported here is that the apoA-I mimetic peptide D-4F reduced the tissue levels of oxidized phospholipids, resulting in significantly less SREBP-1c expression, less renal triglyceride accumulation, less cytokine production, and less inflammation without changing plasma lipid, lipoprotein, or glucose levels, without a change in renal apoB levels, and without a change in blood pressure. This scenario is consistent with the proposed mechanism of action for apoA-I mimetic peptides, in which they provide a thermodynamically stable environment for binding, sequestering, and removing pro-inflammatory oxidized lipids (32).