Cell type-specific control of cardiometabolic disease by the dileucine motif of the LDL receptor-related protein LRP1

Abstract

The LDL receptor-related protein-1 (LRP1) modulates cardiometabolic diseases in a cell type-specific manner and is regulated via several motifs in its cytoplasmic domain. This study compared the cardiometabolic disease phenotype of Ldlr^−/− mice expressing the normal Lrp1 gene (Lrp1^LL) or harboring DVGGVLL⁴⁴⁸⁸ to DVGGVAA⁴⁴⁸⁸ mutation (Lrp1^AA) after feeding a Western-type high-fat high-cholesterol diet. Results showed comparable body weight gain and overall fat mass between Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice, but less adipocyte hypertrophy and adipose tissue inflammation, as well as reduced hepatosteatosis in the mutant Lrp1^AALdlr^−/− mice. The reduced hepatosteatosis coincided with reduced expression of cholesterol synthesis genes and increased expression of PPAR-responsive fatty acid oxidation genes in the liver. Elevated expression of PPAR-responsive genes was also observed in blood cells of Lrp1^AALdlr^−/− mice after oxidized LDL (oxLDL) or LPS activation, resulting in suppression of acute leukocyte inflammatory response. Despite this apparent anti-inflammatory response, the Lrp1^AALdlr^−/− mice displayed exaggerated atherosclerosis. Reciprocal bone marrow transplant experiments revealed that the Lrp1^LL to Lrp1^AA mutation in bone marrow-derived cells was responsible for the atherosclerosis increase. Mechanistically, the data showed that Lrp1^AA mutation in cholesterol-loaded macrophages caused mitochondrial dysfunction, with lower levels of mitochondrial fission proteins and reduced fatty acid oxidation capabilities. Taken together, these results documented a cell-specific role of the LRP1 proximal dileucine motif in LRP1 modulation of cardiometabolic diseases. The cardiometabolic phenotype of Lrp1^AALdlr^−/− mice differs from phenotypes observed in mice with LRP1 inactivation or NPxY motif mutation, thus highlighting a unique role of this LRP1 motif in mediating LRP1 functions.

The LDL receptor related protein-1 (LRP1) is a ubiquitously expressed transmembrane protein that shares structural similarities with other members of the LDL receptor family, including a large 515-kDa extracellular α-subunit that contains 4 clusters of ligand-binding domain of cysteine-rich repeating units with negatively charged amino acids and an epidermal growth factor-like domain, along with a smaller 85-kDa β-subunit that contains a single hydrophobic transmembrane domain and a short cytoplasmic domain. This large protein was first identified as the hepatic receptor for binding apoE-containing lipoproteins to mediate their clearance from the plasma circulation (1). Since its initial discovery, LRP1 was found to be expressed ubiquitously in all tissues and cell types and is capable of interaction with a plethora of different ligands, including protease-protease inhibitor complexes, extracellular matrix proteins, growth factors, apoptotic cells, toxins and viruses (2). In addition to mediating internalization of these ligands and their delivery to lysosomes for degradation, LRP1 is also a liaison receptor in transmitting signals from extracellular ligands to the cell interior for cell regulation. The signal transduction properties of LRP1 are mediated via binding of intracellular adaptor proteins to specific motifs in the cytoplasmic domain of the receptor (3).

The 100-amino acid residue cytoplasmic domain of LRP1 contains 2 NPxY motifs, a YxxL motif, and 2 dileucine motifs that interact with various adaptor molecules to modulate cell signaling, endocytosis, and intracellular trafficking (3). The NPxY motif proximal to the plasma membrane (NPxY⁴⁴⁷³) interacts with sorting nexin 17 to mediate LRP1 cell surface translocation and recycling in the early endosomes (4, 5). Functional mutation in the proximal NPxY motif impairs LRP1 translocation to the cell surface (4, 6) and causes fetal liver destruction and prenatal death (7). The distal NPxY⁴⁵⁰⁷ motif overlaps with the YxxL⁴⁵¹⁰ motif to form the NPxYxxL sequence, which, along with the distal dileucine motif (RELL⁴⁵³¹), is the primary signal for endocytosis (8, 9). The distal NPxY⁴⁵⁰⁷ motif can also be tyrosine phosphorylated to direct cellular trafficking and signaling events (10, 11, 12, 13). In the non-phosphorylated state, the distal NPxY⁴⁵⁰⁷ interacts with the adaptor protein diabled-2 to facilitate LRP1 endocytosis, whereas tyrosine phosphorylation of NPxY⁴⁵⁰⁷ switches its interaction to the signal transducer SH2 domain protein-C1 and promotes activation of phosphoinositide 3-kinase/protein kinase B and peroxisome proliferator-activated receptor-γ/liver X receptor (PPARγ/LXR) pathways in a cell-specific manner (14). Specifically, disabling NPxY⁴⁵⁰⁷ phosphorylation via Y4507F mutation has no effect on LRP1 functions in smooth muscle cells but inhibits cholesterol efflux from macrophages, thereby enhancing cholesterol accumulation and inflammation to accelerate atherosclerosis in Western diet-fed Ldlr^−/− mice (14). Interestingly, mutation of the NPxY⁴⁵⁰⁷ motif to AAxA⁴⁵⁰⁷, which is expected to abolish its interaction with both endocytosis adaptor proteins and signal transducer adaptor proteins, reduces dietary cholesterol-induced hypercholesterolemia, obesity, steatohepatitis and brain inflammation due to compensatory increase in hepatic LDL receptor expression (15). This latter report suggested the interplay between LRP1 and LDL receptor in modulating cholesterol-induced cardiometabolic diseases.

The role of the proximal dileucine motif (DVGGLL⁴⁴⁸⁸) and its physiological significance in LRP1 functions has drawn considerably less attention. Based on its similarities to the conserved LXXLL motif in PPAR coactivators (16), Mao et al. proposed that LRP1 may be a coactivator of PPAR transcriptional activities and showed that mutation of DVGGLL⁴⁴⁸⁸ to DVGGAA⁴⁴⁸⁸ in LRP1 abolished its interaction with the PPAR nuclear receptors and reduces PPARγ transcriptional activity in endothelial cells (17). Their study also showed that endothelial-specific LRP1 inactivation reduces PPARγ transcriptional activity, leading to lower pyruvate dehydrogenase kinase-4 levels and the protection against diet-induced obesity and glucose intolerance. Whether the LRP1 proximal dileucine motif-mediated PPAR activation and metabolic regulation is restricted to endothelial cells remains unclear. However, the report showing pulmonary arterial hypertension in smooth muscle-specific LRP1 knockout mice can be reversed by PPARγ activation suggests that LRP1 may also mediate PPARγ activation in smooth muscle cells (18). The goal of this study was to clarify the physiological role of the proximal dileucine motif in LRP1 modulation of cardiometabolic diseases. To accomplish this goal, we generated a mutant mouse model in which the proximal dileucine motif DVGGLL⁴⁴⁸⁸ has been replaced with DVGGAA⁴⁴⁸⁸. The LRP1 dileucine mutant mice were then mated with Ldlr^−/− mice and then fed a Western type high fat-high cholesterol diet to ascertain the impact of this mutation on diet-induced hyperlipidemia, adiposity, hepatosteatosis, and atherosclerosis.

Results

Generation of LRP1 proximal dileucine mutant mice

The importance of the LRP1 proximal dileucine motif in modulating cell functions and disease manifestation was explored by generating a mutant mouse model in which the proximal dileucine motif in LRP1 has been altered in all tissues and cell types. The Crispr-Cas 9 homologous recombination approach using an LRP1 guide RNA (gRNA) encompassing the DVGGLL sequence and a donor oligonucleotide that encodes a DVGGAA sequence was used to accomplish this task (Fig. 1A). The donor oligonucleotide also introduced a PvuII restriction site for ease of genotyping the mutant allele without changing the coding sequence. Genotyping based on PvuII restriction digest of PCR products surrounding the mutated sequence, as well as genome sequencing of germline mice confirmed transmission of the modified allele and that the appropriate and specific LL→AA mutation had been made (Fig. 1B). Analysis of LRP1 expression levels in several tissues revealed no difference in expression levels between control LRP1 (Lrp1^LL) mice and mice expressing LRP1 with the dialanine substitution in both alleles (Lrp1^AA) (Fig. 1C).

Figure 1.

Generation of LRP1 proximal dileucine motif mutant mice.A, sequences for guide RNA and donor oligonucleotide with Lrp1^LL to Lrp1^AA mutation codons highlighted by italicized letters. B, PvuII restriction digest of DNA obtained from mice with LL, AA, and LA genotype. C, Western blot analysis of LRP1 levels in various tissues of Lrp1^LL (LL) and Lrp1^AA (AA) mice. Expression levels of Lrp1^LL and Lrp1^AA were normalized to β-actin expression levels. The data represent mean ± S.D. from 2 mice in each group.

Proximal dileucine motif mutation in LRP1 limits adipose tissue inflammatory response

The Lrp1^LL and Lrp1^AA mice were mated with Ldlr^−/− mice to produce Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice and the progenies were fed a Western-type, high-fat–high-cholesterol diet to assess the role of the LRP1 proximal dileucine motif in cardiometabolic disease manifestation. Results showed comparable body weight gain between Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice over the course of a 16-week Western diet-feeding period (Fig. 2A). Although body lean mass and fat mass were also comparable between the two groups of mice (Fig. 2, B and C), epididymal white adipose tissue (eWAT) mass was significantly higher in the mutant Lrp1^AALdlr^−/− mice (Fig. 2D). Histological examination of eWAT from these animals revealed a slight but significantly higher number of smaller adipocytes in the mutant Lrp1^AALdlr^−/− mice compared to those observed in the control Lrp1^LLLdlr^−/− mice. However, the number of larger-sized adipocytes was comparable between the two groups (Fig. 2, E and F). Interestingly, more crown-like structures, indicative of monocyte-derived macrophages, were observed in the eWAT of Lrp1^AALdlr^−/− mice compared to those observed in Lrp1^LLLdlr^−/− mice (Fig. 2G). The increased presence of macrophages in eWAT of the mutant Lrp1^AALdlr^−/− mice was confirmed by their higher expression levels of EMR1 (also known as F4/80) and macrophage inflammatory protein-1α (MIP-1α) (Fig. 2, H and I). However, despite the increased presence of macrophages in the eWAT of the mutant Lrp1^AALdlr^−/− mice, expression levels of the macrophage-specific inflammatory cytokine tumor necrosis factor-α (TNFα) were comparable in the eWAT of control Lrp1^LLLdlr^−/− and mutant Lrp1^AALdlr^−/− mice (Fig. 2J). In contrast, expression levels of Triggering Receptor Expressed on Myeloid Cells 2 (TREM2) were higher in the mutant Lrp1^AALdlr^−/− mice compared to the control Lrp1^LLLdlr^−/− mice (Fig. 2K). Since TREM2 is known to mediate anti-inflammatory and pro-resolving responses (19, 20), and TREM2-positive macrophages in adipose tissues have been shown to protect against diet-induced adipocyte hypertrophy (21), these results indicate that mutation in the proximal dileucine motif of LRP1 may limit adipocyte hypertrophy and the consequential effects on metabolic disease manifestation.

Figure 2.

Western diet effects on body weight and adipose tissues of Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− (AA) mice.A, body weights of LL (n = 14) and AA (n = 12) mice in response to Western diet feeding for 16 weeks (B) Lean mass, (C) fat mass, and (D) eWAT tissue mass were obtained from the same animals after 16 weeks of Western diet feeding. E and F, histological images and adipocyte cell size in LL and AA mice after feeding the Western diet for 16 weeks. Bar = 100 μm. G, crown-like structures (CLS) per 1000 adipocytes in eWAT of 7 LL and 5 AA mice after 16 weeks on Western diet. Expression levels of (H) EMR1, (I) MIP-1α, (J) TNFα, and (K) TREM2 from 6 LL and 6 AA mice after feeding the Western diet for 26 weeks. All data were evaluated for statistical significance by Student’s t test with p values shown. ∗ in panelF indicates differences at p < 0.05.

Proximal dileucine motif mutation in LRP1 reduces diet-induced hepatosteatosis

Histological examination of the livers of Western diet-fed Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice showed significant differences in lipid droplet size between the two groups. While the lipid droplets in the livers of Lrp1^LLLdlr^−/− mice were mostly large droplets typical of macrosteatosis observed in Western diet-fed mice, lipid droplets in the livers of Lrp1^AALdlr^−/− mice were mostly of smaller size indicative of microsteatosis (Fig. 3A). Interestingly, while total hepatic triglyceride content was similar between the two groups of mice, hepatic cholesterol content was significantly lower in the mutant Lrp1^AALdlr^−/− mice compared to the control Lrp1^LLLdlr^−/− mice (Fig. 3A). Analysis of expression profile of lipid metabolism genes revealed reduced expression of cholesterol synthesis genes such as HMG-CoA reductase, HMG-CoA synthase, and sterol regulatory element binding protein-2 in livers of Lrp1^AALdlr^−/− mice compared to Lrp1^LLLdlr^−/− mice (Fig. 3B). In contrast, elevated expression of fatty acid oxidation genes such as acyl-CoA oxidase-1 (Acox1), carnitine palmitoyltransferase 1A (Cypt1a), and pyruvate dehydrogenase kinase 4 (Pdk4) was observed in the livers of mice with the dileucine mutation (Fig. 3C). The elevated expression of these fatty acid oxidation genes was not due to increased transcription of peroxisomal proliferator-activated receptor genes Ppara, Ppard, or Pparg (Fig. 3D). However, a 2- and 3-fold higher levels of PPARβ/δ and PPARγ proteins were observed in the livers of Western diet-fed Lrp1^AALdlr^−/− mice compared to Lrp1^LLLdlr^−/− mice (Fig. 3E), thus indicating that elevated levels of PPARβ/δ and PPARγ were responsible for the increased expression of PPAR-responsive fatty acid oxidation genes that hydrolyzed stored triglycerides in the livers of Lrp1^AALdlr^−/− mice compared to Lrp1^LLLdlr^−/− mice (22). The reduced expression of cholesterol synthesis genes along with increased expression of fatty acid oxidation genes may also account for the smaller size of lipid droplets in the livers of Western diet-fed Lrp1^AALdlr^−/− mice compared to control Lrp1^LLLdlr^−/− mice.

Figure 3.

Hepatosteatosis and liver gene expression in Western diet-fed Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− (AA) mice. Age-matched LL and AA mice were fed Western diet for 16 weeks. A, livers were obtained from 7 LL and 6 AA mice for histological examination of steatosis and measurements of triglyceride and cholesterol content. Total RNA was extracted from 6 to 7 LL and 6 to 7 AA mice for RT-PCR to quantify expression of: (B) cholesterol synthesis genes HMGCR, HMGCS, and SREBP2, (C) fatty acid oxidation genes ACOX1, CYPT1α, and PDK4, and (D) PPARα, PPARβ, and PPARγ. The data were evaluated for statistical significance by Student’s t test with p values as shown. E, protein levels of PPARα, PPARβ/δ, and PPARγ in livers prepared from Western diet-fed LL and AA mice were assessed by Western blot analysis of liver lysates using GAPDH as loading control. The data were normalized to the mean levels of expression in LL samples, with fold changes as indicated.

Proximal dileucine motif mutation in LRP1 exacerbates diet-induced hyperlipidemia

Increased PPAR transcriptional activities have also been shown to reduce VLDL production (23). Therefore, we also compared VLDL production between Western diet-fed Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice by injecting fasting mice with Poloxamer 407 to block lipolysis and then measured plasma triglyceride levels at hourly intervals for 4 h. Consistent with results of increased liver PPAR transcriptional activities, lower VLDL production was observed in Lrp1^AALdlr^−/− mice compared to control Lrp1^LLLdlr^−/− mice without the dileucine mutation (Fig. 4A). Interestingly, despite the reduced VLDL production rates observed in the Lrp1^AALdlr^−/− mutant mice, fasting plasma triglyceride levels in these animals were similar to those observed in the control Lrp1^LLLdlr^−/− mice (Fig. 4B). However, fasting plasma cholesterol levels were found to be significantly higher in the Lrp1^AALdlr^−/− mice compared to Lrp1^LLLdlr^−/− mice (Fig. 4B). Analysis of plasma cholesterol and triglyceride distribution among various classes of lipoproteins revealed higher levels of cholesterol in the chylomicron remnants/VLDL and IDL/LDL fractions, whereas triglyceride levels in the chylomicron remnants/VLDL fractions were marginally impacted by the LRP1 dileucine mutation (Fig. 4C). The higher cholesterol levels in chylomicron remnants/VLDL and IDL/LDL fractions despite lower expression of cholesterol biosynthesis genes observed in the livers of Lrp1^AALdlr^−/− mice suggested the possibility of delayed plasma clearance of these atherogenic lipoproteins. This hypothesis was supported by higher apoB48 levels, but not apoB100 and apoE levels, in the plasma of Lrp1^AALdlr^−/− mice compared to those observed in the Lrp1^LLLdlr^−/− mice (Fig. 4D). Since fasting plasma apoB48 levels are indicators of postprandial triglyceride-rich lipoprotein clearance rates (1); taken together, these results indicated that LRP1 dileucine mutation has no influence on plasma triglyceride hydrolysis and lipoprotein remodeling but reduced the plasma clearance of apoB48-containing remnant lipoproteins to exacerbate diet-induced hypercholesterolemia in Lrp1^AALdlr^−/− mice.

Figure 4.

Plasma lipids and lipoproteins in Western diet-fed Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− mice. The LL and AA mice were fed a Western-type diet for 16 weeks. A, VLDL production was assessed in fasting animals (n = 3 per group) by measuring plasma triglyceride levels and determining changes in plasma triglyceride levels (Δtriglyceride) hourly after injection of Poloxamer 407 to inhibit lipolysis. B, plasma cholesterol and triglyceride levels were measured in 14 LL and 12 AA mice after an overnight fast. C, pooled plasma samples from fasting LL and AA mice were fractionated by FPLC for lipoprotein distribution analysis. The elution profiles were compared to elutions of standard chylomicron remnants/VLDL, IDL/LDL and HDL as indicated. D, plasma apolipoproteins were identified by Western blot analysis of plasma samples (n = 8 per group) for apoB (identifying both apoB100 and apoB48), apoE, and apoAI. Levels of apoB100, apoB48, and apoE were normalized to apoAI levels in plasma. All data were evaluated for significance by Student’s t test with p values as shown.

LRP1 proximal dileucine motif mutation reduces LPS-induced blood cell inflammation

In addition to elevated expression of mitochondrial fatty acid oxidation genes in the liver, expression of PPAR-responsive genes, such as Cpt1a and Acox1, was also found to be elevated in oxLDL-activated macrophages isolated from Lrp1^AALdlr^−/− mice compared to those from Lrp1^LLLdlr^−/− mice (Fig. 5, A and B). Since elevated PPAR activities in macrophages are anti-inflammatory, we ascertain the possibility that proximal dileucine mutation in LRP1 may influence inflammatory response of circulating leukocytes. In these experiments, blood samples from Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice were incubated in the presence or absence of LPS in vitro and inflammatory cytokines secreted into the plasma compartment were measured by ELISA. Results showed that whereas IL-6 levels were similar between Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− samples in the absence of LPS incubation, IL-6 secreted into the plasma was significantly lower in Lrp1^AALdlr^−/− blood samples compared to samples from the control Lrp1^LLLdlr^−/− blood cells (Fig. 5C). Plasma levels of another inflammatory cytokine, TNFα, were also lower in Lrp1^AALdlr^−/− samples than Lrp1^LLLdlr^−/− samples when incubated in the presence or absence of LPS (Fig. 5D). These data indicated that LRP1 proximal dileucine mutation limits acute leukocyte inflammatory response to LPS.

Figure 5.

Fatty acid oxidation gene expression and inflammatory cytokine secretion by activated leukocytes in Western diet-fed Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− (AA) mice. Blood obtained from Western diet-fed LL (n = 7–8) and AA (n = 7-8) mice were treated with or without oxLDL prior to RNA isolation for RT-PCR quantification of (A) Cpt1a and (B) Acox1 mRNA levels. Expression levels of CypA mRNA levels were used as loading controls. Blood from these animals were also incubated with or without LPS prior to plasma isolation for ELISA measurements of (C) IL-6 and (D) TNFα secretion levels. Differences between LL and AA samples under various treatment conditions were analyzed for significance by Student’s t test with p values as indicated.

LRP1 proximal dileucine motif mutation exacerbates diet-induced atherosclerosis in Ldlr^−/− mice

In view of the exaggerated hypercholesterolemia observed in Lrp1^AALdlr^−/− mice in comparison with Lrp1^LLLdlr^−/− mice, yet their macrophages displayed anti-inflammatory properties ex vivo, we explored whether the LRP1 proximal dileucine mutation may impact atherosclerosis development by comparing atherosclerotic lesion sizes in Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice after Western diet feeding. En face Oil Red O staining of the whole aorta revealed significant increase in atherosclerotic lesion area in the Lrp1^AALdlr^−/− mutant mice compared to control Lrp1^LLLdlr^−/− mice. The increased lesion area was observed in both the thoracic and the abdominal regions of the aorta (Fig. 6A). Analysis of aortic root cross-sections confirmed increased atherosclerotic lesion size that leads to increased vascular occlusion in Lrp1^AALdlr^−/− mutant mice compared to control Lrp1^LLLdlr^−/− mice (Fig. 6B). Additional analysis of the aortic root lesions by immunohistological staining of CD68 revealed significant increase in macrophages and/or macrophage-like foam cells in the lesion area of Lrp1^AALdlr^−/− mutant mice compared to control Lrp1^LLLdlr^−/− mice (Fig. 6B). These data indicated that proximal dileucine motif mutation in LRP1 accelerates atherosclerosis development despite the apparent ex vivo anti-inflammatory properties exhibited acutely by leukocytes from Lrp1^AALdlr^−/− mice. The accelerated atherogenesis observed in Lrp1^AALdlr^−/− mice was likely due to hypercholesterolemia-induced lipid deposition into monocytes- and/or smooth muscle-derived foam cells.

Figure 6.

LRP1 proximal dileucine motif mutation accelerates atherosclerosis in Ldlr^−/− mice. The Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− (AA) mice were fed Western type-diet for 16 weeks. A, Representative images and quantification of Oil Red O-stained atherosclerotic lesion area in the whole aorta, thoracic aorta, and abdominal aorta of LL and AA mice (n = 8 per group). B, representative images and quantification of lesion area, percent vascular occlusion, and CD68+ areas in the aortic roots of LL and AA mice (n = 14 per group). All data were evaluated for significance by Student’s t test with p values as shown.

Bone marrow cells with LRP1 proximal dileucine mutation accelerate atherogenesis in Ldlr^−/− mice

LRP1 is expressed in numerous cell types including hepatocytes, monocytes/macrophages, and vascular endothelial and smooth muscle cells that may influence atherosclerosis development and progression. Therefore, additional experiments were performed to determine whether the increased atherosclerosis observed in Lrp1^AALdlr^−/− mice was due to impaired LRP1 functions in macrophages or LRP1 defects systemically in other cell types. For these experiments, bone marrow cells derived from Lrp1^LLLdlr^−/− or Lrp1^AALdlr^−/− mice were transplanted into Lrp1^LLLdlr^−/− recipient mice to investigate how the dileucine-to-dialanine mutation in macrophage LRP1 impacts diet-induced atherosclerosis. Results showed that LRP1 proximal dileucine mutation only in bone marrow cells of Lrp1^LLLdlr^−/− mice increased plasma cholesterol and triglyceride levels marginally, although the differences were not statistically significant (Fig. 7, A and B). However, bone marrow cells with the LRP1 AA genotype significantly increased atherosclerotic lesion area in the whole aorta as well as the lesion area in the aortic roots of Lrp1^LLLdlr^−/− recipient mice (Fig. 7, C and D). Conversely, Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice transplanted with bone marrows expressing only the wild type LRP1 (Lrp1^LL) showed insignificant decrease in plasma cholesterol levels but the differences in plasma triglyceride levels were found to be significantly reduced in the mutant Lrp1^AALdlr^−/− mice even with normal LRP1 genotype in their bone marrow cells (Fig. 7, E and F). Despite the difference in plasma triglyceride levels, atherosclerotic lesion area in the whole aorta, as well as aortic root lesion size, were similar in Lrp1^LLLdlr^−/− bone marrow recipient mice regardless of the LRP1 genotype of the host (Fig. 7, G and H). Taken together, these data indicated that LRP1 proximal dileucine motif mutation in bone marrow cells were responsible for the increased atherosclerosis observed in Lrp1^AALdlr^−/− mice and that LRP1 dileucine motif mutation in the liver and vascular cells has minimal impact on atherosclerosis.

Figure 7.

LRP1 with proximal dileucine motif mutation in bone marrow cells exacerbates atherosclerosis. Bone marrows from Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− (AA) mice were transplanted into Lrp1^LLLdlr^−/− recipient mice and then fed the Western type diet for 16 weeks (A) Plasma cholesterol levels in 8 LL and 6 AA mice, (B) plasma triglyceride levels in 8 LL and 6 AA mice, (C) atherosclerotic plaque area in the whole aorta of 7 mice per group, and (D) atherosclerotic lesion area in the aortic roots of 7 mice per group were analyzed. Bone marrows from LL mice were also transplanted into LL and AA recipient mice. After feeding the Western diet for 16 weeks, fasting plasma samples from 10 LL and 10 AA mice were analyzed for (E) cholesterol and (F) triglyceride levels. G, atherosclerotic lesion area in the whole aortas of LL and AA mice (n = 76 per group) and (H) in the aortic roots (n = 10 per group) were characterized. All data were evaluated for significance by Student’s t test with p values as shown.

Proximal dileucine mutation in LRP1 lowers mitochondrial fatty acid oxidation and alters metabolic phenotype in macrophages

The mechanism by which LRP1 proximal dileucine mutation in macrophages exacerbates atherosclerosis despite the apparent increase in oxLDL-induced PPAR activities remains unclear. In view of studies documenting that altered metabolism and bioenergetics also play an important role by which cholesterol-loaded macrophages contribute to atherosclerosis, with reduced mitochondrial oxidative phosphorylation to prevent the repolarization of inflammatory macrophages to an anti-inflammatory phenotype (24), additional experiments were performed to determine if LRP1 proximal dileucine mutation alters cellular bioenergetics in cholesterol-loaded macrophages. For these studies, bone marrow-derived macrophages (BMDM) isolated from Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice were incubated with acLDL for 24 h to induce cholesterol-loaded foam cell formation, while macrophages incubated in the absence of acLDL were used as controls. The cells were then subjected to Seahorse extracellular flux analysis to measure oxygen consumption rate to determine oxidative phosphorylation (Fig. 8, A and B). Results showed no difference in basal or maximal respiration rates between Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− macrophages in the absence of acLDL. Upon preincubation with acLDL to make foam cells, basal respiration rate was significantly elevated in Lrp1^LLLdlr^−/− macrophages but not in Lrp1^AALdlr^−/− macrophages (Fig. 8C). While maximal respiration rate after acLDL preincubation was elevated in both Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− macrophages, the maximal respiration rate increase in Lrp1^AALdlr^−/− macrophages was significantly less than that observed in Lrp1^LLLdlr^−/− macrophages (Fig. 8D). Interestingly, when etomoxir was included in the incubation medium to inhibit fatty acid oxidation, no difference in basal respiration rates was observed between untreated and acLDL-treated Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− macrophages (Fig. 8E). Moreover, the acLDL-induced maximal respiration rate observed in Lrp1^LLLdlr^−/− macrophages was attenuated by etomoxir treatment (comparing data in Fig. 8, D and F). Consistent with the higher respiration rate observed in acLDL-treated Lrp1^LLLdlr^−/− macrophages, ATP production was also increased in Lrp1^LLLdlr^−/− macrophages but not in Lrp1^AALdlr^−/− macrophages after acLDL-induced cholesterol accumulation (Fig. 8G). The increased ATP production observed in Lrp1^LLLdlr^−/− macrophages was also ameliorated in the presence of etomoxir (Fig. 8H). Taken together, these results indicated that the increased oxidative phosphorylation and mitochondrial ATP production observed in acLDL-induced foamy Lrp1^LLLdlr^−/− macrophages was due to elevation of fatty acid oxidation, but the capacity for fatty acid oxidation was significantly lower in Lrp1^AALdlr^−/− macrophages. The reduced mitochondrial oxidative phosphorylation activities of Lrp1^AALdlr^−/− macrophages may be one factor contributing to the increased atherosclerosis observed in Western diet-fed Lrp1^AALdlr^−/− mice.

Figure 8.

LRP1 proximal dileucine mutation reduces mitochondrial respiration in cholesterol-loaded macrophages. Bone marrow-derived macrophages were incubated with or without acLDL prior to determination of oxygen consumption rates (OCR) (A and B). Oligomycin, FCCP, and rotenone/antimycin (rot/AA) were added to measure OCR and determination of basal (C) and maximal respiration (D). Etomoxir was added at 3 μM to inhibit fatty acid oxidation to determine OCR for assessment of basal respiration (E) and maximal respiration (F) due to non-fatty acid sources. Total mitochondrial ATP production (G) and etomoxir-resistant ATP production (H) were calculated from the OCR data. Differences between LL and AA samples under various treatment conditions were analyzed for significance by Student’s t test. ∗ indicates p < 0.05; ∗∗ indicates p < 0.01; ∗∗∗p < 0.001; n.s. indicates not significant.

In contrast to mitochondrial oxidative phosphorylation, the rate of glycolysis measured as proton efflux rate revealed no significant differences between Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− macrophages (Fig. 9, A and B). Both cell types showed similar increase in proton efflux rates after preincubation with acLDL (Fig. 9C). As expected, etomoxir treatment to inhibit fatty acid oxidation did not influence glycolytic rates in either Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− macrophages (compare Fig. 9, C and D). Maximal glycolysis rates determined by measuring proton efflux rates in the presence of oligomycin to block mitochondrial ATP production also revealed increased rate of glycolysis-mediated ATP production in acLDL-treated foamy macrophages compared to untreated cells, but the rates of ATP production were similar between Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− macrophages (Fig. 9E). Etomoxir treatment also had minimal impact on glycolysis-mediated ATP production in untreated and acLDL-treated Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− macrophages (Fig. 9F). Taken together, we interpret these data along with the mitochondrial oxidative phosphorylation data to indicate that acLDL-treatment increased oxidative phosphorylation and glycolysis in Lrp1^LLLdlr^−/− macrophages, but the acLDL-treated foamy Lrp1^AALdlr^−/− macrophages were only glycolytic. Moreover, while foamy Lrp1^LLLdlr^−/− macrophages relied more on oxidative phosphorylation for ATP production, foamy Lrp1^AALdlr^−/− macrophages were defective of mitochondrial oxidative phosphorylation and were more dependent on glycolysis for ATP generation.

Figure 9.

LRP1 proximal dileucine mutation has minimal impact on glycolytic activities in macrophage. Bone marrow-derived macrophages were incubated with or without acLDL prior to determination of glycolysis though measurement of proton efflux rates (PER) (panels A, B). Oligomycin, FCCP, and rotenone/antimycin (rot/AA) were added to measure PER (C). Etomoxir was added at 3 μM to determine etomoxir-resistant glycolysis (D). Total glycolytic ATP production capacity (E) and etomoxir-resistant glycolytic ATP production capacity (F) were calculated from the PER data. Differences between LL and AA samples under various treatment conditions were analyzed for significance by Student’s t test. ∗∗∗ indicates difference at p < 0.0001.

Proximal dileucine motif mutation in LRP1 alters mitochondrial homeostasis in macrophages

The underlying mechanism by which proximal dileucine motif mutation in LRP1 reduces mitochondrial respiration was explored by comparing mitochondrial homeostasis in macrophages prepared from Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice. The macrophages were incubated with acLDL for 24 h and lysates were prepared for Western blot analysis of mitochondrial fission and fusion proteins. Results showed a 50 to 80% reduction of mitochondrial fission proteins mitochondrial fission factor-1 and -2 (MFF) and 50% reduction of the fission protein dynamin-related protein-1 (DRP1) in acLDL-treated Lrp1^AALdlr^−/− macrophages compared to Lrp1^LLLdlr^−/− macrophages (Fig. 10). In contrast, the levels of proteins associated with mitochondrial fusion such as mitofusin-2 (MFN2) and optic atrophy-1 (OPA1) were not significantly different between the two groups (Fig. 10). The reduced mitochondrial fission capacity of cholesterol-loaded Lrp1^AALdlr^−/− macrophages was not due to alterations in cellular uptake of lipoproteins that bind to LRP1 such as aggregated LDL (Fig. 11A). However, the association of LRP1 with mitochondria was found to be reduced by∼20 to 25% in cholesterol-loaded macrophages from Lrp1^AALdlr^−/− mice compared to those derived from Lrp1^LLLdlr^−/− mice (Fig. 11B). Since mitochondrial fission contributes to maintenance of cellular bioenergetic capacity by removal of damaged organelles from the mitochondrial network (25, 26), the reduced mitochondrial respiratory activity in acLDL-treated Lrp1^AALdlr^−/− macrophages is likely due to reduced mitochondrial fission capacity caused by reduced association of LRP1 with mitochondria in the absence of the proximal dileucine motif.

Figure 10.

LRP1 proximal dileucine mutation increased levels of mitochondrial fission proteins in cholesterol-loaded macrophages. Bone marrow macrophages isolated from Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− (AA) mice (n = 2 per group) were treated with acLDL for 24 h for cholesterol loading. Protein lysates were prepared for Western blot analysis of mitochondrial fission proteins: (A) mitochondrial fission factor MFF and (B) dynamin-related protein-1 DRP1; as well as mitochondrial fusion proteins: (C) mitofusin-2 (MFN2) and (D) optic atrophy-1 (OPA1). β-actin level in each sample was used as loading control. The data were normalized to the mean levels of expression in LL samples, with fold changes as indicated.

Figure 11.

LRP1 proximal dileucine motif mutation does not impact aggLDL uptake by reduced LRP1 association with mitochondria in cholesterol-loaded macrophages.A, bone marrow macrophages isolated from Lrp1^LLLdlr^−/− (LL) and Lrp1^AALdlr^−/− mice (n = 6 per group) were incubated with DiI-labeled aggLDL overnight. Intracellular DiI accumulation was determined by fluorometry to determine cellular uptake of aggLDL. B, the bone marrow macrophages were treated with acLDL for 24 h for cholesterol loading. Total cell lysates and mitochondria fractions were prepared for Western blot analysis of total LRP1 levels and LRP1 association with mitochondria. β-actin levels in cell lysates and ATP5A levels in mitochondrial fractions were used as loading control. The data were normalized to the mean levels of expression in LL samples, with fold changes as indicated.

Discussion

Earlier studies have shown that global LRP1 knockout mice are embryonic lethal due to failed blastocyst implantation (27). Tissue-specific Lrp1 gene inactivation revealed its importance in the homeostasis of the vessel wall, central nervous system, liver, adipose tissues and the immune system (3, 28). In the context of cardiometabolic diseases, LRP1 inactivation in smooth muscle cells or perivascular adipose tissues exacerbates atherosclerosis (29, 30), while LRP1 deficiency in hepatocytes promotes metabolic fatty liver disease and its progression to steatohepatitis (31, 32). In contrast, LRP1 inactivation in endothelial cells is protective against diet-induced glucose intolerance and hyperlipidemia (17). The role of macrophage LRP1 in cardiometabolic diseases appears to be more complex. The inactivation of macrophage LRP1 in Ldlr^−/− mice has been shown to reduce hepatic inflammation and protects against diet-induced insulin resistance (33). In contrast, LRP1 dysfunction in macrophages promotes macrophage inflammation and cholesterol accumulation to accelerate atherosclerosis (14, 34, 35). However, the deletion of LRP1 in macrophages has also been shown to accelerate atherosclerosis regression (36). Hence, the influence of LRP1 inactivation on atherosclerosis is dependent on the stage of the atherogenic process as well as the cell type where LRP1 dysfunction occurs. While LRP1 dysfunction in vascular smooth muscle cells and hepatocytes generally exacerbates diet-induced atherosclerosis (29, 37), LRP1 expressed in macrophages has a dual role in atherosclerosis, contributing to the disease by promoting lipid accumulation and foam cell formation but also impeding atherosclerotic plaque progression by accelerating cellular cholesterol export and reducing cell death (14, 34, 36, 38).

The current study showed that the cardiometabolic phenotype of mice with LRP1 proximal dileucine motif mutation is dramatically different from that observed with global or tissue-specific LRP1 gene inactivation. First, in contrast to the embryonic lethality of global Lrp1^−/− mice, Lrp1^AA mice developed normally and thrived postnatally when maintained on a normal low-fat chow diet. Hence, the function of the LRP1 proximal dileucine motif is not necessary for embryonic and neonatal development. Second, while Lrp1^−/−Ldlr^−/− adipose tissues displayed elevated inflammation and accelerated atherosclerosis when transplanted into Ldlr^−/− mice, adipose tissues in Western diet-fed Lrp1^AALdlr^−/− mice were less inflamed than those observed in Lrp1^LLLdlr^−/− mice. Moreover, mice with adipose tissue and vascular cell expression of Lrp1^AALdlr^−/− genotype did not show elevated atherosclerotic response provided that their bone marrow cells expressed the Lrp1^LLLdlr^−/− genotype (Fig. 7, G and H). Thus, the function of the LRP1 proximal dileucine motif is dispensable in adipocytes and smooth muscle cells for atherosclerosis protection. Third, the Ldlr^AALdlr^−/− mice were also less sensitive to diet-induced hepatosteatosis, instead of the increased fatty liver disease observed in mice with liver LRP1 inactivation. The mechanism by which LRP1^AA protects against metabolic dysfunction-associated steatotic liver disease was due to increased expression of PPAR-responsive genes. It is interesting to note that endothelial-specific LRP1 inactivation or LRP1 proximal dileucine mutation has been shown to improve glucose sensitivity and suppress diet-induced obesity (17). However, the mechanism underlying the metabolic benefits of endothelial LRP1 dileucine mutation appeared to differ from that observed with LRP1 dileucine mutation in the liver, adipose tissues, and leukocytes. In the endothelium, LRP1 dileucine motif mutation was reported to reduce PPAR transport to the nucleus and lower pyruvate dehydrogenase-4 expression. In contrast, our current study showed that LRP1 proximal dileucine mutation enhanced PPAR-responsive gene transcription in the liver by increasing the steady state levels of PPARβ/δ and PPARγ proteins. In view of previous studies showing that LRP1 interacts with PPARγ and prevents its degradation in degenerating intervertebral discs to suppress endoplasmic reticulum stress (39), our data reported here suggested that the LRP1 proximal dileucine motif mutation is a gain-of-function mutation that further stabilized PPARγ levels in Western diet-fed mouse livers to prevent its degradation and promote fatty acid oxidation. Furthermore, our data revealed that, in addition to PPARγ, LRP1^AA also increased PPARβ/δ protein levels, thus indicating that LRP1^AA may also interact with PPARβ/δ to prevent its degradation. The increase in both PPARβ/δ and PPARγ protein levels in the livers of Western diet-fed Lrp1^AALdlr^−/− mice likely accounts for the increased PPAR transcriptional activities in their hepatocytes.

Increased expression of PPAR-targeted genes was also observed in Lrp1^AALdlr^−/− leukocytes in response to oxLDL activation (Fig. 5, A and B). The Lrp1^AALdlr^−/− leukocytes were also more sensitive to LPS-induced expression of inflammatory cytokines (Fig. 5, C and D). These observations are consistent with the findings that enhanced PPAR transcriptional activities repolarize M1-like to M2-like macrophages to dampen inflammatory response (40). Additionally, this LRP1-PPAR signaling cascade in macrophages is also essential to limit cholesterol accumulation and foam cell formation (14). While mutation in the NPxY motif of LRP1 abolished this LRP1-PPAR signaling cascade and led to increased atherosclerosis (14), results of the current study showed that mutation in the proximal dileucine motif of LRP1 has the opposite effect of PPAR signal enhancement. Nevertheless, mice with LRP1 proximal dileucine motif mutation in macrophages also displayed accelerated atherosclerosis compared to mice in which their macrophages expressed normal LRP1 with intact proximal dileucine motif. The increased atherogenesis with elevated lipid deposition in the vessel wall of Lrp1^AALdlr^−/− mice, as well as in Lrp1^AALdlr^−/− bone marrow recipient mice, were apparently due to mitochondrial dysfunction with compromised oxidative phosphorylation and fatty acid oxidation capabilities. The importance of macrophage fatty acid oxidation in inhibiting atherosclerosis progression has been established previously (41).

It is noteworthy that high PPAR activities and the resulting repolarization of macrophages to the alternatively activated M2-like state were generally associated with atheroprotection. Hence, the increased atherosclerosis observed in Lrp1^AALdlr^−/− mice despite their increased macrophage PPAR activities in response to acute oxLDL or LPS challenge was an unexpected result. This discrepancy may be explained by the observation of mitochondrial dysfunction with impaired fatty acid oxidation capacity in Lrp1^AALdlr^−/− macrophages. Previous studies have shown that elevated mitochondrial fission promotes de novo lipogenesis and suppresses fatty acid oxidation (42). Moreover, mitochondrial dysfunction also prevents macrophage repolarization to the alternatively activated anti-inflammatory phenotype (24). Thus, the elevated mitochondrial fission and impaired fatty acid oxidation capacity observed in Lrp1^AALdlr^−/− macrophages after cholesterol loading likely led to the increased atherosclerosis observed in animals with macrophages harboring the LRP1 proximal dileucine mutation.

The role of LRP1 in mitochondrial homeostasis maintenance has been noted previously in several tissues. Interestingly, LRP1 enhances mitochondrial integrity, and functions appeared to be cell type-specific. In particular, LRP1 is essential for normal mitochondrial dynamics and bioenergetics in Schwann cells where it limits mitochondrial fission to maintain mitochondrial membrane potentials and reduces glycolysis associated with peripheral sensitization (43). Conversely, LRP1 deficiency in hepatocytes promotes mitochondrial fission and inhibits mitochondrial fusion to increase mitochondrial fragmentation and dysfunction (44). However, and in contrast to these studies, LRP1 deficiency has also been shown to reduce mitochondrial fragmentation and preserve mitochondrial functions following brain injury-induced oxidative stress (45). These cell-specific differences by which LRP1 protects or contributes to mitochondrial integrity and homeostasis may be due to its interaction with specific adaptor proteins expressed in the various cell types (28). For example, we have shown previously that LRP1 interacts with phosphatidylinositol 4-phosphate 5-kinase like protein-1 to maintain mitochondrial integrity (44). In the current study, we showed that LRP1 with proximal dileucine motif mutation reduced its capacity to be associated with mitochondria, resulting in lower levels of mitochondrial fission proteins in cholesterol-loaded Lrp1^AALdlr^−/− macrophages compared to Lrp1^LLLdlr^−/− macrophages. It is of note that LRP1 is predominantly present on plasma membranes, endosomes and intracellular organelles of the biosynthetic pathway. Hence, its association with mitochondria may be due to its presence in mitochondria-associated membranes instead of a direct interaction between LRP1 and mitochondria. Since mitochondria-associated endoplasmic reticulum membranes have been shown to play an important role in balancing mitochondrial fission and fusion (46), it is possible that the association of LRP1 with mitochondria shown here was due to its presence in mitochondria-associated endoplasmic reticulum membranes. Additional future experiments are necessary to clarify this issue. Regardless, the data revealed that the proximal dileucine motif in LRP1 is necessary to limit mitochondrial fission events in cholesterol-loaded macrophages. The impairment of LRP1-mitochondria association leads to macrophage dysfunction that promotes atherosclerosis.

In summary, the data presented in this paper revealed the cell-specific role of the LRP1 dileucine motif in modulating cardiometabolic diseases in response to Western diet feeding. Whereas LRP1 dileucine motif mutation reduced PPAR transcriptional activities in endothelial cells to prevent pyruvate dehydrogenase kinase-4-mediated glucose intolerance (17), mutations of LRP1 dileucine motif also confers metabolic benefits by limiting adipocyte hypertrophy in adipose tissues as well as reducing hepatosteatosis in the liver, in a mechanism related to elevated expression of PPAR-responsive genes. In contrast, LRP1 dileucine motif mutation in macrophages accelerates diet-induced atherosclerosis by compromising respiration and fatty acid oxidation. The Lrp1^AA mice provide an additional model that illustrates the cell-specific role by which LRP1 influences cell function and metabolic diseases.

Experimental procedures

Antibodies and primers

All antibodies and oligonucleotide primers used in this study were obtained from commercial sources as listed in Tables S1 and S2 in the Supporting Information document.

Animal model

The proximal dileucine motif DVGGLL in LRP1 was mutated to DVGGAA by CRISPA-Cas9 (clustered regularly interspaced palindromic repeats-associated protein 9)-mediated homologous recombination. Cas9, donor oligos, and guide RNA (gRNA) were injected into C57BL/6J mouse embryos to generate mutant mice with the Lrp1^LL → Lrp1^AA mutation in the endogenous Lrp1 gene (Fig. 1). Both wild type Lrp1^LL and mutant Lrp1^AA mice were mated with Ldlr^−/− mice (Jackson Laboratories, Bar Harbor, ME) to yield Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice for experiments. The mice were fed a standard rodent chow diet (Purina PicoLab Rodent Irradiated Diet 5053; Cincinnati Lab Supply, Cincinnati, OH) and maintained as breeding colonies on a 12-h light/dark cycle in a pathogen-free environment. Food and water were available ad libitum. All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee at the University of Cincinnati, in accordance with National Institutes of Health guidelines.

Diet studies

Age-matched male Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice were selected randomly for experiments. The mice were fed a Western diet containing 42% fat and 0.2% cholesterol (TD 88137; Teklad Custom Diets) for 16 weeks. Body weights were obtained biweekly and body composition was analyzed using ¹H magnetic resonance spectroscopy (EchoMRI-100, Echo-Medical Systems). Plasma obtained from animals after an overnight fast was used to quantify plasma cholesterol and triglyceride levels using Infinity colorimetric assay kits (TR22421 and TR13421, Thermo Fisher Scientific). At the end of the experimental period, the animals were euthanized by overdose isoflurane inhalation, and tissues were removed for additional analysis.

Lipoprotein characterization

Plasma obtained from Western diet-fed Lrp1^LLLdlr^−/− and Lrp1^AALdlr^−/− mice after an overnight fast was subjected to fast-performance liquid chromatography separation. Individual fractions were analyzed by triglyceride and cholesterol contents and compared with normal mouse plasma to identify VLDL, IDL/LDL, and HDL fractions. The plasma samples were also subjected to SDS-polyacrylamide gel electrophoresis and then transferred onto polyvinylidene fluoride (PVDF) membranes (162-0177, BioRad) for Western blot analysis with antibodies against apoB, apoE, and apoA-I.

VLDL secretion

Fasted mice were injected intraperitoneally with 10% Poloxamer 407 solution in saline at a volume of 10 μl/g body weight to inhibit lipolysis. Blood samples were collected before and hourly for 4 h after injection for triglyceride measurements.

Ex vivo whole-blood acute inflammatory response assay

Acute blood cell response to inflammatory activation was assessed by ex vivo assay as described previously (47). Blood collected from fasted mice was mixed with an equal volume of RPMI medium containing LPS to a final concentration of 100 ng/ml. The samples were incubated at 37 °C for 4 h and then centrifuged to separate the plasma/medium supernatant from blood cells. Cytokines secreted into the plasma/medium were quantified by IL-6 and TNFα ELISA kits (Thermo Fisher Scientific) according to manufacturer’s instructions.

Tissue analysis

Liver and epididymal adipose tissues were fixed in 10% formalin and then embedded in paraffin for preparation of 5-μm sections. The sections were stained with hematoxylin and eosin at 3 different levels. Liver steatosis was assessed based on appearance of intracellular vacuoles in liver sections. Epididymal adipose tissue sections were used to determine adipocyte size and crown-like structures. The total number of crown-like structures per total number of adipocytes was used to quantify adipose tissue macrophage content as described (15). All images were obtained using an Olympus BX61 microscope.

Gene expression analysis

Gene expression in various tissues was analyzed by quantitative real-time PCR. Cellular RNA was extracted using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. The RNA was reverse transcribed using qScript cDNA Synthesis Kit (QuantaBio). Quantitative real-time PCR was performed on a StepOnePlus Fast Thermocycler using Fast SYBR Green Master Mix (Applied Biosystems) with primer sequences as shown in the Supplement (Table S1). The levels of target mRNA were normalized to cyclophilin mRNA levels using the ΔΔCT analysis method.

Western blot

Proteins were extracted from cell culture lysates and tissue homogenates using ice-cold radioimmune precipitation assay buffer containing 50 μM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% sodium deoxycholate, 1% igepal, 0.1% SDS, 1 mM EDTA, 1× phosphatase inhibitor cocktail 2 (Sigma-Aldrich), 1× phosphatase inhibitor cocktail 3 (Sigma-Aldrich), and 1× complete protease inhibitor cocktail (Roche Applied Science). After the removal of cell nuclei and debris by centrifugation at 20,000g for 5 min at 4 °C, proteins in the supernatants were resolved by SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (BioRad). The membranes were blocked in 5% milk in PBS for 1 h at room temperature and then incubated overnight at 4 °C with primary antibodies followed by incubation with HRP-conjugated secondary antibodies for 1 h at room temperature. Immunoreactive bands were visualized by chemiluminescence using Pierce Enhanced Chemiluminescent Western Blotting Substrate (32,106, Life Technologies), digitized, and then quantified using ImageJ software (NIH).

Atherosclerosis assessment

Atherosclerosis was characterized by en face analysis of the whole aorta and cryosections of the aortic roots as described previously (48). Briefly, anesthetized mice were perfused with phosphate-buffered saline, followed by 10% formalin solution. The heart and the entire aorta to the iliac bifurcation were removed. The aortas from the bifurcation of the subclavian and carotid arteries to the iliac bifurcation were opened longitudinally for staining with Oil Red O to identify fatty lesions in the whole aorta. The hearts were fixed in 10% formaldehyde and embedded in Scigen Tissue-Plus OCT compound (Fisher Scientific). Five-μm cryosections were made throughout the aortic sinus starting at the valve nubs and the appearance of the coronary artery branch until the valve separates at the base of the heart. The cryosections were mounted on Premium Superfrost Plus Microscope Slides (VWR International), and serial sections were stained with Oil Red O and counterstained with hematoxylin to measure neutral lipid accumulation. All images were taken using an Olympus BX61 microscope and were subsequently quantified using ImageJ software (NIH).

Bone marrow transplant

Bone marrow recipient mice were fed a standard rodent chow diet and water containing 2 mg/ml neomycin (Sigma-Aldrich) for 1 week prior to irradiation with the Xen-X closed cabinet X-ray irradiator at 2 dosages of 450 cGy 4 h apart to ablate all endogenous bone marrow cells. The mice were then injected with 1 to 5 × 10⁶ bone marrow cells from donor mice via the orbital venous sinus under anesthesia within 24 h of irradiation and continued to be fed a standard rodent chow diet and water containing 2 mg/ml neomycin (Sigma-Aldrich) for 4 additional weeks after irradiation. The mice were then fed the Western diet after a 1-week recovery period on normal, neomycin-free water.

Bone marrow-derived macrophage isolation and characterization

Bone marrow cells were obtained from the tibias and femurs by flushing with Hank’s Buffered Salt Solution, and subjected to red blood cell lysis (eBioscience, Thermo Fisher Scientific) for 10 min at room temperature. The samples were then passed through a 40-μm cell strainer and resuspended in RPMI medium. The isolated bone marrow cells were cultured in RPMI with 10% fetal bovine serum in the presence of 20% macrophage colony stimulating factor-containing L929 medium for 7 days to induce differentiation into macrophages for experiments. Macrophage uptake of aggregated LDL (aggLDL) was performed by overnight incubation of macrophages with increasing concentrations of DiI-labeled aggLDL, prepared by vigorous vortexing of DiI-labeled LDL, followed by determination of cell associated DiI after extensive washing as described (49).

Cellular bioenergetics

Cellular bioenergetics in cholesterol-loaded BMDM were performed as described (50, 51). Briefly, BMDM were plated on XF cell culture plate and incubated with or without 50 μg/ml acLDL for 24 h followed by oxygen consumption rate (OCR) and proton efflux rate (PER) measurements before and after the sequential injections of 1.5 mM oligomycin, 2 mM FCCP, and 1 mM of rotenone/antimycin A. In parallel, etomoxir was added at 3 μM to inhibit fatty acid oxidation followed by OCR and PER measurements. All measures were normalized by cell number.

Statistical analyses

All data are expressed as mean ± SD. Statistical analysis was performed using GraphPad Prism version 5.0 software. Normality was assessed by the Shapiro-Wilk test. Data with normal distribution and with equal variance based on Levene analysis were then subjected to multiple group comparisons by one-way ANOVA with Student-Newman-Keuls post hoc analysis or Student’s t test when evaluating differences between two groups. Data with nonparametric distribution were analyzed by the Kruskal-Wallis test, followed by a Dunn test evaluation of differences for multiple group comparisons or by the Mann-Whitney test for comparison between 2 groups. Differences at p < 0.05 were considered statistically significant.

Data availability

The data supporting this study are available in the article and are available from the corresponding author (huidy@ucmail.uc.edu) upon request.

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by George M. Carman