Bmal1’s role in controlling cholesterol balance in immune cells

We previously demonstrated that global and hepatic-specific Bmal1 deficiency increases atherosclerosis inmice (). Here, we used multiplemouse models to address the role of Mφ-specific Bmal1 in atherosclerosis. First, we transplanted bone marrow cells fromormice into lethally irradiatedmice. After 4 weeks of Western diet (WD) feeding, greater atherosclerotic lesion areas, en face Oil Red O staining in the aortas, and collagen and Mφ content were observed in the aortic roots ofmice transplanted with bone marrow cells frommice rather thanmice (). No significant differences were observed in body weight, plasma cholesterol, or triglycerides inmice receiving cells fromorcontrol mice (; supplemental material available online with this article;). Second, we fed myeloid-specific Bmal1-deficient() mice a chow diet (). Visualization of aortic arches and aortic Oil Red O staining revealed age-dependent increases in atherosclerosis inmice, in contrast tomice (, A and B). En face analysis of aortas at 14 months revealed elevated plaque formation in the aortic root and abdominal aorta inmice (). Furthermore, the lesions at the cardiac/aortic junction contained higher necrotic core, collagen, and Mφ content inmice thanmice. Total plasma, HDL and non-HDL triglyceride, and cholesterol levels did not differ betweenandmice, as determined after separation by precipitation (, C and D) and by fast protein liquid chromatography (FPLC) (, E and F). Third, we studied the effect of WD feeding on atherosclerosis inmice ().mice showed greater atherosclerosis, lipid accumulation, collagen content, and Mφ infiltration thanmice (). Again, no significant differences were observed in plasma lipids in these groups (). These studies indicate that Mφ-specific Bmal1 deficiency mademice susceptible to atherosclerosis without affecting plasma lipid levels. Apoe −/− Apoe −/− Bmal1Apoe −/− −/− Bmal1Apoe +/+ −/− Apoe −/− Apoe −/− Bmal1Apoe −/− −/− Bmal1Apoe +/+ −/− Apoe −/− Bmal1Apoe −/− −/− Bmal1Apoe +/+ −/− M-Bmal1Apoe −/− −/− LysM-Bmal1Apoe Cre fl/fl −/− M-Bmal1Apoe −/− −/− Bmal1Apoe fl/fl −/− M-Bmal1Apoe −/− −/− M-Bmal1Apoe −/− −/− M-Bmal1Apoe fl/fl −/− M-Bmal1Apoe −/− −/− M-Bmal1Apoe fl/fl −/− M-Bmal1Apoe −/− −/− M-Bmal1Apoe −/− −/− M-Bmal1Apoe fl/fl −/− Apoe −/− ⁴⁵ Figure 1A Supplemental Figure 1 Figure 1B Supplemental Figure 2 Figure 1B Supplemental Figure 2 Supplemental Figure 2 Figure 1C Figure 1C Supplemental Figure 3 https://doi.org/10.1172/jci.insight.194304DS1↗

We previously showed that global Bmal1 deficiency increases plasma lipids and lipoproteins and may contribute to enhanced atherosclerosis (). Here, we observed that Mφ-specific Bmal1 deficiency had no effect on plasma lipids or lipoproteins, suggesting that Mφ-Bmal1 affects atherosclerosis by unknown mechanisms. To unravel how Mφ-Bmal1 may contribute to atherosclerosis, we measured cellular cholesterol levels in different mouse models (). Bmal1-deficient Mφs from different mouse models had higher levels of total, free, and esterified cholesterol levels compared with their respective controls (). These studies suggest deregulation of cholesterol homeostasis in Bmal1-deficient Mφs. Attempts were then made to understand molecular mechanisms for increased cellular cholesterol accretions in Bmal1-deficient Mφs. ⁴⁵ Figure 1, D–F Figure 1, G–I

To elucidate the mechanisms for the elevated Mφ cholesterol levels during Bmal1 deficiency, we examined 3 contributing factors: uptake of modified lipoproteins, cellular cholesterol efflux, and lysosomal cholesterol egress. First, we hypothesized that Bmal1-deficient aortas and Mφs would show enhanced oxLDL uptake, thereby contributing to atherosclerosis. Chow-fedandmice were injected with Dil-labeled oxLDL. The aortas harvested frommice showed ~2.5-fold higher assimilation of Dil-labeled oxLDL than controls (). Similarly,aortas showed increased oxLDL uptake compared with controlmice (). Greater uptake of Dil-oxLDL was also observed in the aortas ofmice transplanted withbone marrow cells rather thanbone marrow cells (). Thus, Bmal1 deficiency increased aortic oxLDL uptake. Bmal1Apoe +/+ −/− Bmal1Apoe −/− −/− Bmal1Apoe −/− −/− M-Bmal1Apoe fl/fl −/− M-Bmal1Apoe fl/fl −/− Apoe −/− Bmal1Apoe −/− −/− Bmal1Apoe +/+ −/− Figure 2A Supplemental Figure 4 Figure 2B

Next, we isolated bone marrow–derived Mφs (BMDMs) from these mice and studied the uptake of Dil-oxLDL and [H]-cholesterol–labeled acetylated LDL (acLDL). Bmal1-deficient BMDMs from all mouse models took up more Dil-oxLDL (and) and [H]-cholesterol–acLDL (and), and they contained higher amounts of lipid peroxides compared with their respective controls (and). These studies indicate that Bmal1-deficient BMDMs take up more modified lipoproteins and contain higher amounts of oxidized lipids than control BMDMs. 3 3 Figure 2C Supplemental Figure 5A Figure 2D Supplemental Figure 5B Figure 2E Supplemental Figure 5C

To identify receptors involved in the increased uptake of modified lipoproteins, we measured the mRNA and protein levels of several scavenger receptors (, and, D–F). Bmal1-deficient Mφs had higher mRNA and protein levels of Cd36 and Lox-1 than their respective controls, whereas the Sra1 mRNA and protein levels did not differ. Thus, Bmal1 deficiency specifically increases Cd36 and Lox-1 expression. Figure 2, F and G Supplemental Figure 5

Next, we assessed whether Bmal1 might regulate Cd36 and Lox-1 in Mφs. WT Mφs were treated with siControl or siBmal1. siBmal1 significantly reduced Bmal1 mRNA and protein levels and increased Cd36 mRNA levels, but it had no effect on Lox-1 mRNA levels (and). Moreover, siBmal1 significantly increased oxLDL uptake, and this uptake was inhibited in siCd36-treated cells (). These findings suggest that Bmal1 might regulate Cd36, thereby controlling the uptake of modified lipoproteins. Figure 2H Supplemental Figure 5G Figure 2I

Bmal1 is a transcriptional enhancer. Because knockdown (KD) of Bmal1 increased Cd36 expression, we hypothesized that Bmal1 might decrease the expression of a repressor of Cd36. Bmal1 is known to increase the expression of Rev-erbα, a transcriptional repressor (,–). Indeed, we observed diminished Rev-erbα mRNA levels inMφs and in siBmal1-treated WT Mφs (). Furthermore, siRev-erbα increased Cd36 expression (), thus indicating that Rev-erbα is a repressor of Cd36. ChIP analysis indicates that Rev-erbα interacted with the ROR element in the Cd36 promoter, and this binding was significantly diminished in Mφs obtained frommice (). In WT Mφs, Rev-erbα occupancy on the Cd36 promoter was significantly diminished in cells treated with siBmal1 but to a lesser extent than that observed in siRev-erbα–treated Mφs (). Furthermore, we obtained plasmids for expression of luciferase (pGL4.11-Cd36 promoter) under control of the Cd36 promoter (). siBmal1 and siRev-erbα significantly increased luciferase activity (). These studies indicate that Rev-erbα reduced Cd36 expression. ^{24 46 48} Figure 2J Figure 2K Figure 2L Figure 2M ⁴⁹ Figure 2N Bmal1 −/− M-Bmal1Apoe −/− −/−

Next, we extended these experiments to human monocyte-derived Mφs. Normal peripheral blood mononuclear cells (PBMCs) were differentiated into Mφs and used to study the effects of KD ofandonexpression. KD ofandsignificantly increasedmRNA and protein expression (and), thus indicating thatandregulate CD36 expression in human Mφs. Next, we studied the role of BMAL1 and REV-ERBα in cyclic expression of CD36. In siControl-treated Mφs, CD36 expression showed cyclic expression with peaks at 4 and 24 hours after serum synchronization (). KD of BMAL1 abolished these cyclic changes, whereas KD of REV-ERBα had no effect on the cyclic changes but increased the amounts of mRNA at the peak levels. These studies indicate that Bmal1 determines the cyclic expression of CD36, whereas REV-ERBα determines the extent of expression. Overall, these findings indicate that Bmal1 increased Rev-erbα expression, which in turn decreased CD36 expression and the uptake of modified lipoproteins (). BMAL1 REV-ERBα CD36 BMAL1 REV-ERBα CD36 BMAL1 REV-ERBα Figure 2O Supplemental Figure 5H Figure 2P Figure 2Q

The above studies show that Mφ specific Bmal1 deficiency increased uptake of modified lipoproteins, and Mφs assimilate more cholesterol and lipid peroxides, potentially by increasing Cd36 expression. Cells increase cholesterol efflux and RCT to decrease cellular free cholesterol content. Therefore, we determined whether Mφ-specific Bmal1 might play a role in regulating cholesterol efflux from Mφs and in vivo RCT in multiple mouse models. To study RCT, we injectedH-cholesterol–labeled Mφs from control and Bmal1-deficient mice into WT ormice (). In all cases, Bmal1-deficient Mφs, compared with control Mφs, showed defects in RCT, as evidenced by diminished amounts of cholesterol in the plasma, liver, and feces (). Furthermore, BMDMs from WD-fed Bmal1-deficient mouse models showed defects in cholesterol efflux to extracellular apoA1 and HDL (and). These in vivo and cell culture studies indicate that Bmal1 deficiency decreased Mφ cholesterol efflux capacity. 3 Apoe −/− Figure 3A Figure 3, A and B Figure 3C Supplemental Figure 6A

Upregulated cellular Abca1/Abcg1 expression augments cholesterol efflux and RCT, thereby decreasing cellular free cholesterol content. Acat1 increases conversion of free cholesterol to esterified cholesterol for storage in cytosolic lipid droplets (). Therefore, we measured changes in Abca1, Abcg1, and Acat1 mRNA and protein levels, and we determined the roles of Mφ-specific Bmal1 in their regulation. Protein and mRNA levels of Abca1 and Abcg1 were significantly reduced (, and, B and C) in all mouse models. Srb1 levels were diminished in global Bmal1-KO mice but remained unchanged in Mφ-specific Bmal1-deficient Mφs (, and, D and E). The expression of Acat1, an enzyme involved in cholesterol esterification, did not change (). Therefore, defects in cholesterol efflux in Bmal1-deficient Mφs might be secondary to lower expression of Abca1 and Abcg1 transporters. ¹⁶ Figure 3, D and E Supplemental Figure 6 Figure 3, D and E Supplemental Figure 6 Figure 3, D and E

Next, we used KD approaches to address the role of Bmal1 in the regulation of Abca1 and Abcg1. siBmal1 significantly decreased Abca1 and Abcg1 expression, without affecting Srb1 expression, and additionally decreased cholesterol efflux (). To determine whether overexpression of Bmal1 might also affect Abca1/Abcg1 expression and cholesterol efflux, we transduced the J774A.1 mouse Mφ cell line with adenoviruses for expression of Bmal1. Cells transduced with Adv-Bmal1 showed elevated expression of Abca1 and Abcg1 as well as greater cholesterol efflux to apoA1 and HDL than observed in cells transduced with Adv control (). These studies indicate that Bmal1 KD decreases, whereas Bmal1 overexpression increases, Abca1/Abcg1 expression and cholesterol efflux. Figure 3, F and G Figure 3, H and I

We subsequently studied cyclic expression of Abca1 and Abcg1 in Mφs isolated from control and Bmal1-deficient mice. Serum shock studies indicate robust temporal changes in the expression of Abca1 and Abcg1, with major peaks at 20 and 40 hours in WT Mφs. However, these peaks were absent in Bmal1-deficient Mφs (). Furthermore, similar reductions in Abca1 and Abcg1 expression were observed in WT Mφs treated with siBmal1 (). These studies indicate Bmal1's involvement in the temporal regulation of Abca1/Abcg1. Figure 3J Figure 3K

Next, we extended these experiments to human PBMCs to study the effects of KD ofonandgene expression and cholesterol efflux. KD ofsignificantly decreasedandexpression, as well as cholesterol efflux (), thus indicating that Bmal1 also regulates cholesterol efflux in human Mφs. BMAL1 ABCA1 ABCG1 BMAL1 ABCA1 ABCG1 Figure 3, L and M

We previously demonstrated that Clock modulates Abca1 expression by regulating the Usf2 repressor (). Here, we determined how Bmal1 regulates Abca1 and Abcg1, by examining expression changes in several repressors known to regulate Abca1 and Abcg1. Quantification of various mRNA and proteins inMφs indicated significantly elevated (2- to 4-fold) Znf202 levels (, and). Furthermore, siBmal1 increased Znf202 expression (). Thus, Znf202 might be regulated by Bmal1. ⁴⁴ Figure 4, A and B Supplemental Figure 7 Figure 4C Bmal1 −/−

ZNF202 is a repressor () that controls the tissue-specific expression ofand(). Both human and mouseandpromoters contain GnT motifs recognized by ZNF202 (,,). Znf202 expression is downregulated during monocyte differentiation and foam cell formation. However, whether Znf202 shows diurnal rhythms, and might be involved in the diurnal regulation of Abca1 and Abcg1 expression, is unknown. Therefore, we assessed Znf202's potential involvement in regulating Abca1/Abcg1 and cholesterol efflux. siZnf202 increased Abca1/Abcg1 expression and cholesterol efflux, thus indicating that Znf202 is a repressor (). To further determine whether Znf202 might act at the promoter level, we transfected WT BMDMs with a reporter construct for expression of luciferase under control of the Abca1 promoter, along with various siRNAs (). siBmal1 decreased Abca1 promoter luciferase activity, whereas siZnf202 increased promoter activity by 2.4-fold more than in the siControl group (). The role of Znf202 in regulating Abca1 and Abgc1 was further studied with ChIP (). Znf202 binding on the Abca1/Abcg1 promoters was significantly enhanced after KD of Bmal1. Therefore, Znf202 represses Abca1 expression at the transcriptional level, and Bmal1 regulates Abca1/Abcg1 by regulating Znf202 (). ^{50 51 50 52 53} Figure 4, D and E Figure 4F Figure 4F Figure 4G Figure 4H ABCA1 ABCG1 ABCA1 ABCG1

Next, we extended these studies to human monocyte-derived Mφs. KD of BMAL1 in PBMCs decreased cholesterol efflux to apoA1 and HDL (), decreased the expression of ABCA1 and ABCG1, and increased the expression of ZNF202 (, and). We then studied the binding of ZNF202 to the ABCA1 and ABCG1 promoters in CONTROL and siBMAL1-treated Mφs. siBMAL1 increased ZNF202 binding to these promoters (). Our findings indicate that ZNF202 binding to the ABCA1/G1 promoters was increased in Bmal1 deficiency, thus decreasing the expression of these transporters in human Mφs. Figure 5A Figure 5, B and C Supplemental Figure 8 Figure 5D

After having demonstrated that elevated uptake of modified lipoproteins and decreased cholesterol efflux in Bmal1-deficient Mφs contributes to increasing cholesterol assimilation, we next sought to identify the subcellular organelles assimilating cholesterol. We subjected Mφs to differential ultracentrifugation (,) and determined their purity by detecting specific markers (). Total, free, and esterified cholesterol levels were significantly higher in all subcellular organelles inMφs thanMφs (); however, the highest accretions were in lysosomes and endosomes. Therefore, we sought to understand the mechanisms underlying cholesterol accumulation in lysosomes. Mφs were incubated with [H]-cholesterol–labeled acLDL, and lysosomal cholesterol levels were quantified at various times. Lysosomes ofMφs showed greater cholesterol accumulation over time (). We hypothesized that this accumulation might have been due to defects in cholesterol egress from lysosomes. To examine this possibility, we pulse-labeled Mφs with [H]-cholesterol–acLDL for 4 hours, washed them, and incubated them in serum-free medium for chase experiments. At various times, lysosomes were isolated, and cholesterol egress from lysosomes was quantified (). Lysosomes fromMφs showed less cholesterol egress than controls (). We extended these studies to Mφs isolated from other mouse models (). In all Bmal1-deficient Mφs, lysosomal [H]-cholesterol accretion was elevated (), and egress was diminished (). Because cholesterol egress from lysosomes in Mφs depends on Npc1 and Npc2 (,,), we measured Npc1 and Npc2 mRNA and protein levels in Mφs isolated from various mouse models (, and). All Bmal1-deficient Mφs had diminished Npc1/Npc2 mRNA and protein levels. These results indicate that Bmal1 deficiency decreases the expression of Npc1 and Npc2 in Mφs. Therefore, this decreased expression might contribute to increased cholesterol accumulation in lysosomes in Bmal1-deficient Mφs. ^{54 55} Supplemental Figure 9 Figure 6A Figure 6B Figure 6C Figure 6C Figure 6, D and E Figure 6D Figure 6E ^{10 56 57} Figure 6, F and G Supplemental Figure 10A Bmal1Apoe −/− −/− Bmal1Apoe +/+ −/− Bmal1Apoe −/− −/− Bmal1Apoe −/− −/− 3 3 3

Next, we extended these studies to human monocyte-derived Mφs. PBMCs were treated with siBMAL1, siNPC1, or siNPC2 and incubated with [H]-cholesterol–acLDL for 4 hours; subsequently, the radioactivity in lysosomes was quantified. The various siRNAs specifically decreased their targets' mRNA and protein levels (and). In addition, siBMAL1 decreased the expression ofand. Individual KD of BMAL1, NPC1, or NPC2 significantly decreased cholesterol egress () and increased lysosomal cholesterol accretion (). We then asked whetherandmRNA might show cyclic changes in response to serum synchronization and whether Bmal1 might play roles in their temporal expression. To examine this possibility, we subjected siCONTROL- and siBMAL1-treated PBMCs to serum synchronization. Bmal1 showed cyclic expression, with 2 peaks at 8–16 and 32–40 hours in siControl-treated cells (). Bmal1 expression was not observed in siBMAL1-treated human PBMCs. Both NPC1 and NPC2 showed similar peak expressions at 8–12 and 32–36 hours (). In siBMAL1-treated cells, the expression of NPC1 and NPC2 was significantly diminished and did not exhibit any significant temporal changes. Therefore, decreases in BMAL1 lowered NPC1/NPC2 temporal expression in PBMCs, most likely leading to defects in cholesterol egress from lysosomes. 3 Figure 6H Supplemental Figure 10B Figure 6I Figure 6J Figure 6K Figure 6K NPC1 NPC2 NPC1 NPC2

We subsequently extended these studies to the mouse Mφ J774A.1 cell line. Npc1 and Npc2 showed cyclic expression, whereas these changes were absent in siBmal1-treated J774 cells (). KD of Bmal1 decreased the expression of Npc1/Npc2 (and) and cholesterol egress () as well as increased cholesterol levels in lysosomes (). These studies highlight the importance of Bmal1 in the control of basal and cyclic expression of Npc1 and Npc2 along with its role in lysosomal cholesterol trafficking. Subsequently, we increased the expression of Bmal1 by using Adv-Bmal1. Overexpression of Bmal1 increased Npc1/Npc2 expression (and) and cholesterol egress and decreased lysosomal cholesterol (). These findings indicate that Bmal1 regulates Npc1/Npc2 and lysosomal cholesterol trafficking. Figure 7A Figure 7B Supplemental Figure 11A Figure 7C Figure 7C Figure 7D Supplemental Figure 11B Figure 7E

Because the Npc1 and Npc2 promoters contain E-boxes, we quantified E-box occupancy by Bmal1 and Clock in control and Bmal1-deficient Mφs (). Bmal1 interacted with the E-boxes in WT Mφs, whereas this binding was not observed in Bmal1-deficient Mφs (). In agreement with this finding, no E-box sequences were amplified after anti-Bmal1 ChIP. Use of anti-Clock IgGs during ChIP indicated that Clock also binds E-boxes, and this binding decreased in the absence of Bmal1 (). These studies suggest that Bmal1 directly interacts with E-boxes in the promoters of Npc1/Npc2, thereby increasing their expression and cholesterol egress from lysosomes. Figure 7, F and G Figure 7, F and G Figure 7, F and G

We have used both male and female mice in this project.

[H]-Cholesterol (1 Ci/mL, 9.25 MBq, #NET139250UC) was purchased from NEN Life Science Products. Chemicals and solvents were from various vendors (). 3 Supplemental Table 1

All mice were on a C57Bl6J background.mice were bred to obtainandmice. Various Mφ-specific Bmal1-deficient mouse strains were generated by crossing Bmal1 floxed () mice with C57BL/6J ormice expressing the Cre recombinase transgene under control of the lysozyme M promoter (B6.129-Lyz2/J, Jackson Laboratory).mice were fed a WD containing protein, carbohydrates, fat, and cholesterol at 17%, 48.5%, 21.2%, and 0.2% by weight, respectively (TD 88137, Harlan Teklad). Bmal1Apoe +/– −/− Bmal1Apoe −/− −/− Bmal1Apoe +/+ −/− Bmal1 fl/fl Apoe −/− M-Bmal1Apoeand Bmal1Apoe −/− −/− fl/fl −/− tml(cre)Ifo

BMDMs were obtained through isolation of bone marrow cells from female and male mice and through incubation in RPMI 1640 with 10% FBS and 1% penicillin-streptomycin supplemented with 25% L-cell–conditioned medium in tissue culture plates. Fresh medium was added on days 3 and 5. After 7 days, Mφs were fully differentiated, and the culture medium was changed to 1× DMEM with 1% penicillin-streptomycin supplemented with 5% L-cell–conditioned medium. In some experiments, after 7 days, cells were subjected to serum shock, cholesterol efflux assays, or Dil-oxLDL uptake assays. Cells were transfected with siRNAs or transduced with adenoviruses for KD and overexpression as previously described (,,,,). ^{27 44 45 61 62}

Human PBMCs were differentiated into Mφs by culturing in RPMI-1640 supplemented with 10% human serum and 25 ng/mL recombinant human Mφ colony-stimulating factor (M-CSF, PeproTech) for 7 days. PBMCs (2.0 × 10) were plated in 6-well cell culture plates in RPMI-1640 medium with L-glutamine. After 4 hours, cells were treated with siRNAs for 48 hours, before being subjected to cholesterol efflux assays, isolation of total RNA, or detection of protein levels. 6

Cells were lysed in NP-40 buffer (25 mM Tris, pH 7.5, 300 mM NaCl, 1 mM EDTA, and 2% NP-40) with protease inhibitor cocktail for 30 min on ice. The lysates were centrifuged at 1,000for 5 minutes. The supernatant contained the cytosolic and membrane fractions. The pellets were suspended in NP-40 lysis buffer, passed 10 times through a 25 G needle and subjected to differential ultracentrifugation (,) and used to quantify cholesterol levels. We verified the following subcellular markers in each fraction: calnexin (ER marker), ferritin (mitochondrial marker), ERK1/2 (endosome marker), LAMP1 (lysosome marker), NPC1 (lysosome marker), Na,K-ATPase α1 (plasma membrane), and Gapdh (cytosol). In most experiments, the lysosome fraction was purified with a lysosome isolation kit (Abcam, Ab234047). g ^{44 54}

To study cholesterol accretion, we incubated Mφs in triplicates with 10 μCi/mL [H]-cholesterol–acLDL (50 μg/mL). At different times, Mφs were centrifuged (21,500, 20 min, 4°C) and washed, and homogenates were subjected to cell fractionation to isolate organelles and were counted (,). For egress studies, cells were pulse labeled with 10 μCi/mL [H]-cholesteryl–acLDL (50 μg/mL) for 4 hours. The amounts of cholesterol after 4 hours were set at 100% to calculate time-dependent egress from lysosomes during subsequent chase times. The cells were then washed twice quickly with 5 mL buffer A (150 mM NaCl and 50 mM Tris-chloride, pH 7.4) containing 0.25% BSA, refed with 5 mL medium (DMEM 10% LDS, 1% penicillin-streptomycin, and 20 mM HEPES, pH 7.4), and placed in a COincubator for the indicated chase times. At each time point, cells were collected, and lysosome fractions were isolated using kits. Lipids were extracted and separated on TLC plates, and cholesterol bands were quantified. Amounts in lysosomes were used to calculate accretions and egress. 3 3 g ^{44 54} 2

For gene expression studies, cells were placed in 10% DMEM plus 25% L-cell–conditioned medium for 1 week. Total RNA was extracted and analyzed with quantitative PCR (qPCR) (,). For cholesterol efflux assays, BMDMs were labeled with [H]-cholesterol (5.0 μCi/mL) with acLDL (50 μg/mL) for 24 hours, washed with PBS, and incubated in DMEM containing 0.2% BSA for 1 hour and subsequently in the same medium with or without apoAI (15 μg/mL) or HDL (50 μg/mL) for 8 hours. Radioactivity in the medium and total cell-associated radioactivity were determined by scintillation counting. The assays were performed in quadruplicate and are presented as percentage efflux, as previously described (,). ^{44 45 44 45} 3

BMDMs from various mouse models, such asmice andmice, were cultured in 10% DMEM plus 25% L-cell–conditioned medium for 1 week. BMDMs or PBMCs were incubated for 4 hours at 37°C with 8 μg/mL Dil-labeled oxidized LDL (L34358, Thermo Fisher Scientific, Invitrogen) in α-MEM containing 2.5% lipoprotein-deficient serum (Sigma, S5394). Cells were washed with PBS, then homogenized in 50 mM Tris-HCl buffer, pH 7.4, and 1.15% KCl, and centrifuged (900, 10 min, 4°C). The supernatants were used to measure fluorescence as previously described (). For some experiments, differentiated Mφs from WT mice were transfected with siBmal1 or siControl for 48 hours, before being incubated with Dil-oxLDL (5 μg/mL) in serum-free DMEM at 37°C for 4 hours. Bmal1Apoe fl/fl −/− M-Bmal1Apoe −/− −/− g ⁴⁴

J774A.1 cells, BMDMs, or PBMCs were transfected with various siRNAs. After 48 hours, the cells were washed and starved in the same medium without FBS for 18 hours. Subsequently, cells were treated with medium containing 50% horse serum for 2 hours, and the medium was subsequently changed back to starvation medium (,,,,). Cells were harvested at 4-hour intervals for analysis. ^{27 44 45 61 62}

Plasma and Mφ cholesterol, free cholesterol, and triglycerides were measured with commercial kits. Lipoprotein profiles were determined after FPLC, as described previously (,). Pooled plasma samples from 6 mice per genotype were used for FPLC. ^{44 45}

Lipid peroxides were measured in isolated Mφs as TBARS. Mφs were homogenized in 1.15% KCl in 50 mM Tris-HCl buffer, pH 7.4, and then centrifuged (10,800, 50 min, 4°C). The supernatants were used to measure lipid peroxides with a TBARS assay kit (10009055, Cayman Chemical Company) (,,). g ^{44 45 61}

BMDMs from various mouse models were loaded with cholesterol by incubation with acLDL (50 μg protein/mL) and 5 μCi [H]-cholesterol for 24 hours. The labeled Mφs were injected i.p. into WT mice. Plasma was collected at 0, 6, 12, 24, and 48 hours, and feces were collected at 48 hours to measure tracer counts, as previously described (,). 3 ^{44 45}

The proximal aorta was collected after saline perfusion. The aortic root and ascending aorta were sectioned at a thickness of 10 μm. Alternate sections were used for Oil Red O, H&E, Masson's trichrome, and Mφ staining, as previously described (,). ^{44 45}

mice (age 8 weeks) were lethally irradiated and transplanted with bone marrow cells derived frommice andmice, as previously described (,). Apoe −/− Bmal1Apoe −/− −/− Bmal1Apoe +/+ −/− ^{27 44}

Proteins from tissues or cells were separated under nonreducing conditions, transferred to nitrocellulose membranes, and blocked for 2 hours in TBS buffer containing 0.1% Tween 20 and 5% nonfat dry milk at room temperature. The blots were washed 3 times and incubated overnight at 4°C in the same buffer containing 0.5% dry milk and primary antibodies (1:100–1:1,000 dilution), washed, and incubated with mouse horseradish peroxidase–conjugated secondary antibodies (1:1,000–1:4,000) in 1.0% skim milk for 1 hour at room temperature. Immunoreactivity was detected by chemiluminescence, as previously described (,,). ^{44 45 61}

Total RNA from Mφs, J774A.1, and human PBMCs were isolated with TRIzol. Subsequently, cDNA was synthesized with an OmniScript RT (Qiagen) kit. mRNA levels were measured with a SYBR Green kit for qPCR. The data were analyzed as arbitrary units, as previously described (,,). ^{44 45 61}

ChIP assays using polyclonal antibodies were performed to study the binding of different TFs to gene promoters. Proteins were cross-linked to DNA and sheared, and protein/DNA complexes were immunoprecipitated with specific antibodies to various TFs. DNA samples recovered after immunoprecipitation were subjected to PCR to detect coimmunoprecipitated DNA with specific primers (). Supplemental Table 1

All metabolic and imaging experiments were repeated at least twice on different days and yielded similar results. Data are presented as mean ± SD for= 6–15 animals per time point. Statistical testing was performed with 2-tailed unpairedtest, or multiple unpairedtests, with followed by Holm-Šídák method. Temporal comparisons between 2 groups were performed with 2-way ANOVA using Tukey's multiple-comparison test, as indicated in the figure legends. Three or 4 (multiple) groups were performed using 1-way ANOVA followed by Tukey's test. Differences were considered statistically significant when< 0.05. GraphPad Prism 10 was used for graphing and statistical evaluation. The circadian rhythm patterns were identified by fitting a cosine curve to the data.measures the goodness of fit. n t t P R 2

All animal experiments were approved by the IACUC of SUNY Downstate Medical Center or NYU Long Island School of Medicine.

All theare in thefile. Supporting data values Supporting Data Values