Induction of Lysosomal Biogenesis in Atherosclerotic Macrophages Can Rescue Lipid-Induced Lysosomal Dysfunction and Downstream Sequelae

Abstract

Objective

Recent reports of a proatherogenic phenotype in mice with macrophage-specific autophagy deficiency has renewed interest in the role of the autophagy-lysosomal system in atherosclerosis. Lysosomes have the unique ability to process both exogenous material including lipids and autophagy-derived cargo such as dysfunctional proteins/organelles. We aimed to understand the effects of an atherogenic lipid environment on macrophage lysosomes and evaluate novel ways to modulate this system.

Approach and Results

Utilizing a variety of complementary techniques, we show that oxidized LDL and cholesterol crystals, commonly encountered lipid species in atherosclerosis, lead to profound lysosomal dysfunction in cultured macrophages. Disruptions in lysosomal pH, proteolytic capacity, membrane integrity, and morphology are readily seen. Using flow cytometry, we find that macrophages isolated from atherosclerotic plaques also display features of lysosome dysfunction. We then investigated whether enhancing lysosomal function can be beneficial. TFEB is the only known transcription factor that is a master regulator of lysosomal biogenesis, although its role in macrophages has not been studied. Lysosomal stress induced by chloroquine or atherogenic lipids leads to TFEB nuclear translocation and activation of lysosomal and autophagy genes. TFEB overexpression in macrophages further augments this prodegradative response and rescues several deleterious effects seen with atherogenic lipid loading as evidenced by blunted lysosomal dysfunction, reduced secretion of the proinflammatory cytokine IL-1β, enhanced cholesterol efflux, and decreased polyubiquitinated protein aggregation.

Conclusions

Taken together, these data demonstrate that lysosomal function is markedly impaired in atherosclerosis and suggest that induction of a lysosomal biogenesis program in macrophages has anti-atherogenic effects.

Introduction

Every year over 30% of all deaths in the United States are attributable to cardiovascular disease stemming from myocardial infarction, stroke, or ischemic heart failure ¹. Progressive plaque formation, or atherosclerosis, is the pathogenic mediator of the vast majority of such cases and is primarily caused by the failure of the vascular system to handle increased circulating lipid. Currently, the mainstay of treatment focuses on preventing lipid accumulation through use of drugs such as statins. However, no treatment strategy directly addresses the pathogenic signaling processes that occur in atherosclerosis ².

As lipid is delivered to the vessel wall, cholesterol clearance is largely handled by macrophages owing to their immense phagocytic capacity. In cases of lipid abundance, macrophage cholesterol stores can rise significantly leading to cellular dysfunction, local inflammation in the vessel wall, and the development of instability in plaque architecture ³. It is believed that atherosclerotic progression and the resulting inflammatory response can be ameliorated if plaque macrophages handled lipid more efficiently.

The autophagy-lysosomal system is a highly evolutionarily conserved cellular process with critical roles in the degradation and recycling of long-lived/damaged intracellular material including accumulated lipids ^{4, 5}. Recent work has shown that although macrophages normally display a robust autophagic response, foam cell macrophages appear to develop a dysfunction in autophagy ⁶. This deficiency contributes to a hyperinflammatory state and abnormalities in lipid trafficking, leading to dramatic increases in atherosclerosis ^7–9. Given the essential role of lysosomes in mediating the overall degradative capacity of cells including autophagosome processing, it is possible that progressive dysfunction in the lysosomal apparatus itself underlies such deleterious effects in atherosclerotic macrophages.

Several lines of evidence raise the possibility that lysosomal dysfunction is a critical step in foam cell formation and plaque development. Genome-wide association studies have correlated polymorphisms in lysosomal acid lipase, the enzyme responsible for hydrolyzing cholesterol esters, with atherosclerotic progression ^{10, 11}. Using microscopy, Jerome and colleagues have shown that in early atherosclerosis, lipid flux is maintained such that cholesteryl esters are effectively hydrolyzed in the lysosome and shuttled to cytoplasmic stores as lipid droplets. With disease progression, this process begins to break down, leading to inefficiencies in lysosomal degradative capacity and eventual dysfunction ¹². Most recently, cholesterol crystals either derived extracellularly or via intralysosomal conversion of oxidized LDL have been shown to activate the NLRP3 inflammasome by disrupting lysosomal integrity ^{13, 14}. Thus the notion that modified lipids can have deleterious effects on lysosome function and subsequent atherosclerosis is an attractive one deserving of investigation.

In the initial part of our studies, we utilize oxidized-LDL or cholesterol crystals, modified lipids known to be poorly hydrolyzed in lysosomes and commonly encountered in the plaque, to recapitulate the foam cell phenotype in primary mouse peritoneal macrophages in vitro. We use several independent methods including lysosomal morphology, pH, and proteolytic capacity to determine the effects of such lipids on macrophage lysosome function. Subsequently, we use recently developed methods of reproducibly isolating macrophages from atherosclerotic mouse aortas to determine the extent of lysosomal dysfunction in plaque macrophages.

The recently discovered transcription factor EB (TFEB) is the only known transcription factor that drives the expression of a majority of lysosomal and autophagy genes ¹⁵. A member of the MiT/TFE helix-loop-helix subfamily, TFEB initiates a lysosomal biogenesis program, thus stimulating the overall degradative capacity of cells ¹⁶. TFEB has also recently been demonstrated to increase lysosomal lipid catabolism, lipolysis, and cellular fatty acid oxidation ^{17, 18}. Thus, this provides an exciting new way to address whether foam cell formation and downstream sequelae can be reversed by enhancing lysosomal function. However, it has never been studied in the context of macrophage biology and atherosclerosis. In the latter part of our studies we show that inducing a lysosomal biogenesis program is possible by overexpressing TFEB in macrophages. Furthermore, we demonstrate that this strategy has salutary effects on atherosclerotic macrophages by reducing lipid-mediated lysosome dysfunction, increasing cholesterol efflux, blunting inflammasome hyperactivation, and increasing the clearance of cytoplasmic inclusions.

Materials and Methods

Materials and Methods are available in the online-only Data Supplement.

Results

Atherosclerotic Macrophages have Morphologically Abnormal Lysosomes

The lysosome receives extracellular cargo (via endocytosis) and cytosolic material (via autophagy) for degradation. Failure of the lysosome to process its content efficiently leads to an accumulation of undigested material inside the lumen. For example, nearly all models of lysosomal storage disorders demonstrate noticeably enlarged lysosomes ^{19, 20}. Thus, an assessment of lysosomal morphology is an important first step in an evaluation of lysosomal function. We used immunofluorescence to visualize the lysosomes of peritoneal macrophages (PMACs) treated with atherogenic lipids. LAMP1, a glycoprotein abundantly expressed on the lysosomal membrane is a commonly used marker of lysosomal morphology ²¹. Incubation of PMACs with either oxidized LDL (oxLDL) or cholesterol crystals over a 24 hour period caused a significant and persistent increase in the size of LAMP1+ vesicles as assessed by confocal microscopy (Figure 1A,B).

Figure 1. Atherogenic Lipids Alter Lysosome Morphology and Function.

(A,B) Confocal microscopy of PMACs loaded with oxidized LDL (oxLDL, 50 mg/mL) or cholesterol crystal (CC, 500 mg/mL), and stained with LAMP1 antibody. Lysosome diameter was quantified from n=25 cells. (C,D) FACS analysis of PMACs treated with (C) oxLDL or (D) CC and stained with LysoTracker Red (200 nM). (E) Measurement of lysosomal pH after specified lipid treatment with 50 nM LysoSensor Yellow/Blue. Changes in pH were quantified as the ratio of emission at 530 nm to emission at 460 nm (n=8–10 wells for each treatment). (F) FACS analysis of PMACs loaded with 10 kDa TMR-dextran (25 mg/mL) followed by oxLDL or CC. (G,H) FACS analysis of PMACs loaded with DQ-ovalbumin (10 mg/mL) and treated with (G) oxLDL or (H) CC. For (C,F,G), mean fluorescence intensity for each peak was determined and expressed as a percentage of control (untreated cells). For (D,H) mean fluorescence intensity of the CC-containing cells (subpopulation with lower intensity peak) was determined and expressed as a percentage control (untreated cells). Representative results of at least three independent experiments are shown. For (B,E), graphs show the mean +/− SEM (*p<0.05).

Atherosclerotic Macrophages have Dysfunctional Lysosomes

Optimal lysosome function requires the ability to maintain an acidic pH and sequester a range of prodegradative enzymes operating at low pH away from the cytosol. A tool which permits the monitoring of pH-sensitive indices of lysosomal function is the lysosomotropic dye LysoTracker Red. Decreases in LysoTracker intensity relative to baseline suggest disruptions in lysosomal function, integrity, or quantity. We found that oxLDL-treated PMACs developed a progressive loss of LysoTracker signal with time (Figure 1C). Although a similar result was observed in PMACs treated with cholesterol crystals (Figure 1D), two points are worthy of mention. First, in contrast to oxLDL, the significantly larger size of cholesterol crystals precluded equal uptake by all cultured macrophages thus resulting in a bimodal LysoTracker distribution (i.e. cells with either dramatic losses of fluorescence signal or unaffected cells comparable to untreated control - Figure 1D). Second, cholesterol crystal-treated macrophages developed a loss of fluorescence signal significantly more than oxLDL treatment over a similar 24-hour time-frame (Figure 1C,D). Since oxLDL has recently been proposed to result in increased in situ formation of intralysosomal cholesterol crystals and lysosome dysfunction ¹⁴, we also compared the lysosomal effects of longer-term oxLDL incubation with those of cholesterol crystals. Interestingly, when cells were exposed to 72-hours of oxLDL, the effect on lysosomes was on-par with 24-hours of incubation with cholesterol crystal or the classic lysosomal inhibitor Bafilomycin (Supplemental Figure I-A).

Two primary factors can underlie the observed reduction in LysoTracker Red intensity after atherogenic lipid treatment: either a loss of lysosomal acidity leads to poor retention of the dye or a disruption in membrane integrity leads lysosomal leakage and loss of lysosomes. We desired to evaluate these in the following experiments.

Atherogenic Lipids Increase Lysosomal pH

In order to determine lysosomal pH more accurately we turned to a derivative lysosomotropic dye, LysoSensor Yellow/Blue. Although LysoSensor still diffuses and is selectively retained in lysosomes, it exhibits a dual emission spectra. At high pH (above 6.0) the dye fluoresces at a peak wavelength of 460nm while at low pH, peak emission is 530nm. Fluorometric measurement of the signal intensity at both wavelengths provides an elegant method of distinguishing samples on the basis of pH level. As shown in Figure 1E, oxLDL and cholesterol crystals both led to reductions in the 530/460 nm fluorescence emission ratio indicating a significant rise in lysosomal pH with atherogenic lipid treatment.

Atherogenic Lipids Increase Lysosomal Membrane Permeability

An intact lysosomal membrane is essential for the maintanence of a lysosomal proton gradient and the retention of the various intraluminal proteins and enzymes. Disruption of the lysosomal membrane has been proposed to contribute to the pathogenesis of several lysosomal storage diseases and to the activation of the inflammasome complex ^{13, 22}. We sought to measure the ability of oxLDL and cholesterol crystals to affect membrane porosity by FACS analysis of macrophages loaded with fluorochrome-conjugated dextran molecules. Dextran endocytosis into the lysosomal compartment leads to fluorescence; consequently, loss of fluorescence intensity indicates lysosomal leakage. Using a 10kDa dextran molecule, we found that although oxLDL treatment of macrophages did not alter lysosomal leakage, cholesterol crystals led to an overt signal loss (Figure 1F). In order to estimate the degree of porosity of the lysosomal membrane, we also used larger 70kDa dextran molecules. Significant loss of fluorescence was again seen in a portion of cholesterol crystal- but not oxLDL-treated macrophages (Supplemental Figure I-B). These data suggest that lysosomal membrane integrity is predominantly affected by cholesterol crystals and based on the leakage of both 10kDa and 70kDa dextrans, the degree of lysosomal membrane compromise appears to be significant.

Atherogenic Lipids Diminish the Proteolytic Capacity of Lysosomes

The effect of oxLDL and cholesterol crystals on lysosomal pH and membrane integrity would be predicted to alter the degradative capacity of lysosomes. We employed FACS to measure lysosome function via fluorochrome-conjugated ovalbumin (DQ-ova). Upon endocytosis, DQ-ova is delivered to the late endosome/lysosome and is subject to proteolysis by lysosomal enzyme leading to a quantifiable fluorescence. Figure 1G shows loss of DQ-ova fluorescence upon 24-hour incubation with oxLDL with similar but more dramatic defects seen with cholesterol crystals (Figure 1H). Akin to the LysoTracker experiments demonstrating comparable effects of prolonged oxLDL to shorter courses of cholesterol crystals (Supplemental Figure I-A), 72-hour exposure of macrophages to oxLDL led to dramatic decreases in DQ-ova signal on-par with 24-hours of cholesterol crystals or Bafilomycin (Supplemental Figure I-C).

Plaque Macrophages Display Lysosome Dysfunction

The experiments performed above were an in vitro assessment of lipid-loaded PMACs. Do macrophages present in atherosclerotic plaques in vivo develop lysosomal dysfunction? To answer this question, we needed a method to reproducibly isolate resident macrophages from mouse tissue including the aorta. Recently, the surface markers merTK and CD64 in combination have been shown to specifically label mature resident tissue macrophages ²³. Thus, the use of selective antibodies to these markers can be employed with FACS to isolate and study resident macrophages from any tissue of the mouse (Figure 2A). We first used ApoE−/− mice rendered atherosclerotic after 2 months of Western Diet and isolated macrophages from the spleen, liver, and aorta. LysoTracker Red signal of CD45+/CD64+/merTK+ macrophages from atherosclerotic aortas clearly showed diminished intensity compared to spleen and liver macrophages from the same mice (Figure 2B) indicative of dysfunctional lysosomes in plaque macrophages. We then compared resident macrophages of aortas from wild type and atherosclerotic (ApoE-null) mice fed a Western Diet for 2 months. Similarly, macrophages derived from the atherosclerotic aortas displayed a reduction in LysoTracker Red intensity (Figure 2C). The observed dysfunction appeared to be a result of atherosclerotic progression as aortic macrophages from young adult (6-week old) ApoE−/− mice fed a Chow Diet (where no overt atherosclerosis can be seen) had comparable LysoTracker Red staining to spleen and liver macrophages from the same mice (Supplemental Figure II-A) and wild type control aortas (Supplemental Figure II-B,C).

Figure 2. Atherosclerotic Plaque Macrophages Display Lysosomal Dysfunction.

(A) FACS gating Strategy to isolate live CD45+/merTK+/CD64+ resident macrophages from mouse tissues. (B) FACS analysis of spleen, liver, and aortic resident macrophages isolated from pooled tissue (n=3) atherosclerotic (ApoE−/−) mice after 2 months of Western Diet and stained with LysoTracker Red (200 nM). (C) FACS analysis of aortic resident macrophages isolated from pooled tissue (n=3) wild-type and atherosclerotic (ApoE−/−) mice after 2 months of Western Diet and stained with LysoTracker Red. For all bar graphs, mean fluorescence intensity for each peak was determined and expressed as a percentage of Lysotracker staining in splenic macrophages (B) or macrophages from wild-type (non-atherosclerotic) aorta (C). Representative results of at least three independent experiments are shown.

Lysosomal Stress Mediated by Chloroquine or Atherogenic Lipids Leads to a Compensatory Lysosomal Biogenesis Transcriptional Response

Work in the past few years has led to the discovery of the transcription factor TFEB as the only known master regulator of lysosomal biogenesis with an ability to increase the transcription of numerous lysosomal and autophagy genes in coordinated fashion ^{16, 24}. Increases in TFEB expression can directly lead to upregulation of a cohort of genes involved in autophagolysosome formation and the acid hydrolases involved in degradation of macromolecules ^{15, 16}. Furthermore, the induction of lysosomal stress by potent lysosomal inhibitors such as chloroquine can initiate a compensatory increase in lysosomal biogenesis via TFEB activation and nuclear translocation ²⁵. Although TFEB has thus far not been studied in macrophages, it would be surmised to have an important role given the highly pro-degradative nature of this cell type.

In light of our data demonstrating the development of lysosomal dysfunction in atherosclerotic macrophages, we first evaluated the effects of such stressors on the lysosomal biogenesis transcriptional response. Indeed, such a response appears to be conserved in macrophages as chloroquine treatment for 3 and 12 hours leads to the induction of a panel of lysosomal and autophagy TFEB gene targets (Figure 3A, upper panel) concomitant with nuclear translocation of TFEB (Figure 3B). Interestingly, a similar albeit more blunted response occurs after exposure of macrophages to oxLDL and cholesterol crystals (Figure 3A&B, lower panels). Since the nuclear translocation of TFEB appears to occur similarly in the presence of chloroquine and atherogenic lipids (Figure 3B), the more significant transcriptional response observed with chloroquine is unlikely to simply be a result of chloroquine's relatively higher potency in inhibiting lysosomes. For example, incubation of PMACs with oxLDL and cholesterol crystals over longer periods (12 and 24 hours) demonstrates a gradual reduction in TFEB nuclear staining suggesting desensitization (Supplemental figure III), whereas TFEB appears to remain in the nucleus at 24 hours with both Chloroquine and Bafilomycin (Supplemental Figure IV, upper panels).

Figure 3. Lysosomal Stress Promotes TFEB Nuclear Translocation and Transcriptional Activation of Lysosomal-Autophagy Genes.

(A) Quantitative PCR of PMACs either untreated (0 time-point) or treated with chloroquine, oxLDL, or cholesterol crystals for 3 and 12 hours. The transcription of a cohort of autophagy and lysosomal genes is expressed as fold over untreated cells (n=2–4 wells for each treatment). Graphs show the mean +/− SEM (*p<0.05). (B) Confocal microscopy of PMACs treated for 3 hours with the lysosomal inhibitor chloroquine (10 mm), oxidized LDL (50 mg/mL), or cholesterol crystals (500 mg/mL), and stained with TFEB antibody and DAPI (nuclei). Representative results of at least three independent experiments are shown.

Overexpression of the Transcription Factor TFEB Can Induce a Robust Lysosomal Biogenesis Program in Macrophages

Given the inability of macrophages to mount a robust compensatory transcriptional response to atherogenic lipids, we desired to determine whether TFEB overexpression, a condition that is known to clearly activate lysosomal biogenesis in other cell types ^{15, 16} can rescue lipid-induced lysosomal pathology. Tissue-specific overexpression of TFEB in mice has been successfully demonstrated using a loxP-Cre method ¹⁵. Backcrossing these mice with ones expressing Cre under the control of Lysosomal-M promoter (LysM-Cre), we were able to obtain macrophage-specific expression of TFEB (Figure 4A). PMACs from these mice showed increases in TFEB and target-gene expression (Figure 4B). We also detected concomitant increases in LAMP1 expression by FACS analysis suggesting a TFEB-mediated stimulation of the lysosomal pool (Figure 4C). Furthermore, in contrast to control cells, TFEB overexpression resulted in unabated translocation to the nucleus at baseline that was independent of modulation by lysosomal stressors such as Chloroquine, Bafilomycin, or oxidized LDL (Supplemental Figure IV). Since incubation of macrophages with atherogenic lipids leads to a profound lysosomal dysfunction by FACS analysis, we first tested whether TFEB can ameliorate this process. Figure 4D shows the effect of cholesterol crystals on loss of LysoTracker staining in control PMACs. In contrast, the reduction in LysoTracker fluorescence is less severe in TFEB-overexpressing (TFEB-Tg) macrophages suggesting relative preservation of lysosomal function. We next turned our attention to a variety of functional assays aimed at interrogating the effects of TFEB overexpression in atherosclerotic macrophages.

Figure 4. Macrophage Lysosomal Biogenesis is Induced by TFEB Overexpression and rescues the Lysosomal Dysfunction Mediated by Atherogenic Lipids.

(A) Schema outlining method of macrophage TFEB overexpression in transgenic mice. (B) Quantitative PCR of PMACs derived from macrophage-specific TFEB transgenic mice (n=2–4 wells for each treatment). Graph shows the mean +/− SEM (*p<0.05). (C) FACS analysis of control and TFEB-overexpressing PMACs for LAMP1 expression. (D) FACS analysis of control and TFEB-overexpressing PMACs treated with cholesterol crystals (CC, 500 mg/mL) for 48 hours. (E) Mean fluorescence intensity for each peak was determined and expressed as a percentage of untreated cells (No Tx). Representative results of at least three independent experiments are shown.

TFEB Overexpression Enhances Cholesterol Efflux

A prominent sequela of lysosome dysfunction in atherosclerotic macrophages is an impairment in cholesterol efflux ^{7, 9}. Lipophagy of cholesteryl esters present in macrophage lipid droplets followed by lysosomal hydrolysis via Lysosomal Acid Lipase is a mechanism by which free cholesterol is generated and primed for cellular efflux ^{9, 26}. We loaded control and TFEB-Tg macrophages with Acetylated LDL to generate foam cells and assessed the degree of cholesterol efflux to an Apo-A1 acceptor. Although there is little increase in cholesterol efflux in the early time-points, TFEB leads to a significant increase in efflux at 24 (Figure 5A) that persists to 48 hours (Supplemental Figure V-A). Of note, this TFEB-enhanced efflux is Apo-AI dependent as incubation of loaded cells without ApoA1 acceptor has no appreciable effect (Supplemental Figure V-B). The longer time frame is in keeping with previous data showing that cholesterol efflux generated from the lysosomal pool occurs with slower kinetics ^{7, 9}. Importantly, the enhanced cholesterol efflux clearly involves Lysosomal Acid Lipase (LIPA), as TFEB overexpression selectively upregulates LIPA mRNA and enzyme activity (Figure 5B,C), while a well-known LIPA inhibitor Lalistat-2 (Supplemental Figure V-C) leads to diminished cholesterol efflux in both control and TFEB-Tg macrophages (Figure 5D). Finally, we tested the effect of autophagy-deficiency on TFEB's induction of cholesterol efflux by repeating similar assays in ATG5-deficient (ATG5-KO) as well as dual ATG5-KO/TFEB-Tg macrophages (Figure 5E). In agreement with previous reports, ATG5-deficiency leads to reduced cholesterol efflux ^{7, 9}. Intriguingly, the absence of autophagy in TFEB-Tg macrophages blunted but did not completely abrogate the ability of TFEB to stimulate cholesterol efflux (Figure 5E). This suggests that TFEB's effects are likely mediated by the broader lysosomal biogenesis response that only partly includes the induction of autophagy genes.

Figure 5. TFEB Overexpression Enhances Cholesterol Efflux.

(A) Cholesterol efflux to ApoA1 (100 mg/ml) in control and TFEB-overexpressing (TFEB-Tg) macrophages loaded with Acetylated LDL for the indicated times. (B) Quantitative PCR of a set of genes involved in cholesterol efflux, and (C) Lysosomal Acid Lipase (LIPA) activity from control and TFEB-Tg macrophages. (D) Cholesterol efflux to ApoA1 in control and TFEB-Tg macrophages at 24 hours treated concomitantly with Lalistat (an inhibitor of LIPA). (E) Cholesterol efflux to ApoA1 in control, ATG5-deficient (ATG5-KO), and dual ATG5-deficient/TFEB-overexpressing macrophages (ATG5-KO/TFEBTg) for the indicated times. All graphs reflect n=2–4 wells for each treatment and show the mean +/− SEM (*p<0.05, NS = not significant). Representative results of at least three independent experiments are shown.

TFEB Overexpression Reduces Inflammasome Activation

Recent work has shown that an intact autophagy-lysosomal system and lysosomal membrane integrity are critical factors in suppression of the inflammasome complex and production of IL-1β ^{13, 27}. Additionally, cholesterol crystals in the atherosclerosis are important activators of inflammasomes and hypersecretion of IL-1β in plaque macrophages ^{7, 13}. Given the salutary effects of TFEB on macrophage lysosomal function, we interrogated the inflammasome system in the setting of TFEB overexpression. In control PMACs, the combination of LPS and cholesterol crystals synergistically hyperactivates IL-1β production without significant effects on pro-IL-1β protein levels consistent with inflammasome activation (Figure 6A). In contrast, TFEB-Tg macrophages displayed dramatically lower IL-1β secretion in the presence of LPS and cholesterol crystals, again independent of pro-IL-1β modulation (Figure 6A). We noted a similar beneficial response to TFEB overexpression in macrophages exposed to LPS and ATP, another potent activator of the inflammasome complex (Figure 6B). As autophagy-deficiency has been shown to also synergistically activate the inflammasome complex, we again desired to parse the the role of TFEB in stimulating both lysosomal and autophagy genes by conducting similar assays in ATG5-KO and dual ATG5-KO/TFEB-Tg macrophages (Figure 6C). The presence of TFEB once again led to significantly diminished IL-1β secretion even in the absence of autophagy, suggesting that TFEB's stimulation of the lysosomal pool is most likely the predominant mechanism by which inflammasome signaling is dampened.

Figure 6. TFEB Overexpression Reduces Cholesterol Crystal-mediated Inflammasome Activation and Cytoplasmic Inclusion Body Formation.

(A–C) ELISA of secreted IL-1β in the media of (A) control and TFEB-overexpressing (TFEB-Tg) PMACs treated with LPS (200 ng/ml) +/− cholesterol crystals (500 mg/mL) for 24 hours, (B) LPS +/− ATP for 2 hours, or (C) ATG5-deficient (ATG5-KO) and ATG5-KO/TFEB-Tg PMACs treated with LPS +/− cholesterol crystals for 24 hours. Cell lysates from respective treatments in (A) and (B) were also subjected to immunoblot for pro-IL-1β. (D) Confocal microscopy of control and TFEB-Tg PMACs loaded with cholesterol crystals (500 mg/ml) and stained with DAPI and antibodies against polyubiquitinated proteins and p62. (E) the number of p62+ aggregates were quantified in macrophages (n=25) imaged in (D). (F) Total p62 levels were determined by measurement of total fluorescence intensity. Graphs in (A–C) reflect n=2–4 wells for each treatment and all graphs show the mean +/− SEM (*p<0.05, NS = not significant). Representative results of at least three independent experiments are shown.

TFEB Overexpression Reduces Inclusion Body Formation

Finally, inclusion body formation is also a consequence of an impaired autophagy-lysosomal system ²⁸. In the absence of an intact lysosomal machinery, aged and misfolded proteins aggregate in cytoplasmic precipitates that can be toxic to the cell ²⁹. In support of this, treatment of peritoneal macrophages with atherogenic lipids (oxLDL and cholesterol crystals) can increase inclusion body formation (Figure 6D, Control) likely by disrupting lysosomal function. Consistent with the induction of a prodegradative response, we find that TFEB overexpression is also able to reduce the size and number of inclusion bodies induced by atherogenic lipids (Figure 6D,E). Of note, in the absence of atherogenic lipids, TFEB overexpression actually results in a diffuse cytoplasmic (rather than punctate) increase in p62 levels (Figure 6F). This rise is consistent with p62 being a transcriptional target of TFEB (see Figure 4B) and is an effect distinct from TFEB's ability to reduce the extent of inclusion bodies in atherogenic lipid-treated macrophages. Since p62 is a critical chaperone for removing protein aggregates via the autophagy-lysosomal apparatus ²⁸, it is likely that TFEB's transcriptional stimulation of p62 is a necessary precursor to efficient removal of p62-enriched inclusion bodies.

Discussion

The autophagy-lysosomal system is crucial in the processing and clearance of endocytosed material from the cell periphery as well as intracellular contents including long-lived/dysfunctional organelles and proteins ^{4, 5}. We and others have shown that this system is involved in the macrophage response to lipid in the atherosclerotic plaque and that defective autophagy increases the rate of atheroma progression ^7–9. Our understanding of the mechanism by which autophagy becomes dysfunctional in the plaque is rudimentary. The observation that p62 accumulates in both atherosclerotic plaques and lipid-loaded peritoneal macrophages in the absence of any changes in autophagic flux suggests the defect does not involve autophagy per se but rather the final component of the system, the lysosome. Using a series of complementary experiments, we have shown that atherosclerotic macrophages do indeed develop lysosomal dysfunction including lysosomal engorgement, increased lysosomal pH, decreased proteolytic capacity, and increased membrane porosity. Most importantly, we have demonstrated that TFEB, the transcriptional master regulator of lysosomal biogenesis and function, can be harnessed in macrophages to rescue many of the downstream functional consequences. Specifically, induction of lysosomal biogenesis in macrophages rescued cholesterol crystal induced lysosome dysfunction, temperized inflammasome activation, enhanced cholesterol efflux, and reduced inclusion body formation. A summary of these findings is outlined in Figure 7.

Figure 7.

Model Depicting Lysosomal Dysfunction in Atherosclerosis and The Benefits of Inducing a Lysosomal Biogenesis Program in Macrophages.

The exact mechanism by which the atherogenic lipids oxLDL and cholesterol crystals perturb lysosomal function is not known. Oxidized LDL is taken up by macrophage scavenger receptors and is trafficked to the endolysosomal compartment. Oxidized LDL can then bind and inactivate cathepsins with high affinity ³⁰, inactivate other proteases including the NaβGases ³¹, and produce a form of apolipoproteinB that is highly resistant to hydrolysis ^{32, 33}. OxLDL has also been demonstrated in endothelial and smooth muscle cells to inhibit activity and expression of the enzyme crucial to cholesterol ester hydrolysis, lysosomal acid lipase ³⁴. Most recently, the formation of cholesterol micro-crystals and ensuing disruption of lysosomal integrity has directly been linked to the build-up of oxLDL in the lysosomal compartment ¹⁴. Such a mechanism would favor the notion that the mechanism of lysosomal dysfunction mediated by oxLDL and larger cholesterol crystals lie in a continuum (with the oxLDL pool eventually precipitating as insoluble crystals). Our data supports this premise as the degree of lysosomal dysfunction induced by prolonged exposure to oxLDL eventually matched that seen with shorter courses of cholesterol crystals.

Once cholesterol takes on a crystalline form in lysosomes (either by phagocytosis of exogenous material or in situ formation), the mechanism by which lysosomes are rendered dysfunctional are more apparent. Extensive studies on the lysosomal handling of other types of crystalline material have been performed. Silica crystals are known to induce lysosome permeability ³⁵ with the reactive crystalline surface being postulated to directly perforate the lysosomal membrane and leakage of its contents ^{22, 35}. Interestingly, the sharp edges of crystalline material have been shown to penetrate the lysosome membrane of cells within the intimal layer of atherosclerotic human coronary vessels ^{36, 37}. Our data demonstrating a reduction in TMR-conjugated dextran with cholesterol crystals supports other recent studies where the leakage of fluorescent dextrans was linked to lysosome perforation and loss of its contents ^{13, 14}.

Our data also suggest that an additional consequence of atherogenic lipid loading appears to be a loss of lysosomal acidification (or rise in lysosomal pH). Regardless of the source of lipid (i.e. free cholesterol in crystalline form or that derived from the cholesteryl esters of oxLDL), it is clear that lysosomal cholesterol content increases. The accumulation of lysosomal free cholesterol can directly cause an increase in lysosomal membrane cholesterol content ³⁸. Intriguingly, translocation and attenuation of the activity of vATPase, the proton transporter responsible for lysosomal acidification, is dependent on lysosomal membrane cholesterol levels ³⁸. This could be a major contributor to the loss of acidification observed. It is important to note that we observed no significant differences in the rate of pH increase over a 24 hour period when comparing oxLDL or cholesterol crystals. Yet, one would have expected cholesterol crystals to have a significantly larger effect on lysosomal pH given their ability to also increase lysosomal porosity and destabilization. Detailed temporal comparison of the macrophage's metabolism of these lipids with correlation to pH changes could help answer some of these questions.

An exciting and promising aspect of our study is the ability of the transcription factor TFEB to induce lysosomal biogenesis in macrophages and in turn rescue several of the deleterious responses instigated by atherogenic lipid loading. TFEB has wide ranging effects on numerous processes related to enhanced autophagy and lysosomal function ¹⁵. Importantly, TFEB is translocated to the nucleus in several models of lysosomal storage disorders and under conditions of lysosomal stress, initiating the expansion of the cell's lysosomal compartment and the degradation of accumulated lysosomal contents ^{15, 24, 25}. Our finding that atherogenic lipids lead to the nuclear translocation of TFEB and the induction of a lysosomal biogenesis transcriptional program indicates that macrophages have the capacity to undergo similar regulation as other cell types.

Taken together, these observations make TFEB an attractive therapy for alleviating conditions of lysosomal distress. For example, AAV-mediated overexpression of TFEB provides neuroprotection and enhanced clearance of α-synuclein in an in vivo model of Parkinson's disease ³⁹. We have shown that atherosclerosis behaves as an acquired lysosomal storage disorder. It is very plausible that augmentation of lysosomal degradative capacity by TFEB in macrophages of atherosclerotic plaques can lead to significantly reduced lesion size, inflammation, and instability.

Given the pleiotropic effects of TFEB on the autophagy-lysosomal system, the exact nature of the salutary effects on macrophage function remain unclear. TFEB not only increases lysosome number but also increases the expression of numerous lysosomal enzymes including proteases such as the cathepsins and lipases such as lysosomal acid lipase (Figure 5B) ^15–18. Increased LAMP1 levels in TFEB-overexpressing macrophages (Figure 4C) would indicate an increase in the lysosome pool consistent with at least partial expansion of nascent lysosomes. However, the degree to which an increased quantity of nascent lysosomes or an increased efficiency of the current lysosome pool play a role in macrophages is difficult to ascertain. Furthermore, although autophagy and lysosomal degradation are intrinsically linked, is there an independent contribution of autophagy in mediating TFEB's effects in macrophages? Our initial evaluation of this has involved characterizing peritoneal macrophages from macrophage-specific autophagy-deficient (ATG5-KO) mice and ones with ATG5-KO in the presence of TFEB-overexpression (Figures 5E and 6C). Interestingly, autophagy appeared to play only a partial role in TFEB-mediated cholesterol efflux and a negligible role in mitigating inflammasome activation. This would support the notion that stimulation of lysosomal biogenesis imparts the more critical aspect of TFEB action than its enhancement of autophagy.

Dissecting the role of lysosomes in atherogenesis and macrophages of the atherosclerotic plaque is in its incipient stages. Our study contributes to the emerging body of literature that lysosomes lie at a critical nexus in lipid metabolism and inflammatory signaling. Further studies of TFEB and lysosomal biogenesis in vivo are sure to generate profound insights into the biology and future therapy of atherosclerosis.

Significance.

Lysosomes play an essential role in the degradation of both extracellular and intracellular material. This is especially true in cells with high degradative capacity such as macrophages. Recent studies have implicated disruptions in the macrophage autophagy-lysosomal pathway and atherosclerosis, but it is not clear how this disruption occurs during atherosclerotic progression. An attractive possibility is that endocytosed atherogenic lipids render lysosomes dysfunctional with downstream consequences. We now provide evidence for this in cultured primary macrophages as well as macrophages from the atherosclerotic plaque. These observations led us to hypothesize that stimulation of cellular mechanisms to increase lysosomal function/number might have salutary effects. Indeed, overexpression of TFEB, the only known master regulator of lysosomal biogenesis and a prodegradative response reverses the sequelae of lipid-induced lysosomal dysfunction. These findings highlight the importance of macrophage lysosomes in the pathogenesis of atherosclerosis and provide impetus for harnessing the lysosomal biogenesis response as an atheroprotective measure.

Abbreviations

oxLDL

(oxidized Low Density Lipoproteins)

CC

(cholesterol crystals)

FACS

(Fluorescence Activated Cell Sorting)

PMAC

(Peritoneal Macrophage)