Atherosclerosis is characterized biochemically by the accumulation of excess cholesterol in the artery wall. This process is initiated by migration of low density lipoproteins (LDL) and other apolipoprotein B (apoB)-containing lipoproteins across an injured artery endothelium, where they are retained through a charge-charge interaction with matrix proteoglycans in the subendothelial space (Tabas et al., 2007). Within this space the trapped lipoproteins are subject to modification including oxidation and aggregation, converting them to ligands for scavenger receptors on intimal macrophages and smooth muscle cells (SMCs). Excessive scavenger receptor-mediated uptake of modified lipoproteins leads to accumulation of cytosolic cholesteryl ester (CE) droplets in these cells, giving them a “foamy” appearance microscopically, hence their name foam cells. There is also evidence that at later stages of atherosclerosis, both free cholesterol (FC) and CE droplets accumulate within the lysosome itself (Jerome, 2006). Lysosomal acid lipase (LAL) plays the central role in hydrolyzing lipoprotein CE to generate FC, which after leaving the lysosome can be re-esterified in the endoplasmic reticulum to form cytosolic lipid droplets. If excess FC is retained in the lysosome it can inhibit lysosomal activity including that of LAL. The physiologic role of LAL may therefore lead to eventual inhibition of LAL activity in the presence of excess atherogenic lipoprotein uptake, contributing further to the progression of atherosclerosis.

Uptake of LDL through cell surface LDL-receptors is tightly regulated and therefore insufficient to explain intimal foam cell formation (Goldstein and Brown, 1977, 2009). Within the artery wall LDL and other apoB-containing particles become oxidized, aggregated, or enzymatically-modified (Aviram, 1993; Bhakdi et al., 1995; Hoff and Hoppe, 1995; Chisolm and Steinberg, 2000; Torzewski and Lackner, 2006) (Figure 1). These altered forms of LDL are no longer recognized by the LDL receptor but are recognized and taken up by scavenger receptors expressed on macrophages and SMCs, a cellular process that is not feedback regulated and can lead to massive accumulation of cholesterol in these cells (Brown et al., 1979; Brown and Goldstein, 1983; Moore and Freeman, 2006). Scavenger receptors as well as the LDL receptor utilize the endocytic pathway for delivery of cargo to lysosomes. Lipoprotein cholesterol is primarily in the form of CE, which is hydrolyzed within lysosomes by LAL, the only lysosomal lipase known to perform this function (Brown et al., 1975; Goldstein and Brown, 1977; Aviram, 1993; Dhaliwal and Steinbrecher, 2000; Yancey and Jerome, 2001; Griffin et al., 2005). LAL in the lysosome is most catalytically active at pH 3.5–4.5 (Takano et al., 1974; Haley et al., 1980). The FC released from lipoprotein CE is transported out of the lysosome by the concerted actions of the Niemann-Pick Type C2 (NPC2) and NPC1 lysosomal proteins, with one fate of the FC then being re-esterification in the endoplasmic reticulum by acyl-coenzyme A:cholesterol acyltransferase (ACAT) (Kwon et al., 2009). Re-esterification is a protective mechanism to prevent the toxicity of excess FC in membranes (Tabas, 2002), particularly in the endoplasmic reticulum, with the reformed CE then stored in the cytoplasm as benign lipid droplets (Brown and Goldstein, 1983). CE in foam cell cytosolic lipid droplets can be transported back to lysosomes through the autophagic process termed “lipophagy” (Singh et al., 2009); cholesterol derived from this process appears to form a large part of the substrate pool of ABCA1 for efflux from cells (Ouimet et al., 2011).

During early stages of atherosclerosis foam cells appear as visible fatty streaks in the intima of the artery wall (Haust, 1971). In addition to the presence of macrophage derived foam cells, SMCs present in the intima beginning in the pre-atherosclerotic diffuse intimal thickening stage also take up modified forms of LDL via scavenger receptors to become smooth muscle derived foam cells (Li et al., 1995; Doran et al., 2008). It has been previously thought macrophages were the major contributors to foam cells in the intima, however recent studies from our lab suggest SMCs are the source of more than 50% of the total foam cell population in human coronary artery atherosclerotic lesions (Allahverdian et al., 2014).

Retention of LDL and other apoB-containing lipoproteins in the arterial intima by matrix proteoglycans, as a result of charge-charge interactions (Williams and Tabas, 1995, 2005), means these lipoproteins are susceptible to modification to more atherogenic forms. This modification, either by enzymes, oxidants, or through aggregation converts these lipoproteins to ligands for scavenger receptors and allows for the formation of cholesterol-overloaded foam cells (Hoff and Hoppe, 1995; Torzewski and Lackner, 2006). Within the intima LDL has been demonstrated to undergo both extracellular and intracellular modifications (Aviram, 1993; Bhakdi et al., 1995; Hoff and Hoppe, 1995; Torzewski et al., 1998; Chisolm and Steinberg, 2000; Steinberg and Witztum, 2002; Stocker and Keaney, 2004; Wen and Leake, 2007). LAL and lysosomal enzymes have also been implicated in the modification of LDL, and, conversely, modified forms of LDL have been shown to affect lysosomal function and lead to changes in LAL activity.

Previous studies have reported variable effects of different forms of modified LDL on accumulation of cholesterol either in the cytosol alone or also in lysosomes, with oxLDL and aggLDL causing accumulation of cholesterol in lysosomes, while enzymatically-modified and acetylated LDL cause mainly cytosolic accumulations (Yancey and Jerome, 2001; Griffin et al., 2005; Orso et al., 2011). Treatment of THP-1 macrophages in culture with mildly oxLDL or aggLDL shows similar trends in loss of lysosomal hydrolysis (Yancey and Jerome, 2001; Griffin et al., 2005). Initially CE hydrolysis is not inhibited but lysosomal sequestration of FC occurs. After prolonged exposure to mildly oxLDL or aggLDL (48 h or more) increasing inhibition of CE hydrolysis occurs and lysosomal accumulation switches to mainly CE (Yancey and Jerome, 2001; Griffin et al., 2005).

Some variability between species has been observed (Griffin et al., 2005). Yancey and Jerome (1998) reported that THP-1 human macrophages and pigeon macrophages store FC and CE derived from mildly oxLDL within lysosomal compartments, whereas mouse macrophages stored most of the cholesterol in cytosolic lipid droplets. Interestingly, acLDL was efficiently hydrolyzed in macrophages regardless of the species and led to cytosolic lipid droplet accumulation (Yancey and Jerome, 1998). Maor and Aviram (1994) reported that the addition of oxysterols isolated from oxLDL to medium when loading J-774 mouse macrophages with acLDL led to significant lysosomal FC accumulations.

Cox et al. (2007) have reported a general loss of lysosomal function including reduction in LAL-dependent CE hydrolysis over time following incubation of THP-1 human macrophages with mildly oxLDL or aggLDL. The similar effects using unoxidized aggLDL and oxLDL suggest this inhibition of lysosomal function is not specific to the effects of oxidized lipids. Increasing levels of lysosomal membrane FC during lipid loading were shown to inhibit lysosomal acidification (Cox et al., 2007). It was demonstrated using fluorescence quenching studies that excess FC in the lysosomal membrane leads to loss of acidity as a result of inhibition of vacuolar H⁺-ATPase (v-ATPase) proton pumping activity in the lysosomal membrane, without a change in the amount of v-ATPase protein. The resultant increase in pH within the lysosome corresponded with a time-dependent inhibition of CE hydrolysis and accumulation of apoB, indicating a general decline in lysosomal hydrolytic activity. After 7 days of incubation with mildly oxLDL or aggLDL, the majority of lysosomal vesicles within THP-1 macrophages had pH >4.8. Vacuolar ATPase activity could be partially recovered if isolated lysosomes were treated with cyclodextrins in order to remove membrane cholesterol. As LAL is only active at acidic pH, with maximal activity ~pH 4 and very little activity above pH 4.5 (Burton et al., 1980; Haley et al., 1980), this loss of v-ATPase activity in response to cholesterol loading of the lysosomal membrane may be the main cause of loss of hydrolytic activity of LAL, and hence lead to lysosomal lipid droplet accumulations.

In contrast to oxLDL and aggLDL, incubation of THP-1 macrophages with acLDL, an experimental rather than physiologic form of modified LDL, did not result in an increase in lysosomal membrane FC or an increase in lysosomal pH (Cox et al., 2007). The reasons for the differential effects of acLDL and oxLDL or aggLDL on sequestration of FC in lysosomal membranes following hydrolysis by LAL are unknown. In order to investigate if the observed neutralization of lysosomal pH and inactivation of LAL following oxLDL uptake could be reversed by acLDL, THP-1 macrophages were incubated for 3 days in the presence of oxLDL to accumulate lysosomal CE and then chase incubated with acLDL for an additional 3 days (Jerome et al., 2008). Hydrolysis of CE in the chase acLDL incubation was inhibited and the loss of acidic pH from the initial loading with oxLDL was maintained, indicating lysosomal dysfunction was not readily reversible. Similar results were found in human monocyte derived macrophages. This data indicates that inhibition of LAL activity is the result of a long-lived alteration in lysosomal function under excess lipid loading.

Emanuel et al. (2014) also found a loss of lysosomal acidity occurred in mouse peritoneal macrophages at 72 h of oxLDL treatment and in isolated macrophages from atherosclerotic aortic tissue of ApoE^−/− mice. This group also reported that overexpression of TFEB in mouse peritoneal macrophages was able to preserve lysosomal acidity, by mechanisms that are not yet clear.

An interesting study by Ullery-Ricewick et al. (2009) demonstrated that in THP-1 human macrophages lysosomal CE accumulations from aggLDL could be reduced by 50% by chase incubation with triglyceride (TG)-rich lipid dispersions or VLDL. It was shown that TG treatment reestablished acidic pH by restoring proton pumping by v-ATPases in the lysosomal membrane (Ullery-Ricewick et al., 2009). TGs also increased LAL activity, with no change in LAL expression. The exact mechanism by which TG-containing particles restore lysosomal function in foam cells remains to be determined; however, it was demonstrated to not be a result of competitive uptake of TG-rich vs. CE-rich particles (Ullery-Ricewick et al., 2009). This is further evidence that inhibition of v-ATPase proton pumping in the lysosomal membrane leads to loss of LAL activity and lysosomal CE accumulations, but may be reversible under some conditions. Further investigation is needed to evaluate the role of TGs in lysosomes of arterial wall foam cells, including the signaling role of TG-derived fatty acids on the LXR and PPAR metabolic pathways.

An alternative or additional reason for the increase in lysosomal pH is that excess membrane FC and oxidized FC can cause lysosomal leakiness, leading to a loss of the proton gradient. Li et al. (1998) reported that incubation of J-744 macrophages with oxLDL resulted in leakage of lysosomal enzymes into the cytosol. Yuan et al. (2000) found that a mixture of cholesterol oxidation products found in oxLDL were also able to induce damage to lysosomes and leakage of lysosomal contents into the cytosol, and eventual macrophage cell death. Although leakiness was not implicated in the studies by Cox et al. (2007) using mildly oxLDL or aggLDL, as apoptosis was not observed, it may be a long term contributing factor in vivo. The presence of free cholesterol crystals has been reported in J774 macrophage foam cells (Tangirala et al. (1994), and treatment of mouse peritoneal macrophages with free cholesterol was shown to induce lysosomal membrane leakiness (Emanuel et al. (2014). An increase in lysosomal pH has also been shown to increase extracellular excretion of lysosomal enzymes (Tapper and Sundler, 1990), which may contribute to the extracellular LAL observed in atherosclerotic lesions (Hakala et al., 2003). Overall it appears that increases in lysosomal FC induce both defects in lysosome acidification and leakiness of the lysosomal membrane.

Oxidation of apoB-100 has been reported to increase its resistance to degradation by cathepsin D in mouse peritoneal macrophages and its accumulation in lysosomal compartments (Lougheed et al., 1991; Jessup et al., 1992; Roma et al., 1992; Hoppe et al., 1994; Mander et al., 1994). Decreased apoB-100 protein degradation may also limit LAL mediated hydrolysis of the CE lipid core of oxLDL. Oxidized phospholipids in oxLDL can inhibit cathepsin D, which may explain the reduced ability to degrade apoB (O'neil et al., 2003). Oxidized phosphatidylcholine-apoB complexes have been found in human atherosclerotic lesions and also in lysosomes of cultured mouse macrophages incubated with oxLDL (Itabe et al., 2000). Brown et al. (2000) found that oxidized CE from heavily oxidized LDL loading are resistant to hydrolysis by LAL and accumulate in lysosomal compartments. Therefore, oxidative modifications to apoB containing particles may lead to impaired lysosomal clearance for reasons additional to those induced by non-oxidized aggLDL.

In the absence of genetic deficiency in LAL, in vivo studies including electron microscopy have indicated that, in addition to storage of CE droplets in the cytoplasm, there are substantial accumulations of FC and CE within the lysosomes of foam cells (Jerome and Yancey, 2003). This phenomenon has been previously reported for both human (Coltoff-Schiller et al., 1976) and animal (Peters et al., 1972; Shio et al., 1974, 1979; Goldfischer et al., 1975; Fowler et al., 1980; Jerome and Lewis, 1985) atherosclerosis. It has been indicated that as atherosclerosis progresses there is initially an increasing amount of FC retained in lysosomes, and at later disease stages also retention of lyososomal CE (Jerome, 2006). A possible conclusion that can be drawn from this is that there is a general loss of lysosomal function due to cholesterol overload following unregulated cellular uptake of modified forms of LDL within artery wall cells. As LAL hydrolyzes lipoprotein CE, the loss of LAL function due to excess FC within the lysosome contributes to the formation of foam cells and the altered distribution of lipid stores within these cells. Therefore, atherosclerosis has been proposed as a type of lysosomal storage disorder (Miller and Kothari, 1969; Peters et al., 1972; Peters and De Duve, 1974; Shio et al., 1974; Coltoff-Schiller et al., 1976; Jerome and Lewis, 1985).

As indicated above, the reason for the loss of lysosomal hydrolytic function may not be a result of deficiency in LAL expression as atherosclerosis progresses but rather a decrease in the catalytic activity of LAL as a result of excess lysosomal cholesterol accumulation from modified forms of LDL (Figure 3). LAL remains catalytically active initially but over time FC in the lysosomal membrane increases and inhibits the proton pumping ability of the v-ATPases. As a result, the pH of the lysosome increases and renders LAL catalytically inactivate. Lysosomal CE accumulation then occurs in addition to cytosolic accumulation in later stage atherosclerotic lesions.

Previous studies in atherosclerotic tissue have indicated that LAL activity is increased relative to non-atherosclerotic arteries (Brecher et al., 1977; Takano, 1978; Subbiah, 1979; Haley et al., 1980; Davis et al., 1985). The analytical procedures for those studies, however, involved ex vivo acidic pH adjustment and therefore reflects the total amount of potentially active LAL present, but not what may be actually active in vivo within intact cells under the influence of excess lipids and altered lysosomal pH. As activity is also reflective of protein amount, these studies may indicate that LAL expression increases to deal with the increased cellular influx of apoB containing particles within the artery wall. It can be proposed that LAL activity and protein amount in arterial cells increases with lipid challenge up to a certain point, either from increased LAL expression per cell or an increase in total LAL-expressing cells, after which the lysosome is overwhelmed with lipids and the activity of LAL decreases due to increased pH caused by excess lysosomal membrane cholesterol and leakiness.

The question then arises what is responsible for the elevation of lysosomal FC in response to the excess lipoprotein load in the intima. As NPC1 and NPC2 facilitate the release of FC from the lysosome, investigation of the expression and activity of these proteins may provide useful insight in this area. Previous studies in fibroblasts have suggested that retention of cholesterol within lysosomal compartments may serve to protect cells from endoplasmic reticulum (ER) stress caused by an excess of FC in the ER membrane. Garver et al. (2008) and Jelinek et al. (2009) found that loading normal human fibroblasts with LDL results in reduced SREBP-dependent expression of NPC1 and NPC2. Jelinek et al. (2012) reported that in a mouse model of diet induced obesity, dietary fatty acids but not cholesterol induced down regulation of NPC1 in both hepatic cells and peritoneal fibroblasts via feedback inhibition of the SREBP pathway. In THP-1 macrophages it has been reported that oxLDL treatment for 3 days leads to a buildup of NPC1 protein in the Golgi apparatus (Jerome et al., 2002). Differential responses to cholesterol-dependent manipulations of NPC1 and NPC2 expression have been indicated in the literature. Rigamonti et al. (2005) demonstrated that expression of NPC1 and NPC2 increased in the presence of LXR agonists in human but not mouse macrophages. It has also been demonstrated that treatment of human macrophages with oxLDL leads to increases in NPC1 and NPC2 expression via PPARα- and mitogen-activated protein kinase (MAPK)-dependent pathways (Chinetti-Gbaguidi et al., 2005; Yu et al., 2012). These results indicate that the expression of NPC1 and NPC2 with lipid overload varies between cell types and species. Further investigation is needed to elucidate the reason for FC overload in the lysosomes of atherosclerotic foam cells.

As the rate of release of FC from the lysosome is a regulator of ABCA1 expression, lysosomal dysfunction in atherosclerosis may also lead to decreased expression of ABCA1 in artery wall cells. Choi et al. (2009) have reported that model human intimal SMCs as well as human coronary artery intimal SMCs express lower levels of ABCA1 than medial SMCs. More recently, Allahverdian et al. (2014) reported that intimal SMCs but not myeloid lineage cells express lower levels of ABCA1 in late vs. early stage human atherosclerotic lesions. Further studies are required to determine whether decreased LAL catalytic activity as a result of lysosomal dysfunction is responsible for this defect in ABCA1 expression, and hence further contributing to the overaccumulation of cholesterol in intimal SMC foam cells.

Recombinant human LAL (rhLAL) is currently in phase 3 clinical trials to treat LAL deficiency in Wolman disease and CESD. This treatment is potentially life-saving in Wolman disease. Phase 2 clinical trials in patients with CESD using rhLAL demonstrated safety and improvements in serum lipid profiles (Balwani et al., 2013). Long-term follow up will be needed in CESD patients to determine outcomes in terms of cardiovascular events. Several studies in mice have indicated that rhLAL may be beneficial in reversing atherosclerosis. Studies by Du et al. (2001) and Sun et al. (2014) demonstrated that injection of rhLAL into LAL^−/− mice was able to reverse the pathogenic storage of lipids in multiple tissues. Du et al. (2004) also showed that administration of repeat doses of rhLAL to LDL-receptor-deficient mice fed a high fat/cholesterol diet resulted in complete regression of early stage lesions in coronary and aortic tissue and a significant reduction in late stage lesions.

Although rhLAL treatment seems to be a viable therapeutic strategy for premature atherosclerosis in CESD patients, it is unclear what the effects would be of rhLAL treatment in patients having atherosclerosis unrelated to a deficiency in LAL. The lysosomal acid lipase A (LIPA) gene has been identified as a susceptibility gene for coronary artery disease by several genome-wide association studies (Consortium, 2011; Coronary Artery Disease Genetics, 2011; Wild et al., 2011). For the reasons outlined here, increasing the activity of LAL beyond the normal cellular response may not be an effective strategy. As scavenger receptor mediated uptake of modified forms of LDL in both macrophages and SMCs occurs in atherosclerotic lesions, it is conceivable that rhLAL treatment might increase lysosomal membrane FC in these cells by increasing the rate of lipoprotein CE hydrolysis, and thereby exacerbate lysosomal dysfunction. Reducing the level of atherogenic lipoproteins in plasma and their initial influx into the artery wall, therefore, remains of paramount importance.

The role of LAL in the progression of atherosclerosis is complex. Modified forms of LDL such as oxLDL and aggLDL lead to lysosomal accumulations of first FC and later CE, indicating an acquired loss of LAL hydrolytic activity. In agreement with this observation, tissue studies have indicated increased lysosomal lipid accumulations in addition to cytosolic lipid droplets at later stages of atherosclerosis. Activity assays of atherosclerotic tissue homogenates have indicated an increase in LAL and also other lysosomal enzymes. As these studies are conducted ex vivo using acidic pH adjustment, they are representative of potentially functional LAL present in vivo but not necessarily actual LAL activity. Therefore, although the amount of artery wall LAL may increase in response to increased cellular influx of apoB-100 containing lipoproteins, the hydrolytic activity of LAL may decrease over time as lysosomal function is impaired. Sequestration of cholesterol in the lysosomal membrane has been implicated as a cause of lysosomal dysfunction, specifically through the inhibition of v-ATPase proton pumping. Increasing lysosomal pH reduces the activity of LAL. The result of this is a time dependent switch of lysosomal accumulation of FC to CE. Early stage atherosclerosis may involve normal and perhaps increased LAL activity leading to cytosolic lipid droplet accumulation, whereas later stages of atherosclerosis may have an acquired dysfunction in LAL hydrolytic activity leading to lysosomal lipid sequestration. Therefore, later stages of atherosclerosis, where LAL function is inhibited, may be considered to be morphologically similar to Wolman disease and CESD, albeit for an entirely different reason.

Outside of alleviating premature onset of atherosclerosis associated with genetic deficiency in LAL, it is hard to predict whether augmentation of LAL is a good therapeutic strategy. There are indications that selective upregulation of the autophagic-lysosomal pathway may be able to recover lysosomal function and LAL hydrolysis. It is unclear, however, whether induction of this pathway beyond normal response is a viable in vivo strategy. Further research is necessary to know whether manipulation of arterial cell LAL activity is likely to alleviate or aggravate the consequences of cholesterol accumulation in the artery wall.