The Many Roles of Cholesterol in Sepsis: A Review

Abstract

The biological functions of cholesterol are diverse, ranging from cell membrane integrity, cell membrane signaling, and immunity to the synthesis of steroid and sex hormones, vitamin D, bile acids, and oxysterols. Multiple studies have demonstrated hypocholesterolemia in sepsis, the degree of which is an excellent prognosticator of poor outcomes. However, the clinical significance of hypocholesterolemia has been largely unrecognized. We undertook a detailed review of the biological roles of cholesterol, the impact of sepsis, its reliability as a prognosticator in sepsis, and the potential utility of cholesterol as a treatment. Sepsis affects cholesterol synthesis, transport, and metabolism. This likely impacts its biological functions, including immunity, hormone and vitamin production, and cell membrane receptor sensitivity. Early preclinical studies show promise for cholesterol as a pleiotropic therapeutic agent. Hypocholesterolemia is a frequent condition in sepsis and an important early prognosticator. Low plasma concentrations are associated with wider changes in cholesterol metabolism and its functional roles, and these appear to play a significant role in sepsis pathophysiology. The therapeutic impact of cholesterol elevation warrants further investigation.

Sepsis, the dysregulated host response to infection resulting in organ dysfunction (1), is a major worldwide cause of mortality (2) and morbidity. Current management focuses on adequate fluid resuscitation, organ support, and treating the infection with antibiotics and source control. To date, no available treatments that directly target underlying pathophysiological mechanisms have been clearly demonstrated to improve outcomes.

Cholesterol, a sterol lipid, plays an integral role in multiple body functions, including maintenance of cellular membrane processes, immunity, signaling, and pathway regulation, and acts as a precursor for the synthesis of steroid hormones, vitamin D, bile acids, and oxysterols. Sepsis-induced hypocholesterolemia was first recognized a century ago (3); multiple studies demonstrate a worse prognosis being associated with the magnitude of decline. However, the mechanisms by which plasma concentrations fall, the impact on organ functionality, the relationship of plasma cholesterol to intracellular concentrations, and the potential role of cholesterol as a therapeutic all require elucidation.

There is increasing interest in the therapeutic possibilities of lipoproteins and modulation of cholesterol transport in sepsis, particularly in immunoinflammatory modulation and pathogen scavenging. There has, however, been little focus on cholesterol itself rather than on its carriers. In this article, we provide an overview of the biology of cholesterol, its possible roles in sepsis pathophysiology, and its potential utility as a specific adjunctive treatment.

Cholesterol Synthesis, Structure, Metabolism, and Functional Roles

Cholesterol consists of four linked aromatic hydrophobic rings, a small hydrophilic hydroxyl group, and a hydrophobic chain. Because of its high hydrophobicity, cholesterol is only present within cells predominantly as a component of lipid membranes or bound to lipid-binding proteins (4) (Figure 1). Animals obtain cholesterol through diet and, primarily, by endogenous synthesis. Cholesterol synthesis is a multistep (∼30-reaction) process that is highly energy consuming; synthesis of one cholesterol molecule requires 18 acetyl-CoA, 36 ATP, 16 NADPH, and 11 oxygen molecules. Endogenous cholesterol synthesis is tightly regulated by negative feedback (Figure 2). HMG-CoA (hydroxymethylglutaryl–coenzyme A) reductase, the target of statin therapy, is the rate-limiting enzyme within the pathway and the predominant mechanism by which cells adapt to changes in cholesterol bioavailability.

Figure 1.

Cholesterol structure and location within cell membranes. GPI = glycosylphosphatidylinositol.

Figure 2.

Cholesterol synthesis and metabolism pathways and impact of sepsis. *Plasma concentrations may be normal or raised for adrenocorticoid hormones and bile acids, but this may relate to decreased metabolism and/or excretion rather than to increased production. Cortisol concentrations frequently fail to augment with exogenous ACTH (adrenocorticotropic hormone) stimulation. The question mark represents uncertain effect. ABC1 = ATP-binding cassette transporter 1; AMPK = AMP-activated protein kinase; ApoA1 = apolipoprotein A1; HDL = high-density lipoprotein; HMG-CoA = hydroxymethylglutaryl–coenzyme A; INSIG = insulin-induced gene 1 protein; LCAT = lecithin–cholesterol acyltransferase; LDL = low-density lipoprotein; SCAP = SREBP cleavage–activating protein; SIRT-1 = sirtuin 1; SREBP = sterol regulatory element–binding protein; VLDL = very-low-density lipoprotein.

To enable transport in plasma, cholesterol must be bound to lipoproteins or albumin. Lipoproteins are categorized into chylomicrons, chylomicron remnants, VLDL (very-low-density lipoprotein), LDL (low-density lipoprotein), and HDL (high-density lipoprotein) by density, size, and the types of particle-forming proteins and other associated proteins. Cholesterol bound to LDL is transported from the liver to peripheral tissues, whereas HDL carries cholesterol to the liver and steroidogenic tissues—“reverse cholesterol transport” (4). Mammalian cells lack an enzyme system to catabolize and recycle cholesterol and its derivates. The liver clears cholesterol from the circulation via LDL and HDL receptors (5). It is then metabolized or excreted either unmodified or as bile acids, a large proportion of which are reabsorbed.

Cholesterol and its metabolites provide multiple biological functions (Figure 3):

Figure 3.

Functional roles of cholesterol.

• 1.Cholesterol is an integral part of cell membranes and plays a crucial role in modulating membrane thickness, permeability, fluidity, and functionality (6, 7). Within the membrane, cholesterol distributes nonhomogeneously, accumulating within lipid rafts. These small, highly dynamic, sterol- and sphingolipid-enriched membrane microdomains attract many transmembrane proteins, such as ion channels, transporters, and receptors, including GPCRs (G-protein–coupled receptors) (7). Alterations in membrane cholesterol affect the membrane’s physical properties and influence the presence and activity of transmembrane proteins, such as the sodium–potassium–ATPase and β-adrenergic receptors (7).
• 2.Both cholesterol and its lipoprotein carriers have immunomodulatory properties, including binding of endotoxin and other toxins (8, 9). This scavenging mechanism may play an important role in neutralizing toxins as part of the innate immune system response, preventing activation of TLRs (Toll-like receptors) by pathogen-associated molecular patterns. Of note, key receptors regulating the immune response such as TLRs and T- and B-cell receptors are localized within lipid rafts (10).
• 3.Cholesterol is the only steroidogenic substrate used to synthesize adrenocortical hormones (glucocorticoids, aldosterone), sex hormones (e.g., estrogen, progesterone, testosterone), and vitamin D through multistep processes (11). During a triggered stress response, ∼80% of circulating cortisol may be derived from plasma cholesterol (12). The impact of vitamin D on multiple diseases, including musculoskeletal disorders, insulin resistance, and metabolic syndrome, and on cardiovascular and immunological dysfunction has been studied extensively (13).
• 4.Conversion of cholesterol to bile acids involves 17 distinct enzymatic steps within hepatocytes and is the principal route of cholesterol metabolism. Bile acids undergo enterohepatic recirculation, allowing recycling, with de novo hepatocyte synthesis compensating for physiological intestinal losses. Bile acids aid metabolite excretion by the liver; aid the absorption of lipids, hydrophobic nutrients, and fat-soluble vitamins; and prevent bacterial overgrowth within the small bowel and biliary tree. They also regulate multiple functions within various liver cell types (e.g., cell differentiation and regeneration) (14).
• 5.Oxysterols represent a large family of oxidized derivatives of cholesterol with multiple biological actions, including immunomodulation (15). Cholesterol can be oxidized either enzymatically or nonenzymatically by reactive oxygen species. Oxysterols can exert their functions through GPCRs, nuclear receptors, and other molecular pathways, regulating many processes, ranging from cytokine production to virus entry into cells (16, 17). Oxysterols modulate neutrophil, B-cell, and T-cell functionality, enhance innate immunity, and regulate production of the antiinflammatory cytokine IL-10 (17, 18).

Cholesterol Concentrations Fall during Sepsis, in Line with Severity and Outcome

Reductions in total plasma cholesterol, HDL cholesterol (HDL-C), and LDL cholesterol (LDL-C) are well recognized in sepsis (19–28). Concentrations are decreased at the time of diagnosis (21) and often decline further during the disease course (25). Serum HDL-C concentrations reach a nadir around Day 3 after admission, whereas LDL-C is lowest at the time of diagnosis (21). Variable recovery in serum concentrations occurs over subsequent days (25). The kinetics of VLDL cholesterol (VLDL-C) in sepsis are poorly characterized in human sepsis.

Multiple studies report a greater mortality risk in patients with lower concentrations of total cholesterol, HDL-C, and LDL-C (23–28). Of note, a recent genetic study suggested that low LDL concentrations in sepsis may be associative with rather than causal of an increased mortality risk (27), whereas low HDL-C may be a causal factor (29). Increased LDL clearance may contribute to a lower sepsis mortality via enhanced pathogen lipid clearance (27).

Survivors show a slow return to almost normal values over the disease course. The magnitude of falling is associated with a higher incidence of multiorgan dysfunction, an increased-duration ICU stay, and more nosocomial infection (23, 26). Elevated serum markers of inflammation correlate negatively with cholesterol concentrations (20, 24, 28).

Infusion of recombinant TNF-α or IL-6 into patients with cancer also produced large falls in plasma cholesterol in inverse correlation with markers of inflammation (30, 31). Animal experiments can replicate these findings and can be used as a therapeutic test bed. However, this is model dependent, as some rodent models injected with endotoxin or TNF-α actually demonstrate hypercholesterolemia (32). However, we and others have found large falls in total cholesterol and HDL-C concentrations in rats given a more realistic peritonitis insult (33–35). Hypocholesterolemia has also been demonstrated in septic models by using primates, sheep, and dogs (36–38).

Why Does Serum Cholesterol Fall in Sepsis?

Biological mechanisms leading to hypocholesterolemia in sepsis remain incompletely understood. Apart from decreased intake and impaired intestinal absorption of fat in critical illness (39), decreased synthesis, impaired cholesterol transport, increased metabolism, and depletion through toxin scavenging may be implicated.

Data on the impact of sepsis on cholesterol synthesis are limited and conflicting. Older studies in rodent models reported increased hepatic cholesterogenesis (32, 40) and concurrent hypercholesterolemia (32). Vasconcelos and colleagues (40), however, noted a decrease in HMG-CoA reductase activity in septic rats compared with healthy, fed rats. Our currently unpublished data (A. Kleyman and colleagues, unpublished results) reveal decreased mRNA expression of transcriptional regulators (SREBP-1 [sterol regulatory element–binding protein], SREBP-2, INSIG [insulin-induced gene 1 protein]) and enzymes (HMG-CoA reductase) within the hepatic cholesterol synthesis pathway in our rat peritonitis model.

Proinflammatory cytokines may contribute to hypocholesterolemia by reducing hepatic synthesis of apolipoproteins that bind cholesterol to form lipoproteins (41). Falls in plasma LDL-C are commonly but variably reported, whereas low HDL-C is a consistent finding. Those changes suggest that reverse cholesterol transport (i.e., transfer of cholesterol from peripheral tissues to the liver) may be more affected (19). Figure 4 illustrates different cholesterol metabolic and transfer pathways affected by sepsis. Transporters (e.g., the ABC [ATP-binding cassette transporter] superfamily, which transforms lipid-poor apoA-1 [apolipoprotein A1] particles into mature HDL particles) and enzymes such as LCAT (lecithin–cholesterol acyltransferase), which converts free cholesterol to more hydrophobic cholesterol esters, enabling incorporation into HDL, are affected by sepsis (22, 41). The binding capacity of HDL is also affected by alterations in its structure and protein composition and by the accumulation of oxidized lipids (42).

Figure 4.

Impact of sepsis on cholesterol transport. ABC = ATP-binding cassette transporter; CETP = cholesteryl ester transfer protein; HDL = high-density lipoprotein; LCAT = lecithin–cholesterol acyltransferase; LDL = low-density lipoprotein; LDL-R = LDL receptor; PCSK9 = proprotein convertase subtilisin kexin 9; SR-BI = scavenger receptor B type 1; VLDL = very-low-density lipoprotein.

CETP (cholesteryl ester transfer protein) mediates triglyceride and cholesteryl ester transfer between triglyceride-rich lipoproteins and HDL particles, with lower plasma CETP concentrations increasing the proportion of HDL-C. However, total circulating cholesterol concentrations are unaffected (43). The literature on the relevance of changes in plasma CETP concentrations in sepsis and their relationship to outcomes is conflicting (29, 44–46).

Similarly, conflicting patient data are seen with regard to alterations in plasma concentrations of PCSK9 (proprotein convertase subtilisin kexin 9), an enzyme that degrades hepatic LDL and adipocyte VLDL receptors, resulting in hypercholesterolemia (47–49).

Cholesterol metabolism can be increased in sepsis by enzymatic and nonenzymatic oxidation. Cholesterol-25-hydroxylase is strongly induced by inflammation and its product, 25-hydroxycholesterol (50). The acute-phase protein PLA2 (phospholipase A2) activity rises during inflammation and promotes increased metabolism of cholesterol esters and apolipoproteins, thereby reducing serum cholesterol (51). PLA2 activity is enhanced by another acute-phase reactant, SAA (serum amyloid A), which also affects cholesterol transport (52). Sepsis, however, decreases bile flow (53). Impaired biotransformation and hepatobiliary transport of bile acids occur within hours of the induction of polymicrobial sepsis (54). As a consequence, bile acids can be elevated in the blood compartment.

Impact of Sepsis on the Biological Roles of Cholesterol

As described earlier, cholesterol and its various metabolites exert many complex biological functions, many of which are disrupted during sepsis. The specific contribution of cholesterol deficiency to these abnormalities requires further elucidation, but there is sufficient direct and circumstantial evidence to suggest that cholesterol deficiency may play an important role.

Cell Membrane Function

The cholesterol composition within lipid rafts modifies intrinsic function and downstream signaling, such as within the adrenergic receptor pathway. Cholesterol depletion in human neutrophil cell membranes induced a more proinflammatory phenotype that included priming, enhanced activation, increased adhesion, and oxidant production (55, 56). Raft-dependent signaling of multiple cell types may be altered because of changes in membrane cholesterol concentrations affecting, for example, GPCR density and activity (6, 7). This may be of particular relevance in septic shock, for which myocardial and vascular hyporeactivity to exogenous catecholamines is a defining characteristic, with the magnitude of hyporesponsiveness being associated with increased mortality (57).

Immunomodulatory and Antibacterial Properties of Cholesterol

Notwithstanding the scavenging and immunosuppressive roles of HDL and other lipoproteins, low cholesterol may itself negatively impact innate and adaptive immune cells (58). Intracellular cholesterol plays a pivotal role in TLR signaling in macrophages (59). The cholesterol concentration within membrane lipid rafts significantly impacts the raft concentrations of TLR-4 and TLR-9 (59). Depletion of the ABC-A1 transporter in knockout macrophages, impacting intracellular cholesterol transport, was associated with enlarged, cholesterol-containing lipid rafts that were rich in TLR-4 and hyperresponsive to LPS (59). In lymphocytes, enrichment of cholesterol in lipid rafts was associated with increased formation of an immune synapse between signaling complexes and T-cell receptors. Low serum and low membrane cholesterol concentrations also influence natural-killer-cell function (60).

Steroid, Sex Hormone, and Vitamin D Deficiency

Adrenal insufficiency is a recognized complication in patients with sepsis and septic shock and is associated with increased mortality (61). Even though plasma cortisol concentrations are frequently raised, there is decreased responsiveness to ACTH stimulation, particularly in eventual nonsurvivors (62), suggesting the possibility of diminished reserves. As mentioned earlier, some 80% of circulating cortisol during stress is derived from plasma cholesterol (12). The contribution of hypocholesterolemia in sepsis is uncertain, as the downstream cortisol production pathway may also be compromised because of, for example the expression of StAR (steroidogenic acute regulatory protein), the rate-limiting step in steroidogenesis, which orchestrates transport of cholesterol from outer to inner mitochondrial membranes (63). Pharmacological suppression of HDL-C does, however, disrupt adrenal steroidogenesis (64). Nonetheless, human data are conflicting (65–67).

Falls in sex hormone (68) and vitamin D concentrations (69) are also well recognized in sepsis and carry prognostic and potential therapeutic implications. Pharmacological activation of the estrogen receptor β improved survival in pneumonia and peritonitis models of sepsis (70). Administration of high-dose vitamin D to critically ill patients with severe vitamin D deficiency has produced conflicting outcomes (71, 72). An association has been described between low cholesterol and low testosterone in male patients with septic shock (73); however, the causation remains unclear. Low LDL-C concentrations have also been linked to low testosterone concentrations in chronically ill patients (68).

Bile Acids

Impaired biotransformation and hepatobiliary transport of bile acids occur within hours after the induction of polymicrobial sepsis (54). In septic patients, bile acids are significantly elevated and predictive of poor outcomes (74). This appears to relate to diminished or even obstructed bile flow from the liver rather than to increased synthesis. To what extent changes in cholesterol concentrations in different body compartments during sepsis alter the complex mechanisms of bile acid metabolism remains to be elucidated.

Cholesterol Supplementation and Lipoprotein Therapies

The idea of a lipid treatment for infection is not new, whether this be cholesterol, HDL or analogs, oxysterols, or phospholipid emulsions. Indeed, Bayer took out a patent for cholesterol therapy for blackwater fever (malaria) in 1910. The possible impact of cholesterol therapy on a wide range of infectious diseases was suggested soon after (75).

Published studies remain relatively scanty and are often based on model systems. What benefit derives from the lipoprotein itself or from elevation of cholesterol concentrations is unclear.

Cholesterol nanoparticles elevated intracellular concentrations and prevented the cytotoxic effect of the pneumococcal antigen, pneumolysin, on hepatocytes (76). Administration of 25-hydroxycholesterol decreased the viral load and improved outcomes in a porcine viral pneumonitis model (77). In terms of carriers of cholesterol, intravenous application of reconstituted HDL or HDL mimetics (based on apoA-1) reduced organ damage and improved hemodynamics and survival in a variety of septic or endotoxemic rodent models (35, 78–83). Inhibition of CETP with anacetrapib preserved HDL-C concentrations and improved survival in septic mice (46). Pharmacological inhibition of PCSK9 has, however, delivered variable results. Whereas improved survival was noted in a murine polymicrobial peritonitis model (47), no protection was afforded in a murine endotoxin model (84).

Human studies are limited. Reconstituted HDL decreased proinflammatory cytokine release in human volunteers with endotoxemia (85). A multicenter study enrolling nearly 1,400 patients with presumed Gram-negative sepsis (86) reported that a 10% phospholipid–lipoprotein emulsion that contained no cholesterol, given with the aim of neutralizing endotoxin, failed to deliver any benefit. A two-center phase I/II clinical protocol in which an antiinflammatory lipid emulsion containing fish oil is being administered intravenously to septic patients with the objective of raising plasma cholesterol concentrations has been recently published (87). The impact of cholesterol infusions on lipoprotein concentrations (HDL-C, LDL-C, and VLDL-C) remains unknown. More experimental in vitro and in vivo studies are needed to address mechanisms, feasibility, dose-finding, and possible adverse events.

Statin Therapy for Sepsis—Is There a Paradox?

How can the above arguments related to cholesterol therapy be reconciled with the putative benefits of treating critical illness with statins, agents that are conventionally used to treat hypercholesterolemia? Statins inhibit HMG-CoA reductase, the rate-limiting enzyme in the mevalonate pathway that commences with acetyl-CoA. This pathway later splits into branches that synthesize cholesterol, heme A, ubiquinone, dolichol, and other isoprenoids. Statins also affect other pathways that are either directly related to or are not related to mevalonate, such as endothelial NO synthase activation (88). Thus, other than lowering cholesterol, statins have multiple other immunomodulatory, antinflammatory, and metabolic effects, such as activation of PPARs, increased production of endothelial NO, reduced synthesis of endothelin 1 and thromboxane A2, and NADPH oxidase inactivation (88–90). These may be both beneficial or harmful; for example, statin-induced myopathy has been linked to reductions in ubiquinone and thus mitochondrial functionality or to alterations in sarcolemma and/or membrane-binding proteins (91). The impact of statins on mortality in cardiovascular disease specifically related to cholesterol lowering is questioned (92).

With respect to sepsis, epidemiological studies reported an association with improved survival from sepsis in patients receiving preexisting statin treatment; however, this likely relates to population lifestyle differences (93–95). Two randomized, controlled, multicenter trials found no benefit from de novo statin therapy in sepsis (96, 97). Notably, plasma cholesterol concentrations were markedly subnormal in both the atorvastatin group and the control group (2.4 vs. 2.6 mmol/L, respectively) (96). The HARP-2 (Hydroxymethylglutaryl-CoA Reductase Inhibition with Simvastatin in Acute Lung Injury to Reduce Pulmonary Dysfunction 2) trial of patients with ARDS, of whom 40% had sepsis, showed no outcome effect from simvastatin (98). Of note, a post hoc analysis suggested that patients with a hyperinflammatory phenotype could benefit (99), indicating that non–cholesterol-lowering effects may be more pertinent. On the basis of current evidence, we cannot recommend continuation or addition of statins in sepsis; prospective randomized studies are needed to clarify their potential utility in specific patient subsets.

Conclusions

Low cholesterol concentrations are a well-recognized manifestation of sepsis and septic shock. The magnitude of hypocholesterolemia relates to the disease severity and outcome and is an early prognostic marker. Several pathophysiological mechanisms can participate in the development of hypocholesterolemia in sepsis and its impact on multiple downstream biochemical pathways. Further studies are needed to extend our knowledge about the importance and interactions of these mechanisms and the role of cholesterol with or without lipoproteins as therapeutics.