Bilirubin, a Cardio-Metabolic Signaling Molecule

Introduction

Mostly considered as a marker for hepatic disease, bilirubin, has been revealed as a cardio and metabolic protective factor that has implications for therapy. Reports of bilirubin date back to ancient Greek medicine with Hippocrates and the four humors of human health, which metabolic agents included Choleric-yellow bile, Melancholic-black bile, Sanguine-blood, and Phlegmatic-phlegm. The yellow bile was associated with the gallbladder and black bile the spleen. These intricacies are still under consideration today. Serum levels of bilirubin are thought to be derived from the breakdown of red blood cells that mostly occurs in the spleen. Bilirubin is a naturally occurring product from the catabolism of heme by heme oxygenase (HO), which releases biliverdin, carbon monoxide (CO), and iron^{1, 2}. Then, biliverdin is reduced to bilirubin via biliverdin reductase (BVR). Bilirubin is released into the blood and binds albumin where it travels to other organs or is deposited in the liver and conjugated with glucuronic acid by the UDP-glucuronosyltransferase 1A1 (UGT1A1) enzyme ^{3, 4} Conjugated bilirubin (CBR) is excreted from the liver to bile in the gallbladder and eventually released into the intestine where it is broken down by the gut microbiota into urobilinoids, mainly stercobilin, which is the predominant pigment in feces ³.

High blood levels of unconjugated bilirubin (UCB) is most widely known as the causative agent of jaundice, a yellowing of the skin and eyes. Neonatal jaundice has been described in the medical literature since the late eighteenth century and is a condition reflective of very high levels of UCB in the serum of newborn infants. Severe cases of neonatal jaundice have resulted in a bilirubin-induced neurological condition in newborns known as kernicterus, which can cause brain damage. Despite its role in jaundice, UCB has several beneficial effects on the body ⁵. The results of numerous large-scale population and epidemiological studies have correlated a protective effect of serum bilirubin levels against cardiovascular and metabolic disease. However, the molecular mechanism by which this occurs is not fully understood. The objective of this review is to highlight and discuss the unique signaling actions of BVR generation of bilirubin and how they signal to protect against cardiovascular and metabolic disease, which may lead to new therapeutics.

Bilirubin as a Signaling Molecule

One of the first relevant biological functions previously ascribed to bilirubin has been its ability to act as an antioxidant ^{6, 7}. Bilirubin can directly scavenge reactive oxygen species (ROS) as well as inhibit the production of ROS by NADP(H) oxidases ^8–11. Bilirubin also interacts with other antioxidants such as Vitamin E to further protect against enhanced ROS production^{12, 13}. While there is little doubt that bilirubin is a potent endogenous antioxidant, questions remain as to whether the antioxidant effects of bilirubin can adequately explain all of the beneficial cardiovascular and metabolic actions ^14–16.

Examination of the structure of UCB revealed that it shares similarities to other known activators of peroxisome proliferator-activated receptor α (PPARα) such as WY 14,643 and fenofibrate ¹⁷. Studies in receptor-less COS-7 cells that were transiently transfected with PPARα or empty vector demonstrated that a dose-dependent increase of biliverdin or bilirubin specifically activated a PPARα-dependent reporter gene ¹⁷. However, biliverdin is rapidly produced into bilirubin by Biliverdin Reductase A (BVRA) ¹⁸, which indicates that BVRA generation of bilirubin mediates PPARα transcriptional activity. In a similar analysis comparing the same level of WY 14,643, fenofibrate, and biliverdin, Stec et al. found that biliverdin increased the PPARα-dependent reporter gene to the same extent as the known activator, fenofibrate ¹⁷. However, WY 14,643 at this reporter had significantly higher activity. Fenofibrate and bilirubin have cardiovascular protective properties.

It is important to delineate serum conjugated (CBR) from unconjugated (UCB) bilirubin, as CBR is not preventive of cardiovascular or metabolic diseases, and only elevated UCB levels are correlated with cardiovascular and metabolic protection ¹⁹. Additionally, only UCB has been shown to activate PPARα ¹⁷. UCB has been demonstrated to active PPARα in the 50 μM range which is in the physiological range of UCB reported in patients with Gilbert’s syndrome ^{19, 20}. UCB may signal similarly to fenofibrate and other fibric acid derivatives. Also, bilirubin has anti-diabetic effects that are not typically observed with fibrate class of drugs. However, there has only been one study performed showing this comparison, and it is yet to be thoroughly tested.

Bilirubin as a Selective PPAR Modulator (SPPARM)

Compounds that target the PPARs may simultaneously activate all three PPARs (PPAR pan agonists) or can have selective modulation of a single PPAR (SPPARM) ^21–24. The latter may be a potent inducer of some activities with reduced unwanted effects. When a cognate ligand binds to PPARα, a conformation change occurs in the protein, which results in enhanced phosphorylation, transactivation, and DNA binding to PPAR response elements (PPREs) in promoters or enhancers of regulated genes (Figure 1) ^{25, 26}. PPARα ligands have differential binding affinities, which may result in a slight conformational change that can potentially lead to divergent PPARα gene regulation, which has been shown with fenofibrate and WY 14,643 ^{27, 28}. Stec et al. showed that biliverdin, WY 14,643 and fenofibrate activated PPARα to different extents ¹⁷, which may be due to variations in ligand binding affinity causing different gene regulation. Demonstrating these slight variances in gene regulation, the fibrates have been shown to be better at reducing inflammation than WY 14,643 and are typically used in treating inflammatory hyperlipidemia and fatty liver disease ^29–31. While WY 14,643 does have anti-lipemic activity against hyperlipidemia, it actions at reducing inflammation was shown to be lower ³². However, WY 14,643 has been shown to be more efficient at lowering blood glucose levels ³³. UCB may likewise regulate a unique subset of PPARα target genes and function as a SPPARM to mediate its anti-lipemic, -inflammatory, and -diabetic properties, and potentially its cardiovascular protective functions.

Figure 1: Bilirubin as a Selective PPARα Modulator (SPPARM).

Bilirubin binds to the ligand-binding pocket of PPARα similar to other known activators WY-14,643 and fenofibrate. Also, bilirubin is also a potent antioxidant. Bilirubin has been shown to decrease the phosphorylation of serine 73 (Ser73) to increase transcription activity of PPARα ^{34, 36}. The zinc finger (Z) DNA binding domains of PPARα bind at promoters of target genes. Bilirubin stimulates the promoter of several PPARα target genes including fibroblast growth factor 21 (Fgf21), carnitine palmitoyltransferase I (Cpt1), and cytochrome P450 4A (Cyp4A). WY 14,643 is stronger at inducing the cluster of differentiation 36 (Cd36) and to a lower extent glucose transporter 4 (Glut4) and Fgf21 promoters. Fenofibrate primary target is Cpt1. These genes play a specific role in anti-diabetic, anti-lipemic, anti-inflammatory, and anti-hypertensive actions of each compound.

There is strong evidence from in vivo models demonstrating that UCB acts as a SPPARM. Bilirubin treatment decreased body weight, body fat and increased lean body mass in wild-type but not PPARα knockout mice ¹⁷. Bilirubin treatment also resulted in significant decreases in fasting blood glucose and increases in hepatic and serum fibroblast growth factor 21 (FGF21) levels were blocked entirely in PPARα knockout mice and were not observed in fenofibrate-treated mice ¹⁷. Gilbert’s syndrome is a condition of moderate hyperbilirubinemia (18–58 μM) in humans caused by polymorphisms within the promoter of the UGT1A1 gene ^{34, 35}. Moderately hyperbilirubinemic Gilbert’s mice containing the human UGT1A1*28 polymorphism which have serum bilirubin levels in the 30 μM range are protected against high fat diet-induced increases in adiposity, hyperglycemia, hyperinsulinemia, and hepatic steatosis ³⁶. These mice also exhibit decreased hepatic expression of fatty acid synthase (Fasn), sterol regulatory element binding protein-1 (Srebf1), and acetyl-CoA carboxylase (Acaca) all of which are fundamental genes involved in hepatic lipid synthesis ³⁶. The humanized UGT1A1*28 mice also exhibit enhanced hepatic levels of total PPARα protein and decreased levels of phosphorylated serine 73 (SerP⁷³) PPARα as compared to wild-type mice fed a high-fat diet ³⁶. Ser(P)⁷³ is the primary site which PPARα is targeted for ubiquitin-mediated degradation; thus, lowering phosphorylation at this site would decrease PPARα turnover and increase its activity ³⁴. The decrease in Ser(P)⁷³ PPARα and increased activity was reflected in an elevated expression of several PPARα target genes such as Fgf21 and Cyp4A in the liver of humanized UGT1A1*28 mice as compared to wild-type mice ³⁶. However, the mechanism by which bilirubin decreases Ser(P)⁷³ PPARα is not currently known but may be mediated through regulation of glycogen synthase kinase-3β (GSK3β) ³⁴

Bilirubin Signaling in Metabolic and Cardiovascular Diseases

Bilirubin has a wide variety of cardiovascular actions in many different organs throughout the body (Figure 2). In the heart, bilirubin increases nitric oxide (NO) bioavailability through its effects to lower ROS production ^{35, 37}. In the vascular wall, bilirubin attenuates atherosclerotic plaque formation through its impact on adhesion molecules as well as cholesterol metabolism ^38–42. In the kidney, bilirubin preserves glomerular filtration rate (GFR), and renal blood flow through an increase in NO bioavailability ^{43, 44}. In the liver and adipose, bilirubin acts through PPARα to increase fat burning ^{17, 36}. These studies posit bilirubin as an influential mediator in metabolic and cardiovascular diseases.

Figure 2: Cardiovascular and Metabolic actions of Bilirubin.

In the heart, bilirubin lowers reactive oxygen species (ROS) and increases nitric oxide (NO) levels to improve coronary blood flow. Bilirubin may also act through PPARα to enhance fatty acid metabolism and afford cardioprotection. In blood vessels, bilirubin attenuates cholesterol deposition and immune cell infiltration to decrease plaque formation. In the liver, bilirubin and biliverdin reductase A (BVRA) phosphorylates AKT which inhibits glycogen synthase kinase 3β (GSK3β) preserving PPARα function to increase fibroblast growth factor 21 (Fgf21) and carnitine palmitoyltransferase I (Cpt1) for fat burning and protection against hepatic steatosis. In adipocytes, BVRA production of bilirubin activates PPARα to reduce lipid size. In the kidney, bilirubin improves glomerular filtration rate (GFR), renal blood flow (RBF), and nitric oxide synthase (NOS) levels while inhibiting ROS levels.

Several large-scale population studies have shown that the degree of overweight and obesity are negatively correlated with serum UCB levels in both men and women ^45–48. Serum bilirubin levels are also know to decline with aging which corresponds to increases in body weight and cardiovascular disease risk especially in women ^{49, 50}. Interestingly, higher serum UCB levels are also associated with a more youthful appearance in women but not in men ⁵¹. It was also reported that short-term weight loss during administration of sibutramine in combination with diet and exercise increased bilirubin levels in proportion to weight change with the effect being greater in men than in women ⁵². In addition, moderate hyperbilirubinemia associated with Gilbert’s polymorphism was associated with lower body mass index (BMI), hip circumference (HC), fat mass, and lipid profile as compared to healthy controls and type II diabetic patients ⁵³. These data suggest that plasma UCB levels can have an impact on body weight and that weight loss itself can increase serum bilirubin levels.

Studies of the National Health and Nutrition Examination Survey (NHANES) have found that higher serum bilirubin levels protect against the development of stroke and diabetes mellitus ^{54, 55}. Similar results have also been obtained in smaller population studies ^{56, 57}. The Gilbert’s polymorphism (UGT1A1*28) was reported to protect against coronary artery disease in the Framingham Heart Study⁵⁸. One mechanism for the cardioprotection afforded to Gilbert’s individuals is that the high bilirubin levels reduce ROS production ^{19, 59}. Recent studies have also demonstrated that Gilbert’s individuals have alterations in metabolic pathways such as the AMP-activated protein kinase (AMPK) pathway similar to those observed in humanized UGT1A1*28 mice ^{20, 36}.

There have been several studies which have examined the role of bilirubin in hypertension. Studies in the hyperbilirubinemic Gunn rat have demonstrated a protective role in deoxycorticosterone acetate (DOCA)-salt and angiotensin-II-dependent hypertension ^{60, 61}. However, the serum levels of bilirubin in the Gunn rat are extremely high (~30 mg/dL or 500 μM) due to complete loss of hepatic UGT1A1 activity in this strain. In order to determine the role of a more physiological increase in serum bilirubin similar to what is observed in Gilbert’s patents, we developed a mouse model of moderate hyperbilirubinemia and demonstrated that these mice were protected from angiotensin-II-dependent hypertension and decreased renal blood flow and glomerular filtration rate ^{43, 44}. The effect of moderate hyperbilirubinemia to lower blood pressure in angiotensin-II dependent hypertension was not due to its antioxidant actions suggesting that other actions of bilirubin such as its SPPARM activity may be responsible for the blood pressure lowering actions of moderate hyperbilirubinemia ¹⁴. However, further studies examining the ability of bilirubin to lower blood pressure in angiotensin II-dependent hypertension in the absence of PPARα signaling are needed.

Population studies have also demonstrated a protective role of serum bilirubin levels against the development of diabetes and related pathologies such as diabetic nephropathy and limb amputations^{55, 57, 62–65}. Increasing serum bilirubin levels with protease inhibitors such as atazanavir was also reported to improve vascular function in diabetic patients and improve renal hemodynamics in hypertensive mice ^{43, 44, 66}. Further studies in both diet-induced obese (DIO) and db/db mice have demonstrated that bilirubin administration improves hyperglycemia and obesity ^{40, 67}. While the specific effect of bilirubin signaling was not evaluated in these studies, UCB treatment was associated with activation of insulin-signaling pathways, and decreases in the levels of inflammatory cytokines as well as markers of endoplasmic reticulum (ER) stress ⁶⁷. These studies highlight that bilirubin may act to improve insulin sensitivity through several different mechanisms including anti-ER stress, anti-inflammatory as well as through its actions as a SPPARM (Figure 1).

Alterations in plasma lipids and cholesterol levels can promote cardiovascular disease. Studies in individuals with Gilbert’s polymorphism as well as animal studies in models of moderate hyperbilirubinemia have demonstrated the effects on lipid profile and cholesterol levels which contribute to the protective actions of UCB. Gilbert’s individuals have been reported to have lower serum cholesterol, low-density lipoprotein (LDL), and oxidized LDL (ox-LDL) ^{19, 59, 68, 69}. Similar effects on serum cholesterol and LDL levels were observed in both humanized Gilbert’s mice as well as dietary-induced obese mice treated with UCB ^{36, 40}. The degree to which bilirubin signaling impacts these effects on serum lipids is not known; however, PPARα agonists like fenofibrate have long been known to decrease the levels of serum fatty acids, LDL, and total cholesterol ^{70, 71}.

Hyperlipidemia and increased levels of serum fatty acids and LDL are known risk factors for the development of cardiovascular disease. Several studies in rodent models of hypertension have demonstrated a protective role of the PPARα agonist fenofibrate in lowering blood pressure and protecting against end-organ damage ^72–75. Additional studies in non-human primates have demonstrated the potency of PPARα agonist to reduce plasma lipids and lower cardiovascular disease ⁷⁶. Fenofibrate treatment has been demonstrated to have positive effects on vascular function in patients with type II diabetes; however, the lipid-lowering effects of this drug have not translated into significant protection from cardiovascular disease in human patient populations ^77–79. Bilirubin treatment may offer better therapeutic potential as opposed to treatment with a PPARα agonist since it has antioxidant, anti-lipemic, and anti-inflammatory properties which also contribute to its protection against cardiovascular disease (Figure 2).

While population studies have demonstrated a correlation between serum bilirubin levels and the development of cardiovascular and metabolic disease, it is still possible that serum bilirubin levels may act as a marker of disease rather than play a causative role. In order to definitively prove that genetic mutations which alter serum bilirubin levels play a causative role in the development of cardiovascular and metabolic disease, specific genetic association studies are needed. Mendelian randomization (MR) is an epidemiological approach based on the fact that individuals inherit genetic variants randomly from their parents so that the role of various gene markers can be tested without causal limitations inherent population studies. The results of MR studies examining polymorphisms of UGT1A1 which increase serum bilirubin levels have yielded mixed results with no associations found in stroke, ischemic heart disease or cardiovascular disease in general but positive association found with type 2 diabetes ^80–83. While the result of some of the initial MR studies indicate the elevated serum levels of bilirubin may not be causally related to all disease phenotypes initially reported by population studies, more MR studies looking at other cardiovascular phenotypes in a broader range of populations are needed to fully elucidate the role of bilirubin in the prevention/cause of cardiovascular and metabolic diseases.

Biliverdin Reductase Signaling Actions

BVR is the enzyme responsible for the reduction of biliverdin to bilirubin. Biliverdin reductase A (BVRA) is the major isoform present in adult tissues while the B isoform (BVRB) is the isozyme present during fetal development ¹. Recent studies in BVRA global knockout mice reveal a near absence of plasma bilirubin and green gall bladders indicative of very high levels of biliverdin ⁸⁴. Furthermore, the global BVRA deficient mice exhibit higher plasma concentrations of cholesteryl ester hydroperoxides (CE-OOH) indicative of higher levels of endogenous oxidative stress ⁸⁴. However, the effect of global loss of BVRA on metabolism has not been studied.

Hepatocyte-specific loss of BVRA resulted in marked steatosis decreased in glycogen storage and alterations in liver insulin sensitivity ⁸⁵. Moreover, loss of hepatocyte BVRA results in elevated GSK3β activity by a reduction in Ser9 phosphorylation which is inhibitory ⁸⁶. BVRA regulates hepatic GSK3β activity though Akt1-mediated increases in Ser9 phosphorylation ⁸⁷. The rise in GSK3β activity due to the loss of BVRA-mediated Ser9 phosphorylation resulted in increased Ser(P)⁷³ PPARα and decreased target genes such as Fgf21 and Carnitine Palmitoyltransferase 1A (Cpt1a) ⁸⁵. Thus, BVRA can regulate Ser(P)⁷³ PPARα through its generation of UCB as well as by its direct interaction with Akt1 regulating phosphorylation of GSK3β (Figure 2). The novel regulation of GSK3β phosphorylation by BVRA may be an important protective pathway in hypertension-induced cardiac injury as well as renal ischemia-reperfusion injury as GSK3β activation has been reported to play a significant role in each of these pathologies ^{1, 88–90}. Recently, CRISPR targeting of BVRA in proximal tubules kidney cells showed that the loss of BVRA caused significantly higher lipid accumulation and lipotoxicity ⁹¹. Bilirubin may have numerous benefits for the kidney and kidney damage.

BVRA is an interesting protein in addition to its reductase activity, and it also has other unique domains which allow it to interact with AKT, protein kinase C (PKC), mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK), and insulin receptor pathways ^{1, 88–90}. Several studies have demonstrated that peptide regions of the human BVRA protein can serve as stimulators of insulin receptor kinase (IRK) and glucose uptake as well as inhibitors of IRK and glucose uptake ⁹². The C-terminal peptide, KYCCSRK, was reported to stimulate phosphorylation of IRK and enhance glucose uptake in human embryonic kidney cells ⁹². Further studies using nanoparticles containing the KYCCSRK peptide demonstrated that administration of this peptide markedly improved glucose uptake in control and type II diabetic ob/ob mice ⁹³. The BVRA peptide nanoparticles increase glucose uptake through pathways that mediate phosphorylation of both IRK and GSK3β ⁹³. These studies highlight the potential for the development of BVRA based peptides for the treatment of diabetes. While BVRA can signal alone or through distinct peptide sequences, it is also responsible for the intracellular generation of UCB from the breakdown of heme-containing P450 proteins. BVRA production of UCB from biliverdin is essential for the activation of PPARα, and for this reason, may offer protection against cardiovascular disease ^{15, 85}. However, more work is needed to determine the impact of the BVR isozymes in metabolic and cardiovascular disorders.

Conclusions

In summary, the traditional view of bilirubin as a toxic waste product or just an endogenous antioxidant molecule needs to be revised to include its novel function as a signaling molecule capable of activating nuclear hormone receptors that regulate cardiovascular and metabolic function. The signaling action of bilirubin, especially delineating from conjugated and unconjugated forms, needs to be better defined in future preclinical studies utilizing animal models in which various signaling targets of bilirubin are altered. While data from large population studies have demonstrated a negative relationship between serum bilirubin levels and cardiovascular disease, new genetic approaches such as Mendelian randomization are needed in an extensive section of diverse populations with several different cardiovascular and metabolic phenotypes. In the future, these types of studies will need to be performed on other genes associated with bilirubin metabolism so that the causative nature of increased serum bilirubin levels on the development of cardiovascular and metabolic diseases can be fully elucidated. Strategies to increase serum UCB levels in individuals most at risk for these pathologies need to be developed to take full advantage of the numerous beneficial actions of bilirubin. This may occur by targeting enzymes that regulate the turnover of bilirubin, such as BVRA or UGT1A1 or treatment with novel formulations of bilirubin which are easier to deliver. These types of studies will require carefully planned clinical trials in patients, which will most likely have reduced serum bilirubin levels as well as an enhanced risk of cardiovascular and metabolic disease.