Genetic and microbiome influence on lipid metabolism and dyslipidemia

Abstract

Disruption in the metabolism of lipids is broadly classified under dyslipidemia and relates to the concentration of lipids in the blood. Dyslipidemia is a predictor of cardio-metabolic disease including obesity. Traditionally, the large interindividual variation has been related to genetic factors and diet. Genome-wide association studies have identified over 150 loci related to abnormal lipid levels, explaining ~40% of the total variation. Part of the unexplained variance has been attributed to environmental factors including diet, but the extent of the dietary contribution remains unquantified. Furthermore, other factors are likely to influence lipid metabolism including the gut microbiome, which plays an important role in the digestion of different dietary components including fats and polysaccharides. Here we describe the contributing role of host genetics and the gut microbiome to dyslipidemia and discuss the potential therapeutic implications of advances in understanding the gut microbiome to the treatment of dyslipidemia.

INTRODUCTION

Defects in the metabolism of lipids are broadly classified under dyslipidemia and are characterized by elevation of circulating cholesterol, triglycerides (TGs), or both or by low high-density lipoprotein (HDL) levels.

Observational studies have shown that 60–70% of adults have lipid levels outside of the recommended range (33, 68, 76). Dyslipidemia is an important risk factor for cardio-metabolic diseases (61). Because the early stages of dyslipidemia present no symptoms, dyslipidemia often remains undiagnosed until patients begin presenting vascular complications (55).

Lipid levels are known to have an important genetic contribution and the past decade has seen considerable advances in this field (34). Genome-wide association studies (GWAS) have identified over 500 single-nucleotide polymorphisms (SNPs) in 167 loci associated with altered lipid levels explaining ~40% of the individual variation (34). However, the remaining 60% of the variation in circulating lipids remains unexplained and is either due to undiscovered rare variants or to environmental factors. Some of these other factors that are known to influence lipid levels and metabolism include dietary fat intake, alcohol intake, smoking (14), and levels and intensity of physical activity [e.g., (48, 59)]. In addition, in recent years, the gut microbiome has been shown to be strongly associated with a number of disorders, including inflammation, insulin resistance, and fatty acid metabolism (12), and is emerging as an important candidate to account for the unexplained variation in lipid levels in humans. Therefore, to understand the molecular mechanisms underlying interindividual variation in lipid metabolism it is important to understand genetic and nongenetic mechanisms as both are likely to contribute.

In this review, we first give an overview of the genetics of common forms of dyslipidemia in the general population. Then we discuss the mechanism whereby the gut microbiome acts as a modulator of lipid metabolism, and we present some of the animal and human studies that have been conducted. These two types of influence on lipid metabolism, genetic and nongenetic, may not be directly linked to each other, and further investigation will be needed to establish this.

Finally, we discuss potential future implications of this area of research.

GENETICS OF LIPID METABOLISM

A large proportion, up to 40%, of the interindividual variation in lipid levels can be explained by genetic factors.

Much of the work preceding GWAS relating to the genetics of lipid metabolism derives from the study of uncommon Mendelian traits such as familial hypercholesterolemia (FH). Such studies revealed the important role of mutations in the low-density lipoprotein (LDL) receptor (LDLR) gene leading to the plasma accumulation of cholesterol ester-laden LDL particles. Another form of FH is caused by defects in the APOB gene that lead to decreased clearance of LDL and is now established as a significant cause of coronary heart disease (53). Yet another form is due to mutations in the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene (1). The important role of these genes not only in FH but in common complex trait lipid metabolism has been subsequently confirmed by several GWASes.

In the past few years large-scale human GWASes including thousands of samples [Kathiresan et al. > 15,000 (43), Teslovich et al. >100,000 (74), and The Global Lipids Consortium >90,000 (36)] have identified a substantial number of loci with several common pathways associated with the prevalence, severity, and progression of lipid disorders and lipid markers. In total, more than 70 GWASes found over 500 SNPs in 167 loci across the human genome that are associated with at least one lipid trait (Table 1, Fig. 1). The Circos plot in Fig. 1 presents the list of genes from GWASes associated to lipid levels with a significance threshold of 10⁻⁸. The name of the gene is color coded according to the lipid trait most predominantly associated to it: blue (TGs), red (HDL), green (LDL) or orange [total cholesterol (TC)].

Table 1.

Principal genome-wide association studies reporting SNPs linked to lipid levels

Study	Lipid Traits	Study Design	Participants	Loci Identified	Key Findings
Kathiresan 2008 (42)	LDL, HDL, TG	diabetic vs. nondiabetic	Discovery: 2,758 CEU	18	Confirmed previously identified lipid loci and identified 6 new loci
			Replication: 18,544 CEU
Kathiresan 2009 (43)	LDL, HDL, TG	population and diabetic cohorts	Discovery: 19,840 CEU	30	The 30 loci explain 5% of the variation in lipid traits
			Replication: 20,623 CEU
Aulchenko 2009 (6)	TC, HDL, LDL	twin cohorts and population cohorts	17,083 CEU & 714 Orcadian	22	There is sexual dimorphism in lipid traits
Chasman 2009 (19)	TC, LDL, HDL, TG	population cohorts	Discovery: 17,296 CEU	43	Some variants related only to particle size, indicating a possible new niche for findings
			Replication: >5,000 CEU
Teslovich 2010 (74)	TC, LDL, HDL, TG	population cohorts	Discovery: >100,000 CEU	95	Most of the identified loci are common across populations
			Replication: >24,000 EAS and SAS
			>8,000 AA
			>7,000 CEU
Global Lipids Genetics Consortium (36)	TC, LDL, HDL, TG	population cohorts	188,577 CEU & 7,898 non-CEU	157	There are also nonlipid-related loci associated with lipid levels
Surakka 2015 (69)	TC, LDL, HDL, TG	population cohorts	62,166 CEU	167	Low-frequency variants associated with lipid traits are population specific; also, they identified 10 new loci

TG, triglycerides; HDL, HDL cholesterol; LDL, LDL cholesterol; TC, total cholesterol; CEU, Caucasian; AA, African Americans; SAS, South east Asian; EAS, East Asia.

Fig. 1.

Circos plot of genes associated to lipid levels through published genome-wide association studies (GWASes). The Circos plot displays the genes that have been reported in GWASes with a significance threshold of 10⁻⁸. The name of the gene is color coded according to the lipid trait most predominantly associated to it: blue (triglycerides), red (HDL), green (LDL), or orange (total cholesterol).

The GWASes revealed that lipid disorders share common genetic risk factors, such as mutations in known lipid regulators and proteins including APOE (LDL-C), CETP (HDL-C), and LPL (TG), and with proteins from other metabolic pathways such as GCKR for glucose metabolism (26).

Of particular interest is the case of PCSK9. Secreted into the plasma by the liver, the proteinase K-like serine protease encoded by the PCSK9 gene binds the LDL receptor at the surface of hepatocytes, thereby preventing its recycling and enhancing its degradation in endosomes/lysosomes, resulting in reduced LDL-cholesterol clearance (67). Several PCSK9 variants have been identified; some of them are gain-of-function mutations causing hypercholesterolemia by a reduction of LDLR levels, while others are loss-of-function variants associated with a reduction of LDL-cholesterol (LDL-C) levels and a decreased risk of CHD (1, 23). Since its discovery due to its link to a form of FH it has been the subject of much research and has led to the development of new lipid lowering drugs reducing LDL-cholesterol levels through the inhibition of PCSK9. Two anti-PCSK9 monoclonal antibodies have received Food and Drug Administration and European Medicines Agency approvals: Alirocumab and Evolocumab (29), thus highlighting the value that genetic research has brought to the development of novel treatments for hypercholesterolemia.

Further studies have explored the population heterogeneity and the impact of low frequency variants (minor allele frequency <5%) on lipid metabolism (69, 83). The total variation in lipid levels explained by common and rare variants is ~40% (34), indicating that a large proportion of the variation remains unexplained (21, 36, 60). Emerging evidence from both animal and human studies demonstrates that the gut microbiome plays a key role in lipid metabolism. The gut microbiome affects many pathways, including hepatic lipid and bile metabolism and cholesterol transport (35).

POSSIBLE ROLE OF THE GUT MICROBIOME IN LIPIDS METABOLISM

The gut microbiome comprises 100 trillion microorganisms including eukaryotes, archaea, bacteria, and viruses (78, 79). Its composition is regulated by a complex interplay of the host genome, the colonic milieu, and diet (40). During digestion in the colon, the host and the microbes coproduce a wide range of enzymes, hormones, vitamins, and other chemicals, such as short chain fatty acids (SCFAs), bile acids, conjugated linoleic acids (CLAs), among others, that are essential for regulating multiple host-microbiome pathways including lipid levels (2). A schematic drawing of diet/gut microbiome/lipid interaction metabolism is depicted in Fig. 2.

Fig. 2.

Schematic drawing of diet/gut microbiome/lipid interaction metabolism. Different pathways have been proposed for the influence of diet/gut microbiome interaction on lipid levels. Experimentally, gut microbiome has been shown to modify both dietary components and metabolic precursors producing secondary metabolites. In postprandial state, macronutrients from dietary intake [polyunsaturated fat (PUFA), dietary fiber, and carnitine from red meat] enter the digestion process; at the same time as bile acids are released into the gut from the gall bladder (1). In the gut, microbial enzymes produce short chain fatty acids (SCFAs), conjugated linoleic acid (CLA), trimethylamine N-oxide (TMAO), and secondary bile acids (BAs) from dietary fiber, PUFA, carnitine, and bile acids, respectively (2). Because of cross-feeding (i.e., the process whereby some microbial metabolites can modulate gut microbial composition), CLA production increases SCFA producers, indirectly affecting the overall SCFA production. The secondary metabolites produced in the gut exert their systemic effects in different parts of the body. The metabolites affecting lipid metabolism have their receptors mostly in the liver, triggering metabolic signaling cascades (3). SFCAs and CLA interact with peroxisome proliferator-activated receptors (PPARs) leading to higher HDL levels, higher triglyceride (TG) and VLDL clearance, and higher lipolysis. Interaction of secondary bile acids with FXR receptors correlates to higher HDL and lipolysis and lower VLDL levels. The effects of TMAO in lipid levels are mainly due to HDL lowering disrupting lipid metabolism homeostasis.

One mechanism through which gut microbes affect lipid metabolism involves the fermentation of nondigestible carbohydrates. The human gut is not capable of breaking down many common forms of complex carbohydrates (dietary fiber), whereas a subset of anaerobic gut bacteria can ferment compounds such as pectins, gums, hemicelluloses, and galactose-oligosaccharides (54). SCFAs are among the best known bacterial products derived from fermentation of nondigestible carbohydrates (e.g., dietary fiber) that transits the small intestine unaffected, only to be broken down by a subset of anaerobic bacteria in the cecum and proximal colon. These SCFA, including butyrate, propionate, and acetate, regulate intestinal immune homeostasis, provide rich sources of energy for the host (54), and influence lipid and cholesterol metabolism (25). The host energy balance is mediated via a ligand-receptor interaction of the SCFAs with different G protein-coupled receptors (GPCRs), such as GPR41 [also known as FFAR3 (13)] and GPR43 [also known as FFAR2 (44)]. The gut hormones peptide YY (PYY) and glucagon-like peptide-1 (GLP-1) have also been implicated in this action (9).

SCFAs also regulate the balance between fatty acid synthesis, fatty acid oxidation, and lipolysis in the tissues of the body via the peroxisome proliferator-activated receptors (PPARs) (16). Fatty acid oxidation is activated by SCFAs, while de novo synthesis and lipolysis are inhibited, resulting in a reduction of the concentrations of circulating free fatty acids, as demonstrated by colonic infusions of SCFAs in humans (17). Finally, studies in germ-free (GF) animals have demonstrated that the gut microbiome is essential for immune cell recruitment and differentiation (37). The exact mechanism by which gut bacteria, such as Bifidobacteria, modulate the immune response is not yet fully understood. One potential mechanisms includes SCFAs, which directly increase the abundance of T regulatory cells in the gut (5, 32).

In addition to SCFAs other mechanisms by which the gut microbiome can influence lipid levels include CLAs and bile acids. CLAs are produced by the action of Bifidobacteria, Roseburia, and Lactobacillus on polyunsaturated fatty acids (PUFAs) (10, 28, 52) from omega-3 rich food sources (58). SCFAs influence lipid metabolism by the activation of the PPAR isoforms (8, 57). Such activation triggers an increase in lipolysis, higher β-oxidation rate, and mitochondrial biogenesis that lowers lipid levels (24, 38). CLA can also affect lipid metabolism by promoting higher levels of SCFA bacteria producers, creating a synergistic activation of PPAR (18).

Bile acids are steroid acids that are produced in the liver from cholesterol and secreted in bile. Their main role is to aid digestion of dietary fats in the small intestine and regulate cholesterol homeostasis (51, 82). Evidence from animal studies shows that colonic bacteria produce secondary bile acids from the pool of bile salts secreted into the intestine. A fraction of these bacterial compounds is absorbed into the bloodstream and can modulate hepatic and/or systemic lipid and glucose metabolism through FXR (64) or TGR5, respectively (41). TGR5 signaling induces intestinal glucagon-like peptide-1 (GLP-1) release, leading to improved liver and pancreatic function and enhanced glucose tolerance (75).

Another molecule implicated is angiopoietin‐like protein 4 (ANGPTL4), also known as fasting-induced adipose factor (FIAF). ANGPTL4 is a ubiquitously expressed glycoprotein that plays an important role in lipid metabolism by inhibiting the activity of the enzyme lipoprotein lipase (27). Animal studies have suggested that the ANGPTL4 protein is modulated by the gut microbiome, possibly through increased concentrations of SCFAs (11). However, recent human studies of fiber supplementation have failed to demonstrate an effect on ANGPTL4 (11).

Microbial metabolites may also influence lipid levels directly. An example is trimethylamine N-oxide (TMAO), a metabolite derived from dietary choline and phosphatidylcholine (lecithin) through the action of the gut microbiota. TMAO is associated with atherosclerosis and cardiovascular risk (62, 71, 81), possibly by impacting lipid absorption and cholesterol homeostasis and decreasing the total bile acid pool size (47).

EXPERIMENTAL EVIDENCE OF THE LINK BETWEEN THE GUT MICROBIOME AND CIRCULATING LIPIDS

Much of our knowledge of the functional role of the gut microbiome on various health-related traits, including lipid levels, comes from animal models, though findings have then been back-translated into humans mostly by means of cross-sectional studies.

A comparison of GF vs. conventionally raised (CONV-R) mice has supported a role for the gut microbiome in affecting lipid levels (80) (Fig. 3). Velagapudi and collaborators (80) found that CONV-R mice compared with GF mice had lower triglycerides in blood and higher triglycerides in adipose tissues and liver after a 4 h fast. This difference is consistent with a higher lipid clearance in CONV-R mice that is possibly masking the elevation of serum triglycerides levels. This is consistent with their previous report that the gut microbiota suppresses FIAF, an inhibitor of lipoprotein lipase that promotes higher lipid clearance (7). Dietary intervention changes these patterns. Rabot and collaborators (63) found that feeding GF and CONV-R mice a high-fat diet reversed this association. Indeed, they observed higher circulating TG, HDL, and TC levels in CONV-R mice. Animal studies have shown that both genotype and gut microbiome influence metabolic phenotypes such as liver TG. A study in mice found how the microbiome and host genome interact to influence metabolic phenotype (49).

Fig. 3.

Examples of recent association studies between lipid-related traits and gut microbiome in murine models.* TG, triglycerides; HDL, HDL cholesterol; LDL,  LDL cholesterol; TC,  total cholesterol; CONV-R, conventionally raised; CONV-D,  conventionalized; GF, germfree;↑, high; ↓, low. *Circulating lipid levels unless otherwise specified.

Further experimental work has been conducted on mice treated with pro/prebiotics. Supplementation with probiotic Bifidobacterium spp. and Lactobacillus plantarum in mice led to significant reductions in circulating levels of TG, LDL, and TC and with higher HDL levels (4, 56). Similarly, a decrease in LDL (86) and TG circulating levels and an increase in HDL was observed in obese, dyslipidemic mice taking L. plantarum and Lactobacillus curvatus supplementation (72). These studies suggest that lipid metabolism could be changed through microbiome modulation by diet, which could have important therapeutic implications.

Recent studies in humans have described the association of the gut microbiome and lipid levels and identified several taxa correlating with circulating TG, LDL, TC, and HDL levels (20, 31, 50) (Fig. 4). To date, the majority of these studies have compared Type 2 diabetic (T2D) or obese individuals with healthy controls. They have consistently reported a negative correlation between TG levels and gut microbiome diversity, a general measure of gut health (20, 31, 50). Circulating TG levels in obese and T2D individuals were also found to associate with lower abundance of Clostridium species, independently of body mass index (BMI) (40). Negative correlations with gut microbiome diversity were also reported for LDL (20). On the other hand, HDL cholesterol was found to be positively associated with microbial richness (31, 50) and with abundance of Clostridium species (40).

Fig. 4.

Examples of recent association studies between lipid-related traits and gut microbiome in humans. TG, triglycerides; HDL, HDL cholesterol; LDL, LDL cholesterol; TC, total cholesterol; ↑,  high; ↓, low; spp, subspecies; T2D, Type 2 diabetes.

A recent study by Fu and coworkers (31) examined for the first time the association between the gut microbial composition and lipids in the general population. They assessed gut microbial composition in 893 individuals from the LifeLines cohorts using 16s rRNA sequencing. The gut microbiome explained 6% of the variation in triglycerides and 4% of the variation in HDL at the population level. Gut microbiome diversity was found to correlate negatively with TG levels and positively with HDL levels in line with several reports linking increased diversity to more favorable cardiometabolic outcomes (31). They also identified several taxa to be associated with TG and HDL independently of BMI. These include Eggerthella, Pasteurellacea, and Butyricimonas (31).

One of the most cited microbiome factors in obesity is the shift in the ratio of the Firmicutes and Bacteroidetes phyla. Together they accounts for 90% of the adult gut microbiome. Dietary supplementation with Bacillus subtilis increases the Firmicutes and decreases the Bacteroidetes proportions, and these changes are accompanied by decreases in circulating TGs and free fatty acids (22). Reversal of this ratio has not yet been demonstrated, and whether it can alter cholesterol levels in humans remains disputed.

There is now considerable evidence pointing to the importance of the gut microbiome in influencing lipid levels, and modulation of the gut microbiome may offer an opportunity to address hyperlipidemia. Arguably probiotic supplementation may be seen as one such intervention. A recent meta-analysis of 15 published randomized controlled trials found that probiotic Lactobacillus consumption significantly reduced TC by 0.26 mmol/l [95% confidence interval (CI), −0.40 to −0.12] and LDL-C by 0.23 mmol/l (95% CI, −0.36 to −0.10) (85). Similar results have been seen in individuals with T2D (70) and pregnant women with gestational diabetes (73). It is currently unclear which probiotics establish themselves in the gut and exert a direct effect on the resident community (65). However, the results from these systematic reviews suggest an important metabolic influence and point to the therapeutic potential of probiotic bacteria on lipid levels.

Several studies in humans have suggested that a high-fat diet increases total anaerobic microflora and counts of Bacteroides (84). At the same time, studies in mice suggest that the fatty acid metabolism by gastrointestinal microbes modifies fatty acid composition of the host (45). Saturation metabolism of polyunsaturated fatty acids, a representative mode of lipid metabolism by gastrointestinal microbes, is a detoxifying metabolism of anaerobic bacteria, such as lactic acid bacteria, that reside in the colon and intestine. This process transforms growth-inhibiting free polyunsaturated fatty acids into less toxic free saturated fatty acids (45). A correlation between a decrease in the negative effects of trans fatty acids and an increase in gut Lactobacillus, along with lower levels of proinflammatory cytokines, has also been observed in rats (3). Therefore, functional investigations of lipid metabolisms of gastrointestinal microbes may provide new methods for improving our health by altering lipid metabolism.

CONCLUSIONS AND FUTURE WORK

We have summarized the genetics of lipid metabolism in humans. In addition to the major role that genetic factors play in modulating lipid levels, there is now a substantial body of evidence supporting the role of the gut microbiome in lipid metabolism. Lipid biomarkers, typically thought to be controlled by diet and genetics, are also influenced and altered by the gut microbiome. Unlike genetic factors, the gut microbiome is shaped by environmental factors, including diet, which can indirectly influence lipid metabolism.

We are unaware of studies examining the interactions between host genetic variants and gut microbiome composition in modifying circulating lipid levels in humans; clearly this will require future investigation and will surely add to the complexity of responses to dietary intervention. However, interactions between host genotype and lipid TG levels have been reported in animal studies (49). Since the microbiome varies considerably between individuals the responses to all drugs and diets will likely have variable metabolic consequences. From dietary intervention to lipid-modulating drugs, current medical options for treatment will change with the emergence of the gut microbiome as a potential biotherapeutic tool. Future work will focus on how to integrate an individual’s genetic make-up while manipulating the gut microbiome to improve lipid profile and reduce cardiovascular events. Understanding the effects of changes in microbial composition on metabolism and its interactions with a person’s genetic make-up and lifestyle will be challenging but could provide novel, cheap, and safe therapies in the field of precision medicine.

GRANTS

TwinsUK was funded by the Wellcome Trust; European Community’s Seventh Framework Programme (FP7/2007–2013). The study also receives support from the National Institute for Health Research (NIHR) Clinical Research Facility at Guy’s & St Thomas’ NHS Foundation Trust and NIHR Biomedical Research Centre based at Guy's and St Thomas' NHS Foundation Trust and King's College London. T. D. Spector is an NIHR senior Investigator. This work is also supported by Medical Research Council AimHY Grant MR/M016560/1, by the British Heart Foundation, and by the FP7 project HEALS (Health and Environment-wide Associations based on Large population Surveys) Project no. 603946 of the European Union’s Seventh Framework Programme.

DISCLOSURES

T. D. Spector is cofounder of MapMygut Ltd. A. M. Valdez is a paid consultant for Zoe Global Ltd. All other authors declare no competing financial interests.

AUTHOR CONTRIBUTIONS

M.M.H. prepared figures; M.M.H., A.M.V., and C.M. drafted manuscript; M.M.H., F.M.M.W., T.P., A.M.V., T.D.S., and C.M. edited and revised manuscript; M.M.H., F.M.M.W., T.P., A.M.V., T. D.S., and C.M. approved final version of manuscript; T. D.S. and C.M. conceived and designed research.