Competing Tissue-Specific Functions for the Tribbles-1 plasma lipid GWAS locus

Abstract

Purpose of review:

The pseudokinase Tribbles-1 (TRIB1) remains the focus of intense research since genome-wide association studies (GWAS) associated it with multiple cardiometabolic traits in humans, including plasma lipids and atherosclerosis. This review highlights recent advances in understanding the function of TRIB1 and what questions remain outstanding.

Recent findings:

Studies performed in a myeloid-specific Trib1 mouse model show that Trib1 contributes to foam cell formation, underscoring the importance of continued research into tissue-specific functions of TRIB1. Investigations of TRIB1 function in a 3D hepatic organoid model demonstrate that hepatic TRIB1 functions elucidated in mouse models are recapitulated in these organoid systems. Lastly, a recent study showed berberine, an existing lipid-lowering drug, to be acting via a Trib1-dependent mechanism, highlighting both a novel regulator of TRIB1 expression and the potential of studying TRIB1 through existing therapeutics.

Summary:

TRIB1 remains one of the more fascinating loci to arise from cardiometabolic GWAS, given the constellation of traits it associates with. As genetic studies continue to link TRIB1 to metabolic phenotypes, more functional research on tissue-specific TRIB1, regulation of TRIB1 and its function in current therapies, as well as reproduction of results from mice in human contexts are all necessary to increase our understanding of TRIB1 and its relevance.

Introduction

Cardiovascular disease is a leading cause of death in the developed world, largely due to due atherosclerotic coronary artery disease (CAD) (1). Despite the development of multiple treatments that reduce the risk of cardiac events, worldwide CAD risk remains high (2, 3). Given this high residual risk, there is a need for identifying novel regulators of CAD progression and its associated risk factors. Human genetic studies, and in particular human genome-wide association studies (GWAS), have proven very effective in identifying genomic loci harboring novel regulators of plasma lipid metabolism and CAD, having identified hundreds of loci that associate with plasma lipid traits (4–7). SNPs near the pseudokinase Tribbles-1 (TRIB1), located in the 8q24 GWAS locus, have repeatedly been associated with all four plasma lipid traits as well as CAD, non-alcoholic fatty liver disease (NAFLD), and adiponectin, clearly indicating metabolic relevance (7–11). Specifically, TRIB1 associates with the favorable lipid profile of reduced LDL-C, total cholesterol, and triglycerides, and increased HDL-C (7). These genetic associations have been replicated numerous times, while the list of associations at 8q24 locus keeps growing, including recent associations with ischemic stroke (12), coronary heart disease and serum lipids in Chinese Han patient populations (13), lipid traits in an Arab population (14) and associations between TRIB1 expression and the transcriptome of skeletal muscle and mitochondrial capacity after exercise (15). The immense amount of genetic data associating this locus with numerous cardiometabolic traits across multiple ethnicities has made functional studies of specific TRIB1 mechanisms in human disease of paramount importance.

Given the importance of the liver in lipid metabolism and metabolic health, initial functional studies seeking to understand the mechanisms through which TRIB1 may be modulating plasma lipids and atherosclerotic risk focused on hepatic Trib1. Indeed, hepatic gain and loss of function studies in mice revealed that Trib1 regulates de novo lipogenesis and VLDL production, the former mediated by TRIB1-mediated proteasomal degradation of the transcription factor C/EBPa (16, 17). While hepatic Trib1 clearly plays a role in regulating plasma lipids and atherosclerotic risk, the question of whether extra-hepatic TRIB1 contributes to the genetic associations observed in humans remains unclear. This is especially true given that results from GWAS only identify associations between traits and genes, not the tissue in which these associations are relevant (18). In this review, we will discuss recent work highlighting the importance of myeloid-specific TRIB1 in lipid metabolism and atherosclerosis, as well as ongoing work elucidating the functions of TRIB1 in human and mouse hepatocytes.

Myeloid Trib1 in Metabolism and Human Disease

Trib1 has long been recognized as an important regulator of myeloid cell differentiation and function, as studies nearly 15 years ago determined that TRIB1 and TRIB2 are drivers of acute myelogenous leukemia (19–21). What has been less clear is how this function relates to cardiometabolic traits. Previously, Trib1 was shown to regulate macrophage polarization which in turn can alter other metabolic phenotypes (21–23). This was first brought to attention by Satoh and colleagues, who showed that Trib1 was crucial for differentiation of tissue-resident M2-like macrophages (24). The investigators showed that splenic deficiency of Trib1 resulted in defective differentiation of F4/80⁺MR⁺ tissue-resident macrophages, also termed M2-like macrophages, mediated through the increased expression of C/ebpa, a transcription factor known to regulate myeloid cell differentiation (24). Interestingly, a chimeric mouse model which received a bone marrow transplant from Trib1 whole-body knockout (KO) mice exhibited reduced adipose tissue mass presumably due to increased lipolysis, which was rescued with supplementation of M2-like macrophages. Furthermore, these mice developed insulin resistance, hypertriglyceridemia and hypercholesterolemia when placed on a high fat diet. This was the first evidence suggesting that myeloid Trib1 contributes to the relationship between inflammation and metabolism, specifically through alter adipose tissue function (Figure 1) (24).

Figure 1: Myeloid Trib1 regulates plasma lipids and atherosclerosis through actions in both adipose tissue and the vessel wall.

Previous work has shown that loss of Trib1 in adipose-tissue macrophages causes loss of M2-like macrophages, increasing adipocyte lipolysis which in turn drives lipodystrophy, insulin resistance, and increased plasma cholesterol, all of which contributes to atherosclerosis. More recently, TRIB1 has been shown to promote the expression of macrophage OLR1, which in turns leads to increased oxLDL uptake, foam cell formation, and atherosclerosis in hyperlipidemic mice. These seminal findings suggest a large role for the myeloid niche in metabolic regulation by the pseudokinase TRIB1. This figure was created with BioRender.com.

Later work showed Trib1 to be controlling macrophage function and polarization via the JAK/STAT signaling pathway (25). Additionally, numerous recent publications have reported more and more relevance between TRIB1 and myeloid cells. It was revealed that two different miRNAs binding to TRIB1 in human primary macrophages altered the transcription and secretion of interleukin-8, again supporting a function of TRIB1 in macrophages as well as in overall inflammation (26). TRIB1 SNPs have also been found to have associations with iron recycling, which is relevant to macrophage function (27). In a mouse model of inflammatory bowel syndrome, knockdown of miR-98–5p was shown to alleviate symptoms through upregulating Trib1 expression, thereby promoting M2 macrophage polarization and decreasing inflammation (28). Similarly, Trib1 was found to contribute to self-repair from acute kidney injury in a mouse model by regulating renal macrophage polarization (29). Beyond macrophages, Trib1 is also a regulator of eosinophil lineage commitment. Mice lacking Trib1 in hematopoietic stem cells exhibit decreased eosinophil progenitor cells, due in part to the inability of Trib1 to restrain degrade C/EBPa and restrain the neutrophil gene expression program (30). The same group recently sowed that can regulate T-cell immunity through interactions with MALT1, a T-cell receptor signaling activator (31). Both eosinophils (32) and T-cells (33) have mechanistic links to atherosclerosis, and these findings could present novel avenues of research for understanding the associations between TRIB1 and CAD.

Myeloid Trib1 Promotes Foam Cell Formation and Atherosclerosis

Given the known role of Trib1 in regulating myeloid differentiation and macrophage polarization, and the importance of these cells in atheroprogression, the question of if and how macrophage Trib1 contributes to atherosclerosis has been of high importance. Recently, Johnston and colleagues attempted to answer this question through the generation of mice with macrophage-specific deletion and overexpression of Trib1, and subsequent crossing of these mice onto the atherogenic ApoE −/− background (34). Ultimately, they found that myeloid Trib1 promotes atheroma formation presumably through increased foam cell formation.

The authors established mouse models of both myeloid-specific Trib1 KO (mKO) and Trib1 overexpression (mTg) by crossing the LyzM-Cre transgenic strain to a mouse harboring a floxed allele of Trib1 (mKO), or to a mouse with Trib1 knocked into the Rosa26 locus preceded by a lox-stop-lox cassette (mTg). They then established chimeric ApoE −/− atherosclerotic mouse models via irradiation and bone marrow transplantation from donor Trib1 mKO or mTG mice, or Trib1-floxed wild-type (WT) control counterparts. These mice were subjected to prolonged western-type diet feeding and then the resulting atherosclerosis was quantified. Mice receiving myeloid Trib1 mKO BM cells showed reduced lesion burden in both whole aorta and aortic sinus as compared to WT, whereas Trib1 mTg chimeras had increased lesion area, demonstrating that myeloid Trib1 appears to promote atherosclerosis. Further characterization of the mice and lesions revealed that compared to WT controls, Trib1 mKO chimeras had smaller foam cells, less macrophage content by MAC-3 staining, and more plasma cholesterol, while the opposite was true for the mTg chimeras. Therefore, the authors concluded that it was an increase in macrophage lipid uptake and storage that drove foam cell expansion and atherogenesis in the mTg model. It is worth noting that these mice did not have the adiposity phenotypes observed in the chimeras Satoh and colleagues generated, as described above. This is an important discrepancy that, while potentially explained by differences in animal models, warrants further investigation, as this data would suggest loss of myeloid Trib1 is therapeutically beneficial, while the previous report would suggest the opposite (Figure 1).

Next, to clarify how myeloid Trib1 influences foam cell expansion, the authors investigated differences in gene expression between human monocytes and monocyte-derived macrophages with either high (TRIB1-high) or low (TRIB1-low) TRIB1 expression using microarray RNA data from the Cardiogenics Transcriptome study and applying DAVID Gene Ontology analysis to look for gene cluster enrichment (35). Ultimately, they found that the Oxidized Low Density Lipoprotein Receptor 1 (OLR1) was the most altered and upregulated scavenger receptor in TRIB1-high monocytes compared to TRIB1-low monocyte-derived macrophages. OLR1 is an oxLDL receptor, and can regulate foam cell formation in macrophages, a critical step in atherogenesis. This result was reproduced in their mTg mouse model, which exhibited increased Olr1 expression in bone marrow derived macrophages.

Johnston et al. followed up their observation of induction of Olr1 expression with a mechanistic validation of its contribution to foam cell expansion. Staining for Olr1 in aortic sinus lesions revealed increased Olr1 in macrophages and non-macrophage plaque surfaces in Trib1 mTg lesions compared to WT. Given the role of Olr1 in oxLDL uptake, the authors next incubated nonpolarized BMDMs from their mouse models with oxLDL and observed that Trib1 mTg BMDMs had a significant increase in intracellular total cholesterol and unesterified cholesterol as compared to their WT counterparts. Additionally, upon incubation with oxLDL, three times more Trib1 mTg BMDMs became foam cells than their WT counterparts, accompanied with a visible increase in neutral lipid accumulation. In parallel, these cells were confirmed to have no impairment in HDL-mediated cholesterol efflux, thus the authors concluded that the foam cell expansion induced by mTrib1 in early plaques is due to increased cholesterol and neutral lipid uptake and retention, rather than reduced efflux capacity.

Lastly, the authors did note the BMDMs from their Trib1 mTg mouse model had less expression of Scarb1 despite increased Olr1 expression, prompting further experiments to determine whether or not HDL uptake was part of the mTrib1 mechanism. This reciprocal relationship between Olr1 and Scarb1 transcripts was recapitulated in other models of polarized monocyte-derived macrophages, but not HDL-polarized MDMs, suggesting that myeloid Trib1 induces Olr1 expression and foam cell formation through a non-HDL uptake mediated mechanism, conserved across mouse and humans.

Overall, Johnston and colleagues showed that myeloid-specific Trib1 induces foam cell expansion in early atherosclerosis plaques through a mechanism involving Olr1, resulting in increased cholesterol and neutral lipid uptake that drives this expansion (Figure 1), and that this mechanism is likely conserved between mice and humans. Therefore, inhibiting myeloid TRIB1 could be used therapeutically to reduce coronary heart disease. This is the opposite directionality of hepatic Trib1, where it has previously been seen that loss of hepatic Trib1 drives an unfavorable plasma lipid profile, pointing to different directions of effect of TRIB1 on cardiometabolic health in different tissues (16). These are critical findings as disruption of TRIB1 function, a more traditional therapeutic approach, appears to be therapeutically beneficial in myeloid cells and not the liver. These observations also highlight the question of what the direction of effect for TRIB1 is in other metabolically relevant tissues such as adipose or skeletal muscle.

Hepatocyte 3D organoids recapitulate liver TRIB1 biology

While there is an appropriate growing focus on myeloid TRIB1 in human metabolic disease, hepatic TRIB1 remains an area of intense interest given the critical role of the liver in regulating cardiometabolic health as well as the ease of therapeutic targeting of the liver. Much of what we know about TRIB1’s functional relevance has been shown in mice, pointing to a specific need to investigate if Trib1 mechanisms reproduce in human cells or other relevant human models. Organoids have captured much attention in recent times as a new type of in vitro model that better recapitulate organ development and physiology while also being fairly easy to maintain and manipulate. They are often used for their realistic micro-anatomy, with many proponents emphasizing that they better replicate the functionality of actual organs compared to monolayer cultures (36).

Recently, Abbey and colleagues established and characterized a human hepatocyte-like 3D organoid system which they subsequently used to test if hepatic Trib1 biology previously observed in mice is reproducible in human liver organoids. The group began by establishing and characterizing the scalable system, which was derived from hepatic human induced pluripotent stem cells (hiPSCs) that were exposed to laminin-enriched, nonengineered matrices and growth factors to induce differentiation. Using immunostaining and qPCR, the authors showed that their organoids robustly and specifically expressed early and mature hepatocyte markers at higher levels than the 2D HLCs throughout differentiation. ELISAs and uptake assays showed the organoids to be secreting canonical liver proteins such as albumin, as well as being capable of LDL and CDFDA uptake. Altogether, the authors concluded that their 3D hiPSC-derived organoids were closer to human livers than 2D HLCs.

Next, the authors sought to use their 3D hiPSC-derived hepatic organoid model to recapitulate previously shown Trib1 KO biology. In mice, liver-specific KO of Trib1 results in increased plasma lipids and increased C/EBPa protein, but not increased Cebpa mRNA (17). Using CRISPR/Cas9, the authors knocked out TRIB1 in their hiPSCs by inserting an early stop codon in the first exon of TRIB1, and then growing them into 3D organoids or 2D HLCs. Compared to the TRIB1 KO organoids, the TRIB1 KO 2D HLCs had lower hepatic marker gene expression, while the 2D and 3D WT controls had comparable marker expression, leading the authors to conclude that the TRIB1 KO 2D HLCs have defective hepatic differentiation. They also found that the TRIB1 KO 2D HLCs had decreased C/EBPa protein and mRNA, while the TRIB1 KO organoids reproduced the increased C/EBPa protein. Notably, they were able to rescue the defective differentiation of the TRIB1 KO 2D HLCs by rescuing CEBPA expression via an inducible lentivirus system, indicating a role for the TRIB1/C/EBPa axis in hepatic differentiation. In regards to the lipid phenotype, qPCR showed that the TRIB1 KO 3D organoids had increased expression of lipid synthesis genes compared to their WT counterparts, whereas the TRIB1 KO 2D HLCs had decreased expression of these genes compared to WT 2D HLCs, consistent with C/EBPa driving lipogenic gene expression. Additionally, the authors showed that TRIB1 KO 3D organoids had increased triglyceride mass compared to the WT model, but the TRIB1 KO 2D organoids had decreased TG mass. Taken together, the authors proposed that their 3D hiPSC-derived hepatic organoid model recapitulates TRIB1 KO biology (37).

This work by Abbey et al. is an example of the crucial step of replicating phenotypes from mouse models in a human model, showing the continued functional relevance of TRIB1 in human health beyond mouse biology. This work also raises multiple questions such as is TRIB1 required for proper hepatic development? Current mouse models of hepatic Trib1 deficiency inactivate the gene after hepatic development, while whole-body KO of Trib1 is perinatal lethal (22, 23). Additionally, the results in 2D HLCs, with TRIB1 KO resulting in decreased C/EBPa protein, are in the opposite expected direction given our understanding of the relationship between TRIB1 and C/EBPa. These studies underscore the need for more mechanistic studies of TRIB1 function in human hepatocytes, while also replicating some previous findings in the 3D model. With the scalability of the organoid model and its potentially intact functional biology, this could prove to be an excellent model for the mechanistic interrogation of GWAS loci to prove human relevance.

Berberine decreases plasma triglycerides and upregulates hepatic TRIB1

Most TRIB1 work has focused on either SNP associations or its function in the context of genetic manipulation, but does it have any involvement with existing treatments for metabolic phenotypes? Recently, Singh and Liu reported on the consequences of berberine, a naturally occuring lipid-lowering drug that reduces LDL cholesterol, total cholesterol, and triglycerides in hyperlipidemic patients and mice, likely by upregulating the LDL receptor through enhancing LDLR stability and downregulating PCSK9 transcription (38–43). Considering that Trib1 overexpression in mice lowers plasma lipids, the group looked to see if there was perhaps an overlap in how these phenotypes were occurring.

First, the authors began by feeding their mice a high fat, high cholesterol diet (HFHCD) along with either berberine (BBR) or vehicle for 2 weeks. They saw a decrease in serum LDL, TC and TG, as well as hepatic index and hepatic TC and TG. Interestingly, they also saw a significant increase in Trib1 mRNA in their hepatic samples. To determine if this was due to the diet, they performed the same experiment in normocaloric diet (NCD) fed mice. They again saw a decrease in serum TG in the BBR mice, accompanied with an increase in Trib1 mRNA. To confirm if the induction of hepatic Trib1 expression by BBR was independent of Ldlr status, they fed Ldlr-deficient mice either BBR or vehicle for 1 week. In this experiment, they saw an increase in Trib1 mRNA, no change in LDL-C, and a decrease in serum TG. This led them to conclude that berberine’s lipid-lowering effects are mediated through both Ldlr-dependent and independent mechanisms, and that induction of hepatic Trib1 is associated with plasma TG reduction from BBR.

Next, Singh and Liu checked for human relevance by treating a human liver cell line (HepG2) as well as human primary hepatocytes with berberine. Over the course of a 24-hour treatment, they saw the induction of hepatic TRIB1 expression in their BBR-treated cells, confirming that this effect was conserved between their mouse and human cell models. As Trib1 has previously been shown to promote the phosphorylation of ERK1/2 (44, 45), they investigated ERK phosphorylation status and found that BBR treatment of HepG2s increased phospho-ERK over time, suggesting that BBR induction of TRIB1 expression is regulated by the ERK pathway.

To determine the importance of the ERK pathway as well as the mechanism by which BBR-mediated TRIB1 induction was occurring, the authors repeated the BBR treatment of HepG2s with Act-D, a transcription inhibitor. This did not result in increased TRIB1 expression, and subsequent cloning of the TRIB1 promoter into a luciferase reporter construct revealed increased promoter activity with BBR treatment, supporting a proposed BBR-mediated mechanism of promoter activation of TRIB1 transcription. Lastly, BBR treatment in the presence of a MEK1 inhibitor stopped induction of TRIB1 and the increase in TRIB1 promoter activity. Taken together, all of these experiments led the authors to conclude that berberine lowers lipids in both LDLR-dependent and independent cellular mechanisms, one of which is through inducing TRIB1 by activating transcription and promotion of TRIB1 in an ERK-dependent manner (46).

This report from Singh and Liu from 2019 highlights the potential mechanistic and functional understandings that can be taken from investigating the effect of current therapies on TRIB1. Fairly little is known about the regulation of TRIB1 transcription, a critical gap in the field especially given the lack of eQTLs between GWAS SNPs in the 8q24 lcous and TRIB1 expression in all tissues examined (18). One possible explanation for the lack of eQTLs is that a certain stimulus is required to see the effect of the GWAS SNPs, and this paper presents one of the few stimuli known to upregulate hepatic TRIB1 expression. Exploring how TRIB1 is behaving under current therapies for metabolic disorders could be a possible avenue to reveal more of its own regulation, as well as how it may be acting on other targets to result in metabolic phenotypes.

Conclusion

TRIB1, initially thrust to the forefront of atherosclerosis research through its discovery in GWAS, continues to be a relevant gene of interest in the field of cardiometabolic disease. Tissue-specific functions of TRIB1 are a point of interest, as most early work was largely performed in the liver. However, a growing body of literature is demonstrating the importance of myeloid TRIB1 to metabolic health. As shown by Johnston et al., myeloid Trib1 is distinct from hepatic Trib1 in its role in atherosclerosis progression. A new 3D human hepatic organoid model has recapitulated some of mouse Trib1 biology, bringing more relevance to previous findings, and demonstrates a potential for use in tissue-specific functional work as well as further GWAS validations. Additionally, studying how TRIB1 interacts with existing therapies like berberine may help to unravel its function within different diseases. Clearly, TRIB1 continues to be a metabolically relevant point of research, with numerous recent publications tying it directly to lipid metabolism and atherosclerosis, particularly in myeloid tissue. With a robust continued interest in the mechanisms behind the metabolic relevance of TRIB1, work focused on the regulation and function of TRIB1 remains important in lipidology and CAD pathogenesis.

Key points.

• The TRIB1 8q24 locus remains relevant to metabolic research years after its initial discovery, yet very little is known about the relevant mechanisms and tissues governing these genetic associations.

• Recent research on myeloid-specific Tribbles-1 demonstrates that it drives foam cell formation, a key part of atherogenesis, through increased expression of the Oxidized Low Density Lipoprotein Receptor 1 (OLR1)

• Myeloid Trib1 appears to be proatherogenic, the opposite direction of effect that hepatic Trib1 has in mouse models of gain and loss of function, suggesting that targeted disruption of Trib1 in vessel wall macrophages could be therapeutically beneficial.

• Ongoing work in hepatocytes has recapitulated liver-specific Trib1 findings in a human liver organoid model, while Berberine, a naturally occurring lipid lowering agent, upregulates the expression of Trib1 in hepatocytes