Fight against fibrosis in adipose tissue remodeling

Abstract

Adipose is a key tissue regulating energy homeostasis. In states of obesity, caloric intake exceeds energy expenditure, thereby accelerating lipid accumulation with ongoing extracellular matrix (ECM) remodeling. Excess deposition of lipids and expansion of adipocytes potentially decrease ECM flexibility with local hypoxia and inflammation. Hypoxia and chronic low-grade inflammation accelerate the development of adipose tissue fibrosis and related metabolic dysfunctions. Recent research investigated that some cytokines and proteins are functional in regulating energy homeostasis, meanwhile, are potential targets to fight against adipose tissue fibrosis and insulin resistance. In this review, we focused on the regulatory mechanisms and mediators in remodeling of adipose tissue fibrosis, along with their relevance to clinical manifestations.

INTRODUCTION

Obesity is both an epidemic and a chronic disease. It is closely related to metabolic disorders, such as type 2 diabetes mellitus (T2DM), cardiovascular disease, nonalcoholic fatty liver disease (NAFLD), and certain types of malignancies (1). Obesity is induced by imbalance between energy storage and energy expenditure. Energy stores in white adipose tissue (WAT) and releases in response to increased energy demands, such as during exercise or fasting. In contrast, brown/beige adipose tissue transfers energy into heat through thermogenesis (2). Both brown and beige adipocytes uniquely exhibit uncoupling protein-1 (UCP1), which is the functional marker for fuels oxidation and heat generation in mitochondria, and share similar morphological structures, i.e., multilocular droplets and abundant cristae-dense mitochondria. Both of β3-adrenoceptor activation and high-fat feeding are able to recruit bipotential adipocyte progenitors in WAT (3). Platelet-derived growth factor receptor alpha (PDGFRα⁺) progenitor cells are considered to be a contributor for both of beige adipocytes and fibrotic adipocytes (4).

To regulate energy homeostasis, adipose tissue adapts and remodels in response to metabolic and nutritional stimuli (5). In the remodeling process, however, pathological fibrosis limits plasticity and adipose tissue loses function. Adipose tissue fibrosis refers to excessive pathological accumulation of extracellular matrix (ECM) in adipose tissue, which involves a variety of biological processes, including adipose precursor cell proliferation, mature adipocyte hypertrophy, macrophage polarity, inflammatory immune cell infiltration, and energy metabolism (6). Adipose tissue fibrosis plays critical roles in controlling both energy homeostasis and glucose homeostasis. Collagens, matrix metalloproteinase (MMPs), lumican, and fibronectin are highly enriched ECM components in fibrotic adipose tissue (7, 8). In addition, the roles of cytokines or proteins in adipose tissue fibrosis and related diseases are summarized in this mini-review.

ENERGY HOMEOSTASIS AND ADIPOSE TISSUE REMODELING

Adipose tissue plays an important role in regulating energy homeostasis. Due to the plasticity of adipose tissue, there are morphological and functional changes based on mitochondrial biogenesis and degradation (9). When energy demand increases, such as in response to cold stimuli, beige adipocytes appear in the WAT, a process called “browning,” whereas beige adipocytes can be converted into white adipose cells, a process known as “whitening,” when energy demand is reduced in the absence of stimuli (10, 11). In addition to cold exposure, increased adrenergic signal transduction caused by adrenergic agonist treatment can also lead to adipose beiging. CD81 is a beige fat progenitor cell marker (12). CD81 loss is associated with diet-induced obesity, insulin resistance, and WAT inflammation. It also mediates energy balance through activation of inflammatory integrin-focal adhesion kinase (FAK) signaling, which provides novel insights in fibrotic adipose tissue remodeling (Fig. 1).

Figure 1.

Adipose tissue remodeling during obesity. With increased food intake, adipose tissue undergoes appropriate expansion through hypertrophy and hyperplasia. Adipose tissue remodeling impacts localized adipose tissue metabolism, which including adipogenesis, angiogenesis, insulin sensitivity, cytokine secretion profile, and in turn alters systemic glucose and lipid homeostasis. The activation and maintenance of beige adipocyte is a potential therapeutic strategy for combating adipose tissue fibrosis and insulin resistance.

Adipose tissue has evolved into a highly dynamic organ that can be remodeled at any time to meet the needs of a changing metabolic environment. Under normal conditions and in healthy states, adipose tissue exhibits a remarkable plasticity and can store and mobilize lipids to meet peripheral demand depending on the nutritional status of the body. This expansion is accomplished through increasing both of the size and the number of adipocytes (hypertrophy and hyperplasia) (13), reshaping the vascular system and ECM, and recruiting inflammatory cells (14). Recent research studies have discovered that several cytokines and proteins play regulatory roles during adipose tissue fibrosis (Table 1). The key metabolic effect includes glucose tolerance, insulin sensitivity, hepatic lipid accumulation, lipid metabolism, and energy expenditure.

Table 1.

Roles of cytokines and proteins in adipose tissue fibrosis

Cytokines/Proteins	Mechanism of Action	Key Metabolic Effect	Models Studied	Refs.
Col-VI	Suppression of adipose expansion	Lower energy expenditure, improves insulin sensitivity	Col-VI KO ob/ob miceHuman subjects	(7)
Lumican	Regulation of inflammation	Improves insulin sensitivity, glucose homeostasis	Lumican null mice (Lum-KO), human adipose tissue, Lum overexpression mice (Lum-OE)	(34)
DPT	Dermatopontin regulates ECM and inflammation genes	Increases DPT level in obesity and obesity-related T2D	Human subjects	(16)
HIF-1α	Interleukin-1 receptor-associated kinase M (Irak-M) dependent mechanism	Improves glucose tolerance	Phd2flox/flox/LysMcre miceIrak-M−/− mice	(17)
Cd248	Improves microvascular density	Improves insulin sensitivity, glucose tolerance	Cd248−/− mice, human adipose tissue	(18)
Mincle	Crosstalk between adipocytes and macrophages in CLS	Ameliorates glucose tolerance, hepatic lipid accumulation.	Mincle KO mice	(19)
Postn	Decreases macrophage infiltration	Ameliorates insulin resistance, ectopic lipid accumulation	Postn−/− mice	(20)
TCF21	Promotes interleukin 6 expression and ECM remodeling	siTcf21 reduces COL4 and COL1, but increases COL6 expression in vitro	TCF21 overexpress in mouse visceral WAT adipose stem cells	(21)
iNOS	Decreases leptin-induced tenascin C expression	Improves insulin sensitivity	iNOS knockout	(22)
GTF2IRD1	Through a PRDM16-EHMT1 complex	Improves glucose homeostasis	Fabp4-Prdm16 mice with UCP1−/− background, fabp4 promoter fat-selective Gtf2ird1 mice	(23)
DDR1	Activation of MRTF-A dependent pathway	Improves glucose tolerance and energy expenditure, alleviates WAT fibrosis	Ddr1−/− in a cardiometabolic disease mouse model	(24)
NAMPT	Suppression of adipose expansion	Suppression of appetite, improves glucose tolerance	Fat-specific Nampt knockout (FANKO) mice	(25)
MRTFA	Activation of ITGA5-MRTFA pathway	Improves glucose tolerance at the early stage of obesity Protects the mice from insulin resistance	MRTFA-knockout (KO) mice	(26)
Stk25	Regulates the activities of mitochondria	Improves lipid metabolism	Stk25 overexpression and knockout mice, HIB-1B cell line	(5)

Col-VI, type VI collagen; DPT, dermatopontin; HIF-1α, hypoxia-inducible factor-1α; iNOS, inducible nitric oxide synthase.

Progenitors of Adipose Fibrotic Cells

Adipose tissue fibrosis has also recently been shown to be regulated by specific adipocyte progenitor cell subsets, which are positive for platelet-derived growth factor receptor (PDGFR). PDGFRα⁺ progenitors are an important source of ECM production (4). More specifically, PDGFRα⁺ CD9^high progenitors were a major source of both fibrosis and proinflammatory factors, which participate in visceral WAT fibrosis and are associated with insulin resistance in T2DM. PDGFRα⁺ cells and Pdgfra gene are functionally different and required for WAT development and WAT homeostasis (27). Recent studies have obtained evidence that myocardin-related transcription factor A (MRTFA) is a critical inducer of progenitor fibrotic fate. MRTFA deficiency in mice shifts the fate of perivascular progenitors from Sca1⁻, Sma⁺, ITGA5⁺ and fibrogenic progenitor cells (FPCs) to adipocyte precursor cells, and protects against chronic obesity-induced fibrosis and the accompanying metabolic dysfunction, without changing energy expenditure (26). Notably, inflammation plays a key role in this process. In addition, as a surface marker of profibrotic perivascular progenitor, ITGA5⁺ cells expressing PDGFRα, CD9, and ITGA5 may function together to facilitate recruitment of FPCs from the vascular compartment. Inhibition of the ITGA5-MRTFA pathway may exert therapeutic effects on obesity-related diseases by inhibiting myofibroblast differentiation and promoting the synthesis of adipocyte precursors.

Mitochondrial Function

As the demand-driven terminus of oxidative metabolism, mitochondria play intricate and vital roles in maintaining energy balance. The importance of mitochondrial energetic output to adipose tissue fibrosis has been highlighted. Obesity-induced reactive oxygen species (ROS) in adipose tissue accelerated adipose inflammation and exacerbated fibrosis, resulting in restricted healthy adipose expansion with increased ectopic lipid accumulation (28). Serine/threonine protein kinase 25 (Stk25) exerts a protective effect against diet-induced excessive fat storage, inflammation, and fibrosis through regulating the activities of mitochondria in brown adipose tissues and WAT (5). Nicotinamide adenine dinucleotide (NAD⁺), as a coenzyme with universal functions, is involved in almost all processes of energy metabolism. Nielsen et al. (25) found that nicotinamide phosphoribosyltransferase (NAMPT) functions in regulating adipose fibrosis and the mitochondrial respiratory chain by mediating the biogenesis of NAD⁺ under high-fat diet-fed conditions. Importantly, transitioning from a high-fat diet back to normal chow largely reversed adipose fibrosis and dysfunction in fat-specific Nampt knockout mice, while the improved glucose tolerance persisted.

REGULATORS IN ADIPOSE FIBROSIS

ECM Components

Many studies have shown that fibrosis is related to ECM remodeling (8, 29). The ECM is composed of fibronectin, collagen, proteoglycan, and nonproteoglycan polysaccharides, which can regulate adipose tissue metabolism, immune response, and cellular behavior (6). These ECM components play an important role in maintaining adipose tissue homeostasis (30). Collagen, as the main ECM component, is closely related to adipose tissue fibrosis. In humans, type VI collagen (Col-VI) is elevated in subjects with obesity (31). In mice, Col-VI is the dominant form of collagen (endotrophin) specifically expressed in adipose tissue (1, 32). It is closely related to the systemic energy balance, and functions by interacting with other ECM proteins, such as lumican, fibronectin, and proteoglycans (7). The blockade of endotrophin of which with a neutralizing antibody ameliorated metabolic dysfunction and adverse effects (33). Thus, it is reasonable to speculate that blockade of adipose tissue fibrosis may improve obesity-related metabolic disorders.

Similarly, lumican binds to collagens and participates in ECM remodeling and glucose homeostasis (34). Dermatopontin is a noncollagenous protein with important biological functions in the ECM (16). Interestingly, matrix metalloproteinase 14 (MMP14), a key pericellular collagenase, is able to digest collagen VIα3 to produce endotrophin, which is a potent costimulator of fibrosis and inflammation (8, 33). MMP14 is likely to digest or modify the dense adipose tissue extracellular matrix. It shines new lights into strategies of adipose remodeling.

Hypoxia

In the event of excessive caloric intake, adipose tissue stores energy via enlargement of adipocytes and the recruitment of adipose precursor cells. However, rapid adipocytes expansion will lead to an insufficient local blood supply and thereby result in a locally hypoxic microenvironment in adipose tissue. Hypoxia increases the expression of hypoxia-inducible factor-1 α (HIF-1α) in adipose tissue, which in turn induces inflammatory cell infiltration, and causes adipose fibrosis (35). Overexpression of Hif1a in mouse macrophages is associated with impaired glucose metabolism, adipose tissue fibrosis, inflammation, and increased macrophage infiltration. HIF-1α in macrophages reportedly plays a novel role in obesity-related adipose tissue dysfunction through interleukin-1 receptor-associated kinase M (Irak-M) (17). Transmembrane glycoprotein CD248 mediates the transcriptional response to hypoxia and modulates the vascularity of WAT. Adipocyte-specific Cd248 knockout (KO) mice showed significant improvements in insulin sensitivity and glucose tolerance as well as attenuated hypoxia and fibrosis (18).

Inflammation

Local inflammation is another feature of adipose tissue fibrosis. Excessive fat accumulation promotes chronic low-grade inflammation and a hypoxic microenvironment in adipose tissue, which impairs the normal function of this tissue. Finally, there is an imbalance between the production and degradation of ECM, resulting in adipose fibrosis (29). An unique structure in obese adipose tissue is called crown-like structure (CLS), wherein macrophages are the main inflammatory cells that mediate obesity and adipose fibrosis (14, 19). Macrophage-inducible C-type lectin (MINCLE) is induced in macrophages by lipopolysaccharide through Toll-like receptor 4. Periostin (POSTN) is an ECM/matricellular protein secreted by macrophages. Postn-deficient mice are resistant to diet-induced obesity and insulin resistance, showing low levels of CLS formation and fibrosis in adipose tissue (20). Leptin is a hormone that signals the hypothalamus to regulate appetite and energy balance. Leptin-deficient ob/ob mice show increased macrophage infiltration and collagen deposition in adipose tissue (22). Transcriptional factor 21 (TCF21) is abundantly expressed in the proinflammatory environment in visceral fat depots. TCF21 contributes to ECM remodeling through matrix metalloproteinase-dependent collagen disposition by regulating IL-6 expression (21). In comprehensive experiments in humans and mice, innate lymphoid cells promoted adipose tissue fibrosis by increasing CD11c⁺ macrophages and activating the transforming growth factor-β (TGF-β1)/Smad 3 signaling pathway, and adipose innate lymphoid cells can also promote adipose tissue fiber generation by producing interferon-γ (30).

Transcriptional Factors, Cytokines, and Other Proteins

Biogenesis of brown and beige adipocyte is regulated by a number of transcriptional regulators (10). Transcription factor GTF2IRD1 inhibits fibrosis-related gene expression through the PRDM16-GTF2IRD1 complex (23). Growth hormone plays a key role in fat and glucose metabolism in a sex-independent manner (36). Vascular endothelial growth factor A (VEGFA) is functional in regulating blood vessel permeability and growth. Meanwhile, VEGFA and VEGFB play a balanced role in energy homeostasis in white adipose tissue (37). Interestingly, either inhibition or overexpression of VEGFA presents a phenotype of rapid beiging in the WAT, and glucose tolerance and insulin sensitivity are both improved in the two opposite genetic models (28, 38, 39). Overall, hypoxia and inflammation are both important balancing factors in regulating insulin sensitivity. VEGFA levels are closely related to adipose tissue hypoxia and inflammation. It is important to discover more detailed mechanism of VEGFA-regulated energy homeostasis and fibrosis in WAT.

In another metabolic model, discoid domain receptor 1 (DDR1), a collagen binding tyrosine kinase, was found to play an important role in energy consumption, adipose tissue fibrosis, and glucose homeostasis. Interestingly, by suppressing beige fat formation, DDR1 mediates tissue fibrosis through an MRTF-A dependent pathway (24). Specialized proresolving mediators (SPM) are functional in adipose tissue inflammation and adipose fibrosis. Strategies to increase local SPM production in WAT of subjects with obesity may be of great significance for the treatment of obesity-related diseases (40). Sterol-regulatory element-binding transcription factor (SREBP)-1a is also related to macrophage infiltration, tissue remodeling, and extensive fibrosis (41). ROS accelerated adipose inflammation and adipose tissue fibrosis. ROS inhibited de novo lipogenesis with the suppression of SREBF1 transcriptional activity via a reduction in KDM1A protein expression (28).

FIGHT AGAINST ADIPOSE FIBROSIS IN DISEASES

Type 2 Diabetes and Insulin Resistance

Type 2 diabetes mellitus (T2DM) is related to insulin resistance, along with inflammation and hypoxia in white adipose depots (41). Hyperglycemic glucose level leads to increased collagen VI and other ECM accumulation in WAT of patients with T2DM (29). Lack of collagen VI improved the pancreatic hyperplasia, glucose tolerance, and insulin sensitivity typical of ob/ob mice, which also displayed less tissue fibrosis and inflammation in WAT with higher rate of fatty acid consumption (7). It implied the evidence that there is a bridge connecting adipose fibrosis and insulin resistance. Inflammatory response is a key signal in adipose remodeling and insulin sensitivity. In RID α/β-transgenic mouse model, several inflammatory pathways were inhibited in adipose tissue, including TNFα, IL1β, and TLR4 signaling (13). The mice presented impaired adipose tissue function and promoted insulin resistance, with reduced angiogenesis in the adipose tissue. C-X-C motif chemokine ligand 14 (CXCL14) is a chemokine secreted from brown adipose tissue in response to thermogenic activation (38). Our recent findings revealed that serum CXCL14 levels showed simple correlation with obesity-related parameters in patients with T2DM (42). Transcription factor GTF2IRD1 represses adipose tissue fibrosis through PRDM16-EHMT1 complex and improves systemic glucose homeostasis in mice (23). Brown and beige adipose tissue activity may be functional on reducing of circulating glucose and lipids in patients with T2DM, even on remodeling of fibrotic adipose tissue.

Nonalcoholic Fatty Liver Disease

NAFLD is the most common chronic liver disease worldwide and is intimately linked with the metabolic disorder that is characterized by insulin resistance. Macrophages are the main mediators of obesity-related inflammation in the pathogenesis of adipose tissue fibrosis and NAFLD. In humans, a study found that markers of subcutaneous white adipose tissue (sWAT) inflammation are associated with liver fibrosis independent of obesity and sWAT fibrosis may contribute to diabetic risk through reduction of insulin secretion. Similarly, several proinflammatory changes in subcutaneous fat in obese patients with T2DM treated with insulin are associated with increased liver fat content, suggesting that inflammation and fibrosis in sWAT may be risk factors for NAFLD (43).

Other Adipose Tissue Fibrosis and Diseases

Fibrosis of perirenal adipose tissue (PRAT) is closely related to kidney disease. PRAT-related inflammation and ECM protein can be reduced by blocking plasminogen activator inhibitor-1, which provides evidence that renal injury can be attenuated in mice with obesity (44). In diabetic nephropathy mouse model, the number of inflammatory cells was significantly increased as PRAT fibrosis became more severe. With angiotensin 1–7 treatment, the inflammatory reaction and adipocyte size are reduced, with amelioration of PRAT fibrosis (45). Epicardial adipose tissue is located between the visceral pericardium and the heart. It is the site of local energy storage for the heart, producing many types of adipokines which are transported freely to the adjacent myocardium. Epicardial adipose tissue fibrosis was found to correlate positively with myocardial fibrosis, as well as with atrial fibrillation, which may involve cytotoxic CD8⁺ T lymphocyte-mediated inflammatory infiltration (46). Metformin can reduce susceptibility to atrial fibrillation, while adiponectin may play key roles in preventing metformin-dependent epicardial adipose tissue remodeling associated with atrial fibrillation (47). Metabolic activity in orbital fat is associated with inflammatory orbital diseases (48). In thyroid-associated ophthalmopathy, fibrosis was developed in retrobulbar adipose tissue (49). In patients with Graves’s ophthalmopathy, ECM and CD34 expression were increased in orbital adipose tissue, and sphingosine-1-phosphate mediates the process (50–52). Blocking of S1PR activity and inhibition of S1P synthesis might have therapeutic potential in the disease.

CONCLUSIONS AND FUTURE AVENUES OF RESEARCH

Adipose tissue fibrosis is caused by an energy homeostasis imbalance, which exacerbates obesity and related metabolic disorders. ECM remodeling is the key feature of adipose tissue expansion, which involves local hypoxia, inflammation as well as fibrosis. Ultimately, prevention of adipose tissue dysfunction is an effective strategy for improving systematic energy homeostasis and glucose homeostasis. To date, most of our knowledge has been obtained from animal research, and it remains to be seen whether the metabolic regulation in these models is the same as that in humans. At present, the main way that obesity is treated involves dietary interventions and weight loss surgery. There is as yet no effective drug treatment. Resolving adipose tissue fibrosis can start with ameliorating adipose tissue inflammation and maintaining adipose tissue plasticity. Current research focuses mainly on the effects of adipose tissue-related cytokines and secreted proteins on adipose tissue function, but the study of adipose tissue stem cells and myofibroblasts are gradually deepening. It is worth noting that adipose progenitor cells, which can differentiate into adipocytes or myofibroblasts with strong heterogeneity, have received attention. Studying these pathways is of great significance for understanding and formulating reasonable treatment strategies aimed at tackling obesity and related metabolic diseases.

GRANTS

This work was funded by the National Natural Science Foundation of China 32011540004, 31900823 (to X. Lu). We also acknowledge support from the Health Commission of Jilin Province 2019J069 and Jilin Talents Developing Fund 2019019 (to X. Lu). This work received funding from the Japan Society for the Promotion of Science JPJSBP 120207404 (to Y. Hasegawa).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

X.L. conceived and designed research; S.L., H.G., and X.L. drafted manuscript; H.G., Y.H., and X.L. edited and revised manuscript; Y.H. and X.L. approved final version of manuscript.