The kidney is characterized by a complex anatomy, with millions of nephrons as the functional unit to excrete nitrogen waste and secure fluid homeostasis. The kidney is composed of multiple specialized cell types ensuring vital homeostasis of acid-base and electrolyte balance, blood pressure regulation, nutrient reabsorption, and hormone secretion (Hoenig and Zeidel, 2014; Duann and Lin, 2017; Yu et al., 2019). Therefore, it is one of the most metabolically active organs other than heart and skeletal muscle, with proximal tubules presenting a very high density of mitochondria required for energy consumption (Meyer et al., 1997; Wang et al., 2000). For example, the human proximal convoluted tubules (S1 and S2 combined) contain abundant large mitochondria, which occupy about 16.3% cell volume (Møller and Skriver, 1985). Notably, the mature nephron comprises distinct segments, each utilizing metabolic pathways to varying degrees depending on the specific function (Cargill and Sims-Lucas, 2020).

Complete oxidation of fatty acids (FAs), which are high-energy substrates, to CO₂ and H₂O gives rise to roughly 9 Kcal/g fat, while only 4 Kcal/g are generated from carbohydrates or proteins. The heart possesses metabolic flexibility and powerful catabolic capacity to use various energy substrates, mainly FAs (40–60%) and glucose (20–40%), for ATP production (Karwi et al., 2018). Instead, most proximal tubule epithelial cells (PTEC) have low metabolic flexibility toward glycolysis and rely on FAs as energy source at baseline (Bonventre and Yang, 2011). This was shown by early in vivo studies measuring ATP synthesis by tracking isotope-labeled FAs with NMR in rat kidney, which indicated that FAs are a preferred fuel (Freeman et al., 1986). However, PTEC are able to shift to anaerobic glycolysis to produce ATP required for cellular regeneration after ischemic acute kidney injury (AKI; Lan et al., 2016). In mice, glomerular podocytes display much lower mitochondrial density than in PTECs and rely primarily on anaerobic glycolysis to maintain glomerular filtration barrier and are relatively insensitive to defect in mitochondrial biogenesis during ischemia damage (Brinkkoetter et al., 2019). Instead, as lipid accumulation is commonly observed in patients with chronic kidney disease, podocytes are rather sensitive to cellular cholesterol-mediated glomerular injury (Merscher et al., 2014).

Mitochondria are pivotal for maintaining the health and function of the metabolically active kidney by providing efficient energy support through the process of oxidative phosphorylation (OXPHOS) and aerobic glycolysis. Several factors such as mitochondria biogenesis, bioenergetics, dynamics, and autophagy regulate the mitochondrial physiology (Duann and Lin, 2017). In addition, mitochondria also contribute to production of reactive oxygen species (ROS) free radicals and transduction of metabolic and stress signals (Galvan et al., 2017; Flemming et al., 2018). Persistent mitochondrial damage is a major source of oxidants. Consequently, mitochondrial fitness translates into body’s general health. Mitochondrial dysfunction is involved in various kidney diseases, such as acute kidney injury (AKI), chronic kidney disease (CKD), diabetic nephropathy (DN), and glomerulonephritis (GN; Duann and Lin, 2017; Eirin et al., 2017; Galvan et al., 2017; Flemming et al., 2018).

Dysregulated lipid metabolism with defective cholesterol/free fatty acid (FFA) metabolism leading to dyslipidemia is common in patients of several kidney diseases, including acute kidney disease, CKD, diabetic kidney disease (DKD), nephrotic syndrome, and uremia (Agrawal et al., 2017; Hager et al., 2017; Gai et al., 2019; Nishi et al., 2019; Nishi and Nangaku, 2019; Jang et al., 2020b), and may contribute to end-stage kidney disease. In this review, we summarize the recent advances in understanding lipid metabolism in the function of kidney mitochondria and the molecular mechanisms related to dyslipidemia during kidney disease progression.

In biological systems, lipids include fats, sterols, phospholipids, and triacylglycerides (TAG). In the cell, lipids have numerous functions: they constitute the cell membrane as a protective barrier; form membranous compartments of intracellular organelles; provide energy source and storage; provide building blocks for hormones; and serve as secondary cellular messengers within body. FAs are carboxylic acids with a long aliphatic tail, which constitute building blocks for other lipids such as TAG and phospholipids.

Within the body, lipid metabolism comprises several inter-dependent pathways for the generation, storage, and transport of lipids, which involves plasma lipoprotein particles [chylomicrons, high density lipoproteins (HDL), low density lipoproteins (LDL), intermediate density lipoproteins (IDL), and very low density lipoproteins (VLDL)] in circulation. Dietary lipids, mainly (95%) TAG, some FFAs, and cholesterol, carried by chylomicrons into circulation, are degraded into FFAs and glycerol by lipoprotein lipase (LPL) activity on the capillaries. These FFAs are taken up by muscle, heart, and adipose and peripheral tissues like kidney; remnants of chylomicrons are subsequently cleared in the liver (Florens et al., 2016; Agrawal et al., 2017; Kronenberg, 2018). FFAs are transported by serum albumin to the liver and periphery and could be stored as TAG in kidney capillaries. Additionally, esterified cholesterol could be stored as a lipid droplet within the kidney.

FA β-oxidation (FAO) may occur in both mitochondria and peroxisomes. While mitochondria majorly oxidize LCFAs, and medium-chain and short-chain FAs (MCFAs and SCFAs, referring to FAs with less than 12 carbons chains), peroxisomes oxidize specific carboxylic acids such as very long-chain FAs (VLCFAs), branched-chain FAs, fatty dicarboxylic acids, and bile acid intermediates (in the liver; Cipolla and Lodhi, 2017). Interestingly, peroxisomal FAO provides alternative metabolism of LCFAs and MCFAs in case of mitochondrial long-chain FAO deficiencies (Violante et al., 2019). Mitochondrial FAO is thus the major pathway for the degradation of FAs to sustain cellular energy homeostasis (Houten et al., 2016). This process includes six tightly-regulated steps: (i) FA esterification to acyl-CoA; (ii) mitochondrial CPT shuttle (or the carnitine shuttle); (iii) the FAO pathway; (iv) the OXPHOS pathway; (v) allosteric control of FAO; and (vi) integrated nutrient metabolism in the kidney (Figure 1).

Several transcriptional, epigenetic, and post-translational regulators are involved in the crosstalk between peroxisomes, nucleus, and mitochondria to control the expression/functions of FAO enzymes, mitochondrial biogenesis, and energy reprograming in health and disease-stressed states (Stallons et al., 2013; Bhargava and Schnellmann, 2017; Figure 2). These key molecules include the nuclear hormone receptor, peroxisome proliferator-activated receptors (PPARs, PPARα, and PPARγ as examples; Wu et al., 2009; Corrales et al., 2018); PPARγ coactivator 1α (PGC-1α; Weinberg, 2011; Li and Susztak, 2018; Fontecha-Barriuso et al., 2020); the NAD⁺-dependent deacetylases sirtuins (SIRTs; Wakino et al., 2015; Hershberger et al., 2017; Morigi et al., 2018); AMP-activated protein kinase (AMPK); and nuclear respiratory factors 1 and 2 (NRF1 and NRF2; Akhtar and Siragy, 2019).

PGC-1α, the mitochondrial biogenesis master regulator, is predominantly expressed in proximal tubules and interacts directly with multiple transcription factors to integrate upstream signaling events with mitochondrial biogenesis and functional capacity. Downstream transcription factors control all aspects of mitochondrial function, including biogenesis, energy production, dynamics, and protein homeostasis. PGC-1α regulates the expression of NRF1 and NRF2 to increase NAD⁺ biosynthesis (Tran et al., 2016) and activates genes coding for the OXPHOS system (Zoja et al., 2014). In obesity-related nephropathy models, reduced NRF2 along with suppressed expression of the key FAO enzyme long-chain acyl-CoA synthetase-1 (ACSL1) are associated with elevated renal lipid deposition, further supporting the importance of mitochondria in lipid metabolism and energy homeostasis (Chen et al., 2019).

As the PGC-1α/PPARα axis governs transcriptional regulation of FAO, it was proposed as therapeutic target in AKI and CKD (Simon and Hertig, 2015; Stadler et al., 2015). Defects in the FAO pathway, such as reduced expression of CPT-1 (Kang et al., 2015), CrAT (Kruger et al., 2019), and PPARα (Chung et al., 2018) are associated with CKD and renal fibrosis. PPARα heterodimerizes with its obligate partner, the retinoid-X-receptor (RXR), to regulate FAO and energy metabolism. FAs are natural activators of PPARα, and one of PPARα target genes is CD36 (which increases FA uptake; Portilla, 2003). Moreover, impaired expression of PPARα and specific proteins in FAO pathway are associated with lipid accumulation and fibrosis in renal tubular epithelial cells in aging rats (Chung et al., 2018). Interestingly, Jang et al. demonstrated that proximal tubular mitochondrial interaction of PPARα with cyclophilin D (CypD), a component of the IMM structural protein complex mitochondrial permeability transition pore (mPTP), could repress nuclear PPARα activity and negatively modulate FAO in cisplatin-induced AKI (Jang et al., 2020a). Several PPARα agonists have been shown to enhance FAO activity in kidney (Konig et al., 2008; Lakhia et al., 2018).

PGC-1α is extensively regulated by post-translational modifications. AMPK and SIRT positively regulate PGC-1α through phosphorylation or deacetylation, respectively (Jager et al., 2007; Canto and Auwerx, 2009). Interestingly, liver histone demethylase JMJD3 was identified as a gene-specific transcriptional partner of SIRT1 that epigenetically activates mitochondrial β-oxidation during fasting (Seok et al., 2018). The counterpart of JMJD3 in kidney remains to be uncovered. General control of amino acid synthesis-5-like 1 (GCN5L1), a protein acetylase counteracting the function of SIRT3, was recently shown to negatively modulate hepatic FAO enzyme activities via acetylation (Thapa et al., 2018). Similarly, GCN5L1-mediated hyper-acetylation and impairment of FAO enzymes might be a key pathogenic event underlying lipid overload-induced kidney injury (Lv et al., 2019).

In summary, the integrated regulation of FA metabolism at the genetic, epigenetic, and protein level is tightly associated with mitochondrial homeostasis. Under injury or disease state, deficiency in PGC-1α and associated transcription factors leads to defective FAO, enlarged FA pool and TAG accumulation, massive ROS production, increased PPARα mitochondrial translocation inducing mPTP opening and loss of mitochondrial membrane potential, cytochrome C release, and mitochondrial damage (Figure 2).

Dyslipidemia is an abnormal amount of lipids (e.g., TAG, cholesterol or phospholipids) in the blood. Lipids in excess, which is delivered to organs beyond their energy demands, can be stored mainly as TAG in intracellular lipid droplets (LDs), an ubiquitous organelle that serves as energy stores, dynamic membrane synthesis, and as a hub for further metabolic regulation (Walther et al., 2017). Accumulation of such lipid intermediates or final products in non-adipose tissues, along with the subsequent multi-factorial disturbance of intracellular homeostasis, could result in lipotoxicity of target tissues. Lipotoxicity thus represents a pathologic phenomenon with hallmarks of aberrant lipid accumulation, causing metabolic, inflammatory, oxidative stress in intracellular organelles, and further triggering cell damages (Su et al., 2017; Opazo-Ríos et al., 2020).

Moorhead et al. first hypothesized “lipid nephrotoxicity” in 1982, proposing that dyslipidemia may contribute to the progression of renal dysfunction (Moorhead et al., 1982). This hypothesis had gained supportive evidence in several contexts. For examples, renal lipid accumulation has been shown with high clinical prevalence in patients with CKD, including the insulin resistant obese subjects with diabetic nephropathy (Herman-Edelstein et al., 2014; Escasany et al., 2019; Opazo-Ríos et al., 2020), in nephrotic syndrome (Vaziri, 2016; Agrawal et al., 2017), focal segmental glomerulosclerosis (FSGS; Sasaki et al., 2018), and also as a consequence of acute ischemic renal injury (Zager et al., 2011). Significant alterations in renal lipid metabolism are typified as high TAG, variation in the composition of apolipoproteins and lipids, the accumulation of atherogenic particles VLDL and IDL, and decreased HDL cholesterol (Vaziri, 2006; Stadler et al., 2015; Florens et al., 2016; Kronenberg, 2018; Du and Ruan, 2019; Gai et al., 2019; Thongnak et al., 2020; Jang et al., 2020b). As discussed earlier, systematic lipid metabolism involves multi-organ crosstalk, ultimately also affecting kidney function. Therefore, dyslipidemia and lipid nephrotoxicity could be not only a consequence but also a cause of kidney disease (Florens et al., 2016; Agrawal et al., 2017; Kronenberg, 2018; Czumaj et al., 2019; Nishi et al., 2019).

Excess fat could be derived from either dysfunctional capacity of adipose lipid storage, or from diet-induced hyperlipidemia (high plasma albumin-bound FFAs and cholesterol), or in the condition of renal dysfunction (as commonly exemplified by renal mass reduction in animal model) and defective insulin signaling. Excess kidney ectopic fat deposition and lipid overload in intracellular organelles could lead to ER stress (Zhao et al., 2008), mitochondria dysfunction (Vamecq et al., 2012; Szeto et al., 2016), and lysosomal stress (Yamamoto et al., 2017, 2020). These alterations could change cellular protective mechanisms such as autophagy, mitophagy, lipophagy and contribute to apoptosis and cell damage. These observations thus support the notion of dyslipidemia contributes to the progression of renal injury, and lipid-lowering therapies or shielding mitochondria could provide beneficial effects on lipotoxicity-mediated kidney injury (Izquierdo-Lahuerta et al., 2016; Su et al., 2017).

Dyslipidemia could appear in various forms with different causes and consequences. In lipid-mediated podocyte damage, FFAs and their metabolism affect function and survival of podocytes (Sieber and Jehle, 2014). Dyslipidemia is also a common feature, rather than a complication, of nephrotic syndrome. Excessive urinal protein loss results in hypoproteinemia, in turn leading to low serum oncotic pressure, and even edematous change in severe cases. To rescue the oncotic pressure deficit, the body initiates a reactive hepatic protein synthesis, including lipoproteins (Attman and Alaupovic, 1990; Merscher et al., 2014; Vaziri, 2016; Agrawal et al., 2017). Additionally, reduced plasma levels of lipoprotein lipase results in decreased lipid catabolism. Elevated serum levels of LDL and IDL are filtered through glomeruli and lead to lipiduria, which manifests with fatty casts containing oval fat bodies in the urine sediment (Cavanaugh and Perazella, 2019).

Mutations affecting cholesterol metabolism in the process of lipid trafficking, storage, influx, or efflux, could mediate glomerular injury (Merscher et al., 2014). For example, Tangier disease (OMIM #205400) or DKD caused by mutations in ATP-binding cassette A1 (ABCA1) gene result in reduced HDL in circulation, albuminuria, podocyte phenotype with esterified cholesterol accumulation and dysfunctional mitochondria due to cardiolipin hyperoxidation (Ducasa et al., 2019a,b). For the topics on glomerular diseases-related renal lipotoxicity and mitochondrial dysfunction, please refer to the comprehensive review in the same special issue (Ge et al., 2020).

Furthermore, excess of FFAs leads to TAG accumulation and renal tubular toxicity (Johnson et al., 2005; Scerbo et al., 2017). Increased LCFA-bound albumin induces altered redox balance, high tubular cell apoptosis, and kidney fibrosis (Ruggiero et al., 2014). As lipoprotein abnormalities also correlate with high risk of both cardiovascular and kidney diseases, these modified lipoproteins could be accounted as actual mediators of uremic toxicity (Florens et al., 2016). LDL and oxidized (ox)-LDL uptake by mesangial cells lead to cell proliferation and glomerular matrix expansion, while uptake by PTE results in tubulointerstitial lesions with remarks of heightened expression of extracellular matrix proteins (Nosadini and Tonolo, 2011). HDL is a key player in reverse cholesterol transport to shuttle cholesterol from peripheral cells, such as macrophages, to the liver, therefore relieving the cholesterol burden of these cells. HDL thus exerts its anti-oxidant function through preventing LDL oxidation by ROS and protecting against the adverse effects of ox-LDL on the endothelium. Reduced levels and dysfunction of HDL, which could be due to perturbed HDL proteome composition, are common in CKD patients (Vaziri, 2006; Yamamoto et al., 2012; Agrawal et al., 2017; Kronenberg, 2018; Rysz et al., 2020).

Deficiency of FA metabolism and lipid overload are the main drivers in the progression of both glomerular and tubular kidney diseases. Lipid accumulation, particularly in ischemic proximal tubules, may result in persistent energy depletion with FFA-induced mitochondrial dysfunction, which could play an important role in the AKI to CKD transition (Szeto, 2017). Conversely, mitochondrial protection prevents high-fat diet-induced glomerular and tubular lesions (Szeto et al., 2016).

The pathophysiological changes underlying hyperlipidemia may involve energy shortage from impaired mitochondrial biogenesis or ATP energetics, and systemic oxidative stress due to excessive ROS production accompanied by ER stress and influx of inflammatory cytokines. Without timely intervention, these changes could eventually lead to apoptosis and kidney fibrosis (Agrawal et al., 2017; Du and Ruan, 2019). Less is known about the molecular mechanism of some toxic lipid intermediates (“metabolic poison”) derived from deficiency or decreased expression of FAO-related enzymes in kidney disease development (Stadler et al., 2015; Su et al., 2017); and future research may elucidate this process.

Lipid-lowering therapies in kidney diseases have been studied for many years, although statins is still the first choice of conventional hypolipidemia strategies for its effect on HMGCoA inhibition to block cholesterol synthesis. Cumulative pharmacological efforts have advanced the field to develop classic and novel lipid modifying therapies in kidney diseases, as extensively reviewed recently (Ferro et al., 2018; Sudhakaran et al., 2018; Filippatos et al., 2019; Rosenson et al., 2019; Heine et al., 2020; Opazo-Ríos et al., 2020). These include effective and well-tolerated drugs targeting various lipid synthesis, uptake, trafficking and metabolism pathways. Recent years, compounds that specifically target mitochondria have emerged as promising therapeutic options for patients with renal disease. Here, we discuss molecules targeting mitochondrial lipid metabolism and mitochondrial dysfunction pathways, including pharmacological agents promoting mitochondrial FAO, mitochondrial biogenesis, and ATP synthesis, as well as mitochondrial antioxidants (regulating ROS metabolism) and cardiolipin stabilizers.

Mitochondria are the “powerhouse” of the high-energy demanding kidney cells. Crosstalk between mitochondria, nucleus, endoplasmic reticulum, and peroxisomes impacts numerous cellular functions. Mitochondrial bioenergetics, adaptation of energy metabolism, and mitochondrial biogenesis during physiology or stress conditions are tightly linked to body lipid homeostasis, as well as health and disease states of kidney. Dysfunctional mitochondria could lead to dyslipidemia, microvasculature damage, inflammation, kidney fibrosis, or even kidney failure. The evolving knowledge of the molecular mechanisms modulating mitochondrial energy homeostasis and lipid metabolism suggest that normalizing renal cell mitochondrial function and energy balance could be an important preventative strategy against dyslipidemia and could provide new drug targets in kidney diseases.

All authors contributed to the conception and drafting the work and critically reviewed and revised for the intellectual accuracy of the contents. All authors contributed to the article and approved the submitted version.