Impact of renal lipid droplet dynamics on tubular health

Abstract

Lipid droplets (LDs) are dynamic organelles that regulate lipid storage, metabolism, and organelle crosstalk. In renal tubular cells, which rely heavily on fatty acid oxidation for energy, LDs play a dual role. The transient accumulation of LDs during acute stress may protect against excess fatty acid accumulation and reduce lipotoxicity, but persistent LD overload contributes to mitochondrial dysfunction, oxidative stress, inflammation, and fibrotic remodeling. We compare adaptive versus maladaptive LD-mediated responses across acute kidney injury, chronic kidney disease, and diabetic nephropathy, integrating evidence from human biopsies, animal models, and in vitro studies. Preclinical studies indicate that restoring fatty acid oxidation and improving lipid turnover can reduce tubular injury and fibrosis. However, the effects of directly suppressing LD formation are context-dependent as transient neutral-lipid storage may protect cells from excess free fatty acids, whereas persistent LD accumulation can accompany lipotoxic injury. Clinically, PPAR agonists, SGLT2 inhibitors, and GLP-1 receptor agonists improve selected renal outcomes, but direct evidence that they reduce tubular LD burden in humans is lacking. However, kidney-specific biomarkers and drug delivery strategies remain underdeveloped. Key research gaps include a complete understanding of nephron segment–specific LD biology, LD heterogeneity, organelle crosstalk, and sex- and age-related regulation of LDs. Advances in patient-derived organoids, tubuloids, and kidney-on-a-chip models provide new opportunities to evaluate LD dynamics and identify targeted interventions. With these tools, the understanding of LD biology and pathophysiology will increase, and LDs may become a therapeutic target for protecting tubular health and slowing the progression of kidney disease. This review synthesizes recent advances in LD biology in the kidney, highlighting mechanisms of biogenesis, turnover via lipolysis and lipophagy, and interactions of LDs with mitochondria, the endoplasmic reticulum, and peroxisomes.

Lipid droplets (LDs) are dynamic organelles that are central to renal tubular metabolism. They consist of a neutral lipid core (triacylglycerols, cholesterol esters) surrounded by a phospholipid monolayer enriched with proteins. These proteins include perilipins, adipose triglyceride lipase (ATGL), and cell death–inducing DFFA-like effector (CIDE) family proteins that regulate biogenesis and turnover (1).

Renal proximal tubular cells rely on fatty acid oxidation (FAO) for ATP generation, rendering them vulnerable to lipid overload. LD formation can be protective in response to metabolic stress, resulting in sequestration of fatty acids and preventing their conversion to toxic intermediates, such as ceramides or diacylglycerols (2). LD biogenesis via diacylglycerol O-acyltransferase 1 (DGAT1) preserves mitochondrial function during nutrient stress by buffering free fatty acids (3).

Contacts between LDs and mitochondria contribute to the organization of cellular lipid metabolism. Perilipin-5 (PLIN5), an LD coat protein highly expressed in oxidative tissues, can promote the physical association of LDs with mitochondria. Experimental studies in hepatocytes and muscle cells have shown that PLIN5 enrichment at the LD surface increases LD–mitochondria proximity and supports mitochondrial respiratory capacity (4, 5). However, PLIN5-mediated coupling is not essential for fatty acid oxidation itself, and the evidence does not establish direct transfer of fatty acids from LDs into mitochondria. Although most of these studies were conducted outside the kidney, they provide a mechanistic basis for investigating whether similar spatial coupling contributes to lipid handling in renal tubular cells.

LDs are conceived and tethered to the endoplasmic reticulum (ER), and their functional association affects membrane lipid homeostasis and stress signaling. Proteins destined for LDs often partition from ER bilayers into LD monolayers via hairpin or monotopic topologies (6, 7). Pathological accumulation of LDs in tubular cells may exacerbate ER stress. In renal injury, impaired fatty acid metabolism and lipid accumulation accompany the progression toward chronic damage (8). Experimental evidence also indicates that ER-stress signaling can suppress PPARα-dependent fatty acid oxidation and promote tubular lipotoxicity and fibrosis (9), suggesting reciprocal interactions between altered lipid handling and ER homeostasis.

LD Biogenesis and Turnover in Renal Tubular Cells

LD biogenesis begins in the ER, where neutral lipids such as triacylglycerols (TAGs) accumulate between the two leaflets of the ER membrane. Diacylglycerol acyltransferases (DGAT1 and DGAT2) catalyze the final step in TAG synthesis, forming a neutral lipid lens that buds into the cytoplasm (1, 10). The protein seipin plays a structural role in stabilizing nascent LDs and preventing aberrant droplet clustering (11). Once formed, LDs are coated by structural proteins, notably perilipins (PLIN2 and PLIN5), which regulate the access of lipases to the lipid core (12).

LD turnover occurs through two complementary pathways: cytosolic lipolysis and autophagic degradation (lipophagy). In lipolysis, adipose triglyceride lipase (ATGL, encoded by PNPLA2) initiates the breakdown of TAG into diacylglycerol, followed by hormone-sensitive lipase (HSL) and monoglyceride lipase (MGL), releasing free fatty acids (13). Coactivators, such as comparative gene identification-58 (CGI-58), are essential for ATGL function, whereas PLIN proteins regulate enzyme accessibility.

Lipid clearance is mediated by a specific autophagic pathway known as lipophagy. During nutrient deprivation, LDs are sequestered into autophagosomes marked by LC3-II and degraded upon fusion with lysosomes, releasing fatty acids for oxidation (14). Transcription factor EB (TFEB), a master regulator of lysosomal biogenesis, promotes lipophagy and has been shown to protect against lipid-induced organelle stress (15).

Much of our mechanistic understanding of LD biology derives from hepatocytes, adipocytes, and muscle cells, where core proteins governing LD biogenesis and turnover, such as DGAT1, ATGL, and perilipins, have been extensively characterized. Nevertheless, renal proximal tubular epithelial cells employ comparable machinery when challenged by lipid overload. In experimental models of chronic kidney disease (CKD), the suppression of FAO is accompanied by the transcriptional induction of DGAT1 and PLIN2, promoting neutral lipid storage within LDs (16). Human kidney biopsies have demonstrated tubular lipid accumulation and altered expression of lipid-handling genes in diabetic kidney disease (17). Separately, fatty-acid exposure increased PLIN2 expression and LD accumulation in Caki-1 renal carcinoma spheroids, indicating that PLIN2 participates in lipid storage in this transformed experimental model (18). ATGL initiates triacylglycerol hydrolysis in lipid droplets. Its renal role has been examined using a renal tubule-specific Atgl deletion model, which demonstrated that tubular ATGL contributes to renal lipid metabolism and influences systemic glucagon-like peptide-1 and insulin secretion, although its deletion did not produce overt renal inflammation or lipotoxic injury (19). These observations indicate that proximal tubular cells adapt to lipid excess by mobilizing a conserved LD regulatory program. Thus, LD biology in the kidney is mechanistically aligned with, although less well studied than, that in hepatocytes and adipocytes; the principal mechanisms of LD biogenesis and turnover in renal tubular cells are summarized in Table 1.

Table 1.

Mechanisms of LD Biogenesis and Turnover in Renal Tubular Cells

Step	Key proteins and Mechanisms	Model	References
Biogenesis	-DGAT1/2 (ER diacylglycerol acyltransferases): catalyze TAG synthesis to form LD cores. mTORC1-regulated autophagy channels FAs to LDs via DGAT1. -Seipin: stabilizes nascent LDs. -PLIN2 (ADRP)/PLIN3 (TIP47): coat nascent LDs, stabilizing them. PLIN2/3 induction accompanies FA load; their knockdown in human PTCs increases FA-induced apoptosis, indicating a protective role of LD formation.	Human PTC lines (HK-2, RPTEC) and rodent models	(3, 4, 10,11,66)
Lipolysis	-ATGL (PNPLA2): principal TAG lipase. Renal tubule-specific ATGL KO (mouse) causes TAG/Lipid accumulation but no tubule inflammation or fibrosis, indicating LD buildup. -HSL (LIPE): compensates for ATGL loss by hydrolyzing DAG to MAG. -MAGL (MGLL): completes lipolysis to glycerol and FFA. PLIN2/3 regulate access of lipases to LDs; their degradation (via CMA) enables lipolysis (general mechanism).	Mouse (tubule ATGL-KO) and human cells; in vitro studies	(16, 19)
Lipophagy	-Macroautophagy (Beclin1, LC3): engulfs LDs. TGF-β1–induced autophagy in HK-2 cells paradoxically ↑LDs; Beclin1 knockdown reduced LD number and FA incorporation. -TFEB: master regulator of lysosomal/autophagy genes; promotes lipophagy. Proximal tubule–specific TFEB deficiency (aged mouse) ↓FAO/mitochondrial genes, and impairs clearance. -Chaperone-mediated autophagy (CMA): degrades LD-coat proteins (PLIN2/3) via Hsc70/LAMP2A, unmasking LD for lipolysis (general mechanism).	In vitro HK-2 cells, mouse models (fibrosis, aging)	(15, 67, 68)

Abbreviations used: ADRP, Adipose differentiation-related protein; ATGL, Adipose triglyceride lipase; Beclin1, Autophagy-related protein Beclin 1; CMA, Chaperone-mediated autophagy; DAG, Diacylglycerol; DGAT1, Diacylglycerol O-acyltransferase 1; DGAT2, Diacylglycerol O-acyltransferase 2; ER, Endoplasmic reticulum; FA, Fatty acid; FAO, Fatty acid oxidation; FFA, Free fatty acid; HK-2, Human kidney-2 (proximal tubular cell line); Hsc70, Heat shock cognate 70 kDa protein; HSL, Hormone-sensitive lipase; KO, Knockout; LAMP2A, Lysosome-associated membrane protein 2A; LC3, Microtubule-associated protein 1 light chain 3; LD, Lipid droplet; LIPE, Lipase E, hormone-sensitive lipase; MAG, Monoacylglycerol; MAGL, Monoacylglycerol lipase; MGLL, Monoglyceride lipase; mTORC1, Mechanistic target of rapamycin complex 1; PTC, Proximal tubular cell; PLIN2, Perilipin 2; PLIN3, Perilipin 3; PNPLA2, Patatin-like phospholipase domain-containing protein 2; RPTEC, Renal proximal tubule epithelial cells; TAG, Triacylglycerol; TFEB, Transcription factor EB; TIP47, Tail-interacting protein of 47 kDa; TGF-β1, Transforming growth factor beta 1.

LD Dysregulation in Renal Disease Contexts

In acute kidney injury (AKI), including ischemia–reperfusion injury (IRI) and cisplatin nephrotoxicity, LDs rapidly accumulate in proximal tubular cells. Primary lipidomics and histology show early increases in neutral lipid stores and specific lipid classes after renal IRI: targeted/untargeted MS has detected shifts in lipid species within hours of reperfusion, and classic biochemical studies have demonstrated a three-to fourfold increase in renal cortical cholesteryl esters shortly after in vivo ischemia, with little change in free cholesterol (validated in tubules and ATP-depleted HK-2 cells) (20, 21). In cisplatin models, the accumulation of LDs in the proximal tubules is accompanied by mitochondrial dysfunction and impaired FAO, as shown by Oil Red O/BODIPY staining of primary proximal tubular cells and mouse kidneys (22). An early LD burst may buffer surplus fatty acids and transiently support mitochondrial function, but persistent suppression of FAO and defects in autophagic clearance (lipophagy) favor LD retention and lipotoxic signaling; proximal tubule–specific pathways that curb lipid loading (e.g., FXR signaling) reduce LD build-up and mitigate AKI severity (23). Genetic and pharmacological data also implicate LD coat proteins: PLIN2 expression is induced in stressed proximal tubules, and its modulation alters tubular apoptosis and PPARα-linked responses during AKI (24). AKI features a time-dependent LD response that is initially adaptive but later maladaptive when FAO remains depressed, and lipophagy is inadequate. Analysis of public renal IRI transcriptomic datasets likewise documented ectopic LD accumulation in injured renal tubular epithelial cells, underscoring the consistency of this phenotype across AKI models (8).

In CKD, LD accumulation in renal tubules is a maladaptive hallmark of metabolic stress. In obesity, hypothermic kidney, and other renal injury models, BODIPY and neutral lipid staining in tubular sections revealed marked LD accumulation in cortical tubules, accompanied by upregulation of fatty acid transporters (e.g., FATP1) and suppression of oxidative gene programs (25). Transcriptomic and proteomic profiling of human and animal kidneys revealed suppression of mitochondrial β-oxidation and FAO pathways, and increased expression of enzymes involved in lipid uptake and storage (26). Genetic deficiency of PACS-2 in tubular cells exacerbates diabetic kidney injury by increasing the expression of SREBP1, SOAT1, and cholesterol esterification enzymes, further promoting tubular LD accumulation and interstitial fibrosis (27). In parallel, studies of chronic lipid stress in renal parenchyma have shown that the accumulation of ceramide and diacylglycerol species, which impair mitochondrial function, promote ROS production, and trigger profibrotic signaling cascades, thereby potentiating fibrogenesis in CKD. The capacity of altered renal-cell metabolism to drive LD accumulation is also illustrated in VHL-deficient renal carcinoma cell lines, in which constitutive hypoxia-inducible factor signaling is associated with increased neutral-lipid storage (Fig. 1). These findings provide mechanistic evidence from transformed renal cells and should not be interpreted as direct evidence of normal tubular physiology. Therefore, persistent LD overload is associated with tubular injury and fibrosis in CKD, and experimental manipulation of lipid uptake, oxidation, and storage pathways indicates that disturbed lipid handling can contribute causally to disease progression.

Fig. 1.

Metabolic regulation of lipid droplet accumulation in renal carcinoma cell lines. A: Representative Oil Red O staining of RCC4, RCC10, and 786-O renal carcinoma cell lines with or without VHL expression, showing increased intracellular lipid droplet accumulation in VHL-deficient cells. B: Quantification of extracted Oil Red O staining corresponding to panel (A), demonstrating significantly increased neutral lipid accumulation in VHL-deficient cells. C: Quantification of Oil Red O staining in RCC4 cells treated with linoleic and oleic acids under normoxic (N) or hypoxic (H) conditions, illustrating enhanced lipid droplet accumulation under hypoxia. Reprinted with permission from (69).

Kidney biopsies from patients with diabetic kidney disease (DKD) or obesity-related nephropathy show neutral-lipid and LD deposition, as documented by Oil Red O, BODIPY, and filipin staining alongside altered expression of lipid-handling genes in tubular compartments (e.g., increased lipogenesis programs and reduced FAO signatures) (17). Moreover, dysregulated lipid influx and synthesis converge on proximal tubules. CD36-dependent fatty acid uptake contributes to mitochondrial dysfunction and injury in diabetes, and genetic deletion of Cd36 ameliorates DKD in mice. Albumin-bound fatty acids can also be taken up through KIM-1, directly promoting tubular lipotoxicity (28, 29, 30). High glucose levels increase SREBP-1 activity in the tubular epithelium and in diabetic kidneys, thereby increasing de novo lipogenesis and LD formation, whereas defective FAO in tubules amplifies lipid retention and fibrosis (17, 31, 32). Therefore, metabolic nephropathies display a progressive LD phenotype: early LD accretion may buffer excess fatty acids, but persistent upregulation of proteins involved in lipid uptake/synthesis (CD36, KIM-1, SREBP-1) and concomitant FAO suppression drives sustained LD overload, oxidative stress, and fibrotic remodeling in the diabetic kidney (29, 30, 32). The principal pathophysiological consequences of sustained LD accumulation for tubular cell function are summarized in Table 2.

Table 2.

Pathophysiological Consequences of LD Accumulation on Tubular Cell Function

Consequence	Effect on renal Cells	Model	References
↓ Fatty acid oxidation	Excess LDs (e.g., from ATGL suppression or HILPDA induction) inhibit FAO genes and peroxisomal FA breakdown, causing an energy deficit. TFEB deficiency also suppresses FAO pathways.	Mouse models (HILPDA-overexpression, TFEB-KO); human aged kidneys	(16, 67, 70)
Mitochondrial dysfunction	Lipid overload promotes mitochondrial stress/ROS. In HILPDA-induced LD overload (mouse UUO/IRI), mitochondrial dysfunction and ROS increase. TFEB loss impairs mitophagy/clearance.	Mouse (UUO, IRI), human cells (HK-2)	(43, 67)
Inflammation	Lipid-laden PTCs secrete pro-inflammatory cytokines (e.g., TNFα, IL-6) in vitro. However, RT-ATGL KO mice (high LDs) showed no kidney inflammation or fibrosis, suggesting that LD alone may not be sufficient to trigger inflammation.	Mouse (RT-ATGL KO) and cell studies	(17, 19)
Fibrosis	Persistent LD accumulation is linked to interstitial fibrosis. HILPDA-driven LDs upregulate profibrotic markers (TGF-β1, α-SMA, collagen I) in tubules. In patients with CKD, high HILPDA/LPA levels correlate with fibrosis risk.	Mouse (UUO, IRI), human CKD biopsies	(38)

Abbreviations used: α-SMA, Alpha smooth muscle actin; ATGL, Adipose triglyceride lipase; CKD, Chronic kidney disease; FA, Fatty acid; FAO, Fatty acid oxidation; HILPDA, Hypoxia-inducible lipid droplet-associated protein; HK-2, Human kidney-2 (proximal tubular cell line); IL-6, Interleukin-6; IRI, Ischemia–reperfusion injury; KO, Knockout; LD, Lipid droplet; LPA, Lysophosphatidic acid; PTC, Proximal tubular cell; ROS, Reactive oxygen species; RT-ATGL, Renal tubule–specific adipose triglyceride lipase; TFEB, Transcription factor EB; TGF-β1, Transforming growth factor beta 1; TNFα, Tumor necrosis factor alpha; UUO, Unilateral ureteral obstruction.

Functional impact of LD accumulation on tubular health

The role of LDs in tubular health is not exclusively detrimental. Transient LD formation during acute metabolic or oxidative stress can buffer fatty acids and prevent immediate lipotoxicity, acting as a short-term protective mechanism. When LD accumulation becomes persistent, it shifts toward pathogenic consequences, fueling mitochondrial dysfunction, oxidative stress, inflammation, and fibrosis (33, 34).

Renal tubular cells rely heavily on mitochondrial FAO for energy production. When LDs accumulate in excess, FAO becomes impaired, ATP generation decreases, and the levels of reactive oxygen species (ROS) increase. In both experimental CKD models and human biopsies, FAO defects are closely linked to tubular lipid overload and progressive fibrosis (16, 33). Lipidomic analyses further revealed that toxic intermediates, such as ceramides and diacylglycerols, accumulate under these conditions, disrupting mitochondrial dynamics and exacerbating oxidative stress (9, 35).

In addition to their metabolic roles, LDs can serve as signaling platforms for inflammation. In immune cells, such as macrophages, LDs serve as scaffolds for the synthesis of eicosanoids and other lipid mediators that shape inflammatory responses (36). A similar coupling between LD accumulation and inflammatory signaling has been observed in the kidney (16, 17). In models of diabetic kidney disease and tubulointerstitial injury, proximal tubular LD overload coincides with NF-κB and NLRP3 inflammasome activation, leading to the secretion of proinflammatory cytokines and exacerbation of tissue damage. In renal tubular injury models, lipid overload and mitochondrial stress can coincide with activation of inflammatory pathways, including the NLRP3 inflammasome. However, whether LD formation itself directly activates NLRP3 in tubular epithelial cells remains unresolved. The suppression of FAO and the resulting LD accumulation in proximal tubules further augment NF-κB signaling, directly linking metabolic stress to proinflammatory transcriptional programs (34). These observations indicate that tubular lipid accumulation occurs alongside innate immune activation in kidney disease. Whether LDs directly function as inflammatory signaling platforms in renal epithelial cells requires further experimental investigation (16, 37).

Lipid-loaded tubular cells actively promote fibrotic remodeling in CKD. Excess lipid storage suppresses PPARα activity and FAO, thereby reprogramming proximal tubule metabolism toward profibrotic signaling. In cultured tubular epithelial cells and mouse models, activation of the unfolded protein response transcription factor ATF6α downregulates PPARα, resulting in LD accumulation, mitochondrial dysfunction, apoptosis, and upregulation of connective tissue growth factor (CTGF). Moreover, deletion of Atf6α preserves PPARα, limits LD development, and attenuates tubulointerstitial fibrosis after ischemic injury (9). Complementary in vivo studies have shown that the selective overexpression of PPARα in proximal tubules protects against fibrosis and inflammation in unilateral ureteral obstruction, confirming that the restoration of FAO prevents lipid-driven extracellular matrix deposition (38). Translational evidence supports these findings. In human diabetic nephropathy biopsies, the degree of tubular LD accumulation strongly correlates with the severity of tubulointerstitial fibrosis (17). Therefore, the suppression of PPARα and consequent LD accumulation directly contributes to profibrotic remodeling in the kidney.

Experimental models and tools for studying LD dynamics in renal tubular cells

Classical two-dimensional cultures of HK-2 cells and primary proximal tubular epithelial cells are widely used to investigate renal lipid accumulation, mitochondrial dysfunction, and lipotoxic injury. Exposure to fatty acids or lipid-rich albumin can induce LD formation, mitochondrial dysfunction, and apoptosis in these systems, allowing mechanistic analysis of pathways involved in tubular lipid handling (24, 29). More generally, studies in nonrenal cell systems have shown that sequestration of excess fatty acids into triacylglycerol can protect against free fatty acid-induced lipotoxicity (2). Exposure to palmitate or albumin-bound fatty acids induces LD accumulation, mitochondrial dysfunction, and apoptosis, mimicking proteinuric states (24, 37). Although these models enable the mechanistic manipulation of the DGAT, PLIN2, and autophagy pathways, they fail to recapitulate the complexity of the nephron. Renal carcinoma cell lines, including Caki-1, RCC4, RCC10, and 786-O cells, have also been used to investigate LD regulation and fatty acid metabolism. However, these transformed models exhibit cancer-associated alterations in VHL–HIF signaling and lipid metabolism and therefore provide mechanistic evidence rather than direct models of normal proximal tubular physiology.

Organoids from pluripotent and adult stem cells provide a more complex environment. Human pluripotent stem cell (hPSC)-derived kidney organoids contain proximal tubule-like structures capable of lipid uptake and metabolic analysis (39). In parallel, adult stem cell-derived tubuloids generated from urine or kidney tissue retain epithelial characteristics, nephron-segment markers, and selected donor- and disease-associated features, and can reproduce transport functions relevant to renal physiology (40). These models may provide a useful platform for future studies of patient-specific LD responses and metabolic interventions.

Kidney-on-a-chip platforms have been developed for different nephron compartments, including proximal tubular models that incorporate luminal flow and shear stress (41) and glomerular filtration-barrier models containing podocytes and vascular elements (42). Such platforms could enable the dynamic analysis of LD formation and clearance in response to pathophysiological stimuli or drug exposure, thereby bridging the in vitro and in vivo contexts.

In vivo rodent models, typically those with metabolic syndrome (db/db mice or high-fat diet mice) or experimental kidney injury (such as cisplatin or ischemia–reperfusion), exhibit LD accumulation in proximal tubules (16, 43). Genetic models targeting FAO (PPARα, AMPK) or LD proteins (PLIN2) reveal causal links between LD dysregulation, tubular injury, and fibrosis (24).

Imaging and quantification tools for neutral lipids in kidney (or renal) cells include live-cell stains such as BODIPY and Nile Red. Additionally, Oil Red O staining is widely used to detect LDs or neutral lipid accumulation in renal tissue (e.g., in models of diabetic nephropathy, obesity, or aging); Sudan Black staining is also used occasionally (17, 44, 45). Transmission electron microscopy (TEM) is the most direct method for visualizing LD–mitochondria contacts at the ultrastructural level. In the kidney, TEM has demonstrated that such contacts are altered under disease conditions (43, 46). TEM data revealed, for example, that LDs directly abut mitochondria in renal tubular cells during hypothermic stress, supporting the view that lipid–organelle crosstalk contributes to metabolic adaptation in the tubular epithelium (25). Additional ultrastructural studies of renal injury and glomerular disease models have also revealed the juxtaposition of LDs and mitochondria in kidney tissue (46, 47). Lipidomics by mass spectrometry can quantify triglyceride species and potentially lipotoxic intermediates. LC3-II and p62 can provide information about general autophagic activity or flux, but they are not specific markers of lipophagy. More direct assessment of lipophagy requires evidence that LDs or LD-associated proteins are delivered to autophagosomes or lysosomes, for example by colocalization analyses, autophagic-flux experiments, or measurements of lipid-specific turnover (14, 15).

Therapeutic modulation of LD pathways in renal disease

In renal carcinoma cell lines, restoration of CPT1A-dependent fatty acid oxidation reduced LD accumulation and increased mitochondrial respiration (Fig. 2). This experiment demonstrates the metabolic consequences of restoring fatty acid oxidation in transformed renal cells rather than in normal proximal tubular epithelium. Preclinical strategies include promoting FAO, inhibiting lipid synthesis and droplet formation, and stimulating lipophagy, as summarized in Table 3. Strategies aimed at LD pathways must distinguish between protective sequestration of excess fatty acids into neutral lipids and maladaptive persistence of LDs during chronic metabolic stress. Consequently, interventions targeting DGAT activity, PLIN proteins, or LD biogenesis may have opposite effects depending on disease stage, lipid composition, and cellular capacity for oxidation and clearance. Pharmacological activation of PPARα by fibrates (e.g., fenofibrate and WY-14643) restored FAO pathways in proximal tubules, thereby reducing ectopic LD accumulation and slowing the progression of fibrosis in CKD models. In experimental kidney disease, pharmacological modulation of lipid handling has been associated with reduced tubular LD accumulation and preservation of glomerular and tubular ultrastructure (Fig. 3). Defective tubular FAO has been identified as a central driver of CKD, and activation of PPARα has been shown to normalize FAO, reduce lipid droplet accumulation, and inhibit fibrosis in mouse kidneys (16). In an experimental model of diabetic kidney disease, fenofibrate reduced renal tubular cell apoptosis and increased expression of medium-chain acyl-CoA dehydrogenase, supporting an effect on tubular fatty acid oxidation pathways (48). Therefore, PPARα agonists are useful for metabolic therapies that directly target LD overload in tubules by restoring oxidative lipid flux.

Fig. 2.

Restoration of fatty acid oxidation reduces lipid droplet accumulation and improves mitochondrial respiration in renal cancer cell lines. A: representative Oil Red O staining of RCC4 and 786-O cells following adenoviral expression of CPT1A or GFP control, showing reduced intracellular lipid droplet accumulation upon CPT1A expression. B: quantification of Oil Red O staining corresponding to panel (A). C: oxygen consumption rate (OCR) in RCC4 cells following adenoviral CPT1A expression, demonstrating enhanced mitochondrial oxidative metabolism. Reprinted with permission from (69).

Table 3.

Preclinical Interventions Modulating Renal LD Dynamics

Compound/target	Mechanism	Model	Effects on the Kidneys and LDs	References
DGAT2 inhibitor	Inhibits DGAT2 (alternate TAG pathway)	Human kidney organoids (APOL1 risk variants)	↑ LDs (DGAT2i led to more LDs and APOL1 recruitment to LDs, reducing cytotoxicity)	(52)
5A peptide (ApoA-I mimetic)	CD36 antagonist (inhibits FA uptake via scavenger receptor)	Mouse CKD (five-sixths Nx + AngII and UUO)	↓ macrophage infiltration, ↓ interstitial fibrosis/inflammation	(71)
Fenofibrate (PPARα agonist)	Activates PPARα→ ↑ FAO (via AMPK/FOXA2/MCAD pathway)	Rodent diabetic nephropathy models	Reduces tubular lipid accumulation and apoptosis (improves FAO); shown to reduce microalbuminuria in T2D	(48, 49)

Note: Many targeted interventions (e.g., atglistatin for ATGL or TFEB agonists, such as trehalose) have not been directly tested in renal models; their potential renal LD effects remain to be established.

Abbreviations used: AMPK, AMP-activated protein kinase; AngII, Angiotensin II; ApoA-I, Apolipoprotein A-I; APOL1, Apolipoprotein L1; ATGL, Adipose triglyceride lipase; CD36, Cluster of differentiation 36 (scavenger receptor); CKD, Chronic kidney disease; DGAT1, Diacylglycerol O-acyltransferase 1; DGAT2, Diacylglycerol O-acyltransferase 2; FA, Fatty acid; FAO, Fatty acid oxidation; FOXA2, Forkhead box protein A2; iPSC, Induced pluripotent stem cell; LD, Lipid droplet; MCAD, Medium-chain acyl-CoA dehydrogenase; mito, Mitochondria; NPC1L1, Niemann-Pick C1-Like 1 protein; Nx, Nephrectomy; PPARα, Peroxisome proliferator-activated receptor alpha; TAG, Triacylglycerol; T2D, Type 2 diabetes; TFEB, Transcription factor EB; UUO, Unilateral ureteral obstruction.

Fig. 3.

Ultrastructural evidence that modulation of lipid handling preserves renal architecture. A: Transmission electron micrograph of glomerular capillaries from untreated hyperlipidemic–hyperglycemic mice showing glomerular basement membrane thickening, endothelial injury, and podocyte foot process effacement. B: Glomerulus from lipid-modulator–treated mice displaying preserved ultrastructure and reduced injury. C: Tubular epithelial cells from untreated mice exhibit numerous intracellular lipid droplets. D: Tubular epithelial cells from treated mice showed markedly reduced lipid droplet accumulation. E: Quantification of glomerular basement membrane thickness across experimental groups. Reprinted with permission from (72).

AMPK activation remodels LD handling in renal epithelial cells. In diabetic kidney disease (DKD), tubular lipid accumulation is associated with altered FAO and broader disturbances of cellular quality-control pathways (17). In a mouse model of diabetic kidney injury, pharmacological AMPK activation enhanced mitophagy and attenuated renal tubulointerstitial fibrosis (49). These findings support a role for AMPK-dependent mitochondrial quality control in limiting tubular injury, but they do not directly establish an AdipoR1-mediated effect on lipophagy or renal LD clearance. Genetic loss of renal AMPKα leads to LD accumulation in the proximal tubules, characterized by disordered lipid species, in vivo, underscoring the direct role of AMPK in LD control (43). LD visualization and quantification using in vivo/in vitro tubular models and BODIPY/Oil Red O staining have revealed that perturbing fatty acid uptake or mitochondrial stress modulates LD abundance (25, 46). In kidney epithelial systems, promoting safe lipid buffering via LD biogenesis can be cytoprotective under lipotoxic conditions (45). Metformin activates AMPK-dependent protective pathways in renal epithelial cells and experimental kidney disease. Preactivation of AMPK by metformin reduced the severity of acute renal epithelial injury, whereas AMPK-dependent phosphorylation of acetyl-CoA carboxylase contributed to its antifibrotic effects in chronic kidney injury models (50, 51). However, these studies did not directly establish that metformin promotes tubular lipophagy or mitophagy, restores fatty acid oxidation, or reduces renal LD burden. Its potential role as an LD-modulating intervention remains insufficiently tested.

DGAT enzymes regulate triacylglycerol synthesis and LD formation, but their therapeutic manipulation is context-dependent. In a renal APOL1 risk-variant model, DGAT2 inhibition increased LD formation and reduced cytotoxicity, indicating that limiting DGAT activity does not necessarily reduce LD burden and may instead alter how lipids are safely sequestered (52). Further studies are needed to determine whether DGAT1 or DGAT2 modulation is beneficial in other forms of tubular injury and chronic kidney disease. In experimental acute kidney injury, altered PLIN2 expression influenced tubular apoptosis and PPARα-associated lipid metabolism, indicating that PLIN2 contributes to the renal response to acute metabolic stress (24). Nonrenal studies indicate that autophagy and TFEB activation can enhance lipid droplet turnover (14, 15). Whether these approaches protect against renal tubular injury or diabetic kidney disease requires direct kidney-specific validation.

Clinical and translational approaches include the use of PPAR agonists (fibrates), SGLT2 inhibitors, GLP-1 receptor agonists, statins, and other agents summarized in Table 4. With respect to PPAR agonists, clinical trials using fenofibrate (FIELD and ACCORD) have demonstrated that it reduces albuminuria in patients with type 2 diabetes, consistent with preclinical evidence of FAO restoration. However, fenofibrate-associated increases in serum creatinine can complicate the interpretation of kidney function and require careful monitoring, particularly in patients with CKD (53, 54). Empagliflozin and dapagliflozin provide clinically established renoprotection in diabetic and non-diabetic CKD. Experimental studies suggest that metabolic reprogramming, including effects on fatty acid oxidation and lipid handling, may contribute to these benefits, but renal LD accumulation was not quantified in the major clinical trials (55, 56). GLP-1 receptor agonists such as liraglutide reduce tubular LD accumulation and increase the expression of mitochondrial markers in experimental obesity-associated kidney injury (57). In clinical studies, liraglutide reduced the incidence of a composite renal outcome, predominantly through a lower rate of new persistent macroalbuminuria, whereas the dedicated FLOW trial demonstrated that semaglutide reduced clinically important kidney outcomes in patients with type 2 diabetes and chronic kidney disease (58, 59). These trials did not assess renal LD burden. Statins are used in chronic kidney disease primarily for cardiovascular risk reduction. Available clinical studies, including (60); did not assess renal LD burden or establish statins as LD-targeted renal therapies.

Table 4.

Clinical Interventions With LD-Related Effects

Drug	Mechanism of Action	LD-related renal Effects	Evidence (Trials/studies)
SGLT2 inhibitors (e.g., empagliflozin)	Block SGLT2 in PCT→ glycosuria, altered renal lipid handling, and reduced lipid accumulation	Preclinical: Prevented renal TG accumulation and inflammation in db/db mice. Clinical: Slowed DKD progression and reduced albuminuria (EMPA-REG, DAPA-CKD); LD-specific effects were not measured.	EMPA-REG, CANVAS, DAPA-CKD; Wang et al. (2017) (70)
PPAR agonists (fenofibrate, pioglitazone)	↑PPARα (fenofibrate) or ↑PPARγ (TZDs) → ↑FAO/altered adipogenesis	Fenofibrate increased lipid metabolism in DN models; the FIELD trial showed slower eGFR decline. TZDs moderately alleviated proteinuria. No direct LD data.	FIELD/ACCORD
GLP-1R agonists (liraglutide)	↑GLP-1→ weight loss, improved lipid handling	Animal studies: reduced renal fat/lipotoxicity. RCTs (LEADER) showed ↓nephropathy events; LD effects were not studied.	LEADER, SUSTAIN-6
Mineralocorticoid RAs (finerenone)	MR blockade→ ↓inflammation/fibrosis	Significantly reduced CKD progression in T2D (FIDELIO): ↓ albuminuria & slowed eGFR loss; LD effects were not studied.	FIDELIO-DKD
Statins (HMG-CoA inhibitors)	↓ cholesterol synthesis	Reduce cardiovascular risk; modestly alleviated proteinuria in some analyses. Renal LD effects were not studied.	No clinical data on renal LD burden

Note: In many of the trials exploring these agents (e.g., ACEis/ARB, GLP-1RAs, MRA), renal benefit was demonstrated but LDs were not explicitly assessed. Note that the mentioned agents were not proven to reduce renal LD accumulation in humans, as the clinical trials did not include renal LD quantification and reported only other renal endpoints.

Abbreviations used: ACCORD, Action to Control Cardiovascular Risk in Diabetes trial; ACEi, Angiotensin-converting enzyme inhibitor; ARB, Angiotensin II receptor blocker; CANVAS, Canagliflozin Cardiovascular Assessment Study; CKD, Chronic kidney disease; DKD, Diabetic kidney disease; DN, Diabetic nephropathy; DAPA-CKD, Dapagliflozin and Prevention of Adverse Outcomes in Chronic Kidney Disease trial; eGFR, Estimated glomerular filtration rate; EMPA-REG, Empagliflozin Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients; FAO, Fatty acid oxidation; FIELD, Fenofibrate Intervention and Event Lowering in Diabetes trial; FIDELIO-DKD, Finerenone in Reducing Kidney Failure and Disease Progression in Diabetic Kidney Disease trial; GLP-1, Glucagon-like peptide-1; GLP-1R, Glucagon-like peptide-1 receptor; HMG-CoA, 3-Hydroxy-3-methylglutaryl coenzyme A; LD, Lipid droplet; LEADER, Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results trial; MRA, Mineralocorticoid receptor antagonist; MR, Mineralocorticoid receptor; PCT, Proximal convoluted tubule; PPAR, Peroxisome proliferator-activated receptor; PPARα, Peroxisome proliferator-activated receptor alpha; PPARγ, Peroxisome proliferator-activated receptor gamma; RA, Receptor antagonist; RCT, Randomized controlled trial; SGLT2, Sodium–glucose co-transporter 2; SUSTAIN-6, Semaglutide Unabated Sustainability in Treatment of Type 2 Diabetes trial; T2D, Type 2 diabetes; TG, Triglyceride; TZD, Thiazolidinedione.

Several translational challenges remain. A critical gap is the absence of biomarkers for renal LD burden. Unlike hepatic steatosis, tubular LD burden cannot currently be quantified using a validated noninvasive clinical method, although experimental imaging and metabolomic approaches are under investigation. Candidate noninvasive approaches include urinary metabolomic profiling and extracellular vesicle analyses, but no marker has yet been validated as a measure of tubular LD burden. Another limitation is the lack of kidney-targeted delivery systems; most FAO modulators act systemically. Patient-derived tubuloids and organ-on-a-chip models may support screening of renal-specific LD modulators in the future.

Research gaps and future directions

Despite significant progress, the understanding of renal LD biology remains incomplete. The principal research gaps and emerging experimental tools are summarized in Table 5 and discussed below.

• 1)Segment- and cell type–specific LD biology: Most studies focus on proximal tubules, but distal tubules, the thick ascending limb, and collecting duct cells also undergo metabolic stress. Single-cell transcriptomics and kidney tubuloid models provide segment-resolved systems for studying proximal and distal tubular populations (40, 61), but they have not yet established nephron-segment-specific LD dynamics.
• 2)LD composition and heterogeneity: LDs differ in size, lipid saturation, and protein composition. Lipidomic studies have demonstrated broad alterations of renal lipid composition in diabetic kidney disease, but the spatial distribution of LD-associated lipid species across the cortex, medulla, and individual nephron segments remains insufficiently characterized. Spatial mass-spectrometry imaging combined with segment-specific tissue or cell isolation could address this question. Whether these biochemical differences underlie variable susceptibility to lipotoxicity is unknown.
• 3)Organelle crosstalk: LD–mitochondria contacts regulate fatty acid channeling for oxidation, whereas LD–ER and LD–peroxisome interactions shape stress signaling. PLIN5 can promote LD–mitochondria coupling in oxidative tissues, although this interaction is not essential for FAO (5). Whether PLIN5 or other organelle-contact proteins regulate LD–mitochondria interactions in renal tubular cells remains unknown. Superresolution imaging and proximity labeling could help clarify these pathways.
• 4)Links to inflammation and fibrosis: It remains unclear whether LDs actively promote inflammation and fibrosis or act as epiphenomena. LD accumulation in tubular cells is correlated with TGF-β1 and CTGF expression in CKD (9, 16). Direct causal studies are needed.
• 5)Sex- and age-dependent regulation: Kidney disease progression differs between sexes and worsens with age. Evidence from adipose tissue, liver, and skeletal muscle indicates that sex hormones and aging can modify lipid utilization, mitochondrial capacity, and autophagic activity. These findings provide a rationale for investigating sex- and age-dependent regulation of renal LDs, but they do not establish comparable effects in renal tubular cells (62, 63).
• 6)Biomarker development: Unlike for hepatic steatosis, no noninvasive biomarkers exist for renal LD burden. Urinary metabolomic and extracellular-vesicle approaches may eventually provide indirect information on renal lipid handling, but their ability to quantify tubular LD burden has not been established. Validation against tissue-based measurements will be essential. Metabolomic profiling could further enable the stratification of patients by LD burden.
• 7)Kidney-targeted LD modulators: Systemic agents that influence fatty acid oxidation, including SGLT2 inhibitors, have demonstrated renal benefits in clinical studies, but their ability to reduce tubular LD burden in humans has not been established. Kidney-targeted approaches are therefore still needed to test whether direct modulation of LD handling provides additional benefit. The development of kidney-targeted small molecules, lipophagy inducers, or DGAT inhibitors is a critical translational goal. Patient-derived tubuloids offer a personalized platform for preclinical testing (52, 64, 65).

Table 5.

Research Gaps and Emerging Tools

Area/Tool	Description/Relevance	Notes
Segment-specific LD regulation	Nephron segments (e.g., PT vs. DT vs. glomerulus) may differ in lipid uptake/LD dynamics.	Requires segment-specific models.
LD proteomics	Comprehensive profiling of LD-associated proteins in renal cells (beyond PLIN2/3) is lacking. Proteomic (LC–MS/MS) mapping is needed to reveal novel regulators.
Organoids/tubuloids	Human iPSC-derived kidney organoids or primary tubule organoids can model LD biology. For example, APOL1 variant organoids (52) or Kir5.1−/− organoids (LD+ fibrosis) [Sendino Garví et al. 2024 (73)].	Organoid models enable genetic studies and drug testing.
Advanced LD imaging	Apply label-free imaging (coherent Raman scattering, CARS microscopy, MALDI-MS imaging) to visualize LDs in live tissue or biopsies.	These approaches enable spatially resolved LD/lipid analysis; coherent Raman methods can additionally support live-cell imaging.
Single-cell lipidomics	Emerging mass-spectrometry or fluorescent methods (e.g., BODIPY flow-cytometry, Raman single-cell) to quantify LDs in defined cell types.	Such techniques resolve the issue of heterogeneity in tissues with multiple cell types.
Organ-on-chip models	Kidney microphysiological systems (flow-based chips) to study LD dynamics under shear stress and metabolic gradients.	Recapitulates in vivo environment

Abbreviations used: APOL1, apolipoprotein L1; BODIPY, boron-dipyrromethene; CARS, coherent anti-Stokes Raman scattering; DT, distal tubule; iPSC, induced pluripotent stem cell; Kir5.1, inward-rectifying potassium channel subunit Kir5.1; LC-MS/MS, liquid chromatography–tandem mass spectrometry; LD, lipid droplet; MALDI-MS, matrix-assisted laser desorption/ionization mass spectrometry; PLIN, perilipin; PT, proximal tubule.

Conclusion

LDs in renal tubular cells are dynamic organelles that exert both protective and pathogenic effects. Under acute stress, LDs protect against fatty acid accumulation and reduce immediate lipotoxicity. Persistent LD accumulation, driven by impaired FAO, defective lipophagy, or increased lipid influx, contributes to mitochondrial dysfunction, oxidative stress, inflammation, and fibrotic remodeling (16, 33, 35).

Preclinical studies indicate that interventions targeting fatty acid oxidation and tubular lipid handling can attenuate renal injury and fibrosis in experimental models (16, 38, 48). Potential therapeutic approaches include restoring fatty acid oxidation, improving lipid mobilization and clearance, and reducing pathological lipid influx or synthesis. Direct inhibition of LD formation requires caution because LD biogenesis can be either protective or harmful depending on the metabolic context. Clinical trials of fibrates, SGLT2 inhibitors, and GLP-1 receptor agonists support the therapeutic relevance of renal metabolic pathways, but they do not yet establish tubular LD reduction as a mechanism of benefit in humans. Moreover, kidney-specific biomarkers and drug delivery strategies remain unknown (55, 56, 57).

Future progress requires addressing critical gaps, such as segment-specific LD biology, mechanisms of organelle crosstalk, sex- and age-related regulation, and the development of noninvasive biomarkers. Advances in patient-derived tubuloids, organoids, and kidney-on-a-chip systems provide unprecedented opportunities to study LD dynamics in physiologically relevant contexts (40).

Data and materials availability

All the data are available in the main text.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.