Impairment of PPARα and the Fatty Acid Oxidation Pathway Aggravates Renal Fibrosis during Aging

Abstract

Defects in the renal fatty acid oxidation (FAO) pathway have been implicated in the development of renal fibrosis. Although, compared with young kidneys, aged kidneys show significantly increased fibrosis with impaired kidney function, the mechanisms underlying the effects of aging on renal fibrosis have not been investigated. In this study, we investigated peroxisome proliferator–activated receptor α (PPARα) and the FAO pathway as regulators of age-associated renal fibrosis. The expression of PPARα and the FAO pathway–associated proteins significantly decreased with the accumulation of lipids in the renal tubular epithelial region during aging in rats. In particular, decreased PPARα protein expression associated with increased expression of PPARα-targeting microRNAs. Among the microRNAs with increased expression during aging, miR-21 efficiently decreased PPARα expression and impaired FAO when ectopically expressed in renal epithelial cells. In cells pretreated with oleic acid to induce lipid stress, miR-21 treatment further enhanced lipid accumulation. Furthermore, treatment with miR-21 significantly exacerbated the TGF-β–induced fibroblast phenotype of epithelial cells. We verified the physiologic importance of our findings in a calorie restriction model. Calorie restriction rescued the impaired FAO pathway during aging and slowed fibrosis development. Finally, compared with kidneys of aged littermate controls, kidneys of aged PPARα^−/− mice showed exaggerated lipid accumulation, with decreased activity of the FAO pathway and a severe fibrosis phenotype. Our results suggest that impaired renal PPARα signaling during aging aggravates renal fibrosis development, and targeting PPARα is useful for preventing age-associated CKD.

Age-related changes in kidney function, structure, and their relationships have been extensively recognized recently.¹ The effects of aging on the kidneys are the most dramatic among any other organs because of diverse factors that can accelerate the changes.¹ Epidemiologic studies have suggested that age-related renal changes may be associated mainly with systemic hypertension, diabetes, dyslipidemia, and other environmental causes such as smoking.^2,3 With aging, most subjects show a progressive functional decline that includes decreases in GFR and renal blood flow as well as increases in urinary excretion of proteins such as albumin.^4,5 The prevalence of CKD is significantly higher in older people, with more than one-third of those aged 70 or older exhibiting moderate or severe CKD.⁶ Interestingly, these functional changes occur in concert with structural changes, which gives more accurate information on the status of renal disease.⁷ Morphologic changes of aging kidneys include loss of renal mass (primarily in the cortical region), development of glomerular sclerosis, tubular atrophy, and tubulointerstitial fibrosis.⁷

Fibrosis is characterized by a loss of parenchymal tissue with the accumulation of fibrillary collagens produced by activated myofibroblasts. Although glomerular lesions are significantly associated with renal functional decline and particular renal diseases, renal fibrosis is always accompanied in all forms of CKD.⁸ In renal tubulointerstitial fibrosis, tubular epithelial cells play an important role.⁹ Loss of epithelial integrity causes cell cycle arrest and dedifferentiation, leading to increased expression of mesenchymal markers.¹⁰ During renal aging, it has been well established that structural changes occur with glomerulosclerosis, interstitial fibrosis, and tubular atrophy.¹¹ Interestingly, a recent study has shown that age-related changes in the renal structure occur earlier than functional changes.¹² The authors found a strong relationship between age and nephrosclerosis in healthy adults without functional changes in the kidney, thus implicating the importance of structural changes.

Alterations in lipid metabolism have been implicated in various metabolic diseases.¹³ Although the kidney is not classified as a metabolic organ, renal tubular cells have high basal levels of baseline energy consumption and they prefer fatty acid oxidation (FAO) as a major energy source.^14,15 The high energy yield of FAO (106 ATP equivalents per fatty acid [FA]), which is produced in the mitochondria and peroxisomal compartments, makes it possible to support the high energy consumption of cells. Recently, defects in the FAO pathway have received substantial attention in the context of acute and chronic kidney diseases.^15,16 It has been suggested that a defective FAO pathway induces abnormal lipid accumulation of triglycerides (TG), resulting in lipotoxicity that contributes to the development of kidney diseases.

Peroxisome proliferator–activated receptor α (PPARα) is the key transcriptional factor that regulates intracellular lipids through direct transcriptional control of genes involved in peroxisomal and mitochondrial FAO pathways, FA uptakes, and TG catabolism.^17,18 Accumulating evidence supports a link between PPARα and metabolic diseases including diabetes, obesity, dyslipidemia, and fatty liver.¹⁸ Because the kidney is a metabolically active tissue that uses FAs as a major energy source, the role of PPARα in the development of renal diseases has been extensively investigated recently.^15,19–21 It has been suggested that defects in PPARα and the FAO pathway play important roles during renal interstitial fibrosis development. Defective FAO in tubule epithelial cells has been associated with ATP depletion, cell death, dedifferentiation, and lipid accumulation, which are the phenotypes resembling fibrosis.¹⁵ Furthermore, others have suggested that miR-21-mediated downregulation of PPARα contributes to fibrogenesis and epithelial injury in a mouse fibrosis model.²²

In this study, we hypothesized as to whether age-associated renal fibrosis is associated with defective FAO in the renal tubule epithelium. We found that levels of PPARα and FAO-associated enzymes were reduced in aged rat kidneys with significantly increased lipid accumulation and interstitial fibrosis. Reduced PPARα was associated with increases of miRNAs that regulate PPARα translation. Further animal studies using antiaging calorie restriction (CR) and aged PPARα^−/− mouse models demonstrated the role of PPARα in the regulation of age-associated renal fibrosis.

Results

Aging Increases Renal Fibrosis, Functional Changes, and Epithelial Damage

First, we verified renal fibrosis, functional changes, and epithelial damages that accompany aging. To investigate age-related fibrosis in our aging rat model, four different-aged kidneys (6, 12, 18, and 24 months) were examined. The expression levels of extracellular matrix (ECM) genes including Acta2, Fn, Col1a, Col3a1, Col4a, and Col11a1 were all significantly increased at 24 months of aging (Figure 1A). We next checked markers for mesenchymal and epithelial cells. Interestingly, only markers of mesenchymal cells (Vim and Fsp) were increased during aging, whereas epithelial cell markers (Krt8 and Cdh1) were not affected by aging (Supplemental Figure 1A). The ECM protein levels were further investigated. The ECM proteins started to increase at 18 months of age, and the extent was dramatically increased at 24 months (Figure 1B). A histologic analysis also confirmed age-associated ECM accumulation and fibrosis. Immunohistochemistry (IHC) staining of collagen I was dramatically increased in the interstitial region of aged kidneys (Figure 1C). However, the expression of collagen III did not show any differences between young and aged kidneys (Supplemental Figure 1B). The accumulation of ECM proteins also showed an increase when detected by Sirius red and Masson’s trichrome staining (Figure 1D). Fibrotic-positive regions were significantly increased at 24 months of aging (Figure 1E). We next measured serum urea and creatinine levels to assess age-related kidney function decline. Serum urea levels significantly increased at 24 months of aging (Figure 1F); however, creatinine levels did not change significantly during aging (Supplemental Figure 1C). Renal periodic acid–Schiff staining and evaluation were further implemented in aged kidney. Aged rat showed significantly increased renal damages in both glomerular and tubule regions (Supplemental Figure 1, D–F). To evaluate renal epithelial damage, we measured kidney injury molecule-1 (KIM-1) levels in the serum and kidney. Serum and kidney KIM-1 levels were also significantly increased during aging (Figure 1G, Supplemental Figure 1, G and H). Taken together, we concluded that aging significantly increases renal fibrosis and damage with impaired renal function.

Figure 1.

Aged kidneys show increased renal fibrosis, function decline, and epithelial damages. Four different-aged kidneys (6, 12, 18, and 24 months, n=8) were examined to examine the effect of aging on renal changes. (A) Gene levels of ECM proteins were measured by qPCR. *P<0.05 versus 6 months. (B) Protein levels of ECM proteins were measured by western blotting. Actin was used as the loading control. (C) Collagen I expression in young and aged kidney was visualized by IHC. Scale bar=50 μm. (D) Masson’s trichrome (MT) and Sirius red (SR) staining were performed to visualize fibrosis in kidney. Scale bar=100 μm. (E) Quantification of SR-stained fibrosis in kidneys. *P<0.05 versus 6 months. (F) Urea levels were measured by biochemical method in serum to measure kidney function. *P<0.05 versus 6 months. (G) KIM-1 levels were detected in serum by ELISA to measure renal epithelial damages. *P<0.05 versus 6 months.

Aging Increases Renal Lipid Accumulation

To investigate the relationship between increased fibrosis and changes in lipid metabolism during aging, lipid metabolism–associated changes were further evaluated. Levels of extracted TG were significantly increased at 24 months of aging (Figure 2A). Accumulated TG levels were further visualized via oil red O–staining analysis. As expected, renal lipid accumulation was observed in 24-month-old kidneys (Figure 2B). Oil red O–positive regions were mostly located in renal tubular epithelial cells (Figure 2B). We further detected an overall increase of vacuoles in renal tubules during aging (Figure 2, C and D). These data indicated that lipid accumulation was increased during renal aging, and the accumulation was especially increased in the tubular epithelial region.

Figure 2.

Aging increases lipid accumulation with alteration of various lipid metabolism–associated transcriptional factors. Four different-aged kidneys (6, 12, 18, and 24 months, n=8) were examined to see the effect of aging on renal changes. (A) TGs were quantified in aging kidney model. *P<0.05 versus 6 months. (B) Kidneys (6 months and 24 months) were stained with oil red O to visualize lipid accumulation. Scale bar=50 μm. (C) Representative H&E staining shows increased vacuoles in renal tubules during aging. (D) Number of vacuoles was quantified by counting vacuoles per tubule. *P<0.05 versus 6 months. (E) Gene levels of transcription factors associated with lipid metabolism were measured by qPCR. *P<0.05 versus 6 months. (F) Nuclear protein expressions of transcription factors associated with lipid metabolism were measured by western blotting. TFIIB was used as the loading control. (G) Western blot results from 4 independent experiments were quantified by densitometry. *P<0.05 versus 6 months.

In the regulation of lipid metabolism, various transcriptional factors play important roles.²³ We first checked the mRNA levels of these transcription factors in the four different-aged kidneys; however, there were no significant changes during aging (Figure 2E). Because it is generally accepted that the localization of transcription factors is important, we next checked the nuclear protein expression of these transcriptional factors. Surprisingly, unlike mRNA levels, there were remarkable changes in protein expression of the transcription factors in the nucleus. Nuclear expression of PPARα and Farnesoid X receptor (FXR) were dramatically decreased in 24-month-old rat kidneys, whereas SREBP1 and ChREBP were rather increased during aging (Figure 2, F and G). Taken together, these data implicated that dysregulated lipid metabolism during aging might be associated with changes in nuclear expression of transcriptional factors that regulate lipid metabolism.

Aging Decreases PPARα and FAO Pathway in Renal Tubule Epithelial Region

Recent studies have suggested that defective FAO by impaired PPARα in renal tubular epithelial cells plays a pivotal role in kidney fibrosis development.¹⁵ On the basis of our results and those of previous studies, we hypothesized that impairments in PPARα and the FAO pathway may be important in the development of age-associated renal fibrosis. First, we checked whether the decreased nuclear expression of PPARα was due to defects in translocation. However, cytosolic fractions of aged kidney also showed decreased PPARα expression (Figure 3A). We next checked the expression levels of well known PPARα gene targets by quantitative PCR (qPCR). Most of the PPARα target genes (Cpt1a, Acox1, Lcad, Mcad, Acad10, and Hadh) were significantly decreased at 24 months of aging (Figure 3B). However, other lipid metabolism–associated genes that play important roles in FA synthesis and TG accumulation did not show significant changes (Supplemental Figure 2, A and B). We further checked carnitine palmitoyltransferase 1α (CPT1α) and acyl-coenzyme A oxidase 1 (ACOX1) protein levels. Aged rat kidneys showed significantly decreased CPT1α and ACOX1 protein levels (Figure 3C). The decreased levels of PPARα and its target protein, ACOX1, were further verified by IHC. PPARα and ACOX1 primarily localized to the tubular epithelial regions where the lipids accumulated (Figure 3D). Furthermore, the expression of PPARα and ACOX1 detected by IHC was also decreased during aging (Figure 3D). Taken together, we found that the levels of PPARα and its target proteins, CPT1α and ACOX1, were significantly decreased during aging.

Figure 3.

Aging decreases PPARα and FAO signaling pathways in the renal tubule epithelial region. Four different-aged kidneys (6, 12, 18, and 24 months, n=8) were examined to see the effect of aging on renal changes. (A) Nuclear and cytosolic PPARα levels were measured in the rat kidney. TFIIB was used as the loading control for the nuclear fraction, and actin was used as the loading control for the cytosolic fraction. (B) Gene levels of PPARα target proteins were measured by qPCR. *P<0.05 versus 6 months. (C) Protein expression of CPT1α and ACOX1 was measured by western blotting. Actin was used as the loading control. (D) PPARα and ACOX1 expression in young and aged kidneys was visualized by IHC. Scale bar=50 μm. (E) Protein expression levels of signals upstream of PPARα (AMPK, p-AMPK, and PGC1α) were measured by western blotting. Actin was used as a loading control. (F) The miRNA levels that are known to regulate PPARα expression were detected by qPCR. *P<0.05 versus 6 months.

Because there were discrepancies between gene and protein levels of PPARα during aging, we next checked the upstream signaling of PPARα. However, the levels of AMP-activated protein kinase (AMPK) and activated (phosphorylated) AMPK were both increased during aging (Figure 3E). The protein levels of PGC1α, which is an important coactivator of PPARα, were not changed during aging (Figure 3E). We further checked various miRNAs that can silence PPARα translation by epigenetic mechanisms. Interestingly, among the seven miRNAs that are known to suppress PPARα translation, levels of miR-21, miR-34a, and miR-155 were significantly increased during aging (Figure 3F). In summary, decreased PPARα expression during aging may be affected by miRNA that regulates the translation of PPARα epigenetically.

PPARα Translation and FAO Are Suppressed by miR-21 in Renal Tubular Epithelial Cells

We next performed in vitro experiments to examine whether increased miRNA levels during aging suppress PPARα expression and affect FAO in renal tubular epithelial cells. NRK52E renal tubular epithelial cells were used to examine the effects of miRNAs on PPARα and the FAO pathway. Among the three miRNAs (miR-21, miR-34a, and miR-155) that were increased during aging, only miR-21 efficiently decreased PPARα protein expression in NRK52E cells without changes in PPARα gene levels (Figure 4A, Supplemental Figure 3A). Furthermore, miR-21 also decreased PPRE binding activity as deduced from a luciferase assay (Figure 4B). PPARα target gene levels were further analyzed under miR-21 treatment conditions. Most, but not all, of the PPARα target genes were downregulated by an miR-21 treatment (Figure 4C). Decreased CPT1α and ACOX1 levels were further confirmed by measuring their protein levels (Figure 4D). We next measured oxygen consumption rate (OCR) under the miR-21 treatment conditions. miR-21 treatment decreased OCR levels both under the normal and oleic acid treatment conditions (Figure 4E). Furthermore, consistent with decreased OCR, miR-21 treatment reduced ATP production in cells (Supplemental Figure 3B). However, treatment with miR-21 resulted in marginal changes in the total lipid content of the cells (Figure 4, F and G). Next, we pretreated cells with 200 μM of oleic acid to induce lipid stress in the epithelial cells. Treatment with oleic acid significantly increased intracellular lipid content when measured quantitatively or visualized by oil red O staining (Figure 4, F and G). The miR-21 treatment further increased the accumulation of intracellular lipids (Figure 4, F and G). We further checked FAO-related gene changes in the same conditions. Although oleic acid treatment regulated FAO-related gene expressions differently, miR-21 constantly reduced those gene expressions regardless of oleic acid treatment (Supplemental Figure 3C). Furthermore, oleic acid with miR-21 treatment induced a decrease in cell viability, suggesting the role of FAO in regulating lipotoxicity under the stress condition (Supplemental Figure 3D). These data suggest that miR-21 suppresses the translation of PPARα in renal epithelial cells, thus inhibiting FAO and inducing lipid accumulation.

Figure 4.

miR-21 suppresses PPARα translation and FAO in NRK52E cells. Suppression of PPARα translation and FAO by miR-21 in renal tubular epithelial cells. NRK52E renal epithelial cells were used to investigate the role of miRNAs in PPARα and FAO signaling. (A) NRK52E cells were transfected with selected miRNAs (miR-21, miR-34a, and miR-155) and protein expression of PPAR was measured by western blotting. Actin was used as the loading control. (B) NRK52E cells were transfected with selected miRNAs (miR-21, miR-34a, and miR-155) and PPAR activities were measured by the PPRE luciferase activity. (C) Gene levels of PPARα target proteins were measured by qPCR after miR-21 transfection in NRK52E cells. *P<0.05 versus nontreated control. (D) Protein expression of CPT1α and ACOX1 was measured by western blotting in miR-21-treated NRK52E cells. Actin was used as the loading control. (E) OCR was measured under normal and miR-21-treated conditions. Oleic acid (200 μM) and oligomycin (1 μM) were treated at designated times. (F) Cellular TG levels were extracted and quantified in NRK52E cells treated with 200 μM oleic acid and/or miR-21. #P<0.05 versus nontreated control. *P<0.05 versus oleic acid treatment. (G) Cellular TG was visualized by oil red O staining in NRK52E cells treated with 200 μM oleic acid and/or miR-21. Scale bar=50 μm.

The FAO Pathway Is Important during TGF-β-Induced Fibrosis Signaling in Renal Tubular Epithelial Cells

It has been suggested that TGF-β-induced miR-21 plays an important role during the mediation of fibrosis signaling in renal epithelial cells.²⁴ To this end, we further investigated the roles of miR-21-mediated PPARα silencing and an impaired FAO pathway during fibrosis signaling in renal epithelial cells. First, we measured miR-21 expression changes on TGF-β-treated NRK52E cells. TGF-β significantly induced miR-21 expression as previously described (Figure 5A). We next checked whether TGF-β treatment affected lipid accumulation that was induced by oleic acid. TGF-β treatment further increased oleic acid–induced lipid accumulation (Figure 5, B and C). To further demonstrate the role of miR-21-mediated PPARα silencing and the FAO pathway on TGF-β, an miR-21 inhibitor (anti-miR-21 oligo) was utilized. TGF-β treatment reduced PPARα expression in epithelial cells, and anti-miR-21 abrogated this reduction (Figure 5D). PPARα target genes and protein expression changes were also reduced by TGF-β treatment and blocked by anti-miR-21 treatment (Figure 5, D and E). We next measured the effect of anti-miR-21 on OCR under the TGF-β treatment condition. Treatment with anti-miR-21 blocked TGF-β-induced OCR level decrease (Figure 5F) and rescued ATP levels (Supplemental Figure 4). In addition, anti-miR-21 treatment significantly reduced oleic acid–induced lipid accumulation under the TGF-β treatment conditions (Figure 5, G and H). We next checked whether lipid stress increases TGF-β-induced fibrosis-related signaling in cells. Although oleic acid treatment itself did not alter fibrosis-related gene expression, oleic acid under the TGF-β treatment conditions significantly increased fibrosis-related gene expression (Supplemental Figure 5). The effect of oleic acid was partially blocked by anti-miR-21 treatment (Figure 5I). Collectively, these data implicated the importance of the FAO pathway in TGF-β-induced fibrosis signaling in renal tubular epithelial cells, especially under excessive lipid stresses.

Figure 5.

FAO signaling is important in TGF-β-induced fibrosis signaling in renal tubular epithelial cells. NRK52E renal epithelial cells were transfected with miR-21 under TGF-β treatment conditions to investigate the role of FAO signaling in fibrosis. (A) The miR-21 levels were measured by qPCR under TGF-β treatment conditions. (B) Cellular TG was visualized by oil red O staining in NRK52E cells treated with 200 μM oleic acid or/and TGF-β. Scale bar=50 μm. (C) Cellular TG levels were extracted and quantified in NRK52E cells treated with 200 μM oleic acid or/and TGF-β. #P<0.05 versus nontreated control. *P<0.05 versus oleic acid treatment. (D) Protein expression of PPARα, CPT1α, and ACOX1 was measured by western blotting in TGF-β and anti-miR-21–treated NRK52E cells. Actin was used as the loading control. (E) Gene levels of PPARα target proteins were measured by qPCR after TGF-β and anti-miR-21 treatment in NRK52E cells. #P<0.05 versus nontreated control. *P<0.05 versus TGF-β treatment. (F) OCR was measured under designated conditions. Oleic acid (200 μM) and oligomycin (1 μM) were treated at designated times. (G) Cellular TG was visualized by oil red O staining in NRK52E cells treated with designated conditions. Scale bar=50 μm. (H) Cellular TG levels were extracted and quantified in NRK52E cells treated with designated conditions. #P<0.05 versus nontreated control. *P<0.05 versus TGF-β and anti-miR-21 treatment. (I) Changes in expression of ECM genes in NRK52E cells. Gene expression levels of ECM proteins were measured by qPCR under oleic acid or/and TGF-β treatment conditions. #P<0.05 versus nontreated control. *P<0.05 nontreated control versus TGF-β, oleic acid, and anti-miR-21 treatment.

CR Restores PPARα Expression and Slows Renal Fibrosis

CR is one of the most powerful antiaging regimens that can retard aging and age-related diseases.²⁵ To investigate whether CR can restore defective lipid metabolism during aging, calorie-restricted aged rats were used. CR significantly rescued aging-induced decreases in PPARα expression (Figure 6A). Furthermore, CPT1α and ACOX1 expression levels were also increased in response to CR (Figure 6A). The increased PPARα and ACOX1 levels were confirmed by IHC analysis. Increased expressions of PPARα and ACOX1 were detected especially in the epithelial region of CR kidneys (Figure 6B). Other PPARα target genes that modulate FAO were also increased by CR (Figure 6C). The increased expression of FAO regulators was associated with decreased lipid accumulation in the CR kidneys (Figure 6, D and E). In contrast, other lipid metabolism–associated genes were not affected by CR (Supplemental Figure 6, A and B). We further detected lipid vacuoles in renal tubules during CR (Figure 6, F and G). CR significantly reduced the number of vacuoles in tubules. PPARα-silencing miRNAs were further investigated. CR resulted in significant reductions in the levels of all three miRNAs (miR-21, miR-34a, miR-155) that showed increase during aging (Figure 6H). Together, these data suggested that CR restored PPARα expression through the regulation of miRNAs, implicating the importance of the FAO pathway in the regulation of renal lipid accumulation.

Figure 6.

CR restores PPARα expression and FAO pathways with decreased lipid accumulation in the kidneys. Young (6-month), old (24-month), and calorie-restricted old (24-month) rat kidneys were used to investigate the effects of CR on age-related renal fibrosis (n=7–8). (A) Nuclear protein expression of PPARα and cytosolic protein expression of CPT1α and ACOX1 were measured by western blotting. TFIIB and actin were used as the loading controls, respectively. (B) PPARα and ACOX1 expression in kidney samples were visualized by IHC. Scale bar=50 μm. (C) Gene levels of PPARα target proteins were measured by qPCR. #P<0.05 versus young. *P<0.05 versus old. (D) TGs were quantified in the kidneys. #P<0.05 versus young. *P<0.05 versus old. (E) Kidneys were stained with oil red O to visualize lipid accumulation. Scale bar=50 μm. (F) Representative H&E staining shows increased vacuoles in renal tubules during aging and CR. (G) Number of vacuoles was quantified by counting vacuoles per tubule. #P<0.05 versus young. *P<0.05 versus old. (H) The miRNA levels that were changed during aging were detected by qPCR. #P<0.05 versus young. *P<0.05 versus old.

Because defective FAO is known to play a role in the development of fibrosis, we further examined the effects of CR on age-related fibrosis development. CR significantly reduced gene expression of several ECM components (Figure 7A). CR also reduced protein expression of ECM components including procollagen I, collagen I, and collagen IV (Figure 7B). IHC staining of collagen I, Sirius red, and Masson’s trichrome staining also showed decreased expression of ECM in the interstitial region of the kidneys (Figure 7C). Consistently, fibrotic scores were decreased in the CR group when the area positive for Sirius red staining was calculated (Figure 7D). Furthermore, CR improved renal function with decreased epithelial damage (Figure 7, E and F, Supplemental Figure 6, C–G). In summary, CR significantly delayed aging-induced renal fibrosis with the restoration of PPARα expression and FAO-associated proteins.

Figure 7.

CR prevents the development of age-related renal fibrosis. Young (6-month), old (24-month), and CR-old (24-month) rat kidneys were used to investigate the effects of CR on age-related renal fibrosis (n=7–8). (A) Gene levels of ECM proteins were measured by qPCR. #P<0.05 versus young. *P<0.05 versus old. (B) Protein levels of ECM proteins were measured by western blotting. Actin was used as the loading control. (C) Collagen I expression was visualized by IHC, and Masson’s trichrome (MT) and Sirius red (SR) staining were performed to visualize fibrosis. Scale bar=50 μm. (D) Quantification of SR-stained fibrosis in kidneys. #P<0.05 versus young. *P<0.05 versus old. (E) Urea levels were measured by biochemical methods in the serum to measure kidney function. #P<0.05 versus young. *P<0.05 versus old. (F) KIM-1 levels were detected in the serum via ELISA to measure renal epithelial damages. #P<0.05 versus young. *P<0.05 versus old.

PPARα Deficiency Increases Renal Lipid Accumulation and Accelerates Renal Fibrosis

To confirm the role of PPARα in aging-induced renal fibrosis, we subjected PPARα^−/− mice and sex-matched control littermates to aging. We first compared changes in lipid metabolism. As expected, aged control mice showed decreased PPARα and FAO-associated gene expression (Figure 8, A and B). Compared with aged control littermates, PPARα^−/− mice showed reduced expression of FAO-associated proteins and genes (Figure 8, A and B). In contrast, other lipid metabolism genes were not significantly changed in aged PPARα^−/− mice except for Scd (Supplemental Figure 7, A and B). Aged PPARα^−/− mice showed significantly higher lipid accumulation compared with littermates (Figure 8C). Interestingly, aged PPARα^−/− mouse kidneys showed distinct histologic structures in the renal tubule region when analyzed by hematoxylin and eosin (H&E) staining (Figure 8D). Aged PPARα^−/− mouse kidneys showed increased vacuoles in tubules (Supplemental Figure 8, A and B). These distinct histologic structures were found to be due to excessive lipid droplet formation when the kidneys were subjected to oil red O staining (Figure 8D). These data suggested that PPARα deficiency significantly decreases FAO with high lipid accumulation in the tubule epithelial region during aging.

Figure 8.

PPARα deficiency increases renal lipid accumulation and accelerates renal fibrosis. PPARα^−/− mice and their littermates were subjected to aging to investigate the effect of PPARα on renal lipid metabolism and fibrosis changes during aging (n=5–7). (A) Gene levels of PPARα target proteins were measured by qPCR. #P<0.05 versus 7-month wild type (WT). *P<0.05 versus 20-month WT. (B) Nuclear protein expression of PPARα and cytosolic protein expression of CPT1α and ACOX1 were measured by western blotting. TFIIB and actin were used as the loading controls, respectively. (C) TGs were quantified in the kidneys. #P<0.05 versus 7-month WT. *P<0.05 versus 20-month WT. (D) Kidneys were stained with H&E to visualize structural differences, and oil red O staining was used to visualize lipid accumulation. Scale bar in H&E staining=100 μm. Scale bar in oil red O staining=50 μm. (E) Gene levels of ECM proteins were measured by qPCR. #P<0.05 versus 7-month WT. *P<0.05 versus 20-month WT. (F) Protein levels of ECM proteins were measured by western blotting. Actin was used as the loading control. (G) Collagen I expression was visualized by IHC, and Masson’s trichrome (MT) and Sirius red (SR) staining were performed to visualize fibrosis. Scale bar=100 μm. (H) Quantification of SR-stained fibrosis in the kidneys. #P<0.05 versus 7-month WT. *P<0.05 versus 20-month WT. (I) Urea levels were measured by biochemical methods in the serum to measure kidney function. #P<0.05 versus 7-month WT. *P<0.05 versus 20-month WT. (J) KIM-1 levels were detected in the serum by ELISA to measure renal epithelial damages. #P<0.05 versus 7-month WT. *P<0.05 versus 20-month WT. ORO, oil red O; Ppara, Peroxisome proliferator-activated receptor alpha.

We further investigated whether PPARα-deficient mice exhibit accelerated renal fibrosis during aging. Aged PPARα^−/− mice developed significantly more interstitial fibrosis than control littermates as detected by mRNA and protein expression levels of ECM proteins (Figure 8, E and F). IHC staining of collagen I, Sirius red, and Masson’s trichrome staining were also higher in the interstitial region of aged PPARα^−/− kidneys (Figure 8G). Consistently, fibrotic scores were increased in aged PPARα^−/− kidneys when calculated by using Sirius red–staining–positive area (Figure 8H). Serum urea levels and histologic damages were also increased only in the aged PPARα^−/− mice group (Figure 8I, Supplemental Figure 8, C–E). Similarly, the epithelial injury marker KIM-1 was also expressed at high levels in the aged PPARα^−/− mouse serum and kidneys (Figure 8J, Supplemental Figure 8, F and G). Collectively, these data indicated that PPARα deficiency not only increases renal lipid accumulation but also accelerates renal interstitial fibrosis, suggesting the importance of PPARα on the development of age-associated renal fibrosis.

Discussion

The role of lipid metabolism has been recently implicated in various renal diseases.^26,27 The kidney is not a major metabolic organ that actively participates in the systemic regulation of lipid metabolism. However, recent evidence suggests that lipid accumulation in nonadipose tissues can contribute to organ injury.²⁸ Clinical observations have implicated a potential association between renal lipid accumulation and CKD development.²⁹ Furthermore, there is an abundance of animal data demonstrating an association between renal lipid accumulation and kidney dysfunction, including models of metabolic and nonmetabolic renal diseases.^15,30 In addition, several previous reports have examined the effects of renal lipid metabolism on age-associated renal diseases.^31,32 Aged animals show higher lipid accumulation with increased glomerulosclerosis and tubulointerstitial fibrosis. However, although the occurrence of renal lipid accumulation in aged kidneys was coincident in our aged model, the regions of lipid accumulation were different compared with those reported by previous studies. Our results clearly indicated that the site of lipid accumulation was the renal tubular epithelial region.

Recently, roles of the impaired FAO pathway in the development of renal interstitial fibrosis have been investigated.¹⁵ The authors found that renal tubular epithelial cells depend critically on FAO as their energy source, and depressed FAO was associated with intracellular lipid accumulation. This observation was consistent with previous reports demonstrating that mitochondrial β-oxidation of FAs is a major source for renal energy production, which has a high energy demand with relatively low glycolytic capacity.³³ They also suggested that impaired FAO was owing to a decreased PPARα-PGC1α axis during renal fibrosis development. We also found that age-associated renal lipid accumulation was associated with an impaired FAO pathway. Critically, the protein levels of CPT1α, which is the rate-limiting enzyme of FAO, showed great reduction in the tubular epithelial region. However, we found that upstream regulators of PPARα, AMPK and PGC1α, which were shown to be significantly reduced in the fibrosis model, were actually increased or unchanged during aging. These differences led us to investigate other mechanisms associated with decreased PPARα signaling during aging.

miRNAs are endogenously encoded, evolutionarily conserved small RNAs that regulate gene expression predominantly by facilitating the degradation and translation inhibition of target mRNAs, thus having a powerful role in regulating various biologic processes.³⁴ Additionally, miRNAs have also been implicated in regulating various kidney diseases, including renal carcinoma, diabetic nephropathy, and AKI.³⁵ We found a discrepancy between the gene and protein levels of PPARα, which suggests that there may be epigenetic mechanisms regulating PPARα translation. Among the miRNAs that are known to regulate PPARα, miR-21 showed great efficiency in regulating PPARα expression in renal tubular epithelial cells. Interestingly, miR-21^−/− mice have been reported to exhibit far less interstitial fibrosis in response to kidney injury, and the lipid metabolism pathway regulated by PPARα has been implicated in this model.²² In addition, transfection of an miR-21 shRNA plasmid halted the progression of renal fibrosis in established obstructive nephropathy.²⁴ We demonstrated the importance of the miR-21-PPARα axis in age-related renal fibrosis. Likewise, miR-21-mediated PPARα silencing had significant effects in the regulation of lipid metabolism and TGF-β response in renal tubular epithelial cells. Given the previous reported data and our results, miR-21 has been shown to play a role in the regulation of PPARα and lipid metabolism as an important regulator of renal fibrosis.

Through the regulation of various biologic processes ranging from metabolic changes to epigenetic processes, CR is known to regulate aging and age-related diseases in diverse species.²⁵ We hypothesized that changes in renal lipid metabolism are important in inducing age-associated renal fibrosis; therefore, the effects of CR on renal lipid metabolism and fibrosis were further investigated. We found that CR significantly reduced renal lipid amounts with increased PPARα expression and FAO-associated genes. CR also reduced miRNAs that regulate PPARα silencing during aging. These changes in lipid metabolism were associated with decreased renal fibrosis during aging. The effects of CR on age-associated fibrosis have been reported in other studies. Although the suggested mechanisms vary between the model and tissues used, it is evident that CR can alleviate age-related kidney, heart, liver, and aorta fibrosis.^32,36,37 As a strong metabolic modifier, CR can alleviate renal fibrosis by regulating abnormal lipid metabolism during aging.

It has been suggested that sirtuin 1 (SIRT1) is a main regulator that connects calorie intake and life span.^38,39 Interestingly, SIRT1 has been shown to directly or indirectly regulate PPARα expression and activity, and is also associated with renal fibrosis, aging, and other age-related diseases.^40,41 Furthermore, SIRT1 is also regulated by miRNAs through epigenetic regulation.⁴² Although we did not check the SIRT1 levels in our animal models, previous studies have shown decreased SIRT1 expression and activity during renal aging.⁴³ Because SIRT1 mediates various beneficial roles during CR, a possible role of SIRT1 in age-related renal fibrosis and its association with PPARα and CR should be further considered.

Increasing evidence has also demonstrated an association between PPARα and aging.^44,45 The decreased expression or activity of PPARα has been detected during aging in several tissues including kidney, heart, and spleen.^46–48 More direct evidence on the relationship between PPARα and aging comes from observations of PPARα^−/− mice. Accelerated physiologically aged phenotypes were first reported in PPARα^−/− mice, along with decreased longevity and increased age-dependent lesions.^49–51 In the liver, the roles of PPARα during aging have also been shown. PPARα^−/− mice and hepatocyte-specific PPARα^−/− mice showed significant alterations of systemic lipid metabolism that lead to early hepatic steatosis during aging.⁵² Although it is evident that PPARα deficiency accelerates renal aging with defects in systemic lipid metabolism, the phenotypes of aged PPARα^−/− mice regarding renal lipid metabolism have not been previously reported. We found that renal lipid metabolism was severely impaired in aged PPARα^−/− mice with increased lipid accumulation. Impaired lipid metabolism in the kidney was associated with early onset of renal fibrosis in aged PPARα^−/− mice. However, we cannot exclude other possibilities that can induce an early onset of the fibrosis phenotype in aged PPARα^−/− mice, because we used whole-body PPARα^−/− mice. Performing aging studies in kidney epithelial–specific PPARα^−/− mice will be important.

To conclude, these studies identify PPARα and the FAO pathway as important regulators in kidney lipid metabolism that amplifies age-associated renal fibrosis (Supplemental Figure 9). The post-transcriptional regulator of PPARα, which is miR-21, critically decreased FAO pathways in renal epithelial cells with increased fibrotic responses. Antiaging CR rescued age-associated PPARα and the FAO pathway with slowed renal fibrosis. PPARα^−/− mice showed a severely disrupted renal FAO pathway with high accumulation of lipids, and showed earlier onset of renal fibrosis, demonstrating the importance of PPARα in the development of age-related renal fibrosis. Future studies may explore the effects of impaired PPARα and FAO pathways on human aged kidney samples. It will also be necessary to assess the pharmacologic interventions (e.g., PPARα agonist, miR-21 antagonist) for the treatment of age-associated renal fibrosis. If so, this could provide novel therapeutics to intervene in renal fibrosis and functional decline during aging.

Concise Methods

Animals

To investigate the changes in lipid metabolism and fibrosis during aging, 6-, 12-, 18-, and 24-month-old SD rats (n=8) were used. To compare the effects of CR on lipid metabolism and fibrosis changes, young (6 months), aged (24 months), and CR (60% of ad libitum for 1 month) aged (24 months) rats (n=7–8) were used. To evaluate the effects of PPARα deficiency on age-associated renal fibrosis, age-matched (3, 7, and 20 months) PPARα^−/− mice and their wild-type littermates (n=5–7) were used. Animal studies were designed by the Aging Tissue Bank, approved by the Institutional Animal Care Committee of Pusan National University, and performed in accordance with the guidelines for animal experimentation issued by Pusan National University. Rats and mice were maintained under controlled environmental conditions under a 12-hour light/dark cycle, and allowed ad libitum access to water and a standard laboratory diet except for the CR study. PPARα^−/− mice were genotyped by PCR as previously described.⁵³ For all animal experiments, serum was collected for biochemical analysis. Kidneys were collected and either frozen immediately in liquid nitrogen for qPCR, western blotting, and biochemical tests, or fixed in neutral-buffered formalin for histochemical examination. For long-term storage, tissue samples were moved to a −80°C deep freezer located at the Aging Tissue Bank in Pusan National University.

Histologic Analysis

Kidneys were fixed in 10% neutral formalin and paraffin-embedded sections were stained with H&E. Sirius red staining and Masson’s trichrome staining were performed as described previously to determine the degree of aging-associated fibrosis.⁵⁴

Statistical Analyses

The t test was used to analyze differences between two groups, and ANOVA was used to analyze intergroup differences. P values of <0.05 were considered statistically significant. The analysis was performed using GraphPad Prism 5 (GraphPad software).

Detailed protocols for the other experiments are provided in the Supplemental Material.

Disclosures

None.