Age-related effects of fenofibrate on the hepatic expression of sirtuin 1, sirtuin 3, and lipid metabolism-related genes

Background: Sirtuin 1 (Sirt1) and sirtuin 3 (Sirt3) participate in the regulation of lipid metabolism. Our aim was to investigate the effects of the hypolipemic drug fenofibrate (FN) on hepatic Sirt1 and Sirt3 expression, in relation to the expression of lipid metabolism-related genes and in the context of aging. Methods and Re-sults: Young and old male Wistar rats were fed standard chow or supplemented with 0.1% or 0.5% FN for 30 days ( n =7–10 in each group). In young rats, 0.1% FN did not affect Sirt1 expression, however, 0.5% FN decreased Sirt1 and both doses reduced Sirt3 protein levels. In old rats, 0.5% FN decreased hepatic Sirt1 mRNA and both doses reduced Sirt1 protein levels, but not Sirt3 expression. Although hepatic Pparα protein levels did not change, FN treatment of young rats induced Cpt1b expression, whereas Lcad , Acox1 , Pmp70 , and Hmgcs2 expression increased only after 0.1% FN, and Fas2 expression decreased after 0.5% FN. In the liver of old rats, both doses increased Cpt1b and Lcad expression. Only 0.1% FN increased Pmp70 and Hmgcs2 expression, and only 0.5% FN increased Acox1 and Fas2 mRNA levels. Conclusions: Treatment with fenofibrate at low or high doses may downregulate the expression of Sirt1 and Sirt3 proteins in the rat liver. The dosage of FN affects molecular changes, and aging alters the response to 0.5% FN


INTRODUCTION
Aging decreases the ability to respond to stress and leads to the accumulation of molecular damage (Pyo et al., 2020). Aging-associated changes in energy metabolism and nutrient sensing pathways may lead to disorders in lipid homeostasis, which manifest as elevated concentrations of triglycerides, total cholesterol, and LDL cholesterol in blood plasma . The hypolipemic drug fenofibrate (FN) is used to counteract disturbances in lipid metabolism. FN decreases serum triglyceride and VLDL-and LDL-cholesterol levels, while it increases serum HDL-cholesterol (Sahebkar et al., 2017). In the liver, FN stimulates fatty acid oxidation in mitochondria and peroxisomes through the induction of peroxisome proliferator-activated receptor alpha (PPARα) (Gebel et al., 1992). FN exerts additional beneficial effects through PPARα-dependent and independent mechanisms, e.g. in cardiac, renal, and vascular disorders (Balakumar et al., 2019). FN may thus offer new treatment venues for conditions besides hyperlipidaemia. However, it has not been well investigated whether FN affects the expression of sirtuins in the liver.
Sirtuins (SIRT1-SIRT7) are a family of NAD + -dependent deacetylases, which target histones and non-histone proteins (Zhao et al., 2020). Sirtuins can protect cells against metabolic and oxidative stresses, and may therefore safeguard against some aging-associated alterations, as evidenced by genetic studies in mice (Rodgers & Puigserver, 2007;Pfluger et al., 2008;Zhao et al., 2020). The relationships between the activity of sirtuins and energy metabolism make them a potential therapeutic target for the treatment of metabolic diseases. Two members of the sirtuin family are particularly important for lipid metabolism. SIRT1, the best-studied member of the sirtuin family, decreases blood plasma cholesterol and triglyceride concentrations by inhibitory deacetylation of sterol regulatory element-binding proteins (SREBP) in the liver (Walker et al., 2010). The SREBP belong to transcription factors that control lipid homeostasis by promoting the expression of lipogenic (e.g. fatty acid synthase, FAS) and cholesterogenic genes in the fed state (Walker et al., 2010). Moreover, SIRT1 deacetylates peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α) in the liver (Rodgers et al., 2005;Rodgers & Puigserver, 2007), skeletal muscle and adipose tissue (Feige et al., 2008). This post-translational regulation of PGC-1α leads to the induction of gluconeogenic genes and glucose output in the liver (Rodgers et al., 2005) and enhances oxidative metabolism in muscles and brown adipose tissue (Feige et al., 2008). SIRT1 also interacts with the transcription factor PPARα at the PPAR response element (PPRE) sequence of target genes (e.g. fibroblast growth factor type 21, FGF21). Upon this interaction with SIRT1, PPARα mediates the adaptive response to fasting and starvation by transcriptional activation of genes involved in fatty acid oxidation (Purushotham et al., 2009).
The mitochondrial sirtuin SIRT3 regulates multiple mitochondrial enzymes both at the transcriptional level and through their reversible deacetylation in mitochondria. SIRT3 is induced during fasting, and through deacetylation increases the activity of longchain acyl-coenzyme A dehydrogenase (LCAD), a key enzyme of mitochondrial fatty acid β-oxidation . Similarly, SIRT3 deacetylates and activates 3-hydroxy-3-methylglutaryl-CoA synthase 2 (HMGCS2) , the rate-limiting enzyme in the ketone body β-hydroxybutyrate synthesis. In addition, SIRT3 activates enzymes of the tricarboxylic acid (TCA) cycle and the urea cycle, and regulates reactive oxygen species levels (reviewed by Zhao et al., 2020). Thus, SIRT3 activity is involved in the regulation of multiple metabolic pathways in mitochondria. Importantly, both SIRT1 and SIRT3 may mitigate metabolic dysfunctions which are often associated with aging. Transgenic models have shown that increased SIRT1 activity protects against insulin resistance, type 2 diabetes (T2D) and diabetic complications, high-fat diet-induced hepatic steatosis, and adipose tissue inflammation (Rodgers & Puigserver, 2007;Pfluger et al., 2008;Guo et al., 2015). Mice with whole-body SIRT3 knock-out showed peripheral insulin resistance due to increased fatty acid oxidation and reduced glucose uptake in skeletal muscle (Lantier et al., 2015). When SIRT3 KO mice were fed a high-fat diet (HFD) they developed hyperlipidemia, impaired glucose tolerance, insulin resistance, and steatohepatitis (Hirschey et al., 2011).
Despite the well-established roles of SIRT1 and SIRT3 in the control of fuel metabolism and suppression of some mechanisms of aging, there is limited data on the expression of these sirtuins in response to pharmacological challenge with fenofibrate, and in the context of ageing. It was reported that Sirt1 protein levels were upregulated by FN in TNFα-treated adipocytes (Wang et al., 2013), the kidney (Kim et al., 2016), endothelial cells , and cardiac muscle (Liu et al., 2016) in animal models, and Sirt3 protein abundance decreased in the mouse liver after bezafibrate (100 mg/kg/day for 4 weeks) (Barger et al., 2017). However, it is not clear whether FN affects Sirt1 or Sirt3 levels in the liver. We, therefore, undertook to examine: (1) the effect of aging on hepatic expression of Sirt1 and Sirt3 in the rat; (2) whether FN affects the hepatic expression of Sirt1 and Sirt3 in non-obese, normolipemic rats; and (3) whether this effect of FN is modulated by physiological aging. In our previous work, we have demonstrated that FN had a significant hypolipemic effect in normally-bred old rats (Zubrzycki et al., 2020). Therefore, in the present study, we decided to investigate the molecular changes brought about by FN treatment in rat liver in old as compared to young rats bred in standard, non-obesogenic conditions. We examined the gene expression of factors that play key roles in the regulation of lipid metabolism (Murray et al., 2012): the transcriptional regulators Pparα, Pgc1α, and Fgf21; the mitochondrial enzymes carnitine palmitoyltransferase 1b (Cpt1b), long-chain fatty acid dehydrogenase (Lcad), as well as the rate-limiting ketogenic enzyme Hmgcs2. We also determined the expression of genes encoding the peroxisomal acyl-CoA oxidase (Acox1) and the organelle's marker, peroxisomal membrane protein 70 (Pmp70). Moreover, we investigated the expression of genes encoding the lipogenic enzymes ATP-citrate lyase (Acl) and fatty acid synthase 2 (Fas2). In this paper, we demonstrate that depending on the dose, fenofibrate differently affects the expression of genes involved in lipid metabolism and that aging affects the response to high-dose fenofibrate treatment.

Animals, treatment, and material sampling
Young (4-month-old) and old (24-month-old) male Wistar-Han rats were bred in the Academic Animal Experimental Center in Gdańsk, Poland, and housed one per cage in standard breeding conditions (22±2°C, humidity 55±10%, 12 h/12 h light/dark cycle). For 30 days, the animals were given 0.5% FN (Glentham Life Sciences, Corsham, UK) mixed into standard rodent chow (high dose; n=10 young and 9 old animals). Control animals (n=10 young and 7 old) were fed the same chow without supplementation (Labofeed H, Wytwornia Pasz Morawski, Kcynia, Poland). The dose was expected to provide approximately 100 mg/day FF according to the measured baseline food intake (20.7±2.43 g, mean ± S.D., for the young and 23.4±4.97 g for the old rats). With respect to the mean body weight at the onset of the treatment (395 g for the young and 552 g for the old rats), the 0.5% FN dose was equivalent to 260 and 210 mg/kg body weight in young and old rats, respectively. Because the treatment with 0.5% FN had some negative effects on the histology of the liver in both young and old rats (Zubrzycki et al., 2020), in a separate experiment we treated rats with 0.1% FN for 30 days (low dose; n=9-10 animals in each group) in the same breeding conditions as for 0.5% FN. At the end of the experiments, the animals were sacrificed under full isoflurane anesthesia through exsanguination from heart puncture. Liver samples were collected and immediately frozen in liquid nitrogen for molecular analyses. The experimental procedures were approved by the Local Ethics Committee in Bydgoszcz, Poland (protocols No. 41/2017, 58/2017, 40/2018, and 5/2019, and carried out accordingly.

Quantitative polymerase chain reaction (qPCR)
The mRNA levels of Acox1, Acl,Cpt1b,Fas2,Fgf21,Hmgcs2,Lcad,Pgc1α,Pmp70,Pparα,Sirt1, and Sirt3 were assessed in liver samples by qPCR method (StepOne Plus apparatus, Life Technologies-Applied Biosystems, Grand Island, NY, USA). Total RNA Mini kit (A&A Biotechnology, Gdynia, Poland) was used for RNA extraction according to the manufacturer's protocol, and the obtained RNA was stored at -80°C. 1 µg RNA was reverse-transcribed (RT) with M-MuLV RT enzyme (Thermo Fisher Scientific, Fitchburg, WI, USA) and oligo(dT) 18 primers (Sigma-Aldrich, Munich, Germany). 5× diluted cDNA was used in qPCR reactions with Sen-siFastSybr™ No-Rox reagent (Bioline, London, UK) and primers at 200 nM final concentration. mRNA levels were quantified relative to the expression of the geometric mean of acidic ribosomal phosphoprotein P0 (36B4) and cyclophilin A (CycloA) levels, and analysed using the ΔΔCt method. The primers were designed using Prim-er3Plus software based on BLAST, Ensembl, and Ace-View databases (Table 1).

Statistical analysis
Statistical analysis was performed using GraphPad Prism (v. 6.0; San Diego, CA, USA). Data were analyzed for outliers using the Grubbs' test. Because not all sets of data followed a normal distribution (Shapiro-Wilk test), statistical analysis was performed using the nonparametric Kruskal-Wallis ANOVA with Dunn's multiple comparisons test. Data are presented as mean ± S.D. and normalized to young control animals (expression level=1). Data obtained for control rats on normal diet in the two separately conducted experiments testing 0.1% and 0.5% FF were combined after checking that there were no significant differences between the control groups. Statistical significance was set at p<0.05.

Sirt1 and Sirt3 expression at the mRNA and protein level
Young and old control rats did not differ with respect to the hepatic mRNA and protein expression levels of Sirt1 and Sirt3 (Fig. 1).
In young rats, FN treatment at either dose did not affect Sirt1 and Sirt3 mRNA levels in the liver (Fig. 1A). Even though Sirt1 expression was notably upregulated in a subset of rats treated with 0.1% FN, no statistical significance was found. In old rats, 0.1% FN did not cause any changes in Sirt1 expression, but 0.5% FN decreased Sirt1 expression 6.25-fold (p<0.01; Fig. 1B). FN did not cause any changes in Sirt3 expression in the liver of old rats (Fig. 1B).
In old rats, 0.1% and 0.5% FN decreased hepatic Sirt1 protein level 3.03-fold and 2.86-fold, respectively (p<0.05 for both; Fig. 1D). However, neither dose of FN affected Sirt3 protein level in the liver of old rats (Fig. 1D). In old as compared to young rats, hepatic Pparα gene expression was 4-fold higher (p<0.001), while Pparα protein levels were comparable in the two groups (Fig. 2).
In both young and old rats, neither dose of FN changed the Pparα protein level in the liver (Fig. 2B).

The effects of fenofibrate on the expression of lipid metabolism-related genes in the liver of young and old rats
In our study, aging was not associated with any changes in the hepatic mRNA levels of Cpt1b, Lcad, Pgc1α, Hmgcs2, Acox1, Pmp70, Acl, Fas2, and Fgf21 (Fig. 3).
In young rats, the hepatic expression of the lipogenic gene Acl was not affected by FN treatment, while the expression of the key lipogenic gene Fas2 decreased 4-fold (p<0.05) after treatment with 0.5% FN, with no effect of 0.1% FN (Fig. 3E).
In the liver of old rats, 0.1% and 0.5% FN increased Cpt1b expression 15-fold and 50-fold, respectively (p<0.01 for both) and increased Lcad expression 6-fold and 2.5-fold (p<0.001 and p<0.05, respectively; Fig. 3B). Similarly as in the liver of young rats, the mRNA levels of Pgc1α were not affected by FN treatment (Fig. 3B).
The expression of the ketogenic gene Hmgcs2 in the liver of old rats increased 3.2-fold (p<0.01; Fig. 3D) upon treatment with 0.1% FN (but not 0.5% FN). In old rats, 0.1% FN did not affect Acox1 mRNA level, while 0.5% FN increased its expression 17-fold (p<0.001; The mRNA and protein expression levels were determined by qPCR and Western blotting, respectively, with representative blots shown. Data are mean ± S.D., presented relative to the level in young control rats (set as 1), n=6-10 animals per group. For clarity, data from young and old rats are separated into two graphs, with the scale maintained. * p<0.05, ** p<0.01, *** p<0.001 (Kruskal-Wallis ANOVA). Abbreviations: YC young (4-month-old) control rats; YF1 young rats treated with 0.1% FN; YF5 young rats treated with 0.5% FN; OC old (24-month-old) control rats; OF1 old rats treated with 0.1% FN; OF5 old rats treated with 0.5% FN Age-related effects of fenofibrate on hepatic Sirt 1 and Sirt3  For clarity, data from young and old rats are separated into two graphs (except for G), with the scale maintained. The mRNA expression levels were determined by qPCR and presented as mean ± S.D. relative to young control rats (set as 1), n=7-10 animals per group. *p<0.05, **p<0.01, ***p<0.001 (Kruskal-Wallis ANOVA) Fig. 3D). Only 0.1% FN highly increased Pmp70 expression (p<0.05; Fig. 3D) in the liver of old rats.
The treatment of old rats with 0.5% FN did not alter hepatic Acl expression. However, in contrast to young animals, in old rats treated with 0.5% FN the hepatic expression of Fas2 increased 6-fold (p<0.01; Fig. 3F).
The expression of Fgf21 was not affected by FN treatment in either young or old rats (Fig. 3G).

DISCUSSION
We present a comprehensive analysis of Sirt1 and Sirt3 expression in the liver, the key organ of lipid metabolism, upon 30-day treatment of young and old rats with two doses of FN, i.e. 0.1% (low dose) and 0.5% (high dose). We demonstrate that these two doses of FN differently affected the expression of genes involved in lipid metabolism, with aging affecting the response to 0.5% FN.
Sirt1 and Sirt3 have both been associated with protection against age-related diseases, including metabolic derangements (reviewed by Zhao et al., 2020). Therefore, we first examined whether aging affects the hepatic expression of these sirtuins. We found no age-related differences in either mRNA or protein levels of Sirt1 and Sirt3. Similarly, no effect of aging on Sirt1 protein level was observed in the liver of 19-26 month-old C57BL/6 mice compared to young, 2-4 month-old animals, even though there were reductions in other tissues (Xu et al., 2020). In contrast, Jin and others (Jin et al., 2011) reported that in the liver of old mice Sirt1 mRNA and protein expression decreased due to repression of the gene promoter by CCAAT/Enhancer Binding Protein/histone deacetylase 1 (C/EBPb-HDAC1) complex. Sirt3 protein level was downregulated in the liver of 20-month-old mice (Kwon et al., 2017) and this reduction was attributed to the decreased level of Sirt1, which deacetylates and stabilizes Sirt3 (Kwon et al., 2017). It is plausible that differences in chow composition and time of probing, together with different genetic backgrounds of rat and mouse strains may account for the discrepancies. Moreover, it is important to consider that aging-related changes differ between organs and tissues. For example, in the same rat strain bred under similar conditions, we have found that the old compared to young rats showed lower Sirt3 expression in skeletal muscle, but maintained high expression in cardiac muscle (Ławniczak et al., 2022).
Considering the involvement of Sirt1 and Sirt3 in the control of lipid homeostasis, we examined whether the hypolipemic drug FN affects their hepatic expression. Since dyslipidemia is increasingly common with advancing age , we chose normal breeding conditions to determine the effect of FN during physiological aging. We found that depending on the dose and the rat's age FN decreased the protein abundance of Sirt1 and Sirt3 in the liver, with little effect on the mRNA levels. It is generally accepted that the protein levels rather than mRNA are more important for the activity of a particular bioactive molecule. Sirt1 activity can be regulated by several mechanisms, including activation through autocatalytic deacetylation (Fang et al., 2017) or cAMP-dependent activating phosphorylation (Gerhart-Hines et al., 2011). With regard to Sirt3, the decrease in protein levels observed in our study upon FN treatment of young rats may be related to the reduction in Sirt1 level, since Sirt1 was shown to deacetylate and stabilize Sirt3 (Kwon et al., 2017).
To our best knowledge, this is the first study examining the expression of Sirt1 and Sirt3 in the liver of young and old FN-treated rats under normal breeding conditions. The FN-induced reduction of Sirt1 abundance is consistent with some other studies conducted in different experimental settings. In young Wistar rats with HFD-induced hyperlipidaemia, the highly elevated hepatic Sirt1 protein and mRNA expression was reduced after treatment with FN for 21 days, although to a level significantly higher than in control animals (Le et al., 2014). In contrast, no effect on hepatic Sirt1 protein level was observed in young mice treated with approx. 400 mg/kg FN for 5 days (Li & Wu, 2018). Reports on hepatic Sirt3 expression upon FN treatment showed either no impact of the drug (approx. 400 mg/kg/day, 5 days) (Li & Wu, 2018) or prevention of iron overload-induced downregulation of Sirt3 (100 mg/kg, i.p., 6 weeks) (Mandala et al., 2021).
We next analyzed the hepatic expression of genes encoding enzymes and transcriptional regulators involved in lipid metabolism, the activity of which is directly or indirectly regulated by SIRT1 and SIRT3. We have previously reported that in our cohort of rats FN exerted a hypolipemic effect, which was more prominent in old than in young animals (Zubrzycki et al., 2020). The major molecular target activated by FN (Gebel et al., 1992), the nuclear receptor PPARα, is regulated by ligand binding (Balanarasimha et al., 2014) and post-translationally by phosphorylation (Nakamura et al., 2019). Similarly to our results, various studies demonstrated FN action without altering PPARα level (e.g. Zhao et al., 2015;Liu et al., 2016). We analyzed several transcriptional targets of PPARα (Rakhshandehroo et al., 2010): Cpt1b, an enzyme responsible for the transport of fatty acids across the mitochondrial outer membrane, Lcad, involved in mitochondrial β-oxidation of fatty acids, and Acox1, the rate-limiting enzyme of peroxisomal β-oxidation. Among the Cpt1 isoforms, the Cpt1b gene is expressed in the liver at lower levels than the Cpt1a isoform, which is known to be upregulated by FN (e.g. de la Rosa Rodriguez et al., 2018). The Cpt1b expression in our study was strongly upregulated by both doses of FN in either age group. Lcad expression was induced by 0.1% FN and to a smaller degree and only in old rats by 0.5% FN. Stimulation of mitochondrial fatty acid oxidation-related genes by FN confirms the results of most other studies (Karahashi et al., 2013;Zhao et al., 2017). While Acox1 expression was induced by both FN doses in young rats, its induction by only the higher FN dose in old animals may reflect the general rule that in aging higher doses of activators are needed for the induction of many target genes (e.g. Kobilo et al., 2014;Zubrzycki et al., 2020). The expression of Pmp70, which has been described as a specific marker for PPARα activation and peroxisome proliferation (Colton et al., 2004), was highly induced by 0.1% FN in both age groups.
SIRT1 has been reported to interact with PPARα at the PPAR response element sequence of target genes (Purushotham et al., 2009). Sirt1 levels in young rats did not change upon 0.1% FN treatment but decreased after 0.5% FN. This may explain the induction of PPARα-regulated target genes for mitochondrial and peroxisomal enzymes primarily after 0.1% FN in young animals. However, old rats showed downregulation of Sirt1 protein levels after both doses, indicating Sirt1-independent modulation of genes for fatty acid oxidation. Indeed, Bonzo and others (Bonzo et al., 2014) showed that hepatic SIRT1 is dispensable for FN-induced PPARα function in vivo. In young mice treated with 0.1% FN for 5 days following the HFD, hepatocyte-specific Sirt1 knockout did not affect the induction of PPARα target genes (Bonzo et al., 2014).
PPARα also controls the expression of Fgf21, a hepatokine that promotes lipid oxidation, lipolysis, and ketogenesis in the liver during ketotic states (Badman et al., 2007) and upon treatment with sodium butyrate (Li et al., 2012). We did not observe any significant changes in Fgf21 mRNA expression upon treatment with FN, in contrast to the induction of both mRNA and protein levels of FGF21 reported in hepatocytes treated in vitro with 500 μmol/L FN (Li et al., 2012). On the other hand, FGF21 has been recently demonstrated to be dispensable for FN action on lipogenesis-or autophagy-related proteins in the murine liver (Jo et al., 2017). Furthermore, the ketogenesis-related gene Hmgcs2, whose transcription is regulated by both PPARα (Meertens et al., 1998;Vilà-Brau et al., 2011) and reciprocally by FGF21 (Meertens et al., 1998;Badman et al., 2007;Vilà-Brau et al., 2011), was induced in our study after treatment with 0.1% FN in both young and old rats. In line with our results, a principal transcriptomics analysis of the livers of the FN-treated hepatocyte humanized mice revealed that among many established PPARα targets FN induced Hmgcs2 (de la Rosa Rodriguez et al., 2018). Likewise, dietary obese mice after a single injection of bezafibrate (100 mg/kg) showed elevated hepatic expression and serum concentration of FGF21, elevated expression of Hmgcs2, but no change in the serum concentration of the ketone body β-hydroxybutyrate (Li et al., 2012). In our experimental setting, it is unlikely that the modestly increased Hmgcs2 expression led to any significant change in ketone body production. SIRT1 activates PGC1α by deacetylation (Rodgers et al., 2005), whereas PGC1α stimulates the expression of SIRT3 (Kong et al., 2010). PGC-1α and PPARα cooperate to control the transcription of genes encoding mitochondrial fatty acid oxidation enzymes (Vega et al., 2000), whereas SIRT3 modulates the activity of these enzymes by deacetylation (e.g. Hirschey et al., 2010;Shimazu et al., 2010). Even though we noted no significant changes in Pgc1α mRNA expression, it should be stressed that this transcriptional coactivator is regulated primarily post-transcriptionally through phosphorylation, acetylation, monomethylation, and O-linked β-Nacetylglucosaminylation (Housley et al., 2009). Similarly to Pgc1α, FN did not affect the mRNA expression of Sirt3 in either age group. However, FN caused downregulation of the Sirt3 protein, which suggests decreased stability or increased degradation of Sirt3 upon treatment. Regrettably, in our study, it was not possible to assess Sirt3 function, e.g. by examining the acetylation status of Sirt3 target proteins. In contrast to our results, a study in Pparα knockout mice treated with 100 mg/kg FN for 6 weeks showed downregulation of hepatic Sirt3 expression, indicating the dependence of Sirt3 expression on PPARα in mouse liver (Mandala et al., 2021). Moreover, the FN treatment prevented iron-induced downregulation of hepatic Sirt3 in wild-type iron-overloaded mice (Mandala et al., 2021). Modulation of SIRT3 expression by FN appears to be variable under different experimental conditions.
We observed contrasting, dose-dependent effects of FN on the hepatic expression of the rate-limiting lipogenic Fas2 gene in young and old rats, with Acl expression not significantly affected. In young rats, the decrease in Fas2 expression upon 0.5% FN treatment (though not 0.1% FN) falls in line with similar results reported in adult rats fed a high-fructose diet and treated subsequently with 100 mg/kg FN for 4 weeks (Abd El-Haleim et al., 2016). Because FAS activity is mainly regulated at the transcriptional level by nutrients and hormones, this decreased expression could contribute to reduced accumulation of lipid droplets, which we observed under a transmission electron microscope in hepatocytes of young 0.1% FN-treated rats (Zubrzycki et al., 2021). In contrast to young rats, in the old ones 0.5% FN increased Fas2 expression. Similarly, a 2-week treatment with 0.2% FN strongly induced hepatic de novo lipogenesis and the expression of Fas2 in mice (age not specified) (Oosterveer et al., 2009). Likewise, an increase in FAS protein expression, concurrently with increased mSREB-P1c level was observed in the liver of 10-12-week-old mice fed 50 mg/kg/day FN for 3 weeks (Jo et al., 2017). Further studies are needed to explain the reason for the apparent FN dose-and age-dependent differences with regard to the modulation of lipogenic genes' expression. Nevertheless, it should be noted that the impact on hepatic lipogenesis does not preclude the FN hypolipemic effect through mechanisms including the stimulation of fatty acid β-oxidation and increased lipoprotein lipase activity.
Available evidence suggests that PPARα acts not only as a regulator of fatty acid catabolism but plays a broader role in hepatic lipid metabolism. By using stable isotope techniques, Oosterveer et al. demonstrated that the treatment of mice for 2 weeks with 0.2% FN induced hepatic de novo lipogenesis and fatty acid elongation in vivo in parallel with an increased expression of lipogenic genes (Oosterveer et al., 2009). This lipogenic induction was dependent on SREBP-1c which is a target for inhibitory deacetylation by SIRT1 (Walker et al., 2010). Thus, in terms of our experiment, FN-associated reduction of Sirt1 expression in the liver of old rats may contribute to the increased Fas2 expression. However, the striking contrast between young and old rats with respect to the effect of FN on lipogenic gene expression warrants further studies. One factor contributing to this outcome could be food intake. We previously reported that 0.5% FN was associated with markedly reduced food intake during the first few days, followed by increased consumption, more rapid in young than old rats (Zubrzycki et al., 2020). A re-feeding effect could contribute to the elevation of lipogenesis in old rats.
Our study provides yet another piece of evidence that at least two concentrations of the studied compound should be applied under both in vivo and in vitro experimental conditions. We found clear discrepancies in the effects of low dose, i.e. 0.1% FN (equivalent to ca. 52 and 42 mg/kg/day in young and old rats, respectively) versus high dose, i.e. 0.5% FN (equivalent to ca. 260 and 210 mg/kg/day) in regard to the expression of Sirt1, Pparα, Lcad, Hmgcs2, Acox1, Pmp70, and Fas2. It should be noted that during long-term treatment in humans, FN is administered once daily typically at a dose of 50-150 mg (preceded by higher doses), whereas the dosage applied to studies in rodents varies widely, from 50 mg/ kg/day (Jo et al., 2017) to 400 mg/kg/day (Li & Wu, 2018). Comparison between FN dosage in humans and rodents is difficult because of species differences in the metabolism and elimination patterns of the drug (Caldwell, 1989).
Our study is not without limitations. Firstly, the results of a study examining PPARα agonist effects in rats cannot be directly transferred to human conditions. Upon administration of peroxisome proliferator chemicals, rodents respond with peroxisomes proliferation, hepatomegaly, and increased hepatocyte proliferation, whereas such a response is typically absent in humans (Klaunig et al., 2003;Peters et al., 2005). Secondly, we measured the mRNA and protein levels of Sirt1 and Sirt3. Considering that activation of sirtuins may be achieved by increasing NAD + availability (Zhao et al., 2020), analysis of the hepatic NAD + /NADH ratio would give more insight into Sirt1 and Sirt3 activity upon treatment with FN. Nevertheless, NAD + -independent mechanisms of sirtuins' activation also exist, including phosphorylation of SIRT1 at Ser434 via the cAMP/PKA pathway (Gerhart-Hines et al., 2011). Thirdly, we measured only the mRNA expression of some genes involved in fuel metabolism, selected as representative for particular metabolic processes and/ or under PPARα transcriptional control. Certainly, activity assays, as well as analyses of protein levels and their acetylation status would be more informative with regard to the changes induced by FN. However, these methods were beyond the scope of this project. Lastly, additional studies are needed to examine the mechanisms responsible for the dose-dependent and age-dependent differences in fenofibrate's action in the liver.

CONCLUSIONS
Treatment with fenofibrate at low or high dose affects the hepatic expression of Sirt1 and Sirt3 proteins. The dosage of fenofibrate significantly affects the treatmentinduced changes in the expression of lipid metabolismrelated genes in the rat liver. Aging influences the response to high-dose fenofibrate. The results of our study emphasize the importance of considering the patient's age before adjusting the fenofibrate dose.

Declarations
Ethical approval: All animal experimental procedures had been authorized by the Local Ethics Committee in Bydgoszcz, Poland (protocols No. 41/2017, 58/2017, 40/2018, and 5/2019, and carried out in compliance with the EU Directive 2010/63/EU for animal experiments.
Conflict of interest: The authors have no relevant financial or non-financial interests to disclose.