An HFD (cat. no. D12492) and standard chow diet (cat. no. XTI01SL-002) were procured from Research Diets, Inc. and Jiangsu Xietong Pharmaceutical Bio-Engineering Co., Ltd., respectively. Detailed nutritional composition of the diets is provided in Data S1. Serum lipid profiles, including total cholesterol (TC; cat. no. H202) and triglycerides (TG; cat. no. H201), were analyzed using commercial assay kits from Ningbo Meikang Biotechnology Co., Ltd. Hepatic function biomarkers, alanine aminotransferase (ALT) and aspartate aminotransferase (AST), were quantified with kits from the same supplier. Fetal bovine serum was obtained from PAN-Seratech GmbH. Ready-to-use penicillin-streptomycin and the BCA kit were purchased from Beyotime Biotechnology. AK (cat. no. A113942) was purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. D-(+)-Glucose was purchased from Shanghai HuSHI Co., Ltd. Insulin was purchased from Shanghai Fosun Pharmaceutical (Group) Co., Ltd. A sterile BSA-coupled free fatty acid (FFA) mixture solution was obtained from Shanghai Siduorui Biotechnology Service Co., Ltd. The PPARα agonist fenofibrate was acquired commercially from MedChemExpress. The PPARα agonist Wy-14,643 was purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. The PPARα antagonist GW6,471 was acquired commercially from MedChemExpress. Antibodies against β-Tubulin (cat. no. 80713-1-RR), TAS1R2 (cat. no. 29344-1-AP) and PLCβ (cat. no. 66668-1-Ig) were purchased from Wuhan Proteintech Group, Inc. Antibodies against G protein subunit alpha transducin 3 (GNAT3; cat. no. A15982), p-PLCβ (cat. no. AP0837) and TAS1R3 (cat. no. A10157) were purchased from Wuhan ABclonal Biotech Co., Ltd. The antibody against diacylglycerol kinase eta (DGKH; cat. no. AF0611) was purchased from Affinity Biosciences. Detailed information regarding the product brands, dilution ratios, and catalog numbers for all antibodies can be found in Table SI.

A concentration of 2 mg/ml AK in drinking water was used, this was determined by converting the human ADI (15 mg/kg/day) to a mouse equivalent dose (27). Using the formula HED=human dose × [Km human/Km mouse] (Km human=37, Km mouse=3), the mouse ADI was calculated as 185 mg/kg/day. Based on the measured water intake of the mice on HFD (100-140 ml/kg/day), a concentration range of 1.32-1.85 mg/ml was calculated. Finally, 2 mg/ml was used to ensure effective exposure while keeping close to human ADI.

Wild-type (WT) and PPARα-null (KO) mice (129/Sv strain) were obtained from Dr Frank J. Gonzalez at the National Cancer Institute, Bethesda, MD, USA (28), and subsequently bred in-house at Ningbo University. A total of 88 healthy 6-week male mice were used for the 3 experiments mentioned below. The mean body weights were 18.0±0.5 g (WT) and 17.8±0.6 g (KO) before the experiment. The mice were housed in individual ventilated cages under controlled humidity (60±5%) and temperature (24±1°C) and a 12/12-h dark/light cycle. The mice were monitored daily for general health (activity, posture, fur condition, food/water intake) by trained staff. On the terminal procedure day, monitoring was continuous from carbon dioxide inhalation until confirmation of death. The animal procedures were approved by the Animal Ethics and Welfare Committee of Ningbo University (approval no. AEWC-NBU20230140; Ningbo, China).

In the first experiment, 10 WT mice were randomly divided into WT group (n=5) and WT-AK group (n=5). Another 10 KO mice were divided into the KO group (n=5) and the KO-AK group (n=5). The sample size design followed previous study and the 3R framework (29). A randomization table was utilized to assign experimental groups, with all procedures executed by an independent technician blinded to the animal handling. All the mice were fed a 60% HFD, and those in the WT-AK and KO-AK groups were treated with an additional 2 mg/ml AK in drinking water. Food and water were replaced every 48-72 h to keep them fresh. Food and water intake were recorded to track metabolic changes. After 11 weeks of treatment, oral glucose tolerance test (OGTT) and insulin tolerance test (ITT) were performed with a 3-day interval.

In a second experiment, 25 WT and 25 KO mice were divided into the control (CON) group, F5, F25, F125 and Wy (n=5), where they were treated with 5, 25 or 125 mg/kg fenofibrate (F5, F25or F125, respectively) twice a day and with 160 mg/kg Wy-14,643 once a day by oral gavage for 5 days to explore the direct regulatory effects of PPARα on the STR signaling pathway. All the drugs were suspended in 1% (w/v) sodium carboxymethylcellulose solution before administration. The CON groups were administered the same volume of vehicle during this period.

In a third experiment, 9 WT and 9 KO mice were allocated to 6 groups: WT, WT-AK, WT-AK-F, KO, KO-AK, and KO-AK-F (n=3). All the mice were fed a 60% HFD, and those in the WT-AK, WT-AK-F, KO-AK and KO-AK-F groups were treated with an additional 2 mg/ml AK in drinking water. After 10 weeks of treatment, the WT-AK-F and KO-AK-F groups were treated with 25 mg/kg fenofibrate twice daily via oral gavage for 2 weeks to investigate whether PPARα activation attenuates hepatic steatosis by inhibiting the STR signaling. The fenofibrate was dissolved in 1% (w/v) sodium carboxymethylcellulose solution before administration. The remaining groups were administered the vehicle during this period. The dosage and regimen of fenofibrate were selected based on a previously study that could effectively inhibit NAFLD development (30).

All mice were sacrificed at the scheduled experimental endpoint. The mice were individually placed in a transparent induction chamber prefilled with ambient air. Medical-grade carbon dioxide (CO₂) was introduced at a controlled flow rate (20-30% of the chamber volume per minute), resulting in a gradual increase in CO₂ concentration and loss of consciousness. While the animal remained unconscious, terminal blood collection was performed as part of the procedure. CO₂ exposure was then increased to a higher concentration (~70%) and maintained until complete respiratory arrest, which was sustained for an additional 5 min. Mortality was verified by the absence of respiration and cardiac activity and subsequently confirmed by cervical dislocation. No unscheduled mortality occurred prior to the endpoint. After cervical dislocation, liver tissues were collected. The blood samples were centrifuged at 800 × g for 15 min at 4°C. The serum was cryopreserved at −20°C for analysis within 3 days. After being scaled, half of the biggest lobe was fixed in 10% neutral formalin at room temperature for 24 h, and the remaining tissues were stored at −80°C.

After the 11-weeks of treatment, all the mice (WT, WT-AK, KO, and KO-AK groups) in experiment 1 were fasted for 12 h and gavaged with a glucose solution (2.5 g per kg body weight). Blood glucose levels were measured at 0, 30, 60, 90, and 120 min using a glucometer (ONETOUCH Verio; LifeScan, Inc.) following tail-tip blood sampling. Blood (5-10 μl) was collected at each predefined time point using a sterile lancet. During the fasting period, the mice were allowed to drink control water to eliminate potential AK-induced disturbances in insulin and gastrointestinal hormone secretion (31). The area under the curve (AUC) was calculated to assess glucose tolerance.

For ITT, the WT and WT-AK mice in experiment 1 were fasted for 6 h, followed by intraperitoneal injection of insulin (0.8 U/kg body weight). After insulin injection, the blood glucose levels were measured at 0, 15, 30, 45 and 60 min using a glucometer. Blood sampling and glucose measurement were performed using the same method and sampling volume as those used in the OGTT. Blood glucose concentrations over time were plotted as a glucose-time curve and the AUC was calculated to compare the insulin sensitivity.

Hepa1-6 cells (Suzhou Hysigen Bioscience Co., Ltd.) and Huh-7 cells (Wuhan Pricella Biotechnology Co., Ltd.) were cultured in high-glucose DMEM (4.5 g/l; Corning Life Sciences) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. These two cell lines were treated with 0.2 mM FFA for 48 h to induce lipid accumulation. Simultaneously, AK (0, 1, 3, 10 and 20 mM) was added to the aforementioned growth media to explore the effects of AK on lipid accumulation. 50% AK was added and the incubation continued for 48 h. The TG content of the cells was measured following the procedures described below.

A second-generation lentiviral packaging system was used in this experiment. Lentiviral particles were packaged in 293T cells (ATCC) and produced in 293T cells (ATCC) by co-transfection with 8 μg shRNA plasmid, 6 μg packaging plasmid (Δ8.2) and 4 μg envelope plasmid (VSV-G). Lentiviral supernatants were collected at 48 h post-transfection, filtered through a 0.45-μm filter, and used to infect Hepa1-6 cells at a multiplicity of infection of 10. Hepa1-6 cells were passaged into 10-cm dishes 24 h before infection. A total of 10 ml of DMEM containing lentivirus packaged with short hairpin (sh) NC and shPparα (Wuhan Miaoling Biotechnology Co., Ltd.) plasmids was added to DMEM at a 1:1 ratio during the medium change. At 24 h after infection, the culture medium containing the lentivirus was discarded and replaced with fresh medium. Finally, the infected cells were selected using DMEM supplemented with 30 μg/ml blasticidin (Beyotime Biotech Inc.) for 48 h and maintained in medium containing 3 μg/ml blasticidin to purify stable transfectants. Successful gene knockdown was confirmed by reverse transcription-quantitative (RT-q) PCR and western blot analysis. Hepa1-6 cells induced by FFA were co-treated with 2 μM GW6,471 or AK (0, 1, 3, 10 and 20 mM) for 48 h.

The shNC plasmid was obtained from Wuhan Miaoling Biotechnology Co., Ltd. (cat. no. P47365). The shPparα sequences were as follows: shPparα-1: 5'-AAGAATTCTTACAAGAAAT-3'; shPparα-2: 5'-GGAAAGTCCCTTATCTGAAT-3'; shPparα-3: 5'-GAACATCGAGTGTCGAATAT-3'.

The TC and TG concentrations in the plasma and liver tissues were determined using commercially available TC and TG assay kits. The liver tissues were weighed and homogenized with 9 volumes of saline. For analysis, 2 μl aliquots of plasma or liver homogenate samples were combined with 67 μl of reagent R1. After 5 min of incubation at 37°C, the absorbance was measured at the appropriate wavelengths (TC; 546 and 660 nm; TG; 546 and 700 nm). After 33 μl of reagent R2 was added, the samples were incubated for another 5 min. The absorbance was measured at the aforementioned wavelengths following the manufacturer's protocol.

Concentrations of glycogen in liver tissues were determined using a commercially available glycogen assay kit (cat. no. A043-1-1; Nanjing Jiancheng Bioengineering Institute). The weighed liver tissue was boiled in a metal bath at 100°C after adding 3-fold volumes of the alkaline solution. Glycogen chromogenic solution and double-distilled water were added. The mixture was boiled again and then transferred to a 96-well plate for absorbance measurement at 620 nm.

The hepatic DAG levels were quantified using a commercial ELISA kit (cat. no. KX30762; Shanghai Yaxin Biotechnology Co. Ltd.). The liver tissues were homogenized in ice-cold saline at 4°C. The homogenates were centrifuged at 1,200 × g for 15 min at 4°C. Following the procedures in the kit, the homogenate DAG concentrations were measured and normalized.

Fresh liver tissues were fixed in 10% neutral formalin at room temperature for 24 h. The tissues were then dehydrated in a serial concentration of ethanol (70, 80, 90 and 100%). After clearing with xylene, the tissues were embedded in paraffin and serial 4 μm sections were prepared. To detect polysaccharide content, the liver sections were deparaffinized and stained at room temperature by oxidation with 0.5% periodic acid for 8 min, followed by Schiff reagent incubation for 15 min and H&E counterstaining for 2 min. Representative images were captured using a light microscope (Olympus BX43; Olympus Corporation). Pathological analysis was carried out by a certified pathologist unaware of the study design.

Hepa1-6 and Huh-7 cells were washed twice in PBS and then fixed in 10% neutral formalin buffer for 30 min at 4°C. The cells were stained with fresh oil red O working solution for 30 min, soaked in 60% isopropyl alcohol for 10 sec and finally stained with hematoxylin for 2 min at room temperature. Lipid droplet formation in cells was visualized via inverted phase-contrast microscopy (Olympus CKX41; Olympus Corporation).

Frozen liver and adipose tissues were lysed in TRIzol^® (Invitrogen; Thermo Fisher Scientific, Inc.) and homogenized using a MagNA Lyser (Roche Diagnostics). The RNA concentration was quantified spectrophotometrically, with purity assessed using the OD₂₆₀/OD₂₈₀ absorbance ratio. Total RNA (1 μg) from the liver sample was subjected to reverse transcription in a reaction volume of 20 μl following the manufacturer's protocol of ABScript III RT Master Mix for qPCR (cat. no. RK20428; ABclonal Biotech Co., Ltd.). The synthesized cDNA was stored at 20°C until analysis. The primer sequences are listed in Table SII. Q-PCR amplification on 96-well plates was performed in a 10-μl incubation containing 1 μl cDNA, 5 μl UltraSYBR Mixture, 0.4 μl primer, and 3.6 μl purified water using a LightCycler 480II (Roche Diagnostics). The thermocycling conditions were set as initial denaturation at 95°C for 5 min, followed by 40 cycles of 95°C for 10 sec and 60°C for 30 sec. The 2^-ΔΔCq method was used to quantify the expression of target genes (32). The measured mRNA abundance was normalized to that of GAPDH and expressed as a fold change equivalent to that of WT control mice. All experiments were independently repeated at least three times.

Frozen liver tissues were homogenized in 9 volumes of RIPA lysis buffer (Suzhou NCM Biotech Co., Ltd.) containing 1% PMSF and 1% protein phosphatase inhibitor). The supernatants were collected by centrifugation at 15,000 × g for 15 min at 4°C. Protein concentrations were determined using a BCA protein assay kit (cat. no. B610409; Suzhou NCM Biotech Co., Ltd.). Equal amounts of protein (30 μg per lane) were mixed with 5X SDS-PAGE loading buffer (Beyotime Biotech Inc.) and boiled for 8 min. The 10% SDS polyacrylamide gels were separated for 1 h. The proteins were transferred to 0.45-μM PVDF membranes for 1.5 h in an ice bath. All membranes were blocked with 5% skimmed milk at room temperature for 3.5 h. Subsequently, the membranes were incubated with primary antibodies at 4°C overnight. After three 10-min washes with TBS-0.1% Tween-20, the membranes were incubated with HRP-conjugated secondary antibodies for 2 h at room temperature (primers in Table SI). Finally, the ECL substrate was added to the blotted PVDF, and images were recorded using a chemiluminescent imaging system, ChemiScope 6100 Touch (Clinx Science Instruments Co., Ltd.). Densitometric analysis was performed using ImageJ (version 1.53; National Institutes of Health).

The liver samples from 4 groups (WT-CON, WT-Wy, KO-CON and KO-Wy; n=3/group) in cohort 2 were analyzed. Total RNA was extracted using Trizol reagent (cat. no. B610409; Sangon Biotech Co., Ltd.), treated with DNase I, and assessed for integrity (RIN >8) and purity using a NanoPhotometer spectrophotometer (Implen GmbH) and a Qubit fluorometer (Thermo Fisher Scientific, Inc.). The sequencing libraries were prepared using the VAHTS mRNA-seq V2 Kit (cat. no. NR612-01; Vazyme Biotech Co., Ltd.) for Illumina, including poly-A mRNA enrichment, fragmentation (150-200 bp), RT, adapter ligation, and PCR amplification. Paired-end RNA sequencing (150 bp reads) was performed on an Illumina NovaSeq platform (Illumina, Inc.). The final libraries were quantified using a Qubit fluorometer and Agilent Bioanalyzer and loaded onto the sequencer at a final concentration of 10 pM. The raw reads were filtered using Trimmomatic (version 0.39; https://github.com/usadellab/Trimmomatic) with Q≥20, aligned to the mouse reference genome (GRCm38) using HISAT2 (version 2.2.1; https://daehwankimlab.github.io/hisat2), and quantified as transcripts per million (TPM) using StringTie (version 2.1.4; https://ccb.jhu.edu/software/stringtie). Differentially expressed genes [|FC|≥1.5, False Discovery Rate (FDR) <0.01; DESeq2], including Acot1, Ehhadh, and Dgkh, were validated by qPCR (primers in Table SII).

All experimental subjects were retained for statistical evaluation. Parametric distribution assumptions were confirmed, with results reported as mean ± SEM. Between-group differences were assessed using two-tailed Student's t tests following verification of variance homogeneity. Multigroup comparisons employed one-way ANOVA with Bonferroni adjustment for homogeneous datasets, whereas heteroscedastic data were analyzed through Welch's ANOVA followed by Tamhane's T2 post hoc testing. Statistical significance threshold was set at α=0.05. A post hoc statistical power analysis was performed for the comparisons of liver TG content in experiment 3. Based on the experimentally observed effect sizes, the actual sample size (n=3 per group), and a two-tailed significance level (α) of 0.05, the statistical power was calculated for the comparison between WT-AK and WT-AK-F group. The liver TG content in the WT-AK-F group was significantly lower than that in the WT-AK group [(Mean difference)=3.504-5.532=-2.028 μg/mg; P<0.05]; two-tailed unpaired t-test. The magnitude of this difference was quantified as Cohen's d=8.50, 95% CI [3.0, 14.0]. Analytical procedures were executed in GraphPad Prism 9.5.1 (Dotmatics). P<0.05 was considered to indicate a statistically significant difference.

After 12 weeks of HFD treatment, the WT mice exhibited markedly greater body weight gain compared with the KO mice (Fig. 1A). AK supplementation failed to induce significant changes in body weight gain in either mouse line (Fig. 1A). The WT and the KO mice that consumed AK both presented a reduction in food intake but an increase in water consumption (Fig. 1B-E). The liver weight was markedly higher in the KO mice compared with the WT mice, but AK supplementation did not cause an increase in liver weight in either group (Fig. 1F). These data suggested that AK administration did not affect metabolic phenotype in either the WT mice or the KO mice.

A 12-week HFD regimen induced differential hepatic lipid accumulation between two mouse lines, as indicated by H&E-stained samples in Fig. S1A-D. AK exacerbated hepatic steatosis in the WT mice but not in the KO mice (Fig. 2A). The plasma ALT and AST levels were not altered by AK in either the WT or KO mice (Fig. S2A and B). No significant change in mRNA expression levels of inflammatory genes was visible upon AK treatment in the WT mice, whereas a significant downregulation was observed in the KO mice (Fig. S2C-F). Moreover, AK supplementation increased the hepatic TG levels by approximately 10% in the WT mice, whereas no significant change was observed in the KO mice (Fig. 2B). By contrast, the hepatic TC levels were significantly higher in the KO mice than in the WT mice. However, AK supplementation did not significantly affect hepatic TC levels in either mouse line (Fig. 2C). No significant differences were observed in the serum TC or TG concentrations among the 4 groups (Fig. 2D and E). Thus, chronic AK consumption promotes NAFLD development in the WT mice but not in the KO mice.

After 11 weeks of HFD consumption, an oral glucose tolerance test was performed followed by insulin sensitivity experiments (Fig. 3A). An ~25% higher mean blood glucose peak level was observed in the WT-AK mice as compared with the WT mice (Fig. 3B). AK supplementation impaired glucose tolerance in the WT mice but not in the KO mice (Fig. 3B and C). After the 6-h fasting period, the fasting blood glucose levels of the mice in the KO group were ~25% lower compared with those in the WT group. No significant difference in fasting blood glucose levels was observed between the WT and WT-AK mice (Fig. 3D). AK supplementation did not modify the insulin sensitivity in either the WT mice or the KO mice (Fig. 3E and F). Therefore, AK supplementation may impair the glucose tolerance without altering insulin sensitivity in the WT mice.

Expression of the Gck and G6pc1 genes was markedly upregulated in the WT-AK mice compared with the WT mice (Fig. 3G and H). Gck mRNA expression was increased by ~50%, paralleled by a 1-fold increase in G6pc1 mRNA expression compared with WT mice (Fig. 3G and H). Similarly, the mRNA expression levels of the glucose transport genes Slc2a2 and Slc5a10 were markedly lower in the WT-AK mice compared with the WT mice (Fig. 3I and J). The effect on differential gene expression regulation is lost in mice that are depleted of PPARα. Furthermore, in the WT-AK mice, mRNA levels of the fatty acid synthesis genes Acaca and Elovl6 were higher compared with those in the WT mice, whereas no significant differences were found in the expression of lipid and glycogen metabolism genes (Fig. S3A-H).

Periodic acid-Schiff staining of liver sections revealed more pronounced glycogen deposition in the WT-AK mice compared with in the WT mice (Fig. 3K). The liver glycogen content was elevated by 25% in the WT-AK mice compared with the WT mice, whereas the effect was lost in the KO mice (Fig. 3L). Compared with that in WT-AK mice, the expression of phosphorylated PLCβ in WT-AK mice was elevated, suggesting the activation of STR signaling (Fig. 3M). Although the downregulation of TAS1R2 and TAS1R3 expression was also observed in the WT-AK group, they were supposed to be adaptive modifications. However, none of the aforementioned modifications were observed between the KO and KO-AK mice (Fig. 3M). Therefore, physiologically, PPARα mediates the disruption of glycolipid metabolism by AK in the liver by activating STR signaling, especially by increasing the phosphorylated PLCβ.

Oil Red O staining revealed a dose-dependent increase in lipid accumulation in both Hepa1-6 and Huh-7 cells treated with increasing AK concentrations (Fig. 4A). Supporting results were obtained by direct measurement of the TG contents in both cell lines (Fig. 4B and C). In the Hepa1-6 cell line, the addition of 10 mM AK led to a concentration-dependent increase in PLCβ phosphorylation during treatment (Fig. 4D). A comparable activation of PLCβ was also observed when the same concentrations of AK were added to Huh-7 cells (Fig. 4E).

To functionally prove the involvement of PPARα in the observed AK effects, Hepa1-6 cells were transfected with three shPparα plasmids via lentiviral infection (Fig. 5A). PPARα protein expression was downregulated in shPparα-2-transfected Hepa1-6 cells, as confirmed by western blot analysis (Fig. 5B). No increase in PLCβ phosphorylation was observed in shPparα-transfected Hepa1-6 cells despite exposure to increased AK concentrations (Fig. 5C). Oil Red O staining and intracellular TG analysis revealed that lipid accumulation in shPparα-2-transfected cells was not increased by serially increased concentrations of AK (Fig. 5D and E). Similarly to the knock-down experiment, both the PLCβ activation and the lipid accumulation in AK treatment were also abolished by GW6,471, a selective PPARα antagonist (Fig. 5F and G). Accordingly, lipid accumulation in hepatocytes and the activation of STR signaling component PLCβ is dependent on PPARα.

Next, the effects of PPARα on the STR signaling pathway were explored in mice treated with increasing doses of fenofibrate and Wy-14,643, a more selective PPARα agonist. The mRNA expression levels of the PPARα target genes Acot1 and Ehhadh increased progressively with increasing doses of the PPARα agonist in the WT mice. Such a dose-dependent effect was not detected in the KO mice (Fig. 6A and C). The two mouse lines presented distinct responses to PPARα activation, where the TAS1R2 level increased and the TAS1R3 level progressively decreased (Fig. 6B). GNAT3, the α subunit of the taste receptor G protein involved in intracellular taste signaling, was progressively downregulated upon PPARα activation (Fig. 6B). The activation of GNAT3 facilitated PLCβ-mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate, while PLCβ expression was downregulated upon PPARα activation (Fig. 6B). However, the mRNA transcriptions of the aforementioned components were not modified (Fig. S4A and B). In contrast to the results from the WT mice, the aforementioned modifications of the STR signaling components were absent in the KO mice (Fig. 6D).

Dgkh, a key regulator of DAG catabolism, was identified as the only upregulated gene among the 10 DAG kinase family members analyzed, within the subset of 1,731 genes markedly upregulated by PPARα agonist treatment (Fig. 6E and F). The mRNA expression level of Dgkh exhibited a dose-dependent increase in the WT mice following treatment with Wy14,643 (Fig. 6G). Similarly, the DGKH level was upregulated by the activation of PPARα, which did not occur in KO mice (Fig. 6H and I). A markedly increased expression level of hepatic DAG was observed in the KO mice compared with the WT mice (Fig. 6J). Therefore, PPARα regulates more targets of the STR signaling besides the PLCβ, including TAS1R2, TAS1R3, and DGKH.

The therapeutic action of pharmacologically activated PPARα in AK-induced NAFLD was designed as shown in the flow chart (Fig. 7A). Consistent with the aforementioned observations, hepatic steatosis was markedly exacerbated in the WT-AK mice compared with the control mice. AK-induced lipid accumulation was effectively rescued in the WT-AK-F mice through pharmacological challenge with fenofibrate (Fig. 7B). In the KO mice, neither the pro-steatotic effect of AK nor the therapeutic response to fenofibrate was observed, suggesting both a physiological and pharmacological role of PPARα (Fig. 7B). Hepatic TG content was elevated by 10% in the WT-AK mice compared with the control group. Fenofibrate challenge led to a sharp reduction in this lipid marker compared with that in WT-AK and this was only observed in the WT mice (Fig. 7C). In the KO mice, no increase in TG by AK or decrease in TG by fenofibrate was observed (Fig. 7C).

In western blot analysis, a significant elevation of phosphorylated PLCβ expression was observed in the WT-AK mice compared with the control group (Fig. 7D). In the WT-AK-F group, both the total PLCβ and the phosphorylated PLCβ were markedly reduced (Fig. 7D). Notably, neither expression levels of phosphorylated nor total PLCβ was modified in the KO mice (Fig. 7D). These data suggested the steatotic effect of PLCβ activation mediated by physiological PPARα was inhibited by pharmacologically activated PPARα.