Hepatocyte RAP1A Deletion Impairs Lipid Catabolism and Worsens Steatosis via Autophagy Activation

Animals and ethics statement

The loxP-Rap1a liver-specific knockout (Rap1a^flox/flox Alb^cre, LKO) mice were developed using clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 technology on a C57BL/6 genetic background by Cyagen Biosciences Inc. (Suzhou, China). The LKO mice have loxP sites flanking exon 5 and 6 of the Rap1a gene, specifically designed so that when crossed with Alb-cre mice the Rap1a gene is selectively deleted in the liver. The control mice (Rap1b^flox/flox, fl/fl) in these experiments are typically the mice without the Alb-cre transgene, maintaining intact gene function. Then we typically performed polymerase chain reaction (PCR) on DNA extracted from mouse tail snips and the primer sequences used were detailed in Supplementary Table 1. Both LKO and fl/fl mice were fed a high-fat diet (HFD; Research Diets D12492) for 14 weeks to evaluate the metabolic effects of Rap1a deletion under conditions of nutritional excess. The mice were kept in standard cages within a specific pathogen-free facility, maintained under a 12-hour light/dark cycle. The room temperature was controlled at 22°C±2°C variance, and humidity was maintained at 50% with a 5% variance. After euthanizing the mice using 1.25% tribromoethanol (0.02 mL/g body weight, intraperitoneal injection), tissues and blood were harvested. Ethical approval (approval number: xmsq2023-1021) were reviewed and approved by the Experimental Animal Ethics Committee of Wenzhou Medical University.

Cell culture and transfection

The AML12 cell line, acquired from the Cell Bank of the Chinese Academy of Sciences, was cultured at 37°C in an incubator with an atmosphere of 5% CO₂. Cells were cultured in Dulbecco’s Modified Eagle’s Medium/Ham’s F-12 (Thermo Fisher Scientific, Waltham, MA, USA), enriched with 12% bovine calf serum (D6429, Sigma-Aldrich, St. Louis, MO, USA), 5.5 μg/mL transferrin (T8158, Sigma-Aldrich), 10 μg/mL insulin (abs42019847, Absin, Shanghai, China), 40 ng/mL dexamethasone (A601187-0005, Sangon, Shanghai, China), 5 ng/mL sodium selenite (S5261, Sigma-Aldrich), and 1% penicillin-streptomycin-gentamicin solution (abs9245, Absin). For the manipulation of RAP1A expression in AML12 cells, both KD and overexpression plasmids were synthesized by Tsingke Biotechnology (Beijing, China). The KD vector uses the pLKO.1 lentiviral backbone, with a shRNA sequence targeting coding sequence of Rap1a (NM_ 1455541, 5ʹ-CGGGTAGTTGGCAAAGAACAA-3ʹ). The overexpression plasmid was constructed using the plasmid lentiviral expression vector with internal ribosome entry site and puromycin-resistance gene (pLVX-IRES-Puro) vector, into which the full-length cDNA of mouse Rap1a (NM_145541) was cloned downstream of the cytomegalovirus promoter. Non-specific sequences were inserted into the same vector backbone as the control cell.

Cell treatment and fluorescence microscopy

AML12 cells were serum-starved for 6 hours and then exposed to 200 μM palmitic acid (PA; Sigma-Aldrich, P5585) for at least 18 hours to induce lipid-induced stress. Where indicated, cells were further treated with 25 μM chloroquine (CQ; Sigma-Aldrich, C6628) or 15 μM C16-platelet-activating factor (PAF) (C16, MCE, Monmouth Junction, NJ, USA; HY-108635) to assess the effects of autophagy inhibition and extracellular signal-regulated kinase (ERK) activation on lipid metabolism. For fluorescence staining, AML12 cells were plated onto round glass coverslips (Biosharp, Hefei, China; BS-14-RC) and allowed to adhere prior to treatment. Following PA and/or CQ/C16 exposure, coverslips were washed once with phosphate-buffered saline (PBS) and then incubated with 10 μM dichloro-dihydro-fluorescein diacetate (DCFH-DA; ROS Assay Kit, Beyotime, Haimen, China; Cat. No. S0033) at 37°C for 20 minutes to detect reactive oxygen species. After aspirating the probe and washing twice with PBS, mitochondria were labeled by incubating cells with 350 nM MitoTracker Red FM (Thermo Fisher Scientific, Cat. No. M22425) for 30 minutes at 37°C. Cells were washed twice more with PBS and counterstained with 4ʹ,6-diamidino-2-phenylindole (DAPI; Beyotime; Cat. No. P0131). Finally, coverslips were mounted on glass slides and imaged on a Leica STELLARIS DMi8 confocal microscope under identical acquisition settings for all conditions.

RNA extraction and quantitative real-time polymerase chain reaction

Total RNA was extracted from cells and tissues utilizing TRIzol reagent (#15596018, Invitrogen, Waltham, MA, USA). The quality of RNA was confirmed via NanoDrop One/One^C, with the required A260/A280 ratios set between 1.8 and 2.0. Following the reverse transcription of RNA with the HiScript II Q RT SuperMix kit (R222, Vazyme, Nanjing, China), quantitative real-time PCR (qRT-PCR) was conducted employing the ChamQ Universal SYBR Mix kit (Q311, Vazyme), using β-actin as the internal reference. Primer sequences have been included in Supplementary Table 2 to enhance reproducibility.

Western blotting

Proteins from both tissues and cells were extracted using radioimmunoprecipitation assay (RIPA) lysis buffer (MA0151, Meilune, Dalian, China). Nuclear, cytoplasmic, and membrane proteins from AML12 cells were sequentially extracted using Beyotime P0027 and P0033 kits. To extract mitochondrial membrane proteins according to original research publication, a procedure includes differential centrifugation and ensuring the integrity and purity of the mitochondrial fraction [17]. Quantification was performed with the Pierce^TM BCA Protein Assay Kit (23225, Thermo Fisher Scientific). Samples underwent sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDSPAGE) and were then transferred to polyvinylidene fluoride (PVDF) membranes. Membranes were blocked with 5% skim milk at room temperature for 1 hour and subsequently incubated overnight with primary antibodies at 4°C. After tris-buffered saline with tween-20 (TBST) washes, secondary antibodies were applied at room temperature for 2 hours. Detection was carried out using the Clarity Western ECL Substrate (#1705061, Bio-Rad, Hercules, CA, USA). Band densities were analyzed with ImageJ software (National Institutes of Health, Bethesda, MD, USA), and target protein levels were normalized to internal controls, α-tubulin, or α-actin. The antibodies utilized are detailed in Supplementary Table 3.

Glucose tolerance test and insulin tolerance test

Glucose tolerance test (GTT) and insulin tolerance test (ITT) were conducted according to a well-established protocol [18]. For the GTT, mice were fasted for 16 hours and then administered D-glucose (1 g/kg body weight) via intraperitoneal injection. Blood glucose levels were measured from tail vein blood at 0, 15, 30, 60, 90, and 120 minutes using a Yuwell 580 glucometer (Yuwell, Shanghai, China). For the ITT, mice were fasted for 6 hours and then injected intraperitoneally with insulin (1 U/kg body weight). Blood glucose was monitored at the same time points using the same glucometer.

Metabolic caging and body composition analysis

Body composition metrics, such as percentages of body fat and lean mass, were precisely quantified utilizing a Bruker mq20 Toothpaste Analyzer (Bruker, Billerica, MA, USA). Concurrently, food consumption and activity patterns were continuously monitored via an integrated weighing sensor located within the metabolic cages, employing the Oxymax/CLAMS system (Columbus Instruments, Columbus, OH, USA). All measurements were conducted in strict adherence to the protocols provided by the manufacturers.

H&E staining and Oil red O

Mice were euthanized, and their liver tissues were fixed in 4% paraformaldehyde (No. 158127, Sigma-Aldrich) following the established protocol previously described [19]. Subsequently, tissues were embedded in paraffin and optimal cutting temperature compound (4583, SAKURA, Torrance, CA, USA), sectioned, and stained with hematoxylin and eosin (H&E; C0105M, Beyotime). For lipid visualization, sections were dehydrated using 60% isopropanol and stained with Oil Red O (Product No. A600395, Sangon). Both paraffin-embedded liver sections and cryosections were processed identically. All images were acquired with an EVOS M7000 imaging system (Thermo Fisher Scientific).

Biochemical assays

Triglyceride (TG; No. A110), total cholesterol (TC; No. A111), low-density lipoprotein cholesterol (LDL-C; No. A113), high-density lipoprotein cholesterol (HDL-C; No. A112) concentrations, alanine aminotransferase (ALT; C009), and aspartate aminotransferase (AST; C010) activities were determined using commercial assay kits provided from Nanjing Jiancheng Bioengineering Institute (Nanjing, China), according to the manufacturer’s instructions. Liver tissue samples were homogenized in a mechanical homogenizer (KZ-II-F, Servicebio, Wuhan Optics Valley, China) using a homogenization buffer at a 9:1 volume-to-weight ratio. The homogenates were then centrifuged, and the resulting supernatants were collected for biochemical analysis. For cell samples, disruption was achieved using ultrasonic sonication (JY92-IIN, Scientz, Ningbo, China) at 300 W, with each pulse lasting 3 to 5 seconds followed by 30-second intervals. This cycle was repeated 3–5 times to ensure complete lysis.

CPT1 activity assay

The activity of the carnitine palmitoyltransferase 1 (CPT1) enzyme was assessed using a CPT1 Activity Assay Kit (No. ZIKER558) provided by Suzhou Coming Biotechnology Company (Suzhou, China), following the manufacturer’s instructions. AML12 cell lysates were initially mixed with a reaction buffer that contained 1 to 2 mM ethylenediaminetetraacetic acid (EDTA), 200 μM 5,5ʹ-dithiobis-(2-nitrobenzoic acid) (DTNB), and 80 μM palmitoyl-CoA. The mixtures were pre-incubated at 37°C for 5 minutes to equilibrate. The enzymatic reaction was initiated by the addition of L-carnitine, and the absorbance was immediately measured at 412 nm using a photometer. A blank control was included by substituting L-carnitine with sterilized distilled water. CPT1 activity was determined by normalizing the absorbance readings against the protein concentrations in the samples.

Transcriptome sequencing

Total RNA was isolated from liver tissues using TRIzol reagent and subsequently analyzed by Novogene (Beijing, China). To ensure high-quality RNA, the integrity was assessed using the RNA Nano 6000 Assay Kit on a Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA). mRNA was extracted from the total RNA with poly-T oligo-attached magnetic beads and subsequently fragmented. For cDNA synthesis, random hexamer primers and M-MuLV Reverse Transcriptase were utilized for the first strand synthesis, followed by RNA degradation with RNase H. The second strand synthesis involved DNA polymerase I and deoxyribonucleotide triphosphates (dNTPs) to generate blunt ends by converting overhangs, followed by adenylation and adaptor ligation. cDNA fragments within the 370 to 420 bp size were purified using AMPure XP beads (Beckman Coulter Life Sciences, Brea, CA, USA), amplified via PCR, and subjected to an additional purification step. Library quantification was performed using a Qubit 2.0 Fluorometer (Thermo Fisher Scientific), and quality was assessed with an Agilent 2100 Bioanalyzer before sequencing on the Illumina NovaSeq 6000 platform (San Diego, CA USA), yielding 150 bp paired-end reads. The processed clean reads were mapped to the reference genome using HISAT2 software. Enrichment analysis for differentially expressed genes (DEGs) was conducted utilizing the Database for Annotation, Visualization, and Integrated Discovery (DAVID) database and Gene Ontology resources, with Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis performed using the ClusterProfiler R package version 3.8.1 (R Foundation for Statistical Computing, Vienna, Austria).

Statistical analysis

Statistical analyses were conducted using GraphPad Prism version 9 (GraphPad Software Inc., San Diego, CA, USA). Data were expressed as mean±standard error of the mean, as specified in the figure legends. For comparisons between two groups, unpaired two-tailed Student’s t-tests were employed for normally distributed data. Analysis of variance (ANOVA) followed by Tukey’s post hoc test was applied for multiple group comparisons. The number of biological replicates (N) for each experiment is noted in the respective figure legends. Statistical significance was established at P<0.05, with results denoted as follows: P<0.05, P<0.01, P<0.001, and P<0.0001. All experiments were performed in at least three independent replicates unless otherwise indicated.

Hepatic Rap1a expression is downregulated in metabolic disorders, and its deficiency aggravates metabolic disorders

In this study, we investigated the expression level of Rap1a under various human metabolic conditions. Dataset GSE126848 demonstrated decreased Rap1a levels in patients with non-alcoholic fatty liver disease (NAFL) and non-alcoholic steatohepatitis (NASH) compared to healthy, normal-weight controls (Fig. 1A). Similarly, dataset GSE15653 revealed a significant reduction in Rap1a expression among obese individuals, regardless of T2DM status (Fig. 1B). These findings indicate that Rap1a downregulation is associated with a spectrum of metabolic disorders, including obesity, diabetes mellitus, and fatty liver diseases. To investigate the role of hepatic RAP1A in metabolic regulation, we generated a LKO mouse model (Supplementary Fig. 1A). Due to the lack of reliable commercial antibodies for RAP1A detection, we employed an indirect approach: subtracting RAP1B antibody reactivity from total RAP1 antibody measurements for Western blot analysis, complemented by qRT-PCR for validation. Both Western blot (Fig. 1C and D) and qRT-PCR (Fig. 1E) confirmed a significant reduction in hepatic RAP1A protein and Rap1a mRNA levels in LKO mice. There were no significant differences in RAP1 or RAP1B protein levels, nor in Rap1a and Rap1b mRNA levels in the brain, heart, or adipose tissues of LKO mice compared with fl/fl mice (Supplementary Fig. 1B-D). From the 7th week onward, male mice were subjected to a HFD. By the 9th week, LKO mice exhibited significantly higher body weight compared to fl/fl mice (Fig. 1F). Body composition analysis showed a marked increase in fat mass without significant changes in lean mass (Fig. 1G). Organ and tissue mass measurements revealed increased weights of the liver and various adipose depots, including subcutaneous white adipose tissue (sWAT), gonadal white adipose tissue, and brown adipose tissue (Fig. 1H). Food intake (Fig. 1I) and activity levels (Fig. 1J) were comparable between the two groups. Additionally, LKO mice displayed impaired glucose tolerance and insulin sensitivity, as evidenced by worsened GTT and ITT (Fig. 1K and L). Circulating levels of TG, TC, HDL-C, and LDL-C were all significantly increased in LKO mice compared to fl/fl controls (Fig. 1M-P), indicating that hepatocyte-specific Rap1a deficiency leads to systemic lipid metabolic dysregulation. In addition, serum ALT and AST activities were markedly elevated in LKO mice (Fig. 1Q and R), indicating hepatocellular injury. These results underscore the detrimental metabolic consequences of hepatic Rap1a deficiency, highlighting its critical role in maintaining metabolic homeostasis across a range of metabolic disorders.

Fig. 1.

Hepatic Ras-related protein 1a (Rap1a) expression is downregulated in metabolic disorders, and its deficiency aggravates metabolic disorders. (A) Rap1a mRNA expression in liver biopsies from normal-weight individuals (normal, n=14), obese individuals (obese, n=12), patients with simple steatosis (non-alcoholic fatty liver disease [NAFL], n=15), and those with non-alcoholic steatohepatitis (NASH, n=16), based on GSE126848. (B) Comparison of Rap1a mRNA levels in liver samples from lean controls undergoing cholecystectomy (n=5) and obese individuals undergoing gastric bypass surgery, with or without type 2 diabetes mellitus (T2DM), from GSE15653. (C, D) Western blot analysis (C) and quantification (D) of RAP1 and RAP1B protein expression in total liver lysates from male fl/fl and liver-specific RAP1A knockout (LKO) mice (n=6 per group). (E) Quantitative real-time polymerase chain reaction analysis of Rap1a and Rap1b expression in the livers of male fl/fl and LKO mice (n=6 per group). (F) Body weight of male fl/fl and LKO mice monitored from 4 weeks of age, with high-fat diet (HFD) feeding from 7 to 21 weeks (n=6 per group). (G) Body compositions analysis of HFD-fed fl/fl and LKO mice (n=6 per group). (H) Organ weights of the liver, muscle, heart, brain, brown adipose tissue (BAT), subcutaneous white adipose tissue (sWAT), gonadal white adipose tissue (gWAT), and retroperitoneal white adipose tissue (rWAT) (n=6 per group). (I) Food consumption per 7 days (n=6 per group). (J) Locomotor activity of fl/fl and LKO male mice (n=4 per group). (K, L) Intraperitoneal glucose tolerance test (K) and insulin tolerance test (L) in HFD-fed fl/fl and LKO mice (left: glucose-time curves; right: area under the curve [AUC]) (n=6 per group). Serum levels of (M) triglyceride (TG), (N) total cholesterol (TC), (O) high-density lipoprotein cholesterol (HDL-C), and (P) low-density lipoprotein cholesterol (LDL-C) in HFD-fed fl/fl and LKO mice. Serum (Q) alanine aminotransferase (ALT) and (R) aspartate aminotransferase (AST) activities in HFD-fed fl/fl and LKO mice. NS, not significant. ^aP<0.05, ^bP<0.01, ^cP<0.001, ^dP<0.0001.

RAP1A deficiency enhances hepatic lipid accumulation in both in vivo and in vitro models

Given the liver’s central role in metabolism, we aimed to determine whether systemic phenotypic dysregulation is directly linked to hepatic function by examining liver-related phenotypes. Histological and biochemical assays were performed to evaluate hepatic steatosis. As shown in H&E and Oil Red O staining, liver sections from LKO mice revealed a marked increase in lipid vacuole accumulation and lipid droplet content compared to fl/fl controls (Fig. 2A). Non-alcoholic fatty liver disease (NAFLD) activity score (NAS) indicated significantly elevated steatosis and hepatocellular ballooning in the LKO group (Fig. 2B), consistent with enhanced hepatic lipid accumulation. Quantification of Oil Red O staining further confirmed increased lipid content in LKO livers (Fig. 2C). Moreover, qRT-PCR analysis showed significantly increased expression of key genes related to hepatic fibrosis (transforming growth factor beta 1 [Tgfb1], collagen type I alpha 1 chain [Col1a1], collagen type III alpha 1 chain [Col3a1], tissue inhibitor of metalloproteinases 1 [Timp1], calponin 2 [Cnn2], platelet-derived growth factor receptor beta [Pdgfrb]) and inflammation (interleukin-1 [Il1], tumor necrosis factor alpha [Tnfα], chemokine [C-C motif] ligand 2 [Ccl2]) in LKO mice (Fig. 2D and E), providing additional evidence for the protective role of RAP1A against metabolic dysfunction-associated steatotic liver disease (MASLD) progression. Biochemical analyses revealed significantly elevated hepatic TG and TC levels in LKO mice (Fig. 2F and G). To further elucidate the role of RAP1A, we developed AML12 cell models with RAP1A KD or overexpression, alongside corresponding controls. In RAP1A KD cells, RAP1 protein expression was significantly reduced, with no changes in RAP1B levels (Fig. 2H and I). This reduction in Rap1a mRNA was confirmed by qRT-PCR (Fig. 2J). Consistent with findings in LKO mice, TG levels increased significantly in RAP1A KD cells after treatment of PA (Fig. 2K). Conversely, overexpression of RAP1A was validated by Western blot (Fig. 2L and M) and qRT-PCR (Fig. 2N), and resulted in significantly reduced TG levels compared to control cells (Fig. 2O). Collectively, these findings highlight the critical role of RAP1A in regulating hepatic lipid metabolism. RAP1A deficiency exacerbates lipid accumulation and liver pathology under HFD conditions, underscoring its pivotal function in the pathogenesis of fatty liver disease.

Fig. 2.

Ras-related protein 1a (RAP1A) deficiency enhances hepatic lipid accumulation in both in vivo and in vitro models. (A) Representative H&E and Oil Red O staining of liver sections from high-fat diet (HFD)-fed male fl/fl and liver-specific RAP1A knockout (LKO) mice. Scale bars are shown. (B) Non-alcoholic fatty liver disease (NAFLD) activity scores (NAS), including steatosis, lobular inflammation, ballooning, and total scores. (C) Quantification of Oil Red O staining. (D, E) Quantitative real-time polymerase chain reaction (qRT-PCR) analysis of (D) fibrosis-related and (E) inflammation-related gene expression in liver tissues. Hepatic (F) triglyceride (TG) and (G) total cholesterol (TC) levels in HFD-fed fl/fl and LKO mice (n=6 per group). (H) Western blot and (I) quantification of RAP1 and RAP1B protein levels in indicated alpha mouse liver 12 (AML12) cells. (J) qRT-PCR analysis of Rap1a and Rap1b expression in indicated AML12 cells. (K) TG levels in indicated AML12 cells treated with palmitic acid. (L) Western blot and (M) quantification of RAP1 and RAP1B protein expression in indicated AML12 cells. (N) qRT-PCR analysis of Rap1a and Rap1b expression in indicated AML12 cells. (O) TG levels in indicated AML12 cells after palmitic acid treatment. Tgfb1, transforming growth factor beta 1; Col1a1, collagen type I alpha 1 chain; Col3a1, collagen type III alpha 1 chain; Timp1, tissue inhibitor of metalloproteinases 1; Cnn2, calponin 2; Pdgfrb, platelet-derived growth factor receptor beta; Il1, interleukin-1; Tnfα, tumor necrosis factor alpha; Ccl2, chemokine (C-C motif) ligand 2; Ifnγ, interferon-γ; NS, not significant; pLVX, plasmid lentiviral expression vector. ^aP<0.05, ^bP<0.01, ^cP<0.001, ^dP<0.0001.

Deletion of Rap1a disrupts hepatic lipid catabolism

To investigate how RAP1A modulates lipid metabolism and its potential contribution to fatty liver disease progression, we analyzed gene and protein expression in liver tissues from LKO mice and floxed control (fl/fl) mice. qRT-PCR results revealed increased expression of lipogenic genes, including sterol regulatory element-binding protein 1c (Srebp1c), stearoyl-CoA desaturase-1 (Scd1), and elongation of very long chain fatty acids protein 6 (Elovl6) in LKO mice (Fig. 3A). In contrast, genes critical for lipid catabolism and fatty acid oxidation, such as hormone-sensitive lipase (Hsl), adipose triglyceride lipase (Atgl), peroxisome proliferator-activated receptor α (Pparα), carnitine palmitoyltransferase 1A (Cpt1α), and carnitine palmitoyltransferase 2 (Cpt2), were significantly downregulated in LKO mice (Fig. 3B). Western blot analysis confirmed these findings, showing elevated levels of lipogenic proteins (ATP citrate lyase [ACLY], acetyl-CoA carboxylase [ACC], fatty acid synthase [FASN]) and reduced levels of lipolytic enzymes (HSL and CPT1A) in LKO mice (Fig. 3C and D). We then examined lipid metabolism-related gene expression in AML12 cells following PA treatment. In RAP1A KD cells, Fasn expression was upregulated (Fig. 3E), while genes involved in lipid degradation, such as Atgl, Pparα, and Cpt1a, were downregulated (Fig. 3F). Western blot analysis confirmed these changes (Fig. 3G and H), and we observed reduced CPT1 enzyme activity (Fig. 3I). In RAP1A overexpression cells, lipogenic gene expression remained unchanged (Fig. 3J), but the expression of lipid catabolism genes (Hsl, Pparα, Cpt1a, and Cpt2) was significantly elevated (Fig. 3K). Western blot analysis showed upregulation of ATGL, HSL, and CPT1A (Fig. 3L and M), along with increased CPT1 enzyme activity (Fig. 3N). These findings underscore the crucial role of RAP1A in regulating hepatic lipid metabolism. Specifically, RAP1A KD reduces lipid degradation, leading to increased lipid accumulation and metabolic dysfunction. Conversely, overexpression of RAP1A enhances lipid breakdown, thereby mitigating lipid buildup.

Fig. 3.

Deletion of Ras-related protein 1a (Rap1a) disrupts hepatic lipid catabolism. (A, B) Quantitative real-time polymerase chain reaction (qRT-PCR) analysis of lipogenic (A) and lipolytic (B) gene expression in livers of male fl/fl and liver-specific RAP1A knockout (LKO) mice (n=4 per group). (C, D) Western blot analysis (C) and quantification (D) of lipid metabolism-related proteins in liver lysates from male fl/fl and LKO mice (n=6 per group). (E, F) Quantitative real-time polymerase chain reaction (qRTPCR) analysis of lipogenic (E) and lipolytic (F) genes in control and RAP1A knockdown alpha mouse liver 12 (AML12) cells following palmitic acid treatment. (G, H) Western blot analysis (G) and quantification (H) of lipolysis-related proteins in RAP1A knockdown AML12 cells and controls. (I) Carnitine palmitoyltransferase 1A (CPT1A) enzymatic activity in control and RAP1A knockdown AML12 cells, normalized to protein concentration. (J, K) qRT-PCR analysis of lipogenic (J) and lipolytic (K) genes in control and RAP1A overexpressing AML12 cells following palmitic acid treatment. (L, M) Western blot analysis (L) and quantification (M) of lipolysis-related proteins in RAP1A overexpressing AML12 cells and controls. (N) CPT1A enzymatic activity in control and RAP1A overexpression AML12 cells, normalized to protein concentration. Srebp1c, sterol regulatory element-binding protein 1c; Scd1, stearoyl-CoA desaturase-1; Fasn, fatty acid synthase; Elovl6, elongation of very long chain fatty acids protein 6; Acaca, acetyl-CoA carboxylase alpha; Hsl, hormone-sensitive lipase; Atgl, adipose triglyceride lipase; Ppara, peroxisome proliferator-activated receptor α; Cpt1a, carnitine palmitoyltransferase 1A; Cpt2, Carnitine palmitoyltransferase 2; NS, not significant; ACLY, ATP citrate lyase; p-ACC, phosphorylated acetyl-CoA carboxylase; t-ACC, total acetyl-CoA carboxylase; pLVX, plasmid lentiviral expression vector. ^aP<0.05, ^bP<0.01, ^cP<0.001, ^dP<0.0001.

RAP1A localizes to the mitochondrial membrane and activates autophagy upon knockdown

Given our previous observation that RAP1A KD led to a reduction in CPT1 expression—a key regulator of mitochondrial fatty acid β-oxidation—we hypothesized that RAP1A may influence hepatic lipid metabolism by modulating mitochondrial function. To further explore this possibility, we examined the subcellular localization of RAP1A. Western blot analysis of AML12 cell extracts revealed that RAP1 was detectable in whole-cell lysates and significantly enriched in the cytoplasmic fraction, similar to the cytoplasmic marker glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Fig. 4A). Further analysis of membrane proteins, including mitochondrial membranes, showed RAP1 localization on membranes, comparable with mitochondrial markers TOM70 and voltage dependent anion channel (VDAC) (Fig. 4B). In AML12 cells, RAP1 was enriched in mitochondrial extracts, while the nuclear marker Lamin A/C was restricted to whole-cell lysates (Fig. 4C). Treatment of mitochondrial fractions with Na₂CO₃, which disrupts membrane integrity, revealed RAP1 enrichment in the pellet alongside the mitochondrial membrane marker TOM40, indicating that RAP1 is likely embedded in the mitochondrial membrane (Fig. 4D).

Fig. 4.

Ras-related protein 1a (RAP1A) localizes to the mitochondrial membrane and activates autophagy upon knockdown. (A-D) Subcellular fractionation of alpha mouse liver 12 (AML12) cells. Whole-cell lysates (WCL), nuclear proteins (Nucl), and cytoplasmic (Cyto) fractions were analyzed by Western blot (A). WCL, membrane proteins (Mem), and Cyto fractions were assessed (B). WCL and mitochondrial (Mito) fractions were isolated for immunoblotting (C). Carbonate extraction of mitochondrial fractions following Na₂CO₃ treatment; untreated mitochondria (M), pellet (P), and supernatant (S) fractions were analyzed (D). Specific markers were used to confirm compartmental enrichment: Lamin A/C (nucleus), glyceraldehyde-3-phosphate dehydrogenase (GAPDH; cytoplasm), sodium, potassium-adenosine triphosphatase (Na,K-ATPase; membrane), translocase of the outer mitochondrial membrane 70 and 40 (TOM70/TOM40)/voltage dependent anion channel (VDAC; mitochondrial outer membrane), and glucose-regulated protein 75 (GRP75; mitochondrial matrix). (E–G) Gene Set Enrichment Analysis (GSEA) showing enrichment of pathways related to apoptotic mitochondrial changes (E), regulation of mitochondrial autophagy (F), and positive regulation of autophagosome assembly (G). Enrichment was evaluated using nominal P values, false discovery rate (FDR), enrichment score (ES), and normalized enrichment score (NES). (H, I) Western blot analysis (H) and quantification (I) of autophagy-related protein in liver lysates from male fl/fl and liver-specific RAP1A knockout (LKO) mice (n=6 per group). (J, K) Western blot analysis (J) and quantification (K) of autophagy-related protein in control and RAP1A knockdown AML12 cells following palmitic acid treatment. (L) Left: confocal imaging of dichloro-dihydro-fluorescein diacetate (DCFH-DA; green), MitoTracker (red), and 4ʹ,6-diamidino-2-phenylindole (DAPI; blue) in control and RAP1A knockdown AML12 cells after palmitic acid treatment. Right: quantification of mean fluorescence intensity of DCFH-DA. (M, N) Western blot analysis (M) and quantification (N) of autophagy-related protein in control and RAP1A overexpressing AML12 cells following palmitic acid treatment. (O) Left: confocal imaging of DCFH-DA (green), MitoTracker (red), and DAPI (blue) in control and RAP1A overexpressing AML12 cells after palmitic acid treatment. Right: quantification of mean fluorescence intensity of DCFH-DA. PINK1, PTEN induced kinase 1; PARK2, parkinson protein 2; LC3, light chain 3; NS, not significant; pLVX, plasmid lentiviral expression vector. ^aP<0.05, ^bP<0.01, ^cP<0.001.

To further investigate the role of RAP1A, we performed RNA-seq analysis on liver tissues from LKO and fl/fl mice. Gene Set Enrichment Analysis (GSEA) revealed significant enrichment of gene sets related to mitochondrial regulation (Fig. 4E), including fission (Supplementary Fig. 2A) and fusion (Supplementary Fig. 2B) processes, in the LKO group. We assessed key mitochondrial dynamics genes and proteins, such as dynamin related protein 1 (DRP1), mitofusin 1 (MFN1), MFN2, and Optic atrophy 1 (OPA1), and found no significant alterations across liver tissues (Supplementary Fig. 2C and D), RAP1A KD cells (Supplementary Fig. 2E-G), and RAP1A overexpression cells (Supplementary Fig. 2H-J). In contrast, genes associated with mitochondrial autophagy pathways were upregulated in LKO mice (Fig. 4F and G). Enhanced autophagic activity in LKO liver samples was evidenced by increased levels of PTEN induced kinase 1 (PINK1) and light chain 3II (LC3-II), and decreased P62 (Fig. 4H and I). A similar pattern was observed in RAP1A KD cells treated with PA, with elevated LC3-II and reduced P62 (Fig. 4J and K). Consistently, DCFDA staining followed by confocal microscopy revealed a marked increase in ROS accumulation in RAP1A KD cells (Fig. 4L), indicating enhanced mitochondrial oxidative stress associated with RAP1A deficiency. In contrast, RAP1A overexpression did not produce significant changes in LC3-II or P62 (Fig. 4M and N), and ROS levels were also reduced compared to control cells (Fig. 4O), further supporting a protective role for RAP1A in maintaining mitochondrial homeostasis. In conclusion, these findings demonstrate that RAP1A localizes to the mitochondrial membrane and highlight RAP1A’s critical role in regulating mitochondrial autophagy under conditions of altered fatty acid metabolism.

Mitigation of TG accumulation in RAP1A knockdown cells via phosphorylated ERK activation and autophagy inhibition

Building on previous research demonstrating that RAP1 directly regulates ERK activity through phosphorylation, thereby affecting cell proliferation and migration [20], we hypothesized that RAP1A might modulate lipid accumulation via the ERK signaling pathway. Notably, studies have shown that autophagy deficiency in mice suppresses mitogen-activated protein kinase (MAPK)/ERK activation [21]. Consistent with this, we observed reduced ERK phosphorylation in the liver of LKO mice (Fig. 5A and B). Western blot analyses of RAP1A KD cells following PA treatment revealed decreased ERK phosphorylation in RAP1A KD cells (Fig. 5C and D). Upon treatment with the ERK activator C16, Western blot analysis showed that the reduced phosphorylated ERK (p-ERK) expression in RAP1A KD cells was restored (Fig. 5E and F). Notably, the differences in p62 and LC3-II between RAP1A KD and control cells persisted in the presence of C16 (Fig. 5G and H). This restoration was accompanied by a recovery of lipid degradation proteins ATGL, HSL, and CPT1A to levels observed in control cells (Fig. 5I-K). Additionally, TG levels, which were previously elevated in RAP1A KD cells, decreased to levels comparable to those in control cells (Fig. 5L). These findings suggest that RAP1A modulates lipid accumulation via the ERK signaling pathway. Activation of p-ERK reduces elevated TG levels and restores lipid degradation protein expression in RAP1A KD cells without affecting autophagy.

Fig. 5.

Mitigation of triglyceride (TG) accumulation in Ras-related protein 1a (RAP1A) knockdown cells via phosphorylated extracellular signal-regulated kinase (p-ERK) activation and autophagy inhibition. (A, B) Western blot analysis (A) and quantification (B) of p-ERK and total ERK (t-ERK) protein expression in liver lysates from male fl/fl and liver-specific RAP1A knockout (LKO) mice (n=6 per group). (C, D) Western blot analysis (C) and quantification (D) of p-ERK and t-ERK protein expression in control and RAP1A knockdown alpha mouse liver 12 (AML12) cells following palmitic acid treatment. (E, F) Western blot analysis (E) of lipolysis-related proteins and p-ERK/t-ERK levels in control and RAP1A knockdown AML12 cells treated with 15 μM C16 for 18 hours following palmitic acid (PA) stimulation. Quantification (F) of p-ERK/t-ERK signal ratios (n=3 independent experiments). (G–K) Quantification of Western blot signal intensities for P62 (G), light chain 3II (LC3-II) (H), adipose triglyceride lipase (ATGL) (I), hormone-sensitive lipase (HSL) (J), and carnitine palmitoyltransferase 1A (CPT1A)/β-ACTIN (K) signal ratios were presented (n=3 independent experiments). (L) TG levels in control and RAP1A knockdown AML12 cells with or without 15 μM C16 treatment for 18 hours. (M–P) Western blot analysis (M) and quantification of P62 (N), LC3-II/β-ACTIN (O), and p-ERK/t-ERK (P) in control and RAP1A knockdown AML12 cells treated with 25 μM chloroquine for 18 hours following PA stimulation (n=3 independent experiments). (Q–S) Quantification of Western blot signal intensities for ATGL (Q), HSL (R), and CPT1A/β-ACTIN (S) under the same treatment conditions (n=3 independent experiments). (T) TG levels in control and RAP1A knockdown AML12 cells following PA treatment with or without 25 μM chloroquine for 18 hours. NS, not significant. ^aP<0.05, ^bP<0.01.

Subsequently, we treated AML12 cells with CQ after PA exposure to inhibit autophagy. Western blots showed increased p62 and LC3-II in both RAP1A KD and control cells, indicating successful inhibition of autophagic flux by CQ (Fig. 5M-O). Notably, treatment with CQ alone led to a reduction in p-ERK levels, suggesting that autophagy inhibition may suppress ERK signaling activity independently of RAP1A status (Fig. 5P). Additionally, Western blot analysis revealed that in RAP1A KD cells, the previously observed reductions in ATGL, HSL, and CPT1A were no longer significantly different from control cells after CQ treatment (Fig. 5Q-S). Moreover, the difference in TG accumulation seen in untreated cells was eliminated after CQ treatment (Fig. 5T). Furthermore, combined treatment with CQ and C16 demonstrated that C16 was unable to rescue CQ-mediated suppression of p-ERK (Supplementary Fig. 3A-D), while co-treatment partially restored ATGL and CPT1A levels but failed to recover HSL expression (Supplementary Fig. 3E-G), and still led to a significant reduction in TG accumulation in RAP1A KD cells (Supplementary Fig. 3H). In an additional experiment using AML12 cells cultured without PA (Supplementary Fig. 3I-L), CQ administered at the same dose and duration as in the PA assays still decreased p-ERK, indicating that CQ suppresses ERK phosphorylation as a general phenomenon. Together, these results suggest that lipid accumulation in RAP1A KD cells is mediated, at least in part, by autophagy-dependent mechanisms, and that ERK activation alone cannot fully reverse the metabolic effects of autophagy inhibition. This underscores the critical interplay between autophagy and ERK signaling in regulating hepatic lipid homeostasis downstream of RAP1A.