LARS promotes hepatocellular carcinoma progression via the PI3K/AKT/mTOR pathway and interaction with RPS5, and serves as a prognostic biomarker

Abstract

Background

Hepatocellular carcinoma (HCC) caused many cancer deaths around the world. Its progression involves complex mechanisms, creating an urgent need to identify new therapeutic targets. Leucine-tRNA synthetase (LARS) is a key enzyme for protein synthesis, but its specific role and mechanism in HCC are not well understood.

Purpose

This research aims to investigate the biological function, molecular mechanism, and clinical relevance of the LARS gene in HCC progression, to assess its potential as a treatment target.

Methods

LARS expression was assessed in HCC cell lines (PLC-PRF-5, HCC-LM3) and in mouse subcutaneous tumor models using siRNA and adeno-associated virus (AAV). Techniques including Cell Counting Kit-8(CCK-8), colony formation, EdU, Transwell, wound healing, and flow cytometry were used to measure cell proliferation, migration, invasion, and apoptosis. RNA-seq, proteomics (TMT), western blot, co-immunoprecipitation (Co-IP) with mass spectrometry, molecular docking, and molecular dynamics simulation were employed to study the affected signaling pathway (PI3K/AKT/mTOR) and interacting protein (RPS5). The TCGA (The Cancer Genome Atlas) database and UALCAN platform were used to analyze links between LARS expression and clinicopathological features or prognosis in HCC patients.

Results

Reducing LARS expression significantly inhibited the proliferation, colony formation, migration, and invasion of HCC cells, while promoting apoptosis. In mice, LARS knockdown markedly slowed tumor growth. Mechanistic studies showed that reducing LARS expression levels affected the PI3K/AKT/mTOR signaling pathway and led to decreased levels of the key interacting protein RPS5. Overexpressing RPS5 partly reversed the proliferation inhibition caused by LARS depletion. Molecular docking and dynamics simulations suggested that the environmental contaminant triphenyl phosphate (TPP) might bind to the LARS protein. Clinical data analysis revealed that LARS expression is higher in HCC tissues. High LARS expression was significantly associated with shorter overall survival (OS) in patients and correlated positively with various clinical features like tumor stage, grade, and TP53 mutation status.

Conclusion

LARS helped HCC become worse by affecting the PI3K/AKT/mTOR pathway and working with RPS5. High LARS meant a worse outcome for patients. This suggested LARS could be used to predict disease or as a treatment target in HCC.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s10020-026-01490-9.

Introduction and background

Liver cancer remains a significant global health challenge. Hepatocellular carcinoma(HCC) is the main type of liver cancer seen most often (Vogel et al. 2022, Rich 2024). Its high mortality rate is largely attributed to frequent tumor recurrence and metastasis (Foglia et al. 2023, Niu et al. 2022). Therefore, identifying key molecular drivers of HCC progression is crucial for developing new therapeutic strategies (Wang and Deng 2023, Cao et al. 2024). Recent research has focused on understanding the roles of specific genes and pathways in promoting the malignant behaviors of HCC cells, such as uncontrolled proliferation, resistance to cell death, and increased invasive potential (Wang et al. 2024, Matsumoto et al. 2023). Among various cellular processes, protein synthesis and its regulatory mechanisms are emerging as important areas of study in cancer biology. However, the precise functions of many aminoacyl tRNA synthetases, core components of the translation machinery, in HCC pathogenesis are not fully understood, leaving a significant knowledge gap (Tijaro-Bulla et al. 2023).

Aminoacyl tRNA synthetases are indispensable for protein synthesis, yet accumulating evidence indicates that many synthetases exert noncanonical functions beyond translation, thereby reshaping cell signaling networks and cancer progression (Tennakoon and Cui 2024, Yu et al. 2021, Sung et al. 2022). Among them, leucyl-tRNA synthetase (LARS) is increasingly recognized not only as a charging enzyme but also as a cytosolic leucine sensor that directly potentiates mTORC1 activation, thereby linking amino acid availability to growth-promoting signaling (Kim et al. 2021). Through this sensing role, LARS participates in amino acid-dependent growth control and may serve as an upstream node integrating metabolic cues with oncogenic pathways, including the PI3K/AKT/mTOR axis that is frequently hyperactivated in hepatocellular carcinoma (HCC) and supports tumor growth and survival (Tian et al. 2023, Fan et al. 2024). Importantly, recent HCC-focused studies have already suggested a relevance of LARS-associated signaling to HCC biology (Fan et al. 2024), implying that LARS may contribute to tumor progression through nutrient-sensing mechanisms. However, despite these observations, the impact of LARS on HCC has not been fully established and the molecular mechanisms by which LARS may downstream malignant behaviors in HCC remains insufficiently defined. Thus, it is crucial to clarify whether LARS functions as an actionable upstream regulator of amino acid-driven growth signaling in HCC and to define its therapeutic and prognostic implications.

This study aimed to elucidate the biological function of LARS in hepatocellular carcinoma. The main idea was that LARS helps HCC grow by making cells multiply, move, and spread, while stopping cell death. This might happen because LARS works with the protein RPS5 and changes the PI3K/AKT/mTOR pathway. To check this idea, the research tried to answer: Would lowering LARS stop HCC growth and spread in cells and animals? What are the downstream signaling consequences of LARS depletion? Does LARS physically interact with RPS5, and is this interaction functionally relevant?

Clarifying the role of LARS in HCC could provide new theoretical insights into the link between protein synthesis machinery and cancer progression. In practical use, LARS could be a new sign to predict patient outcome or a possible treatment target for HCC. To reach these goals, the research first looked at what happened when LARS was reduced. To address these objectives, this study evaluated the effects of LARS knockdown on the proliferation, colony formation ability, migration, invasion and apoptosis of liver cancer cells through a series of functional tests. A subcutaneous xenograft model in nude mice was then established to confirm these findings in vivo. Subsequently, transcriptomic and proteomic analyses were employed to identify affected pathways, followed by validation of key signaling molecules. The interaction between LARS and RPS5 was explored through co-immunoprecipitation, mass spectrometry, and molecular docking. Last, patient information from open databases was studied to see if LARS levels linked to disease traits. Together, the findings showed that LARS helps HCC move forward.

Methods

Tissue sample processing

Tissues were fixed in 10% neutral buffered formalin, dehydrated through a graded ethanol and xylene series, embedded in paraffin, and then sectioned for hematoxylin and eosin (H&E) and immunohistochemical staining.

Tissue sample evaluation

For Ki67 assessment, the entire slide was thoroughly examined. Ki67 expression was located in the nucleus. The area with the highest expression was identified, and the percentage of positive tumor cells within that area was calculated.

Cell culture methods

HCC-LM3, Huh-7, and PLC-PRF-5 cells, which are human liver cancer lines, and HEK-293T cells, a human kidney line, came from Saibaikang in Shanghai. PLC-PRF-5 cells grew in MEM medium. The other cells used DMEM medium. The normal medium had 10% fetal bovine serum and 1% penicillin-streptomycin added. All cells stayed at 37℃ with 5% CO₂ in a moist air (Preksha et al. 2021).

Cell passaging

Reagents and supplies were sterilized under UV light for at least 30 min. Trypsin solution was pre-warmed to 37℃ in a water bath. When adherent cells reached over 80% confluency, the old medium was removed. An appropriate amount of PBS buffer was gently added along the side of the dish to wash the cell layer, then aspirated. Warm trypsin-EDTA solution was put on the cells to cover them. The dish then went into the incubator for digestion for 2 to 5 min. Digestion time varied by cell type and was monitored under a microscope every minute. Once cells became rounded and partially detached, digestion was stopped by adding double the volume of complete medium. Cells were mixed again softly using a pipette. This cell mix was gathered and spun at 800 to 1000 rpm for 3 to 5 min. The liquid above was thrown away. The cell clump was mixed into 2 to 3 mL of new full medium. Cells were put into new dishes at a right splitting number. More full medium was added. Dishes were marked and put back in the incubator. After finishing, the safety cabinet was cleaned (Moran et al. 2024).

Medium change

Cell status, medium color, and the number of floating cells were regularly monitored. If the medium turned significantly yellow before cells reached full confluency, or if excessive floating dead cells were observed, the medium was changed. Preparation and sterilization steps were similar to those for passaging. The old medium was aspirated, and PBS was slowly added along the dish wall to gently rinse the cells 2–3 times, avoiding direct disturbance of the cell monolayer. After removing the PBS, fresh complete medium was added. Cell shape was looked at under an upside-down microscope. The dish then went back to the incubator. After that, the workspace was cleaned.

Cell freezing

Cells growing fast with good shape were picked for freezing. After trypsin treatment, the cell mix was gathered and spun at 800 to 1200 rpm for 5 min. The top liquid was removed. The cell clump was mixed into a right amount of freezing medium without serum. This mix was moved into small freezing tubes. The tubes were marked clearly. These tubes went straight into a −80℃ freezer. To keep cells for a long time, tubes were moved to liquid nitrogen after one night at −80℃.

Cell thawing

A water bath was warmed up to 37℃ first. A spin tube with 3 to 5 mL of full medium was made ready and marked. The cryovial was quickly removed from storage and placed in the water bath, gently agitated until completely thawed, usually within one minute. Once the freezing medium turned from opaque to clear, the vial was immediately removed and wiped with alcohol. Inside the biosafety cabinet, the cell suspension was transferred into the prepared centrifuge tube containing complete medium. The tube was spun at 800 rpm for 5 min. The liquid on top was thrown away. The cell clump was mixed again into new full medium. Cells were then moved to a culture dish. The dish was marked and put in the incubator.

Cell counting

The cells were treated with trypsin and spun down at 800 to 1000 rpm for five minutes. After that, they were put into 1 mL of full medium or PBS. The liquid with cells was thinned to the right level. A pipette was used to put 10 µL of this thin liquid slowly into one part of a counting slide. Care was taken to stop bubbles or spilling. The hemocytometer was placed under an inverted microscope. Cells in four big squares at the corners were counted. The rule was to count cells that touched the top line or the left line. Cells that touched the bottom line or the right line were not counted. The cell concentration was calculated using the formula: Cells per mL = (Total cell count in four squares/4) × Dilution factor × 10⁴.

RNA extraction

First, cells were washed softly with PBS two or three times (Decruyenaere et al. 2023). Then, 500–1000 µl of RZ lysis buffer was added to the culture plate or dish. The solution was pipetted up and down 10–20 times until it became clear. After standing at room temperature for 5 min, the lysate was transferred to a nuclease-free EP tube and centrifuged at 12,000 rpm and 4℃ for 5 min. The supernatant was moved to a new nuclease-free EP tube. Next, 200 µl of chloroform was added. The tube was capped tightly and shaken vigorously for about 20 s. It was then left at room temperature for 5 min and centrifuged at 12,000 rpm and 4℃ for 10 min. Three layers formed after centrifugation. The upper aqueous layer was transferred to a new nuclease-free EP tube. The volume was estimated, and 0.5 volumes of absolute ethanol were added and mixed well. The mixture was transferred to a spin column and centrifuged at 12,000 rpm for 30 s. The flow-through was discarded. Then, 500 µl of protein removal buffer containing ethanol was added to the column. It was centrifuged again at 12,000 rpm for 30 s, and the waste liquid was discarded. Next, a wash buffer was added to the column, which stood at room temperature for 2 min before centrifugation at 12,000 rpm for 30 s. The waste liquid was discarded. This washing step was repeated once more. After another centrifugation, the spin column was placed into a new 1.5 ml nuclease-free EP tube. RNase-free ddH2O, preheated to 70℃, was added to the center of the column membrane. After standing at room temperature for 3 min, the tube was centrifuged at 12,000 rpm for 2 min. The eluate was reapplied to the membrane and centrifuged again. The spin column was then discarded. The remaining liquid contained the purified RNA. RNA concentration and purity were measured using a spectrophotometer. Purity was assessed by the OD260/280 ratio. Samples with a ratio between 1.8 and 2.0 were considered acceptable. RNA samples were stored at −80℃ for later use.

siRNA transfection method

siRNA transfection was performed as follows

The siRNAs made to target the LARS gene were bought from Shanghai Sangon Biotech. Table 1 shows the sequences. A siRNA stock solution was prepared. Before opening, the tube with siRNA powder was spun at 12,000 rpm for two minutes at 4℃. Then, the tube was opened slowly inside a safe cabinet. As the maker said, the siRNA was mixed to make a 20 µM concentration. The solution was aliquoted and stored at − 80℃ to avoid repeated freeze-thaw cycles (Bartman et al. 2021).

Table 1.

siRNA sequences targeting the LARS gene

Name	Sequence (5'→3')
Si-LARS#1	GCUGUGCUUAUGGAGAAUAUATT
Si-LARS#2	CCAGGGCUUUACCAAAGACAATT
Si-LARS#3	CCUCACUUUGACCCAAGCUAUTT

The transfection procedure included the following steps

To enhance transfection efficiency, Lipofectamine RNAiMAX was used. Notably, the reagent itself may mildly suppress cell metabolism and proliferation, particularly in PLC-PRF-5 cells. On day two, transfection was carried out after the cells were fully attached. A suitable amount of Opti-MEM medium was warmed to 37℃ in a water bath. For each well, 4 µL of Lipofectamine RNAiMAX was mixed into 200 µL of Opti-MEM. In a different tube, 10 µL of siRNA (20 µM) was mixed into another 200 µL of Opti-MEM. For the control group, a different siRNA that does nothing was used. The two liquids were softly put together and left at room temperature for five minutes. At the same time, the old medium in the wells was changed to 1.8 mL of new full medium per well. Then, the thin Lipofectamine RNAiMAX liquid was dripped slowly into the thin siRNA liquid. The tube was closed and left at room temperature for 15 min. At last, the final mix was dripped evenly into the right wells. The plate was turned softly to mix. On day three, the medium was changed 24 h after transfection. Subsequent experiments could then be performed.

Cell proliferation assay

A CCK8 test was done to see how cell growth and living changed after transfection. Cells that grew well were taken for the test. To get better transfection, cells were usually taken about 24 h after siRNA transfection. The cells were broken up with trypsin containing EDTA, just like usual splitting. After centrifugation, the cell resuspended in 1 ml of full medium and mixed well. A 10 µl pipette tip moved the cell liquid into a counting slide. Cells were counted under an inverted microscope at 50× magnification to calculate cell density. The cell liquid was then thinned with full medium and put into a 96 well plate. Each well received 100 µl of the suspension containing 3000 to 5000 cells. To maintain humidity, any empty wells around the samples were filled with PBS. The plate was labeled and placed in a constant temperature incubator overnight to allow cell attachment. On the following day, cell status was examined to ensure full attachment and healthy growth. Under light protected conditions, complete medium containing 10% CCK8 reagent was prepared and mixed well. The supernatant was removed from the test wells, and the wells were gently rinsed once or twice with PBS. Then, 100 µl of the ready CCK8 medium was put into each well. The plate was kept in the dark for three hours. The optical density value of each well was measured at 450 nm wavelength using a multifunctional microplate reader. After testing, the plate was washed two or three times with PBS, and 100 µl of full medium was put into each well. The same steps were done again the next day. Readings were taken every day for three days in a row. The test was done three separate times, and a line graph of cell growth was made from the numbers.

Scratch assay method

A scratch test was done to look at changes in how cells move after transfection. Cells in good growth condition, typically collected 24 h post-si transfection for higher transfection efficiency, were digested with trypsin containing EDTA following the standard subculture protocol. After spinning, the cells were put back into 1 ml of full medium and mixed well. A small 10 µl bit of the cell liquid was put into a counting slide, and cells were counted under an upside down microscope at 50 times bigger to know cell density. The suspension was diluted with complete medium, and approximately 2 × 10⁵ cells in total were seeded into the two chambers of a 6-well plate pre-inserted with a culture insert (ibidi), 70 µl per chamber. The plate was labeled and incubated overnight in a constant temperature incubator to allow cell attachment. The next day, the insert was taken out softly with tweezers, and the wells were washed with PBS before full medium was put in. Cell migration was immediately observed and photographed under a 50× inverted microscope, showing a clear gap of uniform width. The same locations were re-examined and photographed at 24 and 48 h. A scale was used to calculate the daily migration distance. The experiment was repeated three times.

Transwell assay

To see how transfection changed cell moving ability, migration and invasion tests were done. First, Transwell slides and pipette tips were kept at −20℃ for more than three hours, then put on ice. Matrigel was mixed with medium without serum at 1:8 on ice. Fifty microliters of this prepared Matrigel-containing medium was quickly and evenly dripped onto the upper chamber. The chambers were placed in a 37℃ incubator for three to four hours until the gel solidified. Unsolidified liquid was then carefully removed from the upper chamber using a pipette tip. Cells in good growth condition, typically harvested around 24 h after siRNA transfection to ensure higher transfection efficiency, were digested using trypsin containing EDTA following standard subculture procedures. After spinning down, the cell lump was put into one milliliter of medium with 2% FBS. The suspension was mixed thoroughly, and cells were counted using a hemocytometer under an inverted microscope at 50x magnification to calculate cell density. The cell mix was thinned with medium that had 2% FBS. Then 150 µl of this mix, holding 30,000 cells, went into the top part of the Transwell chamber. Chambers with Matrigel were used for invasion tests. Plain chambers were used for migration tests. The bottom part of a 24-well plate got 600 µl of full medium. The chamber containing the cell suspension was then carefully lowered into the well at an angle to avoid bubble formation and incubated overnight. Medium in the chambers was replaced daily. After 24 h for migration or 48 h for invasion assays, the chambers were removed. The medium was poured off. Chambers were rinsed two times with PBS. They were then placed in 4% paraformaldehyde for 20 to 30 min for fixation. After that, the paraformaldehyde was removed and chambers were rinsed two or three times with PBS. Next, they were put in a 0.1% crystal violet solution for 20 to 30 min. The stain was checked at times. Right after staining, chambers were taken out and rinsed two or three times with PBS. Cells that did not move or invade on the top side were softly wiped off with a cotton swab. Chambers were turned upside down and left to dry in the air. Cells that moved or invaded were looked at and pictures were taken using an inverted microscope. Every test was done three times. Cell numbers were counted and studied with ImageJ software.

Colony formation assay

To evaluate the effect of transfection on cell proliferation and colony-forming ability, cells in good growth condition, typically collected around 24 h after siRNA transfection, were digested and adjusted to a density of approximately 500 cells per milliliter using complete medium. Two milliliters of this suspension was seeded into each well of a 6-well plate. The plate was gently rocked for even distribution and placed in a constant-temperature incubator. Cell attachment and density were checked the next day. Cells were then cultured for seven to ten days, with the medium changed once or twice during this period. When colonies could be seen, the culture was stopped. The medium was thrown away. Cells were rinsed two or three times with PBS. Two milliliters of 4% paraformaldehyde was added to each well for fixation for 20 to 30 min. After rinsing again with PBS two or three times, two milliliters of 0.1% crystal violet solution was put in for staining. This lasted 15 to 20 min. Wells were then rinsed with PBS until the wash solution became clear and air-dried at room temperature. Colony formation was photographed using a flat-panel light box, and colony morphology was captured under a 5X inverted microscope.

EdU assay

Cells in good growth condition, typically harvested around 24 h after siRNA transfection, were digested with trypsin containing EDTA. After counting, cells were thinned with full medium. One milliliter of the mix, with about 100,000 cells, was put into each well of a 12-well plate. After thorough mixing, plates were labeled and incubated overnight. After cells fully attached, EdU reagent A was diluted 1000:1 with complete medium. The medium in each well was replaced with this dilution, and plates were returned to the incubator for three hours. For fixation, wells were washed two to three times with PBS, and 500 µl of 4% paraformaldehyde was added for 10 to 15 min at room temperature. PBS was added, and plates were washed on a decolorizing shaker three times for five minutes each. After throwing away the PBS, 200 µl of PBS with 0.5% Triton X-100 was added to each well. Plates were put on a shaker for 10 min. Then they were rinsed two times with PBS, and leftover liquid was taken off. For Apollo staining, the Apollo staining liquid was made as the kit said. 150 µl was put into each well. A piece of sealing film, cut to fit the plate, was placed on top. This lowered surface tension. Plates were then placed on a decolorizing shaker. They stayed at room temperature in the dark for 30 min. After that, the staining liquid was thrown away. Wells were rinsed with PBS. Next, 200 µl of DAPI mounting medium was added to each well. Plates went back on the decolorizing shaker at room temperature in the dark for 10 min. Extra mounting medium was taken off. Cells were looked at under a fluorescence microscope. Cells that were growing and had EdU showed red light. All cell centers with DAPI showed blue light. Results were quantified using ImageJ software.

Flow cytometry for apoptosis detection

To examine the effect of transfection on apoptosis, cells were collected 48 h after transfection. Culture supernatant was retained. Cells were digested using pre-warmed trypsin without EDTA and neutralized with complete medium. The total cell number was no fewer than 80,000. The liquid above and the cell mix were put together in a 15 ml spin tube. They were mixed softly. Then they were spun at 2000 rpm for 8 min. The cell lump at the bottom was rinsed two times with cold PBS. The 5X Binding Buffer from the assay kit was diluted to 1X working solution using ddH₂O. Each sample was gently resuspended in 100 µl of this working solution and transferred to a labeled flow cytometry tube. 5 µl of Annexin V was added to each tube in the dark. Tubes were mixed on a vortex and left at room temperature in the dark for 15 min. After that, 5 µl of PI was added to each tube. Tubes were mixed again and then left in the dark for 5 to 10 min. Apoptosis was checked with a CytoFlex LX flow cytometer. The information was studied with FlowJo software.

Viral transfection

On day one, the required cells were expanded in vitro. Cells that looked healthy and were growing fast were put into 6-well plates. They covered about 50 to 60% of the plate surface. On day two, the viral stock was taken from the − 80℃ freezer and placed at 4℃ to thaw slowly. The medium in the 6-well plates was replaced with one milliliter of fresh complete medium. Separately, one milliliter of complete medium was added to an EP tube, followed by a virus amount equivalent to 1 × 10⁸ TU and the lentiviral infection enhancer HiTransG P. This mixture was mixed well, evenly dripped into the wells, and mixed gently. Plates were labeled and returned to the incubator. After the procedure, the work surface and pipettes were cleaned with 84 disinfectant, and consumables were autoclaved immediately. On day three, cell status was observed. Usually, the old medium was changed for new full medium 12 to 16 h after transfection. If cell status was poor or many dead cells were present, viral transfection was terminated earlier by changing the medium. On day four, cells were passaged at a 1:2 ratio for further expansion. If the virus carried a GFP fluorescent tag, transfection efficiency could be observed under an inverted fluorescence microscope. On day six, a puromycin kill-off assay was performed. Based on literature and preliminary experiments, puromycin was added to complete medium. Medium in one dish of cells was replaced with this puromycin-containing medium, while another dish was kept as a parallel control without puromycin. After two days of puromycin selection, the concentration was halved and maintained for another seven days. For cells with a Luciferase tag, between days 10 and 14, an appropriate number of cells were placed in a black-walled plate. After putting in D-Luciferin, good transfection was checked with a small animal fluorescence imaging system.

Establishment of a subcutaneous tumor model in nude mice

One week before the experiment, BalB/C nu/nu mice were put into a clean room. The temperature there was 23 ± 2℃,and the humidity was 50 ± 10%. The light and dark periods were 12 h each. The mice drank sterilized water and ate sterilized food. After one week, the mice were used in the experiments, following the rules in the Guide for the Care and Use of Laboratory Animals. PLC-PRF-5 cells that had been given Luciferase lentivirus were grown outside the body. After breaking the cells apart, they were placed in serum-free medium and mixed equally with Matrigel while kept on ice. Using a small syringe, 100 µl of the cell mixture, which had 5 × 10⁷ cells per milliliter, was injected into the right thigh of each mouse. The mice were separated randomly into two groups. One group got injections of LARS KD AAV and the other group got CON AAV. The tumors were looked at seven days later. Then, every two days for 21 days, the weight of the mice and the size of the tumors were recorded. Tumor size was worked out by taking the length times the width squared, divided by two. Lines on graphs were drawn to show how the tumors grew and how the mice body weights changed. Weekly, 30 µl of virus at 1 × 10¹⁰ TU was injected into the tumors. On day 21, mice from both groups were sent for Luciferase imaging. D-Luciferin (sodium salt) stock solution was diluted with PBS to 15 mg/mL. Each mouse received an intraperitoneal injection of 100 µl and was allowed to move freely in its cage for seven minutes for even substrate distribution. Mice were put to sleep with isoflurane. Then, pictures of the tumors inside the mice were taken with an IVIS Spectrum machine. Mice were humanely euthanized under deep anesthesia followed by cervical dislocation. Tumors were dissected, photographed, and weighed. The heart, liver, spleen, lungs, and kidneys were taken out. They were placed in 4% paraformaldehyde to keep them. Then they were put into paraffin blocks, stained with HE, and stained for immunohistochemistry.

Bioinformatics analysis

Data from the TCGA-LIHC project and patient details were used to see how LARS gene levels might affect patient outcomes (Tomczak et al. 2015). To study survival, patients were put into two groups: high LARS or low LARS, based on the middle value of LARS levels. Kaplan-Meier curves and a log rank test were used to check for differences between these groups in several measures: disease-free interval (DFI), disease-specific survival (DSS), overall survival (OS), and progression-free interval (PFI) (Andrade 2023). The UALCAN database helped examine LARS expression in liver cancer (Chandrashekar et al. 2017). The study included 371 primary tumor samples and 50 normal liver tissue samples. Numbers were used to check if LARS levels were different between healthy liver tissues and cancer tissues. Further comparisons were made between normal tissues and cancer tissues at different stages (Stage1, Stage2, Stage3, Stage4). Differences across ethnic groups (Caucasian, African American, Asian) were also evaluated against normal samples. T-tests compared expression between normal tissues and tissues from male or female patients. All statistical tests were performed using the analysis tools within the UALCAN platform. Additional comparisons involved different age groups (21–40, 41–60, 61–80, 81–100 years), various weight categories (normal weight, overweight, obese, severely obese), different tumor grades, and different pathologic N stages relative to normal tissue. The analysis also assessed LARS expression based on TP53 mutation status. Information about TP53 gene changes came from TCGA whole-exome sequencing. It was downloaded as MAF files. Samples that had the TP53 mutation or did not have it were linked to their RNA-seq data. Statistical importance was shown as *P < 0.05, **P < 0.01, and ***P < 0.001.

Molecular docking and molecular dynamics simulations

We have learned from the CTD (Comparative Toxicogenomics Database) that triphenyl phosphate (PubChem CID: 8289) is an environmental endocrine disruptor and toxicant associated with both the LARS protein (PDB ID: 6KIE) and hepatocellular carcinoma (Davis et al. 2023). The protein structure was obtained from the PDB (Protein Data Bank) database and preprocessed using PyMOL 3.1 to remove crystallographic water, non-physiological ions, and irrelevant ligands, while missing residues or atoms were corrected to ensure structural integrity (Bittrich et al. 2023). Triphenyl phosphate was retrieved from PubChem and optimized for docking through structure refinement and format conversion (Kim et al. 2021). The researchers used a computer program called AutoDock Vina v1.1.2 to see how molecules might stick to the LARS protein (Stanzione et al. 2021). The part of the LARS protein where things might bind was marked using PyMOL 3.1. Computer simulations were run to work out two things: the root mean square deviation (RMSD) and how much each part of the protein moved, called RMSF (Filipe and Loura 2022). The simulation also looked at how hydrogen bonds changed over time. The solvent-accessible surface area (SASA) was measured to understand the binding area. Another method, principal component analysis (PCA), was used to simplify the shapes the molecules took. It showed them on a two-dimensional plot. This plot was used to see the Gibbs free energy, measured in kJ/mol. Binding energy contributions were assessed through residue energy decomposition, while the radius of gyration (Rg) was calculated to analyze conformational compactness. Finally, XVG box comparison was conducted to examine system behavior under different simulation conditions.

Western blot analysis

Cells were lysed on ice with RIPA buffer containing protease and phosphatase inhibitors, scraped, briefly sonicated, and centrifuged to obtain clarified lysates. Protein concentration was determined by BCA assay; samples were normalized, mixed with 5× loading buffer, and denatured by boiling. SDS-PAGE gels were prepared with appropriate resolving and stacking phases; equal protein amounts were loaded alongside pre-stained markers and separated by stepwise constant-voltage electrophoresis (80 V → 120 V). Proteins were transferred to methanol-activated PVDF membranes under constant current (250 mA) in a cooled wet-transfer system. Membranes were blocked with 5% BSA, incubated overnight at 4 °C with primary antibodies, washed, and probed with HRP-conjugated secondary antibodies. Signals were detected using ECL reagent on a chemiluminescence imager, and band intensities were quantified by densitometry with ImageJ.

Co-Immunoprecipitation (Co-IP) assay cell lysis and protein extraction

Co-Immunoprecipitation (Co-IP) Assay Cells were lysed on ice using IP/WB buffer containing protease and phosphatase inhibitors, followed by scraping, extended ice incubation (10–20 min), and centrifugation to obtain clarified lysates. Target antibody (or control) was coupled to pre-washed magnetic beads by room-temperature rotation (10–30 min), followed by thorough PBS washing. Clarified lysate was incubated with antibody-beads overnight at 4 °C with gentle rotation to capture interacting proteins. Beads were then magnetically separated, washed multiple times with PBS, and bound complexes were eluted by boiling in 1× SDS loading buffer for 10 min at 100 °C. The eluate was collected after magnetic separation and used directly for downstream SDS-PAGE and immunoblotting.

Results

LARS knockdown suppresses HCC growth

Knockdown of the LARS gene inhibited clonogenic ability in hepatocellular carcinoma (HCC). Researchers used siRNA to lower LARS. After that, PLC‑PRF‑5 and HCC‑LM3 cells made fewer and smaller colonies than control cells. The difference was significant. This meant less LARS lowered the ability of these HCC cells to form colonies (Fig. 1a-d). LARS knockdown also impaired HCC cell proliferation. We repeated the CCK-8 assay and established a standard curve correlating OD values with cell numbers. The results showed that si-NC-treated HCC-LM3 cells proliferated approximately 5-fold within 48 h. si-LARS cells exhibited significantly lower proliferation than si-NC controls at all time points (Fig. 1e). Furthermore, LARS knockdown decreased the number of HCC cells in the proliferative phase. EdU, a thymidine analogue, incorporates into newly synthesized DNA during proliferation, and Apollo fluorescent dye specifically labels EdU, producing red fluorescence under microscopy to directly indicate DNA replication activity. An EdU test was done. It showed fewer red glowing cells in HCC cells with less LARS compared to control cells (Fig. 1f-i). This proved that lowering LARS greatly reduced HCC cell growth.

Fig. 1.

LARS knockdown inhibits HCC clonogenicity, proliferation, and DNA synthesis. a‑d Colony formation assay in PLC‑PRF‑5 and HCC‑LM3 cells after LARS knockdown. e CCK‑8 proliferation assay. f‑i EdU staining for proliferating cells

LARS knockdown inhibits HCC cell migration and invasion and promotes apoptosis

Transwell tests looked at how HCC cells move and invade. The cells used were PLC-PRF-5 and HCC-LM3. When LARS was lowered, far fewer cells moved through the membrane in both tests compared to the control (Fig. 2a, b). This meant less LARS stopped HCC cells from moving and invading. Scratch tests also checked cell movement. Lowering LARS made the scratch close much slower in both cell types (Fig. 2c-e). Also, less LARS led to more HCC cell death. Flow cytometry found many more early and late dying cells in the low LARS group than in the control group (Fig. 2f, g).

Fig. 2.

LARS knockdown inhibits migration, invasion, and promotes apoptosis in HCC cells. a, b Transwell migration and invasion assays in PLC-PRF-5 and HCC-LM3 cells. c-e Wound healing assays in PLC-PRF-5 (c) and HCC-LM3 cells (d) with a statistical chart (e). f, g Flow cytometry analysis of apoptosis

LARS gene effects in a mouse liver cancer model

To investigate the role of LARS in liver cancer in vivo, a subcutaneous tumor model was established in nude mice. The mice got PLC‑PRF‑5 cells that glow. On day 7 after inoculation, tumors were successfully formed. Mice then received local injections of either LARS‑KD AAV or CON AAV on days 7 and 14. Mouse weight and tumor size were checked every two or three days. On day 21, luciferase imaging was performed to evaluate tumor size, after which the tumors were excised, weighed, and analyzed pathologically. The results showed that tumors in the LARS‑KD AAV group were smaller in volume and weight, and luciferase fluorescence was weaker compared with the CON AAV group, while body weight did not differ significantly between groups (Fig. 3a-e). HE staining and Ki-67 immunohistochemical (IHC) staining of the xenografts revealed that CON AAV tumors displayed irregular nest-like structures, polygonal cells with eosinophilic cytoplasm, variable nuclear size, prominent nucleoli, obvious mitoses, and fibrous septa with inflammatory infiltration. In contrast, LARS‑KD AAV tumors showed mucoid degeneration, marked stromal fibrosis, and abundant foam‑cell infiltration with lymphoid aggregation. Index of Ki-67 in LARS‑KD AAV tumors group decreased significantly (Fig. 3f-g). HE staining of major organs (heart, liver, spleen, lung, and kidney) revealed no pathological changes in either group, indicating that LARS knockdown did not cause detectable damage to these tissues (Fig. 3h). Together, these in vivo findings suggest that LARS may be associated with HCC malignancy, and that knocking down LARS reduces tumor growth without affecting major organ histology.

Fig. 3.

LARS knockdown inhibits tumor growth in a subcutaneous xenograft model without damaging major organs. a Tumor volume. b Body weight changes. c Luciferase images of tumors at day 21. d Quantitative luminescence intensity. e Tumor weight. f HE staining and Ki-67 immunohistochemical staining of CON and LARS-KD tumors. The lighter background in LARS-KD Ki-67 images is due to different exposure conditions during image acquisition, which does not affect result interpretation. g Percentage of Ki-67 expression. h HE staining of heart, liver, spleen, lung, and kidney tissues

LARS knockdown alters the PI3K/AKT/mTOR pathway in HCC cells

RNA-seq was performed on PLC-PRF-5 cells with LARS knockdown and control cells. Differential gene enrichment analysis revealed significant enrichment of the PI3K/AKT/mTOR pathway in LARS-knockdown cells (Fig. 4a, b). Protein analysis with TMT matched the RNA findings. It also showed changes in the PI3K/AKT and mTOR pathways in proteins (Fig. 4c, d). These findings suggested the PI3K/AKT/mTOR pathway could be key in how LARS affects HCC. Western blot checked proteins linked to cell death after lowering LARS (Fig. 4e-g). Bcl-2 decreased, while Bax remained unchanged. Based on the RNA-seq and TMT pathway results, key proteins in the PI3K/AKT/mTOR pathway were further tested. LARS knockdown reduced p-mTOR (Ser2448) and PI3K levels, while total AKT did not change. Notably, although PI3K signaling outputs decreased after LARS knockdown, p-AKT(Ser473) was paradoxically increased, which may reflect compensatory mTORC2-driven Ser473 phosphorylation rather than restoration of full AKT activity, a finding consistent with the observed suppression of mTORC1 downstream signaling (Najafov et al. 2012).

Fig. 4.

Effects of LARS knockdown on the PI3K/AKT/mTOR pathway and apoptosis-related proteins in HCC cells. a, b KEGG enrichment of RNA-seq data showing PI3K/AKT/mTOR pathway enrichment. c, d Proteomic analysis confirming changes in PI3K/AKT/mTOR-related proteins. e-g Western blot of apoptosis and pathway proteins after LARS knockdown

Identification of RPS5 as a LARS-interacting protein

Mass spectrometry analysis indicated an interaction between RPS5 and the LARS protein. To further explore the aberrant change in pAKT observed in LARS-knockdown cells, we analyzed the proteins interacting with LARS (Fig. 5a, b). A Co-IP experiment used a tagged LARS to catch proteins that bind to it. The caught proteins were then analyzed by mass spectrometry. The equipment used was an EASY-nLC 1200 UHPLC system with a mass spectrometer, and it worked steadily. Fifty proteins were found to interact with LARS. Using the UniProt database, PI3K-interacting proteins were downloaded and compared with the LARS-interacting proteins, revealing 28 overlapping proteins. These overlapping proteins were further analyzed for interactions, and PPI network was constructed with high-confidence interactions (score > 0.9). Using Cytoscape, the top 10 proteins ranked by MCC were selected to generate the PPI network, identifying RPS5 as a potential core protein. Given the discrepancy between PI3K and downstream AKT expression, literature review suggested that RPS5 might mediate alternative regulation of AKT. These results indicated that RPS5 could be a key target responsible for the aberrant pAKT change upon LARS knockdown. Co-IP in both directions confirmed the interaction between LARS and RPS5 (Fig. 5c, d). Based on earlier Co-IP and mass spectrometry results in Huh7-LARS OE cells, RPS5 was identified as a LARS-interacting protein. HA-RPS5 or Flag-LARS plasmids were transfected into HEK-293T cells, and after 48 h, cell lysates were subjected to Co-IP using tagged magnetic beads. Western blot analysis of the pulled-down proteins showed expression of Flag-LARS and HA-RPS5 in the Input group, and both proteins were detected on antibody-bound agarose beads, while no bands were observed in the control group, confirming the exogenous interaction between LARS and RPS5 in HEK-293T cells. Co-localization of LARS and RPS5 was observed in human hepatocellular carcinoma. To further validate their interaction, immunofluorescence was used to examine the co-localization of LARS and RPS5 in human hepatocellular carcinoma cells. Confocal microscopy showed green fluorescence for LARS and red fluorescence for RPS5. The results (Fig. 5e) showed significant co-localization of LARS and RPS5 signals in PLC-PRF-5 and HCC-LM3 cells (Pearson correlation coefficients of 0.80 and 0.82, respectively).

Fig. 5.

LARS interacts with RPS5. a, b Mass spectrometry identified RPS5 as a LARS-interacting protein in Huh7 cells. c, d Bidirectional Co-IP confirmed the LARS-RPS5 interaction in HEK-293T cells. e Immunofluorescence confirms co-localization of LARS and RPS5 in PLC-PRF-5 and HCC-LM3 cells

LARS and RPS5 protein interaction and functional effects

A computer simulation tested how LARS and RPS5 proteins might stick together (Fig. 6a). The RPS5 protein sequence was retrieved from the Uniprot database and used for a BLAST search in SWISS-MODEL. A template protein (PDB ID: 6SWD) with 66% homology was identified and used for homology modeling. To determine the potential interaction between LARS and the RPS5 model, the ZDOCK program was used to predict a binding model. The calculated interaction surface area was 859.1 Ų with a binding energy of −7.0 kcal/mol, indicating a stable interaction. The ring domain of RPS5 was complementary to residues 276, 369, 373, 425, and 426 of LARS. Hydrophobic interactions were formed between RPS5 residues PHE5, LEU9, VAL14, PHE18 and the mentioned LARS residues, creating a hydrophobic pocket around LARS. Furthermore, four hydrogen bonds were formed between polar groups of RPS5 (VAL7, LEU9, ARG25, LYS118) and LARS residues TYR373, ARG419, PRO428, ASN439, which likely strengthened the binding affinity. Knockdown of the LARS gene in cell lines led to downregulation of RPS5 protein levels. Having established a protein-protein interaction between LARS and RPS5, siRNA knockdown of LARS was performed in PLC-PRF-5 and HCC-LM3 cell lines to further validate this finding. Western blot analysis confirmed that RPS5 protein levels decreased following LARS knockdown (Fig. 6b, c). The effect of overexpressing RPS5 in LARS knockdown cells on proliferation was then investigated. After making RPS5 levels high in cells with low LARS, cell growth and colony formation were tested with CCK-8 and colony assays (Fig. 6d-g). The CCK-8 test found that high RPS5 brought back the cell growth that was lost when LARS was low. Similarly, the colony formation assay demonstrated that RPS5 overexpression restored the impaired clonogenic ability of HCC cells resulting from LARS knockdown.

Fig. 6.

Analysis of LARS and RPS5 interaction and functional rescue. a AutoDock simulation of LARS-RPS5 interaction. b, c Western blot analysis of RPS5 protein levels after LARS knockdown in PLC-PRF-5 and HCC-LM3 cells. d, e CCK-8 proliferation assay of HCC cells with LARS knockdown and RPS5 overexpression. f, g Colony formation assay of HCC cells with LARS knockdown and RPS5 overexpression

LARS expression shows differential associations with clinical outcomes in HCC

In the TCGA-LIHC patient data, how LARS gene levels related to different patient results was not the same. Survival analysis revealed that patients with high versus low LARS expression did not significantly differ in disease-free interval (DFI, p = 0.21, Fig. 7a), disease-specific survival (DSS, p = 0.059, Fig. 7b), or progression-free interval (PFI, p = 0.14, Fig. 7d). However, a significant difference was observed in overall survival (OS), with patients exhibiting high LARS expression having shorter survival times compared to those with low LARS expression (p = 0.01, Fig. 7c).

Fig. 7.

Kaplan-Meier survival analysis of LARS expression in TCGA-LIHC. a Disease-free interval (DFI). b Disease-specific survival (DSS). c Overall survival (OS). d Progression-free interval (PFI)

LARS expression in hepatocellular carcinoma

LARS expression of UALCAN database (https://ualcan.path.uab.edu/) showed significant differences between normal tissues (n = 50) and primary tumor tissues (n = 371, Fig. 8a). Analysis across HCC stages revealed markedly varied LARS expression, especially in early stages (Fig. 8b). Significant differences in LARS expression were also observed across racial groups (Fig. 8c), between genders (Fig. 8d), and among different age groups compared to normal (Fig. 8e). Variations were further evident among body weight classifications (Fig. 8f) and across tumor grades relative to normal tissue (Fig. 8g). In LIHC, comparison of normal tissues with N0 stage showed highly significant differences, while normal versus N1 stage did not reach statistical significance (Fig. 8h). Significant differences were found between normal and both TP53‑Mutant as well as TP53‑NonMutant groups (Fig. 8i). *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 8.

LARS expression in HCC of UALCAN database. a Normal vs. primary tumor. b Across HCC stages. c Among racial groups. d Between genders. e Across age groups. f By body weight classification. g Across tumor grades. h Normal vs. N0/N1 stages. i Normal vs. TP53 mutant status

Interaction between LARS protein and triphenyl phosphate

The binding energy between triphenyl phosphate (TPP) and the LARS protein was − 8.0 kcal/mol (Fig. 9a, b), indicating affinity. Docking analysis revealed a backbone donor interaction between the O = P group of TPP and the Tyr 54 residue of LARS (Fig. 9c). In MD simulations, the RMSD of LARS initially increased and then stabilized, while TPP alone showed low RMSD, and the complex exhibited higher yet stable RMSD (Fig. 9d). RMSF analysis highlighted dynamic residue behaviors, with high-RMSF regions suggesting potential interaction sites (Fig. 9e). A stable hydrogen bond network formed between TPP and LARS (average 2 bonds, peaking at 3 bonds during 20–40 ns) (Fig. 9f). SASA analysis confirmed their tight binding (Fig. 9g), and PCA visualized energy landscapes (red: high energy; blue: low energy) (Fig. 9h, i). Binding free energy decomposition identified A: TYR:52 and A: ASP:676 as major positive contributors, whereas A: LEU:677 and A: PRO:53 hindered binding (Fig. 9j). The Rg and gyration radii (median Gx ≈ 2.3, Gy ≈ 2.6, Gz ≈ 2.3) indicated conformational compactness changes (Fig. 9k, l).

Fig. 9.

Molecular docking and molecular dynamics simulation. a, b Binding energy of the molecular docking for the TPP-LARS complex. c Two-dimensional interaction map. d RMSD trajectory. e RMSF. f Hydrogen bond count. g SASA analysis. h, i Free energy landscape (PC1/PC2). j Energy contribution of hot residue. k Rg value of the complex. l XVG box comparison

Discussion

Lowering LARS gene expression stopped HCC cells from growing and spreading quickly. In lab tests, when LARS was reduced, HCC cells grew less, made fewer colonies, moved slower, and died more often. In a mouse model, LARS knockdown significantly slowed the growth of subcutaneous tumors. The tests showed that LARS worked by changing the PI3K/AKT/mTOR pathway and by binding to the RPS5 protein. Overexpression of RPS5 partially reversed the anti-proliferative effects caused by LARS reduction. Looking at patient data, high LARS levels in tumors linked to shorter life and worse disease signs.

The observed suppression of HCC cell growth and promotion of apoptosis following LARS reduction suggested that this gene plays a crucial role in maintaining tumor cell survival and aggressive properties. The in vivo findings from the mouse model supported this conclusion, showing that targeting LARS could impede tumor progression. To further interpret these phenotypic effects, we explored the molecular mechanisms underlying LARS function. Gene and protein tests both found that the PI3K/AKT/mTOR pathway changed when LARS was lowered. This pathway is known to control cancer cell growth and survival. The downregulation of key components such as PI3K and pmTOR provided a plausible molecular explanation for the reduced cell proliferation and increased apoptosis.

The idea that LARS helps HCC grow via the PI3K/AKT/mTOR pathway fits what is already known about liver cancer signals (Fan et al. 2024, Shi et al. 2025). Many papers showed that a too active PI3K/AKT/mTOR pathway helps HCC grow and resist treatment (Wu et al. 2020, Bang et al. 2023, Li et al. 2022, Sun et al. 2021, Li and Xiong 2022). The new finding here was that LARS acts as an upstream controller of this pathway in HCC. Furthermore, the discovery of RPS5 as a LARS-interacting protein added another layer to the mechanism. The interaction between LARS and RPS5, and the subsequent rescue of proliferation upon RPS5 overexpression, suggested that LARS might exert its oncogenic function partly through stabilizing or regulating RPS5, which in turn could influence downstream signaling events, including the observed atypical activation of AKT. Notably, RPS5 is a key structural component of the 40 S ribosomal subunit involved in translation initiation and has been reported to exert extra-ribosomal functions linked to Akt signaling, suggesting that the LARS-RPS5 interaction may directly couple translational control with growth-promoting pathways (Qiu et al. 2023, Lei et al. 2025) These findings support that LARS links ribosome metabolism to oncogenic signaling in hepatocellular carcinoma. Beyond its canonical role in leucyl-tRNA synthesis, LARS may influence translational capacity through interaction with the ribosomal protein RPS5, thereby modulating protein synthesis programs that are frequently dysregulated in cancer. In parallel, the regulation of the PI3K/AKT/mTOR pathway by LARS places it upstream of a central growth-promoting signaling axis tightly coupled to ribosome biogenesis. Together, these observations suggest that LARS functions as an integrative node coordinating translational control and oncogenic signaling to foster malignant progression in HCC.

The findings of this research have several implications. LARS may serve as a potential novel therapeutic target for HCC, and strategies aimed at inhibiting its expression or function warrant further exploration for drug development. High LARS in tumors linked to poor outcomes, so it might help predict patient survival and choose treatments. The interaction with RPS5 unveiled a previously unrecognized molecular connection between aminoacyl-tRNA synthetase machinery and ribosomal proteins in the context of cancer, offering a new direction for understanding protein synthesis dysregulation in tumors. The molecular docking and dynamics simulation data, suggesting a potential interaction between LARS and the environmental contaminant triphenyl phosphate, offer a new perspective for investigating whether environmental factors might influence HCC risk by interfering with specific molecular targets.

Despite these findings, the study has several limitations. The in vivo evidence relied on a subcutaneous xenograft mouse model, which does not fully replicate the complex liver microenvironment and immune context of human HCC. Orthotopic or genetically engineered mouse models might provide more relevant insights. The mechanistic link between LARS, RPS5, and the precise regulation of the PI3K/AKT/mTOR pathway required further detailed validation.

Despite these findings, the study has several limitations. The in vivo evidence relied on a subcutaneous xenograft mouse model, which does not fully replicate the complex liver microenvironment and immune context of human HCC. Orthotopic or genetically engineered mouse models might provide more relevant insights. The mechanistic link between LARS, RPS5, and the precise regulation of the PI3K/AKT/mTOR pathway required further detailed validation. While the interaction was confirmed and a functional rescue was shown, the exact molecular steps, whether through affecting protein stability, complex formation, or translational efficiency, remained to be fully elucidated. The clinical correlation analysis, while informative, was retrospective and based on public databases. Future studies with more patients are needed to confirm LARS as a prognosis marker. Additionally, the potential binding of triphenyl phosphate to LARS was only predicted computationally and required experimental confirmation to assess its biological significance.

Conclusions

The present study provides evidence that LARS may function as an oncogene in hepatocellular carcinoma. It appears to promote tumor cell growth and invasion, likely by modulating the PI3K/AKT/mTOR signaling pathway and interacting with the ribosomal protein RPS5. High LARS in HCC tissues linked to worse results, showing it might be both a prognosis marker and a treatment target (see Fig. 10 for a schematic summary). These results help understand HCC better and point to new ways to treat it. Future research should focus on validating the mechanistic details in more physiologically relevant models and exploring the translational potential of targeting LARS in hepatocellular carcinoma.

Fig. 10.

Schematic model of LARS in Hepatocellular carcinoma (HCC) progression. Left: normal hepatocyte; right: HCC cell. Compared with normal cells, HCC shows upregulated LARS, enhanced LARS–RPS5 interaction, and overactivated PI3K/AKT/mTOR signaling, leading to increased proliferation, invasion, migration, and reduced apoptosis. Arrows indicate change trends: blue for normal, red for HCC. “×” denotes LARS–RPS5 interaction. LARS: leucyl-tRNA synthetase; RPS5: ribosomal protein S5; mTOR pathway: a core signaling axis regulating cell growth, survival, and metabolism

Authors’ contributions

Hanbin Chen: Conceptualization, Investigation, Formal analysis, Writing - Original Draft. Wei Chen: Methodology, Data Curation, Validation. Shicheng Xie: Resources, Visualization. Bin Wang: Software, Investigation. Shishi Zhu: Validation, Writing - Review & Editing. Athanasios G. Papavassiliou: Supervision, Resources, Writing - Review & Editing. Zhijie Yu: Supervision, Project administration, Funding acquisition. Jinlin Xia: Supervision, Funding acquisition, Writing - Review & Editing, Project administration. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grants 52103196, 32201118, and 81972233), the Wenzhou Institute UCAS startup fund (WIUCASQD2023010), and the Wenzhou Municipal Basic Scientific Research Project (Y20240114).

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request. The public datasets (TCGA-LIHC, UALCAN) used in this study are available through their respective platforms.

Declarations

Ethics approval and consent to participate

All experimental procedures involving animals were approved by the Animal Ethical and Welfare Committee of the First Affiliated Hospital of Wenzhou Medical University (Approval No. WYYY-IACUC-AEC-2021-312). The study was conducted in accordance with the approved experimental protocol and adhered to relevant guidelines and regulations for animal research, including the Guide for the Care and Use of Laboratory Animals.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.