Targeting TLK2 inhibits the progression of gastric cancer by reprogramming amino acid metabolism through the mTOR/ASNS axis

Several recent studies have suggested that TLKs are related to tumor progression. However, the function and mechanism of action of TLK2 in gastric cancer (GC) remain elusive. In this study, TLK2 was found to be significantly upregulated in patients with GC and was identified as an independent prognostic factor for GC. Consistently, TLK2 knockdown markedly reduced the aggressiveness of GC, whereas its overexpression had the opposite effect. IP-MS revealed that the effects of TLK2 on GC were mainly associated with metabolism reprogramming. TLK2 knockdown suppressed amino acid synthesis by downregulating the mTORC1 pathway and ASNS expression in GC cells. Mechanistically, mTORC1 directly interacts with the ASNS protein and inhibits its degradation. Further experiments validated that the ASNS protein was degraded via ubiquitination instead of autophagy. Inhibiting and activating the mTORC1 pathway can upregulate and downregulate ASNS ubiquitination, respectively, and the mTORC1 pathway can reverse the regulatory effects of TLK2 on ASNS. Furthermore, TLK2 was found to regulate the mRNA expression of ASNS. TLK2 directly interacted with ATF4, a transcription factor of ASNS, and promoted its expression. The kinase inhibitor fostamatinib significantly inhibited the proliferative, invasive, and migratory capabilities of GC cells by inhibiting TLK2 activity. Altogether, this study reveals a novel functional relationship between TLK2 and the mTORC1/ASNS axis in GC. Therefore, TLK2 may serve as a potential therapeutic target for GC.


INTRODUCTION
Gastric cancer (GC) is one of the most common cancers and the second leading cause of cancer-related deaths worldwide [1].At present, perioperative care is undertaken in the treatment of GC [2]; however, the median survival time of patients with advanced GC remains <12 months [3].Malignant cells, including GC cells, are distinct from normal epithelial cells.According to the Warburg effect, cancer cells rely on aerobic glycolysis instead of oxidative phosphorylation [4].In addition to abnormal glucose metabolism, several other altered metabolic processes, including reprogramming of amino acid metabolism, facilitate uncontrolled growth and proliferation of GC cells [5].With the increasing knowledge and understanding of metabolic reprogramming, cancer therapies targeting amino acid metabolism have recently emerged [6,7].Elucidating the mechanisms underlying alterations in amino acid metabolism is vital for developing effective therapeutic strategies and identifying potential therapeutic targets for GC.
In mammals, tousled-like kinases (TLKs) are categorised into two subtypes, namely, TLK1 and TLK2, which can bind to each other and are actively expressed in the S phase of the cell cycle.TLKs are associated with cell cycle progression and DNA damage repair [8,9]; however, their role in cancer has received less attention in cancer research.Studies have suggested that TLK1 plays an important role in the progression of various malignant tumours, including prostate cancer, kidney cancer and hepatocellular carcinoma [10][11][12].Small-molecule drugs, such as phenothiazine, can target TLK1 to inhibit the progression of prostate cancer.Therefore, TLK1 is considered a novel therapeutic target for prostate cancer [13].On the contrary, the role of TLK2 in cancer progression remains unclear.To date, only one study has reported that TLK2 is abnormally expressed in breast cancer [14], and the biological functions of TLK2 in most other cancers, including GC, remain elusive.
In this study, we found that TLK2 regulated the expression of the asparagine synthetase (ASNS) protein, a key enzyme in amino acid synthesis.ASNS is involved in the progression of multiple malignancies, including lung cancer, ovarian cancer and GC [15][16][17].In addition, we found that TLK2 was closely associated with the mechanistic target of the rapamycin complex 1 (mTORC1) pathway and amino acid metabolism.As an intracellular protein complex, mTORC1 consists of five components, namely, mTOR, RAPTOR, DEPTOR, mLST8 and PRAS40 [18].mTORC1 can integrate signals from growth factors and nutrients to control biosynthesis, including protein, lipid and nucleic acid synthesis [19].In addition, mTORC1 serves as a core protein that regulates amino acid metabolism and perceives dynamic changes in amino acid levels to regulate protein synthesis in cells [20].Therefore, mTORC1 can substantially promote the proliferation of tumour cells by activating anabolic pathways [21].However, the specific mechanisms underlying ASNS-and mTORC1-mediated amino acid metabolism in GC remain unclear.
In this study, TLK2 was found to function as an oncoprotein in GC.It promoted GC progression mainly by regulating the reprogramming of amino acid metabolism.Mechanistically, TLK2 prevents ASNS from being ubiquitinated to increase its stability.In addition, TLK2 binds to ATF4, thereby regulating its protein expression and promoting its transcription.We speculated that TLK2 can markedly affect the amino acid metabolism of GC cells by regulating ASNS expression.Targeting amino acid metabolism is an effective strategy for the treatment of malignant tumours [22,23].Altogether, this study suggests that TLK2 promotes GC progression by regulating amino acid metabolism and provides potential therapeutic targets and a theoretical basis for targeted therapy of GC.

Patients, tissue microarrays and immunohistochemical analysis
For establishing tissue microarrays (TMAs), a total of 107 GC tissues and 22 randomly selected adjacent normal tissues were collected from the general surgery department of the First Affiliated Hospital of Anhui Medical University from October 2012 to December 2013.The follow-up time ranged from 8 to 71 months.All GC tissues were examined and staged by pathologists according to the tumour-node-metastasis staging system and the American Joint Committee (7 th edition).Immunohistochemical (IHC) staining was performed as described previously [24].Protein levels in TMAs were evaluated by two pathologists who were blinded to patient information.This study was approved by the Ethics Committee of the First Affiliated Hospital of Anhui Medical University.

Western blotting
Total protein was extracted from GC cells and tissues using M-PER™ Mammalian Protein Extraction Reagent (Thermo Fisher Scientific, Waltham, MA, USA; cat.number: 78501).The extracted proteins were quantified using a BCA Protein Assay Kit (Beyotime, Shanghai, China; cat.number: P0012) and separated via sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) (8-12% gels).The separated proteins were transferred to PVDF membranes and blocked using TBST containing 5% milk for 1 h [24].Subsequently, the membranes were incubated with primary antibodies overnight at 4 °C.The following day, the membranes were incubated with secondary antibodies, and protein bands were visualised via enhanced chemiluminescence.

Assessment of protein stability
Cycloheximide (CHX) chase assay was performed to assess the stability of the ASNS protein in rapamycin-treated GC cells.GC cells were pre-treated with rapamycin for 24 h and subsequently treated with 20 μg/mL of CHX (MCE; cat.number: HY-12320) for 0, 2, 4 and 8 h.Cell lysates were collected for western blotting.

Co-immunoprecipitation
Co-immunoprecipitation was performed as described previously [25].Respective antibodies (2 µg) were mixed with 500 µg of protein sample, and the mixture was incubated on a shaker for 8 h.Subsequently, the mixture was incubated with Protein A/G PLUS-Agarose (Santa Cruz Biotechnology) overnight, and immunoprecipitation complexes were detected via immunoblotting.

Coomassie blue staining
Coomassie blue staining was performed to determine the protein content after co-immunoprecipitation.The separated proteins on SDS-polyacrylamide gels were treated with adequate amounts of BeyoBlue™ Coomassie Blue Super Fast Staining Solution (Beyotime, cat.# P0017F) and incubated for 30 min at room temperature.Subsequently, deionised water was used to decolourise the gel background.

Detection of protein ubiquitination
Rapamycin-or MHY1485-treated cells were seeded in 100-mm Petri dishes and cultured overnight.The following day, the cells were incubated with 5-µM MG132 (MCE; cat.number: HY-13259) for 24 h.Immunoprecipitation was performed to enrich the ASNS protein.Cell lysates were collected for western blotting, and the protein was detected using the anti-ubiquitin antibody (Proteintech; cat.number: 10201-2-AP).

Metabolomic analysis
Liquid chromatography-tandem mass spectrometry (LC-MS/MS) was used to analyse changes in metabolites in cells with TLK2 knockdown.
Briefly, precooled methanol/acetonitrile/aqueous solution (2:2:1, v/v) was added to cells (approximately 2 × 10 6 cells per sample), and the samples were mixed on a vortex mixer.The samples were subjected to low-temperature ultrasonication for 30 min and centrifuged at 14,000 × g for 20 min.The supernatant was collected and dried in a vacuum centrifuge.For LC-MS/MS analysis, the samples were eluted in 100 μL of acetonitrile/water solvent (1:1, v/v).The samples were separated via ultra-high-pressure liquid chromatography (HILIC column, 1290 Infinity LC; Agilent, Santa Clara, CA, USA), and their primary and secondary spectra were recorded on a mass spectrometer (AB Triple TOF 6600).Raw MS data were processed using the XCMS software to analyse changes in intracellular metabolites.

Immunoprecipitation-tandem mass spectrometry
Immunoprecipitation-tandem mass spectrometry (IP-MS) was performed to identify the downstream proteins of TLK2.HGC27 cells were transduced with TLK2-OE lentiviruses.Proteins bound to TLK2 were pulled down using Protein A/G PLUS-Agarose (Santa Cruz Biotechnologies).A washing buffer (20-mM Tris-HCl, 150-mM KCl, 1-mM dithiothreitol, 0.05% Nonidet P [NP] 40, 1-mM EDTA, 15% glycerol and 0.2-mM PMSF) was prepared to rinse the protein samples.Finally, the samples were separated via SDS-PAGE and processed for MS analysis.

Cell proliferation assay
Transduced GC cells were seeded in 12-well plates (approximately 5 × 10 3 cells/ well).Beginning from the following day, the cells were counted every 24 h using an automatic cell counter for 6 days.Based on the cell counts cell growth curves were plotted using the GraphPad Prism (version 7.0) software (GraphPad Software, San Diego, CA, USA).

Colony formation assay
Approximately 500 transduced GC cells were seeded in 6-well plates and cultured at 37 °C for 2 weeks.The resulting cell colonies were fixed with 4% paraformaldehyde for 20 min.Subsequently, paraformaldehyde was discarded, and the colonies were stained with 1% crystal violet and counted using a digital camera.The experiment was performed in triplicate, and the Mann-Whitney U-test was used to estimate differences.

Transwell assay
For transwell invasion assay, a transwell chamber (Costar, 8-µM pore size) placed in a 24-well plate was used.Diluted Matrigel (BD Bioscience, San Jose, CA, USA) and transduced GC cells in 100 µL of a serum-free medium were added to the upper chamber, whereas 650 µL of RPMI-1640 medium containing 20% FBS was added to the lower chamber, which acted as a chemoattractant.After approximately 24 h, the chamber contents were fixed with 4% paraformaldehyde for 20 min and stained with 1% crystal violet for 15 min.Cells on the upper surface of the chamber were removed and counted under a microscope.For the transwell migration assay, chambers without Matrigel coating were used, and the other steps were similar to those of the abovementioned invasion assay.

Flow cytometry
To assess apoptosis, cells were harvested and stained with Annexin V-FITC and propidium iodide (PI) for 20 min in the dark.Apoptosis was detected on a Cytoflex flow cytometer (Beckman Coulter, Brea, CA, USA) according to the manufacturer's instructions.To assess the cell cycle, indicated cells were centrifuged and fixed with 75% alcohol overnight.The following day, the cells were centrifuged at room temperature, and the supernatant was removed.Thereafter, the cells were stained with 500-µL PI for 30 min, and changes in the cell cycle were detected on the Cytoflex flow cytometer.

5-ethynyl-2´-deoxyuridine assay
The BeyoClick™ EdU Cell Proliferation Kit (Beyotime; cat.number: C0078S) was used for 5-ethynyl-2´-deoxyuridine (EdU) staining.Briefly, transduced GC cells were seeded in 24-well plates and incubated overnight.The following day, the cells were incubated with 10-µM EdU at 37 °C and 5% CO 2 for 2 h.Subsequently, EdU-labelled cells were incubated with the Click Addictive Solution at room temperature for 30 min and stained with Hoechst 33342 (1:1,000) for 10 min.Finally, the cells were examined and imaged using a microscope.

Immunofluorescence double-staining
Immunofluorescence (IF) double-staining was performed to validate the co-localisation of TLK2 and ATF4 in GC cells.Briefly, GC cells were seeded in a 24-well plate with cell slides and incubated overnight.The following day, the cells were fixed with 4% formaldehyde, blocked with 5% bovine serum albumin at 26 °C for 1 h and incubated with mouse anti-TLK2 antibody (1:100, Santa Cruz Biotechnologies; cat.number: sc-393506) and rabbit anti-ATF4 antibody (1:100, cat.number: 381426; ZenBio, Durham, NA, USA) at 4 °C overnight.The following day, the cells were co-incubated with rabbit and mouse fluorescence-labelled secondary antibodies (1:250, cat.number: A11012; Thermo Fisher Scientific) for 1 h at room temperature and stained with DAPI (1:100, cat.number: D9542; Sigma-Aldrich, Taufkirchen, Germany) for 15 min.Finally, the cells were imaged using a confocal laser microscope (LSM800; Carl Zeiss, Jena, Germany).

TUNEL assay
Processed paraffin-embedded subcutaneous tumour tissues were analysed using a TUNEL kit (Beyotime, Shanghai, China; cat.number: C1098).TUNEL-stained cells were calculated in six random fields at 200fold magnification.

Tumour xenograft experiments
Animal experiments were conducted in accordance with the NIH Guidelines and were approved by Anhui Medical University.Four-weekold male nude mice (BALB/c; GemPharmatech Animal Center) were used for establishing xenograft models.The mice were randomly assigned to groups (n = 6/group).Transduced GC cells (5 × 10 6 cells /100 µL per mouse) were injected into the right flank of each mouse.The mice were monitored for health and tumour growth weekly, whereas tumour size and volume were measured every 4-5 days.The mice were sacrificed after the indicated time point, and tumour growth curves were plotted according to the measurements.Tumour tissues were harvested and fixed in 4% paraformaldehyde for western blotting, real-time reverse transcription PCR (qRT-PCR), IHC analysis and haematoxylin-eosin staining.

Statistical analysis
The GraphPad Prism (version 7.0), Image J and SPSS Statistics (version 22.0) software (SPSS, Inc.; Chicago, IL, USA) were used for statistical analyses.Continuous variables were expressed as the mean ± standard error.Student's t-test or one-way ANOVA was used for analysing differences in experimental data.Survival analysis was performed using the Kaplan-Meier method and Cox proportional hazards regression.A p-value of ≤0.05 indicated statistical significance (*p < 0.05; **p < 0.01; ***p < 0.001).

TLK2 expression is increased in GC and is associated with a worse prognosis
The UALCAN database was used to evaluate the expression of TLK2 in gastrointestinal tumours.The results showed that TLK2 expression was upregulated in various digestive system tumours, including GC (Fig. 1A).Western blotting revealed that TLK2 expression was significantly upregulated in GC cell lines (AGS, SGC7901, MGC803, BGC823 and HGC27) and tissues (Fig. 1B-D).
TMAs were subjected to IHC analysis to examine the relationship between TLK2 expression and tumour aggressiveness in GC.A total of 107 GC tissues and 22 randomly selected adjacent normal tissues were included in the final prognostic and correlation analyses.TLK2 expression was higher in GC tissues than in adjacent normal tissues, which was consistent with the results of western blotting (Fig. 1E).Kaplan-Meier analysis revealed that upregulated TLK2 expression was associated with shorter overall survival and larger tumour diameter in patients with GC (Fig. 1F-G).In addition, Kaplan-Meier curves were plotted to compare TLK2 expression between GC and adjacent normal tissues, which revealed similar results (Fig. S1A, B).Multivariate Cox regression analysis showed that TLK2 expression, lymph node status and tumour differentiation were independent prognostic factors for GC (Table S1).In addition, IF showed that TLK2 was localised in both the nucleus and cytoplasm (Fig. S1C).Altogether, these results suggest that TLK2 expression is upregulated in GC cells and tissues and high TLK2 expression is associated with a worse prognosis in GC.

TLK2 knockdown inhibits the colony-forming, migratory and invasive capabilities of GC cells in vitro
To examine the function of TLK2 in vitro, endogenous TLK2 was knocked down in HGC27 and AGS cells via lentivirus-mediated transduction of three shRNAs.The knockdown efficiency was verified via WB, and sh#2 and sh#3 were selected for further experiments owing to their higher inhibition rates.After the successful establishment of stable polyclonal TLK2-knockdown Fig. 1 TLK2 expression was upregulated in GC, and high TLK2 expression was associated with a worse prognosis.A Analysis of UALCAN data showed that TLK2 was upregulated in various digestive system tumours, including cholangiocarcinoma (CHOL), pancreatic carcinoma (PAAD) and GC.B, C Western blotting showed that the protein expression of TLK2 was significantly upregulated in GC cell lines.D Immunoblotting was performed to examine TLK2 levels in tumour tissues (T) and corresponding non tumour tissues (N) collected from 6 patients with GC.E Immunohistochemical detection of large-sample tissue microarray showed that TLK2 was significantly overexpressed in GC tissues.F Kaplan-Meier analysis suggested that TLK2 expression was significantly correlated with prognosis in GC.G Immunohistochemical scoring indicated that TLK2 expression was correlated with tumour size.cells, their colony-forming capability was examined.The results revealed that TLK2 knockdown significantly inhibited the colonyforming capability of GC cells (Fig. 2B, C).Transwell assay indicated that TLK2 knockdown markedly decreased the invasive and migratory capabilities of AGS and HGC27 cells (Fig. 2D-G).Furthermore, TLK2 was overexpressed in SGC7901 cells via lentivirus-mediated transduction (Fig. 2H).The effects of TLK2 overexpression were contradictory to those of TLK2 knockdown (Fig. 2I, L).Altogether, these results suggest that TLK2 knockdown strongly inhibits the colony-forming, migratory and invasive capabilities of GC cells.

TLK2 regulates the proliferation and apoptosis of GC cells in vitro and in vivo
EdU assay and flow cytometry (FCM) were performed in vitro to examine whether TLK2 regulated the proliferation and apoptosis of GC cells, respectively.EdU assay demonstrated that TLK2 knockdown markedly decreased the proliferation of GC cells (Fig. 3A, B), whereas TLK2 overexpression had contradictory effects (Fig. 3C).Subcutaneous tumorigenesis, IHC analysis and TUNEL assay were performed in vivo to assess the effects of TLK2 on the growth of GC cells.The results of subcutaneous tumour formation showed that TLK2 overexpression promoted the proliferation of GC cells (Fig. 3D, F).Ki-67 staining revealed that TLK2 overexpression significantly increased the proliferation of GC cells (Fig. 3G, H).FCM revealed that downregulated TLK2 expression promoted the apoptosis of GC cells (Fig. 3I).Consistently, the pro-apoptotic protein BAX was upregulated in TLK2-knockdown GC cells (Fig. S2).TUNEL staining revealed that TLK2 overexpression inhibited the apoptosis of GC cells (Fig. 3J).Although these results suggest that TLK2 promotes the progression of GC, the specific regulatory mechanism remains unclear.Therefore, IP-MS was performed to elucidate the mechanisms through which TLK2 regulates the progression of GC.Coomassie blue staining revealed that the protein content of immunoprecipitated samples was higher in the anti-TLK2 group than in the IgG group (Fig. 3K).More than 1,400 proteins were identified after immunoprecipitation of samples with the anti-TLK2 monoclonal antibody.Gene Ontology (GO) enrichment analysis revealed that the downstream proteins of TLK2 were most Fig. 2 TLK2 knockdown inhibited the colony-forming, migratory and invasive capabilities of GC cells in vitro.A Western blotting was used to verify the efficiency of TLK2 knockdown; sh#2 and sh#3 exhibited better knockdown efficiency.B, C The colony-forming capability of GC cells was significantly inhibited after TLK2 knockdown.D-G Transwell assay indicated that TLK2 knockdown markedly decreased the invasive and migratory capabilities of AGS and HGC27 cells.H Western blotting was performed to verify the efficiency of TLK2 overexpression.I, J The colony-forming capability of GC cells was enhanced after TLK2 overexpression.K, L The invasive and migratory capabilities of AGS and HGC27 cells were enhanced after TLK2 overexpression.
enriched in metabolic pathways (Fig. 3L).Altogether, these results suggest that TLK2 regulates the proliferation and apoptosis of GC cells through metabolic pathways.

mTORC1 and ASNS are intermediate molecules of TLK2 regulating amino acid metabolism
To identify the metabolites through which TLK2 promotes GC progression, metabolomic analysis was performed after TLK2 knockdown.A total of 142 metabolites were downregulated and 62 metabolites were upregulated in TLK2-knockdown HGC27 cells compared with control cells (Fig. 4A and S3).These metabolites included amino acids, purines and pyrimidines.TLK2 knockdown exerted the strongest effect on amino acid metabolism, including L-aspartic acid metabolism (Fig. 4B).The level of L-aspartic acid was markedly affected by ASNS (Fig. 4C) [29].Pathway enrichment analysis of metabolomic data revealed that TLK2 knockdown significantly affected the mTOR and phosphoinositide 3-kinase (PI3K)/AKT signalling pathways (Fig. 4D).To verify these results, Fig. 3 TLK2 regulated the proliferation and apoptosis of GC cells in vitro and in vivo through a metabolic pathway.A, B EdU assay demonstrated that TLK2 knockdown markedly decreased the proliferation of AGS and HGC27 cells.C Overexpression of TLK2 increased the proliferation of SGC7901 cells.D Subcutaneous tumorigenesis experiments after TLK2 overexpression in SGC7901 cells.E, F Tumour weight and growth curves suggested that TLK2 overexpression promoted the proliferation of GC cells in vivo.G Western blotting was performed to verify the efficiency of TLK2 overexpression in vivo.H Subcutaneous tumour tissues harvested from mice were stained with Ki-67.I Assessment of apoptosis in AGS and HGC27 cells transduced with shNC and shTLK2 lentiviruses; the proportion of apoptotic cells was increased after TLK2 knockdown.J Haematoxylin-eosin staining was performed to assess the morphology of tumour tissues, and TUNEL staining demonstrated that TLK2 expression was higher in the NC group than in the TLK2-overexpression group in vivo.K Coomassie blue staining was performed after IP using the anti-TLK2 antibody; the protein content was found to be higher in the IP group than in the IgG group.L IP-MS revealed that the downstream proteins of TLK2 were most enriched in metabolic pathways.
Fig. 4 mTORC1 and ASNS were intermediate molecules of TLK2 regulating amino acid metabolism.A Volcano plot of metabolites; 142 metabolites were downregulated and 62 metabolites were upregulated following TLK2 knockdown in HGC27 cells.B Heatmap of metabolites; TLK2 knockdown downregulated amino acid metabolism, including L-aspartic acid metabolism.C ASNS participates in L-aspartic acid metabolism.D Bubble diagram of metabolites; TLK2 knockdown significantly affected amino acid metabolism, purine metabolism, pyrimidine metabolism and the mTOR and PI3K/AKT signalling pathways.E According to the IP-MS data of TLK2, the mTOR protein peptide was found in the immunoprecipitation lysate.F Immunoblotting suggested that the levels of p-P70S6K and p-4EBP1 were decreased and increased after TLK2 knockdown and overexpression, respectively.G Immunoblotting suggested that the expression of ASNS was decreased and increased after TLK2 knockdown and overexpression, respectively.
the IP-MS data of TLK2 were re-analysed, which revealed the presence of the mTOR protein peptide in immunoprecipitation lysates (Fig. 4E).TLK2 knockdown decreased the levels of p-P70S6K and p-4EBP1, suggesting the inhibition of the mTORC1 pathway.However, TLK2 overexpression exerted contradictory effects on the mTORC1 pathway (Fig. 4F).The protein expression of ASNS was evaluated in cells with TLK2 knockdown and overexpression (Fig. 4G), and the results revealed that TLK2 directly regulated the protein expression of ASNS.Altogether, these results suggest that TLK2 regulates amino acid metabolism by regulating mTORC1 and ASNS.mTORC1 interacts with ASNS and inhibits its ubiquitinmediated degradation The abovementioned results indicate that TLK2 regulates the protein expression of ASNS; however, the two proteins do not bind to each other.We speculated that TLK2 regulates ASNS through other molecules.To verify this hypothesis, IP-MS was performed again using the anti-ASNS antibody (Fig. 5A), and the results revealed the presence of mTORC1, RAPTOR and mTOR in immunoprecipitation lysates (Fig. 5B).As mentioned earlier, mTORC1 consists of five components, namely, mTOR, RAPTOR, DEPTOR, mLST8, and PRAS40 (Fig. S4).Therefore, we speculated that TLK2 regulates ASNS levels through mTORC1.To verify this hypothesis, co-immunoprecipitation was performed using the anti-ASNS and anti-mTOR antibodies.The results revealed that mTOR interacted with ASNS (Fig. 4C, D).Subsequently, we examined the physiological significance of the binding between mTOR and ASNS.Inhibition of mTORC1 in rapamycin-treated AGS and HGC27 cells strongly impaired the expression of ASNS (Fig. 5E).Moreover, it decreased the protein expression of ASNS but did not affect its mRNA expression (Fig. 5E).Therefore, we performed CHX chase assay to determine the half-life of the ASNS protein with or without mTORC1 inhibition.Western blotting indicated that degradation of the ASNS protein was accelerated in HGC27 cells with mTORC1 inhibition compared with control cells (Fig. 5F, G).The proteasome inhibitor MG132 prevented the mTORC1-mediated degradation of ASNS; however, the autophagy inhibitor 3-MA and the autolysosome inhibitor bafilomycin A1 did not prevent ASNS degradation (Fig. 5H).The optimal concentration of MG132 required for inhibiting the ASNS protein was determined to be 5 nM (Fig. 5I).Altogether, these results suggest that mTORC1 interacts with ASNS and promotes its ubiquitinmediated degradation in GC.

ATF4 and mTORC1 participate in the dual regulating effects of TLK2-mediated regulation of ASNS
To verify that mTORC1 inhibition enhances the ubiquitinmediated degradation of ASNS, rapamycin or MHY1485 was used to inhibit or activate the mTORC1 pathway, respectively [30].The results showed that inhibiting and activating the mTORC1 pathway downregulated and upregulated the protein expression of ASNS, respectively (Fig. 6A, B).To assess whether TLK2 was involved in mTORC1-mediated regulation of ASNS, HGC27 cells were co-treated with rapamycin and TLK2overexpression lentiviruses or MHY1485 and TLK2-knockdown lentiviruses.TLK2 overexpression did not increase the protein expression of ASNS in rapamycin-treated cells (Fig. 6C, E).Similarly, MYH1485 reversed the TLK2 knockdown-induced reduction in the protein expression of ASNS (Fig. 6D, F).In addition, TLK2 knockdown, but not rapamycin treatment, affected the mRNA expression of ASNS (Fig. 6G).Therefore, we speculated that TLK2 regulates the translation and transcription of ASNS through mTORC1 and other molecules, respectively.ATF4 has been identified as a tumour-associated transcription factor for ASNS [31,32].IF double-staining suggested that TLK2 and ATF4 proteins were co-localised in AGS cells (Fig. 6H).Co-immunoprecipitation performed using the anti-TLK2 and anti-ATF4 antibodies revealed that TLK2 could bind to the ATF4 protein (Fig. 6I, J).In addition, TLK2 was found to affect the expression of ATF4 (Fig. 6K).Altogether, these results suggest that TLK2 maintains the stability of the ASNS protein through mTORC1 and increases ASNS transcription through ATF4.
Fostamatinib inhibits the proliferative, invasive and migratory capabilities of GC cells by targeting TLK2 Fostamatinib has been demonstrated to be effective in treating immune thrombocytopenia in phase III clinical trials [33].A recent study reported that fostamatinib can inhibit the progression of prostate cancer by targeting PKMYT1, suggesting that fostamatinib can be used in the treatment of malignant tumours [34].Similarly, another study reported that fostamatinib can inhibit the expression of oncoproteins in GC [35].Using the Drugbank database, we identified fostamatinib as a potential drug targeting TLK2.p-AKT was used to determine the appropriate dose of fostamatinib, and the lowest effective concentration was found to be 5 uM (Fig. S5).Fostamatinib downregulated the serine phosphorylation of TLK2 in GC cells (Fig. 7A).In addition, it inhibited the growth (Fig. 7B, C) and invasive and migratory capabilities (Fig. 7D-G) of GC cells.Consistently, EdU assay revealed that fostamatinib significantly reduced the proliferation of GC cells (Fig. 7H, I).Subsequently, the changes in GC cells subcutaneous tumour formation were observed after intraperitoneal injection of Fostamatinib, the results showed that Fostamatinib could significantly inhibit the growth of GC cells in vivo (Fig. 7J-K).Altogether, these results suggest that fostamatinib inhibits the proliferative, invasive and migratory capabilities of GC cells by targeting TLK2.

DISCUSSION
This study revealed the mechanism underlying TLK2-mediated reprogramming of amino acid metabolism in GC (Fig. 8).The findings of this study are as follows: (i) TLK2 expression is upregulated in GC cells and tissues, and high TLK2 expression is associated with the poor prognosis of patients with GC.Downregulated TLK2 expression strongly inhibits GC aggressiveness both in vitro and in vivo.(ii) TLK2 participates in the reprogramming of amino acid metabolism mainly by regulating the mTORC1 pathway and ASNS expression.(iii) mTORC1 interacts with ASNS and inhibits its ubiquitin-mediated degradation.(iv) The regulatory relationship between TLK2 and ASNS is influenced by mTORC1.(v) TLK2 regulates the protein and mRNA expression of ASNS through mTORC1 and ATF4, respectively.(vi) The small-molecule drug fostamatinib inhibits the serine phosphorylation of TLK2, thereby inhibiting the progression of GC.To the best of our knowledge, this study is the first to demonstrate the functional significance of the fostamatinib/TLK2/mTORC1/ASNS axis in GC progression.Therefore, the results of this study may help to develop novel strategies for targeted therapy of GC.
Many kinases and related pathways have been identified to perform crucial functions in GC, including the SRC kinase family [36], CDKs [37] and the PI3K/AKT/mTOR pathway [38].However, the effects of the TLK family in GC remain unclear.A pilot study suggested that TLK2 is abnormally expressed in breast cancer [14], whereas the downstream targets and action mechanisms of TLK2 in GC carcinogenesis were not affected.In this study, TLK2 was found to promote the growth and proliferation of GC cells in vitro and in vivo.Multivariate analysis revealed that TLK2 expression was an independent indicator of prognosis in GC.These results suggest that TLK2 functions as an oncogene in GC and is a promising therapeutic target for GC.
Energy metabolism reprogramming is considered a hallmark of cancer [39].Amino acid metabolism has attracted considerable attention in recent years [40].mTORC1 serves as a core protein that regulates amino acid metabolism [20].In this study, we identified a regulatory relationship between TLK2 and mTORC1.The results of IP-MS, co-immunoprecipitation and protein stability assessment validated that mTOR interacted with ASNS, a key enzyme in amino acid metabolism, and promoted its degradation.Further experiments were conducted to elucidate post-translational modifications of ASNS, and ASNS was identified as a novel substrate of mTORC1.In addition, TLK2 knockdown attenuated the promoting effects of mTORC1 on the stability of ASNS protein.Altogether, these findings reveal the crucial role of the TLK2-mTORC1-ASNS axis in GC and may facilitate the development of novel therapeutic strategies for GC.Fig. 5 mTORC1 interacted with ASNS and inhibited its ubiquitin-mediated degradation.A Coomassie blue staining after IP using the anti-ASNS antibody; the protein content was higher in the IP group than in the IgG group.B The protein components of mTORC1, RPTOR and mTOR were found in ASNS immunoprecipitation lysate.C, D ASNS and mTOR proteins were immunoprecipitated using anti-mTOR and anti-ASNS antibodies, respectively.E The mTORC1 inhibitor rapamycin strongly impaired the expression of ASNS in AGS and HGC27 cells.F, G CHX chase assay suggested that mTORC1 inhibition decreased the protein expression of ASNS.H mTORC1-mediated degradation of ASNS was prevented by MG132 but not by 3-methyladenine (MA) or bafilomycin A1 (BafA).I The optimal concentration of MG132 required for inhibiting the ASNS protein was 5 nM, and higher concentrations did not increase the levels of ASNS.
Numerous kinases function as oncoproteins in GC; however, not all kinases can serve as therapeutic targets.In this study, fostamatinib, a kinase inhibitor, was found to inhibit GC progression by blocking TLK2 phosphorylation.Therefore, TLK2 may serve as an active therapeutic target for GC.Furthermore, the results of this study provide a valuable basis for further research on tumour metabolism.Tumour-associated kinases are closely related to tumour cell metabolism, including amino acid and nucleotide metabolism.We have previously demonstrated that U2AF homology motif kinase 1, a ubiquitously expressed nuclear serine/threonine kinase, promotes GC progression by enhancing de novo purine synthesis [41].CDC-like kinase 3 plays a similar role in cholangiocarcinoma cells [42].In addition, well-known oncogenic kinases such as PI3K/AKT are related to metabolic gene expression and metabolic enzyme activities [43].However, relevant studies on GC are limited.Future studies should examine the potential relationship between tumour-associated kinases and metabolism in GC.
In conclusion, this study reveals that TLK2 contributes to the reprogramming of amino acid metabolism by regulating the Fig. 6 ATF4 and mTORC1 participate in the dual regulating effects of TLK2-mediated regulation of ASNS.A Rapamycin, an inhibitor of mTORC1, upregulated the ubiquitination of ASNS in HGC27 cells.B MHY1485, an activator of mTORC1, decreased the ubiquitination of ASNS in HGC27 cells.C, E In HGC27 cells co-treated with rapamycin and TLK2-overexpression lentivirus, TLK2 overexpression failed to increase the protein expression of ASNS.D, F In HGC27 cells co-treated with MHY1485 and TLK2-knockdown lentiviruses, MYH1485 reversed the TLK2 knockdown-induced reduction in the protein expression of ASNS.G TLK2 knockdown, but not rapamycin treatment, affected the mRNA expression of ASNS.H IF double-staining suggested that TLK2 and ATF4 proteins were co-localised in AGS cells.I-J Co-immunoprecipitation performed using the anti-TLK2 and anti-ATF4 antibodies revealed that TLK2 could bind to the ATF4 protein.K TLK2 was found to affect the expression of ATF4.Fig. 7 Fostamatinib inhibited the proliferative, invasive and migratory capabilities of GC cells by targeting TLK2.A IP revealed that serine phosphorylation of TLK2 was downregulated in fostamatinib-treated GC cells.B, C Fostamatinib significantly inhibited the growth of GC cells.D-G Transwell assay revealed that fostamatinib decreased the invasive and migratory capabilities of GC cells.H, I EdU assay revealed that fostamatinib significantly inhibited the proliferation of GC cells.J, K Fostamatinib could significantly inhibit the growth of GC cells in vivo.
interaction between mTORC1 and ASNS in GC.In addition, fostamatinib can inhibit the proliferation of GC cells by targeting TLK2.These findings strongly support the potential of TLK2 as a new therapeutic target for GC and may help to develop effective therapeutic strategies for GC.
Fig. 8 TLK2 promoted the binding between mTORC1 and ASNS, thereby inhibiting the degradation and enhancing the stability of the ASNS protein.However, TLK2 upregulated the transcription of ASNS by binding to ATF4.Consequently, amino acid metabolism was upregulated, and GC progression was promoted.