Skip to main content
NCBI home page
Cell Communication and Signaling : CCS logoLink to Cell Communication and Signaling : CCS
. 2025 Aug 28;23:384. doi: 10.1186/s12964-025-02374-x

Whole-gene CRISPR/cas9 library screen revealed targeting STAT6 increased the sensitivity of liver cancer to celecoxib via inhibiting arachidonic acid shunting

Chujiao Hu 1,2,3,4,#, Zhirui Zeng 4,#, Xin Bao 4,#, Dahuan Li 4, Huading Tai 4, Haohao Zeng 4, Cheng Luo 3,5,6, Lei Tang 1,2,✉, Tengxiang Chen 4,8,✉, Shi Zuo 7,8,✉
PMCID: PMC12392558  PMID: 40877941

Abstract

Celecoxib, a selective COX-2 inhibitor, has demonstrated anti-liver cancer effects in various preclinical models and clinical traits. However, prolonged use of celecoxib can lead to drug resistance, necessitating higher doses to maintain efficacy, which often results in severe side effects, limiting its clinical application. This study aimed to identify strategies to overcome celecoxib resistance in liver cancer. CRISPR/Cas9 screening revealed that liver cancer cells compensated for celecoxib treatment by upregulating ALOX and CYP enzymes, facilitating AA metabolism to produce alternative downstream products. STAT6 was identified as a key regulator of ALOX15, ALOX12, and CYP2E1, acting as a resister to celecoxib. Celecoxib stimulation leaded to increased phosphorylation of STAT6, enhanced binding to the promoters of target genes such as ALOX15, and upregulation of downstream gene expression. Knockdown of STAT6 significantly enhanced celecoxib sensitivity in vitro and in vivo by blocking AA shunting mediated by these enzymes. Furthermore, AS1517499, a STAT6 inhibitor, showed strong synergy with celecoxib in liver cancer cells by inhibiting AA shunting. In conclusion, targeting STAT6 enhances the efficacy of celecoxib in liver cancer by suppressing AA shunting. The combination of AS1517499 and celecoxib holds promise as a novel therapeutic strategy for liver cancer.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-025-02374-x.

Keywords: Liver cancer, Celecoxib, AA shunting, STAT6

Introduction

According to global Cancer Statistics 2022, liver cancer is the sixth most frequent cancer in the world, ranking fifth in incidence and third in mortality. There are more than 900,000 new liver cancer patients in the world every year [1]. Current treatment options for liver cancer patients include surgical resection, orthotopic liver transplantation, radiofrequency thermal ablation, and targeted therapy [2, 3]. In the past decade, numerous immunotherapy drugs and targeted therapy drugs have been gradually tried to apply to the first-line treatment of unresectable liver cancer, such as brivanib, sunitinib, linifanib, sorafenib, sorafenib plus doxorubicin, sorafenib plus erlotinib, lenvatinib, nivolumab, donafenib, and atezolizumab plus bevacizumab. The response rates of systemic therapy, such as sorafenib and lenvatinib, are low, with a high risk of adverse events [4, 5]. Therefore, finding a new therapy for the treatment of liver cancer is urgent.

Chronic inflammatory stimuli that drives the infiltration of immune cells into tissues are well-recognized as key risk factors in the initiation and progression of cancer. Chronic inflammation serves as a critical driver of neoplastic changes, contributing to dysplasia and significantly increasing the risk of developing liver cancer [6]. Arachidonic acid (AA) is a pivotal precursor of potent inflammatory mediators, including prostaglandins and leukotrienes [7]. The metabolism of AA is regulated by three key enzyme families—cyclooxygenases (COXs), lipoxygenases (LOXs), and cytochrome P450 (CYP) enzymes—which act as rate-limiting steps in this pathway [8]. Elevated levels of COX enzymes, particularly COX-2, have been observed in liver cancer tissues and are associated with poor prognosis in liver cancer patients [9, 10]. Thus, targeting AA metabolism has emerged as a potential strategy for inhibiting liver cancer progression.

Nonsteroidal anti-inflammatory drugs (NSAIDs) suppress inflammatory responses by inhibiting the cyclooxygenase pathway of AA metabolism. NSAIDs include traditional nonselective agents as well as selective COX-2 inhibitors [11]. Celecoxib, a selective COX-2 inhibitor with anti-inflammatory properties, has garnered attention for its potential role in cancer therapy [12]. Recent studies have demonstrated significant inhibitory effects of COX-2 inhibitors on liver cancer in preclinical models. Moreover, clinical trials have shown that celecoxib, whether used alone or in combination with chemotherapy, improves overall survival rates in cancer patients, including those with liver cancer [13, 14]. However, clinical evidence also suggests that while low doses of celecoxib initially exhibit significant antitumor effects, prolonged use requires dose escalation to maintain efficacy. This gradual increase in dosage often leads to severe gastrointestinal and cardiac side effects, limiting its long-term clinical application [15, 16]. Park [17] and Ganesh et al. [18] have respectively found in head and neck tumors and colorectal cancer that the expression of ALOXs or CYPs is elevated in tissues from patients resistant to celecoxib. These upregulated enzymes can synthesize similar products that compensate for COX inhibition and promote cell proliferation—a phenomenon known as AA shunting. However, some researchers have reported that AA shunting is not necessarily induced by the use of celecoxib in all cases [19]. Currently, it remains unclear whether AA shunting exhibits tissue specificity or what factors regulate AA shunting during celecoxib treatment.

In this study, we demonstrated that AA shunting contributes to the increased resistance of liver cancer cells to celecoxib. Despite the inhibition of COX-2 activity, liver cancer cells compensate by upregulating various ALOX and CYP enzymes, enabling the continued utilization of AA to synthesize alternative downstream products. We identified STAT6 as a key transcription factor that concurrently regulates ALOX15, ALOX12, and CYP2E1. Targeting STAT6 through gene knockdown or the inhibitor AS1517499 effectively suppressed AA shunting and significantly enhanced the sensitivity of liver cancer cells to celecoxib. These findings suggested that combining AS1517499 with celecoxib could serve as a promising therapeutic approach for liver cancer.

Materials and methods

Cell culture

Human normal hepatocyte THLE-2 cells and liver cancer cell lines (HepG2 and JHH7) were obtained from Procell (https://www.procell.com.cn/; Wuhan, China). THLE-2 cells were maintained in RPMI-1640 medium (Servicebio, Wuhan, China) supplemented with 10% fetal bovine serum (FBS; Hyclone, USA) and 1% penicillin (Servicebio, Wuhan, China). Liver cancer cells (HepG2 and JHH7) were cultured in DMEM medium (Servicebio, Wuhan, China) containing 10% FBS and 1% penicillin. All cell lines were incubated under standard conditions at 37 °C in a humidified atmosphere with 5% CO₂.

In vivo whole-gene CRISPR/cas9 library screen

Stable Cas9-expressing HCC cell lines (HepG2 and JHH7) were generated through lentiviral transduction of a Cas9 coding sequence containing a Blasticidin resistance gene. The Cas9 expression in HepG2 and JHH7 cells were confirmed by western blotting. The functionality of the Cas9 protein was validated by transfecting cells with CTNNB1 sgRNA (sequence: GATGGAGTTGGACATGGCCA). HepG2-Cas9 and JHH7-Cas9 cells were subsequently transduced with the TKO v3 library, comprising 70,948 gRNAs targeting 18,053 protein-coding genes (4 gRNAs/gene) and 142 control gRNAs (targeting EGFP, LacZ, and luciferase) [20], at a low concentration of infection (MOI ~ 0.3) to ensure effective barcoding. Transduced cells were then selected with 1.0 µg/mL puromycin for 7 days in order to create a mutant cell pool.

Following the expansion of cells transduced with the sgRNA library, one-third of the cells were harvested for sgRNA sequencing to establish baseline expression levels. The remaining cells were implanted into the left axilla of nude mice at a density of 2 × 10⁶ cells per mice. The mice were randomly divided into two groups (n = 20 per group): the treatment group received 20 mg/kg celecoxib (dissolved in DMSO) intravenously every three days for a total of seven doses, while the control group received an equivalent dose of DMSO. Tumor tissues were collected from both groups, lysed, and subjected to sgRNA sequencing. The DrugZ algorithm was used to analyze celecoxib resisters, while knockout of them increased cell sensitivity to celecoxib, using cut-off as P < 0.05.

Cell transfection

Short hairpin RNA (shRNA) oligonucleotide sequences of targets were cloned into the pSuper-retro-puro vector and sourced from GeneChem (Shanghai, China). STAT6 overexpression plasmids were obtained from GeneCopoeia (Guangzhou, China). Liver cancer cells were transfected with shRNAs or plasmids using Polybrene and Lipo2000 transfection reagents (GeneCopoeia, Guangzhou, China). To establish stable knockdown cell lines, the transfected cells were treated with 1.0 µg/mL puromycin for 7 days 48 h post-transfection. shRNA sequences used in the current study were exhibited supplement Table 1.

Metabolite detection

HepG2 and JHH7 cells subjected to specific treatments were collected and resuspended in 100 µL of pre-chilled PBS containing 1 µL of protease inhibitor (Sangon Biotech, Shanghai). Cells were lysed via ultrasonication, and the lysates were centrifuged at 1,000 × g for 20 min at 2–8 °C to remove insoluble impurities and cell debris. The supernatant was then analyzed for PGE2, 12-HETE, and 14,15-EET levels using the Prostaglandin E2 ELISA Kit (cat. no. ab287802, Abcam, USA), the 12(S)-HETE ELISA Kit (cat. no. ab133034, Abcam, USA), and the 14,15-EET ELISA Kit (cat. no. ab175812, Abcam, USA), respectively. The relative levels of these metabolites were normalized to the protein concentration.

Western blotting

Cellular proteins were isolated using RIPA lysis buffer (Sangon Biotech, Shanghai, China) containing 10% PMSF (Sangon Biotech, Shanghai, China). Protein concentrations were measured using the BCA assay. Proteins were separated on a 10% SDS-PAGE gel (Sangon Biotech, Shanghai, China) and transferred to PVDF membranes (Millipore, USA) at a constant current of 310 mA. Membranes were blocked with skim milk and incubated overnight at 4 °C with primary antibodies, including Cas9 (1:3000; cat. no. 26758-1-AP, Proteintech), STAT6 (1:5000; cat. no. 82630-1-RR, Proteintech), p-STAT6 (1:1000; cat no. #56554, CST, USA), ALOX5 (1:1000; cat no. 10021-1-Ig, Proteintech), ALOX15 (1:1000; cat. no. A6864, Abconal), ALOX12 (1:1000; cat. no. A14703, Abconal), ALOXE3 (1:2000; cat no. PA5-21833, Thermo Fisher Inc., USA), CYP2A7 (1:1000; cat no. PA5-75499, Thermo Fisher Inc.), CYP2E1 (1:1000; cat. no. A2160, Abconal), CYP2F1 (1:1000; cat no. PA5-76223, Thermo Fisher Inc.), and β-actin (1:10000; cat. no. AC026, Abconal). Afterward, the membranes were washed with TBST (Tris-buffered saline with Tween-20) and incubated with secondary antibodies. Protein bands were visualized using an enhanced chemiluminescence (Millipore, USA) reagent.

qRT-PCR

Total RNA was extracted from HepG2 and JHH7 cells using the TRIzol reagent (Bao Bio) following the manufacturer’s protocol. Complementary DNA (cDNA) was synthesized using HiScript III RT SuperMix (Vazyme, Nanjing, China). Quantitative PCR (qPCR) was performed using Universal SYBR qPCR Premix (Vazyme, Shanghai, China), and the relative mRNA levels were normalized to ACTB. The primer sequences used in this study were exhibited in supplemental Table 2.

Immunofluorescence

First, HEPG2 and JHH7 cells are cultured to the logarithmic phase. After removing the culture medium, the cells are washed twice with PBS to ensure the removal of any residual culture medium. Next, the cells are fixed with 4% paraformaldehyde at room temperature for 10 min, followed by three washes with PBS. Then, the cells are permeabilized with PBS containing 1% Triton X-100 and incubated at room temperature for 8 min. After permeabilization, the cells are washed three times with PBS. To block non-specific binding, the cells are incubated with 5% BSA blocking solution at room temperature for 30 min. Subsequently, specific primary antibodies targeting STAT6 (1:100; cat. no. 82630-1-RR, Proteintech) and p-STAT6 (1:100; cat no. #56554, CST, USA) are added and incubated at 4 °C overnight. The next day, the cells are washed three times with PBS, and a CY3 labeled secondary antibody is added and incubated at room temperature for 1 h. Finally, the nuclei are stained with DAPI, and the cells are observed under a fluorescence microscope to assess the expression of STAT6 and p-STAT6.

Bioinformatics analysis

Transcription factor enrichment analysis for ALOX and CYP genes associated with celecoxib sensitivity was conducted using Metascape (http://metascape.org/gp/). The STAT6 motif was examined through the Human TFDB database (http://bioinfo.life.hust.edu.cn/HumanTFDB#!/), and binding sites of STAT6 within the ALOX15, ALOX12, and CYP2E1 promoter regions were identified with a significance threshold of q-value < 0.05.

ChIP-qPCR

The ChIP experiments were performed using the SimpleChIP Kit (Cell Signaling Technology, USA). STAT6-overexpressing HepG2 and JHH7 cells (2 × 10⁷) were fixed with 1% formaldehyde, and the crosslinking reaction was quenched with 0.1 M glycine. Chromatin was sheared into 200–800 bp fragments via ultrasonication and immunoprecipitated using either an anti-STAT6 antibody (1:50; Cat. no. ab32520, Abcam, USA) or a control anti-IgG antibody. The co-precipitated DNA was purified with a phenol/chloroform extraction method and analyzed by qPCR.

Double luciferase reporting experiment

Wild-type (WT) and mutant (MUT) sequences of STAT6 binding sites in the ALOX15, ALOX12, or CYP2E1 promoter regions were cloned into the pGL4.20 luciferase reporter vector. HepG2 and JHH7 cells, either NC or overexpressing STAT6, were seeded into 96-well plates at a density of 3,000 cells per well. The cells were transfected with 100 ng of luciferase reporter plasmids containing WT or MUT sequences. After 24 h, firefly luciferase activity was quantified and normalized to Renilla luciferase activity.

CCK-8 assay

THLE-2, HepG2, and JHH7 cells were seeded into 96-well plates at a density of 5,000 cells per well and incubated overnight. The following day, the medium was replaced with fresh medium containing the specified treatments. After 48 h of incubation, medium supplemented with 10% CCK-8 reagent was added to each well. Absorbance at 450 nm was measured using a microplate reader. Cell viability curves were generated, and IC₅₀ values were calculated using GraphPad Prism software (version 9).

5-ethynyl-2’-deoxyuridine (EDU) assay

The EDU assay was performed using the BeyoClick™ EDU-555 Cell Proliferation Detection Kit (Beyotime Biotechnology, Nanjing, China) following the manufacturer’s instructions. HepG2 and JHH7 cells, cultured in confocal dishes and subjected to specific treatments, were incubated with the EDU reagent for 2 h. After rinsing with PBS, the cells were fixed with 4% paraformaldehyde for 30 min at room temperature. Apollo staining reaction solution was then applied for 30 min to visualize EDU-positive cells, followed by nuclear staining with DAPI. Immunofluorescence signals were observed at 555 nm using a fluorescence microscope (CKX53, Olympus, Japan).

3D spheroid formation assay

Cells in the logarithmic growth phase were digested and resuspended in complete DMEM medium. A total of 2,000 cells per well were seeded into U-bottom 96-well plates with ultra-low adhesion (Corning, USA) and incubated at 37 °C for 21 days under specific treatments. Images were captured daily to monitor 3D spheroid formation.

Subcutaneous graft tumor model

For the subcutaneous tumor xenograft model, 2 × 10⁶ HepG2 cells were suspended in PBS and subcutaneously injected into the left axilla of 6- to 8-week-old male BALB/c nude mice. Tumor volumes were measured using vernier calipers and calculated using the formula: tumor volume (mm³) = length (mm) × width (mm) × height (mm) × 0.52. Once the tumor volumes reached 40–50 mm³ on day 9, the mice were randomized into treatment groups (n = 5 per group) and subsequently administered 20 mg/kg celecoxib [21, 22], 5 mg/kg AS1517499 [23], or a combination of both drugs via tail vein injection every three days. On day 24, the mice were euthanized, and the tumor tissues were harvested for IHC staining.

Immunohistochemistry (IHC)

Tumor tissues were fixed in 10% neutral buffered formalin, processed through a graded ethanol series for dehydration, and embedded in paraffin. Paraffin-embedded sections were prepared for IHC analysis. Sections were deparaffinized with xylene, rehydrated through descending ethanol concentrations, and rinsed with phosphate-buffered saline. Antigen retrieval was performed, followed by blocking of endogenous peroxidase activity using a hydrogen peroxide (H₂O₂) solution in Tris-buffered saline. The sections were then incubated overnight at 4 °C with the appropriate primary antibody including COX-2 (1: 200; cat no. 27308-1-AP, Proteintech, Wuhan, China), STAT6 (1:1000; cat no. 82630-1-RR, Proteintech), ALOX15 (1:200; cat no. A6864, Abconal, Wuhan, China), ALOX12 (1:200; cat no. A14703, Abconal), CYP2E1 (1:200, cat no. A22629, Abconal), PCNA (1:10000; cat no. 10205-2-AP, Proteintech), and KI67 (1:4000; cat no. 27309-1-AP, Proteintech). The following day, sections were brought to room temperature, incubated with a secondary antibody for 30 min, and visualized using diaminobenzidine staining. Counterstaining was performed with hematoxylin, and the stained sections were imaged and analyzed quantitatively using ImageJ software (version: 1.8.0).

Data statistics

Statistical analysis was performed using SPSS software (version 20.0), with results presented as mean ± standard deviation. Group comparisons were analyzed using a two-tailed unpaired Student’s t-test, while differences among multiple groups were assessed through one-way analysis of variance (ANOVA). A P-value < 0.05 was considered statistically significant. The combination index (CI) for STAT6 inhibitors and celecoxib was calculated using CompuSyn software (version 1.0.1). CI values of 1.1–0.9, 0.9–0.6, 0.6–0.3, and < 0.3 were interpreted as additive effects, weak synergy, moderate synergy, and strong synergy, respectively.

Results

Establish of Cas9 system in liver cancer cell

To identify genes associated with celecoxib sensitivity, the CRISPR/Cas9 system was utilized. Specifically, a Cas9 enzyme was stably expressed in HepG2 and JHH7 cells for CRISPR interference (CRISPRi) (Fig. 1A, B). To ensure the reliability of subsequent screening results, CCK-8 assays (Fig. 1C) and 3D sphere-forming experiments (Fig. 1D-E) were conducted. These experiments demonstrated no significant differences in celecoxib sensitivity between negative control (NC) cells and Cas9-overexpressing cells, indicating that transfection of the Cas9 system does not alter the intrinsic sensitivity of liver cancer cells to celecoxib.

Fig. 1.

Fig. 1

Open in a new tab

Establish of Cas9 system in liver cancer cell. A Western blotting verified the transfection of Cas9 enzyme. B Model of Cas9. C CCK-8 assays indicated that Cas9 loading did not affect the susceptibility of HepG2 and JHH7 cells to celecoxib. D, E The 3D spheroid formation assay indicates that the introduction of Cas9 does not affect celecoxib’s inhibitory effect on the spheroid-forming ability of HepG2 and JHH7 cells. F CTNNB1-targeting sgRNAs significantly reduced CTNNB1 expression by approximately two orders of magnitude in HepG2 and JHH7 loading Cas9. **, P < 0.01; ns, no significant

Additionally, to evaluate the functional efficiency of the Cas9 system, sgRNAs targeting the highly expressed gene CTNNB1 were employed. Results revealed that transfection with CTNNB1-targeting sgRNAs significantly reduced CTNNB1 expression by approximately two orders of magnitude in liver cancer cells (Fig. 1F), confirming the robust activity of the Cas9 system in this experimental setup.

In vivo CRISPR/Cas9 library screen indicated that AA shunting may induce the resistance of liver cancer cell to celecoxib

The whole-genome sgRNA library TKOv3 was transfected into the aforementioned HepG2 and JHH7 cells, which were subsequently injected into the left flank of mice to perform an in vivo CRISPR/Cas9 library screen (Fig. 2A). During the analysis of the in vivo CRISPR/Cas9 library screen, the distribution of sgRNA read counts was found to be uniform in both HepG2 and JHH7 cells (Fig. 2B). Furthermore, 86.44% and 86.94% of genes exhibited intact sgRNA sequences in HepG2 and JHH7 cells, respectively (Fig. 2C). These findings demonstrate that the in vivo CRISPR/Cas9 library screening process was of high quality, with reliable and reproducible results.

Fig. 2.

Fig. 2

Open in a new tab

In vivo CRISPR/Cas9 library screen indicated that AA shunting may induce the resistance of liver cancer cell to celecoxib. A Mode diagram of in vivo CRISPR/Cas9 library screen for selecting genes associated with celecoxib sensitivity. B Readcount distribution of sgRNAs in HepG2 and JHH7 in day0. C A total of 86.44% and 86.94% of genes exhibited intact sgRNA sequences in HepG2 and JHH7 cells. D A total of 1204 resisters for celecoxib sensitivity were found in HepG2 cells, indicated that knockdown of these genes increased cell sensitivity to celecoxib. E A total of 1221 resisters for celecoxib sensitivity were found in JHH7 cells, indicated that knockdown of these genes increased cell sensitivity to celecoxib. F A total of 114 sensitizers for celecoxib sensitivity were overlapped in HepG2 and JHH7. G Pathway enrichment analysis for resisters of celecoxib in HepG2 cells. H Pathway enrichment analysis for resisters of celecoxib in JHH7 cells. I Celecoxib reduced the PGE2 levels in HepG2 and JHH7 cells. J Celecoxib increased the 12-HETE levels in HepG2 and JHH7 cells. K Celecoxib increased the 14,15-EET levels in HepG2 and JHH7 cells. **, P < 0.01

Using the drugZ algorithm, 1,204 and 1,221 resisters were identified in HepG2 (Fig. 2D) and JHH7 cells (Fig. 2E), respectively. Among these, only 114 resisters were found to overlap between the two cell lines (Fig. 2F), indicating that the mechanisms of resistance to celecoxib in the two cell types may differ significantly. Through performing KEGG analysis in HepG2 and JHH7 cells, we found that the 1204 resisters in HepG2 cell were enriched in “Retrograde endocannabinoid signaling”, “Circadian entrainment”, “Glutamate metabolism”, “AGE-RAGE signaling” and “AA metabolism” (Fig. 2G), while the 1221 resisters in JHH7 were enriched in “Neuroactive ligand receptor interaction”, “Spliceosome”, “N glycan biosynthesis”, “Basal transcription factor”, and “AA metabolism” (Fig. 2H). Among them, only AA metabolism was common existed in these two cells (Fig. 2G-H). Previous studies have shown that AA shunting is a key mechanism in head and neck tumors and colorectal cancer to mediate celecoxib resistance [17, 18]. Therefore, we treated liver cancer cell HepG2 and JHH7 with celecoxib to investigate whether AA shunting also exists in liver cancer cells. We found that celecoxib treatment resulted in a reduction of the COX-2 catalytic product PGE2 in both HepG2 and JHH7 cells (Fig. 2I). However, the levels of catalytic products from ALOX and CYP enzymes, specifically 12-HETE (Fig. 2J) and 14,15-EET (Fig. 2K), were notably increased, indicating that AA shunting may contribute to the increased resistance of liver cancer cells to celecoxib.

Liver cancer cells may achieve AA shunting by upregulating different ALOX and CYP enzymes

Based on the above evidences, we further focus on the resisters enriched in the AA metabolic pathway in liver cancer cells. Interestingly, the resisters involved in the AA metabolism pathway differed between the two cell lines. The screening revealed that the genes ALOX15, ALOX5, CYP2E1, ALOX12, and CYP2A7 ranked higher as resisters in HepG2 cells (Fig. 3A), whereas in JHH7 cells, the top-ranked resisters were ALOX5, ALOX15, ALOXE3, CYP2E1 and CYP2F1 (Fig. 3B). Similarly, using qRT-PCR and western blotting, we found that the mRNA and protein levels of ALOX5, ALOX15 and CYP2E1 were both elevated in HepG2 and JHH7 cells after celecoxib treatment (Fig. 3C, D). However, the mRNA and protein levels of ALOX12 and CYP2A7 were only upregulated in celecoxib-treated HepG2 cells, while the mRNA and protein levels of ALOXE3 and CYP2F1 were only upregulated in Celecoxib-treated JHH7 cells (Fig. 3C, D).

Fig. 3.

Fig. 3

Open in a new tab

Liver cancer cells may achieve AA shunting by upregulating different ALOX and CYP enzymes. A Norm Z of resisters in AA metabolism in HepG2 cells. B Norm Z of resisters in AA metabolism in JHH7 cells. C mRNA levels of ALOX5, ALOX12, ALOX15, ALOXE3, CYP2A7, CYP2E1 and CYP2F1 in HepG2 and JHH7 cells after celecoxib treatment. D Protein levels of ALOX5, ALOX12, ALOX15, ALOXE3, CYP2A7, CYP2E1 and CYP2F1 in HepG2 and JHH7 cells after celecoxib treatment. E, F The IC50 of celecoxib in HepG2 and JHH7 cells with target gene knockdown. G-I 3D spheroid assay demonstrated the effects of celecoxib in NC group and target gene knockdown group HepG2 and JHH7 cells. **, P < 0.01

In addition, through a short-term (48 h) CCK-8 assay, we found that knockdown of ALOX5, ALOX12, ALOX15, CYP2A7, and CYP2E1 significantly increased the sensitivity of HepG2 cells to celecoxib. The IC50 values of celecoxib for HepG2 cells in these group relative to the NC HepG2 cells decreased by 2.02, 2.22, 1.91, 1.97, and 1.70 times, respectively (Fig. 3E). Similarly, suppression of ALOX5, ALOX15, ALOXE3, CYP2E1 and CYP2F1 significantly elevated the sensitivity of JHH7 cells to celecoxib. The IC50 values of celecoxib for JHH7 cells in these group relative to the NC JHH7 cells decreased by 1.78, 1.71, 1.58, 1.49, and 1.60 times, respectively (Fig. 3F). However, when we conducted a longer (21-day) 3D spheroid assay, we found that knockout of the aforementioned genes in HepG2 or JHH7 cells also enhanced celecoxib’s inhibitory effect on spheroid growth. However, compared to the reduction in spheroid diameter in the celecoxib-treated NC group, the decrease in spheroid diameter in the celecoxib-treated knockdown groups was less than 1.5 times (Fig. 3G-I), indicating that the long-term sensitization effect of knocking out these individual ALOXs or CYPs is not sufficiently strong. Taken together, all above results may indicated that although COX-2 activity was inhibited by celecoxib, liver cancer cells compensated by upregulating different ALOX and CYP enzymes, enabling the continued utilization of AA to synthesize alternative downstream products. However, targeting individual ALOXs or CYPs was insufficient to achieve a strong long-term sensitization effect.

STAT6 was act as a resister and had potential to concurrently transcribe ALOX15, ALOX12 and CYP2E1 under normal condition

Given that liver cancer cells compensate for COX-2 inhibition by upregulating various ALOX and CYP enzymes to induce AA shunting, targeting individual ALOX or CYP enzymes may not be an effective strategy for different liver cancer cell types and insufficient to achieve a strong long-term sensitization effect. This led us to hypothesize that a central mediator could concurrently regulate these enzymes.

To explore this, transcription factor enrichment analysis was performed in Metascape (https://metascape.org/gp/index.html), identifying two transcription factors, STAT6 and SP1, as regulators of multiple ALOX and CYP genes involved in celecoxib sensitivity (Fig. 4A). Specifically, STAT6 was predicted to concurrently transcribe ALOX15, ALOX12, and CYP2E1, while SP1 was predicted to regulate ALOX12, ALOX5, and CYP2F1 (Fig. 4A). Interestingly, a review of the CRISPR/Cas9 library screening results revealed that STAT6 was a more significant resister than individual ALOX or CYP genes, whereas SP1 did not exhibit a similar effect (Fig. 4B, C). Based on these findings, we considered whether STAT6 increases the resistance of liver cancer cells to celecoxib by concurrently regulating these ALOXs and CYPs transcription.

Fig. 4.

Fig. 4

Open in a new tab

STAT6 was act as a resister and had potential to concurrently transcribe ALOX15, ALOX12 and CYP2E1 under normal condition. A STAT6 and SP1 were predicted as transcription factor for regulating ALOXs and CYPs. B Norm Z score of STAT6 and SP1 in HepG2 and JHH7 cells. C Readcount of targeting STAT6 sgRNAs in day0 sample and tumor tissues treated with DMSO and celecoxib. D Motifs of STAT6 was obtained from TFDB database. E Binding sites of STAT6 for ALOX12, ALOX15 and CYP2E1 promoters were predicted. F Chip-qPCR was used to detect the binding sites of STAT6 for ALOX15 promoters. G Chip-qPCR was used to detect the binding sites of STAT6 for ALOX12 promoters. H Chip-qPCR was used to detect the binding sites of STAT6 for CYP2E1 promoters. I Double luciferase reporter plasmid with WT and MUT promoter sequences of ALOX15, ALOX12 and CYP2E1 were constructed. J Mutation of site 1 in ALOX15 promoter inhibited the transcribe induced by STAT6. K Mutation of site 2 in ALOX12 promoter inhibited the transcribe induced by STAT6. L Mutation of site 1 in CYP2E1 promoter inhibited the transcribe induced by STAT6. M, N Knockdown of STAT6 concurrently reduced the mRNA and protein levels of ALOX15, ALOX12 and CYP2E1 in HepG2 and JHH7 cells. *, P < 0.05; **, P < 0.01

Therefore, we first identify the transcriptional regulation of ALOX15, ALOX12, and CYP2E1 by STAT6 under normal condition. The binding motifs of STAT6 were retrieved in the Human TFDB database (http://bioinfo.life.hust.edu.cn/HumanTFDB#!/) (Fig. 4D). A total of three, two, and three STAT6 binding sites were predicted in the promoters of ALOX15, ALOX12, and CYP2E1, respectively (Fig. 4E). Chromatin immunoprecipitation followed by qPCR (ChIP-qPCR) confirmed that STAT6 binds specifically to site 1 in the promoter of ALOX15 (Fig. 4F), site 2 in ALOX12 (Fig. 4G), and site 1 in CYP2E1 under normal condition (Fig. 4H). To further validate these interactions, binding site mutation plasmids for ALOX15, ALOX12, and CYP2E1 were constructed (Fig. 4I). Luciferase reporter assays showed that mutations in site 1 of ALOX15 (Fig. 4J), site 2 of ALOX12 (Fig. 4K), and site 1 of CYP2E1 (Fig. 4L) significantly decreased the STAT6-induced increase in luciferase activity under normal condition. In contrast, mutations in other binding sites had no such effects. Furthermore, STAT6 knockdown significantly reduced both mRNA (Fig. 4M) and protein levels (Fig. 4N) of ALOX15, ALOX12, and CYP2E1 in liver cancer cells in normal condition. These results demonstrate that STAT6 had potential to directly bind to the promoters of ALOX15, ALOX12, and CYP2E1, concurrently driving their transcription under normal condition.

The activity of STAT6 and its transcriptional promotion effects on target genes were enhanced under celecoxib stimulation

We then analyzed whether the expression and activity of STAT6 would change in liver cancer cell under celecoxib treatment. Through performed qRT-PCR, we found that the mRNA levels of STAT6 had no significant change in HepG2 and JHH7 cells after celecoxib treatment (Fig. 5A). Similarly, through performing western blotting (Fig. 5B) and immunofluorescence (Fig. 5C), we found that the total STAT6 protein levels had no significant change in HepG2 and JHH7 cells after celecoxib treatment. However, p-STAT6 protein levels were obviously elevated in both HepG2 and JHH7 (Fig. 5B-C). These results may indicate that the activity of STAT6 was increased in liver cancer cell under celecoxib stimulation.

Fig. 5.

Fig. 5

Open in a new tab

The activity of STAT6 and its transcriptional promotion effects on target genes were enhanced under celecoxib stimulation. A qRT-PCR experiment was used to detect the mRNA levels of STAT6 in HepG2 and JHH7 cells after celecoxib treatment. B Western blotting experiment was used to detect the protein levels of p-STAT6 and STAT6 in HepG2 and JHH7 cells after celecoxib treatment. C Immunofluorescence analysis was used to detect the protein levels of p-STAT6 and STAT6 in HepG2 and JHH7 cells after celecoxib treatment. D Chip-qPCR experiment was used to detect the binding of STAT6 to the promoter of target genes ALOX15, ALOX12, and CYP2E1 in HepG2 cells after DMSO and celecoxib treatment. E Chip-qPCR experiment was used to detect the binding of STAT6 to the promoter of target genes ALOX15, ALOX12, and CYP2E1 in JHH7 cells after DMSO and celecoxib treatment. F Western blotting was used to detect the effects of STAT6 knockdown on the expression of ALOX15, ALOX12 and CYP2E1 under celecoxib treatment. G The STAT6/ALOX15/ALOX12/CYP2E1 signature score in liver cancer tissues and non-tumor tissues. H KM plot analysis for liver cancer patients with high and low signature. *, P < 0.05; **, P < 0.01

In addition, through ChIP-qPCR analysis, we found that under celecoxib stimulation, STAT6 binding to the promoter regions of target genes ALOX15, ALOX12, and CYP2E1 was enhanced in HepG2 cells (Fig. 5D), while in JHH7 cells, STAT6 binding was enhanced to the promoter regions of target genes ALOX15 and CYP2E1 (Fig. 5E). Furthermore, we found that knockdown of STAT6 significantly reduced the celecoxib-induced upregulation of ALOX15, ALOX12, and CYP2E1 expression in HepG2 cells, while in JHH7 cells, STAT6 knockdown significantly reduced the increase in ALOX15 and CYP2E1 expression induced by celecoxib (Fig. 5F). Moreover, we analyzed the liver cancer tissues from TCGA cohort. Through performing GSEA analysis to calculate the STAT6/ALOX15/ALOX12/CYP2E1 signature in tissues, elevated signature was observed in liver cancer tissues, especially in tissues provided from patients with T3-T4 stage and stage III-IV (Fig. 5G). Similarly, we found that liver cancer patients with high signature had poor overall survival days compared with those with low signature (Fig. 5H). Taken together, these evidences indicated that the transcriptional promotion effects of STAT6 on target genes were enhanced under celecoxib stimulation in liver cancer cell.

Knockdown of STAT6 increased the sensitivity of celecoxib in liver cancer cells via reducing AA shunting in vitro

As STAT6 had potential to regulate ALOX and CYP enzyme under normal condition and celecoxib stimulation, we then evaluated the biological functions of STAT6 in liver cancer cells. It was demonstrated that knockdown of STAT6 significantly reduced the levels of catalytic products from ALOX and CYP enzymes, specifically 12-HETE and 14,15-EET, in HepG2 and JHH7 cells (Fig. 6A). CCK-8 assays demonstrated that STAT6 knockdown markedly increased the sensitivity of liver cancer cells to celecoxib. Compared to NC cells, celecoxib exhibited an approximately 3-fold reduction in IC50 in STAT6-knockdown HepG2 cells, and a 4.4-fold reduction in IC50 in STAT6-knockdown JHH7 cells (Fig. 6B). Additionally, EDU assays and 3D sphere-forming experiments revealed that STAT6 knockdown slightly reduced the EDU-positive rate (Fig. 6C, D) and sphere-forming ability (Fig. 6E) of liver cancer cells. More importantly, STAT6 knockdown significantly enhanced the inhibitory effects of celecoxib on both the EDU-positive rate (Fig. 6C, D) and the sphere-forming ability (Fig. 6E). In the long-term (21-day) 3D spheroid formation assay, the reduction in spheroid diameter induced by celecoxib in sh-STAT6 HepG2 and JHH7 cells remained nearly 2-fold greater compared to the reduction observed in NC cells treated with celecoxib (Fig. 6E).

Fig. 6.

Fig. 6

Open in a new tab

Knockdown of STAT6 increased the sensitivity of celecoxib in liver cancer cells via reducing AA shunting in vitro. A Knockdown of STAT6 reduced the levels of 12-HETE and 14,15-EET in HepG2 and JHH7 cells. B-D CCK-8 and EDU assays indicated that knockdown of STAT6 increased the sensitivity of HepG2 and JHH7 cells to celecoxib. E The 3D spheroid formation assay indicates that knockdown of STAT6 increased the celecoxib’s inhibitory effect on the spheroid-forming ability of HepG2 and JHH7 cells. F STAT6-knockdown cells showed a slight increase in 12-HETE and 14,15-EET levels under celecoxib treatment. G CCK-8 assays indicated that treatment with 12-HETE and 14,15-EET reduced the inhibitory effects of celecoxib on cell proliferation in HepG2 and JHH7 cells with STAT6 knockdown. (H-I) EDU assays indicated that treatment with 12-HETE and 14,15-EET reduced the inhibitory effects of celecoxib on EDU positive rate in HepG2 and JHH7 cells with STAT6 knockdown. *, P < 0.05; **, P < 0.01

Further analysis showed that NC liver cancer cells under celecoxib treatment exhibited reduced PGE2 levels combined with elevated 12-HETE and 14,15-EET (Fig. 6F). In contrast, STAT6-knockdown cells show a slight increase in 12-HETE and 14,15-EET levels under celecoxib treatment (Fig. 6F). Moreover, CCK-8 assays (Fig. 6G) and EDU assays (Fig. 6H-I) revealed that exogenous addition of 12-HETE and 14,15-EET partially alleviated the strong inhibitory effects of celecoxib in STAT6-knockdown cells. In summary, these findings suggested that STAT6 knockdown enhanced the sensitivity of liver cancer cells to celecoxib by reducing AA shunting in vitro.

Knockdown of STAT6 increased the sensitivity of celecoxib in liver cancer cells via reducing AA shunting in vivo

The effects of STAT6 knockdown were further evaluated in vivo. HepG2 cells with STAT6 knockdown and NC cells were subcutaneously injected into mice and treated with celecoxib (Fig. 7A). Tumor tissues derived from STAT6-knockdown HepG2 cells exhibited a significantly lower proliferation rate compared with those derived from NC cells (Fig. 7B, C).

Fig. 7.

Fig. 7

Open in a new tab

Knockdown of STAT6 increased the sensitivity of celecoxib in liver cancer cells via reducing AA shunting in vivo. A In vivo animal assay procedures. (B-C) HepG2 cells with STAT6 knockdown exhibited lower proliferation ability under celecoxib treatment compared with NC cells. D, E Tumor tissues originated from HepG2 cells with STAT6 knockdown exhibited lower STAT6, ALOX15, ALOX12, CYP2E1, KI67 and PCNA expression under celecoxib treatment. F No significant difference of PGE2 between tumor tissues originated from HepG2 cells with STAT6 knockdown and NC cells under celecoxib treatment. G The expression of 12-HETE was significantly reduced in tumor tissues originated from HepG2 cells with STAT6 knockdown under celecoxib treatment compared with those originated from NC cells. F The expression of 14,15-EET was significantly reduced in tumor tissues originated from HepG2 cells with STAT6 knockdown under celecoxib treatment compared with those originated from NC cells. **, P < 0.01

IHC analysis revealed reduced expression of KI67 and PCNA in tumors derived from STAT6-knockdown cells, along with decreased levels of ALOX15, ALOX12, and CYP2E1 compared to NC-derived tumor tissues (Fig. 7D-E). Notably, there was no difference in PGE2 levels between tumors from NC cells and STAT6-knockdown cells (Fig. 7F). However, tumor tissues in the STAT6-knockdown group exhibited significantly lower levels of 12-HETE (Fig. 7G) and 14,15-EET (Fig. 7H). These findings indicated that STAT6 knockdown enhanced the sensitivity of liver cancer cells to celecoxib by reducing AA shunting in vivo.

AS1517499, a STAT6 inhibitor, exhibited high synergistic effects with celecoxib in liver cancer cells in vitro

Since STAT6 knockdown enhanced the sensitivity of liver cancer cells to celecoxib, we explored whether the combination of a STAT6 inhibitor and celecoxib could serve as a potential therapeutic strategy for liver cancer. Eight known STAT6 inhibitors—STAT6-IN-1, STAT6-IN-2, STAT6-IN-3, STAT6-IN-4, STAT6-IN-5, PM-81I, AS1517499, and AS18107222—were evaluated.

First, the nonspecific toxicity of these inhibitors was assessed in normal hepatocyte THLE-2 cells, with their IC50 values determined as 83.886 µM, 39.925 µM, 254.361 µM, 145.662 µM, 291.949 µM, 244.016 µM, 309.468 µM, and 25.573 µM, respectively (Fig. 8A). Next, their inhibitory effects were measured in HepG2 and JHH7 liver cancer cells. The IC50 values for HepG2 cells were 6.091 µM, 12.650 µM, 2.235 µM, 11.369 µM, 35.122 µM, 37.973 µM, 1.373 µM, and 13.247 µM, respectively (Fig. 8B). Similarly, for JHH7 cells, the IC50 values were 7.190 µM, 25.287 µM, 3.068 µM, 29.547 µM, 32.502 µM, 24.320 µM, 1.294 µM, and 10.303 µM, respectively (Fig. 8C). Among these inhibitors, STAT6-IN-1, STAT6-IN-3, and AS1517499 demonstrated higher selectivity for liver cancer cells, suggesting their potential as effective STAT6 inhibitors for combination therapy with celecoxib.

Fig. 8.

Fig. 8

Open in a new tab

AS1517499, a STAT6 inhibitor, exhibited high synergistic effects with celecoxib in liver cancer cells in vitro. A The inhibitory effects of eight STAT6 inhibitor were detected in THLE-2 cell. B The inhibitory effects of eight STAT6 inhibitor were detected in HepG2 cell. C The inhibitory effects of eight STAT6 inhibitor were detected in JHH7 cell. D, E The combination effects of three STAT6 inhibitor (STAT6-IN-1, STAT6-IN-3 and AS1517499) with celecoxib were detected in HepG2 cell. F-G The combination effects of three STAT6 inhibitor (STAT6-IN-1, STAT6-IN-3 and AS1517499) with celecoxib were detected in JHH7 cell

The combined effects of different concentrations of the inhibitors and celecoxib were next evaluated in HepG2 and JHH7 cells. The results showed that most concentrations of STAT-IN-1 primarily exhibited an additive effect with celecoxib, while minority concentrations demonstrated a moderate synergistic effect in both HepG2 (Fig. 8D-E) and JHH7 (Fig. 8F, G) cells. In contrast, most concentrations of STAT6-IN-3 and AS1517499 displayed moderate to high synergistic effects with celecoxib in both HepG2 (Fig. 8D-E) and JHH7 (Fig. 8F, G) cells. Notably, the combination of AS1517499 and celecoxib exhibited the strongest synergistic effects, outperforming the other inhibitors tested (Fig. 8E and G).

AS1517499, a STAT6 inhibitor, exhibited high synergistic effects with celecoxib in liver cancer cells through inhibiting AA shunting in vivo

Given the excellent synergy between AS1517499 and celecoxib observed in vitro, we next evaluated their combined effects in vivo. HepG2 cells were subcutaneously injected into nude mice and treated with DMSO, AS1517499, celecoxib, or their combination (Fig. 9A). Interestingly, while short-term in vitro CCK-8 experiments demonstrated notable inhibitory effects, prolonged treatment with either AS1517499 or celecoxib alone in vivo resulted in only modest tumor growth suppression (Fig. 9B-C). In contrast, tumor tissues derived from HepG2 cells treated with the combination of AS1517499 and celecoxib exhibited significantly slower growth compared to the single-treatment groups (Fig. 9B, C).

Fig. 9.

Fig. 9

Open in a new tab

AS1517499, a STAT6 inhibitor, exhibited high synergistic effects with celecoxib in liver cancer cells through inhibiting AA shunting in vivo. A HepG2 cells were subcutaneously injected into nude mice, and treated with DMSO, celecoxib, AS1517499 and their combination. B, C The proliferation rate of tumor tissues in each group was detected. D, E The expression of COX-2, STAT6, ALOX15, ALOX12, CYP2E1, PCNA and KI67 was detected in tumor tissues in each group. F HE staining showed no detectable damage in the heart, gastrointestinal tract, liver, and kidney organs in combine treatment group. *, P < 0.05; **, P < 0.01

Moreover, IHC analysis revealed that tumor tissues treated with single-agent celecoxib exhibited a slight reduction in KI67 and PCNA expression, accompanied by increased levels of ALOX12, ALOX15, and CYP2E1 (Fig. 9D, E). In contrast, the combination of AS1517499 and celecoxib resulted in a significant reduction in KI67 and PCNA, without any elevation in ALOX12, ALOX15, or CYP2E1 levels (Fig. 9D, E). These findings indicate that AS1517499, a STAT6 inhibitor, synergizes with celecoxib to suppress liver cancer cell growth by inhibiting AA shunting in vivo. Furthermore, to evaluate the safety of the combination treatment, the heart, gastrointestinal tract, liver, and kidney were extracted from mice treated with the combination of AS1517499 and celecoxib. HE staining showed no detectable damage in any of these organs, confirming the safety of the combination therapy (Fig. 9F).

Discussion

Monotherapy often falls short in managing advanced cancers due to the complex and multifaceted mechanisms driving tumor progression and uncontrolled cellular proliferation. Cancer cells possess the ability to adapt and develop resistance, making it challenging for a single drug to fully halt disease progression [24]. Consequently, combination therapy has emerged as a key strategy in cancer treatment. By targeting multiple pathways simultaneously, combination therapy not only enhances therapeutic efficacy but also achieves a synergistic effect where the combined impact exceeds the sum of individual drug effects, often described as “1 + 1 > 2.” This approach has become a cornerstone in anti-tumor research, offering hope for improved outcomes in patients with difficult-to-treat malignancies [25, 26].

Celecoxib, a selective COX-2 inhibitor, has demonstrated significant potential in cancer therapy due to its anti-inflammatory properties. Both preclinical studies and clinical trials have highlighted its efficacy against various cancers, including liver cancer, either as monotherapy or in combination with chemotherapy [27, 28]. However, prolonged use often induces resistance, necessitating dose escalation to maintain therapeutic effects. This escalation frequently leads to severe gastrointestinal and cardiac side effects, limiting its long-term application [15, 16]. To investigate the molecular mechanisms underlying celecoxib resistance in liver cancer, an in vivo whole-genome CRISPR/Cas9 library screen was performed in the current study, revealing AA shunting as a key driver of resistance. This finding in liver cancer was consistent with previously observed phenomena in head and neck tumors [17] and colorectal cancer [18].

The branch of AA metabolism mediated by ALOX, which produces cytokines such as LTB4, not only contributes to celecoxib resistance but also plays a significant role in the progression of liver cancer. For example, Sinha et al. demonstrated that LTB4 produced by ALOX enzymes can bind to its receptor and activate the β-catenin signaling pathway, thereby promoting the proliferation of liver cancer cells [29]. Wang et al. activation of ALOX5-LTB4 pathway in liver cancer would promote CD8 + T cell exhaustion in tissues and mediate the formation of an immunosuppressive microenvironment [30]. In response, dual ALOX/COX inhibitors, such as 2,4-disubstituted thiophene derivatives [31] and licofelone [32, 33], have been developed. While these drugs show efficacy in treating inflammatory diseases, their anti-cancer effects remain limited.

In our study, we further investigated the molecular mediators involved in AA shunting. We found that, in addition to ALOX enzymes, CYP enzymes also play a critical role as resisters mediating celecoxib resistance. However, the specific ALOXs and CYPs that were compensatorily upregulated at the transcriptional level in response to celecoxib differed between the two liver cancer cell lines. Knockout of these ALOXs or CYPs in the corresponding cell lines significantly enhanced celecoxib sensitivity in the short term. However, long-term observations using 3D spheroid formation assays revealed that this strategy had only modest effects in sustaining celecoxib sensitivity. These findings suggested that targeting a single ALOX or CYP enzyme is unlikely to be broadly effective across most liver cancers and may not be sufficient to achieve long-term enhancement of celecoxib sensitivity.

Transcription factors are proteins that regulate gene expression by binding to specific DNA sequences, typically within promoter or enhancer regions, to activate or repress the transcription of target genes. They are pivotal in various biological processes, including cell growth, differentiation, and the development of drug resistance [34, 35]. Given the limitations associated with targeting single ALOXs or CYPs, we sought to identify a regulatory network involving transcription factors capable of modulating multiple targets simultaneously. Through our analysis, we identified STAT6 as a key transcription factor capable of regulating both ALOX and CYP gene families, including ALOX12, ALOX15, and CYP2E1 under normal condition.

STAT6, a member of the STAT family, has been implicated in cancer progression, with clinical evidence demonstrating a strong association between elevated STAT6 expression and advanced malignancies [36]. Recent studies reveal that STAT6 knockdown significantly reduces cancer cell proliferation and metastasis [37]. Additionally, STAT6 has been shown to regulate the transcription of various cytokines in cancer cells, which are subsequently secreted into the tumor microenvironment, promoting macrophage recruitment and M2 polarization [38, 39]. Interestingly, in our study, we found that STAT6 activity was increased in liver cancer cells after celecoxib treatment. Moreover, although STAT6 can bind to the promoters of ALOX15, ALOX12, and CYP2E1 under normal conditions and promote their transcription in both HepG2 and JHH7 cells. However, under celecoxib stimulation, STAT6 binding to the promoter regions of ALOX15, ALOX12, and CYP2E1 was enhanced in HepG2 cells, while in JHH7 cells, STAT6 binding was enhanced only at the promoter regions of ALOX15 and CYP2E1. Previous studies have shown that the transcriptional regulation of ALOX target genes by STAT6 is not only associated with its enhanced phosphorylation activity, but also depends on the expression of cofactors such as histone methyltransferases and histone acetyltransferases, as well as the epigenetic modifications at the promoter regions of the target genes. For example, Liu et al. demonstrated that the levels of histone trimethylation and acetylation at the ALOX15 promoter region are key factors for STAT6 recruitment to this region. Loss of the histone methyltransferase SET and MYND domain-containing protein 3 in cells reduces histone trimethylation and acetylation at the ALOX15 promoter, thereby diminishing the transcriptional activation of ALOX15 by STAT6 [40]. Shankaranarayanan et al. [41] and Namgaladze et al. [42]. indicated that STAT6 can be activated in response to IL-4 stimulation; however, the transcriptional activity of activated STAT6 on its target genes depends on histone acetylation at the promoter regions of those genes. Inhibition of CBP/p300 reduces histone acetylation in cells and significantly diminishes the ability of activated STAT6 to drive the transcription of its target gene, ALOX15. Therefore, we speculate that the differential transcriptional activation by STAT6 observed in the two liver cancer cell lines under celecoxib stimulation may be influenced by cofactor recruitment and epigenetic modifications at the promoter regions.

Moreover, we presented the first evidences that knockdown of STAT6 can simultaneously suppress the transcription of ALOX12, ALOX15, and CYP2E1, which are key enzymes initiating the ALOX and CYP branches of AA metabolism. As a result, when celecoxib blocks COX pathway, the cells are severely impaired in their ability to produce compensatory metabolites through either ALOX or CYP branch, thereby enhancing their sensitivity to celecoxib. Furthermore, AS1517499, a STAT6 inhibitor, demonstrated synergistic effects with celecoxib in liver cancer cells by effectively inhibiting AA shunting. These evidences indicating that targeting STAT6 is an effective strategy to increase the sensitivity of liver cancer to celecoxib.

Certainly, our study has several limitations. First, although we observed that STAT6 phosphorylation levels increased upon celecoxib stimulation, the upstream factors responsible for mediating this activation remain unclear. Second, we found that the extent of STAT6 binding to the promoters of its target genes under celecoxib treatment varied between different cell lines. However, it is not yet known whether this variability is due to differences in epigenetic modifications at the promoter regions or differences in cofactor recruitment. Moreover, our investigation of the combination therapy with AS1517499 and celecoxib focused solely on evaluating toxicity in the heart, liver, kidney, and gastrointestinal tract, without assessing potential effects on immune suppression or skin damage. Lastly, we focused primarily on the common mechanism of AA shunting, while specific resistance mechanisms such as glutamate metabolism in HepG2 cells were not explored. Therefore, further studies are required to address these important aspects.

In conclusion, our study indicated that AA shunting plays a crucial role in driving liver cancer cell resistance to celecoxib. Although COX-2 activity is inhibited, liver cancer cells adapt by upregulating ALOX and CYP enzymes, allowing the continued synthesis of alternative downstream metabolites from AA. STAT6 was a central transcription factor that simultaneously regulates ALOX15, ALOX12, and CYP2E1. Suppression of STAT6, either via gene knockdown or using the inhibitor AS1517499, effectively disrupted AA shunting and significantly increased the sensitivity of liver cancer cells to celecoxib. Combining AS1517499 with celecoxib may be a novel therapeutic strategy for liver cancer (Fig. 10).

Fig. 10.

Fig. 10

Open in a new tab

AA shunting plays a crucial role in driving liver cancer cell resistance to celecoxib. Although COX-2 activity is inhibited, liver cancer cells adapt by upregulating ALOX and CYP enzymes, allowing the continued synthesis of alternative downstream metabolites from AA. STAT6 was a central transcription factor that simultaneously regulates ALOX15, ALOX12, and CYP2E1. Suppression of STAT6, either via gene knockdown or using the inhibitor AS1517499, effectively disrupted AA shunting and significantly increased the sensitivity of liver cancer cells to celecoxib

Supplementary Information

Supplementary Material 1. (9KB, xlsx)
Supplementary Material 2. (9.7KB, xlsx)
Supplementary Material 3. (13.3MB, pptx)

Acknowledgements

This study was supported by the Continuous Support Fund for Excellent Scientific Research Platform of Colleges and Universities in Guizhou Province (QJJ [2022] 020), the National Natural Science Foundation Cultivation Project of the Affiliated Hospital of Guizhou Medical University (gyfynsfc-2021-4), the Affiliated Hospital of Guizhou Medical University Leading Scholar Project (gyfykyc-2023-01), the National Natural Science Foundation of China (82473969, 82160665, 82260535), China Postdoctoral Science Foundation (2022M720929 and 2024M750674), Guizhou High-level Innovative Talents Supporting Program (GCC[2023]002), Guizhou Provincial Science and Technology Projects General (ZK[2024] major 039), and Startup Fund for PhD Scholars of Guizhou Medical University (Xiaobohe J 2022 [061]).

Abbreviations

AA

Arachidonic acid

ALOX

Lipoxygenases

ALOX12

Arachidonate 12-Lipoxygenase

ALOX15

Arachidonate 15-Lipoxygenase

JAMA

Journal of the American Medical Association

ChIP

Chromatin Immunoprecipitation

CI

Combination index

COX

Cyclooxygenasess

COX-2

Cyclooxygenase-2

CYP

Cytochrome P450

CYP2E1

Cytochrome P450 2E1

EET

Epoxyeicosatrienoic acid

HCC

Hepatocellular Carcinoma

HETE

Hydroxyeicosatetraenoic acid

IC50

Half maximal inhibitory concentration

IHC

Immunohistochemistry

NSAIDs

Nonsteroidal anti-inflammatory drugs

PGE2

Prostaglandin E2

shRNA

short hairpin RNA

STAT6

Signal transducer and activator of transcription 6

Authors' contributions

Chujiao Hu, Zhirui Zeng and Xin Bao performed the experiments and conducted result analysis. Dahuan Li, Huading Tai, and Haohao Zeng contribute to part experiment. Cheng Luo, Lei Tang, Tengxiang Chen, and Shi Zuo designed the experiments. Chujiao Hu, Zhirui Zeng, Lei Tang, Tengxiang Chen, and Shi Zuo provided the fund. Zhirui Zeng wrote the manuscript. All authors had read the manuscript and agree to submit.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Chujiao Hu, Zhirui Zeng and Xin Bao contributed equally to this work.

Contributor Information

Lei Tang, Email: tlei1974@163.com.

Tengxiang Chen, Email: txch@gmc.edu.cn.

Shi Zuo, Email: drzuoshi@gmc.edu.cn.

References

  • 1.Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–63. [DOI] [PubMed] [Google Scholar]
  • 2.Demir T, Lee SS, Kaseb AO. Systemic therapy of liver cancer. Adv Cancer Res. 2021;149:257–94. [DOI] [PubMed] [Google Scholar]
  • 3.Anwanwan D, Singh SK, Singh S, Saikam V, Singh R. Challenges in liver cancer and possible treatment approaches. Biochim Biophys Acta Rev Cancer. 2020;1873(1):188314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Li L, Wang H. Heterogeneity of liver cancer and personalized therapy. Cancer Lett. 2016;379(2):191–7. [DOI] [PubMed] [Google Scholar]
  • 5.Yang C, Zhang H, Zhang L, Zhu AX, Bernards R, Qin W, et al. Evolving therapeutic landscape of advanced hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol. 2023;20(4):203–22. [DOI] [PubMed] [Google Scholar]
  • 6.Yang YM, Kim SY, Seki E. Inflammation and liver cancer: molecular mechanisms and therapeutic targets. Semin Liver Dis. 2019;39(1):26–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Gorica E, Calderone V. Arachidonic acid derivatives and neuroinflammation. CNS Neurol Disord Drug Targets. 2022;21(2):118–29. [DOI] [PubMed] [Google Scholar]
  • 8.Yarla NS, Bishayee A, Sethi G, Reddanna P, Kalle AM, Dhananjaya BL, et al. Targeting arachidonic acid pathway by natural products for cancer prevention and therapy. Semin Cancer Biol. 2016;40–41:48–81. [DOI] [PubMed] [Google Scholar]
  • 9.Breinig M, Schirmacher P, Kern MA. Cyclooxygenase-2 (COX-2)—a therapeutic target in liver cancer? Curr Pharm Des. 2007;13(32):3305–15. [DOI] [PubMed] [Google Scholar]
  • 10.Huang M, Wang L, Chen J, Bai M, Zhou C, Liu S, et al. Regulation of COX-2 expression and epithelial-to-mesenchymal transition by hypoxia-inducible factor-1α is associated with poor prognosis in hepatocellular carcinoma patients post TACE surgery. Int J Oncol. 2016;48(5):2144–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Liggett JL, Zhang X, Eling TE, Baek SJ. Anti-tumor activity of non-steroidal anti-inflammatory drugs: cyclooxygenase-independent targets. Cancer Lett. 2014;346(2):217–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Tołoczko-Iwaniuk N, Dziemiańczyk-Pakieła D, Nowaszewska BK, Celińska-Janowicz K, Miltyk W. Celecoxib in cancer therapy and prevention—review. Curr Drug Targets. 2019;20(3):302–15. [DOI] [PubMed] [Google Scholar]
  • 13.Tong H, Wei B, Chen S, Xie YM, Zhang MG, Zhang LH, et al. Adjuvant celecoxib and Lanreotide following transarterial chemoembolisation for unresectable hepatocellular carcinoma: a randomized pilot study. Oncotarget. 2017;8(29):48303–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Meyerhardt JA, Shi Q, Fuchs CS, Meyer J, Niedzwiecki D, Zemla T, et al. Effect of celecoxib vs placebo added to standard adjuvant therapy on disease-free survival among patients with stage III colon cancer: the CALGB/SWOG 80702 (Alliance) randomized clinical trial. JAMA. 2021;325(13):1277–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Duffield-Lillico AJ, Boyle JO, Zhou XK, Ghosh A, Butala GS, Subbaramaiah K, et al. Levels of prostaglandin E metabolite and leukotriene E4 are increased in the urine of smokers: evidence that celecoxib shunts arachidonic acid into the 5-lipoxygenase pathway. Cancer Prev Res (Phila). 2009;2(4):322–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Mao JT, Tsu IH, Dubinett SM, Adams B, Sarafian T, Baratelli F, et al. Modulation of pulmonary leukotriene B4 production by cyclooxygenase-2 inhibitors and lipopolysaccharide. Clin Cancer Res. 2004;10(20):6872–8. [DOI] [PubMed] [Google Scholar]
  • 17.Park SW, Heo DS, Sung MW. The shunting of arachidonic acid metabolism to 5-lipoxygenase and cytochrome p450 epoxygenase antagonizes the anti-cancer effect of cyclooxygenase-2 Inhibition in head and neck cancer cells. Cell Oncol (Dordr). 2012;35(1):1–8. [DOI] [PubMed] [Google Scholar]
  • 18.Ganesh R, Marks DJ, Sales K, Winslet MC, Seifalian AM. Cyclooxygenase/lipoxygenase shunting lowers the anti-cancer effect of cyclooxygenase-2 Inhibition in colorectal cancer cells. World J Surg Oncol. 2012;10:200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Xia H, He Q, Wang H, Wang Y, Yang Y, Li Y, et al. Treatment with either COX-2 inhibitor or 5-LOX inhibitor causes no compensation between COX-2 pathway and 5-LOX pathway in chronic aluminum overload-induced liver injury in rats. Fundam Clin Pharmacol. 2019;33(5):535–43. [DOI] [PubMed] [Google Scholar]
  • 20.Hart T, Tong AHY, Chan K, Van Leeuwen J, Seetharaman A, Aregger M, et al. Evaluation and design of genome-wide CRISPR/SpCas9 knockout screens. G3 (Bethesda). 2017;7(8):2719–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Fan LW, Kaizaki A, Tien LT, Pang Y, Tanaka S, Numazawa S, et al. Celecoxib attenuates systemic lipopolysaccharide-induced brain inflammation and white matter injury in the neonatal rats. Neuroscience. 2013;240:27–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Naeem S, Ikram R, Khan SS, Rao SS. NSAIDs ameliorate cognitive and motor impairment in a model of parkinsonism induced by chlorpromazine. Pak J Pharm Sci. 2017;30(3):801–8. [PubMed] [Google Scholar]
  • 23.Huang C, Wang J, Liu H, Huang R, Yan X, Song M, et al. Ketone body β-hydroxybutyrate ameliorates colitis by promoting M2 macrophage polarization through the STAT6-dependent signaling pathway. BMC Med. 2022;20(1):148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Liang P, Ballou B, Lv X, Si W, Bruchez MP, Huang W et al. Monotherapy and combination therapy using anti-angiogenic nanoagents to fight cancer. Adv Mater. 2021;33(15). [DOI] [PubMed]
  • 25.Bayat Mokhtari R, Homayouni TS, Baluch N, Morgatskaya E, Kumar S, Das B, et al. Combination therapy in combating cancer. Oncotarget. 2017;8(23):38022–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Jin H, Wang L, Bernards R. Rational combinations of targeted cancer therapies: background, advances and challenges. Nat Rev Drug Discov. 2023;22(3):213–34. [DOI] [PubMed] [Google Scholar]
  • 27.Wu K, Ma S, Xu X, Liu Y, Tian C, Zhang C, et al. Celecoxib and cisplatin dual-loaded microspheres synergistically enhance transarterial chemoembolization effect of hepatocellular carcinoma. Mater Today Bio. 2023;24:100927. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Saxena P, Sharma PK, Purohit P. A journey of celecoxib from pain to cancer. Prostaglandins Other Lipid Mediat. 2020;147:106379. [DOI] [PubMed] [Google Scholar]
  • 29.Sinha S, Aizawa S, Nakano Y, Rialdi A, Choi HY, Shrestha R, et al. Hepatic stellate cell Stearoyl co-A desaturase activates leukotriene B4 receptor 2 - β-catenin cascade to promote liver tumorigenesis. Nat Commun. 2023;14(1):2651. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wang G, Shen X, Jin W, Song C, Dong M, Zhou Z, et al. Elucidating the role of S100A10 in CD8 + T cell exhaustion and HCC immune escape via the cPLA2 and 5-LOX axis. Cell Death Dis. 2024;15(8):573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Rakesh KS, Jagadish S, Swaroop TR, Mohan CD, Ashwini N, Harsha KB, et al. Anti-cancer activity of 2,4-disubstituted thiophene derivatives: dual inhibitors of Lipoxygenase and cyclooxygenase. Med Chem. 2015;11(5):462–72. [DOI] [PubMed] [Google Scholar]
  • 32.Mohammed A, Janakiram NB, Li Q, Choi CI, Zhang Y, Steele VE, et al. Chemoprevention of colon and small intestinal tumorigenesis in APC(Min/+) mice by licofelone, a novel dual 5-LOX/COX inhibitor: potential implications for human colon cancer prevention. Cancer Prev Res (Phila). 2011;4(12):2015–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Tavolari S, Bonafè M, Marini M, Ferreri C, Bartolini G, Brighenti E, et al. Licofelone, a dual COX/5-LOX inhibitor, induces apoptosis in HCA-7 colon cancer cells through the mitochondrial pathway independently from its ability to affect the arachidonic acid cascade. Carcinogenesis. 2008;29(2):371–80. [DOI] [PubMed] [Google Scholar]
  • 34.Papavassiliou KA, Papavassiliou AG. Transcription factor drug targets. J Cell Biochem. 2016;117(12):2693–6. [DOI] [PubMed] [Google Scholar]
  • 35.Francois M, Donovan P, Fontaine F. Modulating transcription factor activity: interfering with protein-protein interaction networks. Semin Cell Dev Biol. 2020;99:12–9. [DOI] [PubMed] [Google Scholar]
  • 36.Karpathiou G, Papoudou-Bai A, Ferrand E, Dumollard JM, Peoc’h M. STAT6: a review of a signaling pathway implicated in various diseases with a special emphasis in its usefulness in pathology. Pathol Res Pract. 2021;223:153477. [DOI] [PubMed] [Google Scholar]
  • 37.Lu G, Shi W, Zheng H. Inhibition of STAT6/Anoctamin-1 activation suppresses proliferation and invasion of gastric cancer cells. Cancer Biother Radiopharm. 2018;33(1):3–7. [DOI] [PubMed] [Google Scholar]
  • 38.Rahal OM, Wolfe AR, Mandal PK, Larson R, Tin S, Jimenez C, et al. Blocking Interleukin (IL)4- and IL13-mediated phosphorylation of STAT6 (Tyr641) decreases M2 polarization of macrophages and protects against macrophage-mediated radioresistance of inflammatory breast cancer. Int J Radiat Oncol Biol Phys. 2018;100(4):1034–43. [DOI] [PubMed] [Google Scholar]
  • 39.Yu T, Gan S, Zhu Q, Dai D, Li N, Wang H, et al. Modulation of M2 macrophage polarization by the crosstalk between Stat6 and Trim24. Nat Commun. 2019;10(1):4353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Liu C, Xu D, Han H, Fan Y, Schain F, Xu Z, et al. Transcriptional regulation of 15-lipoxygenase expression by histone h3 lysine 4 methylation/demethylation. PLoS ONE. 2012;7(12):e52703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Shankaranarayanan P, Chaitidis P, Kühn H, Nigam S. Acetylation by histone acetyltransferase CREB-binding protein/p300 of STAT6 is required for transcriptional activation of the 15-lipoxygenase-1 gene. J Biol Chem. 2001;276(46):42753–60. [DOI] [PubMed] [Google Scholar]
  • 42.Namgaladze D, Snodgrass RG, Angioni C, Grossmann N, Dehne N, Geisslinger G, et al. AMP-activated protein kinase suppresses arachidonate 15-lipoxygenase expression in Interleukin 4-polarized human macrophages. J Biol Chem. 2015;290(40):24484–94. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (9KB, xlsx)
Supplementary Material 2. (9.7KB, xlsx)
Supplementary Material 3. (13.3MB, pptx)

Data Availability Statement

No datasets were generated or analysed during the current study.

Articles from Cell Communication and Signaling : CCS are provided here courtesy of BMC

RESOURCES