Squalene synthase induces ERα expression via cholesterol supplementation to confer statin resistance in lung cancer cells

Abstract

Background

Sterol regulatory element-binding proteins (SREBPs) are master regulators of cholesterol and lipid biosynthesis - pathways increasingly linked to cancer progression. Statins, which inhibit HMG-CoA reductase to lower cholesterol, have shown potential in reducing cancer recurrence. However, their efficacy in lung cancer remains uncertain, and predictive biomarkers for statin responsiveness are still lacking.

Methods

We examined the relationship between cholesterol biosynthesis and estrogen receptor alpha (ERα) signaling and assessed the therapeutic potential of targeting this axis in lung cancer models.

Results

Overexpression of farnesyl-diphosphate farnesyltransferase 1 (FDFT1), also known as squalene synthase (SQS), significantly increased the IC₅₀ of lovastatin in lung cancer cell lines, indicating a role in mediating statin resistance. Gene expression profiling revealed enrichment of estrogen receptor alpha (ERα) signaling in SQS-overexpressing cells. Immunohistochemical analysis of 125 NSCLC patient samples showed a positive correlation between SQS and ERα protein expression, and their co-expression was significantly associated with poorer disease-free survival. Mechanistically, SQS upregulated ERα at the protein level via cholesterol replenishment, without altering ESR1 mRNA levels, suggesting a post-transcriptional regulatory mechanism.

Conclusion

SQS promotes statin resistance in lung cancer by stabilizing ERα through cholesterol replenishment. Its co-expression with ERα predicts poor prognosis, highlighting the SQS–ERα axis as a potential therapeutic target and biomarker for stratifying patients likely to benefit from statin–antiestrogen combination therapy.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12935-026-04220-7.

Introduction

Metabolic reprogramming is a hallmark of many cancer types. In fact, cancer cells use metabolic pathways to enhance their malignant properties [1]. In humans, cholesterol is a precursor of the steroid hormones, oxysterol, bile acid, and vitamin D. Cholesterol is also a major structural component of cell membranes. Thus, enrichment of receptors in lipid rafts promotes cancer metastasis [2]. Cholesterol biosynthesis is mediated by the sterol regulatory element-binding protein 2 (SREBP2), which plays a major role in regulating the mevalonate (MVA) pathway and mediating statin resistance in cancer cells [3, 4]. Statins, inhibitors of hydroxymethylglutaryl coenzyme A reductase (HMGCR), are commonly prescribed for hypercholesterolemia and have attracted attention for their potential anticancer properties. Epidemiological studies have linked elevated serum cholesterol to increased cancer risk, particularly in prostate cancer [5–7]. Statin use has been associated with reduced recurrence across multiple tumor types [8–10]. Despite these observations, prospective clinical trials have yielded mixed results. Notably, the LUNGSTAR phase III trial, which evaluated the addition of statins to standard chemotherapy in small cell lung cancer (SCLC), showed no significant clinical benefit. These conflicting outcomes highlight a critical need to identify molecular determinants of statin sensitivity [11]. While various statins —such as pravastatin, simvastatin, and lovastatin—are under investigation in combination with conventional cancer therapies, the underlying mechanisms driving statin response in cancer remain poorly defined [12].

Cholesterol plays critical roles in cancer cell signaling, membrane integrity, and metastasis. Dysregulation of the mevalonate pathway—governed by SREBP2 and its downstream effectors including squalene synthase (SQS/FDFT1)—has been associated with tumor progression in several cancers, including non-small cell lung cancer (NSCLC). Cholesterol-rich lipid rafts facilitate the clustering of oncogenic receptors, while cholesterol metabolites can activate nuclear receptors that support proliferation and survival. These findings suggest that targeting cholesterol biosynthesis may offer therapeutic value, particularly in tumors with elevated pathway activity. HMGCR is the rate-limiting enzyme in the synthesis of MVA, a precursor metabolite for the de novo synthesis of cholesterol, farnesyl pyrophosphate (FPP), and geranylgeranyl pyrophosphate (GGPP). Aberrant activation of the MVA pathway has been linked to cancer progression via cholesterol-, FPP-, and GGPP-dependent mechanisms [2, 13–16]. Preclinical studies have shown that statin-induced apoptosis can be rescued by MVA but not by sterols, indicating that GGPP or FPP may be targeted for statin-induced cytotoxicity [17, 18]. Based on conflicting results, cancer cells with hyperactivated RAS GTPase superfamily members are sensitive to statin treatment [19–21]. However, consistent results have not been reported by other research groups [17, 22, 23]. Therefore, deciphering the molecular mechanisms underlying these discrepancies is important before statins are repurposed as anticancer agents [24].

In this study, we aimed to investigate the molecular features that dictate statin sensitivity in lung cancer by correlating large-scale statin sensitivity data with basal gene expression in lung cancer cell lines [25, 26]. Briefly, we identified a gene signature associated with lovastatin sensitivity and revealed that gene expression in the mesenchymal phenotype, cholesterol biosynthesis pathway, and estrogen receptor alpha (ERα) signaling are key contributors to statin sensitivity in lung cancer cells. Moreover, we found that the overexpression of one of the cholesterol biosynthesis enzymes, farnesyl-diphosphate farnesyltransferase 1 (FDFT1), also known as squalene synthase (SQS), can stabilize ERα and cause resistance to statin-based therapy. Furthermore, the overexpression of SQS was found to be positively correlated with the expression of ERα in patients with lung cancer. The combination of SQS and ERα was also identified as a powerful prognostic predictor in lung cancer.

Materials and methods

In Silico mRNA profiles and Kaplan–Meier analysis

Kaplan–Meier analysis of overall survival was performed using a publicly available lung cancer microarray dataset (http://kmplot.com/analysis/). Correlation analysis of the gene expression levels was performed using Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/).

IPA upstream regulator analysis

The activation/inhibition status of the upstream regulators of the lovastatin sensitivity gene signature was predicted using IPA Upstream Regulator Analysis (Ingenuity Systems, www.ingenuity.com) and the calculated z-scores reflected the overall activation state of the regulator (< 0, inhibited; >0, activated). In practice, a Z-score greater than 2 or less than − 2 can be considered to indicate significant activation or inhibition, respectively.

In Silico datasets and metabolite-dependency

The metabolomic and transcriptomic profiles of the lung cancer cell lines were established using the CLE server. In previous studies, the degree of dependence of each gene on metabolites was determined using statistical analyses [27]. Thus, we selected various cholesterol biosynthesis-related genes (HMGCS, HMGCR, FNTA, FNTB, FDFT1, and SQLE) and target events involving phosphatidylcholine (PC), sphingomyelin (SM), and ceramide (Fig. S2A-2 C). A high gene expression level is represented by a positive value, which indicates high metabolic production, and vice versa.

Specimens

A total of 125 non-small cell lung cancer (NSCLC) samples were collected at Kaohsiung Medical University Hospital following approval from the Institutional Review Board (IRB) and permission from the ethics committee of Kaohsiung Medical University Hospital (KMUHIRB-E(I)−20160099). As all data were analyzed anonymously and no identifying information related to the participants was included, the IRB of Kaohsiung Medical University Hospital approved a waiver of informed consent. All studies were conducted according to the regulations and guidelines for the collection and use of human specimens for research at the Kaohsiung Medical University Hospital. All patients were treated according to standard treatment protocols. In patients with stage I NSCLC, the tumors were resected and no adjuvant chemotherapy was administered. Patients with stage II-III NSCLC were treated with platinum-based chemotherapy after tumor resection. Patients with inoperable locally invasive or metastatic disease were treated with chemotherapy with or without radiotherapy. Histological types were diagnosed according to the 2004 World Health Organization (WHO) classification guidelines for lung cancer. Pathological diagnoses were performed according to the American Joint Committee on Cancer (AJCC) TNM classification of lung cancer. Overall survival was defined as the period between the first treatment and patient death, whereas disease-free survival was defined as the interval from the first treatment to disease relapse or death. The longest follow-up duration was 200 months. Histological diagnosis was performed according to the WHO classification guidelines for lung cancer. Pathological diagnoses of tumor size, local invasion, lymph node involvement, distal metastasis, and final disease stage were conducted according to the AJCC TNM classification of lung cancer.

Tissue microarray and immunohistochemical staining

Tissue cores (1 mm diameter) from each specimen were selected by matching the histology of the original hematoxylin and eosin (H&E)-stained slides with the confirmed histopathological diagnosis. Immunohistochemistry (IHC) was performed using an automated immunostainer (Ventana Discovery XT Autostainer, Ventana, USA). Paraffin sections were dewaxed in a 60 °C oven, deparaffinized using xylene, and then rehydrated in graded ethanol. Antigen retrieval was performed using heat-induced TRIS-EDTA buffer for 30 min. Immunoreactive protein was developed using a 3,3’-diaminobenzidine (DAB) peroxidase substrate kit (Ventana, USA). The slides were counterstained with hematoxylin. The following antibodies were used to detect SQS and ERα expression: SQS (1:500, GTX104091, Genetex) and ERα (1:100, GTX82778, Genetex) (Table S5).

IHC staining assessment

The IHC staining results were evaluated and scored by two independent pathologists who were blinded to the patients’ clinical information. The protein expression of SQS and ERα in tumor cells was scored according to the staining intensity: 0, no staining; 1⁺, weak staining; 2⁺, moderate staining; and 3⁺, strong staining. Scores of 0 and 1 indicated low expression, and scores of 2 and 3 indicated high expression.

Western blot analysis

Protein (20–50 µg) was separated using 10% SDS-polyacrylamide gels via electrophoresis. After electrophoresis, the proteins were transferred onto a nitrocellulose membrane and blocked with 5% non-fat dry milk in PBS containing Tween 20. Immunoblotting was performed using primary antibodies against SQS (1:2000, Genetex, GTX104091) and ERα (1:1000, Cell Signaling, 8644 s) with overnight incubation at 4 °C. Following washing and incubation with the appropriate horseradish peroxidase-conjugated secondary antibodies, the signals were visualized using an enhanced chemiluminescence kit (Amersham ECL Plus™).

Cell lines

The human lung adenocarcinoma cell lines, CL1-0, CL1-5, and H1355, were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS). CL1-0/SQS and H1355/SQS cells were established by infecting cells with the pLenti-6.3-SQS virus. A549/shSQS and CL1-5/shSQS cells were infected with the SQS shRNA virus, as previously described [2]. Overexpression-stable clones were selected with 8 µg/mL blasticidin and knockdown-stable clones were selected with 2 µg/mL puromycin for 2 weeks. A549 and A549/shSQS cells were maintained in DMEM supplemented with 10% FBS and 1% PSG.

Cell viability assay

Cells were seeded in 96-well plates at a density of 2,000 cells per well in 100 µL of complete medium. After overnight incubation, different concentrations (1.23 µM, 3.70 µM, 11.11 µM, 33.33 µM, and 100 µM) of lovastatin or DMSO (as solvent control) were added to the well with 8 replicates for each concentration and incubated for an additional 72 h. Cell viability was analyzed using the MTT assay. The absorbance at 570 nm was measured using a plate reader with an excitation source.

Statistical analysis

Statistical analyses were performed using SPSS software (version 17.0; SPSS, USA). Survival rates were estimated using the Kaplan–Meier method and compared using the log-rank test. For all analyses, a P value < 0.05 was considered to indicate statistical significance. All observations were confirmed by at least three independent experiments. The results are presented as mean ± SD. Two-tailed unpaired Student’s t-tests were used for all pairwise comparisons.

Result

Upregulation of SREBF2 correlates with poor overall survival in patients with lung cancer

In this study, we aimed to identify the potential of targeting the MVA pathway in lung cancer and determine the overall survival rates of patients with lung cancer based on the expression of the SREBF family genes using the KM plotter (http://kmplot.com/analysis/index.php?p=background) dataset. High SREBF2 levels were significantly associated with worse overall survival in patients with lung cancer (Fig. 1A, B). The SREBF2 gene is a locus encoding sterol regulatory element-binding protein 2 (SREBP2), a transcription factor that activates the MVA pathway (Figs. 1C). We further examined the correlation between SREBF2 and MVA pathway genes in patients with lung adenocarcinoma. The mRNA levels of SREBF2 were significantly and positively correlated with the levels of 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), FDFT1 (also known as SQS), squalene epoxidase (SQLE), cytochrome P450 family 51 subfamily A member 1 (CYP51A1), methyl sterol monooxygenase 1 (MSMO1), 7-dehydro-cholesterol reductase (DHCR7), and 24-dehydrocholesterol reductase (DHCR24) in patients with lung adenocarcinoma (Fig. 1D)

Fig. 1.

SREBF2 expression is correlated with an enzyme in the mevalonate pathway in patients with lung cancer. Kaplan–Meier analysis of overall survival, according to the expression levels of SREBF1 (A) and SREBF2 (B) using Kaplan–Meier Plotter. (C) Schematic of the cholesterol pathway. (D) Representation of SREBF2 correlation with HMGCR, FDFT1, SQLE, CYP51A1, MSMO1, DHCR7, and DHCR24 levels in patients with lung adenocarcinoma (GEPIA)

Overexpression of squalene synthase induces Statin resistance by modulating Estrogen receptor alpha signaling

To determine whether SREBF2 expression is related to drug sensitivity profiles, we used the Cancer Therapeutics Response Portal Version 2 (CTRP-V2) (http://www.broadinstitute.org/ctrp/). SREBF2 expression was positively correlated with lovastatin, simvastatin, BRD-K01825520, and avicin D (Fig. 1A). The drug signature related to SREBF2 expression was then defined to be indicated by a Pearson correlation coefficient ≥ 2 (Fig. 1A). In the MVA pathway, FDFT1 (SQS) is the first enzyme in the sterol branch and can be targeted to determine the pathway for cholesterol synthesis. Previously, knockdown of this gene was found to inhibit lung cancer metastasis in vitro and in vivo [2]. As a result, we opted to focus on SQS in subsequent experiments. Briefly, we determined the effects of lovastatin on SQS overexpression in lung cancer cell lines. The MTT assay revealed that the IC₅₀ of lovastatin increased 2.5- to 3-fold in SQS-overexpressing lung cancer cells compared to that in vector-expressing cells (Fig. 2B and Fig. S1A).Conversely, SQS knockdown significantly reduced the IC₅₀ of lovastatin in lung cancer cell lines (Fig. S1B). We also determined whether SQS mediates lovastatin tolerance in lung cancer cell lines. Treatment with lovastatin combined with zaragozic acid A (as SQS inhibitor, 10 µM) reduced the IC₅₀ of lovastatin in SQS-overexpressing cells compared to that in cells treated with the vector control (Fig. 2C).

Fig. 2.

SQS-ERα axis regulates the sensitivity of lung cancer cells to lovastatin. (A) Box-and-whisker plot of SREBF2 expression correlated with drug sensitivity (Cancer Therapeutic Response Portal V2). (B) Effects of SQS overexpression on lovastatin toxicity in CL1-0 cells based on the MTT assay (72 h). Data are presented as mean ± SD; **, P < 0.01. (C) Effect of zaragozic acid A (SQS inhibitor, 10 µM) combined with lovastatin on the viability of CL1-0/SQS cells. Data are presented as mean ± SD; **, P < 0.01. (D) Pathway analysis of differential gene expression in CL1-0/vector (control) and CL1-0/SQS cells (GSE37868). (E) Effect of MPP dihydrochloride hydrate (MPP, an ERα-specific inhibitor), lovastatin, and MMP combined with lovastatin on the viability of CL1-0/SQS cells. CL1-0/SQS cells were treated with MMP (10 µM), lovastatin (20 µM), and MMP combined with lovastatin at the indicated concentrations and subjected to the cell viability assay. Data are presented as mean ± SD; **, P < 0.01

To identify the molecular mechanism by which SQS regulates lung cancer cell resistance to lovastatin, we analyzed GSE37868 (CL1-0/SQS vs. CL1-0/Vector microarray). Based on enrichment p-values, pathway analysis predicted the activation of epithelial-mesenchymal transition (EMT) regulation by the growth factor pathway and estrogen receptor signaling in SQS-overexpressing cells (Fig. 2D and Table S1). Furthermore, the transcription factors that induce Estrogen Receptor 1 (ESR1) signaling was found to be activated in SQS-overexpressing cells (Table S2). To further clarify the effect of estrogen receptor alpha (ERα) on statin sensitivity, we treated SQS-overexpressing CL1-0 cells with MPP dihydrochloride hydrate (MPP, an ERα-specific inhibitor). MPP treatment exhibited minimal cytotoxicity in both CL1-0 and CL1-0/SQS cells (Fig. S1C). Furthermore, the viability of CL1-0/SQS cells was significantly inhibited when treated with lovastatin (20 µM) combined with MPP (10 µM) compared to lovastatin alone or MPP alone (Fig. 2E). Altogether, these results suggest that targeting ERα signaling may overcome the statin-resistant phenotype caused by SQS overexpression in lung cancer.

Basal expression of cholesterol biosynthesis genes is associated with Lovastatin sensitivity in lung cancer cell lines

We proceeded to analyze the gene expression of key enzymes in the cholesterol biosynthesis pathway. Notably, a consistent positive correlation was found between genes in the cholesterol biosynthesis pathway and lovastatin sensitivity AUC, including HMGCS1, HMGCR, FDFT1, and SQLE (Fig. 3A). However, FNTA and PGGT1B, which are non-cholesterol biosynthesis enzymes, were not found to correlate with the lovastatin sensitivity AUC (Fig. 3A). Kaplan–Meier overall survival analyses using open-source lung cancer microarray datasets [28] revealed that high expression levels of HMGCS1 and FDFT1 were significantly associated with adverse outcomes, whereas HMGCR and SQLE expression had marginal or no significant association with these outcomes (Fig. 3B).

Fig. 3.

Expression of cholesterol biosynthesis genes is associated with poor overall survival in patients with lung cancer. (A) Comparison of lovastatin sensitivity AUC and expression of cholesterol biosynthesis genes in CCLE lung cancer cell lines. (B) Schematic of genes positively correlated with lovastatin sensitivity AUC (red) in the cholesterol biosynthesis pathway. Kaplan–Meier overall survival curves using publicly available lung cancer microarray datasets stratified according to HMGCS1, HMGCR, FDFT1, and SQLE expression. HR=hazard ratio

To determine whether these cholesterol biosynthesis enzymes affected metabolic events in lung cancer cell lines, we assessed the dependency score of each enzyme related to cholesterol biosynthesis. Dependency scoring was used to determine the correlation between enzyme expression levels and specific metabolites. In this study, we focused on several cholesterol biosynthesis products, including phosphatidylcholine, sphingomyelin, and ceramides (Figs. S2A-2 C). FDFT1 was identified to exhibit the most significant regulatory effect on these metabolites. However, FDFT1 features were positively correlated with statin compounds, including lovastatin and simvastatin, in the CTRP database (Fig. S2D). Therefore, the mesenchymal phenotype was more prone to respond to statin treatment, whereas cells with aberrant upregulation of cholesterol biosynthesis genes, especially FDFT1 (also known as SQS), tended to be more resistant to statin treatment.

SQS positively correlates with ERα expression in patients with non-small cell lung cancer

As shown in Fig. 1E, the ERα inhibitor (MPP) increased lovastatin sensitivity in SQS-overexpressing cells. We proceeded to determine whether SQS expression correlated with ERα in patients with lung cancer using IHC. A significant positive correlation was found between SQS and ERα expression (Spearman’s nonparametric correlation test, correlation coefficient = 0.420, P < 0.001, Table 1). Representative IHC staining for SQS and ERα in serial sections revealed correlative staining patterns in lung tumor tissues (Fig. 4A). ERα expression alone was not significantly associated with clinicopathological factors (Table S3) or survival (Fig. 4B, left). Univariate analysis of overall survival (OS) and disease-free survival (DSF) revealed that ERα expression was not a significant predictor of survival (Table S4). Moreover, in the analysis of the SQS high/ERα high subset versus the SQS low/ERα low subset, the cumulative survival analysis revealed a significant difference between patients with high SQS/ERα and those with low SQS/ERα (Figs. 4B, C, Right).

Table 1.

Correlation between levels of SQS and ERα in lung cancer

ERαa	SQSa
	Low (0,1)b	High (2,3)	Patient No.
Negative	54	30	84
Positive	8	33	41

^aDetection of the expression of SQS and ERα by immunohistochemistry

^bIntensity of immunohistochemistry staining

Fig. 4.

The combination of SQS and ERα expression predicts adverse survival in patients with non-small cell lung cancer. (A) Representative images of SQS and ERα IHC staining in serial tissue sections from patients with NSCLC. Scale bar: 100 μm. Kaplan–Meier analysis of disease-free survival (B) and overall survival (C) of 125 patients with NSCLC stratified by SQS and/or ERα expression

SQS modulates ERα expression through cholesterol replenishment in lung cancer cells

We proceeded to investigate the mechanism of SQS-activated ESR1 signaling. ERα was found to be significantly upregulated in the SQS overexpression cell lines (Fig. 5A). Conversely, ERα was significantly downregulated in SQS knockdown cells (Fig. 5B). Notably, SQS expression did not affect the mRNA levels of ESR1 in CL1-0/SQS, H1355/SQS, A549/shSQS, and CL1-5/shSQS lung cancer cells (Fig. S3). In previous studies, cholesterol-modulated lipid raft formation was revealed to promote lung cancer metastasis. Cholesterol was also reported to induce EGFR-related pathway-mediated ERRα expression, leading to EGFR-TKIS resistance in lung cancer cells [2, 29]. As a result, we opted to determine whether SQS mediated ERα expression via cholesterol. Treatment with methyl-β-cyclodextrin (MβCD, a detergent to remove cholesterol) resulted in a significant decrease in ERα expression in SQS-overexpressing cells (Fig. 5C). Moreover, lovastatin combined with MβCD significantly increased lovastatin sensitivity in SQS-overexpressing cells (Fig. 5D). Conversely, cholesterol replenishment restored the ERα expression and resistance to lovastatin in SQS knockdown cells (Fig. 5E-F). These data suggest that the upregulation of SQS increased ERα expression via cholesterol replenishment, leading to lovastatin resistance in lung cancer cells.

Fig. 5.

SQS induces ERα expression through cholesterol replenishment in the lung cancer cell lines. Western blot analysis of SQS and ERα in CL1-0/SQS, H1355/SQS, (A) A549/shSQS, and CL1-5/shSQS (B). (C) Relative expression of SQS and ERα in CL1-0/SQS cells treated with or without 10 mM MβCD for 1 h. (D) Effect of lovastatin alone or combined with MβCD (2.5 mM) on the viability of CL1-0/SQS cells. CL1-0/SQS cells were treated with MβCD combined with lovastatin at the indicated concentrations and subjected to the cell viability assay. Data are presented as mean ± SD; **, P < 0.01. (E) Effect of cholesterol replenishment on ERα enrichment of CL1-5/shSQS cells. CL1-5/shSQS cells were treated with or without cholesterol (10 µg/mL) for 6 h and subjected to western blotting analysis. (F) Effect of cholesterol replenishment on the IC₅₀ of lovastatin in CL1-5/shSQS cells treated with or without cholesterol (10 µg/mL) and subjected to the MTT assay. Data are presented as mean ± SD; **, P < 0.01

Discussion

In this study, we identified a lovastatin sensitivity gene signature by correlating basal gene expression and the lovastatin sensitivity AUC for lung cancer. This gene signature revealed that cholesterol biosynthesis-mediated activation of ERα is a critical molecular features that dictate the response to statin treatment in lung cancer. In addition, the overexpression of SQS, a committed step enzyme in the cholesterol biosynthesis pathway, was found to significant induce ERα expression and increase tolerance to lovastatin treatment. Treatment of SQS overexpression cells with the ERα inhibitor increased sensitivity to lovastatin, confirming that SQS-driven lovastatin resistance is dependent on ERα expression. Moreover, a positive association between SQS and ERα, and the detrimental impact of their combined expression on disease-free survival. Therefore, the combination of SQS and ERα status may serve as a powerful prognostic predictor for patients with lung cancer.

Previous studies showed that targeting cholesterol synthesis with statins not only reduces NSCLC aggressiveness but also transforms immuno-cold tumors to inflamed ones, enhancing their response to immune checkpoint blockade (ICB) therapy through mechanisms including the transcriptional inhibition of PD-L1 and induction of ferroptosis in NSCLC cells [30]. Moreover, Guo et al.. demonstrated that chemo-resistant SCLC exhibits metabolic reprogramming that relies on the mevalonate-geranylgeranyl diphosphate (MVA-GGPP) pathway, which can be effectively targeted using statins to induce oxidative stress and apoptosis, overcoming chemoresistance in vivo and showing promising results in combination with chemotherapy in clinical cases [31]. Statin sensitivity was suggested to be mediated by inhibiting the prenylation of RAS family proteins, which are the dominant driver mutations in many types of cancer [15, 17, 18, 32]. However, some studies have shown that the intrinsic sensitivity of cancer cell lines to statin treatment is independent of RAS function [33–35]. Moreover, clinical trials and epidemiological studies do not support this association as RAS mutations fail to predict statin sensitivity [36–39]. These conflicting results suggest that the inhibition of RAS protein prenylation is not the major contributing factor to the determination of statin sensitivity when targeting cancer.

One of the significant differences between statin-sensitive and -resistant lung cancer cells at basal gene expression is the expression level of genes in the cholesterol biosynthesis pathway. Following statin treatment, resistant cells could strongly induce the expression of genes in this pathway via a normal negative feedback loop resulting from the statin-induced inhibition of HMGCR, the rate-limiting enzyme and a key regulator of the MVA pathway [40]. However, this feedback regulation was weaker in sensitive cells, suggesting that these cells may have defects in the regulatory mechanisms of this pathway. The inability to induce cholesterol biosynthesis pathway genes in response to statin treatment has been reported in statin-sensitive multiple myeloma cells [41]. Therefore, this cholesterol biosynthesis signature may be a common feature across many types of cancer and may represent a useful biomarker for predicting statin sensitivity.

The estrogen receptor antagonist fulvestrant is an effective compound in reducing mesenchymal features of lung carcinoma cells, sensitizing them to immune- and chemotherapy-mediated lysis, suggesting a role for estrogen signaling in promoting tumor resistance and supporting further investigations into its role in lung cancer progression [42]. In the present study, overexpression of the step-committed enzyme in the cholesterol biosynthesis pathway, SQS, led to significant upregulation of ERα and was associated with increased IC₅₀ upon statin treatment. Furthermore, the combination of an ERα inhibitor abolished this resistance (Fig. 1E). Cholesterol replenishment restored ERα expression and increased the IC₅₀ of lovastatin in SQS knockdown cells (Fig. 5E-F). These data suggest that targeting the SQS-ERα axis may be a more efficient strategy to eliminate statin-resistant cells. Consistently, Pan et al. also showed cholesterol mediated ERRα expression leading to EGFR‑TKIs resistance in lung cancer cell lines. Lovastatin combined with XCT790 (an ERRα selective inverse agonist) increased EGFR-TKI sensitivity in vitro. In animal models, gefitinib combined with lovastatin or XCT790 inhibited xenograft tumor growth [29]. Liang et al. showed that the combination of simvastatin and tamoxifen significantly increases the percentage of apoptosis of tamoxifen-resistant breast cancer cells by suppressing the DNA replication licensing factor, MCM7 [43]. Similarly, Nguyen et al. demonstrated that long-term estrogen deprivation of ERα-positive breast cancer cells activates the cholesterol biosynthesis pathway by epigenetic reprogramming coupled with an enrichment of the SREBP DNA-binding motif [44]. This epigenetic activation of the cholesterol biosynthesis pathway increased the levels of 27-hydroxycholestrol, which in turn activated ERα signaling and promoted cancer progression [44]. In our lung cancer study, this SQS-ERα axis was found to be active in both male and female lung tumors. Previous studies in a transgenic mouse model using an estrogen-responsive element (ERE)-luciferase reporter revealed that in addition to the reproductive tract, lung tissues had significant induction of luciferase activity upon 17β-estradiol treatment in both male and female mice [45, 46]. In human lung adenocarcinoma, both ERα and ERβ are highly expressed and responsive to estrogen treatment [47, 48]. Moreover, hormone replacement therapy has been reported to accelerate disease progression in lung cancer, while estrogen antagonist therapy has been shown to improve patient survival [49–51]. Estrogen signaling has been linked to the maintenance of the epithelial phenotype by suppressing EMT in hormone-related cancers [52, 53].

Although statin-based therapy is an attractive and economical method for the clinical management of lung cancer, careful selection of patients for treatment is warranted. In addition to demonstrating SQS-mediated ERα expression and treatment, lovastatin combined with MPP (an ERα-specific inhibitor) or zaragozic acid A (SQS inhibitor) in SQS overexpression lung cancer cells significantly increased sensitivity to lovastatin. Our data suggest that the SQS-ERα axis can be used to identify patients who will most likely benefit from the combination of cholesterol-lowering medications and antiestrogen therapy for controlling lung cancer. This study has some limitations. First, while we demonstrated that SQS upregulates ERα via cholesterol replenishment and confers resistance to lovastatin, in vivo validation and mechanistic dissection of this axis in clinical specimens remain to be fully explored. Second, although we observed co-expression of SQS and ERα in NSCLC tissues, prospective clinical data linking this axis to statin treatment outcomes are lacking. Future studies should assess whether the SQS–ERα signature can serve as a predictive biomarker in statin-treated cohorts, and whether dual inhibition of cholesterol biosynthesis and ERα signaling enhances therapeutic efficacy. Furthermore, the potential interplay between SQS–ERα signaling and EMT-related pathways warrants deeper investigation to clarify their respective roles in drug sensitivity.

Abbreviations

HMGCR

Hydroxymethylglutaryl coenzyme A reductase

ERα

Estrogen receptor alpha

FDFT1

Farnesyl-diphosphate farnesyltransferase 1

SQS

Squalene synthase

EMT

Epithelial-to-mesenchymal transition

SCLC

Small cell lung cancer

MVA

Mevalonate

FPP

Farnesyl pyrophosphate

GGPP

Geranylgeranyl pyrophosphate

AUC

Area under curve

GSEA

Gene set enrichment analysis

WHO

World Health Organization

AJCC

American Joint Committee on Cancer

H&E

Hematoxylin and eosin

CCLE

Cancer Cell Line Encyclopedia

IHC

Immunohistochemistry

Author contributions

The authors contributed in the following ways: designing and writing the manuscripts: Yi-Fang Yang, Yi-Hua Jan, and Michael Hsiao; providing materials: Chih-Jen Yang, and Ming-Shyan Huang; performing experiments: Ming-Hsien Chan, Yi-Hua Jan, and Yi-Fang Yang; interpret data: Yi-Hua Jan, Yi-Fang Yang, Yu-Chan Chang, Min-Hsi Lin, and Michael Hsiao; Study supervision: Michael Hsiao.

Funding

This study was supported by Genomics Research Center funds to M.H, Kaohsiung Veterans General Hospital, Taiwan [KSVGH110-144], [KSVGH111-148], and [KSVGH112-150] to Y.F.Y, VGH, TSGH, AS Joint Research Program [VTA111-A-3-2, VTA113-A-3-3 (Y.F.Y) and AS-VTA-111-15 (M.H)], National Science and Technology Council, Taiwan (NSTC 110-2314-B-075B-009 -MY3, 113-2314-B-075B-002, and 114-2314-B-075B-014 -MY3) to Y.F.Y.

Data availability

The datasets used and analyzed during this study are available from the corresponding authors upon reasonable request.

Declarations

Ethics approval and consent to participate

The NSCLC samples were recruited with the Kaohsiung Medical University Hospital with IRB approval (KMUHIRB-E(I)-20160099) by the ethics committees of Kaohsiung Medical University Hospital.

Consent for publication

All the authors have read and approved the final article.

Competing interests

The authors declare no competing interests.

Conflict of interest

The authors declare that they have no conflict of interest.