Skip to main content
Download PDF

Investigating the mechanisms by which low NAT1 expression in tumor cells contributes to chemo-resistance in colorectal cancer

Clinical Epigenetics volume 17, Article number: 77 (2025) Cite this article

Abstract

Background

In the therapeutic landscape of colorectal cancer (CRC), chemo-resistance poses a significant and prevalent obstacle that complicates treatment efficacy and patient outcomes. Over time, cancer cells can develop mechanisms to resist the toxic effects of chemo-therapy drugs, leading to reduced sensitivity or complete insensitivity to these agents. The enzyme Arylamine N-acetyltransferase 1 (NAT1) has emerged as a promising target in strategies aimed at overcoming this challenge. NAT1 is involved in the metabolism of various xenobiotics, including some chemotherapeutic agents. Understanding the complex interactions between NAT1 and chemotherapeutic agents, as well as the molecular mechanisms underlying chemo-resistance, is crucial for the development of novel therapeutic approaches.

Objective

This study aimed to assess the role of NAT1 in mediating chemo-resistance in CRC, with the goal of identifying novel strategies to overcome this clinical challenge.

Methods

We conducted a comprehensive analysis using various bioinformatics tools and in vitro experiments to evaluate the effect of NAT1 expression on chemo-resistance in CRC. Furthermore, we employed a multi-omics approach, including metabolomics and next-generation sequencing, to uncover the mechanisms by which NAT1 influences chemo-resistance. Additionally, we utilized single-cell RNA sequencing (scRNA-seq), the Cellchat assay, and western blot to explore the intercellular communication between tumor and endothelial cells in the context of anti-PD-1 therapy and NAT1’s impact.

Results

Our study reveals that decreased NAT1 expression in CRC tumor tissues, relative to adjacent normal tissues, is significantly associated with a poorer patient prognosis. Experimental data indicate that silencing NAT1 in CaCO2 and HCT116 cell lines results in heightened resistance to five chemotherapeutic agents: vinblastine, docetaxel, gemcitabine, vincristine, and daporinad. Additionally, NAT1 silencing increases the proportion of LGR5+ cells, which are known to be chemo-resistant. Our research further revealed that exposure to these five drugs induces a decrease in NAT1 expression within CRC cells. Mechanistic insights show that NAT1 knockdown triggers a metabolic reprogramming in CRC cells, shifting from oxidative phosphorylation and the tricarboxylic acid cycle to a preference for glycolysis. Furthermore, silencing of NAT1 in CRC cells leads to an up-regulation of VEGFA expression. Notably, the application of anti-PD-1 therapy was demonstrated to significantly disrupt the VEGFA-VEGFR axis signaling, an interaction critical between CRC cells and endothelial cells. This discovery underscores the potential of targeting the VEGFA pathway as a therapeutic approach to mitigate the adverse effects associated with NAT1 down-regulation in CRC.

Conclusion

Our study underscores the multifaceted role of NAT1 in modulating chemo-sensitivity, cellular metabolism, and angiogenesis in CRC. These findings position NAT1 as a compelling candidate for a biomarker and a potential therapeutic target, offering new avenues for CRC management.

Graphical abstract

Introduction

Colorectal cancer (CRC) stands as a major challenge in the global oncology arena, marked by its high incidence and mortality rates [1]. The progress in chemo-therapy has been notable, yet drug resistance remains a critical barrier for many CRC patients, often resulting in treatment failure and an unfavorable prognosis [2]. Addressing chemo-resistance is crucial for enhancing the outcomes of CRC patients.

Aromatic amine N-acetyltransferase 1 (NAT1) has been recognized for its role in drug metabolism and has recently been implicated in cellular processes such as cell proliferation, apoptosis, and DNA repair [3, 4]. Studies suggest a correlation between higher NAT1 levels and improved patient survival rates, as well as increased responsiveness to chemo-therapy [5], indicating its potential in enhancing the efficacy of cancer treatments. Cancer stem cells (CSCs) are known for their self-renewal capabilities and resistance to chemo-therapy [6, 7]. which is attributed to robust DNA repair, drug efflux pump activity, and the tumor microenvironment (TME) [8]. The hypothesis that NAT1 could modulate CSCs’ formation, thereby impacting chemo-resistance, is an unexplored area in CRC.

Chemo-resistance is a multifaceted phenomenon arising from interactions within tumor cells and their surrounding TME [9], which is a complex milieu of various cell types [10]. The influence of NAT1 expression in tumor cells on these other TME constituents is not yet fully understood and could represent a key mechanism through which NAT1 influences chemo-resistance. Additionally, while immunotherapy targeting the programmed cell death 1 (PD-1) pathway has shown clinical benefits in various cancers [11], the connection between NAT1 expression and the efficacy of anti-PD-1 therapies remains unclear, highlighting the need for further research into these mechanisms.

This study is undertaking an extensive and in-depth analysis of NAT1’s influential role in chemo-resistance, exploring the spectrum of its mechanisms within the cellular and extracellular environments. Our aim is to pioneer transformative therapeutic strategies for CRC patients, with the vision of providing effective solutions to combat chemo-resistance.

Materials and methods

Public datasets

RNA sequencing (RNA-seq) data and corresponding clinical profiles for colon adenocarcinoma (COAD) and rectal adenocarcinoma (READ) were extracted from The Cancer Genome Atlas (TCGA) database.

In-house samples

Between January 2021 and September 2021, a total of 20 pairs of adjacent normal tissues and CRC tumor tissues were collected at Suzhou Municipal Hospital. Written informed consent was secured from all participating patients prior to the commencement of the study. The study protocol, including all experiments, was reviewed and approved by the Ethics Committee of Suzhou Municipal Hospital.

RNA-seq

Total RNA from ten pairs of adjacent normal tissues and CRC tumor tissues was isolated using the PROTRIZOL reagent (PRN02, Proteinbio, Nanjing, China). Subsequently, complementary DNA (cDNA) libraries were constructed and sequenced on the Illumina platform at Metware Biotechnology Co., Ltd (Wuhan, China).

Differential analysis

We conducted differential expression analysis utilizing the "Limma" package (version 3.40.6) in R software. Briefly, we employed the "Voom" method for data transformation. Subsequently, a multiple linear regression analysis was conducted using the lmFit function. We then computed the moderated T-statistics, moderated F-statistics, and log odds ratios for differential expression, employing empirical Bayesian moderation to stabilize variance estimates. Genes with a FC ≥ 2 and a P-value ≤ 0.05 were designated as differential expressed genes (DEGs).

Prognostic analysis

The TCGA-CRC dataset was used for prognostic analysis. For overall survival (OS), the R package maxstat (Maximally selected rank statistics with several p-value approximations, version:0.7–25) were used to calculate the optimal truncation value of NAT1, set the minimum sample number of the group to be greater than 25%, the maximum sample number of the group to be less than 75%, and finally obtained the optimal truncation value is: 2.9654. For progression-free survival (PFS), the optimal truncation value of NAT1 is 2.7803. For Disease-free survival (DFS), the optimal truncation value of NAT1 is 2.9654. Based on this, the patients were divided into NAT1high and NAT1low groups separately. The survfit function of R software package survival was further used to analyze the prognostic difference between the two groups. Logrank test was used to evaluate the significance of the prognostic difference between samples from different groups, and a significant prognostic difference was finally observed.

Drug sensitivity analysis

The median inhibition rate (IC50) of the drugs were predicted using the oncoPredict package (version 12) in conjunction with the Genomics of Drug Sensitivity in Cancer (GDSC) database, for both Version 1 and Version 2 [12]. This analysis was based on the RNA-seq data matrix derived from TCGA-CRC database.

Functional enrichment analysis

For KEGG pathway analysis, we harnessed the KEGG REST API available at https://www.kegg.jp/kegg/rest/keggapi.html to retrieve the most up-to-date KEGG data. Utilizing the gene annotations from the pathways as a reference background, we conducted enrichment analysis with the R package clusterProfiler, version 3.14.3 [13]. We established a minimum gene set threshold of 5 and a maximum of 5000, considering P-value < 0.05 as indicative of statistical significance.

Least absolute shrinkage and selection operator (LASSO)

We employed the R package glmnet (version 4.1.8) [14] for the integration of survival time, status, and gene expression data, performing regression analysis via the lasso-cox method. Additionally, we implemented tenfold cross-validation to ascertain the optimal model parameters.

Cell culture

CaCO2 and HCT116 cell lines were sourced from Zhongqiao Xinzhou Biotechnology Co., LTD (Shanghai, China). Both cell lines were cultured in their respective complete media—DMEM for CaCO2 and 1640 for HCT116—each supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin. The cultures were maintained in a humidified 37 °C incubator with an atmosphere of 5% CO2.

Cell transfection

Lentivirus that over-expressing NAT1 (OE-NAT1) or silencing NAT1 (sh-NAT1), and the corresponding negative control (NC) were obtained from Fenghui Biotechnology Co., LTD (Changsha, Hunan province, China). Polybrene (Beyotime Biotechnology, Shanghai, China) was used to transfect lentivirus in cells.

Metabolomics

The metabolomics was conducted in Metware Biotechnology Co., Ltd (Wuhan, China). Briefly, Liquid Chromatography Tandem Mass Spectrometry (LC–MS/MS) was used to determine the metabolome of CaCO2 and HCT116 cells. The data acquisition instrument system mainly includes Ultra Performance Liquid Chromatography (UPLC) (ExionLC AD). https://sciex.com.cn/) and Tandem mass spectrometry (MS/MS) (QTRAP®, https://sciex.com.cn/).

IC50 detection

Vinblastine (HY-13780, molecular weight: 909.05 Da, purity: 99.86%), docetaxel (HY-B0011, molecular weight: 807.88 Da, purity: 99.94%), gemcitabine (HY-17026, molecular weight: 263.20 Da, purity: 99.96%), vincristine (HY-N0488, molecular weight: 923.04 Da, purity: 99.81%), and daporinad (HY-50876, molecular weight: 391.51 Da, purity: 99.72%) were obtained from MedChemExpress (Shanghai, China). These five drugs were each solubilized in dimethyl sulfoxide (DMSO) to prepare stock solutions. CaCO2 and HCT116 cells (5*103) were cultured in 96-well for 24 h. Then cells were treated with 0, 2, 4, 6, 8 micromolar (μmol) vinblastine, docetaxel, gemcitabine, vincristine, and daporinad for 24 h, respectively. Finally, 96-well was cultured with 10μL CCK-8 (PC001-1ML, proteinbio) and 90μL serum-free DMEM medium for 2 h a humidified 37 °C incubator with an atmosphere of 5% CO2. The OD values were detected by microplate reader at 450 nm.

Cell apoptosis

CaCO2 and HCT116 cells (5*105) were cultured in 6-well for 24 h. Then cells were treated with drugs or DMSO for 48 h, respectively. SuperViewTM 488 Caspase-3 Assay Kit (PS6007-25T, proteinbio) and APC Annexin V (640,920, Biolegend) was used to detect the apoptosis of cells using a flow cytometry. SuperView 488: Ex/Em = 500/530 nm. APC: Ex/Em = 640/675 nm.

Detection of CSC markers

APC anti-human CD44 antibody (C44Mab-5, BioLegend), APC anti-human CD133 antibody (W6B3C1, BioLegend), APC anti-human CD166 antibody (343,906, BioLegend), and APC anti-human LGR5 (GPR49) antibody (SA222C5, BioLegend) were used to detect the proportions of different phenotypic CSCs in cells via using the APC channel of a flow cytometer.

Real-time quantitative PCR (qRT-PCR)

Total RNA extraction from both cells and tissues was performed using the PROTRIZOL Reagent. The extracted total RNA exhibited concentration levels of 1700 to 3300 ng/μL for tissue samples and 800 to 1600 ng/μL for cell samples. Total RNA (1000 ng) was reverse-transcribed into cDNA using the PrimeScript™ RT reagent Kit (RR037A, Takara), according to the manufacturer’s protocol. Reverse transcription was carried out under the following conditions: incubation at 42 °C for 15 min to synthesize cDNA, followed by an enzyme inactivation step at 85 °C for 5 s. The reaction mixture was then held at 4 °C for storage or subsequent processing. qRT-PCR was performed using 2 × SYBR Green qPCR MixWith 100 × ROX (PC5902, proteinbio). The Applied Biosystems 7500 Fast Real-Time PCR System was used to perform two-step real-time PCR amplification following a standard protocol: an initial denaturation at 95 °C for 30 s, followed by 40 cycles of PCR amplification with denaturation at 95 °C for 3 s and annealing/extension at 60 °C for 30 s. After the amplification cycles, a dissociation stage was included to verify the specificity of the PCR products through melting curve analysis. The primers used in this experiment are shown in Table 1.

Table 1 The primers used in this experiment
Full size table

Western blot (WB)

Total protein was extracted with RIPA (PWB006, proteinbio) and PMSF (PWB507, proteinbio). Beta Actin Monoclonal antibody (Cat No: 66009-1-Ig, proteintech), NAT1 Monoclonal antibody (Cat No. 67942-1-Ig, proteintech) and VEGFA Monoclonal antibody (Cat No. 66828-1-Ig, proteintech) was used as the primary antibody. UltraPolymer Goat anti-Mouse IgG (H&L)-HRP (Cat No. PR30012, proteintech) or UltraPolymer Goat anti-Rabbit IgG (H&L)-HRP (Cat No. PR30011, proteintech) was used as the secondary antibody. Enhanced ECL luminescent solution (PECL08, proteinbio) and an imaging system (Tanon-5200, Shanghai, China) was used for the detection of proteins.

Molecular docking

The 2D structure of vinblastine, docetaxel, gemcitabine, vincristine and daporinad was obtained from Pubchem database (https://pubchem.ncbi.nlm.nih.gov/). The 3D structure of NAT1 (ID:2PQT) was obtained from PDB database (https://www.rcsb.org/structure/2PQT). AutoDockTools-1.5.7 was used to predict the binding of NAT1 and these five chemo-therapy drugs. The binding sites of NAT1 and these five chemo-therapy drugs were found and shown via PyMol (version 3.11.2).

Single cell RNA-seq (scRNA-seq)

GSE231559 and GSE205506 were obtained from the Gene Expression Omnibus (GEO) database. The GSE231559 dataset included 11 adjacent normal tissues and 15 tumor tissues from CRC patients. The GSE205506 included 10 tumor tissues from CRC patients (without anti-PD-1 treatment) and 14 tumor tissues from CRC patients (anti-PD-1 treatment). The R software package Seurat (5.0.1) [15] was used to analyze scRNA-seq. In short, nFeature_RNA > 200 & nFeature_RNA < 2500 & percent.mt < 15 was set to filter the cell. First, we used tSNE for cell clustering. Then, the HumanPrimaryCellAtlasData-based R software package singleR (2.4.1) [16] was used for cell annotation, followed by the CellMarker database [17] for manual annotation of different clusters. Finally, cellchat (1.6.1) [18] was used for the analysis of cell–cell communication.

Cellchat analysis

The dataset was loaded into R using the readRDS function. The cell types were identified and listed using the command levels(cellchat@idents). The number of cells in each type was calculated and displayed with groupSize <—as.numeric(table(cellchat@idents)). Original cell types were checked by running unique(cellchat@meta$seurat_clusters). For human data analysis, the CellChatDB.human database was used (CellChatDB.mouse would be selected for mouse studies). The category of ligand-receptor interactions in the database was shown using showDatabaseCategory (CellChatDB). Unused factor levels in cellchat@idents were dropped to refine the dataset. The CellChatDB was assigned to cellchat@DB, and data was subsetted appropriately using subsetData (cellchat). Over-expressed genes and interactions were identified through identifyOverExpressedGenes (cellchat) and identifyOverExpressedInteractions (cellchat), respectively. Gene expression data was then projected onto the protein–protein interaction network PPI.human via projectData (cellchat, PPI.human). To infer the cell–cell communication network, the probability of interactions was computed with computeCommunProb(cellchat). Interactions involving fewer than one cell were filtered out using filterCommunication(cellchat, min.cells = 10). Pathway-level interaction probabilities were summarized from ligand-receptor pairs through computeCommunProbPathway (cellchat). An aggregated cell interaction network was generated using aggregateNet (cellchat), highlighting significant pathways via cellchat@netP$pathways. Network centrality scores for signaling pathways were calculated using netAnalysis_computeCentrality (cellchat, slot.name = "netP"), providing insights into the importance of different pathways within the cell–cell communication network.

Statistical analysis

Data were analyzed using statistical analysis software including GraphPad Prism 8.0 software and R software (4.3.2). Data were expressed as mean ± standard deviation. An unpaired t-test, paired t-test, or one-way ANOVA was used to measure differences between groups. Statistical significance was set at P < 0.05.

Results

Low NAT1 expression is associated with a worse prognosis in CRC patients

We performed a comprehensive analysis of datasets from TCGA-COAD and TCGA-READ (Supplemental Figure 1A) in conjunction with our in-house CRC RNA-seq data (Fig. 1A) to discern DEGs between adjacent normal tissues and CRC tumor tissues. A consistent core of 70 DEGs was discerned across all cohorts (Supplemental Figure 1B, Supplementary material 1.xls). Functional enrichment analysis underscored the substantial involvement of these DEGs in metabolic processes, the PI3K-Akt signaling pathway, the MAPK pathway, and additional signaling cascades (Supplemental Figure 1C). LASSO analysis underscored a robust association between NAT1 and CD79A with respect to overall survival (Supplemental Figure 1D). The Kaplan–Meier survival analysis revealed a significant association between reduced NAT1 expression and poorer outcomes in CRC patients, as evidenced by worse overall survival (P = 0.00046), disease-free survival (P = 0.00075), and progression-free survival (P = 0.0013) (Fig. 1B). Similarly, elevated CD79A expression was linked to adverse survival metrics, including overall survival (P = 0.016), disease-free survival (P = 0.017), and progression-free survival (P = 0.032) (Supplemental Figure 1E). Notably, diminished NAT1 expression demonstrated a more robust correlation with unfavorable prognosis in comparison to increased CD79A levels, underscoring its potential significance as a prognostic marker in CRC. Additionally, NAT1 expression was significantly downregulated in CRC tumor tissues relative to adjacent normal tissues, as evidenced by FC of 0.29 (P < 0.0001) in the TCGA-COAD dataset and FC of 0.31 (P < 0.0001) in the TCGA-READ dataset (Fig. 1C). This trend was further validated by our in-house RNA-seq analysis, which reported a FC of 0.06 (P = 0.002) (Fig. 1D), and by qRT-PCR results, showing a FC of 0.47 (P = 0.0014) (Fig. 1E). The FC values indicate the ratio of NAT1 expression in tumor tissues to that in normal tissues, with values less than 1 indicating downregulation in tumor tissues. These consistent findings across multiple platforms highlight the marked reduction in NAT1 expression in CRC tumor tissues compared to normal tissues.

Fig. 1
figure 1

Lower NAT1 expression is associated with a poorer prognosis in CRC patients. A Identification of differential expressed genes (DEGs) through RNA sequencing between adjacent normal tissues (N = 10) and CRC tumor tissues (N = 10). Genes with a fold change (FC) ≥ 2 and a P-value ≤ 0.05 were designated as DEGs using the R package "limma-voom". B Kaplan–Meier analysis of overall survival, disease-free survival, and progression-free survival was used to compare the prognosis of CRC patients based on the optimal cut-off value of NAT1 expression level. Log-rank test was used to calculate the P value. C In both the TCGA-COAD (N = 458 tumor, N = 41 normal) and TCGA-READ (N = 168 tumor, N = 10 normal) datasets, NAT1 expression was significantly down-regulated in tumor tissues compared to normal tissues, as determined by unpaired t-tests. D and E The down-regulation of NAT1 in CRC tumor tissues (n = 10) compared to adjacent normal tissues (n = 10) was validated using in-house RNA sequencing, as assessed by paired t-test. Similarly, qRT-PCR analysis confirmed the down-regulation of NAT1 in CRC tumor tissues (n = 20) compared to adjacent normal tissues (n = 20), also evaluated by paired t-test

Full size image

Low NAT1 expression enhances the resistance of multiple chemotherapeutic drugs in CRC

The IC50 values of drugs were predicted using the oncoPredict package (version 12) in conjunction with the GDSC database, covering both Version 1 and Version 2. This analysis utilized the RNA-seq data matrix from the TCGA-CRC database (N = 592). When comparing the NAT1high group (N = 296) to the NAT1low group (N = 296), we observed a significant reduction in the relative IC50 values for 6 compounds in the GDSC_V1 database and for 22 compounds in the GDSC_V2 database (Table 2). These findings suggest that lower NAT1 expression may enhance the chemo-resistance of CRC to multiple chemo-therapy agents. Based on the FC of relative IC50 values to chemo-therapy drugs in NAT1low and NAT1high groups, we prioritized the top 5 chemo-therapy drugs exhibiting the largest FC difference for further experiments. These drugs were vinblastine (FC = 0.473), docetaxel (FC = 0.472), gemcitabine (FC = 0.441), vincristine (FC = 0.425), and daporinad (FC = 0.397) (Table 2). The correlation between NAT1 and these top five chemo-therapy drugs was validated using Spearman correlation analysis (N = 592) (Fig. 2A). We modulated NAT1 expression in CaCO2 and HCT116 cells, either by silencing or over-expressing NAT1 (Fig. 2B). The CCK-8 assay results revealed that altering NAT1 expression levels significantly impacted the IC50 values of CaCO2 and HCT116 cells for five chemotherapy drugs: vinblastine, docetaxel, gemcitabine, vincristine, and daporinad. Specifically, sh-NAT1 increased the IC50 values compared to sh-NC, indicating enhanced drug resistance, while OE-NAT1 decreased the IC50 values relative to OE-NC, suggesting increased drug sensitivity (Fig. 2C). Additionally, we treated CaCO2 and HCT116 cells with respective IC50 concentrations of each drug for 48 h, the results of flow cytometry analysis demonstrated that NAT1 silencing attenuated the apoptotic effects of chemo-therapy drugs on CaCO2 and HCT116 cells. Conversely, NAT1 over-expression potentiated this apoptotic effect. These observations were supported by the flow detection of annexin-V and caspase-3 (Fig. 2D). Additionally, cells were treated with 5 ng of each compound (equivalent to vinblastine 0.55 μmol, docetaxel 0.62 μmol, gemcitabine 1.90 μmol, vincristine 0.54 μmol, and daporinad 1.28 μmol) for 24 h. This was followed by a 24-h incubation period in fresh complete medium. The treatment cycle, consisting of 24 h of compound exposure and 24 h of recovery in fresh medium, was repeated five times. Notably, these five chemotherapeutic agents exhibited inhibitory effects on NAT1 expression in CaCO2 and HCT116 cells, as confirmed by WB assays (Fig. 2E).

Table 2 Comparison of relative IC50 between NAT1high and NAT1low groups
Full size table
Fig. 2
figure 2

Silencing NAT1 enhances the chemo-resistance of chemotherapeutic agents in CRC cells. A Using the oncoPredict package, we calculated the relative IC50 values for the top five chemotherapeutic agents in each sample based on RNA-seq data from TCGA-CRC (n = 592). Spearman correlation analysis was conducted to evaluate the relationship between NAT1 expression and the relative IC50 values of these top five chemotherapeutic agents. B The expression levels of NAT1 in each group were assessed using Western blot (WB) assays. C The IC50 values (uM) for each group were determined using the CCK-8 assay. D Flow cytometry was employed to quantify the mean fluorescence intensity of annexin-V and caspase-3 in each group. E Cells were exposed to DMSO or 5 ng of each compound—equivalent to 0.55 μM vinblastine, 0.62 μM docetaxel, 1.90 μM gemcitabine, 0.54 μM vincristine, and 1.28 μM daporinad—for an initial 24-h treatment period. This was followed by a 24-h recovery phase in fresh complete medium. The cycle of 24 h of drug exposure followed by 24 h of recovery in fresh medium was repeated for a total of five cycles. WB analysis revealed that this treatment regimen led to a downregulation of NAT1 expression in CRC cells. Statistical significance is denoted as ns represents no significance, *P < 0.05. μM represents micromolar. Comparison of sh-NC with sh-NAT1 and OE-NC with OE-NAT1 via unpaired t-test. The experiments were independently repeated at least three times

Full size image

Suppressing NAT1 enables cancer cells to evade the lethal effects of multiple chemotherapeutic agents in CRC

Evidence from both in vitro and in vivo studies supports the notion that CSCs exhibit a high degree of chemo-resistance [19, 20]. In this study, we have selected four markers known to be associated with CSCs, specifically CD44, CD133, CD166, and LGR5. The aim was to investigate the relationship between NAT1, CSCs, and chemo-resistance in CRC. CaCO2 cells were subjected to a 48 h exposure to 1.68 μmol docetaxel or DMSO (Fig. 3A, Supplement Figure 2A). In comparison to the DMSO group, a noteworthy increase was observed in the proportion of cells expressing CD44+ (FC = 2.36, P < 0.0001), CD133+ (FC = 2.43, P < 0.0001), CD166+ (FC = 2.80, P < 0.0001), and LGR5+ (FC = 2.89, P < 0.0001) markers in the 1.68 μmol docetaxel-treated group. Among these markers, the LGR5+ CaCO2 cells demonstrated the most significant elevation (Fig. 3A, Supplement Figure 2A). Subsequently, CaCO2 cells were treated with 8.52 μmol vinblastine, 1.68 μmol docetaxel, 5.49 μmol gemcitabine, 10.94 μmol vincristine, and 11.71 μmol daporinad or DMSO for 48 h. HCT116 cells were treated with 4.87 μmol vinblastine, 1.68 μmol docetaxel, 5.49 μmol gemcitabine, 10.94 μmol vincristine, and 11.71 μmol daporinad or DMSO for 48 h. Compared to the DMSO control group, a significant increase in the proportion of LGR5+ cells was observed in CaCO2 cells treated with chemotherapy drugs: vinblastine (FC = 4.49, P < 0.0001), docetaxel (FC = 2.89, P < 0.0001), gemcitabine (FC = 3.34, P < 0.0001), vincristine (FC = 6.85, P < 0.0001), and daporinad (FC = 4.39, P < 0.0001). Similarly, in HCT116 cells, the proportions of LGR5+ cells were markedly elevated following treatment with the same drugs: vinblastine (FC = 8.89, P < 0.0001), docetaxel (FC = 12.60, P < 0.0001), gemcitabine (FC = 15.06, P < 0.0001), vincristine (FC = 4.77, P < 0.0001), and daporinad (FC = 11.86, P < 0.0001) (Fig. 3B). Additionally, our research demonstrated that suppressing NAT1 expression led to a significant increase in the proportion of LGR5+ cells: in CaCO2 cells (FC = 2.352, P < 0.0001) and HCT116 cells (FC = 4.620, P < 0.0001). Conversely, overexpressing NAT1 resulted in a notable decrease in LGR5+ cell proportions: in CaCO2 cells (FC = 0.446, P < 0.0001) and HCT116 cells (FC = 0.653, P < 0.0001) (Fig. 3C).

Fig. 3
figure 3

Silencing NAT1 enhances chemo-resistance in CRC by promoting LGR5+ CSC formation. A Flow cytometry was used to quantify the proportions of CD44+, CD133+, CD166+, and LGR5+ cell subpopulations in CaCO2 cells treated with 1.68 μmol docetaxel or DMSO for 48 h. B CaCO2 cells were treated with 8.52 μmol vinblastine, 1.68 μmol docetaxel, 5.49 μmol gemcitabine, 10.94 μmol vincristine, and 11.71 μmol daporinad or DMSO for 48 h. HCT116 cells were treated with 4.87 μmol vinblastine, 1.68 μmol docetaxel, 5.49 μmol gemcitabine, 10.94 μmol vincristine, and 11.71 μmol daporinad or DMSO for 48 h. Flow cytometry was used to quantify the proportion of LGR5+ subpopulation in CaCO2 and HCT116 cells. C Compared to their respective control groups, the proportion of LGR5+ cells was significantly increased in the sh-NAT1 group, whereas it was significantly decreased in the OE-NAT1 group. Statistical significance is denoted as ns represents no significance, *P < 0.05. Comparison of sh-NC with sh-NAT1 and OE-NC with OE-NAT1 via unpaired t-test. The experiments were independently repeated at least three times

Full size image

NAT1 regulates metabolism in CRC

To gain a deeper understanding of the underlying mechanisms involved in NAT1, RNA-seq analysis was conducted. Specifically, we examined the heterogeneity between the sh-NC (N = 3) and sh-NAT1 (N = 3) groups in CaCO2 and HCT116 cells (Fig. 4A). Our findings identified a cluster of 1756 DEGs in CaCO2 and HCT116 cells (Fig. 4B, Supplementary material 2.xls). Functional enrichment analysis revealed that these DEGs were primarily enriched in metabolic pathways (Fig. 4C). The formation of CSC is mainly influenced by metabolic processes including oxidative phosphorylation, tricarboxylic acid cycle (TCA) and glycolysis [21, 22]. Functional enrichment analysis revealed that these DEGs were also enriched in oxidative phosphorylation, TCA and glycolysis pathways (Fig. 4D). Utilizing the qRT-PCR assay, we observed that the overexpression of NAT1 led to an increase in the expression of oxidative phosphorylation related genes NDUFB6 (FC = 1.74, P < 0.0001), NDUFS7 (FC = 2.15, P < 0.0001), NDUFB3 (FC = 2.06, P < 0.0001), and ATP6V0D2 (FC = 2.09, P = 0.00012) and the TCA related genes ACLY (FC = 1.74, P = 0.000372), MDH2 (FC = 2.06, P = 0.000123), SDHA (FC = 2.07, P < 0.0001), ACO1 (FC = 2.21, P < 0.0001), whereas it caused a decrease in the expression of genes linked to glycolysis TPI1 (FC = 0.60, P = 0.000418), ALDOA (FC = 0.33, P = 0.000114), LDHA (FC = 0.36, P < 0.0001), ADH4 (FC = 0.49, P = 0.000129) in CaCO2 cells. Similarly, we observed that the overexpression of NAT1 led to an increase in the expression of oxidative phosphorylation related genes NDUFB6 (FC = 1.68, P = 0.000419), NDUFS7 (FC = 2.06185567, P = 0.000129), NDUFB3 (FC = 2.06, P < 0.0001), and ATP6V0D2 (FC = 2.06, P = 0.000127) and the TCA related genes ACLY (FC = 1.68, P = 0.000418), MDH2 (FC = 3.044, P = 0.000114), SDHA (FC = 2.80, P < 0.0001), ACO1 (FC = 2.06, P = 0.000129), whereas it caused a decrease in the expression of genes linked to glycolysis TPI1 (FC = 0.60, P = 0.00076), ALDOA (FC = 0.34, P = 0.000111), LDHA (FC = 0.39, P = 0.000124), ADH4 (FC = 0.49, P = 0.000152) in HCT116 cells. Conversely, the silencing of NAT1 resulted in a decrease in the expression of genes related to oxidative phosphorylation and the TCA, accompanied by an increase in the expression of genes related to glycolysis (Fig. 4E). Furthermore, a metabolomic analysis has validated that the silencing of NAT1 results in a decrease in metabolites linked to oxidative phosphorylation and the TCA, whereas glycolytic-related metabolites are overexpressed. On the contrary, the overexpression of NAT1 demonstrates an opposing pattern (Fig. 5).

Fig. 4
figure 4

The impact of NAT1 on metabolism in CRC cells at mRNA level. A DEGs were identified between the sh-NC (N = 3) and sh-NAT1 (N = 3) groups through RNA-seq analysis in CaCO2 and HCT116 cells. Genes with a FC ≥ 2 and a P-value ≤ 0.05 were designated as DEGs using the R package "limma-voom". B A total of 1756 DEGs were identified in both CaCO2 and HCT116 cells. C Functional analysis of these DEGs revealed metabolic differences between the sh-NC and sh-NAT1 groups. D DEGs was significantly enriched in glycolysis, tricarboxylic acid cycle and oxidative phosphorylation pathways. E Expression levels of DEGs involved in oxidative phosphorylation, the citrate cycle, and glycolysis pathways were assessed using qRT-PCR assay via unpaired t-test. Comparison of sh-NC with sh-NAT1 and OE-NC with OE-NAT1 via unpaired t-test. The experiments were independently repeated at least three times. *P < 0.05

Full size image
Fig. 5
figure 5

The impact of NAT1 on metabolism in CRC at metabolite level. Metabolomics was used to evaluate the expression levels of metabolites involved in oxidative phosphorylation, citric acid cycle and glycolysis pathways. Statistical significance is denoted as ns represents no significance, *P < 0.05. Comparison of sh-NC with sh-NAT1 and OE-NC with OE-NAT1 via unpaired t-test

Full size image

Identifying the specific cell–cell interactions between tumor cells and other cells in CRC

By utilizing the GSE23159 dataset, we successfully obtained the scRNA-seq expression profile from adjacent normal tissues. Subsequently, we employed the t-SNE method to cluster the cells (Supplement Figure 3A, upper) and annotated them using singleR (Supplement Figure 3A, lower). Additionally, we calculated the outgoing and incoming signaling patterns for each cell (Supplement Figure 3B), along with the likelihood of cell–cell contact between epithelial cells (designated as normal cells) and other cell types, using Cellchat (Fig. 6A). Furthermore, by leveraging both the GSE23159 (Supplement Figure 4) and GSE205506 (Supplement Figure 5) datasets, we retrieved the scRNA-seq expression profile from CRC tumor tissues. The cells were independently clustered and annotated using t-SNE and singleR. Additionally, we calculated the outgoing and incoming signaling patterns for each cell, along with the likelihood of cell–cell contact between epithelial cells (designated as tumor cells) and other cell types, using Cellchat (Fig. 6B, C). Notably, cell–cell contacts mediated by the VEGFA-VEGFR axis are more frequent between epithelial and endothelial cells in CRC tumor tissues compared to adjacent normal tissues.

Fig. 6
figure 6

Identifying the specific cell–cell interactions between tumor cells and other cells. A The cell–cell communication between epithelial cells and other cells in normal tissues in the GSE231559 dataset: this panel focuses on the specific communication between epithelial cells and other cell types. B The cell–cell communication between epithelial cells and neighboring cells in tumor tissues without treatment in the GSE231559 dataset was conducted using Cellchat method. C The cell–cell communication between epithelial cells and neighboring cells in tumor tissues without treatment in the GSE205506 dataset was conducted using Cellchat method. D The cell–cell communication between epithelial cells and neighboring cells in tumor tissues with PD-1 blockade in the GSE205506 dataset was conducted using Cellchat method

Full size image

Anti-PD-1 therapy effectively blocks the communication between tumor cells and endothelial cells via the VEGFA-VEGFR Axis

Using the GSE205506 dataset, we derived the scRNA-seq expression profile from CRC tumor tissues treated with anti-PD-1. Cells were then clustered and annotated independently via tsne and singleR (Supplement Figure 6A). Both incoming and outgoing signaling patterns were computed for each cell (Supplement Figure 6B). Additionally, the potential for cell–cell contact between epithelial cells and other cells was determined using Cellchat (Fig. 6D). Notably, anti-PD-1 treatment reduced the likelihood of VEGFA-VEGFR-based cell–cell interactions in epithelial and endothelial cells. Within the GSE205506 dataset, we isolated epithelial and endothelial cells from CRC tumor tissues, regardless of anti-PD-1 treatment. Epithelial cells were further classified based on NAT1, LGR5, and VEGFA expression (Fig. 7A), while endothelial cells were categorized based on VEGFR1 and VEGFR2 expression (Fig. 7B). Cellchat analysis revealed that in CRC tumor tissues without anti-PD-1 treatment, the probability of cell–cell communication between NAT1low/VEGFAhigh/LGR5low epithelial cells and VEGFRhigh endothelial cells was 0.292. However, in CRC tumor tissues treated with anti-PD-1, this probability decreased to 0.064, representing a FC of 4.533 (Fig. 7C). Moreover, prognostic analysis demonstrated a strong association between the number of endothelial cells in tumor tissue and a poorer prognosis for CRC patients (Fig. 7D). Additionally, our findings suggest that silencing NAT1 enhances VEGFA expression in CRC cells, whereas overexpressing NAT1 leads to decreased VEGFA expression in CRC cells (Fig. 7E).

Fig. 7
figure 7

Analysis of the impact of anti-PD-1 treatment on cell–cell communication in CRC. A and B Re-clustering and annotation of epithelial cells (A) and endothelial cells (B). This section provides a re-clustered and annotated view of epithelial and endothelial cells, enabling a deeper understanding of their interactions. C Cell–cell communication probability between epithelial cells and endothelial cells: this panel quantifies the probability of cell–cell communication between epithelial cells and endothelial cells, shedding light on the efficacy of anti-PD-1 treatment on VEGFA-VEGFR axis. D Kaplan–Meier analysis of overall survival, disease-free survival, and progression-free survival was used to compare the prognosis of CRC patients based on the optimal cut-off value of the proportion of endothelial cells. Log-rank test was used to calculate the P value. The proportion of endothelial cells was calculated based on the TCGA-CRC dataset (N = 592) via the TIMER 2.0 database. E WB analysis revealed that silencing NAT1 promoted the expression of VEGFA, whereas overexpressing NAT1 inhibited VEGFA expression in CRC cells. Comparison of sh-NC with sh-NAT1 and OE-NC with OE-NAT1 via unpaired t-test. The experiments were independently repeated at least three times. Statistical significance is denoted as *P < 0.05

Full size image

Reduced NAT1 expression may upregulate VEGFA via activating the transcription factors ZFP36 and HNF1A

Utilizing the TRRUST 2.0 database (https://www.grnpedia.org/trrust/), we identified candidate transcription factors for VEGFA. Subsequently, RNA-seq analysis was employed to filter DEGs potentially under the regulation of NAT1. Our findings revealed the regulation of five transcription factors by NAT1 (Fig. 8A). In both CaCO2 and HCT116 cell lines, the silencing of NAT1 resulted in the downregulation of EGR1, RUNX1, and NKX3-1, and conversely, the upregulation of ZFP36 and HNF1A (Fig. 8B). Moreover, correlation analysis demonstrated a significant positive correlation between EGR1, RUNX1, ZFP36, and HNF1A with VEGFA expression (Fig. 8C).

Fig. 8
figure 8

The potential regulatory mechanism between NAT1 and VEGFA. A Using the TRRUST 2.0 database (https://www.grnpedia.org/trrust/), we identified candidate transcription factors that may regulate VEGFA. In Fig. 4A, RNA-seq analysis was performed to examine the heterogeneity between sh-NC (n = 3) and sh-NAT1 (n = 3) groups in CaCO2 and HCT116 cells, identifying DEGs potentially regulated by NAT1. Our findings suggest that NAT1 may influence VEGFA expression through five transcription factors. B In both CaCO2 and HCT116 cells, compared to the sh-NC group (n = 3), the expression levels of EGR1, RUNX1, and NKX3-1 were significantly reduced in the sh-NAT1 group (n = 3). Conversely, the expression levels of HNF1A and ZFP36 were significantly increased in the sh-NAT1 group. C Based on data from TCGA-CRC (n = 592), Spearman correlation analysis revealed the relationships between these five transcription factors and VEGFA expression

Full size image

The impact of NAT1 on the TME in CRC

Utilizing the TIMER 2.0 database, we assessed the correlation between NAT1 expression levels and the infiltration of 26 distinct immune cell types in CRC. Our findings indicate a significant positive association between NAT1 expression and the presence of multiple CD4+ T cell subtypes, including activated memory CD4+ T cells, resting memory CD4+ T cells, effector memory CD4+ T cells, and central memory CD4+ T cells (Table 3).

Table 3 The effect of NAT1 on immune cell infiltration was evaluated using TIMER2.0 database
Full size table

Discussion

Low NAT1 expression is associated with decreased responsiveness to chemotherapies in breast cancer patients [5]. In this study, we demonstrated that NAT1 deficiency in colorectal cancer (CRC) cells increases resistance to multiple chemotherapeutic agents, including vinblastine, docetaxel, gemcitabine, vincristine, and daporinad. Importantly, overexpression of NAT1 in these cells restored sensitivity to all five drugs, providing strong evidence that NAT1 deficiency is a key factor driving resistance to these therapeutic agents in CRC. Previous studies have shown that NAT1 expression can be suppressed by medications such as tamoxifen [23], cisplatin [24], and disulfiram [25]. Our findings extend these observations by demonstrating that treatment with vinblastine, docetaxel, gemcitabine, vincristine, or daporinad is associated with decreased NAT1 levels in CRC cell lines. This suggests that patients with CRC might develop chemo-resistance following treatment due to the reduction of NAT1 expression potentially caused by the drugs themselves.

CSCs are a small, quiescent subpopulation within tumors, characterized by their self-renewal ability and capacity to generate diverse tumor cells. Similar to normal stem cells, CSCs contribute to the unlimited clonogenic potential of cancers and help explain why complete tumor removal is rarely achieved, despite initial treatment responses [26, 27]. Previous study has demonstrated that LGR5 inhibits apoptosis by regulating the PDCD5/p53 signaling axis, thereby promoting chemo-resistance in hepatocellular carcinoma [28]. In this study, LGR5+ CSCs exhibit greater resistance to these chemotherapy agents compared to other cells. Additionally, inhibiting NAT1 led to an increase in the proportion of LGR5+ CSCs in CRC. These observations imply that reduced NAT1 expression enhances chemo-resistance by promoting the formation of LGR5+ CSCs. Notably, LGR5 ablation resulted in a greater reduction in tumor volume than expected based on the original LGR5+ tumor population in CRC [29]. Therefore, LGR5+ CSCs-like that evade chemo-therapy are more likely to cause tumor recurrence.

Our study shows that NAT1 primarily regulates metabolic pathways. To explore the relationship between NAT1 expression and cellular metabolism, we conducted comprehensive assays focusing on glycolysis, oxidative phosphorylation, and the TCA cycle. The results indicated that NAT1 suppression in CRC cells increased glycolytic activity while decreasing oxidative phosphorylation and TCA cycle activities. Conversely, NAT1 overexpression had the opposite effect, suggesting NAT1’s role in modulating CRC cell metabolism. Previous studies have reported similar findings: silencing NAT1 in breast cancer cells increased glycolytic reserve capacity [30], and eliminating NAT1 in both MDA-MB-231 breast cancer cells and HT-29 colon cancer cells reduced oxidative phosphorylation [31]. These observations reinforce the link between NAT1 expression and cellular metabolic processes. Glycolysis is known to promote chemo-resistance in cancer cells [32, 33], suggesting that low NAT1 expression may enhance chemo-resistance in CRC cells by increasing glycolytic activity. Cellular metabolism is critical for the maintenance and differentiation of CSCs [34], with glycolysis being particularly important for CSC formation and survival [35]. Our results indicate that reduced NAT1 expression, which shifts cellular metabolism towards glycolysis, likely drives the development of LGR5+ CSCs and their chemo-resistance.

TME plays a crucial role in chemo-resistance [36]. In CRC, scRNA-seq has revealed a critical interaction between tumor cells and endothelial cells, mediated by the VEGFA-VEGFR pathway. Endothelial cells are essential for angiogenesis—the formation of new blood vessels—with VEGFR1 and VEGFR2 signaling being vital for their functional development and maintenance [37, 38]. Our research shows that suppressing NAT1 expression increases VEGFA expression in CRC cells. This amplifies VEGFA-VEGFR signaling between tumor cells and endothelial cells, promoting angiogenesis by activating endothelial cells. Angiogenesis is known to contribute to chemo-resistance in various tumors [32, 39, 40].This suggests that post-chemotherapy, surviving CRC cells with low NAT1 expression may activate angiogenesis via the VEGFA-VEGFR axis, leading to poorer patient outcomes.

Anti-PD-1 therapy activates T cells by disrupting the PD-1/PD-L1 interaction, enhancing the immune system’s ability to target tumor cells [41, 42]. Recent studies show that anti-PD-1 can also regulate VEGF and VEGFR expression indirectly through activated T cell-secreted cytokines (e.g., interferons, IL-2, IL-6, IL-12, TNF-α, TNF-β), which reduce VEGFA production and inhibit tumor angiogenesis [43]. In this study, scRNA-seq data reveals that anti-PD-1 treatment suppresses VEGFA-VEGFR signaling between NAT1low/VEGFAhigh tumor cells and VEGFRhigh endothelial cells. While more research is needed, these findings suggest that CRC patients with high VEGFA-VEGFR signaling may benefit from anti-PD-1 therapy.

Memory CD4+ T cells detect tumor antigens and boost the immune response by activating other immune cells [44]. Anti-PD-1 therapy restores the function of memory CD4+ T cells by blocking PD-1/PD-L1 interactions, enhancing their ability to target and destroy tumors [45, 46]. Our analysis using the TIMER2.0 database shows that reduced NAT1 expression is linked to lower infiltration of these cells, which may reduce the therapeutic effect of anti-PD-1 therapy.

In summary, our study shows that NAT1 loss in CRC cells increases drug resistance by promoting LGR5+ CSC formation via glycolysis activation. Low NAT1 expression in CRC patients is linked to poor chemotherapy outcomes, informing clinical treatment strategies.

Availability of data and materials

No datasets were generated or analysed during the current study.

References

  1. Dekker E, Tanis PJ, Vleugels JLA, Kasi PM, Wallace MB. Colorectal cancer. Lancet. 2019;394:1467–80.

    Article  PubMed  Google Scholar 

  2. Yan H, Talty R, Aladelokun O, Bosenberg M, Johnson CH. Ferroptosis in colorectal cancer: a future target? Br J Cancer. 2023;128:1439–51.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Butcher NJ, Minchin RF. Arylamine N-acetyltransferase 1: a novel drug target in cancer development. Pharmacol Rev. 2012;64:147–65.

    Article  CAS  PubMed  Google Scholar 

  4. Tiang JM, Butcher NJ, Minchin RF. Effects of human arylamine N-acetyltransferase I knockdown in triple-negative breast cancer cell lines. Cancer Med. 2015;4:565–74.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Minchin RF, Butcher NJ. Trimodal distribution of arylamine N-acetyltransferase 1 mRNA in breast cancer tumors: association with overall survival and drug resistance. BMC Genomics. 2018;19:513.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Mai Y, Su J, Yang C, Xia C, Fu L. The strategies to cure cancer patients by eradicating cancer stem-like cells. Mol Cancer. 2023;22:171.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Liu YQ, Wang SK, Xu QQ, Yuan HQ, Guo YX, Wang Q, Kong F, Lin ZM, Sun DQ, Wang RM, Lou HX. Acetyl-11-keto-beta-boswellic acid suppresses docetaxel-resistant prostate cancer cells in vitro and in vivo by blocking Akt and Stat3 signaling, thus suppressing chemoresistant stem cell-like properties. Acta Pharmacol Sin. 2019;40:689–98.

    Article  CAS  PubMed  Google Scholar 

  8. Wu B, Shi X, Jiang M, Liu H. Cross-talk between cancer stem cells and immune cells: potential therapeutic targets in the tumor immune microenvironment. Mol Cancer. 2023;22:38.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Kumari S, Advani D, Sharma S, Ambasta RK, Kumar P. Combinatorial therapy in tumor microenvironment: Where do we stand? Biochim Biophys Acta Rev Cancer. 2021;1876: 188585.

    Article  CAS  PubMed  Google Scholar 

  10. Mao X, Xu J, Wang W, Liang C, Hua J, Liu J, Zhang B, Meng Q, Yu X, Shi S. Crosstalk between cancer-associated fibroblasts and immune cells in the tumor microenvironment: new findings and future perspectives. Mol Cancer. 2021;20:131.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Yi M, Zheng X, Niu M, Zhu S, Ge H, Wu K. Combination strategies with PD-1/PD-L1 blockade: current advances and future directions. Mol Cancer. 2022;21:28.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Maeser D, Gruener RF, Huang RS. oncoPredict: an R package for predicting in vivo or cancer patient drug response and biomarkers from cell line screening data. Brief Bioinform. 2021;22:bbab260.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16:284–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Engebretsen S, Bohlin J. Statistical predictions with glmnet. Clin Epigenetics. 2019;11:123.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Gribov A, Sill M, Luck S, Rucker F, Dohner K, Bullinger L, Benner A, Unwin A. SEURAT: visual analytics for the integrated analysis of microarray data. BMC Med Genomics. 2010;3:21.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Aran D, Looney AP, Liu L, Wu E, Fong V, Hsu A, Chak S, Naikawadi RP, Wolters PJ, Abate AR, et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nat Immunol. 2019;20:163–72.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Zhang X, Lan Y, Xu J, Quan F, Zhao E, Deng C, Luo T, Xu L, Liao G, Yan M, et al. Cell Marker: a manually curated resource of cell markers in human and mouse. Nucleic Acids Res. 2019;47:D721–8.

    Article  CAS  PubMed  Google Scholar 

  18. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q. Inference and analysis of cell-cell communication using cell chat. Nat Commun. 2021;12:1088.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Madden EC, Gorman AM, Logue SE, Samali A. Tumour cell secretome in chemo-resistance and tumour recurrence. Trends Cancer. 2020;6:489–505.

    Article  CAS  PubMed  Google Scholar 

  20. Crea F, Danesi R, Farrar WL. Cancer stem cell epigenetics and chemo-resistance. Epigenomics. 2009;1:63–79.

    Article  CAS  PubMed  Google Scholar 

  21. Huang T, Song X, Xu D, Tiek D, Goenka A, Wu B, Sastry N, Hu B, Cheng SY. Stem cell programs in cancer initiation, progression, and therapy resistance. Theranostics. 2020;10:8721–43.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Zheng XX, Chen JJ, Sun YB, Chen TQ, Wang J, Yu SC. Mitochondria in cancer stem cells: Achilles heel or hard armor. Trends Cell Biol. 2023;33:708–27.

    Article  CAS  PubMed  Google Scholar 

  23. Lee JH, Lu HF, Wang DY, Chen DR, Su CC, Chen YS, Yang JH, Chung JG. Effects of tamoxifen on DNA adduct formation and arylamines N-acetyltransferase activity in human breast cancer cells. Res Commun Mol Pathol Pharmacol. 2004;115–116:217–33.

    PubMed  Google Scholar 

  24. Ragunathan N, Dairou J, Pluvinage B, Martins M, Petit E, Janel N, Dupret JM, Rodrigues-Lima F. Identification of the xenobiotic-metabolizing enzyme arylamine N-acetyltransferase 1 as a new target of cisplatin in breast cancer cells: molecular and cellular mechanisms of inhibition. Mol Pharmacol. 2008;73:1761–8.

    Article  CAS  PubMed  Google Scholar 

  25. Malka F, Dairou J, Ragunathan N, Dupret JM, Rodrigues-Lima F. Mechanisms and kinetics of human arylamine N-acetyltransferase 1 inhibition by disulfiram. FEBS J. 2009;276:4900–8.

    Article  CAS  PubMed  Google Scholar 

  26. Schatton T, Frank NY, Frank MH. Identification and targeting of cancer stem cells. BioEssays. 2009;31:1038–49.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Gupta PB, Chaffer CL, Weinberg RA. Cancer stem cells: mirage or reality? Nat Med. 2009;15:1010–2.

    Article  CAS  PubMed  Google Scholar 

  28. Ma Z, Guo D, Wang Q, Liu P, Xiao Y, Wu P, Wang Y, Chen B, Liu Z, Liu Q. Lgr5-mediated p53 repression through PDCD5 leads to doxorubicin resistance in hepatocellular carcinoma. Theranostics. 2019;9:2967–83.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Shimokawa M, Ohta Y, Nishikori S, Matano M, Takano A, Fujii M, Date S, Sugimoto S, Kanai T, Sato T. Visualization and targeting of LGR5(+) human colon cancer stem cells. Nature. 2017;545:187–92.

    Article  CAS  PubMed  Google Scholar 

  30. Carlisle SM, Trainor PJ, Doll MA, Stepp MW, Klinge CM, Hein DW. Knockout of human arylamine N-acetyltransferase 1 (NAT1) in MDA-MB-231 breast cancer cells leads to increased reserve capacity, maximum mitochondrial capacity, and glycolytic reserve capacity. Mol Carcinog. 2018;57:1458–66.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Wang L, Minchin RF, Essebier PJ, Butcher NJ. Loss of human arylamine N-acetyltransferase I regulates mitochondrial function by inhibition of the pyruvate dehydrogenase complex. Int J Biochem Cell Biol. 2019;110:84–90.

    Article  CAS  PubMed  Google Scholar 

  32. Wang Z, Chen W, Zuo L, Xu M, Wu Y, Huang J, Zhang X, Li Y, Wang J, Chen J, et al. The Fibrillin-1/VEGFR2/STAT2 signaling axis promotes chemo-resistance via modulating glycolysis and angiogenesis in ovarian cancer organoids and cells. Cancer Commun (Lond). 2022;42:245–65.

    Article  CAS  PubMed  Google Scholar 

  33. Wang X, Zhang H, Yang H, Bai M, Ning T, Deng T, Liu R, Fan Q, Zhu K, Li J, et al. Exosome-delivered circRNA promotes glycolysis to induce chemo-resistance through the miR-122-PKM2 axis in colorectal cancer. Mol Oncol. 2020;14:539–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Rodriguez-Colman MJ, Schewe M, Meerlo M, Stigter E, Gerrits J, Pras-Raves M, Sacchetti A, Hornsveld M, Oost KC, Snippert HJ, et al. Interplay between metabolic identities in the intestinal crypt supports stem cell function. Nature. 2017;543:424–7.

    Article  CAS  PubMed  Google Scholar 

  35. Sancho P, Barneda D, Heeschen C. Hallmarks of cancer stem cell metabolism. Br J Cancer. 2016;114:1305–12.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Velaei K, Samadi N, Barazvan B, Soleimani Rad J. Tumor microenvironment-mediated chemo-resistance in breast cancer. Breast. 2016;30:92–100.

    Article  PubMed  Google Scholar 

  37. Eelen G, de Zeeuw P, Treps L, Harjes U, Wong BW, Carmeliet P. Endothelial cell metabolism. Physiol Rev. 2018;98:3–58.

    Article  CAS  PubMed  Google Scholar 

  38. Olsson AK, Dimberg A, Kreuger J, Claesson-Welsh L. VEGF receptor signalling - in control of vascular function. Nat Rev Mol Cell Biol. 2006;7:359–71.

    Article  CAS  PubMed  Google Scholar 

  39. Shibutani M, Nakao S, Maeda K, Nagahara H, Kashiwagi S, Hirakawa K, Ohira M. The impact of tumor-associated macrophages on chemo-resistance via angiogenesis in colorectal cancer. Anticancer Res. 2021;41:4447–53.

    Article  PubMed  Google Scholar 

  40. Nusrat O, Belotte J, Fletcher NM, Memaj I, Saed MG, Diamond MP, Saed GM. The role of angiogenesis in the persistence of chemo-resistance in epithelial ovarian cancer. Reprod Sci. 2016;23:1484–92.

    Article  CAS  PubMed  Google Scholar 

  41. Lei Q, Wang D, Sun K, Wang L, Zhang Y. Resistance mechanisms of anti-PD1/pdl1 therapy in solid tumors. Front Cell Dev Biol. 2020;8:672.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Kornepati AVR, Vadlamudi RK, Curiel TJ. Programmed death ligand 1 signals in cancer cells. Nat Rev Cancer. 2022;22:174–89.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Lee WS, Yang H, Chon HJ, Kim C. Combination of anti-angiogenic therapy and immune checkpoint blockade normalizes vascular-immune crosstalk to potentiate cancer immunity. Exp Mol Med. 2020;52:1475–85.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Kunzli M, Masopust D. CD4(+) T cell memory. Nat Immunol. 2023;24:903–14.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Takeuchi Y, Tanemura A, Tada Y, Katayama I, Kumanogoh A, Nishikawa H. Clinical response to PD-1 blockade correlates with a sub-fraction of peripheral central memory CD4+ T cells in patients with malignant melanoma. Int Immunol. 2018;30:13–22.

    Article  CAS  PubMed  Google Scholar 

  46. Zuazo M, Arasanz H, Bocanegra A, Fernandez G, Chocarro L, Vera R, Kochan G, Escors D. Systemic CD4 immunity as a key contributor to PD-L1/PD-1 blockade immunotherapy efficacy. Front Immunol. 2020;11: 586907.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We would like to thank the Medical Science and Technology Innovation Center of the Affiliated Suzhou Hospital of Nanjing Medical University for its help in the detection of experimental samples.

Funding

This work was supported by grants from National Natural Science Foundation of China (82303086), Suzhou Applied Basic Research and science and technology innovation project (SYWD2024294), Suzhou Applied Basic Research and science and technology innovation project (SYWD2024010), Suzhou Municipal Science and Technology Bureau’s Livelihood Technology-Basic Research on Medical and Health Applications (Grant NO. SYSD2018126), and Postgraduate Research & Practice Innovation Program of Jiangsu Province (SJCX23_0682).

Author information

Author notes

  1. Zheng Yuan, Kai Fang and Xinsheng Miao have contributed equally to this study.

Authors and Affiliations

  1. Affiliated Suzhou Hospital of Nanjing Medical University, Suzhou Municipal Hospital, Gusu School of Nanjing Medical University, Suzhou, China

    Zheng Yuan, Xinsheng Miao, Yan Zhang, Menghui Gu, Wei Xu, Hao Li, Jiahui Zhou & Xinhua Gu

  2. Nanjing Medical University, Suzhou, China

    Dawei Zhu

  3. Affiliated Suzhou Hospital of Nanjing Medical University, Center for Reproduction and Genetics of Suzhou Municipal Hospital, Gusu School of Nanjing Medical University, Suzhou, China

    Jian Sun

  4. College of Basic Medical Sciences, Suzhou University, Suzhou, China

    Kai Fang

Authors
  1. Zheng Yuan
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Kai Fang
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Xinsheng Miao
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Yan Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Menghui Gu
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Wei Xu
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Hao Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Dawei Zhu
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Jiahui Zhou
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Jian Sun
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Xinhua Gu
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Zheng Yuan, Kai Fang, Xinhua Gu, Jiahui Zhou, and Jian Sun conceived of, designed, and supervised the study. Zheng Yuan, Kai Fang and Xinsheng Miao performed most of the experiments and wrote the manuscript. Jian Sun, Yan Zhang, Menghui Gu, and Wei Xua performed some of the experiments. All co-authors have reviewed and approved this version of the manuscript.

Corresponding authors

Correspondence to Jiahui Zhou, Jian Sun or Xinhua Gu.

Ethics declarations

Ethics approval and consent to participate

Human data and samples were obtained with informed consent and the use was approved by the Institutional Review Board of the Suzhou Municipal hospital.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

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

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yuan, Z., Fang, K., Miao, X. et al. Investigating the mechanisms by which low NAT1 expression in tumor cells contributes to chemo-resistance in colorectal cancer. Clin Epigenet 17, 77 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13148-025-01882-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13148-025-01882-4

Keywords

Clinical Epigenetics

ISSN: 1868-7083

Contact us