Sodium Butyrate Induces CRC Cell Ferroptosis via the CD44/SLC7A11 Pathway and Exhibits a Synergistic Therapeutic Effect with Erastin

Simple Summary Sodium butyrate (NaB) is a short-chain fatty acid produced by intestinal microbial fermentation of dietary fiber. It has been shown to be effective in inhibiting colorectal cancer (CRC), but the mechanism is not known. We verified the ability of NaB to induce ferroptosis and the effect on relevant genotypes in normal intestinal cells and colorectal tumor cells, respectively. Moreover, better inhibition of tumor cells was observed when NaB was combined with Erastin (a ferroptosis-positive drug), suggesting that NaB combined with Erastin might have a stronger anti-CRC effect. Abstract Colorectal cancer (CRC) is one of the most common malignancies, and effective treatment and prevention methods are lacking. Sodium butyrate (NaB) is a short-chain fatty acid produced by intestinal microbial fermentation of dietary fiber. It has been shown to be effective in inhibiting CRC, but the mechanism is not known. Methods: Human normal intestinal epithelial cell line FHT and colorectal tumor cell line HCT-116 were treated with NaB alone or in combination with different programmed cell death inhibitors. Cell activity was then assessed with MTT assays and PI staining; ferroptosis with Fe2+, glutathione (GSH), and lipid peroxidation assays; signaling pathway screening with PCR arrays; and CD44, SCL7A11, and GPX4 expression with Western blotting. A CD44-overexpressing HCT-116 cell line was constructed to determine the effect of the overexpression of CD44 on NaB-induced ferroptosis. The synergistic effect of co-treatment with NaB and Erastin was assessed by isobolographic analysis. Results: NaB induced apoptosis and ferroptosis in HCT-116 cells but only induced low-level apoptosis in FHC cells. Moreover, NaB significantly increased intracellular Fe2+ and promoted GSH depletion and lipid peroxidation in HCT-116 cells. Ferroptosis-related qPCR array analysis identified CD44/SLC7A11 as a potential effector molecular of NaB-induced ferroptosis. NaB significantly inhibited the expression of CD44 and SLC7A11 in mouse CRC tissues. A CD44 overexpressed HCT-116 cell line was used to verify that CD44/SLC7A11 was a key signaling pathway that NaB-induced GSH depletion, lipid peroxidation accumulation, and ferroptosis in HCT-116 cells. Examination of whether NaB can increase the effect of ferroptosis agents showed that NaB, in combination with Erastin, a ferroptosis inducer, further promoted HCT-116 cell death and increased changes of ferroptosis markers. Conclusions: Our results suggest that NaB induces ferroptosis in CRC cells through the CD44/SLC7A11 signaling pathway and has synergistic effects with Erastin. These results may provide new insights into CRC prevention and the combined use of NaB and ferroptosis-inducing agents.


Introduction
CRC is one of the most common malignancies worldwide. In 2020, there were about 1.88 million new cases of CRC and 915,000 deaths, making it the third-leading cause of morbidity and the second-leading cause of mortality [1]. The primary treatments for CRC

MTT
The viability of HCT 116 and FHC cells was measured using the MTT method. Cells are placed in 96-well plates (8 × 10 3 cells/well) overnight and then treated with the corresponding drug for 24-72 h. Add 20 µL of MTT (5 mg/mL) per well and incubate for another 4 h. DMSO is then added to dissolve the MTT tetrazolium crystals. Place the plate on an enzyme immunoassay detector and shake for 10 min under low-speed shaking. Absorbance is obtained at an optical density (OD) of 490 nm.

Cell Morphology Observation and Photography
Cells were seeded in 6-well plates and treated with the corresponding drug for 24-72 h. After removing the 6-well plate, discard the old full medium, wash with PBS twice, carefully drain the remaining liquid with a pipette, fix the cells with methanol for about 15-20 min at room temperature and then dump them, and carefully wash with PBS twice. Observe the morphology and number of cells under the white light lens of the microscope and take pictures.

PI Staining and Fluorescence Mapping
Collect adherent and suspension cells during the experiment, centrifuge at 1000 rpm, 3 min, wash twice using 1 mL PBS, count cells under the microscope, adjust the number of cells to about 1 million cells, and resuspend cells with 500 µL of stain solution after centrifugation again. After 5 min of incubating in the light, add 1 drop of the cell suspension to the slide, cover the coverslip and observe its red fluorescence and take a picture.

Ferroptosis PCR Array Analysis and Real-Time Fluorescence Quantitative PCR
The PCR array used in the study was customized to the Shanghai WCGENE gene by the research group according to the research needs. The size of the PCR array was 96 Wells, and the gene sequence of each well was verified by WCGENE. The list of genes was shown in Table 1.  IREB1  ATG5  CDO1  DPP4  GCLM  HARS  B  ACSL4  ATP5G3  CHAC1  ELAVL1  GLS2  HEPH  C  AKR1B1  BBC3  CISD1  EMC2  GOT1  HFE  D  AKR1B10  BECN1  CISD2  EPRS  GPX4  HMOX1  E  AKR1C1  BRAF  CP  FTH1  GSS  HMOX2  F  ALDH1A1  BRD4  CS  FTL  GSTA1  HRAS  G  ALOX15  CA9  CYBB  FTMT  GSTP1  HSF1  H  ATF4  CARS1  DMT1  GCLC  HAMP  HSPB1   HUMAM  7  8  9  10  11  12   A  IREB2  NCOA4  PCBP1  SAT2  STEAP3  VDAC2  B  KEAP1  NFE2L2  PCBP2  SLC1A5  STIM1  VDAC3  C  KRAS  NOX1  PPARG  SLC39A14  TF  MAP1LC3C  D  LOX  NOX3  PRDX6  SLC39A8  TFR1  PANX2  E  LPCAT3  NOX4  PRNP  SLC3A2  TFR2  SAT1  F  MAP1LC3A  NQO1  PTGES2  SLC40A1  TP53  SQSTM1  G  MAP1LC3B  NRAS  RPL8  SLC7A11  TXNRD1  USP7  H  ACTB  GAPDH  HPRT1  18s  CD44  FSP1 Total RNA was isolated and purified using TRIzol, phenol, chloroform and ethanol. After RNA reverse transcription, mix 900 µL of cDNA and SYBR ® Premix Ex TaqTMI according to the instructions, mix well, add the mixture (9 µL per well) into the 96-well plate, seal the plate with transparent sealing plate membrane after adding, and centrifuge again. A real-time PCR detection system (CFX Connect™, BIORAD, Hercules, CA, USA) Cancers 2023, 15, 423 4 of 21 was used to detect gene amplification. The method was also used for real-time fluorescence quantitative PCR, and the primer sequences were shown in Table 2. After calculating the expression results of each gene, they need to be sorted out into tables (Tables S1 and S2) and sent into Rstudio for heat map and volcano map drawing. Geometric herewith's figure, GO pathway enrichment and enrichment of and function network with the protein interaction network analysis respectively by Bioinformatics at http://www.bioinformatics.com.cn (accessed on 14 February 2022) and Metascape at https://metascape.org/gp/index.html (accessed on 14 February 2022), there are two free online platforms for data analysis and visualization.

Population Public Database Analysis
The box plot, violin plot, and scatter plot of CD44 and SLC7A11 gene expression in this part of the results were drawn by the TCGA and GTEx visualization website GEPIA2 http://gepia2.cancer-pku.cn/#index (accessed on 14 February 2022).
The population viability analysis was drawn by the generic cancer prognosis PROGgeneV2 http://www.progtools.net/gene/index.php database (accessed on 14 February 2022).

Western Blotting
HCT 116 and FHC cells were lysed with RIPA solution (Beyotime Biotechnology, Shanghai, China) containing protease inhibitors (Keygen, Nanjing, China) after treatment with appropriate drugs. Total protein was collected by centrifugation at 14,000× g for 15 min. The protein samples were separated by 10% sodium dodecyl sulfate (SDS) -polyacrylamide gel electrophoresis (PAGE) at 80 V and then transferred to a polyvinylidene fluoride (PVDF) membrane (Bio-RAD Laboratories). The membrane was then incubated with primary and secondary antibodies to detect target proteins. In the final step, the membrane was visualized using a chemiluminescence detection kit (Pierce Biotechnology, Rockford, IL, USA) according to the manufacturer's instructions. The target protein was standardized as glyceraldehyde -3-phosphate dehydrogenase (GAPDH) and used as a sample control.

Immunofluorescence and Laser Confocal Imaging
The cells in the confocal petri dish were fixed, permeated and sealed in advance. According to the antibody instructions, the corresponding primary antibody and fluorescent secondary antibody and DAPI dye were used for detection (SLC7A11/xCT rabbit Polyclonal antibody dilution ratio was 1:5000, CD44 mouse monoclonal antibody dilution ratio was 1:4000, Alexa Fluor 488 rabbit antibody or Alexa Fluor 594 mouse antibody dilution ratio of 1:1000). Photographs were taken using a confocal microscope.

Animal Tissue Immunohistochemistry and Prussian Staining
Animal experimental design and tissue sampling are described in our previously published study [14]. The subsequent immunohistochemical and Prussian staining experiments were commissioned by BIOS Biological (Guangzhou, China). CD44 Rabbit mAb (ProteinTech, Wuhan, China) and SLC7A11/xCT Rabbit mAb (ProteinTech, China) diluted at 1:50 are used. Then, 3 non-overlapping fields were randomly selected for each slice under a 40× microscope for macroscopic observation comparison.

GSH Detection
Cells were lysed with Glutathione Buffer, and the supernatant was collected. The GSH content in the supernatant was assessed by a reduced glutathione assay kit (Solarbio, Beijing, China, #BC1175) according to the manufacturer's instructions. The absorbance was measured at 450 nm (A450). The concentration of GSH was determined using a standard GSH calibration curve and was related to the number of cells used.

Fe 2+ Staining
Serum-free basal medium DMEM and FerroOrange working solution were added to the petri dish so that the final concentration of FerroOrange in the basal medium was 1 µM. The cells were then incubated in an incubator for 30 min in darkness. After incubation, the six-well plate was taken out, and the red fluorescence intensity was observed directly with a fluorescence microscope. FerroOrange: Ex: 561 nm, Em: 570-620 nm.

Lipid ROS Detection
The cells were washed twice with PBS and then added with serum-free DMEM basal culture medium and C11 BODIPY 581/591 dye so that the final concentration of C11 BODIPY 581/591 was 10 µM. The cells were incubated in a cell culture box for 30 min without light. After incubation, the cells were taken out and washed with PBS more than 2 times. A confocal fluorescence microscope was used to stimulate the cells with 488 nm and 565 nm lasers, respectively, and relevant fluorescence signal images were collected.

Isobolographic Analysis
After drug combination treatment and MTT assay, isobolographic analysis was performed with reference to the method of Huang et al. [15]. First, a cartesian coordinate system was set up, in which the X-axis and Y-axis represent the doses of the two drugs (Erastin and NaB in this study) to achieve the same efficacy. In addition, A and B, respectively, represent the dose of IC50 when the two drugs are used alone. The two points and the origin can be enclosed into an equivalent triangle. If the two drugs act synergistically, the equivalent points of drug combination fall within the range of an equivalent triangle.

Statistical Methods
Statistical analyses of different sets of experiments were performed using Prism 7 (GraphPad Software, Inc., La Jolla, CA, USA). A 2-tailed t-test or Mann-Whitney's test was used in a single variable with a 2-group comparison, a 1-way ANOVA with Tukey's post-test or a Kruskal-Wallis test was used in single-variable comparisons with more than 2 groups, and 2-way ANOVA with Bonferroni posttest for multivariable analyses. Differences with p < 0.05 were regarded as statistically significant.

NaB Induces CRC Cell Death through Apoptosis and Ferroptosis
To test our hypothesis that ferroptosis may occur during NAB-induced CRC cell death, we used z-VAD (an apoptosis inhibitor), ferrostatin-1 (Fer-1, a potent ferroptosis specific inhibitor), necrostatin-1 (Nec-1, a potent necrotizing apoptosis inhibitor), and chloroquine (CQ, an effective autophagy inhibitor) to determine the types of programmed cell death induced by NaB. First, we found that NaB promotes FHC cell proliferation at low concentrations but inhibits it at high concentrations. In contrast, NaB had a more obvious inhibitory effect on HCT-116 cells. There were also differences in the response of FHC cells and HCT-116 cells to different cell death inhibitors. In FHC cells, only the apoptosis inhibitor z-VAD could partially reverse cell viability inhibited by NaB. In HCT-116 cells, the apoptosis inhibitor z-VAD and ferroptosis inhibitor Fer-1 partially reversed the proliferation of HCT-116 cells inhibited by NaB, while autophagy inhibitor CQ further inhibited the proliferation of HCT-116 cells. Nec-1 did not affect either cell line ( Figure 1A-D). Next, we observed cell morphology and death of FHC cells and HCT-116 cells when different types of cell death inhibitors were combined with NaB, and the results were consistent with the previous results ( Figure 1E).
were consistent with the results of intracellular GSH detection ( Figure 1H). The FerroOrange ferrous ion fluorescence probe was used to measure the Fe 2+ content in FHC cells and HCT-116 cells to determine if ferroptosis was occurring. HCT-116 cells had a higher basal Fe 2+ level than FHC cells, and the intracellular Fe 2+ level increased after treatment with NaB for 24 h; no significant change in the level of Fe 2+ in FHC cells was observed ( Figure  1I).
Based on these results, we hypothesized that NaB caused CRC cell death by inducing apoptosis and ferroptosis in CRC cells but did not induce ferroptosis in normal intestinal epithelial cells.   To further confirm the role of NaB in ferroptosis in CRC cells, levels of GSH, lipid ROS, and Fe 2+ were measured. The results showed that 10 mM NaB significantly down-regulated the GSH content in HCT-116 cells but did not affect the GSH level in FHC cells. The addition of N-acetyl-L-cysteine (NAC), a precursor of GSH, and Fer-1 significantly reversed the reduction of GSH induced by NaB in HCT-116 cells, but the apoptosis inhibitor z-VAD had no protective effect on the depletion of GSH induced by NaB ( Figure 1F,G). A C11-BODIPY fluorescence probe was used to detect lipid ROS, and the findings were consistent with the results of intracellular GSH detection ( Figure 1H). The FerroOrange ferrous ion fluorescence probe was used to measure the Fe 2+ content in FHC cells and HCT-116 cells to determine if ferroptosis was occurring. HCT-116 cells had a higher basal Fe 2+ level than FHC cells, and the intracellular Fe 2+ level increased after treatment with NaB for 24 h; no significant change in the level of Fe 2+ in FHC cells was observed ( Figure 1I).
Based on these results, we hypothesized that NaB caused CRC cell death by inducing apoptosis and ferroptosis in CRC cells but did not induce ferroptosis in normal intestinal epithelial cells.

Ferroptosis-Related qPCR Array Analysis Identified CD44/SLC7A11 as a Potential Effector Molecular of NaB-Induced Ferroptosis
Database FerrDb (http://www.zhounan.org/ferrdb/legacy/index.html, accessed on 14 February 2022) is a collection of authority management ferroptosis-related markers, regulatory factors, and associated diseases. We selected 92 genes for ferroptosis-related gene array analysis based on reliability scores. According to the analysis, there were 18 differentially expressed genes in FHC, and HCT-116 cells and 10 differentially expressed genes in HCT-116 cells after treatment with 10 mM NaB for 24 h (Padj < 0.05, Log2 fold-change [FC] > 1). The genes are illustrated in a gene heat map and volcano map (Figure 2A,B). All the differentially expressed genes (p < 0.05) in the HCT-116 cell control group and HCT-116 cell NaB treatment group were used to perform functional pathway annotation of the Gene Ontologies (GO) database. GO analysis results include molecular function (MF), cellular component (CC) and biological process (BP). The enrichment score of BP was the highest among the 3 analyses, and the annotation results showed that NaB significantly regulated functions related to GSH and Fe 2+ metabolism ( Figure 2C). Therefore, in the PCR array, we sorted out all genes related to Fe 2+ and GSH metabolism. The results showed that NaB changed the expression of GSH-related genes GCLC, GLS2, GOT1, CHAC1, BECN1, SLC7A11, CD44 and iron-related genes IREB1, IREB2, TFR1, SLC40A1, NCOA4, and ATG5 in HCT-116 cells ( Figure 2D). regulated functions related to GSH and Fe metabolism ( Figure 2C). Therefore, in the PCR array, we sorted out all genes related to Fe 2+ and GSH metabolism. The results showed that NaB changed the expression of GSH-related genes GCLC, GLS2, GOT1, CHAC1, BECN1, SLC7A11, CD44 and iron-related genes IREB1, IREB2, TFR1, SLC40A1, NCOA4, and ATG5 in HCT-116 cells ( Figure 2D).
In order to further study the relations between all pathways and functions, all the differentially expressed genes of HCT-116 cells in the control group and HCT-116 cells in the NaB treatment group were presented as a network diagram, in which items with similarity > 0.3 were connected by edges ( Figure 2E). Protein-protein interaction enrichment analysis was performed to obtain the total protein network. Molecular Complex Detection (MCODE) identified HRAS, NRAS, KRAS, SLC7A11, and CD44 and VDAC3, ALDH1A1, GOT1, and CS as 2 groups of key central protein interaction networks ( Figure 2F).
Previous studies have found that CD44 and SLC7A11 are key molecules that affect the transport of extracellular cystine for the intracellular synthesis of GSH and regulate ferroptosis and that they may be interrelated. Therefore, we determined if NaB induced ferroptosis in CRC cells via CD44/SLC7A11.  In order to further study the relations between all pathways and functions, all the differentially expressed genes of HCT-116 cells in the control group and HCT-116 cells in the NaB treatment group were presented as a network diagram, in which items with similarity > 0.3 were connected by edges ( Figure 2E). Protein-protein interaction enrichment analysis was performed to obtain the total protein network. Molecular Complex Detection (MCODE) identified HRAS, NRAS, KRAS, SLC7A11, and CD44 and VDAC3, ALDH1A1, GOT1, and CS as 2 groups of key central protein interaction networks ( Figure 2F).
Previous studies have found that CD44 and SLC7A11 are key molecules that affect the transport of extracellular cystine for the intracellular synthesis of GSH and regulate ferroptosis and that they may be interrelated. Therefore, we determined if NaB induced ferroptosis in CRC cells via CD44/SLC7A11.

CD44 and SLC7A11 Are Highly Expressed and Positively Correlated in Human CRC
In order to confirm the accuracy of the PCR array analysis and explore the role and correlation of CD44 and SLC7A11 in human CRC, population gene expression and survival analysis were performed using the pan-cancer data analysis tools GEPIA and ProGgene V2. The results showed that compared with a healthy population, CD44 and SLC7A11 mRNA levels were higher in CRC patients, but there was no difference in the expression levels of CD44 and SLC7A11 in different stages of CRC (Stage I to Stage IV) ( Figure 3A). The general linear correlation model showed that CD44 was positively correlated with SLC7A11 expression in CRC patients ( Figure 3B) (Pearson correlation coefficient = 0.21, p < 0.001). Survival analysis results showed that CRC patients with high expression levels of CD44 and SLC7A11 had a worse prognosis and shorter survival time ( Figure 3C). ProGgene V2. The results showed that compared with a healthy population, CD44 and SLC7A11 mRNA levels were higher in CRC patients, but there was no difference in the expression levels of CD44 and SLC7A11 in different stages of CRC (Stage Ⅰ to Stage Ⅳ) ( Figure 3A). The general linear correlation model showed that CD44 was positively correlated with SLC7A11 expression in CRC patients ( Figure 3B) (Pearson correlation coefficient = 0.21, p < 0.001). Survival analysis results showed that CRC patients with high expression levels of CD44 and SLC7A11 had a worse prognosis and shorter survival time ( Figure 3C).

NaB Inhibits CD44/SLC7A11 Expression In Vivo and In Vitro
Targets of ferroptosis induced by NaB were studied in vivo and in vitro. Western blotting results showed that CD44 and SLC7A11 exhibited low expression in FHC cells and high expression in HCT-116 cells ( Figure 4A). NaB significantly inhibited CD44 and SLC7A11 protein expression in HCT-116 cells but did not affect the expression of GPX4

NaB Inhibits CD44/SLC7A11 Expression In Vivo and In Vitro
Targets of ferroptosis induced by NaB were studied in vivo and in vitro. Western blotting results showed that CD44 and SLC7A11 exhibited low expression in FHC cells and high expression in HCT-116 cells ( Figure 4A). NaB significantly inhibited CD44 and SLC7A11 protein expression in HCT-116 cells but did not affect the expression of GPX4 protein. NaB did not affect the expression of CD44, SLC7A11, and GPX4 in FHC cells ( Figure 4B). Confocal laser study results showed that after treatment with 10 mM NaB for 24 h, the expressions of CD44 and SLC7A11 in HCT-116 cells were significantly downregulated, and co-localization was inhibited ( Figure 4C). These results indicate that NaB inhibits the expression of key ferroptosis proteins CD44 and SLC7A11 in HCT-116 cells but does not affect GPX4.
To verify the role of NaB in vivo, we established an AOM/DSS-induced murine inflammatory CRC model. 0.1M NaB(oral administration) reduces tumor load and tumor size in AOM/DSS-induced mice. Corresponding results and experimental methods are shown in our previous study [14]. Colorectal sections of the mice (a tumor-prone site) were examined by immunohistochemistry (IHC) staining and iron-specific staining (Prussian blue) for the key proteins CD44 and SLC7A11. IHC staining results showed CD44 and SLC7A11 expression was significantly up-regulated in the CRC model group, and the expression of CD44 and SLC7A11 was significantly decreased after treatment with NaB ( Figure 4D). Prussian blue staining showed that treatment with NaB increased iron deposition at the tumor site ( Figure 4D). These results indicate that NaB can effectively inhibit the expression of CD44 and SLC7A11 in a murine CRC model, which is consistent with the results of the cell-line experiments. lated, and co-localization was inhibited ( Figure 4C). These results indicate that NaB inhibits the expression of key ferroptosis proteins CD44 and SLC7A11 in HCT-116 cells but does not affect GPX4.
To verify the role of NaB in vivo, we established an AOM/DSS-induced murine inflammatory CRC model. 0.1M NaB(oral administration) reduces tumor load and tumor size in AOM/DSS-induced mice. Corresponding results and experimental methods are shown in our previous study [14]. Colorectal sections of the mice (a tumor-prone site) were examined by immunohistochemistry (IHC) staining and iron-specific staining (Prussian blue) for the key proteins CD44 and SLC7A11. IHC staining results showed CD44 and SLC7A11 expression was significantly up-regulated in the CRC model group, and the expression of CD44 and SLC7A11 was significantly decreased after treatment with NaB ( Figure 4D). Prussian blue staining showed that treatment with NaB increased iron deposition at the tumor site ( Figure 4D). These results indicate that NaB can effectively inhibit the expression of CD44 and SLC7A11 in a murine CRC model, which is consistent with the results of the cell-line experiments.

CD44 Overexpression Inhibits NaB-Induced Ferroptosis in CRC Cells
To further validate the role of CD44 and SLC7A11 in NaB-induced CRC ferroptosis, we transfected HCT-116 cells with a pcDNA3.1 (+) carrier construction of CD44 expression (CD44OE) or blank (vector) plasmid. The results showed that CD44 and SLC7A11 mRNA

CD44 Overexpression Inhibits NaB-Induced Ferroptosis in CRC Cells
To further validate the role of CD44 and SLC7A11 in NaB-induced CRC ferroptosis, we transfected HCT-116 cells with a pcDNA3.1 (+) carrier construction of CD44 expression (CD44OE) or blank (vector) plasmid. The results showed that CD44 and SLC7A11 mRNA and protein levels in the CD44 overexpression group were significantly higher than those in the control and vector transfection groups ( Figure 5A,B). Moreover, NaB reversed the mRNA and protein inhibition of CD44 and SLC7A11. Next, cell viability, GSH content change, and lipid ROS accumulation were studied to determine the occurrence of ferroptosis. The results showed that CD44 overexpression significantly restored the proliferation activity and cell morphology of cells treated with NaB for 24 h (Figure 5C-E). Although CD44 overexpressed plasmid transfection did not significantly increase the basal level of GSH in HCT-116 cells, the decrease in intracellular GSH content caused by NaB treatment was significantly alleviated in the CD44 overexpression group ( Figure 5F). Similarly, after treatment with 10 mM NaB, lipid ROS accumulation was significantly increased in the control and vector-transfected cells, while significantly reduced lipid ROS accumulation was observed in CD44-transfected cells ( Figure 5G). These results confirmed that the CD44/SLC7A11 signaling pathway is a key regulatory pathway of NAB-induced ferroptosis in HCT-116 cells. and protein levels in the CD44 overexpression group were significantly higher than those in the control and vector transfection groups ( Figure 5A,B). Moreover, NaB reversed the mRNA and protein inhibition of CD44 and SLC7A11. Next, cell viability, GSH content change, and lipid ROS accumulation were studied to determine the occurrence of ferroptosis. The results showed that CD44 overexpression significantly restored the proliferation activity and cell morphology of cells treated with NaB for 24 h (Figure 5C-E). Although CD44 overexpressed plasmid transfection did not significantly increase the basal level of GSH in HCT-116 cells, the decrease in intracellular GSH content caused by NaB treatment was significantly alleviated in the CD44 overexpression group ( Figure 5F). Similarly, after treatment with 10 mM NaB, lipid ROS accumulation was significantly increased in the control and vector-transfected cells, while significantly reduced lipid ROS accumulation was observed in CD44-transfected cells ( Figure 5G). These results confirmed that the CD44/SLC7A11 signaling pathway is a key regulatory pathway of NAB-induced ferroptosis in HCT-116 cells.

Erastin and NaB Synergistically Induce CRC Cell Ferroptosis
Several studies have shown that a histone deacetylase inhibitor (HDACi) combined with ferroptosis-inducing agents can enhance its antitumor effect [16][17][18]. Our results suggest that a high concentration or long-term exposure to NaB may promote apoptosis in normal intestinal epithelial cells. As such, it is important to optimize the concentration and exposure time of NaB to increase its antitumor effect and reduce the effects on normal colon cells. Erastin is a classic ferroptosis-inducing agent, and a concentration of about 10 μM induces ferroptosis in a number of tumor cell lines [8]. Results of our MTT assays showed that 10 μM Erastin did not significantly affect cell viability, but cell viability decreased rapidly with increasing NaB concentration in the Erastin-NAB treatment group ( Figure 6A). Cell morphology examination and staining of dead cells also indicated enhanced cytotoxicity by Erastin-NAB combined treatment ( Figure 6B).
The effect of Erastin combined with NaB was further tested by isobolographic analysis ( Figure 6C). MTT assay results showed that the IC50 and 95% confidence interval (CI) of NaB and Erastin for inhibiting HCT-116 proliferation were 12.420 mM (95% CI: 11.936-12.941 mM) and 26.978 µ m (95% CI: 25.714-28.381 µ m), respectively. The combination of NaB and Erastin at a concentration ratio of 12.420 mM to 26.978 µ m exhibited a synergistic effect on the inhibition of proliferation of HCT-116 cells (IC50 = NaB: 2.43 mM+Erastin: 4.861 µ m). Combined treatment with NaB and Erastin did not affect the proliferation activity of FHC cells( Figure 6D). In addition, compared with cells treated with NaB alone, the combined treatment group exhibited decreased GSH levels, increased accumulation of lipid peroxidation products, and significantly decreased levels of ferroptosis marker

Erastin and NaB Synergistically Induce CRC Cell Ferroptosis
Several studies have shown that a histone deacetylase inhibitor (HDACi) combined with ferroptosis-inducing agents can enhance its antitumor effect [16][17][18]. Our results suggest that a high concentration or long-term exposure to NaB may promote apoptosis in normal intestinal epithelial cells. As such, it is important to optimize the concentration and exposure time of NaB to increase its antitumor effect and reduce the effects on normal colon cells. Erastin is a classic ferroptosis-inducing agent, and a concentration of about 10 µM induces ferroptosis in a number of tumor cell lines [8]. Results of our MTT assays showed that 10 µM Erastin did not significantly affect cell viability, but cell viability decreased rapidly with increasing NaB concentration in the Erastin-NAB treatment group ( Figure 6A). Cell morphology examination and staining of dead cells also indicated enhanced cytotoxicity by Erastin-NAB combined treatment ( Figure 6B).
The effect of Erastin combined with NaB was further tested by isobolographic analysis ( Figure 6C). MTT assay results showed that the IC50 and 95% confidence interval (CI) of NaB and Erastin for inhibiting HCT-116 proliferation were 12.420 mM (95% CI: 11.936-12.941 mM) and 26.978 µm (95% CI: 25.714-28.381 µm), respectively. The combination of NaB and Erastin at a concentration ratio of 12.420 mM to 26.978 µm exhibited a synergistic effect on the inhibition of proliferation of HCT-116 cells (IC50 = NaB: 2.43 mM + Erastin: 4.861 µm). Combined treatment with NaB and Erastin did not affect the proliferation activity of FHC cells( Figure 6D). In addition, compared with cells treated with NaB alone, the combined treatment group exhibited decreased GSH levels, increased accumulation of lipid peroxidation products, and significantly decreased levels of ferroptosis marker protein SLC7A11 in HCT-116 cells ( Figure 6E-G). These results suggested that NaB and Erastin can synergistically induce ferroptosis in HCT-116 cells.  Figure S4). * p < 0.05, *** p < 0.001.

Discussion
Butyrate produced by dietary fiber promotes colon health and has anti-tumor properties. Butyrate is the preferred energy source in normal colon cells and can maintain mucosal integrity and inhibit inflammation and the development of cancer by affecting immunity, balancing intestinal flora, and regulating gene expression [7,19]. In tumor cells, butyrate acts as HDACi to promote the expression of anticancer genes, inhibit the proliferation of tumor cells, and promote apoptosis [20]. Recent pharmacotranscriptome and molecular biology experimental results suggest that HDACi drugs have the potential to regulate ferroptosis [18]. In a review of the synergistic effects of diet and gut microbes to enhance programmed death in CRC, Chapkin et al. suggested that butyrate and omega-3 polyunsaturated fatty acids may be effective in inducing ferroptosis. This is believed to be due to ROS accumulation induced by butyrate and the entry of long-chain polyunsaturated fatty acids into cell membranes [21]. However, no studies have confirmed the ability of NaB to induce ferroptosis in normal intestinal epithelial or CRC cells at low and high concentrations. In this study, we found that a high concentration of NaB can induce a high degree of apoptosis in CRC cells and a low degree of apoptosis in normal colon cells, and NaB only induces ferroptosis in tumor cells.
Iron is an active redox metal that is involved in free radical formation and expansion of lipid peroxidation. Generally, cancer cells require excess iron to support their high metabolic rate and support this phenotype by altering cellular pathways. Therefore, an increased iron level can make tumor cells more prone to ferroptosis [22,23]. The PCR array results in this showed that NaB significantly upregulated the iron-related genes TFR1 and IREB1/2. IREB1/2 is a primary binding protein of cells when iron is needed; when cells require iron, IREB1/2 directly binds to the RNA stem-loop structure in the 3 -untranslated region (UTR) of TFR1 mRNA and stabilizes its expression, thereby increasing intracellular iron concentration [24]. In addition, the up-regulation of ferritinophagy-related genes NCOA4 and ATG5 also suggests that NaB may degrade ferritin and release iron ions through autophagy [25].
Our results showed that both the GSH synthesis rate-limiting enzyme GCLC and the GSH cleavage-related gene CHAC1 were significantly up-regulated by NaB, suggesting that NaB plays a role in the regulation of GSH synthesis and metabolism [26,27]. This is in addition to the role of the key enzymes GOT1 and GLS2 in the metabolism and utilization of glutamine, an important precursor of GSH. More importantly, SLC7A11, a key protein in the formation of System Xc-which transports cysteine into the cell to synthesize GSH, was significantly inhibited by NaB at both the mRNA and protein levels. This suggests that NaB may effectively inhibit intracellular GSH content through this pathway.
Studies have shown that CD44 promotes the deubiquitination of SLC7A11 by promoting the binding of the deubiquitination enzyme OTUB1 to SLC7A11, thus making SLC7A11 on the cell membrane stable against proteasome pathway degradation [28]. Our results suggest that NaB regulates CD44/SLC7A11-mediated GSH depletion and ferroptosis. GPX4 is another key regulator of ferroptosis and needs to acquire electrons from GSH for its antioxidant capacity. The PCR and protein assay results in this study showed that NaB does not have a role in GPX4 regulation.
Since colorectal tumors are often insensitive to traditional ferroptosis inducers, finding effective drug synergies has become an area of active research in the development of ferroptosis cancer treatments. For example, dihydroartemisinin (DAT) combined with RSL3, the HDAC inhibitor vorinostat combined with Erastin, and the DHODH inhibitor brequinar combined with sulfapyridine all induce ferroptosis in tumor cells [29][30][31]. Yang et al. reported that high expression of SLC7A11 may be related to the sensitivity of CRC cells to ferroptosis [18]. In our preliminary study, Erastin did not effectively inhibit SLC7A11 mRNA and protein levels in HCT-116 cells (data not presented), even though its pharmacological mechanism is explained by its effect on mitochondrial VDAC and System Xcactivity on the cell membrane. We found that the IC50 of Erastin on H-116 cells was over 20 µM for 24 h, which is much higher than that of other tumor cell lines (1-10 µM). However, NaB effectively inhibits SLC7A11 transcription and protein expression in HCT-116 cells. When HCR-116 cells were co-treated with NaB and Erastin, SLC7A11 protein levels were significantly down-regulated as compared to when cells were treated with NaB or Erastin alone. These findings may provide a basis for the development of combination drugs based on an HDACi and ferroptosis inducer.

Conclusions
The results of this study showed that NaB induces apoptosis and ferroptosis in CRC cells. Treatment with NaB significantly changed the expression of iron and GSH metabolite function-related genes. The findings also suggest that NaB may regulate ferroptosis in CRC cells through the CD44/SLC7A11 pathway. Finally, when NaB was combined with Erastin, inhibition of CRC cells was enhanced, suggesting that NaB combined with Erastin might have a stronger anti-CRC effect.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cancers15020423/s1, Figure S1: The original Western blotting figures of Figure 4A; Figure S2: The original Western blotting figures of Figure 4B; Figure S3: The original Western blotting figures of Figure 5B; Figure S4: The original Western blotting figures of Figure 6G; Table S1: The values of fold change, Padj and Log2FC of each gene in FHC group VS HCT-116 group in PCR array; Table S2: The values of fold change, Padj and Log2FC of each gene in HCT-116 group vs. HCT-116 + NaB group in PCR array.