Shu-Ran Chen, Hua-Zhang Wu , Mu-Lin Liu?

1. Department of Gastroenterology,the First Affiliated Hospital of Bengbu Medical College,Bengbu 233030,China

2. School of Life Science,Bengbu Medical College,Bengbu 233030,China;3. Anhui Province Key Laboratory of Translational Cancer Research,Bengbu 233030,China

Keywords:Hexokinase-3 Immunological infiltration Epithelial-mesenchymal transition Cancer metastasis

ABSTRACT Objective: Investigate the effects of Hexokinase-3 on the proliferation, cell cycle, migration and immunologic invasion of RKO and HCT116. Methods: The expression levels of HK3 in tumor and normal colon samples were analyzed by bioinformatics. RKO and HCT116 were transfected with the control group (si-NC) and interference group HK3(si-HK3#1, si-HK3#2).CCK-8, Flow cytometry, wound healing and Transwell were used to examine the effects of interference with HK3 on the proliferation, cell cycle and migration of RKO and HCT116 cells. The expression levels of vimentin and E-cadherin, which are related to epithelialmesenchymal transition(EMT), were detected by Western blot. Results: HK3 is associated with infiltration of multiple immune cells and co-expression with multiple immune checkpoints in colorectal cancer. Compared with the Control group (si-NC),the protein expression level of HK3 in interference group (si-HK3#1, si-HK3#2) was significantly decreased(P＜0.05);the proliferation and migration of RKO and HCT116 cells were significantly inhibited(P＜0.05);the cell cycle of RKO and HCT116 cells was arrested in S phase(P＜0.05) and the expression of E-cadherin protein was increased, but the expression of vimentin protein was inhibited(P＜0.05). Conclusion: HK3 affects the level of immunologic invasion of colon cancer.Interference with HK3 inhibits the proliferation, cell cycle and migration of colon cancer cells by inhibiting EMT behavior.

1. Introduction

The Global Cancer Report 2020 shows that colorectal cancer is in the top three of oncological diseases and the second most deadly [1].Due to advances in diagnostic and endoscopic techniques, more and more patients with early-stage colorectal cancer are being diagnosed and treated, and the response to treatment and prognosis for earlystage colorectal cancer are significantly better compared to advanced colorectal cancer [2].The main cause of disease progression and death in colorectal cancer patients is liver metastases from colorectal cancer. Due to the insidious nature of clinical symptoms of liver metastases from colorectal cancer, most of the patients attending the clinic have already developed liver metastases, resulting in the loss of the best treatment opportunity [3; 4].Screening for new diagnostic markers for clinical use is therefore essential for the early diagnosis and treatment of colorectal cancer.

HK3 is a key enzyme in the process of glycolysis. Studies have shown that HK3 contributes to the malignant progression of a variety of tumours, including renal clear cell carcinoma, breast cancer and haematological tumours [5; 6; 7].In addition, HK3 has been reported to be involved in the construction of a variety of tumour immune microenvironments [8; 9].In colorectal cancer, HK3 was reported to be highly expressed at the mRNA level [10].However,the effects of HK3 on the immune microenvironment of colorectal cancer and how it affects the malignant progression of colorectal cancer have not been fully investigated. Therefore, this study investigated the effect of HK3 on the immune microenvironment of colorectal cancer through bioinformatics, and also revealed the intrinsic mechanism of malignant metastasis of colorectal cancer through cell proliferation, cell cycle and metastasis experiments,providing new ideas and directions for the clinical treatment of colorectal cancer.

2. Material and methods

2.1 Bioinformatics analysis

Expression profile data for colon cancer samples and clinical information on patients were obtained from the TCGA and GEO databases, and data from the TCGA and GEO databases were corrected using the limma package, and colon cancer tumour samples obtained from GEO were combined after removing batch effects. The online website Kaplan-Meier Plotter (http://kmplot.com/analysis/index) explored the relationship between HK3 expression and patient prognosis. Samples were divided into high and low expression groups based on median HK3 expression, and the Metascape online website (https://metascape.org/gp/index) analysed the functional regions in which the differential genes were enriched between the two groups; GSEA enrichment analysis explored the differences in pathways enriched in the high and low HK3 expression groups. TIMER database (https://cistrome.shinyapps.io/timer/) to investigate the relationship between HK3 expression and immune infiltration in colon cancer.

2.2 Materials

Human colon cells HCT116 and RKO were purchased from Shanghai Institute of Cell Science, Chinese Academy of Sciences;DMEM medium and fetal bovine serum were purchased from GIBCO, USA; si-HK3 was purchased from Shanghai Jima Pharmaceutical Technology Co. Ltd.; Transwell was purchased from Conring; primary antibodies HK3, E-cadherin (CDH1), Vimentin,GAPDH and secondary antibodies were purchased from Wuhan Sanying Biotechnology Co.

2.3 Cell culture and transfection

HCT116, RKO cells were cultured in DMEM containing 10% fetal bovine serum by volume in a cell culture incubator at 37℃and 5%CO2. When cell fusion reached 70%-80%, the cells were transfected using Lipo6000 transfection reagent and the transfection steps were carried out strictly according to the transfection instructions.

2.4 CCK-8

Logarithmic growth phase cells were taken and inoculated in 96-well plates at a cell density of 3000/well for RKO and 2500/well for HCT116, with 100μL of PBS around each well to seal the plate. After overnight, the cells were transfected with Lipo6000 transfection reagent and a transfection control was set up. The cells were incubated for 2h in a warm oven after the addition of CCK-8 reagent and the absorbance at 450nm was measured by enzyme marker. The time points were 0h, 24h, 48h and 72h after transfection.

2.5 Flow cytometry analysis

Logarithmic growth phase cells were inoculated in 6-well plates, and when the cells were fused to 50%-60%, the cells were transfected using Lipo6000 transfection reagent and a transfection control group was set up. The cells were collected 48h after transfection, washed 3 times with PBS and slowly added to 70%ethanol and fixed at 4℃ for 24h. The fixed cells were suspended in binding buffer, staining solution was added according to the instructions and incubated for 30min at room temperature in the dark.

2.6 Wound healing experiment

Cells were taken from logarithmic growth phase, inoculated in 6-well plates, and when cell fusion reached 50%-60%, cells were transfected using Lipo6000 transfection reagent and set up transfection control group. 48h after transfection, cells were collected. After overnight incubation in a 37° incubator, the cells were incubated with a 200ul yellow tip perpendicular to the straight line, scraped out, washed twice with PBS and photographed under an inverted microscope (0h photographed), removed after 24h, washed twice with PBS and photographed under an inverted microscope(24h photographed). Image J was used to calculate the scratch area values. Wound healing rate (%) = ((0H scratch area - 24H scratch area)/0H scratch area)*100%

2.7 Transwell assay

Logarithmic growth phase cells were taken, inoculated in 6-well plates, and when the cell fusion reached 50%-60%, the cells were transfected using Lipo6000 transfection reagent and a transfection control group was set up. 48h after transfection, the cells were collected. 5×104 cells were inoculated in the upper chamber of Transwell using DMEM without serum in a volume of 100μl and the lower chamber was incubated with DMEM medium containing 10% fetal bovine serum in a volume of 500μl and the incubation was continued in a 37℃, 5% CO2 cell incubator for 48h. Cells not crossing the bottom membrane in the upper chamber of Transwell were wiped off using cotton swabs. The cells were fixed in methanol for 40 min, stained with 0.5% crystalline violet for 20 min, washed and dried, observed and photographed under an inverted microscope,and the number of cells migrating was calculated using Image J.

2.8 Western blot

Logarithmic growth phase cells were taken, inoculated in 6-well plates, and when cell fusion reached 50%-60%, the cells were transfected using Lipo6000 transfection reagent and a transfection control group was set up. 72h after transfection, proteins were extracted using RIPA according to the instructions, quantified using the BVA kit and the protein concentrations between groups were corrected to the same level using RIPA.

2.9 Data analysis

The data were expressed as mean ± standard deviation and analysed using Graphpad prism 8.0 software for graphing.

3. Results

3.1 Expression of HK3 in colon cancer and its relationship with prognosis

The expression levels of HK3 in colon cancer and corresponding paracancerous normal tissues were analysed by the TCGA database,as shown in Figure 1A, the expression of HK3 in colon cancer tissues was significantly paired with paracancerous tissues, and the difference was statistically significant (p=0.0024). The relationship between HK3 expression and prognosis of colon cancer patients was analysed by Kaplan-Meier Plotter database, as shown in Figure 1B, high expression of HK3 was associated with poor prognosis of patients, and the difference was statistically significant (P < 0.001).

Figure1 Expression of HK3 in colorectal cancer and its relationship with prognosis.

3.2 HK3 is associated with immune function in colon cancer

As shown in Figure 2A, the use of the R package eliminated the batch effect between the GSE17536, GSE29621 and GSE38832 datasets, allowing samples from different batches to be analysed after pooling. Data were grouped according to the level of HK3 expression to obtain differentially expressed genes between high and low HK3 expression groups (Figure 2B). The functional regions in which the differentially expressed genes were enriched were analysed using the online website Metascape, and as shown in Figure 2C, these genes were enriched in several immune-related functional regions, such as: positive regulation of the immune response, natural immune response and immune response factors.

Figure2 The expression of HK3 is related to the immune function of colorectal cancer.

3.3 HK3 affects immune cell infiltration in colon cancer and correlates with immune checkpoint inhibitor expression

The samples were divided into HK3 high expression group and HK3 low expression group, and the functional differences between the two groups were analysed using GSEA_4.2.2 software. As shown in Figure 3A, the HK3 high expression group was enriched in several immune-related pathways, such as T-cell receptor signalling pathway, B-cell receptor signalling pathway, etc. Next, we explored the relationship between HK3 and immune checkpoints, and the results showed that HK3 was co-expressed with most immune checkpoints and showed a positive prior relationship (Figure 3B).Finally, we analysed the effect of HK3 on the expression of six types of immune cells in colon cancer using the TIMER database. Figure 3C shows that HK3 expression was associated with tumour purity(r= -0.386,P= 6.84e-16), B cells (r= 0.067,P= 1.76e-01), CD8+ T cells (r= 0.23,P= 2.97e-06), CD4+ T cells (r= 0.384,P= 1.44e-15),The expression of macrophages (r= 0.524,P= 6.92e-30), neutrophils(r= 0.649,P= 2.23e-49), and dendritic cells (r= 0.654,P= 1.61e-50)all showed strong correlation.

Figure3 HK3 affects immune cell infiltration in colorectal cancer and is associated with checkpoint inhibitor expression.

3.4 Effect of HK3 on the proliferation and cycle of colon cancer cells

The effect of HK3 on colon cancer cell proliferation was analyzed by CCK-8 experiments(Fig 4A). After knockdown of HK3 expression in colon cancer cells using si-HK3#1 and si-HK3#2,the cell proliferation of both RKO and HCT116 was inhibited significantly, and si-HK3#2 had a stronger inhibitory effect than si-HK3#1. The effect of HK3 on the colon cancer cell cycle was analyzed by Flow cytometry(Fig 4B). Meanwhile, both RKO and HCT116 cell cycles accounted for a significantly higher proportion of S-phase, and in combination with cell proliferation results,knockdown of HK3 could arrest the cell cycle of colon cancer cells in S-phase.enriched in several cell migration-related pathways, such as cell adhesion molecules, adherent patch pathways and extracellular matrix receptors. By analysing the TCGA database, we determined that the expression of CDH1 and VIM, key molecules for epithelial mesenchymal transition, correlated significantly with HK3 (Figure 6B). By protein blotting, we examined the effect of knockdown of HK3 on the expression of CDH1 and VIM, and the results showed that knockdown of HK3 caused an increase in CDH1 expression and a decrease in VIM expression in RKO and HCT116 cells (Figure 6C).

Figure4 Effects of HK3 on proliferation and cell cycle of colon cancer cells A: CCK8 assay for cell proliferation; B: Flow cytometry for cell cycle.

Figure6 The effect of HK3 on the key molecule of epithelial-mesenchymal A: GSEA enrichment indicates that HK3 is involved in several cell migration pathways including epithelial mesenchymal transition; B: Expression of HK3 in relation to E-cadherin and vimentin in TCGA database; C: Western blot assay to detect the expression of HK3 in relation to E-cadherin and vimentin.

3.5 Effect of HK3 on the migration of colon cancer cells

Wound healing was used to validate the effect of knockdown of HK3 on the migratory ability of RKO(Fig 5A). Si-HK3#1 and si-HK3#2 were used to knockdown HK3 expression in RKO, the migration ability of RKO cells was significantly inhibited. Transwell was used to validate the effect of HK3 on the migratory ability of HCT116(Fig 5b). Cell migration ability of HCT116 was significantly inhibited using si-HK3#1 and si-HK3#2.

Figure5 Effects of HK3 on migration of colon cancer cells

3.6 HK3 affects the ability of colon cancer cells to migrate via EMT

As shown in Figure 6A, samples with high HK3 expression were

4. Discussion

In addition to its involvement in the classical pathway of glycolysis,there is growing evidence that HK3 is involved in the construction of the tumour immune microenvironment and in tumour metastasis.In the present study, we demonstrate that HK3 is associated with the degree of infiltration of multiple immune cells in colorectal cancer and leads to tumour metastasis via epithelial mesenchymal transformation.

Although the number of colorectal cancer patients who develop metastases is clinically high, only a very small proportion of tumours develop meaningful metastases due to a variety of adverse factors in the body [11].To adapt to these adverse factors, tumours have evolved a range of capabilities during malignant progression, including:sustained proliferative signalling, tissue invasion and metastasis and evasion of immune destruction, with metabolic reprogramming being one of these many capabilities [12; 13].Glycolysis is a common metabolic reprogramming process that begins with the presence of a rate-limiting enzyme: hexokinases (HKs)[14].Five hexokinase isozymes (HK1-5) have been identified in mammals, and numerous studies have shown that abnormal expression of members of the hexokinase family leads to malignant progression of colorectal cancer [15; 16; 17].The study of metabolic enzymes is not limited to the classical pathways that regulate the metabolic reprogramming of tumours, but more and more research is focusing on the nonclassical pathways in which these enzymes are involved [18].Studies report that hexokinase 2 binds to and inhibits mTOR receptor 1(mTOR1) in glucose-deficient cardiomyocytes, thereby reducing the inhibitory effect of mTOR1 on autophagy [19].In addition, HK2 has been reported to inhibit the function of voltage-dependent anion channels, allowing tumor cells to escape apoptosis [20].

Previous studies have shown a close relationship between HK3 expression and myeloid dendritic cell, NK cell and monocyte lineages in non-small cell lung cancer, with the CD8+ T cell composition of the HK3 low expression group differing from that of the HK3 high expression group. The present study likewise found a positive correlation between HK3 expression and the degree of infiltration of these immune cells. Similar to the results of this study, HK3 positively correlated with multiple immune checkpoint expression at the mRNA and protein levels in non-small cell lung cancer, and this study also detailed that HK3 expression correlated with the outcome of patients receiving immunotherapy in clinical practice [9].These results suggest that there may be an important link between HK3 and immune infiltration in colorectal cancer, and even that HK may be a potential new biomarker to guide clinical decisions on immunotherapy for colorectal cancer.

Study suggests important link between epithelial mesenchymal transition and immune escape from tumours [21].Tumour metastasis is made up of several sequential and interrelated steps, each of which is a switch for metastasis, and any one of which affects the process of metastasis [22].Epithelial mesenchymal transition (EMT) is a process that is central to achieving these steps [23].Numerous studies have also confirmed that the development of EMT is inextricably linked to the malignant progression of tumours [24; 25; 26].Therefore, this study first investigated the association of HK3 with the malignant biological behaviour of colorectal cancer. When HK3 expression was knocked down using small interfering RNA, the proliferation and migration of colon cancer cells were inhibited, and the cycle was blocked in the S phase. In combination with GESA enrichment analysis, it was hypothesised that HK3 inhibited tumour metastasis by suppressing epithelial mesenchymal transition. Therefore, we investigated the effect of HK3 on the expression of EMT-related molecules in colon cancer using protein immunoblotting and showed that knockdown of HK3 significantly reduced epithelial molecular markers and significantly increased mesenchymal markers. This result confirms that HK3 inhibits various malignant behaviours of tumour cells by suppressing the process of epithelial mesenchymal transformation in colon cancer.

This study also has certain shortcomings, firstly, it did not use sufficient clinical analysis of the clinicopathological characteristics of HK3 and clinical patients; secondly, it did not use stronger evidence that HK3 leads to the formation of immune escape through epithelial mesenchymal transition. In summary, bioinformatics and a series of experiments have shown that HK3 can affect the immune microenvironment of colon cancer and inhibit the cell proliferation,cell cycle and migration ability of colon cancer cells by suppressing the EMT signalling pathway. This study suggests that HK3 may be a new target for the diagnosis and treatment of colon cancer.