Molecular and phenotypic characterization of 5-FU resistant colorectal cancer cells: toward enrichment of cancer stem cells

Abstract

Cancer stem cells (CSCs) as a subgroup of cells within a tumor capable of self-renewal, thereby driving tumor initiation and spread. Addressing treatment failures in cancer, linked to CSCs and their resistance mechanisms, requires effective preclinical models for testing targeted therapies. Caco2- and HT-29-resistant cells were generated by repeated treatment of cells with growing concentrations of 5-fluorouracil (5-FU) anticancer drug for an extended time. The sensitivity of 5-FU-resistant cells was evaluated by cytotoxicity assay. Stemness, epithelial-mesenchymal transition (EMT), migration and drug resistance characteristics were assessed through gene expression investigation by real-time PCR. The expression of CD44, CD133, and CD66 were evaluated by flow cytometry. To end, the bioinformatic analysis estimated the molecular function and biological pathways considering the differential expression of selected genes and proteins. 5-FU-exposed cells displayed increased resistance to 5-FU. The gene expression analysis showed an upregulation of stemness genes (KLF4, SOX2, OCT4, C-MYC), enhanced scavenging system, and elevated expression of CSC surface markers (CD44 and CD133) compared to parental cells. Additionally, pro-EMT genes (TWIST1, SNAIL1, ZEB1, Vimentin, and N-cadherin) were significantly upregulated compared to parental cells, with the downregulation of E-cadherin as an EMT suppressor gene reflected in increased migration capacity. Moreover, increased expression of ABC transporter genes (ABCB1, ABCC1) was observed, correlating with enhanced drug resistance. The bioinformatic analysis highlighted pathways related to microRNAs in cancer, cells pluripotency, and proteoglycans. Methods of drug exposure take priority over spheroid formation, particularly due to their enhanced efficacy in stemness, EMT, and surface markers. This positions them as a promising protocol for establishing experimental models of CSCs.

Keywords: 5-FU, EMT; ABC, Drug resistance; Cancer stem cell; Colorectal cancer.

Fig. 1

Response of Caco2 and HT-29 CRC cell lines to cytotoxic effects of 5-FU. Caco2 and HT-29 cells exhibited cytotoxic responses at concentrations of 50 ng/ml or higher of 5-FU. The IC50 values for 5-FU in Caco2 and HT-29 cells were determined to be 353.4 ng/ml and 543.3 ng/ml, respectively

Fig. 2

Response of 5-FU-exposed Caco2 and HT-29 CRC cells to cytotoxic effects of 5-FU. The IC50 values for 5-FU in Caco2 and HT-29 cells were determined to be 7,039 ng/ml and 6348 ng/ml, respectively

Fig. 3

Characteristics changes in Caco2 and HT-29 spheroids and 5-FU exposed cells. (A) Parental Caco2 cells showed epithelial morphology, forming tightly packed monolayers resembling epithelial cells. (B) Spheroids derived from Caco2 cells exhibited a compact, spherical morphology. (C) Caco2 cells appear heterogeneous, with most cells showing cuboidal or round shapes and occasional giant cells. However, cultivation conditions favor specific cell subpopulations, altering cell morphology compared to the parental line. Drug-adapted cells display a looser structure compared to the spherical colony-shaped aggregations of parental cells. (D) Parental HT-29 cells display a range of shapes, including polygonal, spindle-shaped, and elongated morphologies. (E) Spheroids derived from HT-29 cells displayed a rounded shape and exhibited a dense, compact morphology. (F) There were no substantial changes noted in their morphology after exposure to 5-FU; however, they displayed a less firm attachment to the culture flask and seemed to accumulate in colony formations. (G) CFSE analysis showed slight increase in proliferation of both CRC cell lines after prolong treatment with 5-FU

Fig. 4

RT-qPCR analysis of stemness genes. When HT-29 and Caco2 cells were exposed to 5-FU, the expression of genes that regulate stemness was found to be higher than in the parental cell population. A) Compared to the spheroids, the isolated Caco2 subpopulation exposed to 5-FU showed a significant increase in C-MYC and OCT4 expression (p-value < 0.0001). B) With the exception of NANOG, the expression levels of stemness genes in the isolated HT-29 subpopulation cells exposed to 5-FU did not differ significantly from those in the spheroids. Interestingly, spheroid cells showed a more marked upregulation of NANOG (p-value < 0.0001) in comparison to the isolated HT-29 subpopulation cells exposed to 5-FU. The baseline gene expression in parental cells is represented by a dotted line. Data are presented as mean ± SD from three independent experiments as **** = p-value < 0.0001

Fig. 5

Flow cytometry analysis of CRC stemness markers expression. The expression of CRC-CSC markers in HT-29 and Caco2 subpopulations exposed to 5-FU was analyzed using flow cytometry in comparison to their parental and spheroid cells. The changes in marker expression were confirmed by flow cytometry, specifically in CD44 and CD133 in Caco2 subpopulations exposed to 5-FU (A, C) and in CD44 in HT-29 subpopulations exposed to 5-FU (B, D). Data are presented as mean ± SD as *** = p-value < 0.001 and **** = p-value < 0.0001

Fig. 6

Analysis of ROS levels. The parental cells exhibit the highest ROS levels, followed by spheroid cells, with the lowest levels observed in 5-FU exposed cells. Data are presented as mean ± standard deviation (SD) from three independent experiments. Statistical significance is indicated as ***p < 0.001, ****p < 0.0001

Fig. 7

RT-qPCR analysis of EMT genes. In both 5-FU-treated Caco2 and HT-29 cells, there was a notable shift in the expression of genes related to EMT. In Caco2 cells, TWIST1, ZEB1, and N-cadherin were significantly increased, while SNAIL1, Vimentin, and E-cadherin were decreased following exposure to 5-FU. Importantly, TWIST1 (p-value = 0.0073) and ZEB1 (p-value < 0.0001) were markedly upregulated in 5-FU-exposed Caco2 cells compared to spheroids. Similarly, in HT-29 cells, all EMT-inducing genes were upregulated, while E-cadherin was downregulated. Notably, TWIST1 (p-value < 0.0001) and SNAIL1 (p-value = 0.0049) were significantly elevated in spheroids compared to 5-FU-exposed HT-29 cells. The baseline gene expression in parental cells is represented by a dotted line. Data are presented as mean ± SD from three independent experiments as **** = p-value < 0.0001, ** = p-value < 0.01

Fig. 8

Wound scratch assay for migration assessment. Images of wound areas for Caco2 (A) and HT-29 (B) cells at 0 h and 24 h post-scratch are shown. The cell areas are highlighted in green, with the mean scratch length marked by a red line. The bar graphs on the right represent the percentage of the remaining scratch area at 0 and 24 h. For both Caco2 and HT-29 cells, 5-FU exposed CRC cells exhibited significant wound closure, indicating higher migratory capacity. However, migration capacity of parental and spheroid cells was insignificant. Data represent mean ± SD from two independent experiments. Statistical significance: *p < 0.05 and **p < 0.01

Fig. 9

RT-qPCR analysis of MDR genes. A) Compared to parental cells, 5-FU-treated Caco2 cells showed a moderate uptick in ABCB1 and ABCC1 expression. Interestingly, spheroids showed a significantly stronger upregulation of ABCC1 than did isolated 5-FU-exposed Caco2 cells (p-value < 0.0001). B) In contrast, HT-29 cells exposed to 5-FU showed only a modest increase in ABCC1 expression when compared to the parent HT-29 cells. HT-29 cells exposed to 5-FU showed significantly lower ABCB1 expression than spheroids (p-value = 0.0012). The baseline gene expression in parental cells is represented by a dotted line. Data are presented as mean ± SD from three independent experiments as ****= p-value < 0.0001, ** = p-value < 0.01

Fig. 10

The most important enriched gene ontologies are based on (A) molecular functions and (B) biological pathways

Fig. 11

Gene interaction network of common differentially expressed genes. The network illustrates the interactions between 11 genes significantly enriched in KEGG pathways, including MicroRNAs in cancer, signaling pathways regulating pluripotency of stem cells, and Proteoglycans in cancer. Each node represents a gene, and the edges represent known or predicted interactions, categorized as experimentally determined (pink), curated from databases (blue), or predicted through gene neighborhood, co-expression, and text mining (green, black, and yellow respectively). Key players such as MYC, CD44, PROM1, and SOX2 are central to the network, indicating their collaborative roles in the pathways analyzed