Liquid-Liquid Phase Separation Links Cholesterol Biosynthesis Inhibition to DNMT1-UHRF1 Condensate formation and Epithelial-to-Mesenchymal Transition Reversal in Colon Adenocarcinoma Cells

Abstract

Background: Over the last two decades, liquid-liquid phase separation (LLPS) has emerged as a state-of-the-art biochemical phenomenon in modern cancer research, precisely enabling spatio-temporal coordination across biochemical processes crucial for epigenetic choreography associated with gene expression. Among several epigenetic alterations, DNA methylation is the most crucial reversible DNA modification, which dictates the possibility of gene expression and chromatin organization. Methods: Colon adenocarcinoma (COAD) cells (HCT15 and HT29) were cultured in DMEM media, supplemented with 10% FBS, along with 100 units/ml penicillin and 0.1 mg/ml streptomycin, and used as a model experimental system throughout this study. Cell viability was checked to determine the sublethal concentrations (Inhibitory Concentration 30 or IC30) of 1,6-hexanediol (1,6-HD) and Polyethylene glycol-8000 (PEG-8000) by using the MTT assay on HCT15 and HT29 cells. Subsequently, total RNA was isolated from both control and drug-treated cells using the TRIzol method. cDNA was then synthesized using a cDNA synthesis kit, and quantitative RT-PCR was performed to check the expression of the target genes at the mRNA level. Moreover, western blotting was performed using the extracted membrane protein, cytosolic protein, and nuclear protein fractions from control and treated cells to check expressions of the target genes at protein level. Additionally, Co-Immunoprecipitation (Co-IP), followed by Western blotting, was performed to check protein-protein physical interactions. Later, confocal microscopy was performed to visualize the formation and disruption of condensates, as well as to assess the colocalization of proteins within both control and drug-treated cells. Flow cytometry was further performed to check the cell cycle distribution patterns using Propidium Iodide (PI) staining. Wound healing assay was performed, and wound closure percentage was quantified to check the cell migration pattern. Furthermore, the expression of EMT markers was checked and quantified by performing qRT-PCR and western blotting. Besides the in vitro experimentations, bioinformatic web tools such as PONDR, PhaSePred, FuzDrop were used to analyze the proportion of disordered regions and the potential of the proteins of interest to undergo phase separation. Molecular docking was also performed to predict the molecular interactions between 1,6-HD and kinase domains of the EGFR-RAS-MAPK pathway effectors. Results: This study precisely focuses on the impact of LLPS on cholesterol biosynthesis and its downstream effects on EGFR signalling in association with DNA methylation dynamics in COAD. This investigation demonstrates that the application of the phase separation disruptor, 1,6-hexanediol (1,6-HD) impairs cholesterol biosynthesis, thereby affecting membrane dynamics and reducing the phosphorylation of the EGFR signalling cascade. On the contrary, the opposite effects were observed after treating the cells with the phase separation inducer PEG-8000. Furthermore, molecular docking reveals a precise binding of 1,6-HD to the kinase domain of EGFR, which ultimately prevents the activation of the RAS-MAPK pathway and also inhibits phosphorylation events at multiple downstream effector levels. Additionally, the inhibition of the RAS-MAPK pathway ultimately results in a partial decrease in DNMT1 expression. Interestingly, our results further reveal that the formation of the DNMT1-UHRF1 condensate at the replication fork is also obstructed by 1,6-HD treatment, which ultimately diminishes the activity of DNMT1 and disrupts the DNA methylation pattern in the CDH1 gene territory. This epigenetic modulation promotes CDH1 upregulation, which ultimately preserves epithelial traits; while the alteration of LLPS changes the epigenetic landscape at various territories associated with mesenchymal stemness genes, resulting in their downregulation. Implications: Overall, our findings, for the first time, implicate a distinct mechanistic relationship among LLPS, membrane dynamics, oncogenic signalling, DNA methylation, and cellular plasticity in COAD. This study opens new avenues for targeting LLPS and cholesterol biosynthetic pathways as a therapeutic strategy against COAD by remodelling epigenetic landscape.