Epigenetic regulation in colorectal cancer: The susceptibility of microRNAs 145, 143 and 133b to DNA demethylation and histone deacetylase inhibitors

Abstract

Globally, colorectal cancer (CRC) is a major health concern. Despite improvements in CRC treatment, mortality rates remain high. Genetic instability and epigenetic dysregulation of gene expression are instigators of CRC development that result in genotypic differences, leading to often variable and unpredictable treatment responses. Three miRNAs, miR-143, -145 and -133b, are most commonly downregulated in CRC and are proposed here as potential tumour suppressors. Although the downregulation of these miRNAs in CRC is largely unexplained, epigenetic silencing has been postulated to be a causative regulatory mechanism. Potential epigenetic modulation of miRNA expression, by means of histone acetylation and DNA methylation, was assessed in this study by treating early (SW1116) and late stage (DLD1) CRC cells with the DNA demethylating agent 5-aza-2'-deoxycytidine (5-Aza-2'C) and the histone deacetylase (HDAC) inhibitor Trichostatin A (TSA), respectively. Subsequent quantification of miRNA expression revealed that while all the selected miRNAs were susceptible to DNA demethylation in early- and late-stage CRC cells, susceptibility to DNA demethylation was significantly pronounced in late-stage DLD1 cells. Conversely, although histone acetylation moderately affected miRNA expression in early-stage CRC, it had a marginal effect on the expression of miRNAs in late-stage CRC cells. Overall, this study provides further understanding of the contribution of epigenetics to the regulation of putative tumour suppressor miRNAs in CRC.

Fig 1. SW1116 and DLD1 cell viability following 5-Aza-2’-C treatment.

Cell viability assays conducted on SW1116 cells following treatment with 5-Aza-2’-C correlated with the visual changes noted with the number of live cells increasing by 1% to a level of approximately 97% viability when treated with 1μm 5-Aza-2’-C (p = 0.5387), compared to 96% viability for the untreated equivalent of the cells. However, with an increased concentration (3μm) of the DNA demethylating agent, cell viability decreased with a 3% drop in viability to 93%, compared to the untreated equivalent (p = 0.1642). The “no treatment control” (NTC) samples for the DLD1 cell line yielded a 97% cell viability while treatment with 1μM 5-Aza-2’-C decreased cell viability by 6% yielding a value of approximately 91% live cells (p = 0.0157). The number of live cells were further decreased to 89% at the increased dose of 3μM 5-Aza-2’-C (p = 0.0058). Overall, the effect of 5-Aza-2’-C was most notable in the late-stage cancer cell line DLD1. * Significant (p<0.05); ** Very significant (p<0.01).

Fig 2. SW1116 and DLD1 cell viability following TSA treatment.

Consistent decreased viability was demonstrated in SW1116 cell cultures after treatment with 300nM TSA. The dramatic percentage decrease in viable cells compared to the untreated control cells was 18% (p = 0.0001). This drop was independent of DMSO as the DMSO treated control cells had diminished the percentage of live cells by 6% (p = 0.0047). Similarly, the cell viability assay of DLD1 cells showed untreated cells to be 99% viable, whilst DMSO carrier control cells had a minimally decreased viability of some 6% (p = 0.0047). TSA treatment in comparison, resulted in a 15% decrease in DLD1 cell viability (p = 0.0001). ** Very significant (p<0.01), *** Extremely significant (p<0.001).

Fig 3. Relative expression of miR-143, miR-145 and miR-133b in SW1116 cells after treatment with 5-Aza-2’-C.

Low dose (1μm) 5-Aza-2’-C yielded increased miR-143 expression by almost 2-fold (p = 0.0157), while there was a marginal increase of 0.018-fold (p = 0.503) after high dose (3μm) 5-Aza-2’-C treatment. Relative expression of miR-145 in SW1116 cells after treatment with 5-Aza-2’-C. Low dose (1μm) and high dose (3μm) 5-Aza-2’-C yield increased miR-145 expression by 2.4-fold (p = 0.0177) and 1.8-fold (p = <0.00001) respectively. Relative expression of miR-133b in SW1116 cells after treatment with 5-Aza-2’-C. Low and high dose 5-Aza-2’-C treatment yield increased miR-133b expression by 1.6 (p = 0.0282) and 1-fold (p = 0.0047) respectively. *Significant (p<0.05), ** Very Significant (p<0.01), *** Extremely significant (p<0.001).

Fig 4. Relative expression of miR-143, miR-145 and miR-133b in DLD1 cells after treatment with 5-Aza-2’-C.

Low dose (1μm) yielding an increased expression of miR-143 by 3.7-fold (p = 0.005), while high dose (3μm) 5-Aza-2’-C exhibits down-regulation of miR-143 by 1-fold (p = 0.0238). Relative expression is shown as log (2^-ΔΔCt, base 2). A Low dose of 5-Aza-2’-C (1μm) yielded an increased expression of miR-145 by 4.3-fold (p = 0.001) while high dose (3μm) 5-Aza-2’-C exhibited marginal increase of miR-145 by 0.1-fold (p = 0.00001). Relative expression is shown as log (2^-ΔΔCt, base 2). Treatment with 5-Aza-2’-C (1μm) yielded an increased expression of miR-133b by 2.3-fold (p = 0.1179, not significant) while a higher dose (3μm) of 5-Aza-2’-C exhibits down-regulation of miR-133b by 2,8-fold (p = 0.0870, not significant). Relative expression is shown as log (2^{-ΔΔCt, base} 2). *Significant, (p<0.05), ** Very Significant (p<0.01), *** Extremely significant (p<0.001).

Fig 5. Relative expression of miR-143, miR-145 and miR-133b in SW1116 cells after treatment with TSA.

It was determined that relative to the no treatment controls, when the cells were treated with 300nM TSA, miR-143 had increased in expression by 3,7-fold (p = 0.0759). DMSO alone only increased the expression of miR-143 by 1.8-fold, almost 1.9-fold less than that of 300nM TSA. The quantification of miR-143 post treatment with DMSO was found to be significant (p = 0.0028). The effect on the expression of miR-145 was much like that of miR-143. Treatment with 300nM TSA resulted in a mean decrease in Ct (p = 0.0508), providing evidence that treatment with TSA demonstrates a trend of up-regulation of miR-145 by 3.8-fold, relative to the no treatment control. DMSO treatment had induced the expression of miR-145 by 1.7-fold, (p = 0.0332). TSA significantly increased the expression of miR-133b by 3.3-fold, (p = 0.0353). DMSO had induced a significant increase in miR-133b expression by 2.5-fold (p = 0.0101). Although DMSO had also induced expression of the miR-133b, this was 0.8-fold less than the expression for miR-133b after treatment with 300nM TSA. Relative expression is shown as log (2^-ΔΔCt, base 2), where p<0.05 is significant.

Fig 6. Relative expression of miR-143, miR-145 and miR-133b in DLD1 cells after treatment with TSA.

miR-143 had decreased by 0.4-fold relative to the no treatment. This mean increase in Ct was found to be non-significant (p = 0.3277). Treatment with DMSO alone had significantly decreased miR-143 expression by 0.08-fold, (p = 0.007515). MiR-145 expression in late-stage colorectal adenocarcinoma was found to be mildly affected by 300nM TSA. The mean Ct was slightly increased compared to the NTC (p = 0.6591). Therefore miR-145 expression decreased slightly by 0.07-fold, relative to the NTC. The DMSO control exhibited a mean increase in Ct (p = 0.3383) with an almost 0.2-fold decrease in expression was noted when treated with DMSO. A decrease in miR-133b expression by 0.5-fold from the NTC was seen (p = 0.32). However, the DMSO alone had demonstrated a mean decrease in Ct, (p = 0.1171), showing increased expression of miR-133b by 0.26-fold. Relative expression is shown as log (2-^ΔΔCt, base 2). ** Very Significant (p<0.01).